Generation and Characterization of a Human Neuronal In Vitro Model for Rett Syndrome Using a Direct Reprogramming Method

Abstract

Rett Syndrome (RTT) is a severe neurodevelopmental disorder, afflicting 1 in 10,000 female births. It is caused by mutations in the X-linked methyl-CpG-binding protein gene (MECP2), which encodes for the global transcriptional regulator methyl CpG binding protein 2 (MeCP2). As human brain samples of RTT patients are scarce and cannot be used for downstream studies, there is a pressing need for in vitro modeling of pathological neuronal changes. In this study, we use a direct reprogramming method for the generation of neuronal cells from MeCP2-deficient and wild-type human dermal fibroblasts using two episomal plasmids encoding the transcription factors SOX2 and PAX6. We demonstrated that the obtained neurons exhibit a typical neuronal morphology and express the appropriate marker proteins. RNA-sequencing confirmed neuronal identity of the obtained MeCP2-deficient and wild-type neurons. Furthermore, these MeCP2-deficient neurons reflect the pathophysiology of RTT in vitro, with diminished dendritic arborization and hyperacetylation of histone H3 and H4. Treatment with MeCP2, tethered to the cell penetrating peptide TAT, ameliorated hyperacetylation of H4K16 in MeCP2-deficient neurons, which strengthens the RTT relevance of this cell model. We generated a neuronal model based on direct reprogramming derived from patient fibroblasts, providing a powerful tool to study disease mechanisms and investigating novel treatment options for RTT.

Introduction

Rett Syndrome (RTT) is a postnatal neurodevelopmental disorder, caused by pathogenic variants in the X-linked human methyl-CpG-binding protein gene (MECP2) encoding the methyl CpG binding protein 2 (MeCP2), a global transcriptional regulator [1]. The cumulative incidence of RTT is 1 in 10,000 female births [2]. The disease is characterized by an apparently normal development in the first 6–18 months of life, followed by a period of regression with the loss of expressive language and purposeful hand use, as well as other debilitating symptoms [3,4].

The nature of RTT neuropathology was elucidated in human brain studies. Postmortem RTT brains weigh less and exhibit increased packing density of neurons, diminished dendritic arborization, and hence, a decreased number of synapses [5,6]. Moreover, in induced pluripotent stem cells (iPSCs)-derived neurons from Mecp2-deficient mice, as well as in human MECP2-null iPSC-derived neurons, alterations in electrophysiological activity were observed [7,8].

Currently, several mouse models carrying various pathogenic variants of the Mecp2 gene exist, which mimic the neurological phenotype of RTT [9]. Several studies have demonstrated a rescue of some RTT phenotypes through the restoration of biologically active MeCP2 levels in Mecp2-deficient mouse models [10 –13]. Nevertheless, these mouse models do not accurately recapitulate disease progression in humans.

One alternative and promising tool to study RTT is through the reprogramming and differentiation of patient-derived cells through iPSCs into neurons, as primary human neurons from patient and control subjects are not readily available. Usually, iPSCs are generated from human or murine fibroblasts, as well as blood cells, by virally delivering certain factors as previously described [14]. These iPSCs can then be differentiated into various cell types, including neurons, under specific growth conditions [14]. It has been demonstrated that neuronal cells derived from RTT-iPSCs exhibit a decrease in the number of synapses, spine density, and soma size, as well as altered electrophysiological properties, which are comparable to the phenotype observed in vivo [8,15]. The process of iPSC reprogramming, however, involves the introduction of epigenetic modifications, due to the fact that cells are reset to a pluripotent state [16]. Adult donor cells are hence rejuvenated, which is not optimal to study age-relevant diseases. Furthermore, reprogramming of fibroblasts into iPSCs is a prolonged process taking several months [17].

In contrast, direct reprogramming of fibroblasts into induced neuronal progenitors (iNPs) and further on into neurons through nonviral expression of reprogramming factors bypasses the pluripotent stage, thereby avoiding undesired epigenetic modifications, and reduces the risk of insertional mutagenesis. Furthermore, direct-to-iNP reprogramming prevents cell rejuvenation and maintains the aging memory of the donor human dermal fibroblasts (HDFs) [18 –21]. A method of somatic cell reprogramming using two transcription factors, SOX2 and PAX6, which play a major role in early neurogenesis and neuronal development [22 –25], to directly generate iNPs from adult human fibroblasts has been previously described [19,26]. This nonviral reprogramming method generates patient-derived iNPs, which can be further differentiated into neuronal cells using a two-step differentiation protocol.

This study presents a transient nonviral transfection-based strategy for the generation of patient-derived neuronal cells to investigate RTT. Direct conversion of fibroblasts [19,26] from a male individual with neonatal encephalopathy carrying a frameshift deletion in the MECP2 gene and a healthy control subject provides a novel opportunity to study phenotypic differences in neurons. Moreover, this in vitro model could be used to evaluate novel therapeutic options.

Materials and Methods

Mutation sequencing

For the reprogramming process, HDFs from a male individual with neonatal encephalopathy carrying a frameshift deletion (c.806delG) in the MECP2 gene (2 years old at sampling, “MeCP2-deficient,” “patient-derived”) resulting in a truncated MeCP2 protein [27] and an unaffected sex-matched HDF cell line (1 day old at sampling, AG21708, obtained from the NIGMS Human Genetic Cell Repository at the Coriell Institute for Medical Research) were used. MeCP2-deficient fibroblasts were established for research purposes, approved by the Human Research Ethics Committee of the Children's Hospital at Westmead, Australia [27].

DNA was isolated from HDFs using the DNeasy Blood & Tissue Kit (No. 69504; Qiagen) according to the manual. Polymerase chain reaction (PCR) was performed using DreamTaq Green PCR Master Mix (No. K1081; Thermo Scientific), applying the forward primer 5′-GCAGCCAGGCAGTGTGACTCTC-3′ and the reverse primer 5′-GATGGGGAGTACGGTCTCCTGCAC-3′. PCR products were cleaned up with ExoSAP-IT™ PCR Product Cleanup Reagent (Applied Biosystems), and Sanger sequencing was performed using BigDye™ Terminator v3.1 Kit chemistry (Applied Biosystems). Sequencing products were run on an ABI 3500 Genetic Analyzer (Applied Biosystems), and the results were analyzed using SnapGene 6.2.1 software (www.snapgene.com).

Reprogramming of HDFs

HDFs were grown in fibroblast proliferation medium consisting of DMEM (No. 41966-029; Thermo Scientific), 15% fetal bovine serum (FBS, No. F9665; Sigma-Aldrich), and 1% Penicillin/Streptomycin (No. 15140122; Thermo Scientific) and passaged 1:3 every 3–4 days when the cultures reached about 90% confluence. HDFs were regularly determined to be mycoplasma negative by quantitative real-time PCR (qPCR; ISO17025 certified, Eurofins Genomics) using 72 h conditioned media samples collected from cultures. Episomal DNA transduction was performed on cells below passage 17. The plasmid constructs EEV600-CAG-SOX2 and EEV600-CAG-PAX6 containing SOX2 or PAX6 complementary DNA (cDNA) were purchased from System Custom Constructs. Plasmid DNA was amplified in One Shot TOP10 chemically competent Escherichia coli cells (No. C404010; Thermo Scientific) and then isolated using the EndoFree Plasmid Maxi Kit (No. 12362; Qiagen).

HDFs were cultured until the desired number of cells had been reached and split 1 day before transfection. For electroporation, cells were trypsinized, collected, counted, and washed with Dulbecco's phosphate-buffered solution (DPBS, No. D8537; Sigma-Aldrich). Pelleted cells were resuspended in NEON resuspension buffer at 1 × 10⁷ cells/mL. Per transfection assay, 10⁶ cells and 15 μg of each plasmid DNA were used. Electroporation was carried out using 100 μL pipette tips of the Neon Transfection System (No. MPK10096; Invitrogen) according to the manufacturer's instructions. Electroporation conditions used were one pulse with 1,000 V for 60 ms. Electroporated cells were seeded in polystyrene six-well plates (No. 140675; Thermo Scientific) containing 2 mL fibroblast proliferation medium and allowed to recover for 3 days.

The medium was changed 3 days post-transfection to neural reprogramming medium comprising Neurobasal-A medium (NBA, No. 10888-022; Thermo Scientific), 0.3% d-glucose (No. G8270; Sigma-Aldrich), 1% penicillin/streptomycin/glutamine (No. G6784; Sigma), 2% B27 supplement, serum-free (No. 17504-001; Thermo Scientific), 20 ng/mL human EGF (No. AF-100-15; PeproTech), 20 ng/mL human FGF-2 (No. 100-18B; PeproTech), 2 μg/mL heparin (No. 07980; Stem Cell Technologies), and 1 mM valproic acid (VPA, No. P4543; Sigma-Aldrich). All stock concentrations and solvents of the supplements are listed in Supplementary Table S1. BET inhibitor GSK1210151A (No. SML0666; Sigma-Aldrich) was added at a final concentration of 1 μM to neural reprogramming medium for 7 days, while 25 ng/μL Midkine (No. 450-16; PeproTech) was substituted for neural reprogramming medium until day 17. On day 17 after transfection, cells were replated once a week, whereby cells were collected by trypsinization and seeded into six-well plates at 3 × 10⁵ cells per well. The medium was changed thrice weekly.

Neuronal differentiation of iNP cells

For neuronal differentiation, the cells were mechanically dissociated and seeded onto poly-l-ornithine/laminin coated (Nos. P4957, L4544; Sigma) surfaces at 2.5 × 10⁴ cells per cm² in neural plating medium, consisting of NBA supplemented with 0.3% d-glucose, 1% penicillin/streptomycin/glutamine, 2% B27 supplement, 25 ng/mL FGF-2, 10 μM retinoic acid (No. R2625; Sigma-Aldrich), and 1% FBS. On the following day, the medium was changed to neural differentiation medium stage 1 supplemented with 0.3% d-glucose, 1% penicillin/streptomycin/glutamine, 2% B27 supplement, 250 ng/mL SHH (No. 100-45; PeproTech), 100 ng/mL DKK-1 (No. 120-30; PeproTech), 20 ng/mL brain-derived neurotrophic factor (BDNF; No. 450-02; PeproTech), 25 ng/mL FGF-2, 10 μM Y-27632 (No. 688001; Calbiochem), and 10 μM retinoic acid in NBA.

After 10 days, culture conditions were changed to neural differentiation medium stage 2 comprising 3% d-glucose, 1% penicillin/streptomycin/glutamine, 2% B27 supplement, 1 mM dibutyryl-cyclic AMP (No. D0260; Sigma-Aldrich), 1 mM VPA, 20 ng/mL BDNF, and 10 μM Y-27632 in NBA. All stock concentrations and solvents of the supplements are listed (Supplementary Table S1).

Treatment with TAT-MeCP2 fusion protein

TAT-MeCP2, purified as previously described [28], was added to the neuronal differentiation medium stage 2 on days 23, 25, and 28 of neuronal differentiation at a concentration of 50 nM. Control cells were incubated with MeCP2 storage buffer [DPBS, 200 mM NaCl, 10% (v/v) Glycerol, 0.05% (v/v) CHAPS, pH = 7.2].

RNA extraction and quantitative reverse transcription polymerase chain reaction

Total RNA was extracted from reprogrammed iNPs and nontransfected control cells using the RNeasy Plus Mini Kit as per the manufacturer's instructions (No. 74134; Qiagen). Total RNA from neurons was isolated on day 30 of neuronal differentiation using the RNeasy Plus Micro Kit (No. 74034; Qiagen). iScript™ genomic DNA (gDNA) Clear cDNA Synthesis Kit (No. 1725035; Bio-Rad) was used to synthesize cDNA from total RNA. For qPCR the iTaq Universal Probes Supermix (No. 1725131; Bio-Rad) was used. qPCR analyses were carried out using comparative Ct (ΔΔCt) method. PPIA was used as the internal standard. The used TaqMan^® Assays are listed in Supplementary Table S2.

Immunofluorescence staining

HDFs and iNPs were fixed with ice-cold 4% (vol/vol) formaldehyde (No. 28906; Thermo Scientific) in DPBS for 20 min at 4°C and permeabilized in DPBS with 0.2% Triton X-100 (No. T8787; Sigma-Aldrich) twice for 5 min. After blocking for 30 min with 3% bovine serum albumin (BSA, No. A9647; Sigma-Aldrich) in DPBS, primary antibodies with 3% goat serum (No. G9023; Sigma) were applied overnight at 4°C. Before and after incubation with secondary antibodies (Goat anti-Mouse IgG Alexa Fluor™ 488 No. 11029, Goat anti-Rabbit IgG Alexa Fluor 568™ No. 11011; Invitrogen, 1:250 in DPBS with 3% goat serum) for 1 h at room temperature, cells were washed with DPBS thrice for 5 min. Coverslips were mounted using ProLong™ Diamond Antifade Mountant with DAPI (No. P36966; Invitrogen).

For immunofluorescence staining of neurons, a slightly different protocol was used. Cells were fixed for 10 min at 4°C and permeabilized thrice for 2 min. Blocking was carried out overnight at 4°C with 30% goat serum in DPBS. The used primary antibodies are listed in Supplementary Table S3.

All fluorescent images were acquired using a confocal microscope (DMI6000; Leica Microsystems) and LAS × 3.5 software. The following objectives were used: HC PL APO CS 20 × /0.70 UV (No. 506513; Leica Microsystems) and HCX PL APO CS 40 × /1.25 OIL PH3 UV (No. 506181; Leica Microsystems). Images were deconvolved with Huygens Essential version 23.04 (Scientific Volume Imaging; http://svi.nl).

Electrophysiological recordings

For whole-cell patch clamp recordings, an Axopatch 200B amplifier, a Digidata 1320A digitizer, and the Clampex 10.7 software (Molecular Devices) were used. The external solution contained 120 mM NaCl, 3 mM KCl, 2 mM MgCl₂6H₂O, 20 mM glucose, 10 mM HEPES, and 2 mM CaCl₂H₂O (pH = 7.3, 290–310 mOsmol). Patch electrodes were filled with 130 mM CsCl, 20 mM TEACl, 0.24 mM CaCl₂H₂O, 5 mM EGTA, 10 mM glucose, and 10 mM HEPES (pH = 7.3). The recordings were performed at room temperature. The resistance of the patch pipettes was between 2 and 5 MOhm. For blocking voltage-gated sodium channels, 1 μM tetrodotoxin (TTX) was added through a drug application system. The holding potential was set to −80 mV, and the current was recorded by a depolarizing voltage step to −60 mV.

Neuronal morphology analysis

For quantification of the number of neurites, branches, and the total neurite outgrowth per neuron, 30-day differentiated cells were stained with TUJ1 as described. For image acquisition, 10 z-stacks with 1 μm z-stack resolution were acquired. Images were analyzed using the machine learning software AIVIA 12.0.0 (Leica Microsystems). First, a three-dimensional (3D) pixel classifier training set was created, which recognized soma and neurites. Before using the “3D Neuron Analysis” recipe to quantify the number of neurites, branches, and total neurite outgrowth per neuron, an average filter was applied. From three independent experiments at least 100 neurons per group were analyzed.

The number of vesicular glutamate transporter 1 (VGLUT1) puncta per μm dendrite was quantified using the Fiji plugin “Analyze Particles” [29], as described previously [30]. Before analysis, images were subjected to background subtraction with a rolling ball radius of 10. Subsequently, thresholding was applied consistently to all images of each experiment. To separate puncta in close proximity to each other, the watershed algorithm was applied. Anti-microtubule-associated protein 2 (MAP2) images served as dendritic mask. Regions of interest (ROIs) were chosen in the dendritic mask at a width of 20 μm. The length of the dendrite in the ROI was measured, and the puncta were quantified. The puncta were subsequently calculated as puncta per μm of dendrite.

In-Cell Western Assay

In-Cell Western Assay was used to quantify the protein acetylation level of H3K9 (No. H9286, 1:200; Sigma) and H4K16 (No. 07-329, 1:200; Sigma). iNPs were seeded in 96-well plates (No. 655090; Greiner Bio-One). After 30 days, differentiated neurons were fixed with Histofix (No. P087.5; Roth) for 20 min at room temperature and permeabilized thrice in Tris-buffered saline (TBS) with 0.1% Triton X-100 at 30 rpm agitation. After blocking for 90 min using the Intercept^® Blocking Buffer (No. 927-60001; LI-COR), primary antibodies were diluted in Intercept Antibody Diluent (No. 927-65001; LI-COR) and incubated overnight at 4°C.

On the following day, cells were incubated with secondary antibody (IRDye^® 800CW Goat anti-Rabbit IgG Secondary Antibody, No. 926-32211, 1:1,000; LI-COR) and CellTag™ 700 (Stain for In-Cell Western™ Assays, No. 926-41090, 1:500; LI-COR), diluted in Intercept Antibody Diluent, for 1 h at room temperature. Before and after the 1-h incubation, cells were washed five times in TBS with 0.1% Tween-20 (No. P9416; Sigma). Once the plate was completely dry, the plate was imaged using the Odyssey CLx Imager (LI-COR). Blank wells were used for background subtraction. Fluorescence intensity was normalized to CellTag 700.

RNA-sequencing and data analysis

Total RNA was extracted from each of three independent samples of MeCP2-deficient and wild-type HDF cells, iNPs, and neurons as described above. Sequencing libraries were prepared at the Core Facility Genomics, Medical University of Vienna, using the SMARTer Stranded Total RNA Seq Kit—Pico v2 according to the manufacturer's protocols (Takara Bio). Libraries were QC-checked on a Bioanalyzer 2100 (Agilent) using a High Sensitivity DNA Kit for correct insert size and quantified using Qubit dsDNA HS Assay (Invitrogen). Pooled libraries were sequenced on a NextSeq 500 instrument (Illumina) in 1 × 75 bp single-end sequencing mode. Approximately 22 million reads were generated per sample. Reads in fastq format were aligned to the human reference genome version GRCh38 [31] with Gencode 29 annotations [32] using STAR aligner [33] version 2.6.1a in 2-pass mode. Reads per gene were counted by STAR, and differential gene expression was calculated using DESeq2 [34] version 1.20.0.

Differentially expressed genes (DEGs) were defined as those with log2-fold change above 1 or below −1 and an adjusted P value threshold of 0.05. Pathway analysis of DEGs was performed using the web-based tool NetworkAnalyst [35,36] or g:Profiler [37]. The principal component analysis (PCA) plot and clustered heatmaps are based on regularized log-transformed count data generated in DESeq2 [34]. Heatmaps were generated using the coolmap function of the R package limma [38]. VolcaNoseR web app was used for creation of the volcano plot [39]. Estimation of the cell identity was performed through https://voineagulab.shinyapps.io/BrainDeconvShiny [40], Transcript per million (TPM) was used as input, the “DM” [41] as signature matrix.

Statistical analysis

All experiments were performed at least independent times from three independent differentiations. Analysis of the data was carried out with GraphPad Prism 9. Data are given as mean ± standard deviation. Unpaired t-tests were used to test for significance, with *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Results

HDFs for reprogramming and differentiation

We used patient-derived HDFs of a male individual with a deletion of a guanine residue at position 806 of the MECP2 gene for reprogramming and neuronal differentiation, henceforth referred to as “MeCP2-deficient” or “patient-derived” (Fig. 1A). This in turn leads to a frame-shift mutation at amino acid residue 269 at the transcriptional repressor domain (TRD) portion of MeCP2, resulting in a truncated protein (Fig. 1B). To confirm this finding, we investigated the cellular localization of the mutated protein and stained sex-matched MeCP2-deficient and wild-type fibroblasts with an N-terminal and a C-terminal MeCP2 antibody as a control. In wild-type fibroblasts, MeCP2 could be detected in the cell nucleus with both the N- and C-terminal MeCP2 antibodies. In MeCP2-deficient cells, MeCP2 was detected using only its N-terminal antibody, while no staining was observed with the C-terminal antibody, pointing to the truncated MeCP2 presence in MeCP2-deficient fibroblasts (Fig. 1C).

FIG. 1.

Mutational analysis of MeCP2-deficient HDFs. (A) Genomic DNA sequencing of MeCP2-deficient and wild-type HDFs. The arrow indicates the mutated nucleotide position. (B) MeCP2 domain architecture: The p.G269AfsX20 (c.806delG) pathogenic variant is located at position 269. (C) Immunofluorescence images of MeCP2-deficient and wild-type fibroblasts with an N-terminal and a C-terminal MeCP2 antibody. Cell nuclei were counterstained with DAPI. Scale: 30 μm. HDF, human dermal fibroblast; ID, intervening domain; MBD, methyl CpG binding domain; MeCP2, methyl CpG binding protein 2; NLS, nuclear localization signal; TRD, transcriptional repressor domain.

Direct reprogramming of HDFs into iNPs through episomal expression of PAX6 and SOX2

MeCP2-deficient and wild-type fibroblasts were transfected through electroporation with two episomal plasmids encoding the transcription factors PAX6 and SOX2. Reprogramming of patient-derived and wild-type fibroblasts resulted in the formation of iNPs, which were then further differentiated into neuronal cells. An overview of the reprogramming and differentiation process is shown in Fig. 2A.

FIG. 2.

Direct conversion of HDFs into iNPs and neurons. (A) Schematic representation of the reprogramming and differentiation process from HDFs to iNPs and neurons. (B) Bright-field images depicted morphological changes from fibroblasts to iNPs during the reprogramming process. Scale: 50 μm. (C) MeCP2-deficient iNPs expressed SOX2 and PAX6, as well as the neuronal progenitor markers NGN2 and NES. Cell nuclei were counterstained with DAPI. Scale: 50 μm. (D) Real-time PCR analysis demonstrated strong upregulation of the prototypic neuronal progenitor associated genes such as FOXG1, NCAM1, and NES in MeCP2-deficient and wild-type iNPs compared to HDFs. In addition, gene expression levels of SOX2 and PAX6 were also elevated in iNPs. PPIA was used as endogenous control (n = 3). iNP, induced neuronal progenitor; NES, nestin; NGN2, neurogenin-2; PCR, polymerase chain reaction.

Three days post-transfection, the cells were stained for PAX6 and SOX2 to assess the transfection efficiency (Supplementary Fig. S1). Positive PAX6 and SOX2 staining was observed, pointing to the fact that the transfection was indeed successful. The medium was changed from fibroblast growth medium to neuronal reprogramming medium. Weekly passaging from day 17 on led to the generation of iNPs. Bright-field microscopy images showed the morphological changes of the reprogramming process from elongated and flat fibroblasts to neurosphere-like colonies of iNPs over time (Fig. 2B).

For validation of the neuronal progenitor identity, iNP protein levels were examined on day 38 of the reprogramming process. Using immunocytochemistry, sustained expression of the transcription factors PAX6 and SOX2, which were originally used to initiate the reprogramming process, was confirmed. In addition, the neuronal progenitor markers Neurogenin-2 and Nestin were expressed in MeCP2-deficient and wild-type iNPs (Fig. 2C). At the same timepoint, RNA was isolated and subsequent qPCR showed increased gene expression levels of neuronal progenitor associated genes FOXG1, NCAM1, and NES in iNPs, compared to HDFs. In addition, the gene expression of PAX6 and SOX2 was strongly upregulated in iNPs (Fig. 2D).

Generation of patient-derived neurons from iNPs

On day 38 of reprogramming, iNPs were seeded for neuronal differentiation according to a two-step differentiation protocol. The morphology of the cells changed to a neuron-like phenotype with a soma and branched extensions (Fig. 3A). MeCP2-deficient, as well as wild-type, neurons expressed neuron-specific proteins such as MAP2 and neuronal nuclei (NeuN) in addition to glutamatergic (VLGUT1) and GABAergic [glutamate decarboxylase (GAD65/67)] markers (Fig. 3B). In addition, the presence of voltage-gated sodium channels in iNP-derived neurons was demonstrated using the patch clamp technique in the whole cell configuration. To isolate inward currents, potassium channels were blocked with an intracellular solution containing cesium chloride. To confirm the presence of voltage-gated sodium channels, currents in response to depolarizing voltage steps were blocked with 1 μM TTX, followed by a washout period of TTX with recovery of the channel activity (Fig. 3C).

FIG. 3.

Differentiation of iNPs into neuronal cells. (A) Bright-field images showed the morphological shift from neurosphere-like cells to neurons with a soma and neurites in MeCP2-deficient, as well as wild-type cells. Scale: 50 μm. (B) iNP-derived MeCP2-deficient neurons expressed the neuronal-lineage markers MAP2 and NeuN. Moreover, there was a positive stain for VGLUT1, a glutamatergic, and for GAD65/67, a GABAergic marker. Scale: 30 μm. (C) Block of voltage-gated sodium channels of iNP-derived neurons with TTX. Black record: sodium current before TTX; red record: inhibitory effect of TTX; green record: recovery. GAD65/67, glutamate decarboxylase; MAP2, anti-microtubule-associated protein 2; NeuN, neuronal nuclei; TTX, tetrodotoxin; VGLUT1, vesicular glutamate transporter 1.

RNA-Seq confirms neuronal identity

RNA-Seq analysis was performed in triplicates in MeCP2-deficient and wild-type HDFs, iNPs, and neurons to demonstrate successful differentiation and to compare patient-derived and wild-type cells. Initially, PCA based on regularized log-transformed count data was used to get an insight of similarities and differences between each sample group. The first principal component (PC1), which represents the largest possible variance, clearly distinguished fibroblasts from iNPs and neurons. The second principal component (PC2) defined iNPs and neurons as two distinct clusters (Supplementary Fig. S2).

To demonstrate gene expression changes during differentiation, pathway analysis was carried out on DEGs between fibroblasts and neurons of MeCP2-deficient, as well as wild-type, cells (Fig. 4A and Supplementary Table S4). Using Protein ANalysis THrough Evolutionary Relationships (PANTHER) gene ontology database, cellular component analysis displayed a significant upregulation of neuron-specific cellular components such as “axon,” “neuronal cell body,” and “dendrite” in both patient-derived and wild-type neuronal cells, whereas expression of genes related to “cytosol,” “cytoplasm,” “cytoskeleton,” and “plasma membrane” was diminished (Fig. 4B). These similar expression patterns showed that patient-derived and wild-type neurons were comparable, and differentiation resulted in enrichment of neuron-specific cellular components.

FIG. 4.

Gene expression profile of the differentiation process from fibroblasts to neurons. (A) Volcano plot representing DEGs in neurons compared to fibroblasts of MeCP2-deficient and wild-type cells. The top 10 hits are labeled. Dark gray marks significantly increased, light gray significantly decreased gene expression. (B) PANTHER gene function analysis regarding cellular component revealed an enrichment of neuron-specific pathways such as “neuronal cell body,” “axon” or “dendrite” in MeCP2-deficient (left panel) and wild-type (right panel) neuronal cells, compared to untransfected fibroblasts (FDR <0.001). (C) RNA-Seq data of fibroblasts, iNPs, and neurons from MeCP2-deficient and wild-type cells exhibited a similar gene expression pattern of 22 common differently expressed genes from a meta-analysis of induced neurons. The darker the gray, the stronger the up- or downregulation. DEGs, differentially expressed genes; FDR, false discovery rate; PANTHER, Protein ANalysis THrough Evolutionary Relationships; RNA-Seq, RNA-sequencing.

To further confirm neuronal identity, we compared our RNA-Seq expression dataset with a meta-analysis of 10 high throughput expression datasets of induced neurons from human and mouse fibroblasts demonstrating a common gene expression pattern of differentiation [42]. Comparing our dataset with the 22 common DEGs described in this meta-analysis, we found similar gene expression changes. Gene expression of iNPs, which already underwent a transformation to a neuronal lineage, was more comparable to neurons than fibroblasts (Fig. 4C).

To estimate the cell identity after the reprogramming and differentiation process, we compared RNA-Seq data from wild-type and MeCP2-deficient neurons with published data from neurons and glia cells [41]. MeCP2-deficient and wild-type cells revealed 74.3% and 72.4% neuronal cell identity, respectively (Supplementary Fig. S3). Hence, our data indicate that there is no significant difference in the differentiation potential of these two cell lines.

Morphological and gene expression analysis of neurons

The c.806delG pathogenic variant was stable throughout the reprogramming and differentiation processes, and MeCP2 could not be detected through immunofluorescence staining using the C-terminal MeCP2 antibody after 4 weeks of neuronal differentiation. For quantification of the morphological structure of MeCP2-deficient and wild-type neurons, cells were stained with TUJ1, a marker for mature neurons (Fig. 5A). Using the machine learning based software AIVIA, the number of neurites, branches, and the total neurite outgrowth were determined (Fig. 5B). Statistical analysis of the morphological structure confirmed a reduction in the number of neurites, branches, and total neurite outgrowth per neuron in MeCP2-deficient cells, compared to their wild-type counterparts (Fig. 5C). These results resemble the observations in postmortem human brain samples of RTT patients [43], which also showed a diminished dendritic arborization.

FIG. 5.

Morphological and gene expression analysis of MeCP2-deficient and wild-type neurons. (A) Both MeCP2-deficient and wild-type neurons expressed the neuronal marker TUJ1 and exhibited a typical neuronal morphology. In contrast, MeCP2 protein could not be detected in mutant cells. Cell nuclei were counterstained with DAPI. (B) Representative images of MeCP2-deficient and wild-type neurons, which were processed using the AIVIA software and used for morphological analysis. (C) Number of neurites and branches per neuron, as well as the total neurite outgrowth per neuron of TUJ1-positive cells, was quantified on day 30 of neuronal differentiation. Significance was determined by unpaired t-test with **P < 0.01 (n > 100 neurons from three independent experiments per group). (D) Representative images of wild-type and MeCP2-deficient neurons showing VGLUT1 puncta on dendrites, which were stained with MAP2. (E) Quantification of VGLUT1 puncta per μm MAP2-positive neurite. Significance was determined by unpaired t-test with *P < 0.05 (n = 3). (F) Gene expression of BDNF, GRIA1, and GRIK2 was evaluated using quantitative real-time PCR. Gene expression was calculated as fold change to wild-type neurons. PPIA was used as endogenous control. Significance was determined by unpaired t-test with *P < 0.05 and ***P < 0.001 (n = 3).

Another characteristic of RTT is the decreased protein expression of VGLUT1, a vesicular glutamate transporter [15,44]. MeCP2-deficient neurons showed a significant reduction of VGLUT1 puncta per μm MAP2-positive dendrite, compared to their wild-type counterparts (Fig. 5D, E).

Impaired MeCP2 function is also known to affect BDNF levels. BDNF is a prominent target gene of MeCP2 [45 –47]. Gene expression analysis using qPCR revealed a significant downregulation of BDNF in MeCP2-deficient neurons, compared to their wild-type counterparts (Fig. 5D), in line with previous work [48]. RTT is also associated with dysfunctions in the glutamatergic system [15,49,50]. GRIA1 and GRIK2, which encode ionotropic glutamate receptors, were significantly downregulated in MeCP2-deficient neurons (Fig. 5F).

KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis of DEGs between MeCP2-deficient neurons revealed enrichment in “calcium signaling pathway,” “ECM-receptor interaction,” or “axon guidance” (Supplementary Fig. S4). These pathways were also found to be enriched in motor cortex and cerebellum samples of RTT patient autopsies [51].

TAT-MeCP2 treatment reduces histone hyperacetylation in MeCP2-deficient neurons

A hallmark of RTT pathophysiology on the cellular level is the hyperacetylation of histone H3 and H4 [52 –54]. For investigation of acetylation levels in patient-derived neurons, an In-Cell Western Assay was performed. Increased acetylation levels of H3K9 and H4K16 were detected in MeCP2-deficient neurons (Fig. 6). As treatment with TAT-MeCP2 was shown to be sufficient to ameliorate histone hyperacetylation in RTT-relevant cells [53], we investigated whether our cell model is a suitable tool to quantify the therapeutic potential of TAT-MeCP2 fusion protein by treating MeCP2-deficient neurons with TAT-MeCP2 fusion protein on days 23, 25, and 28 of neuronal differentiation. Following treatment, H4K16 amounts have reverted to wild-type levels (Fig. 6B). This reduction of hyperacetylation indicates that the observed differences between wild-type and MeCP2-deficient occur due to MeCP2 deficiency.

FIG. 6.

(A) H3K9 and (B) H4K16 acetylation levels of wild-type and MeCP2-deficient neurons were analyzed using the In-Cell Western Assay. Expression levels of H3K9 and H4K16 were significantly elevated in MeCP2-deficient neurons. Treatment of MeCP2-deficient neurons with TAT-MeCP2 resulted in a reduction of histone acetylation levels. Fluorescence intensity of H3K9 and H4K16 was normalized to CellTag 700. Acetylation levels were normalized to wild-type neurons. Significance was determined by unpaired t-test with *P < 0.05 (n = 3).

However, the morphological abnormalities of MeCP2-deficient neurons were not ameliorated within 7 days of TAT-MeCP2 treatment at the used dose (Supplementary Fig. S5).

Discussion

Disease recapitulating cell culture models are a vital tool for the investigation of RTT in vitro, as postmortem human brain samples are scarce and cannot be used for preclinical testing of novel treatment approaches. Furthermore, mouse models cannot accurately reflect human diseases due to species specific differences [55,56]. In this study, we present for the first time the application of a direct reprogramming method [19,26] to generate neurons from fibroblasts of an individual carrying a pathogenic variant in the X-linked MECP2 gene. To initiate fibroblast reprogramming, the transcription factors PAX6 and SOX2 were exogenously introduced to initiate the reprogramming process, as described previously [19,26].

To date, most of the human neuronal cell culture models for RTT were based on iPSCs or embryonic stem cells (hESCs) [15,57,58]. Using a direct reprogramming method, in contrast to iPSCs, the pluripotent cell state is bypassed and epigenetic aging signatures are maintained, leading to a more accurate reflection of the genetic and epigenetic background of the patient [21,59]. In the case of MeCP2, it is particularly important as this protein is a critical epigenetic regulator [60,61], and as such, a direct programming method—without the generation of a pluripotent intermediate—is preferential.

We used HDFs from a male individual with neonatal encephalopathy carrying a frameshift deletion in the MECP2 gene (c.806delG), although RTT is primarily diagnosed in females. As the MECP2 gene is located on the X chromosome, the use of male cells with only one X chromosome is beneficial to study pathophysiological changes caused by the complete loss of functional MeCP2. A case report about a male individual with an insertional MECP2 mutation presents with a clinically similar phenotype to female RTT patients [62].

In this study, we could demonstrate the successful reprogramming and differentiation of MeCP2-deficient and wild-type fibroblasts. iNPs showed a typical neurosphere-like morphology with the expression of neuronal progenitor markers Neurogenin-2 and Nestin. Furthermore, when comparing these iNPs with untransfected HDFs, an increase in gene expression of the neuronal precursors FOXG1, NCAM1, and NES could be shown (Fig. 2), consistent with similar reprogramming methods [19,20].

The successful differentiation of iNPs to neuronal cells was confirmed using immunocytochemistry depicting the expression of neuronal markers such as TUJ1 or MAP2. Further evidence for differentiation of iNPs toward a neuronal lineage was the PANTHER cellular component analysis, which showed an enrichment of neuron-specific cell structures, indicating the presence of a neuronal cell population, compared to untransfected fibroblasts. Moreover, we could demonstrate that wild-type, as well as MeCP2-deficient, neurons exhibit a typical gene expression pattern, which was also found in other direct reprogramming approaches for the generation of induced neurons from mouse and human fibroblasts [42].

So far, the c.806delG pathogenic variant was only studied in human iPSC-derived neural cells by Andoh-Noda et al. [63]. Human iPSCs were derived from RTT-monozygotic female twins. Differentiation of these two pairs of isogenic mutant and wild-type iPSCs resulted in an enhanced astrocytic differentiation of MeCP2-deficient cells. In contrast, using the direct reprogramming approach here described did not result in a significant difference in the astrocyte population between MeCP2-deficient and wild-type cells. However, according to the cell identity population estimation of our cell model, the percentage of astrocytes is very low.

Analysis of neuronal morphological structure of MeCP2-deficient neurons showed a decrease in the number of neurites, branches, and total neurite outgrowth per neuron in comparison to their wild-type counterparts. Similar observations were made in postmortem human brain samples from RTT patients [43], mouse models [64,65], and iPSC-derived neurons [15,58].

In addition, gene expression levels of BDNF, a gene important for neuronal development and survival [66], were downregulated in MeCP2-deficient neurons, in line with a meta-analysis of RTT transcriptomic data from different species, models, and MECP2 pathogenic variants [48]. In MeCP2-deficient neurons, a significant downregulation of GRIA1 and GRIK2 gene expression was also observed. Analysis of postmortem human brain samples of RTT patients also showed a decrease of GRIA1 gene expression [67]. A similar decrease was found in patients with schizophrenia [68]. Pathogenic variants in the GRIK2 gene, which led to loss of function of the encoded protein, are linked to autosomal recessive intellectual disability [69]. Moreover, puncta density of the vesicular glutamate receptor VGLUT1 was diminished in MeCP2-deficient neurons, which is in line with previous findings in other RTT disease models [15,44] and points to a general dysregulation of the glutamatergic system.

Another characteristic of RTT is the hyperacetylation of H3K9 and H4K16 [52 –54], which was also present in the used MeCP2-deficient neurons. Furthermore, the treatment of MeCP2-deficient neurons with recombinant TAT-MeCP2 fusion protein resulted in a decrease of H4K16 acetylation levels, underlining the applicability of this neuronal cell model for the study of novel treatment options such as protein replacement therapy [28,53,70,71], reactivation of the inactive X-chromosome [72], or gene therapy [73 –75]. In addition, the use of a direct reprogramming strategy would be beneficial for the study of DNA methylation patterns in RTT, as the epigenetic state of the cells is maintained [21,76,77]. Considering the benefits of direct reprogramming, this method serves as a valuable tool to study neurological diseases as it was shown for Huntington's disease [78], fragile X syndrome [30], or Parkinson's disease [18].

Although TAT-MeCP2 treatment ameliorated acetylation levels in MeCP2-deficient neurons, no significant improvement of the dendritic arborization was observed. For the rescue of morphological abnormalities, we hypothesize that a more extended treatment period may be required. This would be in line with findings of treatment with VPA for 20 days [57] or Nefiracetam for 14 days [79], which showed beneficial effects on the morphology of iPSC-derived RTT neurons.

One limitation of the study is the lack of further electrophysiological characterization of MeCP2-deficient and wild-type neurons. Although the presence of voltage-dependent sodium channels was confirmed, no action potentials could be recorded. To quantify action potentials, a longer differentiation time could be required. However, in RTT-iPSC-derived neurons, a significant impairment of action potential properties was observed [7,15]. Therefore, a more rigorous electrophysiological analysis of directly reprogrammed MeCP2-deficient and wild-type neurons will be the subject of further studies.

Conclusion

In conclusion, we have demonstrated that MeCP2-deficient HDFs can be reprogrammed and differentiated into neuronal cells through transient expression of the transcription factors SOX2 and PAX6. The generated MeCP2-deficient neurons express typical neuronal markers and the appropriate morphology. The MeCP2-deficient neurons exhibit properties typical for RTT, including diminished dendritic arborization and histone hyperacetylation, compared to their wild-type counterparts. Therefore, this cell model can act as a reliable proxy for the study of RTT and serve as an important platform for investigating prospective RTT treatment approaches.

Footnotes

Acknowledgments

The authors acknowledge the Core Facilities of the Medical University of Vienna, a member of VLSI, for the contribution of RNA-Seq experiment. The authors also thank Britta Kluge, Mario Mikula, Reinhard Lehner, and Klaus Schicker for their technical and consultative assistance. The research conducted at the Murdoch Children's Research Institute (MCRI) was supported by the Victorian Government's Operational Infrastructure Support Program. The Chair in Genomic Medicine awarded to J.C. is generously supported by The Royal Children's Hospital Foundation. was created with BioRender.com

Author Disclosure Statement

F.L. is inventor of the patent WO2007115578 A8, “Synthetic mecp2 sequence for protein substitution therapy”; holder of the patent: Georg August University, Universitätsmedizin Göttingen. All other authors declare no conflicts of interest.

Funding Information

This research was mainly funded by the Italian Rett Syndrome Association (AIRETT ETS) grant (H.S.). The initial phase of the study was supported by the German “Elternhilfe für Kinder mit Rett Syndrom.” I.G. was funded by Austrian Science Fund (FWF) grant I 3778, which is part of the ERANET programme iPS&Brain.

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

References

Amir

, Van den Veyver

, Wan

, Tran

, Francke

and Zoghbi

. (1999). Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet, 23:185–188.

Tarquinio

and Percy

. (2016). Rett syndrome. In: Neuronal and Synaptic Dysfunction in Autism Spectrum Disorder and Intellectual Disability. Sala C, Verpelli C, eds. United States: Elsevier, pp 301–323.

Hagberg

, Aicardi

, Dias

and Ramos

. (1983). A progressive syndrome of autism, dementia, ataxia, and loss of purposeful hand use in girls: Rett's syndrome: report of 35 cases. Ann Neurol, 14:471–479.

Neul

, Kaufmann

, Glaze

, Christodoulou

, Clarke

, Bahi-Buisson

, Leonard

, Bailey

MES

, Schanen

, et al. (2010). Rett syndrome: revised diagnostic criteria and nomenclature. Ann Neurol, 68:944–950.

Jellinger

and Seitelberger

. (1986). Neuropathology of Rett syndrome. Am J Med Genet Suppl, 1:259–288.

Bauman

, Kemper

and Arin

. (1995). Microscopic observations of the brain in Rett syndrome. Neuropediatrics, 26:105–108.

Farra

, Zhang

, Pasceri

, Eubanks

, Salter

and Ellis

. (2012). Rett syndrome induced pluripotent stem cell-derived neurons reveal novel neurophysiological alterations. Mol Psychiatry, 17:1261–1271.

Mok

RSF

, Zhang

, Sheikh

, Pradeepan

, Fernandes

, DeJong

, Benigno

, Hildebrandt

, Mufteev

, et al. (2022). Wide spectrum of neuronal and network phenotypes in human stem cell-derived excitatory neurons with Rett syndrome-associated MECP2 mutations. Transl Psychiatry, 12:450.

Guy

, Hendrich

, Holmes

, Martin

and Bird

. (2001). A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat Genet, 27:322–326.

10.

Garg

, Lioy

, Cheval

, McGann

, Bissonnette

, Murtha

, Foust

, Kaspar

, Bird

and Mandel

. (2013). Systemic delivery of MeCP2 rescues behavioral and cellular deficits in female mouse models of Rett syndrome. J Neurosci, 33:13612–13620.

11.

Robinson

, Guy

, McKay

, Brockett

, Spike

, Selfridge

, De Sousa

, Merusi

, Riedel

, Bird

and Cobb

. (2012). Morphological and functional reversal of phenotypes in a mouse model of Rett syndrome. Brain, 135:2699–2710.

12.

Lang

, Wither

, Colic

, Wu

, Monnier

, Bardakjian

, Zhang

and Eubanks

. (2014). Rescue of behavioral and EEG deficits in male and female Mecp2-deficient mice by delayed Mecp2 gene reactivation. Hum Mol Genet, 23:303–318.

13.

Guy

, Gan

, Selfridge

, Cobb

and Bird

. (2007). Reversal of neurological defects in a mouse model of Rett syndrome. Science, 315:1143–1147.

14.

Takahashi

, Tanabe

, Ohnuki

, Narita

, Ichisaka

, Tomoda

and Yamanaka

. (2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 131:861–872.

15.

Marchetto

MCN

, Carromeu

, Acab

, Yu

, Yeo

, Mu

, Chen

, Gage

and Muotri

. (2010). A model for neural development and treatment of Rett syndrome using human induced pluripotent stem cells. Cell, 143:527–539.

16.

Polo

, Anderssen

, Walsh

, Schwarz

, Nefzger

, Lim

, Borkent

, Apostolou

, Alaei

, et al. (2012). A molecular roadmap of reprogramming somatic cells into iPS cells. Cell, 151:1617–1632.

17.

Miller

, Ganat

, Kishinevsky

, Bowman

, Liu

, Tu

, Mandal

, Vera

, Shim

, et al. (2013). Human iPSC-based modeling of late-onset disease via progerin-induced aging. Cell Stem Cell, 13:691–705.

18.

Playne

, Jones

and Connor

. (2018). Generation of dopamine neuronal-like cells from induced neural precursors derived from adult human cells by non-viral expression of lineage factors. J Stem Cells Regen Med, 14:34–44.

19.

Maucksch

, Firmin

, Butler-Munro

, Montgomery

, Dottori

and Connor

. (2012). Non-viral generation of neural precursor-like cells from adult human fibroblasts. J Stem Cells Regen Med, 8:162–170.

20.

Connor

, Firmin

, McCaughey-Chapman

, Monk

, Lee

, Liot

, Geiger

, Rudolph

and Jones

. (2018). Conversion of adult human fibroblasts into neural precursor cells using chemically modified mRNA. Heliyon, 4:e00918.

21.

McCaughey-Chapman

, Tarczyluk-Wells

, Combrinck

, Edwards

, Jones

and Connor

. (2023). Reprogramming of adult human dermal fibroblasts to induced dorsal forebrain precursor cells maintains aging signatures. Front Cell Neurosci, 17:1003188.

22.

Graham

, Khudyakov

, Ellis

and Pevny

. (2003). SOX2 functions to maintain neural progenitor identity. Neuron, 39:749–765.

23.

Pevny

and Placzek

. (2005). SOX genes and neural progenitor identity. Curr Opin Neurobiol, 15:7–13.

24.

Zhang

, Huang

, Chen

, Pankratz

, Xi

, Li

, Yang

, Lavaute

, Li

X-J

, et al. (2010). Pax6 is a human neuroectoderm cell fate determinant. Cell Stem Cell, 7:90–100.

25.

Sansom

, Griffiths

, Faedo

, Kleinjan

D-J

, Ruan

, Smith

, van Heyningen

, Rubenstein

and Livesey

. (2009). The level of the transcription factor Pax6 is essential for controlling the balance between neural stem cell self-renewal and neurogenesis. PLoS Genet, 5:e1000511.

26.

Connor

, Connor

, Firmin

, Maucksch

, Liu

, Playne

, Jones

and Dottori

. (2015). Direct conversion of adult human fibroblasts into induced neural precursor cells by non-viral transfection. Protoc Exch; doi: 10.1038/protex.2015.034

27.

Gold

, Lacina

, Cantrill

and Christodoulou

. (2015). MeCP2 deficiency is associated with reduced levels of tubulin acetylation and can be restored using HDAC6 inhibitors. J Mol Med, 93:63–72.

28.

Beribisky

, Steinkellner

, Geislberger

, Huber

, Sarne

, Christodoulou

and Laccone

. (2022). Expression, purification, characterization and cellular uptake of mecp2 variants. Protein J, 41:345–359.

29.

Schindelin

, Arganda-Carreras

, Frise

, Kaynig

, Longair

, Pietzsch

, Preibisch

, Rueden

, Saalfeld

, et al. (2012). Fiji: an open-source platform for biological-image analysis. Nat Methods, 9:676–682.

30.

Edwards

, Combrinck

, McCaughey-Chapman

and Connor

. (2023). Directly reprogrammed fragile X syndrome dorsal forebrain precursor cells generate cortical neurons exhibiting impaired neuronal maturation. Front Cell Neurosci, 17:1254412.

31.

GRCh38 human reference genome downloaded from NCBI, 2018. Available from: ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/000/001/405/GCA_000001405.15_GRCh38/seqs_for_alignment_pipelines.ucsc_ids/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz [Last accessed: September 27, 2018].

32.

Human genome annotations downloaded from Gencode, 2018. Available from: ftp://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_29/[1]]gencode.v29.chr_patch_hapl_scaff.annotation.gtf.gz [Last accessed: November 22, 2018].

33.

Dobin

, Davis

, Schlesinger

, Drenkow

, Zaleski

, Jha

, Batut

, Chaisson

and Gingeras

. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics, 29:15–21.

34.

Love

, Huber

and Anders

. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol, 15:550.

35.

Xia

, Gill

and Hancock

REW

. (2015). NetworkAnalyst for statistical, visual and network-based meta-analysis of gene expression data. Nat Protoc, 10:823–844.

36.

Zhou

, Soufan

, Ewald

, Hancock

REW

, Basu

and Xia

. (2019). NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res, 47:W234–W241.

37.

Raudvere

, Kolberg

, Kuzmin

, Arak

, Adler

, Peterson

and Vilo

. (2019). g:Profiler: a web server for functional enrichment analysis and conversions of gene lists (2019 update). Nucleic Acids Res, 47:W191–W198.

38.

Ritchie

, Phipson

, Wu

, Hu

, Law

, Shi

and Smyth

. (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res, 43:e47.

39.

Goedhart

and Luijsterburg

. (2020). VolcaNoseR is a web app for creating, exploring, labeling and sharing volcano plots. Sci Rep, 10:20560.

40.

Sutton

, Poppe

, Simmons

, Walsh

, Nawaz

, Lister

, Gagnon-Bartsch

and Voineagu

. (2022). Comprehensive evaluation of deconvolution methods for human brain gene expression. Nat Commun, 13:1358.

41.

Darmanis

, Sloan

, Zhang

, Enge

, Caneda

, Shuer

, Hayden Gephart

, Barres

and Quake

. (2015). A survey of human brain transcriptome diversity at the single cell level. Proc Natl Acad Sci U S A, 112:7285–7290.

42.

Mohammadnia

, Yaqubi

and Wee

. (2018). Common gene expression patterns in the generation of induced neurons from fibroblasts. J Cell Biochem, 119:237–239.

43.

Armstrong

, Dunn

, Antalffy

and Trivedi

. (1995). Selective dendritic alterations in the cortex of Rett syndrome. J Neuropathol Exp Neurol, 54:195–201.

44.

Chao

H-T

, Zoghbi

and Rosenmund

. (2007). MeCP2 controls excitatory synaptic strength by regulating glutamatergic synapse number. Neuron, 56:58–65.

45.

Chang

, Khare

, Dani

, Nelson

and Jaenisch

. (2006). The disease progression of Mecp2 mutant mice is affected by the level of BDNF expression. Neuron, 49:341–348.

46.

Abuhatzira

, Makedonski

, Kaufman

, Razin

and Shemer

. (2007). MeCP2 deficiency in the brain decreases BDNF levels by REST/CoREST-mediated repression and increases TRKB production. Epigenetics, 2:214–222.

47.

Chen

, Chang

, Lin

, Meissner

, West

, Griffith

, Jaenisch

and Greenberg

. (2003). Derepression of BDNF transcription involves calcium-dependent phosphorylation of MeCP2. Science, 302:885–889.

48.

Haase

, Singh

, Gloss

, Tam

and Gold

. (2022). Meta-analysis identifies BDNF and novel common genes differently altered in cross-species models of Rett syndrome. Int J Mol Sci, 23:11125.

49.

Meng

, Wang

, Lu

, He

L-J

, Chen

, Chao

, Fiorotto

, Tang

, Herrera

, et al. (2016). Manipulations of MeCP2 in glutamatergic neurons highlight their contributions to Rett and other neurological disorders. ELife, 5:e14199.

50.

Odabasi

, Yanasik

, Saglam-Metiner

, Kaymaz

and Yesil-Celiktas

. (2023). Comprehensive transcriptomic investigation of Rett syndrome reveals increasing complexity trends from induced pluripotent stem cells to neurons with implications for enriched pathways. ACS Omega, 8:44148–44162.

51.

Gogliotti

, Fisher

, Stansley

, Jones

, Lindsley

, Conn

and Niswender

. (2018). Total RNA sequencing of Rett syndrome autopsy samples identifies the M4 muscarinic receptor as a novel therapeutic target. J Pharmacol Exp Ther, 365:291–300.

52.

Wan

, Zhao

, Lee

and Francke

. (2001). MECP2 truncating mutations cause histone H4 hyperacetylation in Rett syndrome. Hum Mol Genet, 10:1085–1092.

53.

Steinkellner

, Kempaiah

, Beribisky

, Pferschy

, Etzler

, Huber

, Sarne

, Neuhaus

, Kuttke

, et al. (2022). TAT-MeCP2 protein variants rescue disease phenotypes in human and mouse models of Rett syndrome. Int J Biol Macromol, 209:972–983.

54.

Shahbazian

, Young

, Yuva-Paylor

, Spencer

, Antalffy

, Noebels

, Armstrong

, Paylor

and Zoghbi

. (2002). Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron, 35:243–254.

55.

Pound

and Bracken

. (2014). Is animal research sufficiently evidence based to be a cornerstone of biomedical research?. BMJ, 348:g3387.

56.

Perlman RL. (2016). Mouse models of human disease: an evolutionary perspective. Evol Med Public Health, 2016:170–176.

57.

Y-J

, Liu

P-P

, Yan

Z-Z

, Mi

T-W

, Wang

Y-Y

, Li

, Teng

Z-Q

and Liu

C-M

. (2022). KW-2449 and VPA exert therapeutic effects on human neurons and cerebral organoids derived from MECP2-null hESCs. Stem Cell Res Ther, 13:534.

58.

Perego

, Alari

, Pietra

, Lamperti

, Vimercati

, Camporeale

, Garzo

, Cogliati

, Milani

, et al. (2022). Modeling RTT syndrome by iPSC-derived neurons from male and female patients with heterogeneously severe hot-spot MECP2 variants. Int J Mol Sci, 23:14491.

59.

Mertens

, Herdy

, Traxler

, Schafer

, Schlachetzki

JCM

, Böhnke

, Reid

, Lee

, Zangwill

, et al. (2021). Age-dependent instability of mature neuronal fate in induced neurons from Alzheimer's patients. Cell Stem Cell, 28:1533.e6–1548.e6.

60.

Good

, Vincent

and Ausió

. (2021). Mecp2: the genetic driver of rett syndrome epigenetics. Front Genet, 12:620859.

61.

Olson

, Zachariah

, Ezeonwuka

, Liyanage

VRB

and Rastegar

. (2014). Brain region-specific expression of MeCP2 isoforms correlates with DNA methylation within Mecp2 regulatory elements. PLoS One, 9:e90645.

62.

Liang

, Xin

, Wang

and Wu

. (2023). An insertion mutation of the MECP2 gene in severe neonatal encephalopathy and ocular and oropharyngeal dyskinesia: a case report. BMC Med Genomics, 16:181.

63.

Andoh-Noda

, Akamatsu

, Miyake

, Matsumoto

, Yamaguchi

, Sanosaka

, Okada

, Kobayashi

, Ohyama

, et al. (2015). Differentiation of multipotent neural stem cells derived from Rett syndrome patients is biased toward the astrocytic lineage. Mol Brain, 8:31.

64.

Nerli

, Roggero

, Baj

and Tongiorgi

. (2020). In vitro modeling of dendritic atrophy in Rett syndrome: determinants for phenotypic drug screening in neurodevelopmental disorders. Sci Rep, 10:2491.

65.

Jentarra

, Olfers

, Rice

, Srivastava

, Homanics

, Blue

, Naidu

and Narayanan

. (2010). Abnormalities of cell packing density and dendritic complexity in the MeCP2 A140V mouse model of Rett syndrome/X-linked mental retardation. BMC Neurosci, 11:19.

66.

Barnabé-Heider

and Miller

. (2003). Endogenously produced neurotrophins regulate survival and differentiation of cortical progenitors via distinct signaling pathways. J Neurosci, 23:5149–5160.

67.

Colantuoni

, Jeon

, Hyder

, Chenchik

, Khimani

, Narayanan

, Hoffman

, Kaufmann

, Naidu

and Pevsner

. (2001). Gene expression profiling in postmortem Rett Syndrome brain: differential gene expression and patient classification. Neurobiol Dis, 8:847–865.

68.

Eastwood

, McDonald

, Burnet

PWJ

, Beckwith

, Kerwin

and Harrison

. (1995). Decreased expression of mRNAs encoding non-NMDA glutamate receptors GluRl and GluR2 in medial temporal lobe neurons in schizophrenia. Mol Brain Res, 29:211–223.

69.

Motazacker

, Rost

, Hucho

, Garshasbi

, Kahrizi

, Ullmann

, Abedini

, Nieh

, Amini

, et al. (2007). A defect in the ionotropic glutamate receptor 6 gene (GRIK2) is associated with autosomal recessive mental retardation. Am J Hum Genet, 81:792–798.

70.

Zhang

, Cattoglio

, Zoltek

, Vetralla

, Mozumdar

and Schepartz

. (2023). Dose-dependent nuclear delivery and transcriptional repression with a cell-penetrant MeCP2. ACS Cent Sci, 9:277–288.

71.

Steinkellner

, Schönegger

, Etzler

, Kempaiah

, Huber

, Hahn

, Rose

, Duerr

, Christodoulou

, et al. (2019). An electrochemiluminescence based assay for quantitative detection of endogenous and exogenously applied MeCP2 protein variants. Sci Rep, 9:7929.

72.

Qian

, Guan

, Xie

, Xu

, Niu

, Tang

, Li

, Colecraft

, Jaenisch

and Liu

. (2023). Multiplex epigenome editing of MECP2 to rescue Rett syndrome neurons. Sci Transl Med, 15:eadd4666.

73.

Rastegar

, Hotta

, Pasceri

, Makarem

, Cheung

AYL

, Elliott

, Park

, Adachi

, Jones

, et al. (2009). MECP2 isoform-specific vectors with regulated expression for Rett syndrome gene therapy. PLoS One, 4:e6810.

74.

Gadalla

KKE

, Bailey

MES

, Spike

, Ross

, Woodard

, Kalburgi

, Bachaboina

, Deng

, West

, et al. (2013). Improved survival and reduced phenotypic severity following AAV9/MECP2 gene transfer to neonatal and juvenile male Mecp2 knockout mice. Mol Ther, 21:18–30.

75.

Sinnett

, Boyle

, Lyons

and Gray

. (2021). Engineered microRNA-based regulatory element permits safe high-dose miniMECP2 gene therapy in Rett mice. Brain, 144:3005–3019.

76.

Mertens

, Paquola

ACM

, Ku

, Hatch

, Böhnke

, Ladjevardi

, McGrath

, Campbell

, Lee

, et al. (2015). Directly reprogrammed human neurons retain aging-associated transcriptomic signatures and reveal age-related nucleocytoplasmic defects. Cell Stem Cell, 17:705–718.

77.

Herdy

, Schafer

, Kim

, Ansari

, Zangwill

, Ku

, Paquola

, Lee

, Mertens

and Gage

. (2019). Chemical modulation of transcriptionally enriched signaling pathways to optimize the conversion of fibroblasts into neurons. ELife, 8:e41356.

78.

Monk

, Lee

, Jones

and Connor

. (2021). Directly reprogrammed Huntington's disease neural precursor cells generate striatal neurons exhibiting aggregates and impaired neuronal maturation. Stem Cells, 39:1410–1422.

79.

Trujillo

, Adams

, Negraes

, Carromeu

, Tejwani

, Acab

, Tsuda

, Thomas

, Sodhi

, et al. (2021). Pharmacological reversal of synaptic and network pathology in human MECP2-KO neurons and cortical organoids. EMBO Mol Med, 13:e12523.