Loss of the fragile X mental retardation protein causes aberrant differentiation in human neural progenitor cells

Fragile X syndrome (FXS) is caused by transcriptional silencing of the FMR1 gene during embryonic development with the consequent loss of the encoded fragile X mental retardation protein (FMRP). The pathological mechanisms of FXS have been extensively studied using the Fmr1-knockout mouse, and the findings suggest important roles for FMRP in synaptic plasticity and proper functioning of neural networks. However, the function of FMRP during early development in the human nervous system remains to be confirmed. Here we describe human neural progenitor cells (NPCs) as a model for studying FMRP functions and FXS pathology. Transcriptome analysis of the NPCs derived from FMR1-knockout human induced pluripotent stem cells (iPSCs) showed altered expression of neural differentiation markers, particularly a marked induction of the astrocyte marker glial fibrillary acidic protein (GFAP). When induced to differentiate, FMRP-deficient neurons continued to express GFAP, and showed less spontaneous calcium bursts than the parental iPSC-derived neurons. Interestingly, the aberrant expression of GFAP and the impaired firing was corrected by treatment with the protein kinase inhibitor LX7101. These findings underscore the modulatory roles of FMRP in human neurogenesis, and further demonstrate that the defective phenotype of FXS could be reversed at least partly by small molecule kinase inhibitors.

. Dots frames indicate fragments presented in the Figure 1E in the main text.

EdU incorporation assay
EdU incorporation assay was performed using Clik-iT EdU Alexa Fluor 488 Imaging kit (Thermo Fisher Scientific) according to the manufacturer' instructions.

Plasmid constructs and lentiviral-mediated transduction
To generate FMRP-expressing lentivirus vector, human FMR1 (NM_001185076.1) was PCR-amplified from a synthetic DNA sequence (Eurofins Genomics, Ebersberg, Germany) and the PCR fragments were then introduced into a BamHI-digested pLVSIN-EF1α Neo vector (Clontech, Palo Alto, CA) using In Fusion Cloning kit (Clontech) according to the manufacturer's instructions. Empty vector was used as control. The PCR primer sets used were as follows: forward primer 5'-GAGCGGCCGCGGATCATGGAGGAGCTGGTGGTG, and reverse primer 5'-CGGTAGAATTGGATCTTAGGGTACTCCATTCACG. Lentivirus containing FMRP-expressing lentivirus vector or empty vector was produced in Lenti-X 293T cells (Clontech) by Lentiviral High Titer Packaging Mix (Clontech) transfected via Lipofectamine 2000 (Thermo Fisher Scientific), and supernatants containing lentivirus were concentrated with Lent-X concentrator (Clontech). Transduction of the lentivirus into iNPC-KO was performed with polybrene (Sigma-Aldrich) according to the manufacturer's instructions.

Screening and compounds treatment
The ReN-WT and KO cells (2 x 10 4 cells/well) were seeded into 96 well plates (#6005558, Perkin Elmer) and kept overnignht in CO 2 incubator. The randomly selected 20,000 compounds (10 μM) were added to the cells, and after 24 h, the cells were fixed with 4% paraformaldehyde at room temperature for 10 min. After PBS washes, these cells were used in immunofluorescence staining analysis.

Optimization of compound concentration
We pre-tested the hit-compounds at various concentrations, and optimized concentration that did not induce cell death in iNPC-KO and ReN-KO cells, followed by performing experiment. In brief, the cells (2 x 10 4 cells/well) were seeded into 96 well plates and kept overnignht in CO 2 incubator. The hit-compounds at various concentrations (20 μM-0.02 μM) were added to the cells, and after 24 h, the cells were fixed with 4% paraformaldehyde at room temperature for 10 min. We performed DAPI staining and detected the number of living cells using CQ1.

Calcium imaging and quantification of the spontaneous firing of cells in iNSs
iNPC lines were dissociated to single cells in StemPro Acctase cell Dissociation Reagent (Thermo Fisher Scientific), and reaggregated using Sumilon PrimeSurface plates (40,000 cells/well) in BrainPhys Neuronal Medium (STEMCELL technologies, Vancouver, British Columbia, Canada) supplemented with N2, B27 and 10 μM DAPT (Sigma-Aldrich). LX7101 or DMSO were present continuously by analysis (day 1-14). After two weeks, Cal-520 (AAT Bioquest, Sunnyvale, CA, USA) was added into medium and then incubated for 2 h at 37 ℃. iNS were washed twice in HHBS, and analysed using CQ1.The activities of neurons in iNS were monitored by calcium imaging for 120 sec. The signal intensities were obtained from the time-lapse calcium imaging data of arbitrary cross-sectional surface of iNS (frame rate: 1 fps). To identify the individual cells in each record, the data were analysed using image software (CQ1 software, Yokogawa). The candidate regions of interest (ROIs) were created by segmentation of pixels based on thresholding the accumulated signal intensities over observation period, which were reviewed manually to ensure correct mapping of the location of cells. Then the activities of individual ROIs were calculated. The baseline intensity (F) was estimated by the intensity of each ROI based on the exponential decay model and ⊿F/F was calculated 1 . The firing is defined as the spike with ⊿F/F > 0.2. The ROIs exhibited at least one firing during the observation period were counted as active cells and the rate of active cells were calculated as indicator of the firing activity.