The Brown Foundation Institute of Molecular Medicine for the Prevention of Human Diseases, University of Texas Medical School at Houston, Houston, Texas, USA
Correspondence: Dachun Wang, M.D., The Brown Foundation Institute of Molecular Medicine for the prevention of Human Diseases University of Texas Medical School at Houston, 1825 Pressler Street/IMM 437D, Houston, Texas 77030, USA. Telephone: +1-713-500-3608; Fax: +1-713-500-2420; e-mail: email@example.com
Human induced pluripotent stem cells (hiPSCs) have great therapeutic potential in repairing defective lung alveoli. However, genetic abnormalities caused by vector integrations and low efficiency in generating hiPSCs, as well as difficulty in obtaining transplantable hiPSC-derived cell types are still major obstacles. Here we report a novel strategy using a single nonviral site-specific targeting vector with a combination of Tet-On inducible gene expression system, Cre/lox P switching gene expression system, and alveolar epithelial type II cell (ATIIC)-specific NeomycinR transgene expression system. With this strategy, a single copy of all of the required transgenes can be specifically knocked into a site immediately downstream of β-2-microglobulin (B2M) gene locus at a high frequency, without causing B2M dysfunction. Thus, the expression of reprogramming factors, Oct4, Sox2, cMyc, and Klf4, can be precisely regulated for efficient reprogramming of somatic cells into random integration-free or genetic mutation-free hiPSCs. The exogenous reprogramming factor transgenes can be subsequently removed after reprogramming by transient expression of Cre recombinase, and the resulting random integration-free and exogenous reprogramming factor-free hiPSCs can be selectively differentiated into a homogenous population of ATIICs. In addition, we show that these hiPSC-derived ATIICs exhibit ultrastructural characteristics and biological functions of normal ATIICs. When transplanted into bleomycin-challenged mice lungs, hiPSC-derived ATIICs efficiently remain and re-epithelialize injured alveoli to restore pulmonary function, preventing lung fibrosis and increasing survival without tumorigenic side effect. This strategy allows for the first time efficient generation of patient-specific ATIICs for possible future clinical applications. Stem Cells2014;32:402–413
Alveolar epithelial type II cells (ATIICs) are essential for maintaining homeostasis of alveoli. They synthesize and secrete pulmonary surfactant, playing a crucial role in reducing surface tension and preventing collapse of alveoli at end expiration. As progenitor cells for alveolar epithelial type I cells (ATICs), ATIICs are particularly crucial during re-epithelialization of the alveoli after lung injury. Loss of normal functions of ATIICs due to injuries or genetic deficiencies is thought to play an important role in the development of many life-threatening pulmonary diseases, including idiopathic pulmonary fibrosis , fatal neonatal respiratory distress syndrome [2, 3], and progressive familial and sporadic interstitial lung disease . Currently available treatments for these diseases at best alleviate symptoms within a limited time range, with poor long-term outcomes. Lung transplantation is the only therapeutic option for many patients. Therefore, there is an urgent need of novel therapeutic strategies. Recently, potential use of induced pluripotent stem cells (iPSCs) to regenerate injured lungs has attracted a lot of interest for disease modeling and repair of injured/diseased alveolar epithelium.
Previous techniques use retrovirus or lentivirus to integrate multiple copies of each reprogramming factor at random into the chromosomal DNA to reprogram somatic cells into iPSCs [5-7]. However, each random integration of viral DNA may cause unpredictable critical gene dysfunction, and the generated iPSCs may develop into tumors or lose their self-renewal capacity or the potential to differentiate into the cell type needed for therapeutic transplantation. Recently, a technique using nonviral vector containing loxP targeting sequence in combination with inducible gene expression system has been developed for iPSC reprogramming . However, random insertion of DNA or part of DNA sequences left behind after removal of factors is still a potential drawback. Several alternative techniques have been developed to improve safety in iPSC generation, including the use of adenovirus , sendai virus , minicircle vector , PiggyBac transposon , and episomal vectors . Yet most of these techniques suffer from impractically low reprogramming efficiency, and the possibility of vector integrations remains. More recently, techniques to directly deliver proteins , RNAs , or mature microRNAs  for reprogramming have been developed but require special treatment and multiple times of transduction/transfection with low reprogramming efficiency. Thus, efficient generation of iPSCs without causing genetic abnormalities continues to be a challenge. In addition, as iPSCs tend to spontaneously differentiate to various cell types in vitro, the spontaneously differentiated cultures of iPSCs are certainly not suitable for clinical application as they could cause deleterious side effects, even tumor formation. Therefore, lack of effective methodologies to derive iPSCs into transplantable cell lineages is also a major hurdle to their clinical application.
To overcome the above problems, here we report a novel β-2-microglobulin (B2M)-specific insertion targeting strategy using a single nonviral vector for reprogramming human somatic cells. By using tetracycline-controlled expression system (Tet-On) to conditionally express Oct4, Sox2, cMyc, and Klf4 immediately downstream of B2M gene for efficient generation of random integration-free human iPSCs (hiPSCs), this site-specific insertion does not cause B2M gene dysfunction. As the single targeting vector contains loxP targeting sequence and NeomycinR (NEOR) transgene controlled by ATIIC-specific surfactant protein C (SPC) promoter (SPCP-NEOR), the reprogramming factor transgenes can be subsequently removed, and the hiPSCs can be selectively differentiated into a homogenous population of ATIICs for further exploration of their therapeutic potential.
Materials and Methods
Construction of 3′hprt.OSMK-LoxP.rtTA.SPCPNEOR.B2M Targeting Vector
One 4.2 kb AscI-BssHII DNA fragment homologous to 3′ region of B2M gene was cloned into AscI site of the 3′ hprt insertion targeting vector (a gift from Dr. Allan Bradley, The Wellcome Trust Sanger Institute, Cambridge, U.K.). The AscI- and SalI-digested TRE-PminCMV/Oct4/IRES/Sox2 fragment was isolated from pTRE-OIS vector (Supporting Information Fig. S1A) and subcloned into the engineered MluI and SalI site downstream of PUROR. Similarly, XhoI-and SalI-digested TRE-PminCMV/Klf4/IRES/cMyc/LoxP fragment from pTRE-KIcML vector (Supporting Information Fig. S1B) was subcloned into SalI site downstream of TRE-PminCMV/Oct4/IRES/Sox2 fragment with correct orientation. A 4.9 kb SPCP-NEOR transgene isolated from SPCP-NEOR vector (Supporting Information Fig. S1C) was subsequently added into SalI and engineered NdeI site of the vector. In addition, EFaP-rtTA transgene was subcloned from EFαP-rtTA vector (Supporting Information Fig. S1D) into AscI site of the insertion targeting vector. The resulting 3′hprt.OSMK-LoxP.rtTA.SPCPNEOR.B2M targeting vector is depicted in Figure 1. The EcoRI was used to linearize the vector and delete a 216 bp gap within Exon 4 of B2M fragment before transfection for gap repair targeting at the B2M gene locus .
Transfection of Human Skin Fibroblasts for Induction of Pluripotency
Approximately 5 × 105 human adult fibroblasts (passages 4, provided by National Disease Research Interchange) were resuspended in 100 µL of supplemented Nucleofector Solution (VPD-1001, Lonza), mixed with 2 µg of the EcoRI-linearized targeting vector, and then transfected following manufacturers' instruction. The transfected cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) (Gibco, Invitrogen, Grand Island, NY, http://www.invitrogen.com). Puromycin (0.5 µg/mL, Sigma) was added next day. After 3-day selection, the stably transfected cells that had survived puromycin selection were replaced at 0.5 × 104 per 10-cm dish on mitomycin-treated mouse embryonic fibroblast (MEF) feeder cells in human embryonic stem cell (hESC) culture medium  in the presence of 2 µg/mL doxycycline (doxy, Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com) .
Southern Blot and Polymerase Chain Reaction Analysis of B2M-Targeted hiPSC Clones
Genomic DNA samples were digested with HindIII, SalI, KpnI, MluI, and ClaI, respectively, for Southern blot analysis. A 216-bp Polymerase chain reaction (PCR) fragment homologous to the gap region within the exon 4 of B2M (Fig. 1) or a 425-bp NEOR gene fragment was used as a probe. In addition, PCR was performed to confirm the site-specific targeting event by using primer A and B (Fig. 1).
Total RNA samples were prepared for real-time quantitative reverse transcriptase-PCR (QRT-PCR) analysis using TaqMan One-Step RT-PCR Master Mix Kit (AB Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com) as previously reported  using 18S as endogenous control. All primers used in the study were listed in Supporting Information Table S1.
Differentiation and Selection of hiPSC-Derived ATIICs
To directly differentiate hiPSC-26B into ATIICs (hiPSC-ATIICs), cells were cultured on Matrigel-coated plate in differentiation medium (DM) with or without G418 (20 µg/mL, Gibco Invitrogen) for 14 days . The DM contains 80% Knockout DMEM (Gibco Invitrogen), 20% FBS (HyClone, Logan, UT, http://www.hyclone.com), 1% nonessential amino acid, 1 mM l-glutamine, 100 µg/mL penicillin, and 100 µg/mL streptomycin.
Differentiated and undifferentiated hiPSC-26B cells were washed twice with phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde for 15 minutes. Washed cells were permeabilized in 0.2% Triton X (Sigma) for 30 minutes before blocking in 10% goat or donkey serum for 2 hours. Differentiated cells were stained with 1:50 diluted mouse anti-human major histocompatibility complex class I (MHC-I) (W6/32, Abcam, Cambridge, U.K., http://www.abcam.com), and visualized with Alexa Fluor 488-conjugated goat anti-mouse IgG (1:1,000, Molecular Probes, Eugene, OR, http://probes.invitrogen.com) with Draq5 (1:500, Biostatus) counterstaining. Undifferentiated cells were stained with either goat anti-human Nanog, goat anti-human Oct4, mouse anti-human SSEA4, mouse anti-SSEA1 (1:10 diluted, R&D Systems, Minneapolis, MN, http://www.rndsystems.com), or 1:500 diluted mouse anti TRA-1–60 (Abcam). Alexa Fluor 546-conjugated donkey anti-goat IgG (1:1,000, Molecular Probes) was used to visualize expression of Nanog and Oct4, and Alexa Fluor 488-conjugated goat anti-mouse IgG to visualize expression of TRA-1–60, SSEA4, and SSEA1. To examine expression of surfactant protein A (SPA), B (SPB), and C, the G418-selected hiPSC-ATIICs on day 14 and human primary control ATIICs (hATIICs) were stained with the rabbit anti-human SPA, proSPB, or proSPC (Chemicon, Temecula, CA, http://www.chemicon.com) as previously reported . The hATIICs were prepared as previously described . The surfactant protein expression was visualized with Alexa Fluor 488- or 546-conjugated goat anti-rabbit IgG with Draq5 counterstaining. In addition, the rabbit anti T1α (Abgent), rabbit anti-human proSPC and mouse anti-human nuclei monoclonal antibody (Chemicon) were used for immunofluorescent staining of bleomycin (BLM)-challenged lungs transplanted with hiPSC-ATIICs as previously reported .
Karyotype Analysis and Teratoma Formation
Chromosomal study using standard protocols for high-resolution G-banding was performed at the Clinical and Research Cytogenetic Laboratory, Texas Children's Hospital. To examine in vivo pluripotency, hiPSC-26B cells were resuspended at 0.5 × 107 in hESC medium. The isoflurane anesthetized severe combined immunodeficiency (SCID) mice were intramuscularly injected with 0.5 × 107 cells on left hind leg. Tumors were surgically dissected from mice 8 weeks after injection for histological analysis.
The G418-selected hiPSC-ATIICs on day 14 were prepared for ultrastructural examination using previously published protocol , which was performed by Electron Microscopy Laboratory, Department of Pathology, University of Medical School at Houston.
Regulated Secretion of Surfactant from Cultured hiPSC-ATIICs
G418-selected hiPSC-ATIICs on day 14 were trypsinized and then seeded back onto fresh Matrigel-coated 10 cm culture plates with DM containing 3H-choline (1 mCi per plate, PerkinElmer) for 24 hours. Cells were then rinsed three times with PBS and cultured for 2 hours at 37°C with or without secretagogue, 12-O-Tetradecanoylphorbol-13-acetate (TPA, 50 ng/mL; Sigma-Aldrich) . The 3H-labeled phosphatidylcholine (PC) in the medium and cells were extracted and counted, respectively . To examine surfactant protein secretion, the selected hiPSC-ATIICs were trypsinized on day 9 and then cultured by using air-liquid interface culture system in small airway epithelial cell growth medium (SAGM, Chemicon Millipore, Billerica, MA, http://www.millipore.com) containing G418, with or without dibutyryl cAMP (Bt2cAMP, 1 mM) + dexamethasone (Dex, 10−10 M) for 5 days . The proteins were harvested from culture medium and analyzed by Western blot using rabbit polyclonal antisera against SPB and SPC.
Transplantation of hiPSC-ATIICs into BLM-Injured Mice Lungs
Pathogen-free, 8–10 week old, female SCID mice on a C57BL/6 genetic background (Jackson Laboratories, Bar Harbor, ME, http://www.jax.org) were exposed to BLM (3.5 units/kg, Bristol-Myers Squibb Company, New York, NY, http://www.bms.com) or sterile normal saline endotracheally via oropharynx intubation. On day 2 after BLM-challenge, mice received saline, human monocytes (hmonos) or G418-selected hiPSC-ATIICs harvested on day 14 (n = 8). All cells were gently washed in sterile normal saline for three times and then resuspended at 1.0 × 106 cells in 50 µL of sterile normal saline for transplantation via an endotracheal catheter . The mice were weighed every another day for 10 days. The spontaneous lung tidal volumes and blood arterial oxygen saturation levels were measured at end time point as previously described . Each group of mice was sacrificed on day 10 after euthanized and lungs were harvested for histological analysis. All procedures were approved by the Animal Welfare Committee, University of Texas Health Science Center at Houston.
Histological Analysis and Hydroxyproline Assay
BLM-challenged lungs with and without transplantation of hiPSC-ATIICs or hmonos were harvested, lavaged, and fixed for embedding as previously described . Tissue sections from three different levels of each lung were stained with hematoxylin-eosin for routine histology. BLM-induced lung damage was graded according to the injured lung area (0, < 10, 10–25, 25–50, 50–75, and >75%) involved with cellular infiltration, interstitial thickening, and structure distortion . Hydroxyproline assay was performed to evaluate content of collagen deposition in the BLM-injured lungs using previously described method .
The data were analyzed with the Student's t test or the χ2 frequency test, and the resulting pairwise p values less than .05 were considered statistically significant.
Efficient Generation of Random Integration-Free and Exogenous Reprogramming Factor-Free hiPSCs
This 3′HPRT.OSMK-LoxP.rtTA.SPCPNEOR.B2M targeting vector (Fig. 1) is designed to target the 3′ region of the B2M gene at a high targeting frequency, without disruption of the B2M function . Of approximately 2.5 × 104 stably transfected fibroblasts seeded on MEF feeder cells, 96 hESC-like hiPSC colonies were picked up 3 weeks after transfection, representing a reprogramming frequency of 0.384%. Of these, 10 were found to contain a single copy of the correctly inserted transgenes immediately downstream of B2M, of which hiPSC clone 26 (hiPSC-26) was selected for further study. Southern blot hybridization demonstrated that hiPSC-26 contains one 4.6 kb wild-type (wt) and one 2.1 kb targeted B2M HindIII fragment hybridized to the gap probe (Fig. 2A upper panel), indicating that the targeting vector had been knocked into immediately downstream of B2M locus. To confirm this, PCR was performed with primer A and B. As primer A is located within rtTA transgene and primer B in the gap region of the B2M (Fig. 1), the PCR can only detect the B2M-targeted vector but not those randomly inserted vectors. In support of Southern blot hybridization analysis, one 2.2 kb transgene fragment located in B2M locus was detected in the hiPSC-26 (Fig. 2A, lower panel). To analyze the number of integration sites, genomic DNA samples were prepared and digested by SalI, KpnI, MluI, and ClaI, respectively, for Southern blot hybridization using a 425-bp DNA fragment homologous to neomycin transgene as probe. The hiPSC-26 was found to have a single copy of integrated vector in one site (Fig. 2B). Collectively, these results demonstrate that a single copy of B2M-targeted nonviral vector can be achieved for efficient reprogramming.
To remove the four reprogramming factor transgenes including the cancer-promoting genes Klf4 and cMyc, the hiPSC-26 was transfected with Cre-IRES-PUROR vector (Addgene) by using Nucleofector II as previously described . To reduce vector integration, uncut vector was used. Six hiPSC-26 colonies (A–F) were randomly selected on day 6 after transfecion to assess expression of exogenous reprogramming factors by QRT-PCR. Although significantly induced expression of each reprogramming factor can be demonstrated in the cultures of hiPSC-26 clones, A, C, D, and E in the presence of doxy, there was no evidence of induced expression of those reprogramming factors in the cultured hiPSC-26 clone B and F (Fig. 2C), representing exogenous reprogramming factor-free hiPSC clones. PCR using Cre forward and reverse primer demonstrated no Cre vector integration in the hiPSC-26 clone B (hiPSC-26B; Supporting Information Fig. S2), indicating that transient expression of Cre recombinase had occurred for Cre-mediated recombination in this clone. Thus, hiPSC-26B was used in the following study. To examine if this site-specific insertion has caused B2M gene dysfunction, QRT-PCR was performed using B2M forward and reverse primer to assess B2M expression. The iPSC-26B and hESC (H9.2) were cultured in DM and left to spontaneously differentiate for up to 14 days. In comparison to H9.2, the hiPSC-26B exhibited similar expression level of B2M RNA in the differentiated as well as undifferentiated culture (Fig. 2D). Since B2M is essential for cell surface expression of MHC-I molecules, aberrantly expressed B2M could lead to lack of MHC-I expression on cell surface. To further evaluate B2M function of the hiPSC-26B, immunocytochemistry was performed to examine MHC-I expression. In support of the B2M expression data, normal expression level of MHC-I was demonstrated on the surface of the differentiated hiPSC-26B cells (Fig. 2D). Taken together, these data demonstrate that hiPSC-26B is genetic mutation-free and exogenous reprogramming factor-free clone.
The hiPSC-26B cells propagated robustly and maintained their self-renewal capacity, exhibiting typical hESC morphology for at least 68 passages in hESC supporting conditions in the absence of doxy. As demonstrated by immunocytochemistry, the hiPSC-26B uniformly expressed the pluripotency markers Nanog, Oct4, SSEA4, and TRA-1–60 but not SSEA1, (Fig. 3A). The cytogenetic analysis using standard protocols for high-resolution G-banding showed the normal karyotype in all analyzed metaphase cells of hiPSC-26B (Fig. 3B). To examine in vivo pluripotency of hiPSC-26B, cells were transplanted intramuscularly on left hind leg of SCID mice to test for teratoma formation. Tumors were noted in 3 weeks and harvested in 8 weeks after injection. The histological analysis revealed that the teratomas were composed of a wide variety of cell types derived from all three embryonic germ layers (Fig. 3C). Collectively, these data demonstrate that genetic mutation-free and exogenous reprogramming factor-free hiPSCs can be successfully generated with this novel, site-specific targeting strategy.
Differentiation and Purification of hiPSC Derived ATIICs
When cells were cultured on Matrigel-coated plate in DM, approximately 13.2% of cells in differentiated cultures of hiPSC-26B spontaneously differentiated into ATIICs on day 14 (Table 1), which was demonstrated by immunostaining for ATIIC-specific SPC expression. The percentage of SPC positive cells was comparable to that (14.0%) in differentiated H9.2 cell cultures. To enrich and purify hiPSC-ATIICs, hiPSC-26B cells were subjected to spontaneous differentiation in DM containing 20 µg/mL of G418 for 14 days. As mentioned above, hiPSC-26B cells already harbor the SPCP-NEOR transgene downstream of the B2M gene. Since SPC protein is specifically expressed by ATIICs, NEOR should be only expressed when hiPSC-26B cells are differentiated into ATIICs. Therefore, only hiPSC-ATIICs can survive G418 selection. To determine the purity of G418-selected hiPSC-ATIICs in the differentiated cultures of hiPSC-26B, the differentiated cells were stained with anti-human proSPC. The number of SPC positive cells was counted per 1,000 cells based on 4',6-diamidino-2-phenylindole (DAPI) staining on each plate. The results showed that the G418 selection leads to an enrichment of hiPSC-ATIICs to greater than 99% (Table 1). The few SPC negative cells were ATICs due to spontaneously differentiation of the ATIICs to ATICs after removal of G418.
Table 1. Relative content of hiPSC-alveolar epithelial type II cells in G418 selected and nonselected cultures of differentiated hiPSC-26B
Ultrastructure and Biological Functions of hiPSC-ATIICs
Ultrastructural examination by transmission electron microscopy demonstrated that hiPSC-ATIICs are morphologically normal and exhibit typical lamellar bodies with normal concentric, tightly packed lamellae (Fig. 4A), which is a characteristic hallmark of hATIICs. Papanicolaous staining demonstrated that all G418-selected hiPSC-ATIICs contain lamellar bodies (Supporting Information Fig. S3B). As hATIIC control, G418-selected hiPSC-ATIICs expressed SPA, SPB, SPC, and thyroid transcription factor one (TTF1) (Fig. 4A). Recently, SPC+uteroglobin (CCSP)+ bronchioalveolar stem cells (BASCs) have been identified at bronchioalveolar duct junction . These SPC+CCSP+ BASCs may serve as progenitor for both ATIICs and Clara cells. To examine the possibility if the hiPSC-26B derived BASCs were also present in the G418-selected cultures, the cells were stained with CCSP. No CCSP positive cells were observed in the G418-selected cultures on day 14 (Fig. 4A), suggesting that production of the more primitive lung precursor cells derived from hiPSCs/hESCs may be difficult. It has been reported that hATIICs in primary culture system retain regulated expression and secretion of surfactant in a manner that recapitulates that found in lungs . To examine if the hiPSC-ATIICs possess the same ability, the cultured cells were labeled with 3H-choline and examined for stimulated surfactant lipid secretion in response to secretagogue. We found that secretion of the 3H-labeled PC (as percent of total 3H-labeled PC) by hiPSC-ATIICs was in a regulated fashion as that seen in the control hATIICs (Fig. 4B), which demonstrated the active surfactant metabolism pathway in the hiPSC-ATIICs. To further examine if the hiPSC-ATIICs can be induced to express and secrete surfactant proteins, the selected hiPSC-ATIICs were cultured by using air-liquid interface culture system in SAGM in the absence or presence of Bt2cAMP + Dex for 5 days. We found that hATIICs and hiPSC-ATIICs synthesize and secrete SPC and SPB at very similar levels, which can be significantly induced in response to the treatment of Bt2cAMP + Dex (Fig. 4C).
One of the key functions of ATIICs is to serve as progenitor cells for maintaining alveolar homeostasis. It is critical that the hiPSC-ATIICs also possess the ability to proliferate and differentiate to ATICs for developing cell-based therapies. To demonstrate this ability, the G418-selected hiPSC-ATIICs were cultured in six-well culture dishes with DMEM containing 10% FBS and left to spontaneously differentiate for up to 8 days . Total RNA was isolated from the cultures on days 0, 2, 4, 6, and 8. Differentiation to ATICs was assessed by QRT-PCR to detect gene expression of Aquaporin-5 (AQP5) and Podoplanin (T1α), which are frequently used as markers for ATICs [26-30]. As hATIICs, differentiated hiPSC-ATIICs exhibited increased expression of AQP5 and T1α at a similar level, indicating differentiation over time into ATICs (Fig. 4D). To examine their proliferation capacity, the selected hiPSC-ATIICs were cultured on Matrigel-coated six-well plates at 1.0 × 104 cells per well in MEF-conditioned DMEM for up to 7 days to test the colony formation ability . A few small colonies were observed on day 4, but more colonies (approximately 16 colonies per well) composed of 15–21 cells were present on day 7's cultures, indicating that hiPSC-ATIICs proliferated in vitro. Proliferation capacity of hiPSC-ATIICs can be promoted by addition of recombinant keratinocyte growth factor (30 ng/mL; R&D Systems) in the culture medium, leading to significantly increased number and size of colonies (average 45 colonies per well and 27–40 cells per colony, Supporting Information Fig. S3). Taken together, these data have demonstrated that hiPSC-26B cells can be derived into an enriched population of ATIICs that possess normal important biological functions and ultrastructurural characteristics of hATIICs. Thus, the hiPSCs generated by our novel strategy will provide reliable source of cells to explore their therapeutic potential in lung tissue regeneration.
Functional Repair of BLM-Induced Acute Alveolar Injury in Mice by Transplanted hiPSC-ATIICs
We used BLM-induced acute lung injury model, which is well-characterized in mice and used routinely as a general model of pulmonary fibrosis [31, 32], to evaluate the capacity of hiPSC-ATIICs to engraft and repair damaged alveoli. To avoid graft rejection of hiPSC-ATIICs, immune deficient SCID mice were used in the study. As described in Materials and Methods section, mice were transplanted with hiPSC-ATIICs (1.0 × 106 in 50 µL of saline) endotracheally on day 2 after BLM exposure. In addition, BLM-challenged mice transplanted with same number of hmonos were used to verify the specificity of hiPSC-ATIICs in engrafting and repairing the injured alveoli. The engraftment efficiency as well as the capacity of transplanted hiPSC-ATIICs to differentiate into ATICs in lungs was determined by immunohistochemical staining (IHC) of end point lung sections using anti-human nuclei, anti-proSPC, and anti-T1α antibody . There was no human nuclei positive cell observed in control mice and BLM-treated mice receiving either saline or hmonos (Fig. 5A, Table 2). In contrast, numerous human nuclei positive cells were identified in the BLM-challenged lungs transplanted with hiPSC-ATIICs, most of which were costained with proSPC (indicated by arrows), representing the retention of transplanted hiPSC-ATIICs. QRT-PCR analysis of expression of human TTF1 in the BLM-challenged lungs with and without transplantation of hiPSC-ATIICs supported this observation (Supporting Information Fig. S3A). In addition, costaining of human nuclei positive cells with T1α (Fig. 5B) in the BLM-challenged lungs transplanted with hiPSC-ATIICs demonstrated that some of transplanted hiPSC-ATIICs had differentiated into ATICs. To determine percentage of remained hiPSC-ATIICs that had been differentiated into ATICs, 500 SPC positive cells were counted in at least three slides per mouse lung receiving hiPSC-ATIICs; the number of human nuclei positive cells was determined in the same counted area of 500 SPC positive cells. Of these human nuclei positive cells, approximately 59.1% of cells were costained with proSPC and 40.9% of cells proSPC negative (Table 2), indicating numerous remained hiPSC-ATIICs had differentiated into ATICs. The percentage of hiPSC-ATIICs differentiated into ATICs was confirmed by counting human nuclei positive cells costained with T1α in BLM-challenged lungs. Collectively, these data demonstrate the capacity of transplanted hiPSC-ATIICs to remain and differentiate into ATICs in injured alveoli.
Table 2. Relative content of human-derived ATIICs and ATICs in BLM-mouse lung tissue
Exposure to this relatively high dose of BLM severely impaired the endogenous repair capacity, thus damaged alveolar epithelium could not be effectively repaired, resulting in severe alveolar structure damage in 73.1% ± 5.3% of area showing extensive accumulation of inflammatory cells and interstitial thickening on day 10 (Fig. 5A). To evaluate BLM-mediated collagen deposition in the lungs, level of hydroxyproline, a modified amino acid specifically derived from collagen, was assessed. BLM-exposed lungs showed a significantly increased hydroxyproline content when compared with control lungs (Fig. 5C1). Transplantation of hmonos after BLM challenge did not significantly reduce BLM-induced lung damage (71.8% ± 5.9%) and collagen deposition. In contrast, the extent of parenchymal damage within BLM-challenged lungs transplanted with hiPSC-ATIICs was greatly reduced to only 6.2% ± 2.3% of the lung area as evidence by a few isolated, small destructed regions, whereas much larger areas showed tissue with normal alveolar organization (Fig. 5A). In addition, the level of hydroxyproline in the BLM-injured lungs that received hiPSC-ATIICs was significantly reduced to control level, suggesting that transplanted hiPSC-ATIICs reversed or prevented development of BLM-induced fibrosis.
BLM-induced destruction of alveolar architecture could result in severely impaired pulmonary function associated with significant loss of body weight. As illustrated in Figure 5A, 5C1, BLM-induced alveolar damage and collagen deposition were significantly repaired by transplanted hiPSC-ATIICs; therefore, it was expected that impaired lung function of BLM-exposed mice can be restored by transplanted hiPSC-ATIICs. To test this hypothesis, body weight and lung function were examined. The body weight of BLM-challenged mice receiving either saline or hmonos significantly decreased on day 2 and continued to decrease over time to approximately 70% of their starting body weight on day 10 (Fig. 5C2). Although BLM-challenged mice transplanted with hiPSC-ATIICs experienced an initial decrease in body weight, their body weight was significantly increased on day 6 and almost completely recovered on day 10. The lung tidal volumes of BLM-challenged mice without transplantation of hiPSC-ATIICs significantly decreased. Likewise, blood oxygen levels of these mice declined drastically (Fig. 5C3, 5C4), indicating that the mice experienced respiratory failure. As expected, the BLM-challenged mice transplanted with hiPSC-ATIICs showed normal lung tidal volumes and blood arterial oxygen saturation levels, demonstrating functional repair of injured alveoli by the transplanted hiPSC-ATIICs.
Any pluripotent cells or incompletely differentiated hiPSCs remained in G418 selected hiPSC-ATIIC cultures may cause either teratomas or lung epithelial tumors after transplantation into recipients. To assess the possibility, 24 female SCID mice were subjected to BLM-induced acute lung injury, and treated with saline, hmonos, or hiPSC-ATIICs (eight mice per group) as above. All the BLM-challenged mice without the treatment of hiPSC-ATIICs died of severe respiratory failure before day 12 after BLM-challenge (Supporting Information Fig. S4A). Impressively, all the BLM-challenged mice transplanted with hiPSC-ATIICs remained healthy with normal pulmonary function and no evidence of teratoma formation during the study period of 11 months after transplantation (Supporting Information Fig. S4), indicating that the selection procedure produces pluripotent cell-free hiPSC-ATIICs. Immunofluorescent staining demonstrated that approximately 5% of the hiPSC-ATIICs and approximately 8.4% of the hiPSC-ATIIC-derived ATICs remained in the BLM-challenged lungs 330 days after transplantation (Supporting Information Fig. S4C, Table S2). The percentage of remaining hiPSC-ATIICs as well as their derived ATICs reflects the turnover of transplanted hiPSC-ATIICs in the BLM-injured lungs, which is comparable to what we found in our previous study . Collectively, these data demonstrate the capacity of transplanted hiPSC-ATIICs to functionally repair injured alveolar epithelium with a long-term benefit, without side effect.
With our novel strategy, a single copy of all of the required transgenes can be specifically knocked into a site immediately downstream of B2M gene locus at a high frequency to generate random integration-free and exogenous reprogramming factor-free hiPSCs, without causing B2M dysfunction. Thus the risk of genetic abnormalities caused by vector integrations, a major concern for clinical use of hiPSCs, is completely eliminated. Previously, it was reported that a significant number of iPSCs generated with vector-based methodologies was not fully reprogrammed to pluripotency [33, 34], possibly because of stochastic expression variation of reprogramming factors and/or disrupted reprogramming pathways due to insertion mutations. Nongenetic techniques have been developed to generate vector integration-free iPSCs, yet they require repeated delivery of reprogramming factors, which is still stochastic. A major advantage of our new strategy is that a single copy of four reprogramming factor transgenes can be inserted downstream of the B2M gene locus, thereby facilitating the efficient reprogramming somatic cells into genetic mutation-free hiPSCs. As expected, the derived hiPSCs express a panel of stem cell markers, including TRA-1–60 previously reported not to be expressed in partially reprogrammed clones , and strongly resemble characteristics of hESCs with a long-term self-renewal capacity.
Development of safe and efficient methodologies to derive pluripotent stem cells into transplantable tissue cell types is critical to explore the therapeutic potential of hiPSCs for tissue regeneration. Efficiency of differentiation of ESCs into ATIICs is very low . Use of factors to manipulate signaling pathways for efficient differentiation of ESCs into ATIIC cells has been reported  but not yet successful for generating transplantable ATIICs. In addition, it remains unclear whether or not those ESC-derived ATIIC cells possess normal biological functions of hATIICs. Recently, a pure population of ATIICs derived from genetic modified hESCs has shown their therapeutic potential for functional repair of injured alveolar epithelium . However, immune rejection toward hESC-derived ATIICs and random genetic modification limit their therapeutic application. As our novel technique combines the ATIIC-specific NEOR expression strategy, the derived random integration-free and exogenous reprogramming factor-free hiSPCs can be selectively differentiated into a homogenous population of ATIICs without the need of additional genetic modification. Our data reported for first time that the hiPSC-ATIICs are fully functional with characteristics of hATIICs. A relatively low dose of G418 was used during the entire time course of differentiation for ATIIC selection, which allows more efficient in vitro production of hiPSC-ATIICs compared with a high dose of G418 used to eliminate all other cell types in the last two days (Supporting Information Table S3). Treatment with low dose of G418 from the beginning may promote activation of the signaling pathways required for derivation of ATIICs. Although the underlying mechanism still awaits deciphering, our results demonstrate that this novel strategy is reliable for producing sufficient number of functional hiPSC-ATIICs for regeneration of injured alveolar epithelium.
To explore the therapeutic potential of hiPSC-ATIICs, a relative high dose of BLM was used to eliminate the endogenous repair capacity of the injured alveoli. Therefore, no ATIICs survived in the damaged alveolar areas of BLM-challenged mice treated with saline or hmonos, causing extensive alveolar structure damage. All of these mice died 12 days after BLM challenge. Our data have demonstrated the retention and differentiation of hiPSC-ATIICs in the BLM-injured lungs that received hiPSC-ATIICs on day 10, suggesting that functional transplantation of hiPSC-ATIICs had prevented or reversed development of pulmonary fibrosis. It was observed that numerous endogenous mouse ATIICs survived in the BLM-injured lungs of hiPSC-ATIIC transplanted mice, suggesting that structural engraftment and differentiation of hiPSC-ATIICs is not the only beneficial event provided by transplanted hiPSC-ATIICs. In vitro studies have suggested that reciprocal interaction between ATIICs and mesenchymal cells may play an essential role in maintaining the homeostasis of vascular and alveolar epithelial compartment in lung [37, 38]. Therefore, the remained hiPSC-ATIICs may provide an appropriate signal microenvironment via their paracrine effect to promote endogenous repair capacity. Thus, structural engraftment and differentiation of hiPSC-ATIICs coupled with activated endogenous ATIICs accounted for robust re-epithelialization of injured alveoli. Further investigations are needed to characterize signaling communication between ATIICs for maintaining alveolar homeostasis. In addition, unlike BLM-challenged mice treated with saline or hmonos, all BLM-injured mice transplanted with hiPSC-ATIICs maintained normal pulmonary function well with no evidence of tumorigenesis for at least up to 11 months, suggesting that the strategy is safe and efficient to generate transplantable hiPSC-ATIICs.
In conclusion, we have successfully established a novel site-specific insertion targeting strategy for efficiently generating genetic mutation-free and reprogramming factor-free hiPSCs. Importantly, the derived hiPSCs can be subsequently differentiated into a homogenous population of functional ATIICs. This novel strategy makes it plausible for the first time to explore the clinical application of patient-specific iPSC-derived ATIICs. Furthermore, this technology can be applied generally for generation of hiPSC for transplantation in different organs. For example, the ATIIC-specific NEOR expression transgene in the vector can be easily replaced with any cell type-specific NEOR expression transgene to develop a platform for generating patient-specific neuron progenitor, heart cells, pancreatic β cells, and other for tissue regeneration or disease modeling.
This work was supported by The Brown Foundation Institute of Molecular Medicine, University of Texas Medical School at Houston, U.S. Public Health Service National Institutes Health Grants R21 HL102833-01(D.W) and R01 HL168116 (J.L.A), and the Nancy and Clive Runnells Embryonic Stem Cell Research Fund (R.A.W).
Q.Y., Y.Q., and D.W.: conception and design, collection and assembly data, data analysis and interpretation, and manuscript writing; H.S., X.P., Z.Z., and J.L.A.: collection of data and data analysis and interpretation; R.A.W.: administrative support. Q.Y. and Y.Q.: co-first authors contributed equally to this article.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflict of interests.