Introduction

Zebrafish (Danio rerio) represents a promising biomedical model in cancer research and regenerative medicine (Yang et al. 2013; Santoriello and Zon 2012). Although non-mammalian, this animal model has many advantages when compared to other established mammalian animal models. Zebrafish is easy to manipulate, cost-effective, easy to breed, genetically well-defined, and optically transparent at embryo stage (Stoletov and Klemke 2008). The unique optical properties of zebrafish embryos allows researchers to perform direct imaging of grafted cells for several days, and initially without the use of immunosuppressive drugs (Stoletov et al. 2007). Zebrafish also effectively absorbs small chemicals directly from the water and can be used for screening candidate agents with potential therapeutic properties. Therefore, zebrafish represents a great bioassay tool that allows disease-driven drug target identification and in vivo drug validation (Peal et al. 2010; Kari et al. 2007).

The discovery of technology for reprogramming somatic cells into pluripotent stem cells (also known as induced pluripotent stem cells or iPS cells), with four defined factors originally pioneered by Shinya Yamanaka in 2006, has received a lot of attention from the scientific community (Takahashi and Yamanaka 2006; Takahashi et al. 2007). iPS cells can be generated by the ectopic expression of a defined set of transcription factors, and they seem to be indistinguishable from ethically controversial embryonic stem cells regarding their differentiation potential (Stadtfeld and Hochedlinger 2010). The autologous cells derived from induced pluripotent stem cells hold great promise for future tailored or patient-specific therapies (Jung et al. 2012). The iPS cell technology allows generation of patient- and disease-specific cell lines and therefore represents a powerful tool for drug discovery and human disease modeling (Egawa et al. 2012; Marchetto et al. 2010). To the best of our knowledge, this is the first report describing grafting, differentiation, and survival of human iPS cell-derived neural precursor cells, transplanted into zebrafish. We assume that in the field of regenerative medicine, zebrafish can be used for drug-screening studies using patient-specific, induced pluripotent stem cell (iPSc)-derived cells (Singh et al. 2015; Chamberlain et al. 2008). The aim of this study (to establish, evaluate, and optimize protocol for short-term grafting of human iPSc-derived neural precursors into zebrafish) is therefor of interest.

Material and Methods

iPS Cell Isolation

Human neural progenitor cells (hNPCs) were derived from iPS cells induced from a skin biopsy of a normal, healthy individual at the University of California, San Diego, UCSD. Derivation of hNPCs was approved by UCSD Internal Review Board (Approval ID No. 100887). Briefly, skin biopsy from a healthy individual was washed twice with sterile PBS (Cat. No.10010023, Gibco, USA), supplemented with antibiotics (1% penicillin/streptomycin, Cat. No.15140122, Gibco, USA), and transferred to a sterile plastic polystyrene Petri dish (Cat. No. FB0875712, Fisher Scientific, USA) with DMEM/F12/GlutaMAX Supplement media (Cat. No. 31331093, Gibco, USA), 10% FBS (Cat. No. 16000044, Gibco, USA), and 1% penicillin/streptomycin. Primary skin fibroblasts started to appear around the tissue biopsy in about 7–12 days. Isolated fibroblasts were expanded, and the fourth passage was used for reprogramming into iPS cells (Fig. 1(A, B)).

Fig. 1
figure 1

The scheme of experiment showing generation of induced pluripotent stem cells (iPS cells) from human skin biopsy with retroviruses encoding four Yamanaka factors. Stable RFP expressing reporter line from subsequently generated neural precursor cells was prepared, and cells were transplanted intracranially or into blastocyst stage (4-h post-fertilization) embryo (A). The typical phenotype of isolated human fibroblasts (B), reprogrammed cells (iPS cell colony) (C), embryoid bodies (D), and neural precursor cells (E) is shown. Immunocytochemical staining of transduced Yamanaka factors in reprogrammed cells (iPS cell colonies, grown on MEF cells) showed cells positive for Oct-4, Sox-2, Klf-4, and c-Myc (G–J). Note that the feeder cells (MEFs) were negative for these factors. DAPI DNA fluorescent probe was used for staining nuclei. Bar represents 10 μm. Freshly FACS sorted (K) human neural precursor cells and the same cells plated on ORN/LAM plates in differentiation media (L and M). Typical changes in phenotype of cells are observed (L and M) together with expression of neural markers like TUJ (neuron-specific class III beta tubulin), DCX (doublecortin), MAP-2 (microtubule-associated protein 2), and ChAT (choline acetyltransferase) (N–R). Differentiated human iPSc-derived NPCs expressing GFP driven through synapsin promoter 5 days after induction of differentiation (S)

Full size image

Transduction with Yamanaka Factors

Fibroblasts were transduced with retroviral reprogramming vectors encoding four Yamanaka factors. Oct4, c-Myc, Klf4, and Sox2 human cDNAs were obtained from Addgene, and viruses were generated as described previously (Marchetto et al. 2010). Briefly, to increase efficiency of transduction, cells were washed in PBS containing SureEntry Transduction Reagent (Cat. No. NC0259653, Qiagen, Germany) and transduced by mixing cell pellet with 20 μL of virus stock. Cells were then incubated for 30 min at 37 °C in cell incubator and replated into a 6-well plate in DMEM/F12/GlutaMAX media containing 10% FBS and 1% penicillin/streptomycin. Three days after transduction, cells were transferred onto mitotically inactivated mouse embryonic fibroblasts (feeder cells or MEFs, CF-1, irradiated, Cat. No. GSC6001G, Globalstem, USA) in DMEM/F12 GlutaMAX medium containing 20% KnockOut Serum Replacement (Cat. No. A3181502, Gibco, USA), 0.11 mM 2-mercaptoethanol or beta-mercaptoethanol (Cat. No. 21985023, Gibco, USA), 0.1 mM non-essential amino acids (NEAA; Cat. No. 11140035, Gibco, USA), and 20 ng/mL b-FGF recombinant human protein (Cat. No. 13256029, Gibco, USA). The emerging iPS cell colonies (Fig. 1(C)) were passaged by manual picking and expanded. Embryoid bodies (Fig. 1(D)) were isolated by transferring colonies onto Corning Costar Ultra-low Attachment plates (Cat. No. 07-200-601, Fisher Scientific, USA) with media containing DMEM/F12 GlutaMAX Supplement, 1% penicillin/streptomycin, and 10% FBS (Gibco, USA) and cultured in this media for 4 days for NPC isolation (Fig. 1(E)).

NPC Generation

Embryoid bodies were exposed to retinoic acid (10 ng/mL, Cat. No. R2625-50MG, Sigma-Aldrich, USA) in media containing DMEM/F12 GlutaMAX Supplement medium with 1% penicillin/streptomycin and 10% FBS) for 4 days and transferred to 6-well poly-l-ornithine/laminin-coated plates (Cat. No. 734-0255, VWR, USA) with NPC media (DMEM/F12 GlutaMAX, 0.5× B27 Supplement (Cat. No. 17504001, Gibco, USA), 0.5× N2 Supplement (Cat. No. 17502048, Gibco, USA), 1% penicillin/streptomycin, and bFGF (20 ng/mL).

Rosette-like structures were manually harvested and isolated cells were further expanded. The expanded NPCs were presorted for PSA-NCAM+ population on magnetic column with magnetically labeled antibody (Cat. No. 130-097-859, Miltenyi Biotec, Germany), expanded for 10 days, and subsequently sorted for CD184+/CD24+/CD44-/CD271-cell population with FACS Aria II sorter (BD Bioscience) as described (Yuan et al. 2011). FITC Mouse anti-human CD24 (Cat. No. 555427), PE Mouse anti-human CD44 (Cat. No. 555479), APC Mouse anti-human CD184 (Cat. No. 560936), and PE-CF594 Mouse anti-human CD271 (Cat. No. 563452) antibodies were from BD Biosciences, USA. Sorted cells were collected into a tube containing NPC media, plated into a 6-well plate, and allowed to recover in culture for 3–4 days. Recovered and expanded cells (Fig. 1(K)) were transduced with EF1A-RFP virus (Cat. No. PLV-10072-50, Cellomics Technology, USA), and stable fluorescent cell line was generated by puromycin-containing selection medium (selection was performed for 14 days; conc. of puromycin was 1 μg/mL; puromycin Cat. No. A1113803 from Thermo Fisher Scientific, USA, was used). Fluorescent NPCs were used for grafting into zebrafish.

SYN-GFP Reporter Cell Generation

Briefly, cells (NPCs) were washed in PBS containing SureEntry Transduction Reagent and transduced by mixing cell pellet with 15 μL of HIV1-SYN-GFP (10 M.O.I) virus stock (made and kindly provided by Dr. Atsushi Miyanohara, the director of the Vector Core Laboratory, School of Medicine, University of California, San Diego). Cells were then incubated for 30 min at 37 °C and replated into a 6-well plate with NPC media (DMEM/F12 GlutaMAX, 0.5× B27, 0.5× N2, 1%P/S, and bFGF (20 ng/mL)). To test the reporter cells, transduced NPCs were expanded and plated on ORN/LAM-coated plates and differentiation was induced with media containing 10% FBS and BDNF and GDNF (conc. of both factors was 10 ng/mL, Cat. No. 450-02 (BDNF) and No. 450-10 (GDNF), Peprotech, UK). Differentiated neurons expressing GFP driven through synapsin promoter were detected in 4–5 days after plating (Fig. 1(S)).

Cell Grafting

As mentioned already, human iPS cell-derived NPCs were transduced with EF1A-RFP lentivirus and stable RFP-expressing cells were generated with antibiotic selection (puromycin, conc. = 1 μg/mL, cells were selected for 14 days). The Tg(gfap:GFP) and wild-type AB zebrafish were obtained from ZFIN (Zebrafish International Resource center, http://zfin.org/ZDB-GENO-080606-263) and maintained according to the University of California animal welfare guidelines (described and approved by UCSD Institutional Animal Care and Use Committee, protocol number S06008). Embryos were obtained as described previously (Stoletov et al. 2007). For cell grafting, two approaches were used: (i) grafting into early-stage zebrafish embryo (blastocyst stage, 4 h after fertilization) and (ii) intracranial grafting into 3-day-old zebrafish larvae. The Eppendorf FemtoJet injector (Eppendorf, Germany) equipped with 0.75-mm borosilicate glass needle (Sutter instrument Cat. No. B100-75-10, diameter of the needle opening = 15 μm) was used in both cases to deliver cells into the zebrafish. The NARISHIGE micromanipulator (MN-151-08144) was used to precisely control the movement of the needle. Briefly, for an early embryo grafting, zebrafish embryos at blastocyst stage were placed and aligned well on a 2% agarose pad under the stereomicroscope (Leica MZFLIII). Approximately 100–300 cells in PBS (pH 7.4) were injected into the marginal zone. For later stage embryo grafting, zebrafish were anesthetized using tricaine (0.02% (w/v)), placed on the agarose pad, and 200–500 (for brain ventricle injection) of human neural progenitor cells diluted in PBS (pH 7.4) were injected into the ventricular zone of the forebrain and midbrain of the animals. After transplantation, zebrafish were maintained in fish water, in an incubator set to 35 °C. This temperature was sufficient for supporting the growth of neural precursor cells while it was still well-tolerated by zebrafish (Fig. 2(J)). To test the ability of human iPSc-derived NPCs to differentiate in vivo in zebrafish, SYN-GFP reporter system was used. Reporter cells (NPCs) were injected into an early-stage embryo (blastocyst stage, 4 h after fertilization) and injected animals were observed daily. Interestingly, early signs of differentiation (detected by GFP expression driven through synapsin promoter) have been detected in less than 48 h after grafting (Fig. 3(A–D)).

Fig. 2
figure 2

Transplantation of human iPS cell-derived neural precursor into zebrafish. Naïve, non-injected embryo (A) and blastocyst-stage embryo injected with RFP-labeled human NPCs are shown (B). Developing zebrafish without (C) or with (D and E) grafted cells. Note the localization of transplanted cells in zebrafish body. Three-day-old transgenic GFAP:GFP zebrafish injected intracranially (F, G, and H) with human iPS cell-derived NPCs. 3D reconstruction of intravital confocal images in Volocity software shows red human fluorescent cells clustering around zebrafish brain, which was labeled with GFP (Tg(gfap:GFP) (I). Graphs showing effect of incubation temperature on zebrafish survival (J, a total of 300 animals were used for this experiment) and percentage of fish with detectable fluorescent signal from grafted cells several days after transplantation (for both approaches, a total of 100 animals were used for cell grafting). Data from intracranial, orthotopic transplantation and early embryo injection (K and L). Black bar represents 250 μm; yellow bar represents 150 μm

Full size image
Fig. 3
figure 3

Transplantation of reporter iPSc-derived NPCs into early-stage wild-type zebrafish embryos. Reporter cells were generated with HIV1-SYN-GFP lentivirus vector. Less than 2 days after grafting, GFP expression driven through synapsin promoter was detected (a–d, a total of 50 animals were used for this study). In contrast to human breast cancer cells grafted into an early-stage embryo ((E–H), a total of 100 animals used for cancer cell grafting), human iPSc-derived NPCs do not cause developmental abnormalities in zebrafish. Bar represents 150 μm. (I) Injection setup for injection of cells into zebrafish: (a) Eppendorf FemtoJet injector (Eppendorf, Germany) equipped with a 0.75-mm borosilicate glass needle (diameter of the needle opening = 15 μm), (b) stereomicroscope (Leica MZFLIII), and (c) NARISHIGE micromanipulator (MN-151-08144) was used to precisely control the movement of the needle. (I) (d) Fluorescent lamp for detection of fluorescently labeled injected cells

Full size image

Imaging

Grafted cells were recorded by intravital imaging of anesthetized live animals on Nikon Eclipse TE2000 confocal microscope (Nikon, Japan) in a drop of E3 medium, without mounting into agarose. After recording, animals were immediately returned back to the incubator. To minimize stress, the animals were imaged every second day. Recorded animals were checked daily for vitality, and dead fish were removed from the tanks. A total of 100 animals were used for each grafting approach.

Image Processing and 3D Reconstruction

Collected image stacks were digitally reconstructed into high-resolution, 3D image using Volocity 6.3 (PerkinElmer, USA) software. Fluorescent intensity was analyzed with ImageJ software, and animals showing positive fluorescent, cell-related signal were recorded.

Immunocytochemistry

iPS cell colonies grown on feeder cells (mouse embryonic fibroblasts (MEFs)) from Globalstem (CF-1, irradiated, Cat. No. GSC6001G, Globalstem, USA) or Millipore (EmbryoMax, Mouse Primary Fibroblasts, Cat. No. PMEF-CF, Millipore, USA) and NPCs were washed with sterile PBS and fixed with 4% paraformaldehyde (pH 7.4) for 5 min at laboratory temperature. After fixation, cells were washed twice with PBS and blocked with 5% donkey serum (Cat. No. NC9624464, Jackson Immunoresearch Labs, USA) for 1 h. Primary antibodies from StemLight iPSc reprogramming antibody kit (Cat. No. 9092T, Cell Signaling Technology, USA) and companies listed here were diluted in PBS containing 1% donkey serum in PBS (pH 7.4) and 0.3% Triton X-100 (Cat. No. 1610407, Bio-Rad, USA). MAP2 antibody was from Millipore (Cat. No. MAB3418), TUJ and ChAT antibodies from Abcam (Cat. Nos. ab14545 and ab137349), and DCX antibody from LSBio (Cat. No. C204512). Fixed cells were incubated with primary antibodies overnight at 4 °C in refrigerator. Subsequently, cells were washed three times with PBS+0.3% Triton X-100 and incubated for 1 h at lab temperature with secondary antibody (Alexa Fluor 594 AffiniPure F(ab’)2 Fragment Donkey anti-Rabbit IgG (H+L), Cat. No. 711-586-152, Jackson Immunoresearch Labs, USA), diluted in PBS+0.3% Triton X-100 and 1% donkey serum. After incubation, cells were washed three times with PBS+0.3% Triton X-100 and mounted with ProLong Gold Antifade Mountant with DAPI (Cat. No. P36931, Thermo Fisher Scientific, USA). Images were taken with Olympus iX70 fluorescence microscope, with Orca camera (Hamamatsu, Japan), and processed in Wasabi software (Wasabi System Inc., USA) (Fig. 1(G–J, N–R)).

Statistical Analysis

All quantified data were analyzed and plotted in GraphPad Prism 6.0 (GraphPad Software, USA) with Student’s t test. Data reported as mean ± SD are representative of at least five independent experiments.

Results and Discussion

Several studies have been published based on zebrafish as an animal model in the field of cancer research (Zhao et al. 2015; Yen et al. 2014; Cao et al. 2013). However, there is still a lack of evidence for the use of zebrafish for transplantation of human iPS cells or iPS cell-derived neural precursors. The optical transparency of zebrafish allows tracking of fluorescently labeled grafted cells to study their behavior in real time. Therefore, in this study, we evaluated the possibility of using zebrafish for grafting of human-induced pluripotent stem cell-derived neural precursors to test their survival and differentiation. First, we successfully generated human iPS cells from skin biopsy (Fig. 1(A)) by reprogramming isolated skin fibroblasts (Fig. 1(B)) and verified the expression of transduced Yamanaka factors in iPSc colonies (Fig. 1(G–J)). Then, we generated neural precursor cells through embryoid bodies (Fig. 1(D, E)). The ability of these cells to differentiate into neurons was described previously by the co-author of this study (Marchetto et al. 2010) and also shown in this study (Fig. 1(N–S)). After differentiation (see “Material and Methods”), cells showed expression of TUJ (neuron-specific class III beta tubulin), DCX (doublecortin), MAP-2 (microtubule-associated protein 2), and ChAT (choline acetyltransferase) (Fig. 1(N–R)). To test the ability of transplanted cells to differentiate into neurons, we also injected human iPSc-derived NPCs transduced with reporter virus (HIV1-SYN-GFP) into early-stage wild-type zebrafish embryos (Fig. 1(S)). To characterize their survival, homing, and integration, high-resolution confocal in vivo imaging was used. For cell visualization, we utilized Volocity software to reconstruct images into 3D objects. Before cell grafting, we also analyzed the optimal temperature for our experiments. For zebrafish, a water temperature of 28.5 °C is considered to be an optimal temperature for growth, staging, and breeding, whereas mammalian cells are typically maintained at 37 °C. Several authors have shown successful short-term survival of zebrafish at 37 °C previously; however, in our experiments, this temperature has never been well-tolerated by zebrafish embryos. Therefore, we performed our own experiments to determine the optimal temperature that had to be close to 37 °C (for supporting the survival of iPS cell-derived neural precursors), but at the same time well-tolerated by zebrafish embryos. We found that there was no significant difference between survival of fish embryos at 33 versus 35 °C (Fig. 2(J)). However, we decided to continue our experiments at 35 °C, which is a temperature closer to an optimal temperature for mammalian cells.

For cell grafting, we applied two different approaches: (a) cell transplantation into early (blastocyst)-stage zebrafish embryos (Fig. 2(A–E) and (B) intracranial grafting into the ventricular zone of the forebrain and midbrain areas of 3-day-old zebrafish larvae (Fig. 2(F–I)). Both approaches have their own advantages. Injection of cells into an early-stage embryo is typically faster and does not require the use of anesthesia. On the other hand, intracranial injection into 3-day-old embryos allowed us to perform orthotopic, fluorescence-guided injection into the ventricular zone of the brain (Figs. 1(A) and 2(F–I)) and to graft more cells (approximately 2×) when compared with early embryo grafting (see “Material and Methods”). The mortality of fish after intracranial injection was surprisingly low, only about 15% of injected animals were lost because of trauma caused by injection. Therefore, we consider intracranial approach as a safe delivery method for cell grafting into the zebrafish. In our experiments, we used Tg(gfap:GFP) zebrafish that allowed us to perform fluorescence-directed cell delivery into GFP-labeled brain. The embryo injection was associated with higher mortality (20–30%, observed 1 day after injection) compared to intracranial injection.

For both grafting approaches, our data indicated that transplanted human neuronal progenitor cells survive in zebrafish for more than 2 weeks. This conclusion is based on visual scoring of fluorescent signal from transplanted cells (Fig. 2(K, L)). Approximately 15–17 days after cell grafting, signal intensity from fluorescent RFP+ cells started to decline, suggesting active involvement of developing immune system in xenograft rejection (Fig. 2(K, L)). We predict that for experiments in which longer survival of grafted cells is necessary (for example, teratoma formation from iPS cells as a confirmation of the true pluripotency of human-induced pluripotent stem cells), the use of water-soluble immunosuppressant (dexamethasone) or irradiation (immune ablation) will be necessary. Such experiments will probably require some optimization; however, benefits can be tremendous. Current mammalian models for teratoma assay are quite expensive, and their use requires substantial surgical skills (Buta et al. 2013). On the contrary, zebrafish is cheap and an easy-to-use animal model for preliminary in vivo experiments.

In this study, we also tested the ability of transplanted cells to differentiate into neurons in non-mammalian organism. To test their differentiation potential, we injected human iPSc-derived NPCs transduced with reporter virus (HIV1-SYN-GFP) into early-stage (wild-type) zebrafish embryos. GFP protein expression (driven through synapsin promoter) was surprisingly detected in less than 2 days after transplantation (Fig. 3(A–D)). Although more experiments and IHC evaluation for grafted differentiated cells need to be performed, our data showed that zebrafish embryo represents a supportive environment for mammalian iPSc-derived precursor differentiation.

Importantly, we observed that transplanted human iPSc-derived neural progenitor cells did not significantly affect normal development of zebrafish. On the contrary, when cancer cells (human breast cancer cells MDA-BR-231/GFP) were injected into an early-stage embryo, they showed negative effect on normal fish development and caused severe abnormalities in individual animals (40–60% of affected animals, Fig. 3(E–H)).

Conclusion

In conclusion, zebrafish is a suitable animal model that allows real-time, high-resolution detection of transplanted human, iPS cell-derived neural precursor cells and will be useful for high-throughput screening of new candidate therapeutics using patient-derived, disease-specific iPS cells.