Generation of a Well-Characterized Homozygous Chromodomain-Helicase-DNA-Binding Protein 4G1003D Mutant hESC Line Using CRISPR/eCas9 (ULIEGEe001-A-1)

The chromatin remodeler Chromodomain-helicase-DNA-binding protein 4 (CHD4) is crucial for the development of multiple organ systems. Functional mutations of CHD4 have recently been described in a developmental disorder, namely Siffrim-Hitz-Weiss syndrome (SIHIWES). Herein, we have generated a homozygous CHD4G1003D hESC line (WAe025-A-1) using CRISPR/eCas9-based gene editing in the WA-25 hESC line. The edited hESC line maintains normal karyotype, pluripotency, and ability to differentiate into three germ layers. This cell line will be a valuable resource for studying the functional role of CHD4 during the development and disease modeling of SIHIWES in vitro.


Introduction
The ATP-dependent chromatin remodeler CHD4 is a core component of the nucleosome remodeling and histone deacetylation (NuRD) complex and an essential epigenetic factor to regulate spatiotemporal gene expression during development [1,2]. The ATPdependent chromatin remodeler proteins are implicated in many developmental disorders, including hearing loss [2,3]. Normal CHD4 activity was found to be essential for maintaining stem cell pluripotency [4], differentiation capacity towards neuronal lineage [5], T cell development and hematopoietic stem cell [6,7], cardiac development [8,9], renal progenitor cell self-renewal, and differentiation [10]. It has also been shown that CHD4 can regulate gene expression for maintaining a distinct cellular state and pledge for accurate cell fate decisions upon developmental/external cues [11]. CHD4 was shown to be very important during mammalian embryogenesis as it helps in the formation of functional trophectoderm and lineage commitment of unspecified cells at the preimplantation stage by regulating the frequency of lineage-specification gene expression [12]. Recently, de novo mutations in CHD4 have been implicated in a distinct developmental disorder, 'Sifrim-Hitz-Weiss' (SIHIWES) syndrome [13]. De novo mutations in CHD4 were also found in syndromic and non syndromic congenital heart defect patients and moyamoya angiopathy [14,15]. To study the role of wild-type and mutant CHD4 protein in development and disease, we have generated a mutant CHD4 (c.3008G>A; p. G1003D) hESC line and performed functional characterization. This CHD4 G1003D/G1003D hESC model would provide ample opportunity to differentiate towards targeted cell types in 2D and 3D organoids and study the functional deficit of the specific CHD4 mutant version in cardiac, retinal, neuronal, skeletal, gonadal, and auditory system development.
Here, we have targeted exon 20 of CHD4, specifically at c. 3008G through CRISPR/ eSpCas9 system and allowed the precise insertion of c. 3008A by favoring the homologous recombination (HR) mediated DNA repair ( Figure 1A, the single guide RNA (sgRNA) target sequence is indicated in red, followed by the protospacer adjacent motif (PAM) sequence in blue; the c.3008G is in bold, and the silent mutation to alter the PAM site is underlined; the two single nucleotide polymorphism (SNP) markers flanking the desired insertion site are indicated by blue color). This in vitro hESC model can help study the role of CHD4 mutation during the development of the 'Sifrim-Hitz-Weiss' (SIHIWES) syndrome. The targeted differentiated cell types and organoids for cardiac, renal, ocular, and neuronal systems will provide a unique opportunity to study the functional role of CHD4 in the development and related disorders. The gRNA target sequence is indicated in red color followed by the PAM sequence in blue; the c.3008G is in bold, and the silent mutation to alter the PAM site is underlined; the two SNP markers flanking the desired insertion site are indicated by blue color. (B) Sanger sequence chromatogram of wild type WA25 hESC and edited C3 hESC line. Highlighted in blue is the position of the desired base change (c.3008G>A). (C) Western blot images of wild type CHD4 and mutant CHD4 (p. G1003D) proteins (upper panel) extracted from WA25 hESC and C3 line. The lower panel is for loading control αtubulin. (D) Karyotype of WA25 (R12C02) and C3 (R02C02) hESC line as visualized through KaryoStudio v1.4 software. (E) Immunofluorescence imaging of pluripotency markers OCT4, NANOG and SOX2 for both WA25 and C3 hESC colonies. Images were captured in Zeiss LSM 980 confocal microscope with 20X objective, Z-stacked with 1.3 µm optical slices. Images were processed in ImageJ software (scale bar = 100 µm). (F) qPCRbased quantification of pluripotency marker gene expression in WA25 and C3 hESC lines, statistical analysis was performed as 2-way ANOVA with Sidak's test, *-p < 0.05, **-p < 0.01, ***-p < 0.001, ****-p < 0.0001, ns-p > 0.05. (G) qPCR based quantitative analysis of three germ layer marker gene expression (Ectoderm: PAX2, PAX6, TFAP2A; Endoderm: GATA3, EOMES, and Mesoderm: MSX1, TBXT) in C3 hESC derived embryoid bodies compared to the C3 pluripotent stages.

Efficient Gene-Editing Generates Multiple Heterozygous and Homozygous Single-Cell Derived Mutant Clones
The WA25 hESCs were nucleofected with sgRNA oligo cloned pLentiCRISPR-E (Addgene #78852) and DNA repair template single-stranded oligonucleotide (ssODN), and subjected to puromycin selection for positively transfected cells. The puromycin-resistant single-cell derived colonies were manually picked up and expanded in two separate batches. We have picked around 160 colonies for further analysis. These colonies are morphologically pluripotent in nature, devoid of any spontaneous differentiation, and well-separated from surrounding colonies. this is confirmed further ( Table 1). Genomic DNA samples were isolated from each colony from one batch and genotyped through PCR, restriction fragment length polymorphism (RFLP) assay, and Sanger sequencing (primer sequences are listed in Table 2). We have identified 14 heterozygous CHD4 c.3008G>A clones and one clone with homozygous insertion of c.3008A, as confirmed by the Sanger sequencing chromatogram ( Figure 1B). That made our CRISPR experiment 9.37% efficient for homology-directed repair (HDR). As we envision to delineate the functional role of CHD4 G1003D homozygous mutant in development and disease, we only further characterize the homozygous mutant clone and banked the heterozygous clones. The CHD4 c.3008G>A (homozygous) cell line [WA25-CHD4 c.3008A/A (C3) and mentioned as C3 henceforth] was expanded for further characterization and validation (Table 1).

C3 Line Maintains Pluripotency and Normal Karyotype
To find out if c.3008G>A mutation has any effect on CHD4 RNA and protein expression, we assessed the CHD4 p. G1003D homozygous mutant protein expression level through immunoblotting along with the wild-type CHD4 protein and found no significant change in protein expression ( Figure 1C). At the RNA level, mutant CHD4 has also comparable expression like wild-type CHD4 mRNA as shown through qPCR analysis ( Figure 1F). The chromosomal integrity of C3 cells was confirmed by digital karyotyping through wholegenome SNP genotyping and copy number variation analysis, which found a normal female karyotype without any chromosomal aberration ( Figure 1D, Supplementary Figure S1). In Figure 1D, the X-chromosome images are not evident as the KaryoMap was obtained from KaryoStudio v1.4 (Illumina). There is a 'software bug' in KaryoStudio v1.4, which prevents the software drawing the X/Y chromosome in the KaryoMap. To circumvent this issue, we have provided Supplementary Figure S1, where the B-allele frequency and LogR ratio for X-chromosome is depicted for both the WA25 and C3 hESC lines. This dataset also showed the absence of a Y chromosome, proving the C3 is a WA25 cell of origin (Supplementary Figure S1). Positive immunostainings (>95%) against standard pluripotency markers SOX2, NANOG, and OCT4, as well as a tightly packed morphology of C3 cells, confirmed the pluripotent characteristics comparable with wild type WA25 hESCs ( Figure 1E, Supplementary Figure S2). RNA Expression levels of pluripotent genes (OCT4, NANOG, SOX2, and DNMT3B) have been analyzed by qRT-PCR ( Figure 1F). The expression of OCT4 and NANOG was found to be comparable to that of WA25. In contrast, SOX2 and DNMT3B expressions were found to be significantly higher in C3 cells. This can be explained by the fact that CHD4 deficiency has previously been found to affect the pluripotency gene expression [4]. SOX2 and DNMT3B genes are direct targets of CHD4, and the p. G1003D mutation could impair the repressive activity of CHD4 on SOX2 and DNMT3B expression.

C3 Line Is Able to Differentiate into Three-Germ Layers
The ability of the C3 line to differentiate into three germ layers was assessed by in vitro embryoid body (EB) formation followed by three germ layers markers expression analysis by qRT-PCR and compared with the expression levels of those markers in the C3 pluripotent stage ( Figure 1G). We have found that the ectodermal markers (PAX2, PAX6, and TFAP2A), endodermal markers (GATA3, EOMES) and mesodermal markers (MSX1, TBXT) significantly upregulated in expression compared to that of C3 pluripotent cell population. The significant fold-change in the expression (2-way ANOVA with Sidak's test, *-p < 0.05, **-p < 0.01, ***-p < 0.005, ****-p < 0.001) of the marker genes confirmed that the C3 hESC line is capable of differentiating into three germ layers confirming its pluripotency characteristic again.

C3 Line Qualified through Several Additional Quality Control Experiments
We have first checked for any undesired off-target editing that happened in C3 line. The 'CCTop' algorithm predicted the top five exonic off-targets were screened for potential off-target editing through targeted PCR followed by Sanger sequencing and they were found to be unaffected in all five genomic sites (Supplementary Figure S4). Luminescencebased mycoplasma testing confirmed that the C3 cell line is free from any mycoplasma infection (Supplementary Figure S3). Cell type identity of C3 and WA-25 hESCs were validated by STR analysis of 33 loci (archived with journal).

Discussion
The CHD4 protein belongs to the SNF2/RAD54 helicase family and is one of the core components of NuRD complex [1,16,17]. CHD4 also belongs to CHD subfamily II and characteristically contains PHD fingers and chromodomains as well as an SNF2-like AT-Pase/helicase core [18]. The PHD and chromodomains were found to bind nucleosome and histone H3 tail to regulate the ATPase and chromatin remodeling activity [19,20]. CHD4 has a ubiquitous expression and primarily acts as a transcriptional repressor [21][22][23]. CHD4 protein activity is involved in certain key developmental pathways like spermatogenesis and gonadal development [24][25][26], neuronal and brain development [27,28], cardiac development and maintenance [8,9,29], and vascular development [30][31][32]. So, it is conceivable that any pathogenic loss-of-function or gain-of-function mutations and functional impairment of CHD4 activity would severely impact the spatiotemporal transcriptional network and normal developmental pathways.
The G1003D mutation was found in an adolescent-onset male patient affected with 'Siffrim-Hitz-Weiss' syndrome with multiple organ systems affected [13]. Phenotypically, the patient has a developmental delay with delayed speech and learning difficulties in school, hypotonia, and mild intellectual disability. The brain MRI identified an enlarged lateral and third ventricle. The patient developed glaucoma and sensorineural hearing loss. His cardiac system is also affected by atrial septal defect, patent ductus arteriosus, and ventricular septal defect. The patient has micropenis with delayed sexual development with decreased gonadotropin level. As other phenotypic features, the patient has a short stature without growth hormone deficiency, hypertelorism, a squared face, widow's peak, skeletal osteopenia, short palpebral fissures with microcornea, small ears with dysmorphic helices, and joint hyperlaxity. A singleton, whole genome sequencing identified only CHD4 c.3008G>A as a candidate pathogenic mutation. It is evident from this study that CHD4 p.G1003D pathogenic mutation affects significantly the normal CHD4 activity during development alongside other pathogenic mutations. In a recent study, Cryo-Electron microscopy-based structural analysis at a 1.3 Å resolution was undertaken to determine the nucleosome-CHD4 complex interaction [33]. Structurally, the G1003 residue is located in ATPase lobe 2 of the CHD4 protein. The G1003D mutation of CHD4 was shown to disrupt the interaction with H3 alpha helix loop [33]. Loss of such interaction or deletion of the loop in closely related protein CHD1 was shown in a loss of chromatin remodeling activity [34], which may stand true for CHD4 also.
With the advent of hESC/hiPSC derived targeted cell type differentiation in 2D and 3D organoid, microfluidic-based organ-on-a-chip/disease-on-a-chip engineering, the understanding of disease mechanism and drug discovery against developmental disorders gained an unprecedented boost [35][36][37]. To complement such efforts, parallel development of appropriate disease models with hESC/hiPSC derived organoids has a pivotal importance [38][39][40]. Here, we have generated a well-characterized clinically relevant CHD4 G1003D homozygous mutant hESC line to contribute to the effort of disease modeling of SIHIWES.

Cell Culture
The WA-25 hESC line was cultured on Vitronectin (Gibco™, # A31804) coated plates (Corning) using mTeSR™ Plus media (STEMCELL™ Technologies, # 100-0276) under standard cell culture conditions (37 • C, humidified air with 5% CO 2 ). Cell colonies were routinely checked for intact pluripotent stage morphologies and the presence of any random differentiation, and if any, mechanically cleaned immediately. Cells were passaged with 0.05 mM EDTA solution at around 80% confluency in a 3-5 cell micro-colony format.

Precise Gene Editing with CRISPR/eCas9-Based HDR Repair
CRISPR sgRNA oligos were designed using the 'CCTop' algorithm (https://crispr.cos. uni-heidelberg.de/ (accessed on 22 May 2023)), targeting the c. 3008 G (NM_001273.3) at exon 20 of the CHD4 gene. Synthesized sgRNA oligos were cloned into pLentiCRISPR-E (Addgene #78852) containing eSpCas9 and a puromycin selection cassette in a lentiviral vector backbone. The sequence-verified successful sgRNA/eSpCas9 clone was purified by EndoFree ® Plasmid Maxi Kit (Qiagen #12362). For the HDR template, a 101 base long ssODN template was designed having the desired base change c. 3008 G>A and a 'codon usage bias' optimized PAM altering c. 2998 C>A silent mutation. Before nucleofection, WA-25 cells were treated with 200 nM Nocodazole (STEMCELL™ Technologies, # 74072) to synchronize the cells in the G2 phase, as the HDR happens more frequently in G2 cell cycle phase, and RevitaCell™ Supplement (Gibco™, # A264450) for 16 h. The treated WA-25 cells were dissociated with Accutase™ solution (STEMCELL™ Technologies, # 07920), and 1 × 10 6 live cells were nucleofected with P3 Primary Cell 4D-Nucleofector ® X Kit (Lonza # V4XP-3024) and Amaxa ® 4D Nucleofector ® (Lonza # AAF-1003X) following optimized CA-137 program and the manufacturer's protocol. A total of 4 µg of sgRNA/eSpCas9 construct and 1 µM of ssODN were nucleofected. Immediately after nucleofection, cells were plated on a vitronectin-coated 6-well plate with a density of 1.66*10 5 cells/well using pre-warmed mTeSR plus media supplemented with Revitacell™. Post 24 h of recovery, cells were subjected to puromycin (Sigma #P8833-10MG) selection for positive transfection using 0.5 µg/mL puromycin and Revitacell™-supplemented mTeSR plus media. To favor the HR-based repair, 1 µM Scr7 (DNA Ligase IV inhibitor/NHEJ inhibitor) was used during puromycin selection. Puromycin selection was continued for 48 h with the everyday change of supplemented media, and then the positively selected single-cell derived colonies were expanded for another 10 days with unsupplemented media. Colonies were manually picked and seeded on 96-well plates and allowed to grow until 70% confluency. Each colony was then passaged to duplicate, where one batch was used for genotyping and the other was maintained in culture.

Genotyping through RFLP and Sanger Sequencing
The genomic DNA of each clone was isolated using 50 µL of QuickExtract™ DNA Extraction Solution (Lucigen # QE09050) following the manufacturer's protocol and PCR was performed with Taq DNA polymerase (Qiagen #201203) using the pre-optimized protocol. PCR amplicons were then subjected to RFLP with Tsp45I restriction enzyme (NEB # R0583L) where the bi-allelic edited product has one cut site to generate 341 and 280 bp long fragments. In contrast, the non-edited product stayed undigested (621 bp). The purified PCR amplicons of positively edited clones were then bi-directionally sequenced using BigDye ® Terminator v3.1 sequencing kit (Applied Biosystems TM ) at the GIGA-Genomics platform (https://www.gigagenomics.uliege.be/cms/c_4164592/en/gigagenomics (accessed on 22 May 2023)). Sequence chromatograms were analyzed by SnapGene Viewer (GSL Biotech) software. Primer and ssODN sequences are provided in Table 2.

Karyotyping and STR Analysis
The chromosomal integrity of edited cells was assessed with digital karyotyping using Illumina Infinium OmniExpress-24 v1.3 Beadchip for whole genome SNP genotyping at the GIGA-Genomics platform. SNP genotyping data were analyzed and visualized by Illumina KaryoStudio Software v1.4. Copy number variation (CNV) was called through B-allele frequency and Smoothed LogR ratio. Karyomap visualization showed no gross chromosomal structural alteration.
The WA-25 and C3 cell lines were authenticated by STR analysis for 33 STR loci. The STR analysis was performed by the Spits group at Reproduction and Genetics Research, Vrije Universiteit Brussels (VUB).

Immunocytochemistry and Western Blot
WA-25 and C3 cells were grown on vitronectin-coated glass coverslips in a 24-well culture plate and fixed with 4% paraformaldehyde (PFA) solution, permeabilized with 0.25% TBS/Triton X-100, blocked by 10% donkey serum (Sigma # D9663-10M) in PBS-T (0.1% Tween 20), and incubated with primary and secondary antibodies (listed in Table 2). Nuclear counterstaining was performed using Hoechst 33342 (CST # 4082S). Images were acquired with Confocal Nikon A1R hybrid resonant microscope and processed and analyzed by ImageJ software.

RNA Isolation and qPCR Analysis
Total RNA from WA-25, C3 hESCs, and EBs was isolated using TRIzol™ Reagent (Invitrogen™ #15596026) following the manufacturer's protocol. 1 µg of total RNA from each sample was subjected to cDNA synthesis using RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific™ # K1622) and random hexamer following the manufacturer's protocol. qPCR was performed using GoTaq ® qPCR master mix (Promega # A6002), 1 µL of 1:10 diluted cDNA samples, and pre-validated qPCR primer pairs on a LightCycler ® 480 System (Roche). The qPCR data was calculated based on the 2 −∆∆Ct method; statistical analysis and graphs were performed using GraphPad Prism v.7 software.

Conclusions
In a nutshell, we have generated a CHD4 homozygous mutant hESC cell line through CRISPR-based gene editing, which preserved normal karyotype and pluripotency characteristics with the ability to differentiate into three germ layers, which could be a valuable resource to study the pathophysiology of SIHIWES syndrome in vitro.