|HSP||heavy strand promoter|
|LSP||light strand promoter|
|MGME1||mitochondrial genome maintenance exonuclease 1|
|mtSSB||mitochondrial single-stranded DNA-binding protein|
- Mitochondrial DNA (mtDNA) is a 16.6 kilobase double-stranded molecule.
- One strand of mitochondrial DNA is rich in guanines, allowing density centrifugation in alkaline CsCl2 gradients to distinguish between a heavy (H) and a light (L) strand.
- The mitochondrial DNA (mtDNA) comprises a lengthier noncoding region (NCR), also known as the regulatory region.
- In the NCR, each mtDNA strand has its own promoter for polycistronic transcription: the light strand promoter (LSP) and the heavy strand promoter (HSP) (HSP).
- Additionally, the NCR contains the origin for H-strand DNA replication (OH). A second origin for L-strand DNA replication (OL) is found outside of the NCR, in a tRNA cluster about 11,000 bp after OH.
Structure of Human mtDNA
- Thirteen messenger RNAs (mRNAs; green), twenty-two tRNAs (tRNAs; violet), and two ribosomal RNAs (rRNAs; pale blue) are all encoded by the genome.
- The top shows an enlargement of a significant noncoding region (NCR).
- In addition to the H-strand origin of replication (OH) and the termination-associated sequence, the main NCR also includes the light-strand promoter (LSP), the heavy-strand promoter (HSP), three conserved sequence boxes (CSB1-3, orange), and the HSP (TAS, yellow).
- Disruption of ongoing DNA synthesis at TAS results in the formation of the triple-stranded displacement-loop (D-loop) structure.
- 7S DNA is the name given to the short H-strand replication product generated in this way.
- The L-strand replication origin of DNA can be found in a small NCR about 11,000 base pairs (bp) after the OH (OL).
Factors for mtDNA replication
- Mammalian mitochondrial DNA (mtDNA) is copied by proteins that are different from those that copy nuclear DNA, and many of these proteins are similar to replication factors found in bacteriophages.
- DNA polymerase γ (POLγ) is the polymerase in mitochondria that makes copies of DNA. In human cells, POLγ is a heterotrimer with one subunit that does the work (POLγ A) and two subunits that help it do its job (POLγB).
- Taking away POLγA and POLγB from mice has shown that both are important for embryonic development.
- At least four more polymerases (PrimPol, DNA polymerase θ, DNA polymerase β, and DNA polymerase ζ) are thought to play a role in mitochondria.
- These polymerases are not needed to keep mtDNA alive, and none of them can replace POLγ.
- Most likely, they help fix mtDNA, but the exact role of these extra polymerases in keeping mtDNA in good shape needs to be figured out.
- POLγA is a DNA polymerase from the family A. It has a 3′–5′ exonuclease domain that checks the newly made DNA strand for mistakes.
- POLγ is a very accurate DNA polymerase that makes mistakes less than 1 x 10^-6 of the time.
- The POLBγ subunit is an add-on that helps POLγA interact better with the DNA template and increase both its catalytic activity and its speed.
- POLγ can’t use double-stranded DNA as a template, so at the mitochondrial replication fork, there needs to be a DNA helicase.
- The DNA helicase TWINKLE is similar to the T7 phage gene 4 protein. During mtDNA replication, TWINKLE moves in front of POLγ and unwinds the double-stranded DNA template.
- TWINKLE forms a hexamer and needs a fork structure (a single-stranded 5′-DNA loading site and a short 3′-tail) to load and start unwinding.
- Mitochondrial single-stranded DNA-binding protein (mtSSB) binds to newly formed single-stranded DNA (ssDNA), protects it from nucleases, and stops the formation of secondary structures.
- mtSSB helps make more mtDNA by making TWINKLE’s helicase activity stronger and by making POLγ more active.
Mode of Mitochondrial DNA Replication (mtDNA)
- In 1972, Vinograd and his colleagues showed a model for how mtDNA copies itself. Their strand displacement model says that DNA is made all the time on both the H- and L-strands.
- Each strand, OH and OL, has its own start point.
- First, replication starts at the OH site, and then DNA synthesis makes a new H-strand.
- During the first phase, no L-strands are made at the same time, and the displaced, parental H-strand is covered by mtSSB.
- By binding to single-stranded DNA, mtSSB stops the mitochondrial RNA polymerase (POLRMT) from starting random RNA synthesis on the displaced strand.
- When the replication fork has moved about two-thirds of the way through the genome, it passes the second origin of replication, OL.
- When the parental H-strand at OL is shown in its single-stranded form, it folds into a stem–loop structure.
- The stem stops mtSSB from binding very well, so a short stretch of single-stranded DNA in the loop region is still accessible. This lets POLRMT start making RNA.
- POLRMT does not work on DNA templates with only one strand. After about 25 nt, POLγ takes its place, and L-strand DNA synthesis starts.
- From this point on, H-strand and L-strand synthesis keep going until the two strands have gone all the way around.
- Replication of the two strands is linked because the H-strand must be made before the L-strand can be made.
- Both in vivo and in vitro studies of the structure and sequence requirements of mammalian OL have shown that a functional human OL must have a stable double-stranded stem region with a pyrimidine-rich template strand and a single-stranded loop of at least 10 nucleotides (nt).
Replication of the human mitochondrial genome
- Mitochondrial DNA replication starts at OH and goes in one direction until the nascent H-strand is full length.
- mtSSB binds to the exposed parental H-strand and keeps it safe.
- When the replisome passes OL, a stem-loop structure is made that prevents mtSSB from binding. This leaves a single-stranded loop-region where POLRMT can start making primers.
- After about 25 nt, when POLγ replaces POLRMT at the 3′ end of the primer, the process changes to making L-strand DNA.
- The two strands are made in a continuous process until there are two full, double-stranded DNA molecules.
What is the D-loop?
Strangely, not all replication events that start at OH go all the way back to where they started. Instead, 95% are already over after about 650 nt at the sequences that signal the end (TAS). The short piece of DNA that is made this way is called 7S DNA. It stays attached to the parental L-strand, but the parental H-strand moves. As a result, a D-loop structure is made, which is made up of three strands of displacement loops.
- It’s not clear what the D-loop structure does, and it’s also not clear how replication ends at the TAS.
- It seems, though, that stopping at TAS is a controlled event that acts as a switch between mtDNA replication that ends early and replication that goes all the way through the genome.
- In vivo occupancy analysis showed that POLγ stops moving at the 3′-end of the D-loop, while TWINKLE occupancy is low in this area, which supports this idea.
- When mtDNA is used up, the situation changes. TWINKLE occupancy goes up, and 7S DNA levels go down at the same time.
- Because there is more demand for mtDNA replication, these data have been seen as proof that TWINKLE reloads.
- If the helicase binds to the 3′ end of 7S DNA, the stuck POLγ can keep copying 7S DNA until it goes full circle.
- Experiments with the genes of mice support the model and show that TWINKLE is important for controlling the number of copies of mtDNA. When the levels of mtDNA go up or down, so do the levels of TWINKLE. So, it’s possible that mtDNA replication is controlled at the pretermination level instead of the initiation level.
- The switch could fine-tune the number of mtDNA copies based on what each cell needs.
- On each side of the D-loop region, there are two ATGN9CAT sequence motifs that are closely related and have stayed the same over time.
- One motif is near the 5′ end of the 7S DNA, where it is part of the CSB1 (Conserved Sequence Box 1) sequence.
- The second motif, called core-TAS, is in the TAS region, right after the 3′ end of 7S DNA (Figure 1, top panel).
- It’s still not clear what these motifs do in the body, but proteins that bind to DNA often recognise and bind to palindromic sequences.
- Organello footprints have been found in the TAS region, which supports this idea. However, despite a lot of work in different labs, a TAS-binding protein has not yet been found.
- It’s possible that traditional methods make it hard to purify the proteins that bind to CSB1, core-TAS, and other parts of TAS. Maybe the protein that is missing is stuck to a membrane and hard to keep in solution during chromatography.
- Binding could also be a controlled process that needs specific redox conditions or nucleotide concentrations.
- Lastly, it can’t be ruled out that secondary structures in mtDNA, like stem-loops or G-quadruplexes, could also play a role and add to the DNA footprints seen in living cells.
Initiation of Mitochondrial DNA Replication (mtDNA) at OH
- We know that the primers needed to start H-strand synthesis OH are made by POLRMT.
- When transcription starts at LSP, RNA 3 ends are made, which POL can use to start DNA synthesis.
- In human mitochondria, there are several RNA-to-DNA transition points located downstream of LSP. These points are clustered around two conserved sequence motifs, CSB3 and CSB2.
- These conserved sequence elements are high in guanine, and a G-quadruplex structure can form between new RNA and the nontemplate DNA strand at CSB2 during transcription.
- The new transcript is then attached to the mtDNA in this way, making an R-loop structure.
- The G-quadruplex structure also stops transcription before it should at sites in the CSB2-region that are similar to RNA-to-DNA transition sites.
- Based on these facts, it was thought that primer formation at OH might be caused by sequence-dependent transcription termination.
- The transcription elongation factor TEFM strongly decreases transcription termination and R-loop formation at CSB2. This suggests that active TEFM may change the ratio between primer formation and full-length, productive transcription.
- Also, there is no direct evidence from experiments that R-loop-forming transcripts that end too soon can be used directly by POLγ to start DNA synthesis.
- The ways in which DNA replication starts at OH may be similar to the ways in which DNA replication starts in the E. coli plasmid ColE1.
- In the plasmid, an RNAII transcript binds to the template strand to make an R-loop, which is used to start the process of making DNA.
- Also, the ColE1 origin of replication is located after a guanine-rich stretch that is needed for both the beginning of replication and the formation of an R-loop.
- RNase H cuts the ColE1 R-loop before it is used to start DNA synthesis. It is still unknown if the mitochondrial RNASEH1 plays the same role in mammalian cells.
Termination of Mitochondrial DNA Replication (mtDNA)
- DNA ligase III joins the new DNA strands together after POLγ has finished making them.
- For ligation to work well, the 5-ends and 3-ends of the new DNA must be put next to each other. This means that the RNA primers that were used to start mtDNA synthesis must first be taken out.
- RNASEH1 is a likely candidate for primer removal because RNA primers are still found in the mitochondrial origin regions of mouse embryonic fibroblasts that don’t have Rnaseh1 and Rnaseh1 knockout mice lose their mtDNA.
- After a full circle-replication, POLγ meets the 5-end of the new full-length mtDNA strand it just made.
- At this point, POLγ starts a cycle of 3–5 exonuclease degradation and polymerization at the nick. Idling is a process that must happen for ligation to work right.
- POLγ that doesn’t have exonuclease activity can’t stop making DNA. Instead, it keeps making DNA in the dsDNA region past the 5-end, making a flap-like structure that can’t be joined.
- The fact that mice with exonuclease-deficient POLγ have strand-specific nicks at OH may be caused by the inability to make DNA ends that can be joined together.
- In an interesting twist, there is a major 5-end of nascent DNA about 100 bp after the RNA-to-DNA transition sites that have been found.
- Even though it was first thought that the free 5-end at position 191 was where mtDNA replication began, it may be made in other ways. For example, the nascent H-strand may go through a lot of 5′-end processing when the primer is removed. Not only is the RNA primer removed, but also about 100 nt of DNA further down the strand.
- This would separate the place where RNA turns into DNA from the place where nascent H-strands join at the end of replication.
- The mitochondrial genome maintenance exonuclease 1 (MGME1), a RecB-type exonuclease in mitochondria that is part of the PD-(D/E)XK nuclease superfamily, could be the cause of this effect.
- Human cells that don’t have active MGME1 have trouble joining at OH and make linear deleted mtDNA molecules that span OH and OL. There are also more 7S DNA molecules than before.
- The 5 ends of these 7S DNA products are closer to CSB2 than they are in normal cells, which suggests that MGME1 is involved in processing the 5-end of the new H-strand.
Separation Mitochondrial DNA (mtDNA)
- During DNA replication, the original molecule stays the same, which causes a problem for the moving machinery that does the replication.
- By letting one of the strands go through the other, type 1 topoisomerases can relieve the torsional strain caused by this.
- In the mitochondria of mammals, a type IB enzyme called TOP1MT can work with the mitochondrial replisome as a “swivel” for DNA. When the Top1mt gene is taken out of a mouse, it produces viable offspring with different mtDNA supercoiling.
- In other systems, when intact, circular DNA is copied, it makes daughter molecules that are mechanically linked together as catenanes. This means that the DNA circles are not yet finished.
- So, for complete separation of daughter molecules, replication of circular genomes needs decatenation.
- To resolve the hemicatenane structure, a mitochondrial isoform of Topoisomerase 3αα (Top3) is needed. Loss of Top3α causes mtDNA to decrease and form large networks of linked mtDNA.
- The fact that the hemicatenanes that hold these mtDNA networks together are in the OH-region is interesting. This suggests that these structures are made when mtDNA replication is finished.
- Even if Top3 is needed to separate newly copied mtDNA, it is likely that other proteins are also needed.
- There is a nuclear version of Top3 that works with three other proteins: the helicase BLM and the OB-fold proteins RMI1 and RMI2.
- Together, these proteins make the BTR complex, which breaks down double Holliday junctions. For Top3 to do its job as a topoisomerase, it needs the other subunits.
- But since neither BLM nor RMI1 nor RMI2 have mitochondrial isoforms, other proteins may work with Top3 in mitochondria to control and/or speed up its activity.
Replication of Nucleoid
- mtDNA is not a single molecule. Instead, it is part of large nucleoprotein complexes called nucleoids. There are different ways to use fluorescent microscopy to see nucleoids.
- The average size of a nucleoid is about 100 nm, and most of the time there is only one mtDNA molecule per nucleoid.
- The most important structural protein in the nucleoid is called TFAM. It has one subunit for every 16–17 bp of mtDNA.
- TFAM is a member of the HMG box domain family, and it can bind to DNA regardless of the sequence.
- TFAM is also an important part of the machinery that controls transcription in the mitochondria.
- During the beginning of transcription, the protein binds upstream of the transcription start site and makes the DNA bend sharply.
- By mixing TFAM and mtDNA, nucleoid-like particles can be made again. This suggests that TFAM can fully compact mtDNA on its own.
- TFAM has two places where it can bind to DNA, and it seems to make mtDNA more compact by binding across strands and making loops.
- TFAM also works with other proteins to bind to DNA, making protein patches on mtDNA.
- The idea that TFAM controls the replication of mtDNA through epigenetics is very interesting.
- Super-resolution microscopy has shown that nucleoids come in different shapes. Maybe the smaller nucleoids are a form of mtDNA storage, while the larger ones are active in replication and/or transcription.
- Nucleoids that are involved in active DNA replication have been found at places where the endoplasmic reticulum (ER) meets the mitochondria. At these places, mitochondria divide, which leads to the idea that contacts between the endoplasmic reticulum and mitochondria can coordinate the synthesis of mtDNA with division to make sure that the newly made nucleoids are spread out evenly in the mitochondrial network.
- Small changes in the ratio of TFAM to DNA in a test tube can have big effects. At physiological ratios, there are big differences in how tightly mtDNA is packed, and both fully packed nucleoids and naked DNA can be seen at the same time.
- Under these conditions, a small increase in the amount of TFAM can make a big difference in how many mtDNA molecules are fully compacted.
- This model explains why the amount of TFAM in vivo stays roughly proportional to the amount of mtDNA. It also suggests that small changes in the amount of TFAM can have big effects on both gene expression and mtDNA replication.
- This idea is supported by the fact that longer patches of TFAM stop DNA from unwinding, which stops the mtDNA replication and transcription machinery from moving forward.
- So, TFAM might be an epigenetic regulator that controls the number of mtDNA molecules that can be used for active transcription and/or mtDNA replication.
- mtDNA in mammals is replicated by proteins separate from those employed for replication of nuclear DNA.
- In accordance with the strand displacement concept, replication begins from two separate origins, OH and OL.
- The primer used by POL to trigger DNA synthesis at OH is derived from LSP-initiated transcripts.
- OL produces a stem–loop structure, and POLRMT begins primer production in the loop region of single-stranded DNA.
- The function of the mitochondrial D-loop is unknown.
- RNASEH1 and MGME1 play crucial roles in primer elimination, but the specifics of this mechanism remain unclear.
- Top3α is necessary for the resolution of hemicatenane structures between newly replicated mtDNA molecules.
- mtDNA is not a free-floating molecule, but is instead packed into nucleoprotein complexes called nucleoids.
- mtDNA synthesis is linked to mitochondrial division.
- Apostolova, N., & Esplugues, J. V. (2011). Mitochondrial DNA Replication in Health and Disease. In (Ed.), DNA Replication – Current Advances. IntechOpen. https://doi.org/10.5772/19162
- Holt IJ, Reyes A. Human mitochondrial DNA replication. Cold Spring Harb Perspect Biol. 2012 Dec 1;4(12):a012971. doi: 10.1101/cshperspect.a012971. PMID: 23143808; PMCID: PMC3504440.
- Falkenberg M. Mitochondrial DNA replication in mammalian cells: overview of the pathway. Essays Biochem. 2018 Jul 20;62(3):287-296. doi: 10.1042/EBC20170100. PMID: 29880722; PMCID: PMC6056714.
- Gustafson, Margaret & McCormick, Elizabeth & Perera, Lalith & Longley, Matthew & Bai, Renkui & Kong, Jianping & Dulik, Matthew & Shen, Lishuang & Goldstein, Amy & Mccormack, Shana & Laskin, Benjamin & Leroy, Bart & Ortiz-Gonzalez, Xilma & Ellington, Meredith & Copeland, William & Falk, Marni. (2019). Mitochondrial single-stranded DNA binding protein novel de novo SSBP1 mutation in a child with single large-scale mtDNA deletion (SLSMD) clinically manifesting as Pearson, Kearns-Sayre, and Leigh syndromes. PLOS ONE. 14. e0221829. 10.1371/journal.pone.0221829.