Figure 1.
POLRMT can initiate primer synthesis from a single dT and mtSSB is abundant in vivo.
(A). POLRMT can initiate primer synthesis from a linear template containing one or more dT (lanes 2 to 7). Deletion of the poly-dT stretch abolishes primer synthesis (lane 1). (B) Representative quantitative Western blot measurement of endogenous mtSSB protein in human Hela cells. Protein extracts (5, 10 or 20 µl) were loaded from a determined number of cells. Purified recombinant mtSSB was used to create standard curve with known protein concentrations.
Table 1.
Quantification of mtSSB to mtDNA ratio in HeLa cells.
Figure 2.
mtSSB governs OriL specificity.
(A) In vitro rolling circle DNA replication reaction with increasing concentrations of mtSSB (0, 10, 100, 500 fmol, 1, 5, 10, 20 and 40 pmol) on the SK+OriL dsDNA template. DNA replication was performed in the presence of [α- 32P]-dCTP in order to label newly synthesized DNA as described previously [6]. The weak labeling of input template in lanes 1 and 2 is most likely due to POLγ idling on the free 3′-end, in the absence of active rolling circle DNA replication. (B) Schematic illustration explaining the replication products formed on lagging-strand DNA. At higher mtSSB levels, primers synthesis is restricted to OriL, but at lower levels primer synthesis can take place also at other sites. During the first round of DNA synthesis, the OriL-depending lagging-strand products have a length of about 2100 nts. In later rounds, the fragments will span the entire distance between two OriL sequences (about 3900 nts) and migrate with the same size as the input template. (C) Reactions were performed as in panel A, but replication products were analyzed by Southern blotting using strand-specific probes to detect leading- or lagging-strand DNA synthesis. For comparison, we used a mutant template (OriL-del) in which the OriL sequence had been deleted.
Figure 3.
The stem-loop structure of OriL prevents mtSSB binding.
MtSSB binding to ssDNA oligonucleotides was monitored by EMSA as described in experimental procedures. Increasing mtSSB concentrations (0, 5,10, 25, 50 100 fmol mtSSB calculated as a tetramer) were incubated for 15 minutes at room temperature with 25 fmol of [γ- 32P]-ATP 5′-end labeled DNA. As templates we used WT OriL, an OriL derivative lacking a stem or an OriL derivative with a 6 bp increase in stem length.
Figure 4.
mtSSB in vivo occupancy reflects strand-displacement mtDNA replication.
(A) Occupancy of mtSSB and TFAM analyzed by strand-specific qPCR amplification of ChIP samples. (B) Strand-specific ChIP-seq profile of mtSSB binding to mtDNA. The origins of replication are indicated. The short black bars indicate the location of the primers used for strand-specific qPCR. (C) Schematic illustration of expected occupancy of mtSSB accordingly to the different mtDNA replication models. SDM (Strand displacement mode), SC (Strand coupled mode), and RITOLS.
Figure 5.
Neutral 2D-AGE analysis of mtDNA replication products.
(A) Extracted DNA analyzed on 1% Agarose. (B) Purified DNA cut with HincII was analysed using 2D-AGE. A fragment (mtDNA 13636-1006 bp) spanning the OriH region was visualized with probe located in CYTB (14641-15590 bp). Upper panel; untreated DNA (containing both RNA and DNA), Middle panel; RNase A and RNase H treated DNA (containing only DNA); Lower panel, DNA treated with RNaseA and RNaseH remixed with the RNA still present after DNase I treatment. (C) Schematic illustration of how Y and bubble arcs are expected to run in 2D-AGE. The bubble arc observed here is dependent on RNA (indicated in red).