(a) Representative voltage clamp recordings showing increased sIPSC rates in Ts65Dn CA1 pyramidal neurons (bottom) compared to those from both euploid (top) and TsOlig1/2^+/+/− (middle). (b) The cumulative probability distribution function for the intervals between sIPSC is shifted rightward for Ts65Dn CA1 pyramidal cells. In contrast, the curves for TsOlig1/2^+/+/− and euploid mice are similar. (c) In the presence of TTX, IPSC rates are similar in CA1 pyramidal neurons from all mice. (d) The cumulative probability distribution functions for each genotype in the presence of TTX are similar.

(a, b) Confocal images of parvalbumin expressing cells in the Ts65Dn and euploid hippocampus at P15. (e, f) Higher magnification of parvalbumin⁺ cells in the CA1 region of the Ts65Dn and euploid hippocampus. (c, d) Confocal images of somatostatin expressing cells in the Ts65Dn and euploid hippocampus at P15. (g) Analysis of cell counting uncovered an overabundance of parvalbumin and somatostatin-expressing inhibitory interneurons in the Ts65Dn brain. This is a phenocopy of the results in the Ts65Dn neocortical wall (Figure 1) and implicates changes in production of these specific interneuron subtypes. Data points represent mean ± s.d. (n=4 mice for each age and group). *p<0.05 by unpaired two-tailed t-test. Scale bars, 200 μm.

(a–d) Confocal images of somatostatin expressing cells in the neocortex of Ts65Dn, euploid, euploidOlig1/2^+/− and TsOlig1/2^+/+/− littermates at P15. (e–h) Confocal images of parvalbumin expressing cells in the neocortex of the same animals. Immunoreactive cells were counted within a fixed area (500μmx150μm) for all the different layers of the cortex. The layer boundaries were determined using molecular layer markers. (i–j) Comparing the number of somatostatin and parvalbumin positive cells demonstrates that the increased interneuron phenotype in Ts65Dn neocortex is prevented in TsOlig1/2^+/+/− animals. Moreover, significantly reduced numbers of somatostatin and parvalbumin expressing cells were observed in the cortical layer II–III of euploidOlig1/2^+/− heterozygotes. Data points represent mean ± s.e.m. (n=4–6 mice for each age and group). *p<0.05; **p<0.01 by two-tailed t-test. Scale bar: 100 μm.

(a–f) Postmitotic interneuron markers Lhx6 and somatostatin were analyzed in brain sections of E15.5 euploid, Ts65Dn and TsOlig1/2^+/+/− embryos by in situ hybridization. (a–c) The Lhx6 expression pattern in TsOlig1/2^+/+/− MGE is similar to euploid. (d–f) Somatostatin expression in TsOlig1/2^+/+/− telencephalon is comparable to euploid. The boxed areas g and h represent the cell counting areas for dorsal and ventral telencephalon respectively. (g, h) Comparing the number of somatostatin-expressing cells demonstrates an increase in Ts65Dn telencephalon while the TsOlig1/2^+/+/− rescued animals exhibited a normal number of somatostatin-expressing cells in dorsal telencephalon and no significant difference in the ventral telencephalon. Data points represent mean ± s.d. (n=3 mice for each age and group). *p<0.05; **p<0.005 by unpaired two-tailed t-test. Scale bar: 200 μm.

(a) Brain sections of P15 Ts65Dn and euploid littermates were immunostained for different interneuron markers (viz. Calretinin, Calbindin, Parvalbumin and Somatostatin). Cell counts were performed within a fixed area (500μmx150μm) for all the different layers of the cortex. The layer boundaries were determined using molecular layer markers. (b) Comparing the number of calretinin, calbindin, somatostatin and parvalbumin positive cells at this postnatal developmental time point demonstrated a significant increase in calbindin, somatostatin and parvalbumin expressing cells in Ts65Dn cortex with no change in calretinin+ cells. Data points represent mean ± s.d. (n=6 mice for each age and group). *p<0.02 by unpaired two-tailed t-test. Scale bars, 100 μm. See Supplementary Fig. 2 online for analyses at other ages.

(a, b) Olig2 expression was increased in the MGE of Ts65Dn brain at E13.5 and E14.5. Higher magnification shows an increased number of Olig2 expressing cells in the MGE of Ts65Dn mice at E13.5 (c) and E14.5 (d). Scale bar, 200 μm. (e) Semi quantitative RT-PCR indicates higher expression of Olig1 and Olig2 genes in the MGE of Ts65Dn at E14.5. (f) Quantification of Olig2 expressing cells demonstrates an increase in the number of Olig2⁺ cells in the VZ and SVZ of Ts65Dn MGE compared to euploid littermates at E13.5 and E14.5. (g) Substitution of C57BL/6J Olig1/2^+/− females for the C57BL/6JEi females yields B6C3^Olig1/2+/− males for crossing with Ts65Dn females. Thus, the resulting Ts65DnOlig1/2^+/+/− (TsOlig1/2^+/+/−), Ts65DnOlig1/2^+/+/+ (Ts65Dn) and euploid pups have identical background strain potential compared to the normal Ts65Dn production techniques. We compared TsOlig1/2^+/+/, Ts65Dn and euploidOlig1/2^+/+ littermate pups to remove any strain and construct-dependent effects on neuronal population and differentiation. (h) The increase in MGE neurogenesis found in Ts65Dn is prevented in TsOlig1/2^+/+/− animals at E14.5. Moreover, reduced neurogenesis was found in the MGE of euploid heterozygotes (euploidOlig1/2^+/−), suggesting that normal ventral telencephalic neurogenesis is dependent on Olig1 and Olig2 gene copy number and expression. (i, j) To determine the q-fraction 24 h after BrdU injection, sections from euploid, Ts65Dn and TsOlig1/2^+/+/− brains were stained with antibodies to Ki67 (red) and BrdU (green). Scale bar, 50 μm. The fraction of cells labeled only with BrdU (BrdU⁺/Ki67⁻; no longer dividing) after 24 h period (q-fraction) is higher in Ts65Dn MGE compared to euploid. However, the q-fraction increase found in Ts65Dn MGE is corrected in TsOlig1/2^+/+/− animals. Data points represent mean ± s.d. (n=3–4 mice for each age and group). *p<0.05; **p<0.005 by two-tailed t-test.

(a, b) Lhx6 expression was increased in the MGE of Ts65Dn brain at E13.5 and E14.5. Scale bar, 200 μm. See Supplementary figure 5 for additional markers. (c, d) E13.5 timed pregnant Ts65Dn females were injected (intraperitoneally) with BrdU to label a cohort of neural progenitors in S-phase and embryos were harvested 24 h later. (c) A representative section stained with antibodies to BrdU (green) and TUJ1 (red). Scale bar, 200 μm. (d) The fraction of cells labeled with both BrdU and TUJ1 (BrdU⁺/TUJ1⁺; newly generated neurons) 24 h after pulse label is 36% higher in the MGE of Ts65Dn. No significant change in LGE or CGE neurogenesis was observed in Ts65Dn brains, suggesting that increased neurogenesis was specific to the Ts65Dn MGE during this period of time. (e, f) The cumulative labeling plots for MGE progenitors at E13.5 and E14.5 demonstrate that the labeling index (percentage of BrdU⁺ cells) rises precipitously until reaching saturation at the GF value, which represents the proportion of the precursor cells that are participating in the cell cycle. The time to reach saturation defines the T_c-T_s value, which was normal in Ts65Dn (see also Supplementary Table 2). (g, h) Distribution of PH3-labeled M-phase cells in the MGE at E14.5 and E15.5. Starting from the ventricular border, the MGE was divided into 15 bins each 20 μm in thickness. Scale bar, 100 μm. The number of mitotic cells in the SVZ of Ts65Dn MGE was increased compared to euploid indicating an increase in abventricular mitoses in Ts65Dn. Data points represent mean ± s.d. (n=3–6 mice for each age and group). *p<0.05; by unpaired two-tailed t-test.