Cell transplantation of retinal ganglion cells derived from hESCs

2.1 hESC culture

Human embryonic stem cell lines H9 and H7-A81 (Sluch et al., 2015), and hiPSCs were cultured in mTeSR1 (StemCell Technologies) on 6-well tissue culture plates coated with Matrigel (Corning) or Synthemax (Corning), incubated in 37°C at 5% CO₂, fed daily, and manually passaged approximately every 5 days or until about 80% confluence with ethylenediaminetetraacetic acid solution (EDTA, Sigma-Aldrich). For differentiation, hESCs were treated with Dulbecco’s phosphate-buffered saline (DPBS, Thermo Fisher Scientific) without Ca²⁺ and Mg²⁺ for 4 minutes and manually passaged and filtered using a 0.22 μm Steriflip Filter for detachment and embryonic body (EB) formation.

2.2 hESC differentiation

This chemically defined protocol for neural differentiation was modified from a previously described protocol (Lamba et al., 2006). Once stem cell cultures in 6-well plates were 80% confluent, media was changed to DMEM/F-12 medium supplemented with 1% N2 (both Thermo Fisher Scientific) on day 0. On day 1, medium was supplemented with ALK 2/3 inhibitor dorsomorphin (1 μM, Tocris) and ALK 5 inhibitor SB-431542 (10 μM, Sigma-Aldrich) to block BMP signaling, and Wnt/β-catenin signaling inhibitor IWP-2 (2 μM, Sigma-Aldrich), which we named N2+Dor+SB+IWP2 (N2DSI). On Day 2, to create embryonic bodies (EBs), we washed cells with PBS (without Ca²⁺/Mg²⁺) for 5 min at 37°C. Cells were then mechanically detached with cell scraper (BD Falcon) into N2DSI medium and transferred to ultra-low binding plates (Corning). From day 3, N2DSI medium was changed every other day, using a strainer to avoid the loss of EBs, through day 9. On day 10, EBs were dissociated with Accutase (Innovative Cell Technologies) and cultured on Matrigel-coated plates using DMEM/F-12 supplemented with 1% PS, 0.5% N2, 1% B27 and bFGF, 20 ng ml^–1 (NBF medium). RPCs formed from EBs over the next 6 days were then dissociated with Accutase and cultured on PDL/laminin-coated plates in NBF medium. Subsequent passaging of RPCs was noted to enrich these cultures; Rock inhibitor Y-27632 (10 μM; EMD Millipore, Billerica, MA) was added to prevent cell death after single cell dissociation.

For RGC differentiation, we unitized previously published Full Sato (FS) medium (Apara et al., 2017; Galvao et al., 2018) containing Notch signaling γ-Secretase inhibitor DAPT (10 μM; EMD Millipore, Billerica, MA). Half volumes of FS+DAPT medium were exchanged every other day until cell harvest.

2.3 Animals

All animal studies followed the Association for Research in Vision and Ophthalmology (ARVO) statement for the use of animals in research, and were approved by the Institutional Animal Care and Use Committee (IACUC) and the Institutional Biosafety Committee of the University of California, San Diego. Three-month-old Sprague-Dawley rats, both male and female, and 14-day-timed pregnant C57BL/6 mice were obtained from Charles River (Portage, MI).

2.4 Retinal progenitor cell isolation and differentiation

Retinal progenitor cells (RPCs) were isolated from embryonic day 14 (E14) mice, as previously described (Chang et al., 2017). Briefly, retinas were dissected, dissociated with papain (Worthington) in Dulbecco’s PBS (Life Technologies) incubated at 37°C for 30 min, and then gently triturated into single-cell (RPC) suspensions with ovomucoid inhibitor (Roche). RPCs were cultured in DMEM/F12 medium containing glucose (0.6%, Sigma), GS21 supplement (1:100, GlobalStem), Sato supplement (1:100), insulin (5 μg ml^–1, Sigma), epidermal growth factor (EGF; 20 ng ml^–1, Peprotech), fibroblast growth factor (FGF; 20 ng ml^–1, Peprotech), and penicillin/streptomycin (1%, ThermoFisher Scientific), in a humidified atmosphere at 5% CO₂, 37°C. RPCs were cultured 7 days before calcium imaging.

2.5 Quantitative real-time PCR

Total RNA was isolated from hESC-derived RGC-like cells (RNeasy; Qiagen, Valencia, CA) and treated with DNase before reverse transcription (RT). Then equal amount of RNA was reverse transcribed to cDNA (iScript cDNA Synthesis Kit, Bio-Rad, Hercules, CA). Quantitative real time PCR (qRT-PCR) was performed using primers against GAPDH, Ki67, Nanog, Rx, Pax6, Brn3a, Thy1 and Recoverin (Taqman Gene expression Mix Kit; ThermoFisher Scientific) in the iQ7 Multicolor Real-Time PCR Detection System (Bio-Rad). Gene expression levels were normalized to reference gene GAPDH.

2.6 Calcium imaging

For intracellular calcium recordings, E14 RPCs-derived or hESC-derived RGC-like cells were cultured on PDL/laminin-coated glass coverslips as previously described (Trakhtenberg et al., 2017). Coverslips with cultured cells were incubated for 30 min at room temperature with 5 μM fura 2-AM (F-1221; Life Technologies) in Tyrode’s solution: (in mM) NaCl 129, KCl 5, CaCl₂ 2, MgCl₂ 1, glucose 30, and HEPES 25, pH 7.4, containing 1 mg/ml bovine serum albumin. The cells were washed with Tyrode’s solution and placed in the imaging chamber (RC-26GLP, Warner Instruments). Calcium imaging was performed on an inverted fluorescence microscope (AxioObserver, Zeiss) with a fast switching light source (DG-4, Sutter Instruments) using MetaFluor software (Molecular Devices) working within the linear range of the camera. Regions of interest (ROIs) were drawn around brightly labeled cell bodies corresponding to cells with neuronal morphology observed under DIC illumination. Calcium responses within these regions of interest were observed as changes in fluorescence ratio (340/380) upon sequential stimulation with muscimol (100 μM, Worthington) and KCl (30 mM) in Tyrode’s solution. Test solutions were washed out with Tyrode’s solution following each stimulus. A response was considered positive if the change in fluorescence ratio exceeded six times the standard deviation in baseline values. With this criterion, cells were scored as responsive or non-responsive to a specific stimulus. More than 25 cells were tested and analyzed in each experiment.

2.7 Stem cell derived Retinal ganglion-like cell transplantation

hESC-differentiated RGC-like cells were intravitreally transplanted into adult rats, following our previous protocol (Hertz et al., 2014; Venugopalan et al., 2016). Briefly, cells were centrifuged at 80 rcf for 10 minutes and resuspended in 100 μl of FS medium. 4 μl of cell suspension (~50,000 cells) or vehicle (medium) was injected into the vitreous of anaesthetized rats using a 33-gauge Hamilton syringe. The rats were sacrificed 7 days after transplantation.

2.8 Immunostaining

In vitro-differentiated cells were fixed in 2% paraformaldehyde (PFA) for 20 minutes and then permeabilized with 0.1% Triton X-100 (Sigma) for 10 min, blocked with 10% normal goat serum (NGS; Sigma) in PBS for 1 h, and incubated with mouse monoclonal primary antibody anti-Brn3b (1:300; Santa Cruz) or rabbit polyclonal antibodies anti-Nestin (1:600; Cell signaling technology), anti-Ki67 (1:500; Abcam), anti-Pax6 (1:400; Covance), anti-Rx (1:200; Abcam), anti-synaptophysin (1:500; Sysy), anti-Nanog (1:500; Cell signaling technology), anti-Tra-1-60 (1:500; Cell signaling technology), anti-β-III-Tubulin (1:200; Abcam) overnight at 4°C. The cells were rinsed with PBS and incubated with fluorophore-conjugated secondary antibodies overnight, and then analyzed by inverted microscopy imaging.

After in vivo transplantation, rats were sacrificed and perfused with PBS followed by 4% PFA. Eyes were dissected and fixed in 4% PFA for 3 hours, then dissected and flat-mounted on glass slides. The flat-mounted samples were permeabilized with 0.3% Triton X-100 (Sigma) for 20 min, blocked with 10% normal goat serum (NGS; Sigma) in PBS for 1 h, and then incubated with a mouse monoclonal anti-human nucleus primary antibody (1:200; Millipore), RBPMS (1:500; PhosphoSolution) and β-III-Tubulin (1:200; Abcam) overnight at 4°C. Retina samples were rinsed three times with PBS and incubated with Alexa-Fluor 488-conjugated secondary antibody (1:500; Life technologies) overnight. The explants were then rinsed twice, stained with DAPI (1:5000 in PBS) for 15 min, rinsed twice again, and sealed under 0.17 mm coverslips with anti-fade mounting medium (ProLong Gold, Life Technologies) before confocal microscopy imaging.

2.9 Statistical analyses

For qRT-PCR, one-way ANOVA and post-hoc t-test with Tukey correction were used to analyze gene expression differences between stem cells and differentiated progeny. P < 0.05 was considered statistically significant. Means and standard errors of the mean (SEM) are shown in all graphs with at least 3 biological replicates per data point.

3.1 Differentiation from hESCs to RPCs

Following our chemically defined protocol (Fig. 1A), we first asked if RPCs could be differentiated from hESCs (H7-A81) that expressed stem cell markers including Tra-1-60 and Nanog (Fig. 1B). After differentiation for 15 days, the cells changed their morphology from growth in tight clusters to spread out single cells, and expressed markers Rx, Nestin and Pax6, and also a proliferation maker, Ki67 (Fig. 1C, D), consistent with RPC differentiation [7]. We quantified this differentiation with single- and double-label markers, and found at 15 days in retinal progenitor induction media, 61% of progeny were Ki67+/Rx+, and 83% were Nestin+/Pax6+ in H7-A81 (Fig. 1E); 52% of progeny were Ki67+/Rx+, and 81% were Nestin+/Pax6+ in H9 (Fig. 1F) and 62% of progeny were Ki67+/Rx+, and 90% were Nestin+/Pax6+ in iPSC (Fig. 1G). Thus this chemically defined approach promotes differentiation of hESCs into RPCs.

OPEN IN VIEWER

Fig.1

Chemically defined protocol for hESC-to-RPC-to-RGC differentiation. (A) hESCs were differentiated into RGCs using a retinal progenitor induction protocol for 15 days and then an RGC differentiation protocol for another 15 days. (B) hESCs express stem cell markers, Tra-1-60 and Nanog. (C, D) hESC-derived progeny express RPC markers such as Rx, Nestin and Pax6 as well as cell proliferation marker, Ki67. All cells were counterstained with DAPI for nucleus identification. (E-G) Quantification of the fraction of cell progeny expressing RPC genes at 15 days from H7-A81 (E), H9 (F) and iPSCs (G), as marked. Scale bars = 200 μm.

Next, we asked whether hESC-derived RPCs were able to differentiate into functional RGC-like cells. After culturing RPCs for 15 days with RGC culture medium supplemented with BDNF, CNTF, and DAPT, the cells became morphologically differentiated with extensive neurite growth, and expressed neuronal markers Brn3b, and a neurite marker, β-III-Tubulin (Fig. 2A), as well as a synapse marker, synaptophysin (Fig. 2B). After 15 days in this differentiation medium, 9% of progeny were Brn3b+/β-III-Tubulin+. Thus withdrawal of RPC-promoting media and additional of media that supports differentiation and survival of RGC progeny promotes RGC differentiation as assessed by multiple markers.

OPEN IN VIEWER

Fig.2

Differentiation of hESCs to RGC-like cells. (A-C), hESC-derived RPCs were further differentiated to RGC-like cells and express neuron markers such as β-III-Tubulin and Brn3b (A), or β-III-Tubulin and synapse marker, synaptophysin (B). (C) qRT-PCR was performed to detect markers of proliferation (Ki67), stem cells (Nanog), RPCs (Rx and Pax6), RGC-like cells (Brn3a and Thy1) and photoreceptor precursor cells (Recoverin). N = 3, ^*P < 0.05, One-way ANOVA post-hoc t- test with Tukey correction. Scale bars = 50 μm (A), 100 μm (B).

We further investigated the expression patterns of genes associated with changing differentiation stages by qRT-PCR (Fig. 2C). First, the cell proliferation marker, Ki67, was significantly reduced in the transition from RPC to RGC. Second, the stem cell marker, Nanog, was undetectable after the hESC stage in either RPCs or RGCs. The RPC marker, Rx, was expressed in RPCs but was not detected in hESCs and was significantly downregulated upon RGC differentiation. Finally, increasing gene expression of Pax6, Brn3a and Thy1 was observed as hESCs progressed through RPC and particularly into RGC differentiation stages, consistent with the importance of Pax6 in particular on RGC differentiation in vivo (Kayama et al., 2010). In addition, Ki67 as a marker of proliferating cells went down by half, and we did not detect any Recoverin expression at the RGC differentiation stage, suggesting that photoreceptors were not differentiated in this protocol, or at least not at this time point. For the purposes of this paper we did not undertake further exploration of what cell progeny other than progenitor cells and RGCs might be generated in these cultures. Thus, qRT-PCR confirms quantitatively the coordinated transitions from hESC to RPC to RGC differentiation in this culture model.

Going beyond molecular markers, we also studied the electrophysiological differentiation of these stem cell-derived RGCs, compared with primary RGCs. Newly differentiated neurons undergo several morphological and physiological changes as they mature functionally through development. One such developmental process involves a shift from excitatory depolarization to inhibitory hyperpolarization in response to the neurotransmitter gamma-aminobutyric acid (GABA). For example, early postnatal RGCs depolarize whereas mature RGCs hyperpolarize in response to GABA_A receptor activation; only the immature RGCs thus demonstrate a calcium influx in response to GABA_A-specific agonists like muscimol. To characterize this measure of electrophysiologic maturity, Ca²⁺ levels were monitored using ratiometric Fura-2 imaging in response to muscimol in RGCs differentiated from E14 mouse RPCs, in which we observed a muscimol-evoked increase in [Ca²⁺]_i similar to that seen in KCl-induced depolarization (12 of 16 cells; Fig. 3A). Similarly, muscimol stimulated a depolarization-induced increase in [Ca²⁺]_i in hESC-derived RGCs (10 of 12 cells; Fig. 3B). Thus after differentiation in vitro at this time point, hESC-derived RGC-like progeny are physiologically immature. For our hESC RGCs, we could specifically image Brn3-mCherry+ cells. However for the E14 mouse retinal progenitor-derived RGCs, we did not have that option. It is possible that some of the “GABA immature cells” are other non-RGC neurons, but given that the majority of the progeny were thought to be RGC-like, this likely does not affect themain conclusion.

OPEN IN VIEWER

Fig.3

Calcium imaging of rodent RPC-derived and hESC-derived RGCs. (A-B) Cultures of E14 mouse RPC-derived RGCs (A) or hESC-derived RGC (B) were loaded in culture with calcium indicator Fura-2, and then treated with muscimol and KCl as indicated in the example traces. Calcium-evoked response was monitored in real time selecting large nuclei and mCherry-positive RGC progeny in the RPC-derived and hESC-derived cultures, respectively. Scale bars = 20 μm (A), 40 μm (B).

Finally, we asked if the hESC-derived RGC-like cells survive in vivo and migrate into the ganglion cell layer (GCL) after intravitreal transplant in adult rat eyes. Seven days after injection, a significant number of transplanted hESC-derived RGCs were localized in the GCL, as identified by expression of a human nuclear marker that was not detectable in control medium-injected eyes (Fig. 4A-C). Although in these experiments the RGC marker RNA-binding protein with multiple splicing (RBPMS) was not detectable in donor human cells, a subset of donor cells that remained on the retinal surface or in the vitreous clearly expressed β-III-Tubulin (Fig. 4C). In terms of quantifying efficacy, we attempted the transplants into 5 eyes and found cells were detectable in all 5 attempts. Although some areas of the retina naturally had no cells detectable, in retinal sections where donor cells were found, the flatmount density ranged from 19 to 25 cells/mm² (Fig. 4D), and the cross section donor cells ranged from 3.3 to 25 donor cells per linear mm of retinal section. Thus at least in these initial studies, hESC-derived RGC-like cells survive and migrate into the GCL after intravitreal transplant into the normal adult rodent eye.

OPEN IN VIEWER

Fig.4

Differentiated RGC-like cells survive in the adult retina after transplant. Vehicle (O.S. Control) or hESC-derived progeny (O.D. Transplant) were intravitreally injected into 6-week-old rat eyes. One week after transplant, rats were sacrificed and cryosections (A-C) and flatmount retinas (D-E) were probed with human nuclei, RBPMS and β-III-Tubulin, as marked. White arrows indicate ganglion cell layer (GCL). Scale bars = 200 μm (A-C), 50 μm (D-E).