Neuronal maturation defect in induced pluripotent stem cells from patients with Rett syndrome

Abstract

Rett syndrome (RTT) is one of the most prevalent female neurodevelopmental disorders that cause severe mental retardation. Mutations in methyl CpG binding protein 2 (MeCP2) are mainly responsible for RTT. Patients with classical RTT exhibit normal development until age 6-18 mo, at which point they become symptomatic and display loss of language and motor skills, purposeful hand movements, and normal head growth. Murine genetic models and postmortem human brains have been used to study the disease and enable the molecular dissection of RTT. In this work, we applied a recently developed reprogramming approach to generate a novel in vitro human RTT model. Induced pluripotent stem cells (iPSCs) were derived from RTT fibroblasts by overexpressing the reprogramming factors OCT4, SOX2, KLF4, and MYC. Intriguingly, whereas some iPSCs maintained X chromosome inactivation, in others the X chromosome was reactivated. Thus, iPSCs were isolated that retained a single active X chromosome expressing either mutant or WT MeCP2, as well as iPSCs with reactivated X chromosomes expressing both mutant and WT MeCP2. When these cells underwent neuronal differentiation, the mutant monoallelic or biallelelic RTT-iPSCs displayed a defect in neuronal maturation consistent with RTT phenotypes. Our in vitro model of RTT is an important tool allowing the further investigation of the pathophysiology of RTT and the development of the curative therapeutics.

Fig. 1.

Characterization of RTT- iPSCs derived from RTT patients. (A) Structure of MeCP2 and location (amino acid residues) of mutations. (B) Representative immunofluorescence images of RTT-iPSC-w and RTT-iPSC-m showing expression of pluripotency markers of SSEA3, SSEA4, Tra-1-81, Tra-1-60, NANOG, and OCT4. (C) Relative expression of OCT4, SOX2, KLF4, MYC, and NANOG in monoallelic WT and mutant RTT-iPSCs compared with parental fibroblasts and H1 hESCs by real-time qPCR. qPCR was performed in triplicate, and the relative gene expression was calculated by normalizing against the parental RTT fibroblast cells. Mean ± SEM are shown with error bar. (D) Images for normal karyotype of iPSCs. (E) Representative images of teratomas generated in immunodeficient mice injected with RTT-iPSCs. RTT-iPSCs composed a well-defined cystic teratoma with tissues of all three germ layers including endoderm (respiratory epithelium), mesoderm (smooth muscle and bone), and ectoderm (neuronal rosettes).

Fig. 2.

XCI status of iPSCs. (A and B) Sequences of MeCP2 expressed in RTT-iPSCs. RTT1-iPS-13w, RTT4-iPS-24w, and RTT1-iPS-15m express only WT or mutant MeCP2, whereas RTT4-iPS-19bi expresses both mutant and WT MeCP2. (C and D) Relative expression of MeCP2 (C) and EZH2 (D) in different iPSCs and human ESCs. iPSCs showing the X chromosome reactivation display approximately twice the MeCP2 expression as iPSCs retaining the inactive X chromosome (C). The difference in MeCP2 expression was not observed between WT and mutant monoallelic RTT-iPSCs. EZH2, a polycomb group protein essential for XCI, shows an expression pattern inversely correlated with MeCP2 expression (D). Interestingly, mutant monoallelic RTT-iPSCs has significant less expression of EZH2 than WT monoallelic RTT-iPSCs. Histograms show the average levels of MeCP2 and EZH2 distributions quantified by qPCR and normalized with β-actin. Data are mean ± SEM. *P < 0.01; **P < 0.001. (E and F) Nuclear pattern of H3K27me3 in RTT fibroblasts and iPSCs. RTT1 fibroblast cells and RTT1-derived iPSC lines monoallelically expressing MeCP2 (RTT1-13w and RTT1-15m) exhibit predominant H3K27me3 nuclear foci (E). RTT4 fibroblasts and RTT4-iPS24w cells monoallelically expressing MeCP2 show condensed H3K27me3 staining, whereas RTT4-iPS19bi cells display diffuse staining (F). Nuclei were detected with DAPI, and pluripotency marker (OCT4) was stained. The arrow points to H3K27me3 Xi enrichment in the fibroblast cells. (Insets) Higher-resolution images.

Fig. 3.

Defect in neuronal maturation of mutant RTT-iPSCs. (A) Schematic depicting the neural differentiation procedure. See Materials and Methods for a detailed description. Representative images showing morphological changes during neural differentiation of hESCs (H1), and RTT-iPSCs (RTT4-iPS-24w, RTT4-iPS-19bi). (B and C) H1 hESCs and RTT-iPSCs were induced for neuronal differentiation and stained with neuronal stem cell marker (NESTIN), mature neuronal marker (TuJ), and astrocyte marker (GFAP) during the differentiation. H1 hESCs and WT and mutant RTT-iPSCs all show a similar neuronal commitment, represented by NESTIN staining after 11 d of neural differentiation (B). H1 hESCs and WT RTT1-iPS-13w cells are fully differentiated into neuronal differentiation for 25 d and display a robust expression of the mature neuronal marker TuJ, whereas mutant RTT1-iPS-15m cells demonstrate less neural maturation. In contrast, glial differentiation (GFAP⁺ cells) is similar in all cell lines (C). **P < 0.01. Bar graphs show the percentages of TuJ⁺ and GFAP⁺ cells. The densities of TuJ⁺ and GFAP⁺ cells vs. DAPI per square mm were analyzed using Image J analysis software. Data are mean ± SEM of one experiment performed in triplicate.

Fig. 4.

Expression of neuronal markers during differentiation of hESCs and iPSCs. (A–F) RNA was isolated from each time point during neuronal differentiation, and real-time qPCR was performed with primers to detect SOX2, PAX6, GFAP, TuJ, SCN1A, and SCN1B. hES cells (H1), normal iPSCs (551-iPS-K1), WT RTT1-iPS-13w cells, and mutant RTT1-iPS-15m iPSCs were used. Gene expression was normalized against β-actin and calculated against H1 at day 0. Data are mean ± SEM in one experiment performed in triplicate.