Brain Research 1014 (2004) 131 – 144 www.elsevier.com/locate/brainres

Research report

Neural differentiation of mouse embryonic stem cells in vitro and after transplantation into eyes of mutant mice with rapid retinal degeneration Jason S. Meyer a, Martin L. Katz b, Joel A. Maruniak a, Mark D. Kirk a,* a b

Division of Biological Sciences, 103 Lefevre Hall, University of Missouri, Columbia, MO 65211, USA Department of Ophthalmology, Mason Eye Institute, University of Missouri, Columbia, MO 65212, USA Accepted 3 April 2004

Abstract Embryonic stem (ES) cells can differentiate into many specialized cell types, including those of the nervous system. We evaluated the differentiation of enhanced green fluorescent protein (EGFP)-expressing B5 mouse ES cells in vitro and in vivo after transplantation into the eyes of mice with hereditary retinal degeneration. After neural induction with retinoic acid, the majority of cells in embryoid bodies expressed markers for neural progenitors as well as for immature and mature neurons and glial cells. When induced ES cells were plated in vitro, further differentiation was observed and the majority of cells expressed h-III Tubulin, a marker for immature neurons. In addition, many plated cells expressed markers for mature neurons or glial cells. Four days after intravitreal transplantation into the eyes of rd1 mice (a model of rapid retinal degeneration), donor cells appeared attached to the vitreal surface of the retina. After 6 weeks in vivo, most transplanted cells remained adherent to the inner retinal surface, and some donor cells had integrated into the retina. Transplanted cells exhibited some properties typical of neurons, including extensive process outgrowth with numerous varicosities and expression of neuronal and synaptic markers. Therefore, after induction B5 ES cells can acquire the morphologies of neural cells and display markers for neuronal and glial cells in vitro and in vivo. Furthermore, when placed in the proper microenvironment ES-derived neural precursors can associate closely with or migrate into nervous tissue where differentiation appears to be determined by cues provided by the local environment, in this case the degenerating neural retina. D 2004 Elsevier B.V. All rights reserved. Theme: Development and regeneration Topic: Cell differentiation and migration Keywords: Stem cell; Differentiation; Transplant; Retina; rd1 mouse

1. Introduction Stem cells are undifferentiated, primordial cells capable of two types of self-renewal. They can divide symmetrically to form copies of themselves [15,28] or they can divide asymmetrically to give rise to additional clonal stem cells and one or more specialized cell types [14,24,40]. Mammalian embryonic stem (ES) cell lines are derived from the inner cell mass of the blastocyst prior to its implantation in the uterus [38,41]. Clonal ES cell lines have been estab* Corresponding author. Division of Biological Sciences, 103 Lefevre Hall, University of Missouri – Columbia, Columbia, MO 65211, USA. Tel.: +1-573-882-6507; fax: +1-573-884-5020. E-mail address: [email protected] (M.D. Kirk). 0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2004.04.019

lished for several mammals, including mice and humans [18,38,45]. Embryonic stem cells are capable of forming many specialized cell types of the organism, including cells that constitute all three embryonic germ layers [3,26,29] and as such, they are termed pluripotent. Embryonic stem cells can proliferate extensively in an undifferentiated state, are accessible for genetic manipulation, and can be induced to differentiate into neural cells (a process referred to as neuralization) [16,51]. Stem cells, including ES cells, can respond to environmental cues by differentiating into unique cell types [32]. An important issue is how different ES cell lines respond to environments within the central nervous system (CNS) after transplantation and whether their fate can be modified to achieve specific cellular phenotypes in vivo [39,54,57].

132

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

Mouse ES cells have been induced to form a number of different cell types, including dopaminergic and serotonergic neurons [21] as well as insulin-secreting pancreatic isletlike cells [26]. Typically, induction is carried out when the ES cells are grown in spherical clusters known as embryoid bodies (EBs). ES cells are also capable of generating Nestinpositive neural precursors that can incorporate into a mouse blastocyst, ultimately contributing to all embryonic germ layers [46]. In this work we used the B5 ES cell line, which expresses enhanced green fluorescent protein (EGFP) [19] that enables one to distinguish between transplanted cells and cells of host origin. Several methods exist to differentiate ES cells in vitro and to direct their fate to that of neuron-like and glial-like cells. However, there is currently no known method to generate a homogenous population of a specific cell type after induction of ES cells in vitro. For instance, after retinoic acidmediated neural induction, mouse ES cells can generate neuron-like, astrocyte-like, and oligodendrocyte-like cells both in vitro and in vivo [13,27]. Similarly, induction protocols applied to tissue-derived neural stem cells produce heterogeneous populations of cells. Neuronal progenitor and glial progenitor cells are derived from neural stem cells, have more restricted fates, and differentiate into specialized neuron-like and glial-like cell types, respectively [40]. To test the response of neuralized B5 ES cells in the environment of nervous system, we used the rd1 (Pde6brd1) mouse, an important model for studies of rapid retinal degeneration [7]. The rd1 mouse is characterized by photoreceptor cell degeneration due to a mutation in the gene encoding the h-subunit of rod photoreceptor cyclic GMP phosphodiesterase [30,35]. This mutation results in the disappearance of rod photoreceptors from the retina shortly after birth (by 5 weeks postnatal only a few cone photoreceptors remain), while the remaining neural retina endures largely intact [6,7,17,34]. Neural stem cells fail to integrate within the retina after injections into the vitreous of healthy rodents, while neural stem cells transplanted in the same way into rodents with retinas lesioned mechanically and/or with inherited retinal degeneration exhibit retinal incorporation [25,33]. Transplanted neural stem cells can also integrate within normal immature neural retinas [43,48]. The rd1 mouse retina provides a good model to study fate determination of transplanted stem cells within the environment of the nervous system as well as to test the ability of transplanted neural stem and progenitor cells to survive and integrate within a CNS structure compromised by injury and/or degeneration. In the present study, we quantify and compare the number and types of neuronal-like and glial-like cells derived from mouse B5 ES cells in vitro after neural induction. In addition, we assess the survival and fate in vivo of neuralized ES cells after transplantation into the vitreous of the rd1 mouse eye. Some of this work has appeared in abstract form [31].

2. Materials and methods 2.1. Cell culture Mouse B5 embryonic stem cells, expressing enhanced green fluorescent protein (EGFP), were used in all experiments (the B5 ES cell line was kindly provided by Dr. Andras Nagy, Mt. Sinai, Toronto). Undifferentiated ES cell cultures were maintained and expanded in the presence of leukemia inhibitory factor (LIF; Chemicon, Cat. #ESG1106) as described previously [2,5,42]. The cells were induced using the 4 /4+ protocol developed by Bain and colleagues [2]. Briefly, undifferentiated ES cells were grown in gelatin-coated T25 tissue culture flasks in ES cell growth medium (ESGM). ESGM contained DMEM, 10% newborn calf serum, 10% fetal bovine serum, nucleosides, h-mercaptoethanol (1 mM), and LIF (1000 U/ml). The 4 /4+ protocol was performed by growing the ES cells as unattached embryoid bodies (EBs) for 4 days in uncoated Petri plates using ES cell induction medium (ESIM). ESIM was similar to ESGM but lacked h-mercaptoethanol and LIF. The EBs were then cultured for an additional 4 days in the presence of retinoic acid (all-trans retinoic acid, 500 nM, Sigma Cat. #R2625). For transplantation experiments, EBs were dissociated to a nearly single cell suspension on the 8th day in culture using 0.25% trypsin with EDTA (8 min, 37 jC) and mild trituration. For plating induced ES cells, the EBs were dissociated on the 9th day, seeded onto either gelatin-coated 24-well plates or ECL (Upstate Biotechnologies)-coated, 8-well culture slides and fed every 2 days with ESIM for up to 8 days in vitro [22,27]. 2.2. Animal care and ES cell intravitreal transplantation rd1 (Pde6brd1) and C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME). At 5 weeks of age, mice were anesthetized with intraperitoneal injections of 80 mg/kg ketamine, 8 mg/kg xylazine, and 1.6 mg/kg acepromazine. The neuralized B5 ES cells were concentrated to 30,000 cells/Al, and 1.5 Al of the cell suspension was transplanted into the vitreous of each eye using a 10Al Hamilton syringe and a 31-gauge needle. Mice were allowed to recover from the anesthesia and were examined for donor cell survival and differentiation at 4 days and 6 weeks post-transplantation. All animal experiments were approved by the University of Missouri – Columbia Animal Care and Use Committee and were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 2.3. Preparation of eyes for transmitted light, epifluorescence, and electron microscopy To assess the fate of transplanted cells, retinas were prepared as either unfixed whole mounts or in sections of fixed tissue. Eyes were enucleated immediately after the

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

animals were euthanized via carbon dioxide inhalation. For whole mounts, neural retinas were dissected from the eyes and mounted as fresh tissue in 0.17 M sodium cacodylate, pH 7.4 on glass slides. Four radial cuts were made to enable the retina to lie flat on the slide before a coverslip was applied. Retinas were mounted on slides with the vitreal surface facing up. To obtain retinal sections for epifluorescence microscopy, eyes were enucleated and the corneas, irises, and lenses were removed immediately. The remaining posterior portions of the eyes were fixed in 4% paraformaldehyde, 50 mM sodium cacodylate and 8% sucrose (pH = 7.4) for 60 min. The eyecups were examined with an epifluorescence stereomicroscope and were bisected through regions containing the greatest concentrations of EGFP-expressing cells. The isolated retinas were then washed in 0.17 M sodium cacodylate buffer, equilibrated in 20% sucrose and embedded in OCT (Tissue Tek) for frozen sections. To view EGFP-expressing donor cells, cross-sections of the embedded tissues were cut at a thickness of 8 Am by using a cryostat. A minimum of five eyes was examined as whole mounts, and at least an additional five eyes were examined as cryostat sections for viewing EGFP expression at each time post-transplantation. To obtain retinal sections for light and electron microscopy, the eyes were placed immediately in a fixative consisting of 1.25% glutaraldehyde, 2% paraformaldehyde, 0.13 M sodium cacodylate, and 0.13 mM CaCl2 at pH 7.4. With the eyes immersed in fixative, the corneas, irises, and lenses were removed. The remaining eyecups were fixed for at least 2 h at room temperature with gentle agitation. As with sections for epifluorescence microscopy, eyecups were examined with an epifluorescence stereomicroscope and were bisected through regions containing the greatest concentrations of EGFP-expressing cells. Following dissection, the samples were washed with 0.17 M sodium cacodylate (pH 7.4), post-fixed with osmium tetroxide and uranyl acetate and embedded in epoxy resin. For light microscopy, 1-Am-thick sections were obtained from the regions containing the EGFP-expressing cells. The sections were stained with Toluidine blue and were examined with conventional transmitted light microscopy. Ultrathin sections (70 – 90 nm thick) were obtained from the same regions of the specimens (in three separate preparations), stained with uranyl acetate and lead citrate, and were examined with a JEOL 1200 EX transmission electron microscope (TEM). In one preparation, thick frozen sections were viewed for EGFP expression, and then the retina was fixed for electron microscopy and immediately adjacent ultrathin sections were obtained and examined with TEM. 2.4. Immunocytochemistry Embryoid bodies were pelleted and fixed for 45 min at room temperature in 4% paraformaldehyde in 0.1 M PBS (PBS = 137 mM NaCl, 2.7 mM KCl, 5.4 mM Na2HPO4,

133

0.56 mM KH2PO4, pH = 7.4). The EBs were then equilibrated in 25% sucrose in 0.1 M PBS, embedded in OCT and frozen. For plated cells, the cultures were fixed for 30 min in 4% paraformaldehyde and 0.1 M PBS (pH = 7.4) and washed three times for 5 –10 min per wash with 0.1 M PBS. Cryostat sections of retinas and EBs were taken at 8and 10-Am thicknesses, respectively, mounted on SuperFrost slides, washed in 0.1 M PBS for 15 min, and permeabilized for 1 h at room temperature using 0.3% Triton X-100, 0.1 M PBS, and 10% normal goat serum. Primary antibodies specific for neural precursors (mouse monoclonal Nestin antibody, 1:100, DSHB, Cat. #Rat401), immature neurons (mouse monoclonal for h-III Tubulin, 1:100, Promega, Cat.# G7121, or mouse monoclonal for Map2ab, 1:200, Sigma-Aldrich, Cat.# M1406), mature neurons (mouse monoclonal for NF-M, 1:200, Zymed, Cat.# 34-1000, or mouse monoclonal for NeuN, 1:10, Chemicon, Cat.# MAB377), oligodendrocytes (mouse monoclonal for O4, 1:50, Chemicon, Cat.# MAB345) and astrocytes (rabbit polyclonal for GFAP, 1:100, Sigma-Aldrich, Cat.# G-9269) were diluted in 0.1 M PBS containing 1% normal goat serum and 0.5% Triton X-100 and applied overnight at 4 jC. Appropriate fluorescence-tagged goat anti-mouse secondary antibodies (Jackson ImmunoResearch Laboratories) and goat antirabbit secondary antibody (Molecular Probes) were diluted 1:200 in 0.1 M PBS containing 10% normal goat serum and 0.3% Triton X-100 and applied individually for 3– 4 h at room temperature. Plated cells were permeabilized for 30 min with a solution of 0.1 M PBS, 2% normal goat serum, 1% bovine serum albumin and 0.01% Triton X-100, then washed thoroughly with 0.1 M PBS containing 2% normal goat serum and 1% bovine serum albumin; the latter solution was also used to dilute all antibodies. As for labeling of retinal and EB sections described above, primary antibodies specific for neural precursors, neurons, oligodendrocytes, and astrocytes were used to label the plated cells. Individual primary antibodies were applied overnight at room temperature and then exposed to an appropriate fluorescencetagged, goat anti-mouse or goat anti-rabbit secondary antibody (1:200) for 3 – 4 h at room temperature. Fluorescent images were obtained by using one or more of three epifluorescent image capture systems. The first system was a BioRad Radiance 2000 Confocal microscope (BioRad, Hercules CA) equipped with an Olympus IX70 microscope. The software for this system was BioRad Lasersharp 2000. The second system was equipped with an Olympus IX70 microscope but a Photometric Sensys CCD camera was used. The software for the latter system was Image Pro Plus Version 4.1 by Media Cybernetics (Silver Spring, MD, USA). Within Image Pro, Vay Tek Volume Scan Version 3.1 by Vay Tek, was used for capturing images. The third system was a Zeiss Axiophot microscope equipped for epifluorescence, and fluorescent

134

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

emissions were stimulated with light from a 50-W, highpressure mercury vapor source. In the latter system, examination and photography of the specimens was performed using a 40  Plan-Neofluor objective with a 1.30 numerical aperture. Photography using the latter system was performed with Kodak EliteChrome 100 ASA film and the color slides were digitized for data analysis and reproduction. 2.5. Quantification of cultured and transplanted ES cells Counts of different cell phenotypes found in whole EBs, after plating of cells from dissociated EBs, and following intravitreal transplantation were quantified after immunolabeling as described above. Frozen sections of EBs and retinas were used, and nuclei were stained with Hoechst 33358. The presence of a Hoechst-labeled nucleus in an EGFP-expressing cell enabled us to distinguish between adjacent donor cells and to obtain the total number of donor cells found in each section. The number of cells expressing a particular cell marker was counted and compared to the total number of cells found in each section as determined by Hoechst-labeled nuclei. The total number of plated cells was obtained by counting the total number of EGFP-positive cells in each well of 24-well plates. Again, the cells were labeled with antibodies specific for either neural precursors, neurons, oligodendrocytes, or astrocytes as described above. The number of immunoreactive cells in each well was counted and compared to the total number of EGFPexpressing cells found in the same well.

neural precursors, neurons or glia. Immunolabeling of embryoid bodies (EBs) collected immediately after induction (Fig. 1) revealed a large percentage of cells labeled for Nestin, a marker for neural precursors (Fig. 1B, Table 1). In adjacent sections, numerous cells were immunopositive for neuronal markers, such as h-III Tubulin and Map2ab, while fewer cells labeled for the mature neuronal markers Neurofilament-M (NF-M) and NeuN (Fig. 1C – F, Table 1). Relatively few cells in the EBs labeled for the astrocyte marker GFAP or the oligodendrocyte marker O4 (Fig. 1G and H, Table 1). The mean percentage of cells expressing either Nestin or the neuronal markers h-III Tubulin, Map2ab or NF-M in EBs was significantly greater than that for either glial marker. Control experiments in which

2.6. Statistical analysis of samples Data were analyzed using SAS with assistance from the University of Missouri – Columbia Statistical Consulting Service or using SPSS on an iMAC G3. The sample size (i.e., number of separate preparations) for quantification of immunolabeling in EBs, plated cells and retinal sections was three. A chi-square test for homogeneity of proportions was used for all statistical tests. In cases where the number of positively labeled cells was small, Fisher’s exact test was used as a second method to confirm significant differences. Data in the text are expressed as mean F S.E., and significance levels stated in the text are a minimum of p < 0.01. N values given in text represent the number of different eyes tested.

3. Results 3.1. Embryoid bodies and plated cells derived from neuralized B5 ES cells display markers for neural precursors, neurons and glial cells ES cells responded to retinoic acid induction by developing modified morphologies and expressing markers for

Fig. 1. Cells in embryoid bodies (EBs) on day 8 of the 4 /4+ induction protocol express markers for neural precursors, neurons and glia. (A) EBs derived from ES cells express enhanced green fluorescent protein (EGFP). (B – H) EB sections were labeled with antibodies against neural precursors [anti-Nestin (B)], immature neurons [anti-h-III Tubulin (C) and antiMap2ab (D)], mature neurons [anti-NF-M, (E) and anti-NeuN (F)], astrocytes [anti-GFAP (G)], and oligodendrocytes (anti-O4 (H)]. Arrowheads in (F) indicate nuclei co-labeled for Hoechst and NeuN. No labeling was observed in any EBs when the primary antibody was omitted. All images in (A – H) were taken from frozen sections cut at 8 Am. Scale bar in (H) is 200 Am and applies to all panels.

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144 Table 1 Mean percentages of B5 cells labeled for specific neural markers in vitro (EBs and plated cells) and after intravitreal transplantation (retinal sections) Stage

Marker

Total no. of cellsa

# Labeled

Mean (%) F S.E.

Embryoid bodies (EBs)

Nestin h-III Tubulin Map2ab NF-M NeuN GFAP O4 Nestin h-III Tubulin Map2ab NF-M NeuN GFAP O4 Nestin h-III Tubulin Map2ab NF-M NeuN GFAP O4 Nestin h-III Tubulin Map2ab NF-M NeuN GFAP O4

4453 4852

2448 1829

55.2 F 1.1 37.6 F 0.7

3915 4993 4006 4288 4076 3397 3515

1347 1523 385 305 280 115 1951

34.4 F 2.1 30.5 F 1.2 9.6 F 1.8 7.2 F 0.2 7.0 F 0.3 3.6 F 0.4 56.4 F 2.5

4184 2947 3007 3384 3293 297 299

1628 984 842 1309 137 17 270

38.9 F 1.9 33.4 F 2.3 28.0 F 1.4 38.7 F 0.7 4.1 F 0.6 5.7 F 0.3 90.4 F 1.0

184 196 208 298 294 281 283

137 94 45 7 0 13 262

74.5 F 2.4 48.0 F 2.1 21.6 F 1.0 2.3 F 0.2 0.0 F 0.0 4.6 F 0.1 92.5 F 0.8

263 228 240 286 276

205 117 49 23 0

77.9 F 1.7 51.3 F 1.5 20.4 F 0.8 8.0 F 0.6 0.0 F 0.0

8 days in vitro (plated cells)

4 days posttransplantation (retinal sections)

6 weeks posttransplantation (retinal sections)

135

were accompanied by a significant decrease in the percentage of cells expressing Nestin (Fig. 2B, Table 1), relative to that observed in EBs. Similar percentages of Map2ab positive and NF-M positive cells were observed after plating (Fig. 2D and E), when compared to that observed for EBs (Table 1). Control experiments in which either the primary or secondary antibody was excluded from the protocol revealed only background labeling (data not shown).

a

The total numbers of cells sampled for EBs and retinal sections were determined using nuclear labeling with Hoechst 33358; the total numbers of plated cells were determined based on visualization of EGFP expression.

either the primary antibody or secondary antibody was omitted revealed only background fluorescence (data not shown). In subsequent experiments, retinoic acid-induced EBs were dissociated and the cells plated and cultured for 8 days in vitro (Fig. 2). Among cells that had been induced and plated, a majority expressed the neuronal marker h-III Tubulin (Table 1). This level of expression was significantly greater than the expression of all other markers tested. Substantial expression of Map2ab, NF-M, NeuN and GFAP by plated cells was also observed, and in all cases was significantly greater than the expression of either Nestin or O4 (Table 1). After plating the induced B5 cells, there was a significant increase in the mean percentage of cells expressing h-III Tubulin compared to those in the undissociated EBs (Fig. 2C, Table 1). The percentages of NeuN-positive (Fig. 2F) and GFAP-positive (Fig. 2G) cells also increased significantly after plating (Table 1). The increases in hIII Tubulin- and GFAP-positive labeling of plated cells

Fig. 2. Neuralized ES cells display neural markers in vitro. After induction with retinoic acid, embryoid bodies were dissociated and the cells plated in gelatin-coated dishes or ECL-coated glass culture slides for 8 days. (A) After 8 days in vitro, the cells continue to express EGFP and some cells have extended numerous processes. (B – H) The plated cells display markers for neural precursors [anti-Nestin (B)], immature neurons [anti-hIII Tubulin (C) and anti-Map2ab (D)], mature neurons (anti-NF-M (E) and anti-NeuN (F)], astrocytes [anti-GFAP (G) and oligodendrocytes (anti-O4, H)]. Note that clustering of GFAP-labeled cells (as shown in G) was frequently observed, while no such trend was observed for cells positively labeled for the other markers. In panel (F), EGFP expression is shown to indicate the morphology of cells, and in all panels, nuclei were stained with Hoechst 33358 (Blue). Control experiments in which the primary antibody was omitted revealed no labeling. Scale bar in (H) equals 40 Am and applies to all panels.

136

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

3.2. Degeneration of the rd1 retina at 5 weeks of age As reported previously [6,7,17,34], degeneration of the rd1 retina is rapid, beginning at approximately 3 weeks postnatal and complete by 7 weeks of age. We observed that in contrast to retinas of genetically normal C57BL/6J mice at 5 weeks of age, age-matched rd1 retinas were devoid of any morphologically recognizable rod photoreceptors. However, in the rd1 retina at 5 weeks postnatal, the ganglion cell layer, inner plexiform layer, and inner nuclear layer remained largely intact (Fig. 3). 3.3. Neuralized B5 ES cells survive, differentiate and can integrate within the neural retina To determine whether induced B5 ES cells can acquire a neural fate in vivo and integrate into the retina, neuralized ES cells were injected into the vitreous of 5week-old rd1 mice. Host eyes were removed at either 4 days or 6 weeks post-transplantation and the retinas viewed in whole-mount preparations to assess the distribution of donor cells along the retina. At 4 days posttransplantation (Fig. 4A), EGFP-expressing donor cells were found as distinct clusters of flattened cells in the vitreous, in close proximity to the retina. We also examined cryostat sections taken of retinas at 4 days post-transplantation. At this early time point, most of the donor cells were found within the vitreous fluid close to the inner surface of the retina (Fig. 4B). At 4 days posttransplantation, we did not observe donor cells that had integrated within the retina proper.

By 6 weeks post-transplantation, most of the donor cells were attached to the inner retinal surface (Fig. 5). Based on observations from whole-mount preparations, they were more widely distributed across the retina than at 4 days post-transplantation. At 6 weeks post-transplantation, some clusters of the donor cells still possessed flattened morphologies (Fig. 5A), though most donor cells possessed neuronal-like morphologies, extending numerous fine processes along the retina and creating an apparent network of cells (Fig. 5C,E). Many of the donor cell processes possessed numerous varicosities (Fig. 5C – F). At the ultrastructural level, electron micrographs confirmed that many of transplanted cells were adherent to the inner limiting membrane of the retina (Fig. 6A) which remained intact. Some donor cells extended long processes along the surface of the retina (Fig. 6B). As in normal C57BL/6J mouse eyes, eyes from sham-injected, control rd1 mice displayed a retinal ultrastructure with no cells found within the vitreous fluid or on the vitreal side of the inner limiting membrane (Fig. 6C). We obtained serial thick and ultrathin sections of host retina from regions where clusters of EGFP-expressing B5 donor cells were present. Cells within the same donor cell cluster that exhibited EGFP fluorescence in thick sections were found in immediately adjacent, ultrathin sections. Electron micrographs revealed that these donor cells were localized on the vitreal side of the inner limiting membrane and displayed morphologies similar to those shown in Fig. 5. Also, at both 4 days and 6 weeks post-transplantation, all cells observed on the

Fig. 3. Photoreceptors degenerate in retinas of the rd1 mice by 5 weeks of age. (A) A normal C57BL/6J mouse retina displays the full complement of photoreceptors and retinal layers. (B) The rd1 retina degenerates rapidly and extensively by 5 weeks postnatal, with selective loss of virtually all rod photoreceptors by this age. Note that the outer plexiform layer and outer nuclear layer are virtually absent in the rd1 mouse at this age. GCL = ganglion cell layer, IPL = inner plexiform layer, INL = inner nuclear layer, OPL = outer plexiform layer, ONL = outer nuclear layer, RPE = retinal pigment epithelium. These abbreviations apply to subsequent figures. Retinal sections were stained with Toluidine blue. Scale bar in (B) also applies to (A).

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

137

3.4. Transplanted, neuralized ES cells display neuronal, glial and synaptic markers

Fig. 4. At 4 days post-transplantation, donor cells derived from neuralized B5 ES cells were located in the vitreous or associated with the inner retina but for the most part have not integrated into the neural retina. (A) When viewed in retinal whole mounts, neuralized B5 cells expressing EGFP formed distinct clusters. Some degree of process formation is evident by the donor cells. (B) When viewed in cryostat sections, donor cells were found either adjacent to the vitreal side of the inner limiting membrane or floating nearby in the vitreous. Streams of cells were often found within the vitreous in whole mounts at this early time post-transplantation, using a fluorescence stereomicroscope (data not shown). Little evidence of integration into the retina was seen at this early time point. The cross-section in (B) was taken in the region containing the optic nerve head (ONH), delineated by the dashed line. The neural retina (R) and associated layers are labeled. Calibration in (B) equals 100 Am and also applies to (A).

vitreal surface of the host retina viewed with phase contrast microscopy, expressed EGFP fluorescence. Therefore, these results indicate that all cells found at the electron microscopic level on the vitreal surface of the inner limiting membrane are donor B5 cells. Also at 6 weeks post-transplantation, a small fraction of the transplanted cells had integrated into the retina and were located largely in the ganglion cell or inner plexiform layers (Figs. 5D,F and 7). Like cells on the inner limiting membrane, cells that had integrated within the retina displayed neuron-like morphologies, with abundant processes that projected horizontally and exhibited numerous varicosities (Figs. 5F and 7A, F). Very rarely were transplanted cells observed in the inner nuclear layer.

Immunolabeling of retinas at 4 days and 6 weeks posttransplantation revealed that the donor cells express general neural markers. At 4 days post-transplantation, a large majority of donor cells labeled positive for the markers of immature neurons, h-III Tubulin or Map2ab (Table 1). At this same time post-transplantation, separately labeled sections revealed large percentages of donor cells expressing the mature neuronal markers, NF-M or NeuN (Table 1). A small percentage of donor cells in retinal sections at 4 days post-transplantation were immunoreactive for Nestin or GFAP, while no expression of the oligodendrocyte marker O4 was observed in any retinal sections (Table 1). The mean percentage of donor cells expressing h-III Tubulin, Map2ab or NF-M was significantly greater at 4 days post-transplantation, when compared to the prevalence of cells expressing these markers in EBs or in plated cells after 8 days in vitro (Table 1). Also, at 4 days post-transplantation, there was a significantly greater percentage of cells labeled for Nestin compared to that for GFAP. The percentages of cells labeled for glial markers at 4 days posttransplantation was significantly lower than that observed for induced cells plated and grown in vitro (Table 1). No cells at 4 days post-transplantation labeled positively for O4, consistent with the absence of oligodendrocytes in the normal retina [10]. Similar patterns of immunoreactivity were observed at 6 weeks post-transplantation, (Figs. 5 and 7), and most donor cells expressed h-III Tubulin or Map2ab (Fig. 7A and B, Table 1), while many donor cells expressed more mature neuronal markers such as NF-M or NeuN (Fig. 7C and D, Table 1). At 6 weeks post-transplantation, the percentage of cells expressing any neuronal marker was significantly greater than the labeling seen for Nestin or GFAP. Only a small fraction of EGFP-expressing donor cells labeled positive for GFAP (Fig. 7E, Table 1) or the neural precursor marker Nestin in cryostat sections (Table 1). As observed at 4 days post-transplantation, no evidence of O4 immunoreactivity was present in host retinas at 6 weeks post-transplantation. The percentages of donor cells expressing glial markers were significantly less than that for cells plated and grown in vitro (Table 1). We found no evidence for colocalization of glial markers (GFAP, O4) and neural precursor/neuronal markers (Nestin, h-III Tubulin) in cells in vitro (i.e., EBs and plated cells) or in donor cells within the host eye (at 4 days and 6 weeks post-transplantation—data not shown; N = 4 for each condition). We also tested for expression of the synaptic vesicle protein SV2 (Fig. 7F), using immunocytochemistry (N = 3 at both 4 days and 6 weeks post-transplantation). Labeling for SV2 was only observed in host eyes at 6 weeks posttransplantation. The host IPL exhibited extensive expression of SV2, consistent with the high density of synapses located in this plexiform layer. In addition, SV2 immuno-

138

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

Fig. 5. At 6 weeks post-transplantation, donor B5 cells incorporated into the rd1 retina and assumed neuronal-like morphologies. As viewed in whole-mount samples of the retina (A, C, E), donor cells assumed varied morphologies, with some donor cells retaining a flattened morphology (A), similar to that seen at 4 days post-transplantation while other donor cells extended numerous long, fine processes across the retinal surface (C, E). Many of these long processes possessed numerous varicosities (E). As viewed in cryostat sections (B, D, F), donor cells were found closely associated with the inner limiting membrane of the retina (B). In other areas of the retina, cells had penetrated the neural retina, with some cell bodies found within the IPL (D, F). (F) Those donor cells found within the IPL (and in rare cases with cell bodies in the adjacent INL) often possessed many fine, lengthy processes that extended horizontally within the IPL and exhibited numerous varicosities. Scale bar equals 100 Am in (A – E), 50 Am in (F).

reactivity in donor B5 cells was localized to processes in the IPL and to peripheral regions of the donor cell bodies. Note in Fig. 6F that the cell body of a donor cell and several of its primary processes were located in the IPL. The latter result was typical of donor cells that had incorporated into the host retina.

4. Discussion Our results show that retinoic acid induction of B5 ES cells efficiently produces neural progenitors that can differentiate into neuron-like and glial-like cells in vitro. In contrast to other ES cell lines [37], a majority of B5 cells express neuronal markers after retinoic acid-mediated neural induction and plating in vitro. We also demonstrate that in a mouse model of rapid retinal degeneration, neuralized ES cells survive at least 6 weeks following transplantation,

associate closely with or incorporate into the degenerating retina without overtly affecting the retinal architecture, differentiate in vivo primarily into cells with neuronal-like morphologies and express largely neuronal markers, including a marker for synaptic vesicles. Therefore, the environment found within the eye of rd1 mice promotes long-term survival and greater neuronal-like rather than glial-like differentiation of transplanted, neuralized B5 ES cells, when compared to in vitro conditions. 4.1. In vitro differentiation of B5 ES cells Embryonic stem cells from several mammalian species can be induced in vitro to achieve a neural fate using a variety of protocols. For example, NGF promotes nerve cell differentiation [52], a combination of FGF2, EGF, and PDGF generates glial precursor cells [4], a combination of insulin, transferrin, selenium, and fibronectin supports dif-

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

ferentiation into dopaminergic and serotonergic neurons [21]. Stromal cell-derived inducing activity acts on mouse ES cells to generate largely dopaminergic neurons [39], and neural precursors are generated in the presence of LIF when ES cells are grown at low density [46]. Retinoic acid is an important neuralizing factor during development and efficiently induces some mouse ES cell lines to differentiate into various neural cell types, including specialized neurons and glia [2,8,12,13,22]. Using the 4 /4+ neural induction protocol, Liu et al. [22] demonstrate that ROSA26 ES cells become more glial-like in nature, based on in vitro labeling for neuronal and glial markers in sectioned EBs. In contrast, using the same induction protocol we show here that B5 ES cells differentiate primarily into neuronal-like cells, as determined by immunocytochemical labeling. The basis for this difference in response to induction treatment is not clear but may reflect an intrinsic difference between the two ES cell lines. Differences in the properties of ES cell lines are suggested by differences in their responses to neuralizing protocols as well as the varying degrees of difficulty reported in the derivation of ES cells from certain mouse strains [37,41]. As stated above, our results demonstrate that B5 ES cells generate relatively more neuronal-like cells than other mouse ES cell lines when treated with retinoic acid. It will be important in future studies to explore differences in intracellular and/or extracellular signaling that can explain the different developmental potentials of various ES cell lines. It appears that ES cells from different donors may respond differently to the same induction cues. In some applications, it may be beneficial for the cells to assume primarily a neuronal fate, whereas in other cases, one may wish to direct the fate of ES cells to that of a glial lineage. Our results show a correlation between the developmental fates of B5 ES cells in vitro and in vivo after using the 4 / 4+ induction protocol. This suggests that in vitro differentiation may predict the in vivo fate of transplanted, neuralized B5 ES cells in retina and potentially other areas of the CNS. After induction, the B5 EBs contain a large number of cells that express Nestin, a characteristic of neural progenitor cells with the potential to become neurons and/or glial cells. A large number of cells in the EBs also express h-III Tubulin, Map2ab, and NF-M, cytoskeletal markers for neurons. These observations indicate that induction of stem cells in vitro produces a heterogeneous population of cells that are primarily neuronal-like that differ in their degree of differentiation. Immediately after the 4 /4+ induction process, the large number of Nestin-positive cells indicates the existence of a population of cells whose fate is yet to be determined exclusively. For instance, Nestin is typically used as a marker for neural precursors; however, it is also expressed by precursors to pancreatic cells generated from ES cell lines (including the B5 cell line) using a different induction

139

protocol [26]. Therefore, it was of importance to determine the fate of these cells following further differentiation. In the present study, 8 days after plating the B5 cells, Nestin expression substantially decreased and expression of h-III Tubulin and GFAP greatly increased, indicating that at the end of the 4 /4+ induction protocol Nestin-positive cells differentiate primarily into neuron-like or astrocyte-like cells (in the absence of further specific cues). Therefore, Nestin expression of cells following this induction protocol is a good indicator of future differentiation to a mature neural fate. 4.2. Survival, retinal integration and phenotypic differentiation of transplanted cells Intravitreal transplantations produce a widespread distribution of donor stem cells across the retina. In fact, the vitreous (a fluid-filled chamber) supports extensive migration and long-term survival of transplanted cells [36,43]. In the present study, transplanted cells were widely dispersed across the inner retinal surface (especially at later times post-transplantation) in what appeared to be a tendency for the cells to spread out to cover as large an area of the retina as possible. However, the global distribution across the retina of donor cells was not always uniform; for instance, some areas of the retina did not contain any transplanted cells on the surface or within the retina, whereas large patches of donor cells were present in other areas of the same retina. The close association of transplanted B5 cells with the inner retinal surface suggests the rd1 retina produces factors that attract these cells. These factors do not appear to be present at significant levels in the vitreous of normal retinas, since B5 cells transplanted into the vitreous of mice with normal retinas remain in the vitreous relatively far from the retinal surface. However, since many cells remained attached to the vitreal surface of the retina at 6 weeks post-transplantation (as seen with electron microscopy), even in the rd1 mouse the inner limiting membrane of the retina appears to act as a significant barrier to migration of the transplanted cells. By 6 weeks posttransplantation, a fraction of the injected cells had breached this barrier, incorporated into the neural retina and differentiated into neural-like cells based on cellular morphologies and the expression of neural markers. These results are consistent with the possibility that in some retinal degenerative diseases, the integrity of the inner limiting membrane is compromised. The presence of cells within the rd1 retina at 6 weeks post-transplantation is also consistent with previous studies showing that neural stem cells can incorporate into the developing or damaged mammalian retina [33,36,43, 48,56]. For instance, a greater degree of retinal incorporation by hippocampal-derived adult neural progenitor cells occurs in early postnatal dystrophic rats at a time when photoreceptor cells were actively degenerating [56].

140

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

No retinal integration of brain-derived neural stem cells occurs in intact, normal rat retinas [33], and we found the same to be true after transplantation of neuralized B5 ES cells into normal, C57BL/6J mice at 5 weeks postnatal (data not shown). These results suggest that the undamaged retina may either inhibit migration of transplanted cells from the vitreous into the retina. Alternatively, active neural degeneration may lead to changes in the structural integrity of the retina (in particular the inner limiting membrane) and to cause the release of factors that stimulate migration of transplanted cells from the vitreous into the neural retina. Upon incorporation into the retina, donor B5 cells migrated as far as the inner plexiform layer (IPL), but rarely penetrated deeper into the retina. It is possible that after migration into the retina, donor cells fused with host cells, as suggested by previous studies [44,53,55]. This could impart EGFP expression to host cells. If cellular fusion contributes to EGFP expression by cells in the host IPL, it likely accounts for a very small percentage of cases observed because previous studies that demonstrated cellular fusion involving mouse ES cells state that fusion events are extremely rare [44,49,53,55]. Furthermore, one would expect that fusion events would exhibit morphologies typical of the host cells as demonstrated by previous studies [50]. While donor cells in the current study assumed highly neuronal morphologies, they did not assume retinal morphologies very often. Regardless, the presence of EGFP expression within the host retina indicates that some transplanted cells were able to migrate across the inner limiting membrane and incorporate into the retina. The absence of EGFP expressing cells beyond the IPL after transplantation suggests that the INL of the host rd1 retina may present molecular cues that are repulsive for donor-derived cells or acts as a physical barrier to cell migration. Another possibility is that the IPL provides an especially attractive environment for induced cells. For instance, cells that migrate into the IPL may interact with host retinal cells in such a manner that they develop into specific neuronal cell types and establish long-term associations with host retinal cells within this layer. The presence of numerous varicosities along extensive networks of processes elaborated by incorporated cells and the expression of synaptic vesicle protein SV2 suggests that donor cells differentiate into neuron-like cells and may establish synaptic contacts within the host retina.

141

In the present study, we used rd1 mice in which most photoreceptor degeneration had occurred prior to the time of transplantation, yet we obtained no evidence that transplanted cells migrated to the outermost layer of the retina or differentiated into photoreceptors, the cell type that is primarily lost in rd1 mice. There are few reported cases in which stem cells transplanted into developing, injured or dystrophic eyes achieve a location and morphology consistent with differentiation into photoreceptors, and in most cases the donor cells do not express photoreceptor-specific markers, such as rhodopsin. However, a recent report by Dong et al. [9] demonstrates that donor human neural stem cells pretreated with TGF-h3 can integrate into the outer nuclear layer (ONL) and express rhodopsin. We have preliminary data using a mouse mutant with delayed photoreceptor degeneration (i.e., the mnd mouse) that neuralized B5 ES cells can integrate within the outer regions of the neural retina and can acquire morphologies similar to that of bipolar and horizontal cells; although no photoreceptor-like cells have been observed to date (Meyer et al., unpublished observations). Therefore, neuralized B5 cells and their processes are competent to penetrate the outer retinal layers and appear to differentiate into layer-specific neuron-like cells when presented with the proper microenvironment. Within the normal retina, a vast majority of cells are neuronal, and a small yet significant number of cells express GFAP [23]; the GFAP-positive cells are astrocytes in the ganglion cell layer as well as reactive Mu¨ller cells [20]. It is noteworthy that GFAP-positive donor B5 cells in host rd1 eyes appear to be confined to the vitreous and ganglion cell layer. No O4-positive, oligodendrocyte-like cells are observed after transplantation into the rd1 eye. This is consistent with the lack of oligodendrocytes in the normal retina [10]. We observed host Mu¨ller cells that expressed GFAP (Fig. 7E). These host Mu¨ller cells are likely reactive to the degenerating retina because we also observed GFAPexpressing Mu¨ller cells in untreated and sham-injected rd1 eyes (data not shown). In addition, we found no GFAPimmunoreactive Mu¨ller cells in untreated C57BL/6J (control) mouse eyes. The distribution of Mu¨ller cells that express GFAP was not uniform in host rd1 retinas, in fact GFAPexpressing Mu¨ller cells were found both in regions of the host eye that contained donor B5 cells as well as in regions without donor cells, indicating that GFAP expression by host Mu¨ller cells was not related to the transplantation.

Fig. 6. Electron microscopic images of retinas at 6 weeks post-transplantation revealed the location of transplanted cells on the vitreal side of the inner limiting membrane. (A) Transplanted cells (T) and their processes were found associated with the vitreal side of the inner limiting membrane of the neural retina (arrowhead, inner limiting membrane; R, neural retina). Note the prominent nuclei of cells and numerous processes extending into the vitreous. Inset: Highmagnification view of a transplanted cell (T) and numerous processes viewed in cross-section, adjacent to the inner limiting membrane (arrowhead) and neural retina (R). (B) Transplanted cells closely adhered to the vitreal side of the retina and extended long processes across the inner surface of the retina. This image demonstrates a thin process emerging from a donor cell (T) that remains in close association with the inner limiting membrane (arrowhead). Note the ganglion cell body (GCB) in the neural retina, also adjacent to the inner limiting membrane. (C) In age-matched, sham-injected rd1 eyes, there was no evidence of cells found on the vitreal side of the inner limiting membrane. Arrowheads in (A – C) indicate the location of the inner limiting membrane. Calibration in (C) equals 3 Am and applies to the main figure in (A) and (B); calibration bar in inset of (A) equals 1 Am.

142

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

Fig. 7. Transplanted cells differentiate into neuron-like and glial-like cells based on labeling for neuronal and glial markers at 6 weeks post-transplantation. Note that all panels represent merged images and that positive labeling of donor B5 cells is shown as yellow (a merger of green from EGFP expression and red from the tagged secondary antibody). Commonly, transplanted cells and their processes labeled positively for the early neuronal marker h-III Tubulin (A). In adjacent sections, many donor cells labeled for other neuronal markers such as Map2ab (B), NF-M (C), and NeuN (D). Note that donor cells expressing Map2ab and NF-M were located on the vitreal surface of the inner limiting membrane. Donor cells in adjacent sections were labeled for GFAP (E) and these cells were also located on the vitreal side of the inner limiting membrane. Note the numerous reactive host Mu¨ller cells (projecting vertically in panel E) labeled positively for GFAP. Incorporated cells were also occasionally observed to express the synaptic vesicle marker SV2 (F). Note the location of SV2 expression within the inner plexiform layer, is consistent with the possibility that donor B5 cells exhibit synaptic interactions within the host retina. Calibration bar in (F) equals 25 Am in (A), (C), and (E), and equals 12.5 Am in (B), (D), and (F).

In summary, the current study focused on the differentiation potential of neuralized ES cells under various conditions. Our results indicate that the host rd1 retina promotes further maturation of the donor neural progenitors and suggests that the donor cells achieve a neuronal-like cell fate, including the possible formation of synaptic connections with host retinal neurons based on immunocytochemical labeling. In the future, it will be of interest to identify factors that regulate the degree of migration and integration of transplanted cells into the retina or CNS and test for potential therapeutic benefits due to the presence of the donor stem cells. For some applications, it may be desirable to have donor cells act as a source of secreted therapeutic agents [1,11,47]. However, if replacement of cells lost due to disease or injury is required for functional repair, it will be desirable to maximize incorporation of transplanted cells into the neural retina or brain parenchyma and promote their differentiation into the appropriate cell types with appropriate synaptic connectivity. In either case, it will be important to identify intrinsic retinal factors that regulate the degree of migration of transplanted cells into

the retina or CNS and the state of differentiation of these cells.

Acknowledgements We thank Dr. John McDonald (Washington University, St. Louis) and his entire lab for helping us to master the culture techniques and protocols to neuralize ES cells. Additionally, we thank Dr. Andras Nagy (Mt. Sinai Hospital, Toronto) for kindly providing us with the B5 ES cells. The monoclonal antibody for Nestin generated by Susan Hockfield was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. Grant support was provided by the U.S. National Institutes of Health (NS 38987), the Batten Disease Support and Research Association, Research to Prevent Blindness, Inc., Children’s Brain Diseases Foundation, Rockefeller Brothers Fund, a University of Missouri – Columbia

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

PRIME grant and a University of Missouri Research Board Grant. [23]

References

[24]

[1] K.S. Aboody, A. Brown, N.G. Rainov, K.A. Bower, S. Liu, W. Yang, J.E. Small, U. Herrlinger, V. Ourednik, P.M. Black, X.O. Breakefield, E.Y. Snyder, From the cover: neural stem cells display extensive tropism for pathology in adult brain: evidence from intracranial gliomas, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 12846 – 12851. [2] G. Bain, D. Kitchens, M. Yao, J.E. Huettner, D.I. Gottlieb, Embryonic stem cells express neuronal properties in vitro, Dev. Biol. 168 (1995) 342 – 357. [3] K.R. Boheler, M.Y. Fiszman, Can exogenous stem cells be used in transplantation? Cells Tissues Organs 165 (1999) 237 – 245. [4] O. Brustle, K.N. Jones, R.D. Learish, K. Karram, K. Choudhary, O.D. Wiestler, I.D. Duncan, R.D. McKay, Embryonic stem cell-derived glial precursors: a source of myelinating transplants, Science 285 (1999) 754 – 756. [5] T. Burdon, I. Chambers, C. Stracey, H. Niwa, A. Smith, Signaling mechanisms regulating self-renewal and differentiation of pluripotent embryonic stem cells, Cells Tissues Organs 165 (1999) 131 – 143. [6] L.D. Carter-Dawson, M.M. LaVail, R.L. Sidman, Differential effect of the rd mutation on rods and cones in the mouse retina, Invest. Ophthalmol. Vis. Sci. 17 (1978) 489 – 498. [7] B. Chang, N.L. Hawes, R.E. Hurd, M.T. Davisson, S. Nusinowitz, J.R. Heckenlively, Retinal degeneration mutants in the mouse, Vision Res. 42 (2002) 517 – 525. [8] J. Dinsmore, J. Ratliff, T. Deacon, P. Pakzaban, D. Jacoby, W. Galpern, O. Isacson, Embryonic stem cells differentiated in vitro as a novel source of cells for transplantation, Cell Transplant. 5 (1996) 131 – 143. [9] X. Dong, J.S. Pulido, T. Qu, K. Sugaya, Differentiation of human neural stem cells into retinal cells, NeuroReport 14 (2003) 143 – 146. [10] J.E. Dowling, in: J.E. Dowling (Ed.), The Retina: An Approachable Part of the Brain, Belknap Press, Cambridge, 1987, p. 282. [11] M. Ehtesham, P. Kabos, A. Kabosova, T. Neuman, K.L. Black, J.S. Yu, The use of interleukin 12-secreting neural stem cells for the treatment of intracranial glioma, Cancer Res. 62 (2002) 5657 – 5663. [12] M.F. Finley, N. Kulkarni, J.E. Huettner, Synapse formation and establishment of neuronal polarity by P19 embryonic carcinoma cells and embryonic stem cells, J. Neurosci. 16 (1996) 1056 – 1065. [13] A. Fraichard, O. Chassande, G. Bilbaut, C. Dehay, P. Savatier, J. Samarut, In vitro differentiation of embryonic stem cells into glial cells and functional neurons, J. Cell Sci. 108 (Pt 10) (1995) 3181 – 3188. [14] F.H. Gage, Stem cells of the central nervous system, Curr. Opin. Neurobiol. 8 (1998) 671 – 676. [15] F.H. Gage, Mammalian neural stem cells, Science 287 (2000) 1433 – 1438. [16] D.I. Gottlieb, Large-scale sources of neural stem cells, Annu. Rev. Neurosci. 25 (2002) 381 – 407. [17] P. Gouras, T. Tanabe, Ultrastructure of adult rd mouse retina, Graefes Arch. Clin. Exp. Ophthalmol. 241 (2003) 410 – 417. [18] A. Hadjantonakis, V. Papaioannou, The stem cells of early embryos, Differentiation 68 (2001) 159 – 166. [19] A.K. Hadjantonakis, M. Gertsenstein, M. Ikawa, M. Okabe, A. Nagy, Generating green fluorescent mice by germline transmission of green fluorescent ES cells, Mech. Dev. 76 (1998) 79 – 90. [20] K.R. Huxlin, A.J. Sefton, J.H. Furby, The origin and development of retinal astrocytes in the mouse, J. Neurocytol. 21 (1992) 530 – 544. [21] S.H. Lee, N. Lumelsky, L. Studer, J.M. Auerbach, R.D. McKay, Efficient generation of midbrain and hindbrain neurons from mouse embryonic stem cells, Nat. Biotechnol. 18 (2000) 675 – 679. [22] S. Liu, Y. Qu, T.J. Stewart, M.J. Howard, S. Chakrabortty, T.F. Hole-

[25]

[26]

[27]

[28] [29] [30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38] [39] [40]

[41] [42]

[43]

143

kamp, J.W. McDonald, Embryonic stem cells differentiate into oligodendrocytes and myelinate in culture and after spinal cord transplantation, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 6126 – 6131. F.J. Livesey, C.L. Cepko, Vertebrate neural cell-fate determination: lessons from the retina, Nat. Rev. Neurosci. 2 (2001) 109 – 118. S. Lowell, Stem cells show their potential, Trends Cell Biol. 10 (2000) 210 – 211. B. Lu, T. Kwan, Y. Kurimoto, M. Shatos, R.D. Lund, M.J. Young, Transplantation of EGF-responsive neurospheres from GFP transgenic mice into the eyes of rd mice, Brain Res. 943 (2002) 292 – 300. N. Lumelsky, O. Blondel, P. Laeng, I. Velasco, R. Ravin, R. McKay, Differentiation of embryonic stem cells to insulin-secreting structures similar to pancreatic islets, Science 292 (2001) 1389 – 1394. J.W. McDonald, X.Z. Liu, Y. Qu, S. Liu, S.K. Mickey, D. Turetsky, D.I. Gottlieb, D.W. Choi, Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord, Nat. Med. 5 (1999) 1410 – 1412. R. McKay, Stem cells in the central nervous system, Science 276 (1997) 66 – 71. R. McKay, Stem cells—hype and hope, Nature 406 (2000) 361 – 364. M.E. McLaughlin, T.L. Ehrhart, E.L. Berson, T.P. Dryja, Mutation spectrum of the gene encoding the beta subunit of rod phosphodiesterase among patients with autosomal recessive retinitis pigmentosa, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 3249 – 3253. J.S. Meyer, M.L. Katz, J.A. Maruniak, M.D. Kirk, Transplantation of mouse embryonic stem cells into the degenerating murine retina, Soc. Neurosci. 27 (San Diego, CA). A. Morizane, J. Takahashi, Y. Takagi, Y. Sasai, N. Hashimoto, Optimal conditions for in vivo induction of dopaminergic neurons from embryonic stem cells through stromal cell-derived inducing activity, J. Neurosci. Res. 69 (2002) 934 – 939. A. Nishida, M. Takahashi, H. Tanihara, I. Nakano, J.B. Takahashi, A. Mizoguchi, C. Ide, Y. Honda, Incorporation and differentiation of hippocampus-derived neural stem cells transplanted in injured adult rat retina, Invest. Ophthalmol. Vis. Sci. 41 (2000) 4268 – 4274. W.K. Noell, Aspects of experimental and hereditary retinal degeneration, in: C.N. Graymore (Ed.), Biochemistry of the Retina, Academic Press, New York, 1965, pp. 51 – 72. S.J. Pittler, W. Baehr, Identification of a nonsense mutation in the rod photoreceptor cGMP phosphodiesterase beta-subunit gene of the rd mouse, Proc. Natl. Acad. Sci. U. S. A. 88 (1991) 8322 – 8326. S. Pressmar, M. Ader, G. Richard, M. Schachner, U. Bartsch, The fate of heterotopically grafted neural precursor cells in the normal and dystrophic adult mouse retina, Invest. Ophthalmol. Vis. Sci. 42 (2001) 3311 – 3319. Y. Qu, F. Schottler, S. Liu, J. Lu, M. Platik, A. Lu, A. Clair, D.S. Adams, D.I. Gottlieb, J.W. McDonald, Line specific differences in parameters of neural stem cells derived from embryonic stem (ES) cells, in: Society for Neuroscience vol. 27,Society for Neuroscience, San Diego, CA, 2001. E.J. Robertson, Derivation and maintenance of embryonic stem cell cultures, Methods Mol. Biol. 75 (1997) 173 – 184. Y. Sasai, Generation of dopaminergic neurons from embryonic stem cells, J. Neurol. 249 (Suppl 2) (2002) II41 – II44. R.M. Seaberg, D. van der Kooy, Stem and progenitor cells: the premature desertion of rigorous definitions, Trends Neurosci. 26 (2003) 125 – 131. A.G. Smith, Embryo-derived stem cells: of mice and men, Annu. Rev. Cell Dev. Biol. 17 (2001) 435 – 462. A.G. Smith, J.K. Heath, D.D. Donaldson, G.G. Wong, J. Moreau, M. Stahl, D. Rogers, Inhibition of pluripotential embryonic stem cell differentiation by purified polypeptides, Nature 336 (1988) 688 – 690. M. Takahashi, T.D. Palmer, J. Takahashi, F.H. Gage, Widespread integration and survival of adult-derived neural progenitor cells in the developing optic retina, Mol. Cell Neurosci. 12 (1998) 340 – 348.

144

J.S. Meyer et al. / Brain Research 1014 (2004) 131–144

[44] N. Terada, T. Hamazaki, M. Oka, M. Hoki, D.M. Mastalerz, Y. Nakano, E.M. Meyer, L. Morel, B.E. Petersen, E.W. Scott, Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion, Nature 416 (2002) 542 – 545. [45] J.A. Thomson, J. Itskovitz-Eldor, S.S. Shapiro, M.A. Waknitz, J.J. Swiergiel, V.S. Marshall, J.M. Jones, Embryonic stem cell lines derived from human blastocysts, Science 282 (1998) 1145 – 1147. [46] V. Tropepe, S. Hitoshi, C. Sirard, T.W. Mak, J. Rossant, D. van der Kooy, Direct neural fate specification from embryonic stem cells: a primitive mammalian neural stem cell stage acquired through a default mechanism, Neuron 30 (2001) 65 – 78. [47] M.H. Tuszynski, J. Roberts, M.C. Senut, H.S. U, F.H. Gage, Gene therapy in the adult primate brain: intraparenchymal grafts of cells genetically modified to produce nerve growth factor prevent cholinergic neuronal degeneration, Gene Ther. 3 (1996) 305 – 314. [48] S.J. Van Hoffelen, M.J. Young, M.A. Shatos, D.S. Sakaguchi, Incorporation of murine brain progenitor cells into the developing mammalian retina, Invest. Ophthalmol. Vis. Sci. 44 (2003) 426 – 434. [49] G. Vassilopoulos, D.W. Russell, Cell fusion: an alternative to stem cell plasticity and its therapeutic implications, Curr. Opin. Genet. Dev. 13 (2003) 480 – 485.

[50] J.M. Weimann, C.B. Johansson, A. Trejo, H.M. Blau, Stable reprogrammed heterokaryons form spontaneously in Purkinje neurons after bone marrow transplant, Nat. Cell Biol. 5 (2003) 959 – 966. [51] M. Wernig, O. Brustle, Fifty ways to make a neuron: shifts in stem cell hierarchy and their implications for neuropathology and CNS repair, J. Neuropathol. Exp. Neurol. 61 (2002) 101 – 110. [52] A.M. Wobus, R. Grosse, J. Schoneich, Specific effects of nerve growth factor on the differentiation pattern of mouse embryonic stem cells in vitro, Biomed. Biochim. Acta 47 (1988) 965 – 973. [53] A.E. Wurmser, F.H. Gage, Stem cells: cell fusion causes confusion, Nature 416 (2002) 485 – 487. [54] H.Q. Xian, D.I. Gottlieb, Peering into early neurogenesis with embryonic stem cells, Trends Neurosci. 24 (2001) 685 – 686. [55] Q.L. Ying, J. Nichols, E.P. Evans, A.G. Smith, Changing potency by spontaneous fusion, Nature 416 (2002) 545 – 548. [56] M.J. Young, J. Ray, S.J. Whiteley, H. Klassen, F.H. Gage, Neuronal differentiation and morphological integration of hippocampal progenitor cells transplanted to the retina of immature and mature dystrophic rats, Mol. Cell Neurosci. 16 (2000) 197 – 205. [57] X. Zhao, J. Liu, I. Ahmad, Differentiation of embryonic stem cells into retinal neurons, Biochem. Biophys. Res. Commun. 297 (2002) 177 – 184.