Transcript availability dictates the balance between strand-asynchronous and strand-coupled mitochondrial DNA replication

Abstract Mammalian mitochondria operate multiple mechanisms of DNA replication. In many cells and tissues a strand-asynchronous mechanism predominates over coupled leading and lagging-strand DNA synthesis. However, little is known of the factors that control or influence the different mechanisms of replication, and the idea that strand-asynchronous replication entails transient incorporation of transcripts (aka bootlaces) is controversial. A firm prediction of the bootlace model is that it depends on mitochondrial transcripts. Here, we show that elevated expression of Twinkle DNA helicase in human mitochondria induces bidirectional, coupled leading and lagging-strand DNA synthesis, at the expense of strand-asynchronous replication; and this switch is accompanied by decreases in the steady-state level of some mitochondrial transcripts. However, in the so-called minor arc of mitochondrial DNA where transcript levels remain high, the strand-asynchronous replication mechanism is instated. Hence, replication switches to a strand-coupled mechanism only where transcripts are scarce, thereby establishing a direct correlation between transcript availability and the mechanism of replication. Thus, these findings support a critical role of mitochondrial transcripts in the strand-asynchronous mechanism of mitochondrial DNA replication; and, as a corollary, mitochondrial RNA availability and RNA/DNA hybrid formation offer means of regulating the mechanisms of DNA replication in the organelle.


Supplementary Information
Transcript availability dictates the balance between strandasynchronous and strand-coupled mitochondrial DNA replication

Figure S2
. Transgenic Twinkle DNA helicase expressed in HEK293T cells remodels mtDNA replication to render the intermediates resistant to RNase H1 and single-stranded nuclease. 30 ng/mL doxycycline was added to control HEK293T cells and cells carrying Twinkle-HA (mutant or wild-type) transgenes: total cellular DNA was harvested 72 hours later. DNA was digested with Hinc2 and treated additionally with RNase HI and single-stranded nuclease (SSN) as indicated. The products of the reactions were separated by 2D-AGE, transferred to solid support and hybridized to a radiolabeled probe (h1) that detected the fragment spanning nt 13,636-1,006. Y -replication fork arc, SdY -supra-double Y arc; b -bubble or initiation arc, eeyebrow, e m -modified eyebrow. Note that transgenic Twinkle expression is associated with Y and bubble arcs that are RNase H and SSN nuclease resistant, which indicates the RIs comprise duplex DNA on all branches. In contrast, the SdY arc is sensitive to these enzymes.  (B) Illustrations of replication forks entering the same (Dra1) restriction fragment of mtDNA from opposite ends, and the consequent effects of subsequent cleavage with Nci1 (in-gel) on individual replication intermediates and on Y arc mobility (insets). If forks enter from both ends a composite pattern will be produced (gray broken line). (C) Control HEK293T cells yielded arc c that is incompatible with initiation of replication from Ori-H, and this was greatly enhanced by transgenic HA tagged Twinkle (Twk). The supra-Y arc (SY) (and its partially degraded derivative -dSY) results from unidirectional replication from the CR (5). 'Splaying' of the prominent descending portion of the Y is attributable to some branch migration occurring during the in-gel digestion step.

Figure S5. Strand-asynchronous replication is re-established in the minor arc of mtDNA in cells expressing transgenic wild-type Twinkle. (A)
The Bgl1-associated prominent double Yarc (covering the minor arc of human mtDNA), induced by transgenic wild-type Twinkle (Fig. 5B), is modified by RNase H, whereas the bubble arc (corresponding to the major arc of mtDNA) is unaffected by RNase H. (B) A novel arc of mtRIs appearing in cells expressing transgenic Twinkle DNA helicase is a double Y species with an extended arm owing to restriction site blockage. 2D blots equivalent to Fig. 1E, 1F were probed to determine the extent of restriction site blockage. The SdY arc is detected by probes h1, h6 and h7, but not h5. The position of the probes and the deduced blocked restriction sites are illustrated on a representative intermediate of the SdY arc (inset). These results indicate that the Hinc2 SdY arc associated with HEK293T cells expressing wild-type Twinkle spans nt 13636-5693, and because it is sensitive to RNase HI and SSN (Fig. S2) it is not formed of duplex DNA on all branches. (C) A fragment of human mtDNA (nt 2,650-6,286) spanning much of the minor arc detected by probe h5 yields a strong strand-asynchronous replication arc in samples derived from cells expressing HA tagged Twinkle wild-type (Twk-HA). Loss of tracts of RNA from the lagging-strand branch of mtRIs creates a Ylike arc with a lower trajectory than a conventional Y arc (sub-Y) in fragments spanning the minor arc and including the light-strand 'origin' of replication Ori-L (for further details see (6)), b -bubble structures, dY -double Y arc.

Figure S6. Mutant Twinkle causes mtDNA depletion, accompanied by an increase in transcripts per mtDNA.
Whole cellular DNAs were isolated from control HEK293T cells or cells expressing a mutated Twinkle-HA (D311A). Relative mtDNA copy number was calculated from the abundance of the cytochrome b gene of mtDNA relative to the single copy nuclear gene APP1, using real-time PCR quantification as previously described (7). (B) The abundance of five mitochondrial transcripts, cytochrome b, cytochrome c oxidase II, NADH dehydrogenase 1, 16S and 12S rRNA was measured by RT-q-PCR.

Figure S7. Transgenic Twinkle DNA helicase is inducibly expressed in HEK293T cells. (A)
RNA was extracted from HEK293T cells transformed with pcDNA5 plasmids carrying cDNAs to wild-type Twinkle without a tag (Twk), wild-type Twinkle with a haemagglutinin tag (Twk-HA) or a mutant variant of Twinkle D311A (Twk-mut2-HA) and untransformed (control) cells. Twinkle and GAPDH mRNAs levels were determined by Q-RT-PCR and the ratio was expressed relative to control cells (panel A, n = 2; panel B, n = 3) that was arbitrarily set as 1. The transgene was induced for 24 h or 72 h with 0, 3 or 30 ng/ml doxycycline (Dox). (B) Dose and time dependent expression of transgenic Twinkle (wild-type, untagged) via immunodetection. (C) Expression of transgenic Twinkle wild-type and mutant (D311A) after 72 h induction with 10 ng/mL doxycycline (FLAG tagged) via immunodetection. This difference in expression was observed in independent transformant cell lines, suggesting HEK cells are less tolerant of the mutant variant, which is further suggested Twk-mut2 causing (slightly) more depletion of mtDNA (Fig. 7A vs S6A). Further increasing the level of transgenic mutant protein would be expected to exacerbate the replication stalling phenotype and accelerate the mtDNA depletion. Moreover, replication stalling was never seen at earlier time points or with lower doses of doxcycline in the cells with transgenic wild-type Twinkle (data not shown). Therefore it is unlikely the difference in effects on mtDNA replication and mtRNA between mutant and wild-type Twinkle (Figs 2-7) are attributable to the amount of transgenic protein. Hence, the data suggest that a considerable excess of functional Twinkle protein is required to repress strand-asynchronous replication and transcripts, and enhance bidirectional strand-coupled replication in cultured cells.

Supplementary Materials and Methods
Primers and probes used for Southern hybridization and RNA quantitation.

Probe
Primer  PCR products were digested with Dra1 to create a consistent end at np 16,010 prior to 1D-AGE fractionation of the markers. Another marker spanning the BsaW1 site at np 15,924 to 50 was generated by digesting 143B crude mtDNA with BsaW1 and BstX1 to create a product of 695 bases in length.