Figure 1: Generation of the astrocyte-specific Ndufs4(KO) mouse model.

A. Schematic showing the crossing strategy used. The heterozygous Ndufs4^Δ/+ mice do not display hypersensitivity to volatile anesthetics or the Leigh syndrome pathogenesis displayed by the global KO, indicating that Ndufs4 is haplo-sufficient. The Pgfap-CreERT2 allele has a dominant phenotype; the breeding recommendation from the JAX lab for these mice is to maintain them as heterozygotes. Pgfap-CreERT2/+ mice were crossed to Ndufs4^Δ/+ mice. From the resulting F1, Pgfap-CreERT2/+;Ndufs4^Δ/+ offspring were selected and crossed to Ndufs4^lox/lox mice to generate F2 Pgfap-CreERT2/+;Ndufs4 ^Δ/lox (conditional astrocyte-specific Ndufs4(KO)) and Pgfap-CreERT2/+;Ndufs4^+/lox mice (controls). We used Ndufs4^lox/Δ instead of Ndufs4^lox/lox to avoid possible recombination due to the leakiness of the promoter (as seen with the Pvglut2 driven Cre mice). Both genotypes were injected with 4-hydroxytamoxifen intraperitoneally at a concentration of 50 μg/g bodyweight daily from post-natal day 33 to 39. Red stars indicate the loxP sites. B. i: Schematic of generation of mice to confirm loss of full length Ndufs4 upon 4-hydroxy tamoxifen induced activation of Cre recombinase in astrocytes. Pgfap-CreERT2/+ mice were crossed to Ndufs4^lox/lox mice; from the progeny, Pgfap-CreERT2/+;Ndufs4^lox/+ animals were selected by genotyping and crossed again with Ndufs4^lox/lox to generate Pgfap-CreERT2/+;Ndufs4^lox/lox mice, which were injected with 4-hydroxy tamoxifen or vehicle control (sham). The brains of the injected mice were harvested after 3 and 10 days post injection (dpi) and genotyped. B. ii: Genotyping gel of brains of Pgfap-CreERT2/+;Ndufs4^lox/lox after 4-hydroxy tamoxifen or sham injections, using PCR (lanes 2–5, dpi – days post injection of 4-hydroxy tamoxifen or sham oil control). Sham injections yield only a large uncut band at 1300bp, while 4-hydroxy tamoxifen injections show both the large band (uncut in neurons) as well as a smaller band (excised in astrocytes) at 270bp. Sequencing of excised bands outlined in white rectangles (lane 5), confirmed a full length allele as well as a truncated allele lacking the second exon, as predicted by size, in the 4-hydroxy tamoxifen injected mice. Genomic DNA isolated from the tails of Ndufs4^lox/lox mice and total Ndufs4(KO) (TKO) were used as controls (lanes 6 and 7 respectively). Lane 1 shows the DNA ladder size markers (base pairs).