Thursday, 3 May 2018

Linear mitochondrial DNA is rapidly degraded by components of the replication machinery

Peeva V, Blei D, Trombly G, Corsi S, Szukszto MJ, Rebelo-Guiomar P, Gammage PA, Kudin AP, Becker C, Altmüller J, Minczuk M, Zsurka , Kunz WS

  • Tools such as mitoTALENs and mitochondrial ZFNs are able to cut mtDNA in a sequence-specific manner. Whilst these methods are known to be able to reduce the proportion of mutant mtDNAs, the mechanism by which linear mtDNAs are degraded is unknown.
  • The authors show that linear mtDNA is eliminated within hours
  • Inactivation of the mitochondrial exonuclease MGME1; inhibition of the exonuclease activity of POLG; or knockdown of the mitochondrial DNA helicase TWNK leads to severe imediment of mtDNA degradation. 
  • Failure to remove damaged mtDNA leads to the accumulation of abnormal linear and rearranged molecules.
  • Degradation of linearized mtDNA is performed by the same machinery that is responsible for mtDNA replication.

Mitochondrial nicotinamide adenine dinucleotide reduced (NADH) oxidation links the tricarboxylic acid (TCA) cycle with methionine metabolism and nuclear DNA methylation

Lozoya OA, Martinez-Reyes I, Wang T, Grenet D, Bushel P, Li J, Chandel N, Woychik RP, Santos JH

  • The authors use an inducible dominant-negative mutant of the mtDNA polymerase POLG in human embryonic kidney cells (HEK293), see here (DN-POLG cells).
  • The authors perform RNA-seq on cells with induced DN-POLG at 0, 3, 6 and 9 days after treatment, with approximately 100%, 30%, 10% and 0% mtDNA copy number respectively.
  • Loss of mtDNA resulted in DNA hypermethylation. 57% of differentially expressed genes showed significant alteration in their promoter methylation compared with day 0.
  • The authors expressed two non-mammalian proteins (NDI1/AOX, see here) which by-pass the mitochondria allowing a normal NAD+/NADH balance to be maintained. In doing this, the authors found far fewer differentially expressed genes when comparing day 0 to day 9 cells. Furthermore, no significant changes in DNA methylation were observed in cells expresing NDI1/AOX. 
  • The authors suggest that mtDNA depletion induces imbalance in the NAD+/NADH pool, causing DNA methylation and pathological gene transcription

Wednesday, 18 April 2018

Amelioration of premature aging in mtDNA mutator mouse by exercise: the interplay of oxidative stress, PGC-1 a, p53, and DNA damage. A hypothesis

Safdar A, Annis S, Kraytsberg Y, Laverack C, Saleem A, Popadin K, Woods DC, Tilly JL, Khrapko K

(Not a new article, but thought-provoking)

  • The PolG mutator mouse accumulates single nucleotide polymorphisms in its mtDNA and shows premature ageing phenotypes
  • The threshold effect states that heteroplasmy must reach very high levels before a phenotype is shown. Although each molecule of mtDNA will possess several mutations, the activity of several mitochondrial enzymes are unaffected (whilst others such as Complex IV show activities of about 35%). In this sense, mtDNA mutations can be thought of as recessive.
  • The authors here suggest that an accumulation of reactive oxygen species from a diversity of different mutations could explain how the mutator mouse displays premature ageing phenotypes.
  • Endurance exercise is able to rescue the mutator phenotype in large part
  • Somatic tissues and the stem cell pool suffer from oxidative stress in PolG mice. Anti-oxidants have been shown to ameliorate some phenotypes of the mutator mouse.
  • Mitochondrial ROS induces telomere erosion
  • ROS (in particular, hydroxyl free radicals) can react with the nucleotide guanine to form 8-OHdG. This is a marker of oxidative stress, and is referred to by the authors as "non-mutational oxidative DNA damage). In the PolG mutator mouse, endurance exercised mice have lower 8-OHdG levels by ~x3.
  • A muscle-specific knockout of p53 abolishes the amelioration of the PolG phenotype by exercise. Importantly, p53 is a mtDNA repair protein, and can repair oxidative damage.
  • The authors also highlight that non-mutational mtDNA damage (e.g. 8-OHdG) can be converted into spurious mutations during PCR amplification
  • Note that the transcription machinery of the cell is also prone to making mistakes at sites of oxidative DNA damage
  • The authors revive the idea of the classical "vicious cycle" of mtDNA damage (mtDNA damage -> ROS -> more mtDNA damage) but instead of nucleotide substitutions the authors suggest that non-mutational mtDNA damage could be a potentially explanatory hypothesis.
  • Furthermore, the authors suggest that mitochondria compete with the nucleus for p53 during oxidative stress: mtDNA damage -> ROS -> nuclear DNA damage -> translocation of p53 to nucleus -> prevention of mtDNA repair -> mtDNA damage. The authors label this a "malicious cycle"
  • The authors suggest that exercise promotes PGC-1a, which promotes the expression of antioxidants, which lowers the rate of nuclear DNA damage, allowing p53 to leave the nucleus. 

  • In the "malicious cycle" hypothesis, the authors speculate that ROS induces nuclear DNA damage, causing p53 to translocate to the nucleus, meaning that mtDNA is repaired less. So implicit to this assumption is that the nucleus takes higher precedence over mtDNA in the context of oxidative stress. If so, that's interesting. Why not upregulate p53 so that it is not limiting, and both the nucleus and mtDNA can be repaired?

Thursday, 12 April 2018

So Happy Together: The Storied Marriage Between Mitochondria and the Mind.

Ruth F. McCan, David A. Ross

Many neuronal functions need mitochondria. Therefore, one would expect mitochondrial damage to to affect the nervous system. Indeed, we see a higher than normal incidence of psychiatric illnesses in people with genetic mitochondrial disorders and depressive episodes have been observed in mouse model of genetic mitochondrial diseases.

Another area of research is exploring the other direction of causality: can psychological stress and depression cause mitochondrial dysfunctions?
One proposed mechanism involves glucocorticoids. Experiments with cultured mouse neurons suggest that mitochondria are impaired by long-term exposure to glucocorticoids, which may be overproduced in states of stress and depression.
Another hypothesis involves oxidative stress (OS), caused by reactive oxygen species (ROS), which are produced by mitochondria. Biomarkers of OS are increased in people with depression, and other mood and anxiety disorders seem associated to OS. It might be that stress leads to a hyper-metabolic state in which mitochondria produce more ROS. These are toxic to mitochondria themselves, which are very vulnerable to oxidative damage, potentially causing more ROS production in a vicious cycle.
A study highlighting potential connections between stress, depression and mitochondria was published in 2015 by Cai et al., who collaborated with more than 60 scientists to look at a cohort of 11,670 women from China, through whole-genome sequencing of saliva samples.
It was found that women who had experienced stressful life events and depression had shortened telomeres, something which can be seen in settings of OS. It might be that stress acts on mitochondria, triggering a cascade which leads to depression. However, another possible explanation can be that stress take people who are more prone to depression and triggers an overdrive state in which mitochondria become overwhelmed, leading to OS. Therefore, it is not clear whether mitochondrial dysfunctions cause stress or the other way around.

Another question which is attracting interest concerns our mitochondria are involved in synaptic health and dysfunction in depression. It is thought that in depression neurons atrophy, synapses vanish and dendrites shrink. Although it is not clear whether or not (and how) these morphological changes are causally connected to the disease, it is worth considering the underlying mechanism. It is easy to imagine that large amounts of energy are required to create new neurons and synapses, so it is likely that mitochondria play a role. Moreover, mitochondria are involved in the regulation of intracellular calcium leves, which is crucial at synapses, since calcium stimulates neurotransmitter release.

It is clear that the more we learn about mitochondria, the more they can help unravel the connections between neurotransmitters, mood states, genetic diseases and psychiatric symptoms, life experiences and mental health.

Tuesday, 27 March 2018

Mice lacking the mitochondrial exonuclease MGME1 accumulate mtDNA deletions without developing progeria

Stanka Matic, Min Jiang, Thomas J. Nicholls, Jay P. Uhler, Caren Dirksen-Schwanenland, Paola Loguercio Polosa, Marie-Lune Simard, Xinping Li, Ilian Atanassov, Oliver Rackham, Aleksandra Filipovska, James B. Stewart, Maria Falkenberg, Nils-Göran Larsson & Dusanka Milenkovic
  • The authors developed a knockout mouse model for the gene MGME1.  Loss of MGME1 expression in siRNA treated cells, or patient fibroblasts, leads to an accumulation of 7S DNA.
  • 7S DNA is a single-stranded, ~650 nt long, nascent DNA species that creates a characteristic triple-stranded DNA structure in mtDNA called the D-loop. 7S DNA is formed by premature replication termination of mtDNA. 
  • Human patients with loss-of-function MGME1 mutations show depletions/ rearrangements of mtDNA, and a number of devastating phenotypes.
  • MGME1 is not essential for embryonic development, but its loss leads to accumulation of multiple deletions and depletion of mtDNA.
  • In the heart, these mice possessed ~50% less wild-type mtDNA (Southern blot analysis), and ~x5 more 7SDNA than mtDNA (which is around ~x5 higher ratio than seen in wild-type mice).
  • The authors found severe wt-mtDNA depletion in: kidney, liver, brain and heart. Skeletal muscle was least severely affected (if at all). (NB: this is from visual inspection of the Southern blots in Fig 2D).
  • Mice with both germline and tissue-specific knockout of Mgme1 are viable and appear healthy, despite the existence of deletions and rearrangements of mtDNA.
  • The authors found that the deleted species was a linear mtDNA molecule of ~11kb (~67% mass of normal mtDNA) due to the stalling of mtDNA replication
  • Mgme KO mice had similar levels of point mutations as wild-type mice
  • There were no clear OXPHOS defects in Mgme KO mice, despite depletion of mtDNA, accumulation of linear subgenomic mtDNA, and replication stalling in young animals. The mice did not show premature ageing.
  • In liver tissue, a prominent stalling site is observed, whereas in heart, a range of replication intermediates is observed. This shows the existence of tissue-specific stalling profiles in this system.
  • The authors suggest that MGME1 might be part of a regulatory switch acting at the end of the D-loop region that controls mtDNA replication and heavy-strand transcription termination. 

  • Do the authors still see that the mice are susceptible to arrhythmias, as observed here (blog here)? In other words, are there phenotypes in these mice which only become evident when the animals are challenged e.g. through stress?
  • How do we resolve the observation that loss of function of MGME1 in humans causes devastating phenotypes, whereas in mice it does not? 
  • If these deletions are all linear, can they still transcribe? If not, to what extent are these deletions representative of deletions which occur naturally, for instance the common deletion, which (as I understand it) are circular, see here?

Tuesday, 20 March 2018

Age-Associated Impairments in Mitochondrial ADP Sensitivity Contribute to Redox Stress in Senescent Human Skeletal Muscle

Graham P. Holloway, Andrew M. Holwerda, Paula M. Miotto, Marlou L. Dirks, Lex B. Verdijk, and Luc J.C. van Loon

  • The authors sought to determine whether there is an age-associated increase in mitochondrial reactive oxygen species (ROS) in vitro, using permeabilized muscle fibres.
  • They find that the capacity of mitochondrial H2O2 emission does not increase with ageing
  • However, ADP sensitivity does reduce with age. Consequently, H2O2 levels increase with age.
  • Increasing muscle mass, strength, and maximal mitochondrial respiration through exercise in older individuals did not alter H2O2 emission rates, the fraction of electron leak to H2O2 or the redox state of muscle.
  • In summary, reduction in mitochondrial ADP sensitivity increases mitochondrial H2O2 emission, which cannot be rescued through resistance training in later life (although there were other benefits to health of these individuals).

Optimized Mitochondrial Targeting of Proteins Encoded by Modified mRNAs Rescues Cells Harboring Mutations in mtATP6

Randall Marcelo Chin, Tadas Panavas, Jeffrey M. Brown, and Krista K. Johnson

  • Allotopic expression where a gene ordinarily encoded by mitochondrial DNA (mtDNA) is placed inside the nucleus, and modified such that the resultant protein is correctly transported into the mitochondria.
  • It is hoped that allotopic expression may be able to rescue pathologies which arise due to mutations in mitochondrial DNA: indeed, allotopic expression-based gene therapy is in phase 3 clinical trials for the mitochondrial disease LHON. 
  • mtDNA-encoded proteins are highly hydrophobic, causing them to often fold into import-incompetent states, thereby preventing them from entering the mitochondria. 
  • Mitochondrial targeting sequences (MTSs) and 3' untranslated regions (3' UTRs) have been used to target proteins or mRNA to the mitochondria. 
  • In this study, the authors performed a screen of 31MTSs and 15 UTRs in their ability to localize up to 9 allotopically expressed proteins to the mitochondrial DNA (note that mtDNA encodes 13 proteins, 22 tRNAs and 2 rRNAs).
  • Cybrid cells harbouring the 8993T>G point mutation in the mtATP6 gene were transiently transfected with a construct which was able to allotopically express mtATP6 and rescue the mtATP6-deficient cells.

Wednesday, 14 March 2018

Mitochondrial DNA as an inflammatory mediator in cardiovascular diseases

Hiroyuki Nakayama and Kinya Otsu

In this review, the authors discuss the role of mitochondria, and especially mitochondrial DNA (mtDNA) in triggering and maintaining cardiac inflammation. In this blost post we only summarise some parts of the review directly related to mtDNA.

We all know that the immune system provides protection against microorganisms such as bacteria, viruses, and fungi. This is achieved by sensing both pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs).

One major mechanism for activating the innate immune system is the sensing of pathogen-derived nucleic acids, and this is where mitochondria come into play. Due to their bacterial origin, mtDNA shares similarities with bacterial DNA (e.g. it contains cardiolipin and a predominantly unmethylated CpG motif). Mitochondria also release other DAMPs which can bind and activate multiple pattern recognition receptors similar to those activated by PAMPs.

MtDNA can leave the mitochondria and enter the cytoplasm or leave the entire cell. Opening of the mitochondrial transition pore plays an important role of mtDNA release from mitochondria, as inhibition of pore opening reduced levels of mtDNA in the cyotosol. MtDNA release is also controlled by other regulatory proteins such as the voltage dependent anion channel, Bax, and Bak.

MtDNA released after cell death functions as a DAMP. The mechanism of releasing mtDNA from non-nectrotic cells remains unclear, though exosomal release is proposed to be involved in this mechanism.  MtDNA enters the endocytic pathway by endocytosis and stimulates pattern recognition receptors which eventually leads to inflammasome formation.

It is important to degrade extracellular mtDNA to inhibit unnecessary inflammatory responses. It could be that mtDNA, like other non-host DNA in circulation, is digested in part by circulating nucleases. It is, however, unclear whether this occurs in physiological conditions, especially when mtDNA exists in microvesicles such as exosomes. Inside cells, DNasell plays an important role in mtDNA degradation

Levels of circulating mtDNA increase with age and correlate with levels of pro-inflammatory cytokines. Therefore, mtDNA-induced inflammatory responses can be involved in age-related cardiovascular disease, heart failure and atherosclerosis.

During heart failure, multiple endogenous DAMPs (including mtDNA) are released and recognized to induce an inflammatory response. However, no associated was found between the severity of heart failure and mtDNA levels in serum of patients (patients do show much higher serum mtDNA levels compared to controls).


Monday, 12 March 2018

Efficient termination of nuclear lncRNA transcription promotes mitochondrial genome maintenance

Dorine Jeanne Mariëtte du Mee, Maxim Ivanov, Joseph Paul Parker, Stephen Buratowski, Sebastian Marquardt

  • Most of the DNA of eukaryotes does not code for protein, yet many such regions of non-coding DNA are still transcribed into RNA (these are called lncRNA). Understanding the biological functions of these large expanses of non-coding DNA is an active area of current research.
  • Here, the authors show that a particular non-coding RNA in budding yeast (CUT60) is required for the proper transcription of its neighbouring gene ATP16. Mutations in CUT60 could result in the fusion of the lncRNA with the RNA of ATP16, and consequently the ATP16 protein could not be produced.
  • Interestingly, this had the consequence of yeast cells losing their mitochondrial DNA. ATP16 constitutes a subunit of ATP-synthase, which perhaps explains this striking phenotype.
  • The authors speculate that loss of mtDNA triggered by controlled, transient, transcription termination efficiency of CUT60 could allow cells to detoxify themselves of deleterious mtDNA 
The authors raise an interesting idea that yeast cells may be able to cleans their mtDNA through reducing CUT60 transcription termination efficiency. They suggest that cells may shed their mtDNA, and then gain healthy copies of molecules through mating. We know from animals that, during development, the mtDNA bottleneck serves to cleanse the developing embryo of deleterious mtDNA mutations. I wonder whether yeast cells could use this mechanism to a gentler extent (not completely losing their mtDNA, just reducing their mtDNA copy number) to bottleneck their mtDNA?

Thursday, 1 March 2018

Hallmarks of Cellular Senescence

Alejandra Hernandez-Segura, Jamil Nehme, Marco Demaria

A senescent cell is one which permanently stops dividing. In vitro, this can be caused by various stimuli, although it is unclear which amongst these cause senescence in vivo. The accumulation of senescent cells is observed through ageing, and a growing body of evidence is pointing towards the removal of senescent cells as a strategy to combat ageing.

Types of senescence currently known:

- DNA damage-induced senescence. This can be induced in vitro through radiation or drugs
- Oncogene-induced senescence. Activation of oncogenes (e.g. Ras or BRAF) or inactivation of tumour suppressors (e.g. PTEN) can induce senescence
- Chemotherapy-induced senescence. Drugs such as bleomycin or doxorubicin induce DNA damage. Drugs such as abemaciclib and palbociclib can inhibit cyclin-dependent kinases which regulate the cell cycle
- Mitochondrial dysfunction-associated senescence. The so-called "senescence associated secretory phenotype" (SASP) appears to be characteristic of this kind of senescence
- Epigenetically induced senescence. Inhibitors of DNA methylases or histone deacetylases can cause senescence
- Paracrine senescence. Senescence can be induced via the SASP produced by primary senescent cells

The senescence phenotype is often characterised by:

- Activation of a chronic DNA damage response
- Engagement of various cyclin-dependent kinase inhibitors
- SASP (which comprises, in part, various proinflamatory and tissue-remodelling factors)
- Induction of anti-apoptotic genes
- Altered metabolic rates
- Endoplasmic reticulum stress
- Consequent to the above, senescent cells are: enlarged and more flattened; have altered plasma membrane composition; accumulate lysosomes and mitochondria.

Current methods which are used to detect senescent cells include:

- DNA damage response: Immunostaining for γ-H2AX, p53
- Cell cycle arrest: Measurement of colony-formation potential or DNA synthesis rate via BrdU/EdU-incorporation. Expression level of the cyclin-dependent kinase inhibitors p16 and p21.
- Secretory phenotype: Cytokines (IL-1a, IL-6 and IL-8) , chemokines (CCL2) and metalloproteinases (MMP-1, MMP-3). However, the SASP is heterogeneous.
- Apoptosis resistance: Upregulation of BCL-proteins, BCL-2, Bcl-w or Bcl-xL.
- Cell size: Enlarged cell body and irregular shape using bright-field microscopy. Immunofluorescence targetting vimentin, actin or other cytoplasmic proteins have been used.
- Increased lysosomal content: e.g. SA-βgal, SSB, GL13, LysoTrackers, orange acridine
- Accumulation of mitochondria: MitoTrackers

Transitional correlation between inner-membrane potential and ATP levels of neuronal mitochondria

In this paper, the authors simultaneously measure mitochondrial membrane potential (Δψ) and mitochondrial ATP production (ATPmito) in dorsal root ganglion neurons from rat embryos. They measure the dynamics of Δψ and ATPmito, as well as their correlation, during physiological neuronal activity and focus on the following questions:

Is there a relation between Δψ, ATPmito and 
  • mitochondrial size?
  • mitochondrial transport velocity?
  • mitochondrial transport direction?
 Furthermore, they ask the question
  • What happens to Δψ and ATPmito during mitochondrial fusion and fission events? 20 fusion events and 20 fission events were investigated.

Some of the Results:
  • Δψ and ATPmito were compared among anterogradely transported, retrogradely transported, and stationary mitochondria in axons. Retrogradely transported mitochondria had slightly lower Δψ compared to anterogradely transported mitochondria.
  • No correlation was found between Δψ, ATPmito and mitochondrial velocity and transported distance.
  • Post-fusion mitochondrial membrane potential Δψ seemed to be higher than the average of the pre-fusion potentials (i.e. Δψfused > 0.5 (Δψpre-1 + Δψpre-2)).*
  • ATPmito was higher in the post-fusion mitochondrion compared to the average of the two pre-fusion mitochondria
  •  The two post-fission mitochondria tended to have different values for
    Δψ and their average was typically lower than the pre-fission potential (again, no information on size was provided as discussed below*).
  • No changes in ATPmito were observed upon fission
  • Mitochondrial density was higher in growth cones (an extension of a developing or regenerating neurite seeking its synaptic target) compared to axons. 
  • Average ATPmito levels were slightly lower in growth cones, though integrated ATP levels (over all mitochondria) were higher. 
  • Average Δψ was higher in growth cones compared to axons
  • Higher ATP levels in growth cones led to faster elongation of the axon, though no correlation between elongation speed and Δψ was found.
  • ATPmito tends to follow a change in Δψ (i.e. the change in Δψ occurs first). 
    ATPmito and Δψ are not necessarily always correlated.
  • Various other results were obtained which you can find by reading the paper!

*We note that no information was provided regarding mitochondrial size. If one of the pre-fusion mitochondria is much larger than the other, we might expect the membrane potential of the former to have more influence on the final potential of the post-fusion mitochondrion. In this case, one would not expect  Δψfused to be the arithmetic average of the two pre-fusion potentials.

Wednesday, 28 February 2018

Circadian Control of DRP1 Activity Regulates Mitochondrial Dynamics and Bioenergetics
Schmitt K, Grimm A, Dallmann R, Oettinghaus B, Restelli LM, Witzig M, Ishihara N, Mihara K, Ripperger JA, Albrecht U, Frank S, Brown SA, Eckert A

  • Circadian regulation of dynamin-related protein 1 (DRP1), a key mitochondrial fission protein, results in daily cycles of fission and fusion which are essential for circadian oscillations in ATP production
  • Genetic and pharmacological abrogation of DRP1 activity abolished circadian network dynamics and eliminated circadian ATP production

Wednesday, 14 February 2018

Myosin VI-Dependent Actin Cages Encapsulate Parkin-Positive Damaged Mitochondria

Antonina J. Kruppa,Chieko Kishi-Itakura, Thomas A. Masters, Joanna E. Rorbach, Guinevere L. Grice, John Kendrick-Jones, James A. Nathan, Michal Minczuk, Folma Buss

  • The authors identify a protein MYO6 which triggers the formation of F-actin cages to form around damaged mitochondria (mitochondria were damaged using a variety of pharmacological means)
  • These cages form a physical barrier, preventing damaged mitochondria from refusing with the network
  • MYO6 interacts with other proteins known to recruit autophagosomes to damaged mitochondria

Friday, 9 February 2018

Affinity purification of cell-specific mitochondria from whole animals resolves patterns of genetic mosaicism

Arnaud Ahier, Chuan-Yang Dai, Andrea Tweedie, Ayenachew Bezawork-Geleta, Ina Kirmes & Steven Zuryn

  • The authors demonstrate a technique called Cell-specific mitochondrial affinity purification (CS-MAP) to yield intact, functional, mitochondrial with >96% enrichment, >96% purity at single-cell and single-animal resolution in C. elegans.
  • CS-MAP consists of tagging the outer mitochondrial membrane protein TOMM-20 with a fluorophore and a particular epitope (which is something that an antibody can bind to) called HA. They placed the fusion protein under the control of a tissue-specific promoter so that e.g. muscle-specific mitochondria could be tagged. Mitochondria can then be purified using magnetic beads coated with anti-HA antibody and performing immunoprecipitation.
  • The authors crossed worms containing the CS-MAP construct with animals containing mitochondrial DNA deletions, and analysed the heteroplasmy in the individual mitochondria purified from different cell types and compared this to homogenate heteroplasmy across the whole animal. The authors found that intestine and neurons had significantly lower heteroplasmy than the homogenate, whereas the germline had significantly higher heteroplasmy than the homogenate.
  • The authors also quantified mtDNA copy number per mitochondrion in a number of different tissues, finding that the germline had ~3.5 mtDNAs per mitochondrion whereas tissues such as neurons and the intestine had ~1.5 mtDNAs per mitochondrion.
  • Using three-dimensional reconstruction from fluorescence images, the authors found that individual germ cells contained 71.2+/-6.5 mtDNAs per cell, whereas neurons contained 14.4+/-0.5 mtDNAs per cell.
  • The authors suggest that mtDNA turnover is higher in germ cells, which may account for their observations of increased copy number and heteroplasmy in the germline

Check out a pre-print from our group with some ideas on how increases in copy number may be related to heteroplasmy, and thoughts about comparisons between homogenate heteroplasmy and cellular heteroplasmy