Deficits in human trisomy 21 iPSCs and neurons

Abstract

Down syndrome (trisomy 21) is the most common genetic cause of intellectual disability, but the precise molecular mechanisms underlying impaired cognition remain unclear. Elucidation of these mechanisms has been hindered by the lack of a model system that contains full trisomy of chromosome 21 (Ts21) in a human genome that enables normal gene regulation. To overcome this limitation, we created Ts21-induced pluripotent stem cells (iPSCs) from two sets of Ts21 human fibroblasts. One of the fibroblast lines had low level mosaicism for Ts21 and yielded Ts21 iPSCs and an isogenic control that is disomic for human chromosome 21 (HSA21). Differentiation of all Ts21 iPSCs yielded similar numbers of neurons expressing markers characteristic of dorsal forebrain neurons that were functionally similar to controls. Expression profiling of Ts21 iPSCs and their neuronal derivatives revealed changes in HSA21 genes consistent with the presence of 50% more genetic material as well as changes in non-HSA21 genes that suggested compensatory responses to oxidative stress. Ts21 neurons displayed reduced synaptic activity, affecting excitatory and inhibitory synapses equally. Thus, Ts21 iPSCs and neurons display unique developmental defects that are consistent with cognitive deficits in individuals with Down syndrome and may enable discovery of the underlying causes of and treatments for this disorder.

Keywords: cerebral cortex; developmental disorders.

Fig. 1.

Reprogramming of mosaic fibroblasts yields Ts21 iPSCs and an isogenic control. (A) Fibroblast line AG05397 was mosaic for Ts21, with (Upper) ∼89% of cells carrying three copies [runt-related transcription factor 1 (AML1)/Down syndrome critical region (DSCR); orange], whereas (Lower) 10% were disomic for HSA21 (AML1/DSCR; orange). The telomere (TEL) marker (green) was used as a control probe. (B) All iPSC lines had morphological characteristics of pluripotent stem cells, and karyotype analysis showed that DS1 and DS4 iPSCs are trisomic, whereas DS2U is disomic for HSA21 (red circles). (C) 2DS3 iPSCs from a second DS individual carry Ts21. (D) SNP analysis revealed no absence of heterozygosity of HSA21 in the euploid DS2U iPSC line. (E) Short terminal repeat analysis revealed that Ts21 and control lines are isogenic at all loci tested. (F) Table of different iPSC lines used in this study. (G) A heat map shows that genes changed more than fivefold in DS1 and DS4 iPSCs compared with the isogenic DS2U control iPSCs. (H) qPCR verification in all Ts21 iPSC lines of various genes that are changed in microarray results. (I) Ts21 iPSCs did not exhibit increased oxidative stress compared with their respective controls, which were assayed by DHE. (J) The proportion of Ts21 cells that underwent apoptosis was similar to controls, which were assayed by TUNEL⁺ cells. Error bars represent SEM. (Scale bars: 100 µm.)

Fig. 2.

Generation of cortical neurons is not affected by Ts21. (A) FACS analysis reveals no difference in the propensity of Ts21 iPSCs to generate Pax6⁺ neuroepithelia. (B) Ts21 iPSC-derived neural progenitor cells exhibited increases in many HSA21 and oxidative stress genes, which were assayed by qPCR. (C) Transcript expression of dorsal telencephalic markers was evident in neurons differentiated from all lines, but ventral forebrain [NK2 homeobox 1 (Nkx2.1)], hindbrain [gastrulation brain homeobox 2 (Gbx2)], and spinal cord [homeobox B4 (Hoxb4)] markers were not readily observed. (D and E) Immunostaining of cultured cells shows no difference in β_III-tubulin⁺ neurons or forebrain markers [forkhead box G1 (FoxG1) and orthodenticle homeobox 2 (OTX2)] across iPSC lines. (F) Heat map depicts global gene expression changes more than threefold in isogenic Ts21 (DS1 and DS4) vs. control (DS2U) iPSC-derived neuronal cultures. (G) qPCR verification of various genes up-regulated in microarray results in all iPSC-derived neuronal cultures. (H) DS neurons exhibited increased oxidative stress, which was assayed by DHE, and (I) mitochondrial membrane potential. (J) The proportion of Ts21 cells that underwent apoptosis was similar to controls, which were assayed by TUNEL⁺ cells. (Scale bars: 50 µm.) Error bars represent SEM. *P < 0.05. (A, B, D, E, and J) For measures where no significant differences were found between groups, dashed lines indicate the average of the control groups (DS2U and IMR90).

Fig. 3.

Forebrain Ts21 neurons display synaptic deficits across transmitter phenotype. (A and B) Representative whole-cell patch clamp traces illustrate that Ts21 (DS) did not show differences in (A) sodium (Na⁺) and potassium (K⁺) currents or (B) the number of APs in response to current injection. (C) Representative voltage clamp (−70 mV) traces show that both control and DS neurons displayed sPSCs. Expanded timescale (traces 3 and 4) illustrates sPSCs with different kinetics in both control and DS neurons. (D) Significantly fewer DS neurons display synaptic activity relative to controls (DS2U: 86 ± 3.9%; IMR90: 81 ± 3.2%). (E) DS neurons also displayed significantly lower frequencies of sPSCs compared with controls (DS2U: 0.48 Hz; IMR90: 0.92 Hz). (F) Representative images of control and DS β_III-tubulin⁺ neurites (red) displaying synapsin⁺ puncta (green; arrows). (Blue) Hoechst. Pooled data revealed fewer synapsin⁺ puncta in DS neurons compared with controls (DS2U: 2.9 ± 0.6/100 µm; IMR90: 4.4 ± 0.6/100 µm). (G) The proportion of excitatory and inhibitory sPSCs was not changed in DS cultures relative to controls [excitatory PSCs (ePSCs): DS2U: 0.54 ± 0.07 Hz; IMR90: 0.55 ± 0.11 Hz; iPSCs: DS2U: 0.46 ± 0.06 Hz; IMR90: 0.45 ± 0.09 Hz]. No differences were observed in (H) the fraction of GABA⁺ neurons compared with controls (DS2U: 47.7 ± 1.9%; IMR90: 46.2 ± 2.2%) or (I) the fraction of VGAT⁺/synapsin⁺ puncta (arrowheads) in DS cultures compared with controls (DS2U: 37.1 ± 6.1%; IMR90: 41.5 ± 4.5%). Error bars represent SEM. *P < 0.05. (Scale bars: F and I, 10 µm; H, 50 µm.)