Replication machinery degrading linear mitochondrial DNA

Linear mitochondrial DNA (mtDNA) molecules with deletions are a curious case. To my knowledge, they are only present at high levels in a few genetic models, including the mtDNA mutator mouse(Trifunovic et al. 2004, Bailey et al. 2009), mtDNA mutator flies (Bratic et al. 2015), patients carrying mutations in mitochondrial genome maintenance exonuclease 1 (MGME1)(Nicholls et al. 2014) and MGME1 knockout mouse (Matic et al. 2018).

In the case of mtDNA mutator mouse, the level of this linear mtDNA with deletion remains stable with time (Kukat et al. 2009). Additionally, it has been shown using mitochondrially targeted restriction enzymes that linear mtDNA is rapidly degraded (Bayona-Bafaluy et al. 2005). Therefore the question remains, who is degrading these mtDNA fragments?

Moretton et al. tried to answer this question by linearizing mtDNA with mitochondrial targeted restriction enzyme followed by systematic knockdown of exonucleases known/suggested to be in mitochondria (Moretton et al. 2017). These exonucleases included ExoG, EndoG, MGME1, DNA2 and FEN1. EXD2 was not tested, but that was later shown to be localized on the mitochondrial outer membrane (Hensen et al. 2018). Perhaps surprisingly, none of the exonucleases studied by Moretton et al. seemed to participate in degradation of linear mtDNA fragments.

There is still at least one exonuclease that Moretton et al. did not study, which is the exonuclease activity of mitochondrial DNA polymerase (POLG). Indeed, in the absence of nucleotides POLG can engage in so called exonuclease mode where it starts degrading the primer in 3’-5’ direction (Bratic et al. 2015). A recent study in fruit flies tried to answer the question what enzyme degrades mtDNA in fly sperm (Yu et al. 2017). Yu et al. concluded that DmPOLG is involved in this degradation but it takes place independent of its exonuclease activity. I was not convinced by this paper because they saw mtDNA replication in punctae without DmPOLG-GFP and more importantly all the models were based on knocking down various proteins of the minimal mitochondrial replisome with variable efficiencies. Also, when I read this paper I was aware of some in-house work taking place in the group of Martin Graef.

Medeiros et al. showed in yeast that under starvation there is a strong mtDNA depletion if autophagy is inhibited and that this depletion can be rescued by increasing nucleoside pools through supplementation or genetically (Medeiros et al. 2018). Therefore, the authors hypothesized that under starvation the autophagy-deficient cells might have changes in nucleotide pools causing MIP1 (yeast POLG) to enter the above mentioned exonuclease mode and degrade mtDNA. Using an exonuclease-deficient POLG the authors showed that indeed it seems to be POLG which is degrading mtDNA.

The most recent paper on the subject comes from Peeva et al. (Peeva et al. 2018). Peeva et al. tested multiple exonucleases (EXOG, APEX2, ENDOG, FEN1, DNA2, MRE11, RBBP8, MGME1 and POLG) to see whether the knockout or knockdown of these enzymes would affect the degradation of linear mtDNA molecules. Somewhat similar to Medeiros et al., Peeva et al. also suggest that POLG together with MGME1 and TWINKLE would be degrading linear mtDNA.

So there it is. Both in yeast and human cells mtDNA seems to be degraded by the exonuclease activity of POLG. It would be interesting to see how the yeast mtDNA actually looks like at the molecular level when degraded as yeasts do not possess MGME1. One could therefore predict that these degraded molecules would not have blunt ends similarly to the results from humans where POLG is suggested to degrade mtDNA to the 3’-5’ direction and MGME1 to the 5’-3’ direction. It is still open question what controls this degradation, if it is controlled to begin with. Both Medeiros et al. and Peeva et al. hypothesize that low nucleotide pools might be a controlling factor. Although this hypothesis would make sense in the case of yeast under starvation and inhibited autophagy, I don’t see how this would work with these human cells under normal culture conditions. Clearly just linearizing mtDNA is sufficient to promote mtDNA degradation without manipulation of the dNTP pools.

Some open questions for the future:

• Is this kind of degradation present also in more in vivo conditions or whether it is only present in somewhat artificial conditions where most of mtDNA is linearized by restriction enzymes.
• What, if any, is controlling the balance between mtDNA synthesis and degradation?
• Is there a specific endonuclease cleaving the yeast mtDNA before POLG can start degrading mtDNA?
• How is mtDNA degraded in organisms without MGME1?

Perhaps this is a stretch but in Peeva et al. study the authors showed how in WT cells linear mtDNA is degraded piece by piece. Where mitophagy selective for damaged mtDNA to exist, which I think it does not (Kauppila et al. 2017), one would not expect to see these slowly degrading molecules but instead a bulk disappearance of linear mtDNA fragments.

References:

Bailey LJ, Cluett TJ, Reyes A, Prolla TA, Poulton J, Leeuwenburgh C, Holt IJ. Mice expressing an error-prone DNA polymerase in mitochondria display elevated replication pausing and chromosomal breakage at fragile sites of mitochondrial DNA. Nucleic Acids Res. 2009. PMID: 19244310

Bayona-Bafaluy MP, Blits B, Battersby BJ, Shoubridge EA, Moraes CT. Rapid directional shift of mitochondrial DNA heteroplasmy in animal tissues by a mitochondrially targeted restriction endonuclease. Proc Natl Acad Sci U S A. 2005. PMID: 16179392

Bratic A, Kauppila TE, Macao B, Grönke S, Siibak T, Stewart JB, Baggio F, Dols J, Partridge L, Falkenberg M, Wredenberg A, Larsson NG. Complementation between polymerase- and exonuclease-deficient mitochondrial DNA polymerase mutants in genomically engineered flies. Nat Commun. 2015. PMID: 26554610

Hensen F, Moretton A, van Esveld S, Farge G, Spelbrink JN. The mitochondrial outer-membrane location of the EXD2 exonuclease contradicts its direct role in nuclear DNA repair. Sci Rep. 2018. PMID: 29599527

Kauppila TES, Kauppila JHK, Larsson NG. Mammalian Mitochondria and Aging: An Update. Cell Metab. 2017. PMID: 28094012

Kukat A, Trifunovic A. Somatic mtDNA mutations and aging--facts and fancies. Exp Gerontol. 2009. PMID: 18585880

Matic S, Jiang M, Nicholls TJ, Uhler JP, Dirksen-Schwanenland C, Polosa PL, Simard ML, Li X, Atanassov I, Rackham O, Filipovska A, Stewart JB, Falkenberg M, Larsson NG, Milenkovic D. Mice lacking the mitochondrial exonuclease MGME1 accumulate mtDNA deletions without developing progeria. Nat Commun. 2018. PMID: 29572490

Medeiros TC, Thomas RL, Ghillebert R, Graef M. Autophagy balances mtDNA synthesis and degradation by DNA polymerase POLG during starvation. J Cell Biol. 2018. PMID: 29519802

Moretton A, Morel F, Macao B, Lachaume P, Ishak L, Lefebvre M, Garreau-Balandier I, Vernet P, Falkenberg M, Farge G. Selective mitochondrial DNA degradation following double-strand breaks. PLoS One. 2017. PMID: 28453550

Nicholls TJ, Zsurka G, Peeva V, Schöler S, Szczesny RJ, Cysewski D, Reyes A, Kornblum C, Sciacco M, Moggio M, Dziembowski A, Kunz WS, Minczuk M. Linear mtDNA fragments and unusual mtDNA rearrangements associated with pathological deficiency of MGME1 exonuclease. Hum Mol Genet. 2014. PMID: 24986917

Peeva V, Blei D, Trombly G, Corsi S, Szukszto MJ, Rebelo-Guiomar P, Gammage PA, Kudin AP, Becker C, Altmüller J, Minczuk M, Zsurka G, Kunz WS. Linear mitochondrial DNA is rapidly degraded by components of the replication machinery. Nat Commun. 2018. PMID: 29712893

Trifunovic A, Wredenberg A, Falkenberg M, Spelbrink JN, Rovio AT, Bruder CE, Bohlooly-Y M, Gidlöf S, Oldfors A, Wibom R, Törnell J, Jacobs HT, Larsson NG. Premature ageing in mice expressing defective mitochondrial DNA polymerase. Nature. 2004. PMID: 15164064

Yu Z, O'Farrell PH, Yakubovich N, DeLuca SZ. The Mitochondrial DNA Polymerase Promotes Elimination of Paternal Mitochondrial Genomes. Curr Biol. 2017. PMID: 28318978

The presence of rNTPs decreases the speed of mitochondrial DNA replication

In this time and age, it is delightful to see a publication which is not doing poorly controlled correlative stuff or omics for the sake of omics, but instead strong hypothesis driven biochemistry. The paper I am referring to comes from the lab of Sjoerd Wanrooij in Umea (Forslund et al. 2018).

Now, it has been known since the 70's that mitochondrial DNA (mtDNA) has ribonucleotides, but it has remained unclear whether they possess a function and how they affect mtDNA maintenance and gene expression. Until very recently, there has been very little attention given to these ribonucleotides, but it seems this subject is becoming trendy again. A recent paper from the Clausen lab showed for the first time the amount and distribution of ribonucleotides in whole mtDNA and also how the amount of ribonucleotides might be different in some patients with changes in nucleotide pools (Berglund et al. 2017). In this paper, Forslund et al. try to understand whether the mitochondrial DNA polymerase (POLG) is inhibited by ribonocleotides in the template strand or by changes in the relative ration of NTP to dNTP.

In the beginning of the paper Forslund et al. make a good point, which might be missed by a lot of people not working with POLG. Back in the 90's and early 00's, people performed all POLG in vitro experiments using the exonuclease-deficient mutant because it makes the experiments easier. However, it was later recognized that the exonuclease-deficient POLG might be doing some funky things such as increased strand-displacement (Farge et al. 2007). For this reason, one should interpret the old papers with some grain of salt.

I personally like these kind of experiments, where the models used are first established using previously published systems. Not only is this helpful in reproducing already published findings but it really makes you trust the experiment. I highly recommend reading the whole thing.

The weirdest finding of the paper is that only WT POLG but not the exonuclease-deficient POLG is sensitive to low NTP/high rNTP conditions. The authors suggest in the discussion that this could be caused idling of the WT POLG between polymerase/exonuclease modes. It is not obvious to be why the exonuclease-deficient POLG would not idle also. Could it be that the exo- mutation somehow alters the protein structure to have different nucleotide selection dynamics? Or perhaps the low NTP/high rNTP condition alters the efficiency of the WT POLG so that it will have increased removal of correct bases leading to decreased full length synthesis? I hope they will follow this up.

In the discussion the authors stated that their in vitro estimation of ribonucleotide content in mtDNA is somewhat lower than the values obtained from in vivo samples. One explanation for this could be that because the nucleotide pool sizes are so challenging to measure, it could be that the in vitro measurements are over-estimating the dNTP pools or under-estimating the NTP pools. To me, this is the most parsimonious answer. They also continue hypothesizing that some other polymerases, like Polβ or PrimPol, could introduce ribonucleotides in vivo. I'm yet to be convinced that either of these polymerases would be in mitochondrial matrix or have a mitochondrial function. For instance, depending on the study, small amount of PrimPol localizes to mitochondria or doesn't. Even when a fraction of PrimPol has been shown to localize within mitochondria, in subcellular fractination it seems to be degraded together with OPA1 (Torregrosa-Muñumer et al. 2017), suggesting PrimPol is in the inner membrane space. I put my money on over/under-estimated nucleotide pools.

In overall, once again strong biochemistry from Sjoerd's group.

References:

Berglund AK, Navarrete C, Engqvist MK, Hoberg E, Szilagyi Z, Taylor RW, Gustafsson CM, Falkenberg M, Clausen AR. Nucleotide pools dictate the identity and frequency of ribonucleotide incorporation in mitochondrial DNA. PLoS Genet. 2017. PMID: 28207748

Farge G, Pham XH, Holmlund T, Khorostov I, Falkenberg M. The accessory subunit B of DNA polymerase gamma is required for mitochondrial replisome function. Nucleic Acids Res. 2007. PMID: 17251196

Forslund JME, Pfeiffer A, Stojkovič G, Wanrooij PH, Wanrooij S. The presence of rNTPs decreases the speed of mitochondrial DNA replication. PLoS Genet. 2018 PMID: 29601571

Torregrosa-Muñumer R, Forslund JME, Goffart S, Pfeiffer A, Stojkovič G, Carvalho G, Al-Furoukh N, Blanco L, Wanrooij S, Pohjoismäki JLO. PrimPol is required for replication reinitiation after mtDNA damage. Proc Natl Acad Sci U S A. 2017. PMID:29073063