Function and Immunogenicity of Gene-corrected iPSC-derived Hepatocyte-Like Cells in Restoring Low Density Lipoprotein Uptake in Homozygous Familial Hypercholesterolemia

Gene correction of induced pluripotent stem cells (iPSCs) has therapeutic potential for treating homozygous familial hypercholesterolemia (HoFH) associated with low-density lipoprotein (LDL) receptor (LDLR) dysfunction. However, few data exist regarding the functional recovery and immunogenicity of LDLR gene-corrected iPSC-derived hepatocyte-like cells (HLCs) obtained from an HoFH patient. Therefore, we generated iPSC-derived HLCs from an HoFH patient harbouring a point mutation (NM_000527.4:c.901 G > T) in exon 6 of LDLR, and examined their function and immunogenicity. From the patient’s iPSCs, one homozygous gene-corrected HoFH-iPSC clone and two heterozygous clones were generated using the CRISPR/Cas9 method. Both types of iPSC-derived HLCs showed recovery of the function of LDL uptake in immunofluorescence staining analysis. Furthermore, these gene-corrected iPSC-derived HLCs showed little immunogenicity against the patient’s peripheral blood mononuclear cells in a cell-mediated cytotoxicity assay. These results demonstrate that LDL uptake of iPSC-derived HLCs from HoFH can be restored by gene correction without the appearance of further immunogenicity, suggesting that gene-corrected iPSC-derived HLCs are applicable to the treatment of HoFH.

. Gene correction of iPSCs using the CRISPR/Cas9 method. (A) Schematic overview of the genetargeting strategy for the human LDLR mutation. After introduction of a double-strand DNA break at LDLR exon 6 near the point mutation, a normal amino acid sequence along with a neomycin-selection cassette flanked by two loxP sites was inserted using a homologous recombination vector. Finally, after selection and sequencing of the targeted vector, the selection cassette was excised by transient Cre expression. (B) An sgRNA with a 23-base pair target sequence corresponding to bases adjacent to the mutation site in exon 6 of human LDLR was designed. A donor sequence, containing a T > G correction for the point mutation, was used as a template for the homology-directed repair process induced by Cas9 cleavage. Blue characters indicate silent mutations. SV40, simian virus; Neo, neomycin resistance gene; PGK phosphoglycerine kinase; DT-A diphtheria toxin A; PAM, protospacer adjacent motif. (C) Genomic sequencing showing retention of the LDLR mutation in the HoFH-iPSC line and correction of the target sequence in the gcHoFH-iPSC lines (arrows). Wild-type-derived iPSCs (WT-iPSCs), homozygous FH-derived iPSCs (HoFH-iPSCs), homozygous gene-corrected HoFH-iPSCs (gcHoFH +/+ -iPSCs), and heterozygous gene-corrected HoFH-iPSCs (gcHoFH +/− -iPSCs).
Western blotting detected the mature form of LDLR (130 kDa) in all lines of HLCs, particularly in the presence of 5 µM rosuvastatin (Wako Chemicals, Osaka, Japan) (Fig. 2D). By contrast, the immature form of LDLR (85 kDa) was detected in HoFH-HLCs and gcHoFH +/− -HLCs. Quantitative evaluation of LDLR protein by western blotting showed that the mature and immature forms of LDLR were not significantly different in all cell lines (Fig. 2E). On the other hand, the immature form was present in significantly larger amounts in HoFH-HLCs than in WT-HLCs and gcHoFH +/+ -HLCs (Fig. 2F).

Discussion
Here, we demonstrated that iPSCs could be generated from lymphocytes of a patient with HoFH and differentiated into HLCs with characteristics of the lipid metabolic disorder of the patient. Gene correction using the CRISPR/Cas9 method restored LDL uptake ability in HoFH-HLCs to levels similar to those of WT-HLCs, and HoFH-HLCs and gcHoFH +/+ -HLCs showed little immunogenicity against the patient's PBMCs in vitro.
The most common cause of the HoFH phenotype is a mutation in both LDLR alleles 3 . The patient studied here harboured a point mutation in exon 6 of both LDLR alleles. Although exon 6 encodes the ligand-binding domain of LDLR 26 , there are no reports on the mutation classes of the c.901 G > T point mutation. We successfully generated iPSCs from this patient and obtained iPSC-derived HLCs. Consequently, the iPSC-derived HLCs expressed an LDLR with impaired LDL uptake. Previously, we reported that the LDLR activity of fibroblasts derived from this patient was approximately one-tenth of that of fibroblasts from a healthy subject 27 . These findings are consistent with the LDL uptake of the iPSC-derived HLCs presented here. LDLR protein is usually glycosylated in the endoplasmic reticulum and transported to the Golgi, where it is converted into mature LDLR, and immature LDLR proteins are thought to be insufficiently modified. In present study, western blotting showed that HoFH-HLCs expressed both mature and immature LDLR proteins, which suggest that c.901 G > T is a class IIB mutation 28 .
Using the single CRISPR/Cas9 transfection method 29 , we generated not only gcHoFH +/− -iPSCs but also gcHoFH +/+ -iPSCs that expressed mature LDLR with LDL uptake function, although the gcHoFH +/+ -iPSC clone was determined to be heterozygous for all intentionally inserted silent mutations. A possible explanation is that www.nature.com/scientificreports www.nature.com/scientificreports/ the homologous recombination-mediated knock-in occurred only in a small range of donor sites, as observed in another study 25 . One-base substitution to the wild-type sequence with nonhomologous end joining might occur in one allele, although the precise mechanism of this phenomenon is still unclear. www.nature.com/scientificreports www.nature.com/scientificreports/ One of the advantages of the present study is that the iPSC-derived HLCs were obtained not from patient fibroblasts but from peripheral lymphocytes, which might be less invasive than previous procedures to analyse their function and immunogenicity. Under our conditions, the capacity of LDL uptake after homozygous gene  www.nature.com/scientificreports www.nature.com/scientificreports/ correction seemed to be similar to that of WT-HLCs. In addition, it is interesting that the patient's iPSC-derived HLCs after gene correction exhibited little immunogenicity, similar to that observed before gene correction, whereas WT-HLCs from the healthy volunteer showed significant immunogenicity against the patient's PBMCs. It is intriguing to point out that an auto-immune response was similarly not observed after hepatic cell transplantation with gene therapy 30 . This is the first study to perform a cell-mediated cytotoxicity assay for investigating the utility of gene-corrected iPSCs as autologous transplanted cells in vitro. Changes in the LDLR steric structure caused by gene correction may not have affected overall immunogenicity in either a previous study 30 or the present study. This suggests that the gcHoFH-HLCs described here are suitable for transplantation without aggressive immune suppressive therapy.
The present study has several limitations. First, we did not quantify the LDL uptake ability of the gcHoFH-HLCs. Therefore, the extent to which iPSC-derived HLCs can assimilate LDL as compared with human hepatocytes remains unclear. However, the fact that gcHoFH +/+ -HLCs showed LDL uptake equal to that in WT-HLCs suggests the functional restoration of these cells. Second, we did not further evaluate the immune response against the newly-developed LDLR in gcHoFH-HLCs, although we confirmed little immune response by human leukocyte antigen mismatch. However, as observed in our study, total cell damage might be minimized when the patient's iPSC-derived HLCs are used. Finally, we did not investigate whether the gene correction affected insertional oncogenesis. A previous study demonstrated few possibilities of insertional mutagenesis in oncogenes using CRISPR/Cas9 gene correction 31 . However, this might vary with the type of transplanted cell, insertion location, and method. Further in vivo studies are necessary to address these issues.

Conclusion
LDL uptake ability of iPSC-derived HLCs can be restored by CRISPR/Cas9-mediated gene correction without altering immunogenicity. We suggest that gene correction using the CRISPR/Cas9 system can be used to treat HoFH patients with LDLR dysfunction.

Material and Methods
Generation of ipsCs from the HoFH patient's t Cells. This study was reviewed and approved by the Research Ethics Committee of Kanazawa University (Kanazawa, Japan), and written informed consent was obtained from the patient and the volunteer. All procedures were conducted in accordance with the Helsinki Declaration of 1996. The animal study was approved by the Animal Care and Use Committee of Kanazawa University. Use of animals was conducted according to the "Basic Guidelines for Conduct of Animal Experiments" published by the Ministry of Health, Labor, and Welfare, Japan.
T cells were harvested from a 66-year-old male patient with HoFH and a healthy volunteer. The patient's profile has been reported elsewhere 27 . This patient has a point mutation (NM_000527.4:c.901 G > T) in exon 6 of LDLR. From these T cells, iPSCs were generated using a published protocol 32 . The pluripotent potential of iPSCs was assessed by their ability to spontaneously differentiate into three germ layers (endoderm, mesoderm, and ectoderm) in vitro as described previously 33 . The generated iPSCs were maintained on mitomycin C-inactivated mouse embryonic fibroblasts (MEFs) in Dulbecco's modified Eagle's medium/nutrient mixture F-12 Ham (DMEM/F12, Sigma-Aldrich, MO, USA) supplemented with 1 mM L-glutamine (Invitrogen, CA, USA), 1 mM non-essential amino acids (NEAAs) (Sigma-Aldrich), 1% penicillin/streptomycin (pen/strep), 20% knockout serum replacement (Invitrogen), 0.1 mM β-mercaptoethanol (Sigma-Aldrich), and 4 ng/ml of basic fibroblast growth factor (bFGF; Wako Chemicals). The iPSC culture medium was changed every 2 days, and the cells were passaged using 1 mg/ml of collagenase IV (Invitrogen) every 7 days. On day 4, the cells were differentiated in Roswell Park Memorial Institute (RPMI)-1640 medium containing GlutaMAX, 2% B-27 supplement containing insulin (Invitrogen), 1% NEAAs, 1% pen/strep, 100 ng/ml of activin A, and 80 ng/ml of bFGF. On days 5-7, the cells were differentiated in RPMI medium supplemented with 50 ng/ml of activin A. On days 8-12, the cells were differentiated in RPMI-B27 medium containing 20 ng/ml of BMP4 and 10 ng/ml of FGF10 (Autogen Bioclear, Nottingham, UK). After day 12, the cells were differentiated in hepatocyte basal medium (HBM; Lonza, Basel, Switzerland) supplemented with 30 ng/ml oncostatin M (R&D Systems) and 50 ng/ml of hepatocyte growth factor (PeproTech, NJ, USA) with a medium change every other day. iPSC-derived HLCs were treated for 24 h with 5 µM rosuvastatin or vehicle (0.025% dimethyl sulfoxide) before RNA isolation or protein extraction.

Differentiation of iPSCs into
Genomic sequence of ipsCs. Genomic DNA was isolated from iPSCs using a Gentra Puregene Cell kit (Qiagen, Venlo, Netherlands), and the region encoding exon 6 of LDLR was amplified by PCR. The PCR products were electrophoresed on a 2% agarose gel and purified using a Wizard SV gel and PCR clean-up system www.nature.com/scientificreports www.nature.com/scientificreports/ (Promega). The purified PCR product was sequenced using a BigDye ® Terminator (v1.1 or v3.1) (Applied Biosystems, CA, USA) and an Applied Biosystems 3130 Genetic Analyzer (Applied Biosystems).
short tandem repeats-pCR analysis. Short tandem repeats (STRs) are tandemly repeated simple sequences of a few bases in length that vary in number of repeat units among alleles from different individuals 35 . STR-PCR analysis was performed to facilitate individual identification. Briefly, PCR amplification was performed for the STR in the D12S391 locus using AmpliTaq Gold (Thermo Fisher Scientific). The primer sequences were forward, AACAGGATCAATGGATGCAT, and reverse, TGGCTTTTAGACCTGGACTG. The forward primers were labelled at the 5′ end with 6-carboxyfluorescein. Amplification was performed in a thermal cycler (ASTEC, Fukuoka, Japan) for 30 cycles. The PCR products (1 μl) were mixed with 0.3 μl of size standards (GeneScan 500LIZ dye Size Standard; Thermo Fisher Scientific), 23.5 μl of Hi-Di formamide (Thermo Fisher Scientific), and 1.2 of μl nuclease-free water. DNA bands were detected using an ABI PRISM 310, and peak patterns were visualized using GeneMapper software (Applied Biosystems).
Gene expression analysis. Total RNA was isolated from cells using an RNAeasy kit (Qiagen). For real-time PCR analysis, cDNA was synthesized using SuperScript III reverse transcriptase and oligo (dT) primers (Invitrogen). Primers used for amplification are listed in Supplemental Table 1 , rabbit IgG polyclonal isotype control (Thermo Fisher Scientific) for A1AT, mouse IgG2a isotype control (R&D Systems) for ALB, rabbit IgG monoclonal isotype control (abcam) for LDLR, and mouse IgG2b isotype control (Thermo Fisher Scientific) for ASGPR1. In the ER localization experiment only, cells were treated for 24 h with a green fluorescent protein fused to an ER retention signal (CellLight ER-GFP BacMam 2.0, Invitrogen) before immunostaining according to the manufacturer's instructions.
Fluorescence-labelled LDL uptake assay. The immunofluorescence LDL uptake assay was performed according to the modified protocol of Cayo et al. 18 . iPSC-derived HLCs were washed three times with ice-cold PBS, and then incubated in ice-cold HBM containing 5 µg/ml of BODIPY ® FL LDL (Invitrogen) at 37 °C for 4 to 6 h. After termination of the assay, iPSC-derived HLCs were washed with PBS and treated as described below for flow cytometry. Fixation and permeabilization of iPSC-derived HLCs for imaging were performed as described above. Cells were incubated for 1 h at room temperature with ASGPR1 antibody diluted in 1% BSA in PBS (1:100; Santa Cruz Biotechnology). The cells were washed twice with PBS for 5 min each, and then incubated for 1 h at room temperature with secondary antibody (Alexa Fluor 568; 1:1000; Invitrogen) diluted in 1% BSA in PBS. Cells were visualized using a fluorescence microscope (BZ-9000, Keyence).
Flow cytometry-based LDL uptake assay. iPSC-derived HLCs were treated for 24 h with 5 µM rosuvastatin before the flow cytometry-based LDL uptake assay was performed. iPSC-derived HLCs were dissociated using TrypLE Select (Invitrogen), washed with 1% BSA in PBS, and labelled with primary antibody (ASGPR1; 1:50; Santa Cruz Biotechnology) for 1 h at room temperature. Next, the cells were washed twice with PBS for 5 min each and then incubated for 1 h at room temperature with secondary antibodies (Alexa Fluor 647; 1:500; Invitrogen) diluted in 1% BSA in PBS. Experiments were performed using a Guava EasyCyte Mini Flow Cytometer (Millipore, Billerica, MA, USA). Data analysis was performed using FlowJo software (Tree Star, Inc., Ashland, OR).
Western blotting. Proteins were extracted using Pierce radioimmunoprecipitation assay buffer (Thermo Gene correction of iPSCs using CRISPR/Cas9. Gene correction was performed according to the method of Li et al. 29 with some modifications. To correct the LDLR mutation in HoFH-iPSCs, we constructed a donor www.nature.com/scientificreports www.nature.com/scientificreports/ template vector using pMA-RQ (ampR) (Thermo Fisher Scientific) in order to replace the mutated sequence with a wild-type sequence harbouring some silent mutations encoding a normal amino acid sequence and a neomycin-selection cassette flanked by two loxP sites (Fig. 1A). We designed a single guide RNA (sgRNA) with a 23-base pair target sequence corresponding to the bases adjacent to the mutation site in exon 6 of human LDLR (underline in Fig. 1B shows the sgRNA binding site) to produce a Cas9-induced double-strand break near the mutation on the LDLR allele using the CRISPR design tool (Thermo Fisher Scientific). We obtained the sgRNA from the template DNA using the TranscriptAid T7 High Yield transcription kit (Thermo Fisher Scientific). A repair oligodeoxynucleotide, containing a T > G correction for the point mutation, was used as a template for the homology-directed repair process induced by Cas9 cleavage (Fig. 1B).
Cells were plated on a 100-mm dish with neomycin-resistant MEFs (Cell Biolabs, CA, USA) as feeder cells in the presence of 10 µM Y-27632. G418 (50 µg/ml; Nacalai Tesque, Kyoto, Japan) selection was applied after iPSC colonies appeared 48 h post-transfection. The resulting neomycin-resistant colonies were dissociated into single cells and plated at 200-500 cells per 100-mm dish with MEFs. Each subclone was screened by genomic PCR, followed by DNA sequencing. After establishing the single-copy knock-in clones, we electroporated the cells with 10 µg of Cre expression vector pCAG-Cre:GFP, a gift from Connie Cepko (Addgene plasmid # 13776, MA, USA), using a NEPA 21 electroporator as described above.
All iPSC-derived HLCs were loaded with 10 ng/ml of interferon gamma (IFNγ) and incubated overnight at 37 °C in HBM supplemented with 30 ng/ml of oncostatin M and 50 ng/ml of hepatocyte growth factor. For labelling, PBMCs were incubated in PBS with 2 µM CellTrace Violet (Invitrogen) for 4 min at 37 °C. After co-culturing of iPSC-derived HLCs with labelled PBMCs for 24 h in HBM supplemented with 10 ng/ml of IFNγ and 0.5 ng/ml of interleukin 2 (BioLegend), the mixture of floating and attached cells was collected using TrypLE select. The cells were treated with propidium iodide (PI) (Invitrogen). Cell number was counted by flow cytometry using a BD FACS Canto II system (BD Biosciences, NJ, USA). We evaluated HLCs except for PBMCs using CellTrace Violet, and identified PI-positive HLCs as dead cells. Percent cell death represents the number of dead cells divided by the total number of HLCs evaluated. We calculated the mean value of percent cell death in HLC lines that were not co-cultured, and then calculated delta cell death as the difference between percent cell death in each HLC line and this mean value. Data analysis was performed using FlowJo software. statistical analysis. mRNA levels were determined and cell-mediated cytotoxicity assay was performed with three independent biological samples, and the results were expressed as mean ± standard error (SE). Protein expression levels were determined with five independent biological samples, and the results were expressed as mean ± standard error (SE). Statistical analysis was performed using analysis of variance followed by Dunnett's test for independent samples. Cell-mediated cytotoxicity assay was performed using Tukey-Kramer's test. The mean difference was considered significant at the p < 0.05 level.