Generation of Induced Pluripotent Stem Cells from Human Nasal Epithelial Cells Using a Sendai Virus Vector

The generation of induced pluripotent stem cells (iPSCs) by introducing reprogramming factors into somatic cells is a promising method for stem cell therapy in regenerative medicine. Therefore, it is desirable to develop a minimally invasive simple method to create iPSCs. In this study, we generated human nasal epithelial cells (HNECs)-derived iPSCs by gene transduction with Sendai virus (SeV) vectors. HNECs can be obtained from subjects in a noninvasive manner, without anesthesia or biopsy. In addition, SeV carries no risk of altering the host genome, which provides an additional level of safety during generation of human iPSCs. The multiplicity of SeV infection ranged from 3 to 4, and the reprogramming efficiency of HNECs was 0.08–0.10%. iPSCs derived from HNECs had global gene expression profiles and epigenetic states consistent with those of human embryonic stem cells. The ease with which HNECs can be obtained, together with their robust reprogramming characteristics, will provide opportunities to investigate disease pathogenesis and molecular mechanisms in vitro, using cells with particular genotypes.


Introduction
Induced pluripotent stem cells (iPSCs) are generated from somatic cells by transducing them with reprogramming factors [1]. Initially, human dermal fibroblasts were used to derive human iPSCs (hiPSCs) [2,3], and the majority of iPSC research in humans has focused on fibroblasts as a source of somatic cells. However, recent studies have shown that other human somatic cells can be used to generate iPSCs such as those from blood [4,5,6,7], teeth [8], adipose tissues [9], and oral mucosa [10]. Obtaining these cells from the aforementioned sources, except blood, requires biopsy with local anesthesia, making it cumbersome for generating patient-specific stem cells. Additionally, although obtaining blood cells does not require local anesthesia, the rearrangement of the T-cell receptor (TCR) chain genes in T cells and the VDJ region in B cells means that they are not identical to naïve lymphocytes at the genomic level.
In the present study, we generated iPSC cells using human nasal epithelial cells (HNECs). This is a less invasive method to obtain human somatic cells, since neither anesthesia nor biopsy are required. In addition, we used Sendai virus (SeV) vectors to introduce reprogramming factors. SeV is an important respiratory pathogen of rats and mice, and it has been reported that SeV vectors efficiently transduce the respiratory tract of mice as well as humans [11]. Therefore, we speculated that HNECs would be highly amenable to efficient gene transduction with SeV vectors.

Results
Freshly obtained HNECs were maintained on collagen-coated matrix, and they got attached within 4-6 hours, forming small colonies. The HNECs reached confluence within2 weeks ( Figure 1) with typical epithelial morphology. We also confirmed that primary HNECs can be cultured and expanded after cryopreservation in liquid nitrogen.
We first determined the infection efficiency using a SeV vector that expressed green fluorescent protein. HNECs seeded at 1.0610 5 cells per 35-mm dish were infected by green fluorescent protein vectors over a range of multiplicities of infection (MOI, number of viral particles per cell; Figure 2). We determined that a MOI of 3 or 4 was sufficient to induce the transgenes for HNECs.
The scheme for generation of iPS from HNECs is presented in Figure 3. We observed the appearance of colonies with an embryonic stem (ES) cell-like morphology at 20 days after infection of SeV vectors carrying 4 reprogramming factors ( Figure 4A), and reprogramming efficiency was 0.1% at MOI 4, and 0.075% at MOI 3 ( Figure 4B).
To characterize colonies generated from SeV-infected HNECs, we picked a total of 74 colonies, and randomly chose 7 lines (iPS-B2, iPS-B3, iPS-C6, iPS-2B1, iPS-2B6, iPS-3A1, and iPS-4B1). We first performed a high-density single nucleotide polymorphism (SNP) genotyping assay to evaluate structural variations of HNECderived colonies. As shown in Figure S1, chromosomal duplications were observed on chromosome 2p in iPS-2B6 and chromosome 12p in iPS-C6. We excluded these colonies from further analysis. None of the other 5 colonies (iPS-B2, iPS-B3, iPS-2B1, iPS-3A1, and iPS-4B1) harbored duplications or deletions on chromosomes, and the genotype concordance rate in each colony with those of original HNECs was greater than 99.99%. This value was similar to the rate of technical replicates (i.e., concordance rate of the same genomic DNA), showing that colonies were derived from parental HNECs. Then, we examined the gene expression of the reprogramming factors and the expression of SeV vectors in HNEC-derived colonies ( Figure 5). We used a temperature-sensitive mutant SeV vector in these experiments in order to shut off transgenes efficiently by temperature shift [12]. Colonies generated from HNECs showed endogenous OCT4, SOX2, KLF4, and c-MYC gene expression levels that were similar to those of human ES cells, and we confirmed that SeV-derived gene expressions were not detected by reverse transcription-polymerase chain reaction (RT-PCR) using SeVspecific primers ( Figure 5A). Protein expression of pluripotency markers was confirmed by immunofluorescence staining ( Figure 5B and Figure S2). DNA methylation analysis revealed that CpG dinucleotides at the OCT4 and NANOG promoter region in the HNEC-derived colonies were demethylated while those in original HNECs were mostly methylated ( Figure 6). Furthermore, global gene expression pattern in the HNEC-derived cell lines strongly correlated with that of human ES cells (r = 0.99, Figure 7), while correlation of gene expression between HNEC-derived cell lines and parental HNECs was weak (r = 0.87 in iPS-B3 and r = 0.86 in iPS-2B1).
In order to evaluate the differentiation potential of HNECderived iPSCs, we performed in vitro differentiation using HNECderived cell lines. We generated embryoid bodies (EBs), which were spontaneously differentiated for 10,13 days. RT-PCR analysis revealed that these cells were positive for three germ cell markers ( Figure S3), although the gene expression pattern of these cells varied in different HNEC-derived iPS cell lines. Next, in order to investigate in vivo differentiation, we injected HNECderived cell lines into immunocompromised mice and evaluated their ability to form teratomas. In the experiments, we used 4 cell lines (iPS-B2, iPS-B3, iPS-2B1, and iPS-3A2). Histological examination of the teratomas revealed that tissues were from the endoderm, mesoderm, and ectoderm lineages, and we observed neural and epithelial tissues, muscle, cartilage, bone, gut-like structures, and various glandular structures ( Figure S4).

Discussion
In the present study, we demonstrated that efficient reprogramming of HNECs can be achieved without integration of the reprogramming transgenes or a viral genome. iPS cells generated in the present study showed characteristics similar to human ES cells, and had the potential to differentiate into the 3 germinal layers of endoderm, mesoderm, and ectoderm in a teratoma formation assay in vivo.
One of the many benefits that iPS technology offers is the development of robust and personalized models of human disease. iPSCs can be generated from subjects with various genetic diseases [13,14]. Skin fibroblasts are frequently used to generate iPSCs. However, because this requires skin biopsy with local anesthesia, it can be challenging to recruit donors. An alternative is to use blood cells as a source for iPSCs. This has been performed using both peripheral B cells [15], and T cells [5,7], but in this scenario, the genomic region of TCR chain genes in T cells and the VDJ region in B cells are rearranged. Therefore, iPSCs derived from T and B cells can exhibit only 1 pattern of the TCR gene and VDR genomic regions, respectively. CD34-positive cells have also been used to generate iPSCs [6,16,17]. This has a potential advantage over using the mature T and B cells, since CD34-positive cells do not have their genomes rearranged. However, these cells comprise only a small proportion (0.1%) of total peripheral hematopoietic cells, which poses a challenge for the separation of pure CD34positive cells. Our current method using HNECs maintains the genome in the same state as the original somatic cells, does not require complicated methods for culture or a high MOI for gene transduction, and achieves reprogramming efficiencies similar to those found using skin fibroblasts (0.08-0.13% [12]). Because of the minimal invasiveness involved in obtaining these somatic cells and their amenability to reprogramming, using SeV vectors to reprogram HNECs may be a more suitable approach for generating patient-specific iPS cells. This will provide opportunities to investigate disease pathogenesis and molecular mechanisms in vitro, using cells with particular genotypes.

Ethics statement
This study was approved by the Ethical Committee of University of Tsukuba and followed International guidelines (i.e., the Helsinki Declaration). Informed written consent was taken from a tissue sample donor. All animal work was conducted according to university and international guidelines.

Primary cell culture
HNECs were obtained from a healthy female donor by the brushing technique. Nasal brushing was performed by an otolaryngologist using a soft sterile brush (Medical Packaging Camarillo, CA). A brushing was performed in both nostrils by a gentle circular movement. Brushes were immersed in 3 mL of small airway epithelial cells basal medium (SABM) (Lonza, Basel, Switzerland) with small airway epithelial cell growth medium (SAGM) SingleQuots (Lonza, Basel, Switzerland) and 100 U penicillin and 0.1 mg/mL streptomycin (Sigma-Aldrich, St. Louis, MO), followed by a manual shake and pipetting. SABM containing HNECs were seeded on type I collagen-coated 35mm dishes (BD Biocoat, BD, Franklin Lakes, NJ), and cultured at iPS Cells Generated from Nasal Epithelial Cells PLOS ONE | www.plosone.org 37uC in an atmosphere of 5% CO 2 . After 24 and 48 h of culture, non-adherent HNECs were collected and transferred to a new type I collagen-coated dish. SABM was changed every 2 days, and cells were cultured with SABM with SAGM SingleQuots and Reagent Pack (Lonza, Basel, Switzerland).

Reprogramming efficiency
Reprogramming efficiency was calculated as the number of iPS colonies formed per number of infected cells seeded. iPS colonies were identified based on ES cell-like morphology and crystal violet staining. Crystal violet staining was performed with 0.1% crystal iPS Cells Generated from Nasal Epithelial Cells PLOS ONE | www.plosone.org violet (Nakaraitesk, Kyoto, Japan) in methanol, as described previously [18].

SNP genotyping
Genomic DNA was isolated from HNEC-derived iPS cells by DNeasy Blood & Tissue kit (Qiagen, Hilden, Germany). Highdensity SNP genotyping was performed using Human610-Quad (Illumina, San Diego, CA) that contains more than 610,000 rationally selected tag SNPs and markers per sample, according to the manufacturer's protocol. The scanned image was imported into GenomeStudio (Illumina) for analysis. Genotype calls and analysis were performed by GenomeStudio (Illumina).

RT-PCR
Total RNA was isolated using the RNeasy Mini kit (Qiagen), according to the manufacturer's instructions. The concentration and purity of the RNA were determined using the ND-1000 spectrophotometer (Nanodrop, Wilmington, DE). The cDNA was   Table S1.

Global gene expression analysis
Total RNA was isolated from HNEC-derived iPS cells using the RNeasy Mini Kit. We used Illumina Bead Array with single-color array (Illumina) as a microarray platform. For the Illumina BeadArray assay, cRNA was synthesized with an Illumina RNA Amplification kit (Life Technologies), according to the manufacturer's instructions. In brief, 500 ng of total RNA were reverse transcribed to synthesize first-and second-strand cDNA, purified with spin columns, and then in vitro transcribed to synthesize biotin-labeled cRNA. A total of 750 ng biotin-labeled cRNA was hybridized to each Illumina Human-Ref8 v3.0 BeadChip array (Illumina) at 55uC for 18 h. The hybridized BeadChip was washed and labeled with streptavidin-Cy3 (GE Healthcare, Buckinghamshire, UK) and then scanned with the Illumina BeadStation 500 System (Illumina). The scanned image was imported into GenomeStudio software (Illumina) for analysis. Twenty-two thousand transcripts representing 8 whole-genome samples can be analyzed on a single BeadChip. GenomeStudio output of the microarray data was processed with lumi package for the R language [19] on R version 2.10.0 (http://www.R-project.org/).

Bisulfite sequencing
Genomic DNA was isolated from HNECs, human ES cells, and iPS cells derived from HNECs using DNeasy Blood & Tissue kit (Qiagen), and was treated with sodium bisulfite using EpiTect Bisulfite kit (Qiagen), according to the manufacturer's instructions. Converted DNA was used as the template for PCR using primer sets previously described to amplify the promoter regions of OCT4 [20] and NANOG [2]. The purified PCR products were TA-cloned into pCR4-TOPO vector using TOPO TA Cloning Kit for Sequencing (Life Technologies). The insert sequences of randomly picked clones were analyzed using the ABI 3130 DNA analyzer (Life Technologies).

In vitro differentiation
iPS cells were cultured on Petri dishes for one days in hiPSC medium, then were cultured free floating to induce EBs for 10-13 days in EB medium, consisting of DMEM/Ham's F12 contain-ing5% KSR, 2 mML-glutamin, 1610 24 M nonessential amino acids, and 1610 24 M 2-mercaptoethanol.