An Alternative In Vivo Model to Evaluate Pluripotency of Patient-Specific hiPSCs

The generation of autologous human induced pluripotent stem cells (hiPSCs) from a patient’s somatic cells and the subsequent differentiation of these cells into desired cell types offer innovative treatment options for tissue regeneration. The hiPSCs obtained are usually implanted in immunodeficient mice, and teratoma formation is analyzed after 4 to 6 weeks to assess the cells’ pluripotency. In this study, an alternative in vivo model based on chicken egg chorioallantoic membrane (CAM) was established to analyze the pluripotency of newly created hiPSCs. 0.5, 1, 2, 4 x 106 hiPSCs generated from urine-derived renal epithelial cells were seeded on CAM and incubated for 9 days. Teratoma formation was detected in 70% of eggs inoculated with 2 x 106 hiPSCs and in 100% of eggs inoculated with 4 x 106 hiPSCs. All teratomas exhibited vascular structures. The robustness of the CAM model was confirmed using two additional hiPSC lines derived from human fibroblasts (NuFFs) or jaw periosteal cells. The presence of all three germ layers within the teratomas was successfully verified by histochemical and immunofluorescence staining and gene expression analysis of germ layer-specific markers. Urine-derived renal epithelial cells were used as negative control and showed no teratoma formation. The CAM-based in vivo model provides an optimal in vivo test environment for the pluripotency evaluation of newly generated hiPSC lines. This simple, fast, inexpensive and reproducible method reduces the suffering of animals and thus implements the principles of the 3Rs (replacement, reduction, and refinement). This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International license (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the original work is appropriately cited. # contributed equally


Introduction
The groundbreaking discovery of the reprogrammability of somatic cells into human induced pluripotent stem cells (hiPSCs) opened up new opportunities in the field of tissue engineering and the development of personalized cell therapies. Since hiP-SCs are derived from patient's somatic cells, their generation and use avoid ethical concerns related to embryonic stem cells. Thus, hiPSCs are a promising cell source to generate patient-specific cell types. Yamanaka and colleagues first reprogramed murine fibroblasts into iPSCs (Takahashi and Yamanaka, 2006), and shortly thereafter reprogramming was also demonstrated in human fibroblasts using retroviral vectors encoding four transcription factors, Klf4, c-Myc, Oct4 and Sox2 (Takahashi et al., 2007). However, retroviral vectors are inserted into the host genome, and this is associated with a significant risk of insertional mutagenesis, incomplete transgene silencing or reactivation, and residual expression of reprogramming factors, which can lead to reprogramming were performed according to our recent study (Steinle et al., 2019b). The obtained hiPSCs were cultivated on T25 culture flasks coated with 0.5 mg/cm 2 vitronectin (Thermo Fisher Scientific, Waltham, MA, USA) in E8 stem cell medium (Essential 8, Thermo Fisher Scientific) at 37°C and 5% CO 2 . After reaching 70% confluence, hiPSCs were washed once with Dulbecco's phosphate-buffered saline (DPBS) (Thermo Fisher Scientific) and incubated for 5 min at 37°C with DPBS containing 0.5 mM ethylenediaminetetraacetic acid (EDTA, Sigma-Aldrich, Steinheim, Germany). After detachment, cells were suspended in E8 medium containing 10 µM ROCK inhibitor Y-27632 (Enzo Life Sciences, Lausen, Switzerland) and passaged at a 1:10 split ratio into a new vitronectin-coated T25 culture flask. hiPSCs were cultivated at 37°C and 5% CO 2 (normoxia). After 24 h, the medium was changed to E8 medium without ROCK inhibitor Y-27632, and daily medium changes were performed.

Chorioallantoic membrane (CAM) assay
Fertilized chicken eggs of the Lohmann White x White Rock breed chicken were obtained from a breeding facility (Matthias Sittig, Buchholz, Germany). Feathers, dirt and excrement were removed from the eggshells by wiping with a wet tissue. The eggs were placed in an egg incubator (Heka-Brutgeräte, Rietberg-Varensell, Germany) and incubated at 37°C and 60% relative humidity (Day 0). The eggs were completely rotated twice a day. At day 3 after fertilization, an 18G needle was inserted at the tip of the egg without harming the yolk, and 2-3 mL albumen was removed to lower the level of the CAM. A semi-permeable adhesive tape, Suprasorb F (Lohmann & Rauscher, Rengsdorf, Germany) was fixed on the eggshell. Then, under a sterile bench, using sterile surgical scissors, a circular opening (Ø 1-1.5 cm) was cut into the eggshell without damaging the CAM. Unfertilized eggs without a beating heart or without vasculature were discarded. Afterwards, the window of the eggshell of viable eggs was After the successful generation of hiPSC lines, a detailed characterization of the cells is essential. In addition to the exclusion of genetic abnormalities and the expression of proteins associated with stem cell properties, the confirmation of pluripotency is required. To confirm the pluripotency of newly generated hiPSC lines, their differentiation into three germ layers is analyzed by subcutaneous or intramuscular injection of hiPSCs in immunodeficient mice (DeBord et al., 2018;Nelakanti et al., 2015). Animals are monitored for approximately 6 weeks and sacrificed to explant the teratoma before it is larger than 1 cm 3 . Because of its tumor-like growth, it can cause the animals pain and suffering and therefore raises ethical concerns (Buta et al., 2013). In addition, keeping immunodeficient mice is time-consuming and expensive.
In recent years, the chorioallantoic membrane (CAM) assay was applied to investigate angiogenesis (Steinle et al., 2018;Naik et al., 2018), tumorigenesis (Li et al., 2015;Dexter et al., 1983;Durupt et al., 2012;Rovithi et al., 2017), bone and cartilage generation (Moreno-Jimenez et al., 2016), and irritant potential of chemicals (Gilleron et al., 1996;Ying et al., 2010). CAM is formed by the fusion of the mesodermal layer of the allantois with the mesodermal layer of the chorion by the third day of development of the chicken embryo (Kunz et al., 2019;Dohle et al., 2009). The highly vascularized CAM mimics a perfect environment for cell transplantation, and it can maintain engrafted cells (Deryugina and Quigley, 2008). Furthermore, the CAM assay provides a highly reproducible, cost-effective, immunodeficient, and non-innervated extra-embryonic test environment (Kunz et al., 2019;Kunzi-Rapp et al., 2001) and can be used to implement the 3R principles (reduction, replacement, and refinement) (Petrovova et al., 2019). An ethical advantage of the CAM assay is that the CAM itself is not innervated, allowing the growth of xenografts without pain or impairment of the embryo (Kunz et al., 2019). Furthermore, the lack of nociception in chicken embryos due to incomplete neuronal differentiation until day 14 of the gestation period makes the model a favorable alternative to rodent models (Buhr et al., 2020). Therefore, in some countries, ethical approval is not required for CAM assays with chicken embryos until 14 days of development. In other countries, the chicken embryo is not considered an independent living animal until day 17 or until hatching (Winter et al., 2020), so ethical approval for animal experiments is not required. Thus, depending on the country, the legal requirements should be followed and, if necessary, the approval of an ethics committee for animal experiments should be obtained.
In this study, we describe the applicability of a CAM assay-based in vivo model as an alternative to conventional rodent models for analyzing the pluripotency of patient-specific hiPSCs by spontaneous teratoma formation.

Cultivation of hiPSCs
Footprint-free hiPSCs were generated by reprogramming human renal epithelial cells (RECs) isolated from 100-200 mL urine of healthy human donors using VEE-OKSiM-GFP srRNA encoding OCT4, KLF4, SOX2, cMYC, and GFP. Transfection and closed with the adhesive tape to prevent dehydration and to minimize the risk of contamination. Subsequently, eggs were incubated without rotation at 37°C and 75% relative humidity. On day 7, the semi-permeable adhesive tape covering the circular opening was carefully opened with sterile scissors and the development of the eggs was analyzed. Insufficiently developed or contaminated eggs without clearly visible vasculature and movement of the embryo were removed. Sterile silicone sealing rings of cryovials with an inner diameter of 0.85 cm (neoLab, Leonberg, Germany) were carefully placed onto the CAM. 0.5, 1, 2, or 4 x 10 6 hiPSCs (passage 11) suspended in 50 µL cell culture medium and mixed with 50 µL Matrigel (hECS qualified, Corning, NY, USA) were applied into the silicone ring. As a control, 1 x 10 6 RECs were applied onto the CAM. After the application, the opening in the eggshell was sealed with the adhesive tape again and further incubated at 38°C and 80% relative humidity. On day 16, the CAMs including the generated teratoma were excised around the silicone ring and fixed overnight at 4°C with 4% paraformaldehyde (PFA, Merck, Darmstadt, Germany) for further analysis.
To verify the robustness of the CAM model, 2x10 6 hiPSCs generated from newborn human foreskin fibroblasts (NuFFs, Amsbio, Milton Park, UK) or human jaw periosteal cells (obtained from Prof. Dr Dorothea Alexander-Friedrich, Department of Oral and Maxillofacial Surgery, University Hospital Tübingen) using srRNA were also applied onto CAM, and the generated teratomas were analyzed.
To prove the presence of cells from all three germ layers, immunohistochemical staining was performed using an automated immunostainer (Ventana Medical Systems, Inc., Arizona, USA) and specific antibodies against early cell types of three germ layers according to the company's protocols with slight modifications. Monoclonal mouse anti-human CD34 antibody (Dako, Agilent Technologies, Santa Clara, USA) was used to show the presence of mesoderm tissue. CD34 is a marker for hematopoietic stem cells and endothelial progenitor cells. Monoclonal mouse anti-human SALL4 antibody (M03), clone 6E3, (Abnova, Taipeh, Taiwan) was used to detect endoderm tissue. The presence of ectodermal tissue was assessed using the monoclonal mouse anti-human vimentin (V9) (ROCHE, Basel, Switzerland) antibody, which is a neural and pancreatic progenitor cell marker. All antibodies were visualized by the automated immunostainer using DAB (deaminobenzidine). Counterstaining was performed using hematoxylin. Images were taken using an Axiovert135 microscope (Carl Zeiss, Oberkochen, Germany) and EOS Utility software (Canon, Tokyo, Japan).
To perform immunofluorescence staining, the teratoma sections were washed with DPBS and blocked with 4% BSA in DPBS for 1 h at RT. Sections were then incubated for 1 h at RT with fluorescently labeled antibodies in DPBS with 2% BSA or in DPBS with 2% BSA and Triton X-100 (Sigma-Aldrich) for intracellular staining. After washing, the coverslips were mounted using Fluoroshield mounting medium with DAPI (Abcam, Cambridge, UK). Alexa Fluor ® 488-mouse anti-human β-tubulin (BD Biosciences, Franklin Lakes, USA), PE-mouse anti-human CD31, and PE-mouse anti-human CXCR4 (both from Miltenyi Biotec, Bergisch Gladbach, Germany) antibodies were used to detect the three different germ layers. Staining with the respective isotypes was performed as control.
qRT-PCR analysis RNA was isolated from the explanted teratomas generated from 2 x 10 6 hiPSCs or hiPSCs cultivated in cell culture plates (control) using the Aurum™ Total RNA Mini Kit (Bio-Rad, Munich, Germany). 300 ng RNA was reverse transcribed into complementary DNA (cDNA) using the iScript kit (Bio-Rad). The primers used for the specific amplification of transcripts, ordered from Eurofins (Luxembourg, Luxembourg) or ELLA Biotech (Martinsried, Germany), are listed in Table 1. They were used at a final concentration of 300 nM. Real-time qRT-PCR reactions were performed in a CFX Connect Real-Time PCR Detection System (Bio-Rad) using IQ SYBR Green Supermix (Bio-Rad). Expression of the constitutively expressed gene GAPDH (glyceraldehyde 3-phosphate dehydrogenase) was used as an internal control for the amount of RNA input. Primers were designed by using the Primer-Blast tool from NCBI (Ye et al., 2012). Melting temperatures and self-complementarities were checked using the Oligonucleotide Properties Calculator from Northwestern University Medical School (Kibbe, 2007).
The qRT-PCR amplification of cDNA was performed under the following conditions: 3 min at 95°C for one cycle, followed by 40 cycles of 95°C for 15 s, 60°C for 30 s, and 72°C for 10 s. After 40 cycles, melt curve analyses were performed to detect the specific amplification of transcripts. The qRT-PCR reactions led to an increased teratoma size (24.47 ± 11.71 mm²) compared to the seeding of 1 x 10 6 (12.79 ± 1.91 mm²) or 0.5 x 10 6 hiPSCs (12.5 ± 0.69 mm²) (Fig. 2C). Histological examinations showed the formation of vascular structures within the teratomas, which enable the blood supply to the cells (Fig. 2D).

Analysis of the tri-lineage differentiation of the hiPSCs
H&E staining of the explanted teratomas that were generated from 2 x 10 6 hiPSCs revealed their tri-lineage differentiation (Fig. 3). Differentiation into mesodermal tissue was demonstrated by the presence of bone-like structures (Fig. 3A I) and differentiation into endodermal tissue was shown by the detection of the primitive gut-like epithelium (Fig. 3A II). The presence of squamous epithelial tissue (Fig. 3A III) showed the differentiation of hiPSCs into ectoderm tissue.
Immunohistochemical analyses demonstrated the expression of the mesodermal marker CD34 (Fig. 3B I), endodermal marker SALL4 (Fig. 3B II), and the ectodermal marker vimentin (Fig. 3B III) associated with the respective morphological structures. The antibodies showed no binding to tissue sections of CAMs that were seeded with RECs (Fig. S2A 1 ).
The immunofluorescence analyses of cryosections with specific antibodies revealed a strong expression of CD31 within the mesodermal tissue structures (Fig. 3C I), CXCR4 within endodermal tissue (Fig. 3C II), and β-tubulin within ectodermal structures (Fig. 3C III). Isotype controls showed no binding (Fig.  S2B 1 ). Furthermore, the immunofluorescence analyses of CAMs seeded with RECs showed no binding to CAM (Fig. S2C 1 ).
were run in triplicates with a total volume of 15 μL per well. Levels of mRNA for each gene were normalized to GAPDH. The results are shown relative to control mRNA levels.

Analysis of teratoma formation upon application of hiPSCs onto CAM
To analyze the teratoma formation of newly generated hiPSCs, a CAM-based assay was established (Fig. 1A). The viability of embryos was evaluated using 176 viable eggs. After the opening of eggs on day 3 and incubating the eggs until day 7, 15 embryos died, which corresponds to a survival rate of 91%. After 16 days, 63% of the initial embryos were still viable (Fig. 1B).

Discussion
Patient-specific hiPSCs offer the possibility to regenerate destroyed cell types and tissues for personalized treatment. After the generation of hiPSCs and their differentiation, an extensive characterization of the cells including confirmation of their pluripotency is required. Typically, cells are implanted in immuno-To further prove the functionality of the CAM model, the teratoma formation potential of two other hiPSC lines, which were generated in our laboratory by reprogramming of NuFFs or JPCs, was assessed by seeding of 2 x 10 6 cells on CAM. The H&E staining of the explanted teratomas, which were generated from NuFF-derived (Fig. 4A) or JPC-derived (Fig. 4B) hiPSCs, demonstrated tri-lineage differentiation potential similar to that of hiPSC derived from RECs (Fig. 3). Differentiation into mesodermal tissue was demonstrated by the presence of bone-like structures (Fig. 4A I) and adipose tissue (Fig. 4B I). Furthermore, gutlike epithelium showing endodermal differentiation (Fig. 4A II, Although the established CAM assay-based method has several advantages over the in vivo mouse teratoma assay, it also has some limitations. Teratoma formation on CAM can only be performed for 9-10 days after the application of hiPSCs, as the chick hatches 21 days after incubation. In contrast, in mice the incubation period can be extended to increase the size or tissue maturation of the teratoma if the animal's behavior is not negatively affected. In a further study, kinetic trajectories showed that around 37 days are required to see and measure the size of the teratoma externally (McDonald et al., 2020). Thus, we assume that the tissue structures formed in teratomas on CAM are less mature than those in mice.
However, the established CAM assay is clearly appropriate to demonstrate the tri-lineage differentiation capability of the generated hiPSCs. The robustness of the established CAM assay has been demonstrated by the successful generation of teratomas after the application of hiPSC lines generated from NuFFs or JPCs in addition to hiPSCs derived from RECs. The tri-lineage differentiation was successfully demonstrated with all three hiPSC lines. RECs seeded on CAM did not cause teratoma formation.
The Matrigel used to apply the hiPSCs to CAM is obtained from the murine Engelbreth-Holm-Swarm (EHS) tumor. Therefore, replacing Matrigel by synthetic hydrogels, such as polymers and scaffolds based on polyacrylamide and polyethylene glycol (PEG) (Aisenbrey and Murphy, 2020), will further reduce and prevent pain and suffering of animals in the future.
To test the influence of inoculated cell numbers on teratoma formation, different numbers of hiPSCs were seeded on CAM. The increase of the cell number from 0.5 x 10 6 to 2 x 10 6 hiPSCs led to a higher teratoma formation efficiency (30% versus 70%). The further increase of hiPSC number to 4 x 10 6 cells resulted in teratoma formation on all CAMs. However, 2 x 10 6 hiPSCs were sufficient to obtain teratomas containing deficient mice to demonstrate the ability of newly created hiPSCs to differentiate into all cell types of the three germ layers, i.e., mesoderm, endoderm, and ectoderm (Nelakanti et al., 2015). In this study, we tested the applicability of the CAM assay as a new in vivo model for the evaluation of the pluripotency of hiPSCs to reduce animal suffering.
Commonly, subcutaneous (Cao et al., 2007), intramuscular (Lee et al., 2009), intramyocardial (Cao et al., 2006), or sub-renal capsule implantation of hiPSCs is performed into 15 to 20 immunodeficient mice aged 6 to 10 weeks. The animals are maintained for a period of approximately 4 to 6 weeks to allow the growth of the implanted cells and teratoma formation (Nelakanti et al., 2015;Aldahmash et al., 2013). If the cells are transplanted subcutaneously or intraperitoneally, easily palpable teratomas are formed around 6 weeks post-injection (Zhang et al., 2008). The teratomas can be measured using the traditional caliper method (Hentze et al., 2009). Mice should be sacrificed before the teratoma is larger than 1 cm 3 or earlier if the dimension reached impairs the animal's behavior, motility or food and water intake due to pain or distress (Nelakanti et al., 2015). The composition of the explanted teratoma is then analyzed using histological and immunohistochemical staining.
Using the established CAM assay, the incubation of implanted hiPSCs for 9 days was sufficient to form characteristic tissue structures of all three germ layers. Thus, the established CAM model is cost-effective and far less time-consuming than the mouse model. Moreover, since the CAM is not innervated, it allows the growth of xenografts without pain or impairment of the embryo (Kunz et al., 2019). The highly vascularized CAM efficiently supports the growth of the teratoma. Furthermore, the immaturity of the chicken embryo's immune system enables the use of cells from different species (Deryugina and Quigley, 2008) as the seeded cells are not rejected. Fig. 4: Detection of the three germ layer-specific tissue types in teratoma sections generated from NuFF-or JPCderived hiPSCs Representative microscopic images of H&E stained teratoma sections of (A) NuFF-derived hiPSCs: mesoderm: (I) bone-like tissue; endoderm: (II) primitive gut-like epithelium; ectoderm: (III) squamous epithelium, and (B) JPC-derived hiPSCs: mesoderm: (I) adipose tissue; endoderm: (II) primitive gutlike epithelium; ectoderm: (III) squamous epithelium. The arrows indicate the described germ layer-specific structures. These images are taken from three teratomas, respectively. cells of all three germ layers. Thus, we recommend using 2x10 6 or more hiPSCs for the teratoma analysis. In immunodeficient mice, usually, 1 x 10 6 cells are used per injection, and it has been shown that injections of 2 or 4 x 10 6 hiPSCs can increase the chances of successful teratoma formation in mice (Nelakanti et al., 2015).
Cells of all three germ layers were detected within the formed teratoma from 2 x 10 6 hiPSCs. The presence of endoderm, mesoderm, and ectoderm tissue cell types was detected using specific antibodies as well as by detection of gene expression using qRT-PCR. In addition to using the established CAM model to analyze the pluripotency of hiPSCs, it can also be applied to determine whether unwanted hiPSCs remain in the differentiated cells. The implantation of remaining, not fully differentiated hiPSCs can lead to the formation of teratomas and must be avoided for clinical use. Thus, the CAM assay also can be applied to analyze the safety of hiPSC-derived cell types and the complete differentiation of hiPSCs into somatic cells.

Conclusion
In this study, an alternative in vivo model was established and tested to evaluate the pluripotency of newly generated hiPSC lines within 9 days. The CAM model is a valuable method to bridge the gap between in vitro cell culture and in vivo animal experiments. In contrast to the immunodeficient mouse model, it is simple, inexpensive, and time-saving. The application of this CAM assay can reduce the number of required animals and their suffering. Besides the pluripotency evaluation of the reprogrammed cells, the CAM assay can be also applied to test the safety of differentiated cells from hiPSCs.