Figure 4
Comparison of GABA A R subunit gene expression in wild-type and Ts65Dn GCs by single-cell real-time PCR. (A) Harvesting of a GC into a glass pipette. Calibration bar (5 μm) applies to all panels. 1: pipette tip (containing RNA-protecting solution) touching a chosen cell within a cerebellar slice; 2: suction draws the cell into the pipette. The pipette is raised away from the slice (out of focus) so as to minimise entry of contaminating material; 3: complete cell in the tip, prior to ejection. (B) Percentages of wild-type (empty bars) and Ts65Dn (filled bars) GCs in which cDNAs encoding different GABAAR subunits (α6, α1, β2, δ, β3 and γ2) were detected. Numbers of cells are shown on the bars. The various cDNAs were detected with dissimilar frequencies in both wild-type (χ2(5, n = 226) = 52.73, p < 0.0001) and Ts65Dn (χ2(5, n = 174) = 65.83, p < 0.0001) GCs. The frequency with which each subunit cDNA was detected did not differ between wild-type and Ts65Dn (Fisher’s exact test, p > 0.2020 for all comparisons). (C) Box plots of numbers of cDNA copies calculated for cDNA-positive GCs (log10 scale). The numbers were not uniform in wild-type (p < 0.0001 Kruskal Wallis test) or Ts65Dn (p < 0.0001 Kruskal Wallis test) GCs but followed the approximate order: α6 ≥ α1 ≥ β2 > δ ≈ β3 ≈ γ2 in both (Dunn’s multiple comparison test p < 0.001 or < 0.01). Copy numbers for α6, α1, β2, and δ cDNA were not different between wild-type and Ts65Dn (p > 0.2800 for all comparisons, Mann Whitney U test), the number of β3 cDNAs was reduced in Ts65Dn (*p = 0.0125, Mann Whitney U test), and the number of Ts65Dn GCs in which γ2 cDNAs were detected was too small for statistical comparison.