Transplantation of human pluripotent stem cell-derived GABAergic neurons generated by neuronal programming could be a potential treatment approach for neurodevelopmental disorders. This protocol describes the generation and transplantation of human stem cell-derived GABAergic neuronal precursors into the brains of neonatal mice, allowing long-term investigation of grafted neurons and evaluation of their therapeutic potential.
A reduced number or dysfunction of inhibitory interneurons is a common contributor to neurodevelopmental disorders. Therefore, cell therapy using interneurons to replace or mitigate the effects of altered neuronal circuits is an attractive therapeutic avenue. To this end, more knowledge is needed about how human stem cell-derived GABAergic interneuron-like cells (hdINs) mature, integrate, and function over time in the host circuitry. Of particular importance in neurodevelopmental disorders is a better understanding of whether these processes in transplanted cells are affected by an evolving and maturing host brain. The present protocol describes a fast and highly efficient generation of hdINs from human embryonic stem cells based on the transgenic expression of the transcription factors Ascl1 and Dlx2. These neuronal precursors are transplanted unilaterally, after 7 days in vitro, to the hippocampus of neonatal 2-day-old mice. The transplanted neurons disperse in the ipsi- and contralateral hippocampus of a mouse model of cortical dysplasia-focal epilepsy syndrome and survive for up to 9 months after transplantation. This approach allows for investigating the cellular identity, integration, functionality, and therapeutic potential of transplanted interneurons over an extended time in developing healthy and diseased brains.
The establishment, maturation, and refinement of neuronal networks occur during the perinatal and early postnatal period and represe a crucial time windows for brain development1. From an after-birth exuberance of connectivity, the brain evolves to a fine-tuning of connections that extend until adolescence2. Consequently, alterations in genes expressed during these periods, as well as external factors or insults, define an individual's predisposition to multiple neurodevelopmental disorders. Impairments in cognition and motor function unfold over time, and pharmacological treatments are limited, the majority targeting symptoms with the risk of severe side effects.
Dysfunction in gamma-aminobutyric acid (GABA)ergic inhibition has been shown to be a major contributor to the underlying cause of various neurodevelopmental disorders3, such as fragile X syndrome, Angelman syndrome, epilepsy, schizophrenia, and autism. GABA is the major inhibitory transmitter of the central nervous system and is instrumental for maintaining excitatory/inhibitory (E/I) balance, the synchronization of neuronal firing, and computation. GABAergic interneurons are a heterogeneous population of neurons, with increasing functional intricacy in more complex brain regions4 and with evolution5,6. Considering the limited endogenous regenerative capacity of the human brain and the implication of interneuron dysfunction in several neurological disorders, the transplantation of GABAergic interneurons may be a promising therapeutic avenue to explore. Along this line, human stem cell-derived GABAergic interneuron-like cells (hdINs) seem to be the most translational and viable source for this purpose compared to rodent allogenic neuronal precursors or other sources used elsewhere7. Protocols for generating GABAergic neurons from diverse cell sources are available8,9,10,11,12,13, but more knowledge is required on how hdINs mature, integrate, and function over time in a developing pathologic brain. Several studies have identified alterations in genes active during cortical patterning, establishing neuronal connectivity14, and tuning physiological E/I balance15. Neonatal transplantation of hdINs into mouse models with corresponding genetic perturbations allows us to follow the interplay between host and graft, which is knowledge necessary to determine potential therapeutic strategies.
Immunomodulation is commonly and successfully used in xenograft transplantations to avoid triggering a host immune response and rejection16. However, the administration of immunosuppressive drugs, such as cyclosporine A, causes renal toxicity after chronic administration, is labor-intensive because of the need for daily intraperitoneal injections to achieve stable systemic concentrations, causing animal stress17, and has off-target effects that can interact with pathology18. In addition, compromising the immune system has been shown to change behavioral phenotypes19, with alterations in the corresponding neuroanatomical regions20. Transplantation in the first week of life has been shown to allow for adaptation to transplanted cells21,22, while others have reported initial survival followed by rejection of the grafts within the first postnatal month23,24.
This protocol describes procedures from hdIN generation to cell transplantation in neonatal mice resulting in long-term graft survival and allowing for the investigation of the neuronal specificity, synaptic integration, function, and therapeutic potential of transplanted human interneurons during physiological and pathological development.
All experimental procedures were approved by the Malmö/Lund Animal Research Ethics Board, ethical permit number 12548-19, and conducted in agreement with the Swedish Animal Welfare Agency regulations and the EU Directive 2010/63/EU for animal experiments. C57BL6/J and Contactin-associated protein-like 2 (Cntnap2) knock-out (KO) mice, both males and females, were used at postnatal day (P) 2 for the present study. Human embryonic stem cells (hESCs) were used. The animals and stem cells were obtained from commercial sources (see Table of Materials).
1. Generation of the hdIN precursors
NOTE: All the steps in this section are done in a cell culture hood. The hESCs were maintained as feeder-free cells on coated plates using a stem cell culture medium and passaged as colonies.
Figure 1: Generation of hdIN precursors from hESCs by overexpressing Ascl1 and Dlx2. (A) Schematics of the differentiation protocol used for the generation of hdIN precursors. (B–E) Immunocytochemistry of hdIN precursors at 7 DIV for (B) the neuronal marker MAP2, (C) the proliferative marker Ki67, (D) general nuclear staining, and (E) a merge of the previous markers. Scale bar: 50 µm. Please click here to view a larger version of this figure.
2. Preparation of the single-cell suspension for transplantation
NOTE: All the steps in this section are done in the cell culture hood. On 7 DIV, hdIN precursors are dissociated and used for transplantation.
3. Intrahippocampal cell transplantation
NOTE: All the steps in this section are performed outside the cell culture hood in the animal facility. Early postnatal transplantation of cells into the brain was performed on P2, considering P0 the day of birth.
Figure 2: Stereotaxic transplantation in newborn mice pups at P2. (A) A Play-Doh-like stage for holding the pup's body in position and inverted ear bars (magenta arrows). (B) White front paw (blue arrow) indicative of the reduced blood flow in that area so that the pup is experiencing anesthesia by hypothermia. (C) Overview of the setup with the pup already covered by ice over the soft tissue paper. (c#) Closed-zoom of the head of the pup, with the injection needle already inserted into the brain (yellow dashed line indicating lambda and lambdoid sutures). This figure is adapted from Gonzalez Ramos et al.27. Please click here to view a larger version of this figure.
Following the protocol presented here and illustrated in Figure 1A, hdIN precursors were not proliferative yet at 7 DIV as defined by (i) negative immunoreactivity for the cell cycle marker Ki67 and (ii) expressing neuronal markers such as microtubule-associated protein 2 (MAP2) (Figure 1B–E). This characterization was performed on leftover cells replated for 24 h after having undergone all the procedure steps. In addition, the gene expression analysis published previously indicated that a rapid transition from the pluripotent state to a neuronal phenotype occurs around 4 DIV and 7 DIV8. Overall, these results confirmed the presence of postmitotic cells and the absence of risk for teratoma formation.
Next, the hdIN precursors' survival after early postnatal transplantation into the hippocampus of wild-type (WT) mice was tested by immunohistochemistry against the human cytoplasmatic marker STEM121. The hdIN precursors were transplanted into the right dorsal hippocampus of naïve immunocompetent mice at P2, which were then sacrificed at P14 and 2 months PT. Grafted cells were found across the whole dorsal hippocampus, as well as dispersed through the corpus callosum and the contralateral hippocampus, at both time points. Moreover, at both time points, grafted hdINs expressed Ascl1, one of the induction transcription factors (Supplementary Figure 1), and were not proliferative, as indicated by the absence of Ki67 expression (Supplementary Figure 2).
Importantly, no immune reaction or local inflammation against the transplanted cells was found either at P14 or 2 months PT, as assessed by the absence of reactive microglia identified using Iba1, CD68, and galectin-3 (Gal3) (Figure 3), the extent of astrogliosis determined by the glial fibrillary acidic protein (GFAP) and inflammatory cytokines such as interleukine-1 (IL-1), and the absence of cytotoxic T lymphocytes (CD8) (Figure 4).
Figure 3: hdINs at P14 and 2 months PT into the hippocampus of newborn WT mice without triggering immune rejection from the host tissue. Immunofluorescence for Iba1, CD68, and Gal3 markers in brain tissue from (A–D) the proximity of an ischemic core area in an electrocoagulation stroke mouse model (positive control, Ctrl +), (E–H) negative control animals (Ctrl-) at 2 months and (M–P) P14, and (I–L) animals that have undergone cell transplantation at 2 months and (Q–T) P14. The white arrows indicate some examples of blood vessels visible at all channels due to autofluorescence. Abbreviations: Ctx = cortex; Hip = hippocampus; DG = dentate gyrus; CA1 = cornu ammonis 1. Scale bar: 50 µm. Please click here to view a larger version of this figure.
Figure 4: hdINs at 2 months PT into the hippocampus of newborn WT mice without triggering immune rejection from the host tissue. Immunofluorescence for IL1, GFAP, and CD8 markers in brain tissue from (A–D) the proximity of an ischemic core area in an electrocoagulation stroke mouse model (positive control, Ctrl +), (E–H) negative control animals (Ctrl-) at 2 months, and (I-L)animals that have undergone cell transplantation at 2 months. Abbreviations: Ctx = cortex; Hip = hippocampus; DG = dentate gyrus. Scale bar: 50 µm. Please click here to view a larger version of this figure.
Similarly, hdIN precursors were also transplanted into the hippocampus of Cntnap2 KO mice, a model for autism spectrum disorder and cortical dysplasia-focal epilepsy syndrome. In the Cntnap2 KO mice, indeed, the hdINs survived up to 9 months PT and were localized at the injection site, although they were also dispersed across the ipsilateral and even the contralateral hippocampus as observed in the WT mice (Figure 5). Moreover, most of the grafted hdINs were immunoreactive for interneuron markers, as expected from previous results in vitro8,26 and in adult rodents in vivo25.
Figure 5: Grafted hdINs in the hippocampus of Cntnap2 KO mice at 9 months PT. (A) Immunochemistry against the cytoplasmatic human marker STEM121 in cell-transplanted (left) and sham (right) mice. (a1 and a2) Magnified images of STEM121 positive cells. (B) Immunofluorescence for STEM121 (magenta) and the interneuron markers parvalbumin (PV) and somatostatin (SST). White arrows indicate double positive cells for STEM121 and the respective interneuron marker. (C) Orthogonal view of a grafted hdIN immunoreactive for STEM121 and PV. Scale bar: 200 µm (A and B), 100 µm (a1 and a2), 20 µm (small square magnification in a1 and a2, and C). Please click here to view a larger version of this figure.
Supplementary Figure 1: Grafted hdINs at 2 months PT into the hippocampus of newborn WT mice expressing Ascl1. Immunofluorescence against Ascl1 and the cytoplasmatic human marker STEM121 at the (A) CA3 and (B) DG in cell-transplanted WT mice. (a) Magnified image of a STEM121 positive cell. The white arrows indicate double-positive cells for STEM121 and Ascl1. Scale bar: 100 µm. Please click here to download this File.
Supplementary Figure 2: Post-mitotic grafted hdINs at 2 months PT into the hippocampus of newborn WT mice. Immunofluorescence against the proliferative marker Ki67 and the cytoplasmatic human marker STEM121 in (A) naïve and (B) cell-transplanted WT mice. (b) Magnified image of a STEM121 positive cell. The yellow arrow indicates a cell positive for Ki67 and negative for STEM121. The white arrow indicates cells positive for STEM121 and negative for Ki67. The white asterisk points out a lateral ventricle. Scale bar: 100 µm. Please click here to download this File.
The present protocol describes a robust, fast, simple, and widely accessible methodology to generate hdIN precursors in vitro and its use as early interventional cell therapy in preclinical models of neurodevelopmental disorders.
Even though some of the characteristic phenotypes of neurodevelopmental disorders arise during adolescence or adulthood, pathophysiological alterations are already present during early development. For this reason, early intervention would be highly warranted for achieving beneficial effects by acting in critical brain developmental periods before symptomatology or clinical manifestation. In the future, genetic screening and the development of biomarkers will afford prophylactic or pre-symptomatic treatment, representing a game changer for those patients. Therefore, hdIN precursors were transplanted early after birth in the Cntnap2 KO mouse model when epileptogenic changes in the neuronal network might be ongoing28 and at a timepoint at which cellular alterations have been described in this animal model29. It is important, however, to consider the potential pitfalls of age extrapolation and the timing of certain processes in the mouse versus the human brain.
Focusing on the procedure itself, the differentiation protocol presented here is based on the use of transcription factors, which allow for fast and highly efficient programming of stem cells compared with other protocols elsewhere based on small molecules10,30. A potential drawback of this approach could be the requirement for lentiviral vectors, which carries a risk of insertional mutagenesis. Two critical steps in the protocol are the addition of the antibiotics and the anti-mitotic agent to the medium to select for cells expressing the transcription factors and eliminate proliferative cells, avoiding the risk of teratoma formation, respectively. Although only hdIN precursors were tested in this study, the procedure is expected to be feasible with other cell sources and programming/differentiation protocols. Nevertheless, other neuronal subtypes and/or models should be validated.
The hdIN precursor's age for transplantation, 7 DIV, was decided based on (i) the absence of proliferative cells, assessed by immunoreactivity against Ki67, (ii) together with the previously reported observation of decreases in the expression of pluripotency genes such as POU5F1 and the appearance of the neuronal marker MAP2 and the interneuron marker GAD1 at that timepoint8. However, the original work describing this protocol performed transplantations at 14 DIV after DOX withdrawal9. This raises questions about whether DOX in the mother's drinking water can reach cells grafted into the brains of nursing pups via the milk, or if 7 DIV of DOX induction is enough to establish the GABAergic fate. Although Yang et al. identified 14 days of DOX as sufficient to generate stable neuronal cells in vitro9, Gonzalez-Ramos et al.8 detected GAD1 gene expression already at 7 DIV, indicating that the downstream activation of GAD67 by Ascl1 and Dlx2 has already occurred at this time point. Hence, patterning has begun at 7 DIV and might be less dependent on the DOX treatment. Moreover, evidence in rodents and humans indicates the presence of DOX in breast milk31,32, and the results presented here show that grafted hdINs were immunoreactive for Ascl1 at 2 weeks and 2 months PT and interneuron markers later on at 9 months PT. Within the grafted population, besides PV and SST positive neurons, other markers for subpopulations of interneurons were also found in lower amounts, such as calretinin (CR) and calbindin (CB).
A challenging aspect of this procedure is the coordination of the timings for both differentiation and the age of the pups. Usually, mouse gestation takes 21 days after setting up the mating cage, albeit this can vary sometimes. This scenario does not occur when performing cell transplantations in adult rodents when everything can be carefully planned and arranged. Nevertheless, this can be easily mitigated by setting up two to three mating cages with a 2 day interval or two to three differentiation batches with a 2 day time-lapse from each other.
Although the mice used in this study were neither immunodeficient nor immunosuppressed, the transplanted cells survived up to 9 months in vivo, and markers of immune reaction against xenogeneic cells or local inflammation were not observed at either P14 or 2 months PT. Immune rejection of grafted xenogeneic cells is triggered against MHC/peptides, and the key cellular mediators of graft rejection are T lymphocytes and microglial cells33,34. Therefore, immunoreactivity to markers of T cells, as well as reactive microglia, was explored. No signs of immune rejection of the grafted cells in the host tissue were detected either by levels of reactive microglia or by the presence of T lymphocytes in WT mice at P14 or 2 months. Moreover, no local inflammation was observed based on the assessed levels of astrogliosis and inflammatory cytokines. This outcome could partly be dependent on neonatal immune tolerance35,36,37, observed by other cell identities, locations, and animal models35,38. Englund et al. identified regional differences in the outcome of the grafted cells in terms of migration and maturation, including the observation of grafted cells in the adjacent white matter35.
Finally, a greater dispersion of the grafted cells within the hippocampus was observed compared to other studies transplanting into adult rodents, where hdINs remained as a grafted core25. This dispersion also differed from results observed previously by Yang et al.9, which could be explained in this case by the age of the cells at the time of transplantation.
The authors have nothing to disclose.
This project was funded by the Swedish Research Council (Grant Number: 2016-02605, M.A.), the Swedish Brain Foundation F02021-0369 (M.A.), the Crafoord Foundation (M.A.), and the European Union Horizon 2020 Programme (H2020-MSCA-ITN-2016) under the Marie Skłodowska-Curie Innovative Training Network project Training4CRM No. 722779 (M.K.). We are extremely grateful for the help from Andrés Miguez, from Josep Maria Canals' lab (Laboratory of Stem Cells and Regenerative Medicine, University of Barcelona), for teaching stereotaxic cell transplantation in P2 newborn mice, and Mackenzie Howard, group leader at the University of Texas at Austin, for the advice and preliminary coordinates for cell transplantation into the hippocampus of P2 newborn mice. We thank Susanne Geres for assisting with animal care and Ling Cao for the help with processing tissue, as well as students that have contributed in one way or another to the study and specifically Diana Hatamian. Finally, some of the graphics used to illustrate this paper were created with BioRender.com.
30 G needle | B Braun | 4656300 | |
33 G needle for Hamilton syringe | Hamilton | 7762-06 | |
4-well plates | Thermo Scientific | 176740 | |
Accutase | STEMCELL Technologies | 7920 | Cell detachment solution use for splitting cells (hESC and hdIN precursors) |
Adjustable volume pipettes 10, 20, 200, 1000 µL | |||
Alexa Fluor Plus 488/555/647 | Thermo Fisher | 1:1000 | |
Anti-CD68 (Rat) | Bio-Rad | MCA1957 | 1:200 |
Anti-CD8 (Rabbit) | Abcam | 203035 | 1:200 |
Anti-Galectin 3 (Goat) | R&D systems | AF1197 | 1:500 |
Anti-GFAP (Guiena Pig) | Synaptic systems | 173004 | 1:500 |
Anti-Iba1 (Rabbit) | WAKO | 19119741 | 1:500 |
Anti-IL1 (Goat) | Santa Cruz Biotech | SC-106 | 1:400 |
Anti-Ki67 | Abcam | ab16667 | 1:250 |
Anti-Ki67 (Rabbit) | Novocastra | NCL-Ki67p | 1:250 |
Anti-MAP2 (Chicken) | Abcam | ab5392 | 1:2000 |
Anti-Mash1 (Ascl1) | Abcam | ab74065 | 1:1000 |
Anti-Parvalbumin (Rabbit) | Swant | PV 27 | 1:5000 |
Anti-Somatostatin (Rat) | Millipore | MAB354 | 1:150 |
Anti-STEM121 (Mouse) | Takara Bio | Y40410 | 1:400 |
Avidin/Biotin Blocking Kit | VECTOR Laboratories | SP-2001 | |
B6.129(Cg)-Cntnap2tm1Pele/J | Jackson Laboratory | 17482 | Animal model |
Biotinylated Horse anti-Mouse | VECTOR Laboratories | BA-2001 | 1:200 |
Burker Chamber | Thermo Fisher Scientific | 10628431 | |
C57BL/6J | Janvier Labs | Animal model | |
Centrifuge | For 15 mL tubes | ||
Confocal microscope | Nikon | Confocal A1RHD microscope | |
Costar 6-well Clear TC-treated | Corning | 3516 | |
Cy3 Stretavidin | Jackson ImmunoResearch | 016-160-084 | 1:200 |
Cytosine β-D-arabinofuranoside (AraC) | Sigma | C1768 | 4 µM |
DAB Substrate Kit, Peroxidase (With Nickel) | VECTOR Laboratories | SK-4100 | |
Digital Stereotax | KOPF | Model 940 | |
DMEM/F12 | Thermo Fisher Scientific | 11320082 | Use for the N2 medium |
DNase I Solution | STEMCELL Technologies | 7900 | 1 µg/mL |
Doxycyclin | Sigma-Aldrich | D9891 | 2 µg/mL |
DPBS -/- | Gibco | 14190144 | |
Epifluorescence microscope | Olympus | BX51 Microscope | |
Ethanol | Solveco | 70%, 95%, 99.8% | |
FUW-rTA | Addgene | 20342 | Lentiviral vector |
FUW-TetO-Ascl1-T2A-puromycin | Addgene | 97329 | Lentiviral vector |
FUW-TetO-Dlx2-IRES-hygromycin | Addgene | 97330 | Lentiviral vector |
H1 (WA01) ESC | WiCell | WA01 | Human embryonic stem cell line under a MTA agreement |
H2O2 | Sigma-Aldrich | 18304 | |
Hamilton Syringe | Hamilton | 7634-01 | 5 µL |
HBSS | Gibco | 14175095 | No calcium, No magnesium – Transplantation medium |
Hoechst 33342 | Invitrogen | H3570 | 1:1000 |
Hygromycin B | Gibco (Invitrogen) | 10687010 | |
Incubator | 5% CO2, 37 °C | ||
Isoflurane Baxter | Apoteket AB | ||
Manual cell counter | VWR | 720-1984 | |
Matrigel hESC-Qualified Matrix, LDEV-free | Corning | 354277 | For the coating |
Methanol | Merck Millipore | 1060091000 | |
Microscope Coverslips 24 x 60 mm | Thermo Scientific | BBAD02400500#A113MNZ#0## | |
Microscope Slides | VWR | 631-1551 | |
Microscope Software | Olympus | CellSens | |
Mounting media | Merck | 10981 | PVA-Dabco |
Mouse adaptor to stereotax | RWD | 68030 | |
mTeSR1 | STEMCELL Technologies | 85850 | Kit Basal Medium and 5X Supplement – Stem cell culture medium |
N2 supplement | Gibco | 17502048 | |
NaOH | Sigma-Aldrich | S8045 | 1M |
Penicillin-Streptomycin | Sigma-Aldrich | P0781 | |
Pertex | HistoLab | 811 | |
Pipet Filler | |||
Play-Doh | |||
Puromycin (Dihydrochloride) | Gibco | A1113803 | |
Round cover glasses thickness No. 1.5H (tol. ± 5 μm) 13 mm Ø | Marienfeld | MARI0117530 | For immunocytochemistry |
Serum | Thermo Fisher | Goat, Donkey, Horse | |
Sterile pipette tips | For volumes 0.1-1000 µL | ||
Sterile serological pipettes | 5, 10, 25 mL | ||
Sterile water Braun | B Braun | 3626873 | |
Sucrose | Sigma-Aldrich | S8501 | For 0.5% Sucrose solution |
Triton X-100 | Sigma-Aldrich | X100 | |
Trypan Blue Solution | Gibco | 15250061 | |
Tubes | Sarstedt | 15 ml, Eppendorf 1.5 mL | |
Tweezer | VWR | ||
Ultra pure water | MilliQ Water System | ||
Xylene | VWR | 28973.363 | |
Y-27632 (ROCK inhibitor) | STEMCELL Technologies | 72304 | 10 µM |