Generation of brain region-specific inducible mitoPstI mice. (A) Dox-regulated induction of mitoPstI in tet-off HeLa cells. Cells were transiently transfected with the mitoPstI construct in the absence (upper panels) or presence (lower panels) of Dox. In the absence of Dox, PstI was expressed (upper left panel) and co-localized with the mitochondria stained with Mitotracker Red (middle upper panel). Nuclear staining is non-specific. (B) Schematic diagram of tet-off system. In the double transgenic mice, tTA expressed under CamKIIα promoter binds to TRE in the absence of Dox and induces the expression of mitoPstI. The application of Dox prevents the binding of tTA to TRE and silences the expression of mitoPstI. Mouse mtDNA carries two PstI sites: One is at nt 8420 and the other is at nt 12 238. (C) Levels of mitoPstI in the cerebral cortex of mitoPstI/tTA mice obtained from four different lines of mitoPstI mice. Cortical tissues from 10-week-old mice were subjected to western blotting for PstI. Purified PstI available from NEB was used as a positive control.

Detection and characterization of ΔmtDNA. (A) To avoid mtDNA depletion, mitoPstI was induced for only 3 days at P28. Animals were studied after 1 or 10 weeks. (B) Detection of mouse mtDNA in the cerebral cortex of mitoPstI/tTA mice after a transient (3 days) induction and 10-week suppression of mitoPstI. Total DNA isolated from induced 14-week-old mice (n = 2) as well as age-matched non-induced controls (n = 3) were subjected to Southern blotting for mtDNA and the nuclear 18S rDNA. In the upper part of the panel, a DNA probe complementary to the 3000–4200 region of mtDNA (4000 Probe) was used, whereas the 9000 Probe (complementary to the 8931–10 660 region) was used in the lower part. (C) PCR to amplify ΔmtDNA. Using 8300-forward and 12 812-backward primers, full-length fragment (∼4.5 kb) of wild-type mtDNA was amplified from cortical DNA from non-induced mice. In induced samples, smaller (∼670 bp) fragments were preferentially amplified. (D) Cloning and sequencing analysis of the 670 bp bands. Three types of ΔmtDNA (Deletions 1–3) with ∼3.8 kb deletions were identified. Sequences around the breakpoints are shown. Deletion 2 had direct repeats (underlined bold letters) shared by both sides of the breakpoint (one direct repeat is located inside and the other is outside of the deleted region). The predicted mechanism of formation of Deletion 3 through NHEJ is shown at the bottom. (E and F) Detection of ΔmtDNA in the cortical tissues of induced 14-week-old mice. Primer sets used in these PCR reactions are shown at the bottom.

Quantification of ΔmtDNA generated by a transient induction of mitoPstI. (A) Semi-quantitative PCR of ΔmtDNA (primers indicated in the figure) in the cerebral cortex of 5- and 14-week-old mice that underwent a transient induction of mitoPstI at P28. Age-matched non-induced cortical samples (Mice A–D) were used as negative controls. (B) Quantification of smaller deletions (∼3.8 kb deletions). Using a primer set (8378-forward and 12 281-backward primers) that amplified ΔmtDNA with ∼3.8 kb deletions (amplicon size 80–90 bp), real-time qPCR was performed for cortical DNA samples used in (A). Signals obtained from this PCR were normalized by the total levels of mtDNA (ND1-encoding sequence located in non-deleted regions of mtDNA), which were also quantified by qPCR. The fold difference in the relative levels of ∼3.8 kb deletions between the average of 5-week-old mice (Mice 1 and 2) and each 14-week-old mice (Mice 3 and 4) is shown. (C) Quantification of larger deletions (∼10 kb deletions). A set of 5549-forward and 15 501-backward primers was used to quantify ΔmtDNA with ∼10 kb deletions (amplicon size ∼100 bp).

Functional assessment of mitoPstI. (A) When constitutively expressed in mice, mitoPstI/CamKIIa-tTA caused ‘clasping’, a well recognized neurological phenotype. (B) To avoid developmental problems, mitoPstI was induced only at P21. With this regimen, lifespan was significantly shortened. (C) Analysis of mitoPstI expression in different brain regions of 10-week-old mice. Nuclear-coded mitochondrial proteins (SDH and VDAC1) were not altered, but the mtDNA-coded CoxI was reduced in mitoPstI expressing mice. (D) DNA from cerebral cortex, hippocampus, striatum and cerebellum was isolated after perfusion and subjected to Southern blotting for mtDNA. The same membrane was later blotted for nuclear 18S rDNA, and the ratio of mtDNA/nDNA was determined in a phosphoimager. Ctrl: mitoPstI+ single transgenic mouse. Tg: mitoPstI/tTA double transgenic mouse.

Characterization of ΔmtDNA breakpoints. Amplification of mtDNA deletion breakpoints were sequenced and analyzed. The PstI sites (when close to the breakpoint) are indicated. Ends of deleted regions are shown for the analysis of homologies (underlined).