LABORATORY SCIENCES
Retinal Progenitor Cell Xenografts to the Pig Retina
Morphologic Integration and Cytochemical Differentiation
Karin Warfvinge, PhD; Jens F. Kiilgaard, MD, PhD; Erin B. Lavik, DSc; Erik Scherfig, MD; Robert Langer, PhD;
Henry J. Klassen, MD, PhD; Michael J. Young, PhD
Objective:: To investigate the survival, integration, and
differentiation of mouse retinal progenitor cells after transplantation to the subretinal space of adult pigs.
Methods:: Green fluorescent protein–positive (GFP⫹)
murine retinal progenitor cells were transplanted subretinally as single cells, spheres, or biodegradable polymerprogenitor composites into 24 nonimmunosuppressed
adult pigs. Of these, 14 pigs received laser lesions (n=11)
or outer retinal scraping injury (n = 3). Recipients were
killed at 30 minutes to 5 weeks after grafting.
Results: The GFP⫹ murine retinal progenitor cells survived well for up to 14 days after transplantation to the
pig retina. After 5 weeks, fewer GFP⫹ cells were found.
In the pigs that received laser treatment before grafting
of cell suspension, GFP⫹ cells integrated into the retinal pigment epithelium and all layers of the retina. The
GFP⫹ cells exhibited morphologic evidence of differentiation into mature retinal neurons, although evaluation of marker expression found only nestin and glial fibrillary acidic protein colocalization. In noninjured pigs,
cells mainly integrated into the retinal pigment epithelium. In pigs that received composites, cells appeared to
Author Affiliations: Department
of Ophthalmology, Lund
University Hospital, Lund,
Sweden (Dr Warfvinge); Eye
Department, Rigshospitalet and
Eye Pathology Institute,
Copenhagen University,
Copenhagen, Denmark
(Drs Kiilgaard and Scherfig);
Department of Electrical
Engineering and Computer
Science, Massachusetts Institute
of Technology, Cambridge, Mass
(Drs Lavik and Langer); Stem
Cell Research, Children’s
Hospital of Orange County,
Orange, Calif (Dr Klassen); and
Schepens Eye Research Institute,
Department of Ophthalmology,
Harvard Medical School, Boston,
Mass (Dr Young).
T
mature and extended processes through pores in the polymer matrix.
Conclusions: Retinal progenitor cell xenografts survive for a sufficiently long period to integrate into areas
of injury and exhibit morphologic differentiation. By 5
weeks, survival diminishes. Biodegradable polymers may
be useful for transplanting retinal progenitor cells in a
structurally organized manner.
Clinical Relevance: Central nervous system (CNS) diseases may cause long-term disabilities. Substantial tissue
destruction can be sustained by the complex structures of
the brain, spinal cord, or retina without loss of life, yet the
lack of effective CNS regeneration frequently results in disruption of activities of daily living and marked degradation in quality of life. It has become clear that an enormous potential for repair is present within the mammalian
CNS. The challenge is to harness this potential to treat disease. Transplantation of neuronal tissue to the CNS represents a promising, albeit challenging, approach to the
replacement of neurons lost owing to injury or disease.
Arch Ophthalmol. 2005;123:1385-1393
HE MAJOR DISEASES OF THE
retina, including retinitis
pigmentosa, age-related
macular degeneration, and
diabetic retinopathy, are
quite heterogeneous with respect to etiology, pathology, and demographics, yet have
in common the unifying feature of neuronal loss. As in the brain and spinal cord, loss
of neurons in the retina is permanent, and
there are at present no restorative treatments available for these conditions. One
approach to neuronal repopulation that has
been actively investigated is the transplantation of embryonic or fetal retinal tissue.
Achieving widespread functional integration of transplanted retinal tissue in the diseased retina has proved to be elusive1; however, neural progenitor cells have recently
demonstrated considerable potential as an
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1385
alternative to fetal tissue in the setting of
transplantation.
In rodents, a variety of neural precursor
cells have been transplanted to the intact
retina of normal neonatal and adult animals, to the degenerating retina, and to the
injured retina, with markedly varying results in terms of integration, migration, and
differentiation.2,3 In all studies of this type,
an important consideration that has emerged
is the need to distinguish donor cells from
host tissue after transplantation. The use of
soluble proteins such as green fluorescent
protein (GFP) as a label confers the advantage of demonstrating the detailed configuration of the grafted cells.
Recently, isolation and amplification of
stem cells from the retina of neonatal mice
expressing the enhanced version of the
GFP transgene were described.4 After ex-
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
pansion in vitro, these cells were transplanted to the retina
of mice with retinal degeneration. The grafted cells survived, maintained GFP expression, migrated, and showed
evidence of integration within the host retina. Herein we
describe the transplantation of these GFP mouse progenitor cells to the subretinal space of adult pigs, as cell
suspension, spheres, or a composite graft incorporating
a biodegradable polymer. The pig is of interest as a recipient because the anatomy, size, and vasculature of the
porcine eye more closely approximate the human eye than
does that of the rodent, and the techniques and instrumentation routinely used in the clinic are directly applicable to the pig model. Moreover, the porcine retina has
an area centralis that contains a large number of cones.
METHODS
PROGENITOR CELLS
Retinal progenitor cells (RPCs) were isolated from pooled retinas of postnatal day 1 GFP transgenic C57Bl/6 mice (a kind
gift of M. Okabe, PhD, University of Osaka, Osaka, Japan). Retinas were dissected free from the posterior eyecup, and the ciliary marginal zone and optic nerve head were removed under
microscopic visualization. Pooled retinal tissue was finely minced
and digested with 0.1% type 1 collagenase (Sigma-Aldrich Corp,
St Louis, Mo) for 20 minutes. The supernatant containing liberated cells was forced through a 100-µm mesh strainer, centrifuged, and seeded into culture vessels in basal medium (Neurobasal; GIBCO BRL Life Technologies, Rockville, Md)
supplemented with 2-mmol/L L-glutamine, 100-µg/mL penicillin-streptomycin, 20-ng/mL epidermal growth factor (Promega, Madison, Wis), and B-27 neural supplement (GIBCO BRL
Life Technologies). This cycle was repeated until all retinal tissue was digested. Cells were re-fed on alternating days. Within
2 to 3 weeks, RPCs were visible as nonadherent spheres and
continued to expand in the presence of epidermal growth factor. Cultures were split 1:5 every 7 to 10 days. Although capable of expansion through more than 60 passages, cells of less
than 20 passages were used in this study.
POLYMER SCAFFOLDS
The biodegradable polymer was formed into highly porous scaffolds. These were fabricated from blends of poly(lactic-coglycolic acid) and poly(L-lactic acid) by means of a freezedrying technique that led to pores being oriented normal to the
plane of the scaffold. The RPCs were seeded onto the polymers and cultured under expansion conditions (20-ng/mL epidermal growth factor) for 14 days before transplantation. The
fabrication of the biodegradable polymer scaffolds has recently been described by Lavik et al.5
ANIMALS AND SURGERY
Twenty-four female domestic pigs of the Danish Landrace breed
(age, 4 months; weight, approximately 30 kg) were used in the
experiments. All animals were preanesthetized with intramuscular injections of 15 mg of midazolam (DormicumA; Roche,
Hvidovre, Denmark) and a combination of zolazepam hydrochloride, 11.9 mg/mL, and tiletamine hydrochloride, 11.9 mg/mL
(Zoletil 50 Vet; Virbac SA, Carros, France) mixed with xylazine hydrochloride (12.38 mg/mL; Intervet, Skovlunde, Denmark), ketamine hydrochloride (14.29 mg/mL; Intervet), and
methadone (2.38 mg/mL; Nycomed, Roskilde, Denmark). The
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1386
pigs underwent endotracheal intubation and were artificially
ventilated and anesthetized with 2% to 3% isoflurane (Abbott,
Solna, Sweden) in combination with oxygen. The stroke volume (300 mL/stroke) and respiratory frequency (12/min) were
kept constant throughout the surgery. The left pupil was dilated with an eyedrop combination of 0.4% benoxinate hydrochloride (oxybuprocaine; SAD, Copenhagen, Denmark), 10%
phenylephrine hydrochloride (metaoxedrine chloride; SAD),
0.5% tropicamide (Mydriacyl; Alcon, Heinaut, Belgium), 1%
atropine sulfate (SAD), and 5% povidone-iodine (SAD).
At surgery, the central and posterior vitreous was removed
together with the posterior hyaloid membrane by means of a
localized 3-port pars plana vitrectomy. With the goal of promoting integration of grafted cells into the host retina, focal
damage was induced in 11 animals via the application of green
argon laser burns in a grid pattern to the area centralis. The
laser lesions were placed as a triangle with the apex close to
the optic disc. The energy of the lesions was chosen from peripheral test lesions to give a uniform, white color of the spot
without signs of perforation or necrosis. Thereafter, a retinal
bleb was elevated in the same area by the injection of 0.25 to
0.5 mL of 0.9% sodium chloride through a 41-gauge needle.
The use of endodiathermy on the detached retina preceded further enlargement of the retinotomy before transplantation. In
3 animals, mechanical scraping of the outer retina was performed before grafting to the area centralis of the polymerprogenitor composite (no laser treatment was given).
The murine GFP-positive (GFP⫹) RPCs were injected into
the retinal bleb as a single-cell suspension by using a 27-gauge
needle (n=12), as spheres of cells by using a 20-gauge needle
(n=6), or as a biodegradable polymer-progenitor composite by
using fine forceps (n=6). The single-cell suspension and sphere
injections contained approximately 2⫻107 cells, while the biodegradable polymer composite graft contained approximately
2⫻106 cells. Immediate reflux of some cells into the vitreous
cavity was observed in most animals injected with cell suspension or spheres. A small air bubble was placed in the subretinal bleb under the retinotomy to prevent further reflux after
withdrawal of the needle. Chloramphenicol (SAD) was given
locally at the end of surgery to prevent infection. The pigs were
examined by ophthalmoscopy on a weekly basis.
The research protocol was reviewed and approved by the Danish Animal Experiment Inspectorate, and was in accordance with
the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research.
TISSUE PROCESSING
Eyes were enucleated with the animal under anesthesia at 30
minutes (suspension, n=1), 1 week (suspension, n=3; spheres,
n=2; polymer, n=2), 2 weeks (suspension, n=2; spheres, n=2;
polymer, n=2), and 5 weeks (suspension, n=6; spheres, n=2;
polymer, n=2) after transplantation. After enucleation, pigs were
killed by intravenous injection of 2 to 4 g of pentobarbital sodium (200 mg/mL; KVL, Copenhagen). Globes were placed in
4% paraformaldehyde for 10 to 20 minutes. The anterior segment and the lens were then removed and the posterior segment was postfixed for 2 hours in 4% paraformaldehyde, with
subsequent rinsing in increasing concentrations of sucrose containing Sörensen phosphate buffer. A horizontal cut was made
that extended from the temporal retinal margin to 2 to 3 mm
nasal to the optic disc, thus comprising the temporal ciliary margin, the area centralis, and the optic disc. The tissues were embedded in a gelatin medium and serially sectioned at 12 µm on
a cryostat. During the sectioning process, every 15th section
was examined by epifluorescence microscopy for GFP⫹ cells
and every 10th slide was stained with hematoxylin-eosin.
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
Table 1. Primary Antibodies
Antigen
Species
Description
GFP
Chicken polyclonal
PCNA
Mouse monoclonal
Vimentin
Mouse monoclonal
Green fluorescent
protein
Proliferating cell nuclear
antigen
Vimentin
Used for Detection of
Dilution
Grafted cells
1:5000
Chemicon, Temecula, Calif
GFAP
Rabbit polyclonal
Proliferative cells
1:500
1:10
1:100
DAKO
1:100
1:200
Mouse monoclonal
Microtubule-associated
protein 2
RPE-specific protein 65
Müller cells and
astrocytes
Müller cells and
astrocytes
Immature cells and
activated glial cells
Neural precursor cells
Santa Cruz Biotechnology,
Santa Cruz, Calif
DAKO, Copenhagen, Denmark
Nestin
Mouse monoclonal
RPE
1:200
Cytokeratin
Recoverin
Mouse monoclonal
Rabbit polyclonal
Cytokeratin
Recoverin
RPE
Cones
1:100
1:10 000
Transducin
Rho 4D2
Rabbit polyclonal
Mouse monoclonal
G protein G-␥-c subunit
N-terminal of rhodopsin
Cones
Rods
1:1000
1:100
Parvalbumin
Mouse monoclonal
Parvalbumin
1:1000
Calbindin
PKC
Mouse monoclonal
Mouse monoclonal
Calbindin
Protein kinase C
Bipolar and
amacrine cells
Horizontal cells
Bipolar cells
BD Biosciences Pharmingen,
Heidelberg, Germany
Sigma-Aldrich Corp, St Louis,
Mo
Gift of D. Thompson, PhD,
Ann Arbor, Mich
DAKO
Gift of A. Dizhoor, PhD,
Detroit, Mich
CytoSignal, San Diego, Calif
Gift of R. S. Molday, PhD,
Vancouver, British Columbia
Sigma-Aldrich Corp
MAP-2
Mouse monoclonal
RPE65
Glial fibrillary acidic
protein
Intermediate filaments
1:200
1:1000
Source
Sigma-Aldrich Corp
Nordic Biosite AB, Täby,
Sweden
Abbreviation: RPE, retinal pigment epithelium.
IMMUNOHISTOCHEMISTRY
The retinal sections were exposed to primary antisera (Table 1)
in a moist chamber for 16 to 18 hours, at 4°C, followed by rinsing in 0.1M phosphate-buffered saline with 0.25% Triton X-100.
Sections were then incubated with secondary Texas red–
conjugated antibodies (1:200; Jackson Immunoresearch, West
Grove, Pa) for 1 to 2 hours at room temperature in the dark. Normal eyes, processed in parallel, were used as controls. In addition, negative controls with omission of the primary antisera were
performed. The specimens were examined with an epifluorescence microscope. Colocalization of Texas red–labeled primary
antibodies and GFP⫹ cells was assessed by superimposition of
separate digital images of each fluorochrome.
Sections were frequently stained with chicken anti-GFP, either
to examine the quality of the GFP expression or to disclose possible down-regulated grafted cells. The endogenous GFP expression was always high, and enhancing with anti-GFP resulted in
blurring of the cell boundaries. In the pigs with few or no surviving cells, the anti-GFP did not disclose any additional cells;
therefore, these results will not be further discussed.
associated protein 2 (MAP-2) were expressed at some
point, but not vimentin, retinal pigment epithelium
(RPE)–specific protein 65, cytokeratin, recoverin, transducin, Rho 4D, parvalbumin, calbindin, or protein kinase C. Specific data concerning pretreatment conditions and GFP⫹ cell integration and migration are
summarized in Table 2 and Table 3.
POSTTREATMENT REACTIONS
IN THE HOST RETINA
In all pigs killed 1 to 2 weeks after grafting, the host retina
showed up-regulation of both GFAP and vimentin. This
up-regulation corresponded to the location of the subretinal bleb induced at the time of transplantation and thus
the area of transient focal detachment. Eleven of the 24
animals were pretreated with laser burns in an effort to
enhance donor cell integration. Histologically, areas of the
retina with laser burns exhibited decreased thickness, laminar disorganization, and occasional intraretinal RPE cells.
RESULTS
Retinal progenitor cells derived from postnatal day 1 GFP
transgenic mice were transplanted to the subretinal space
of 24 mature pigs, as single-cell suspension, as spheres,
or as a biodegradable polymer-progenitor composite. After death of the animal at various time points ranging from
30 minutes to 5 weeks after transplantation, surviving
GFP⫹ cells were found at or near the transplantation site
in 16 (67%) of the 24 animals. A vast number of markers (Table 1) were examined before and after grafting.
Only proliferating cell nuclear antigen (PCNA), nestin,
glial fibrillary acidic protein (GFAP), and microtubule(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1387
GFP TRANSGENIC MURINE RPCS:
BEFORE GRAFTING
Cells prepared as single-cell dissociates, as well as cells
grown as spheres, displayed immunoreactivity to the
markers PCNA, nestin, GFAP, and MAP-2. No baseline
reactivity was detected with regard to the other antibodies tested, as listed in Table 1. Mouse RPCs seeded onto
a scaffold were found to be situated largely at the surface of the polymer, were rounded, and maintained immunoreactivity to PCNA, nestin, GFAP, and MAP-2
(Figure 1).
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
Table 2. Method of GFPⴙ Cell Delivery, Survival Times, Pigs Grafted, Surviving Cells, and Retinal Damage Before Transplantation
Cell Deliverables, Pig No./No. of Surviving Cells (Pretreatment)
Survival Time
Suspension
30 min
1 wk
64/⫹⫹⫹ (None)
55/⫹⫹⫹ (None)
73/⫹⫹⫹ (Laser)
74/⫹⫹⫹ (Laser)
56/⫹⫹⫹ (None)
75/⫹⫹⫹ (Laser)
57/⫹ (None)
76/− (Laser)
107/− (Laser)
108/− (Laser)
109/⫹ (Laser)
110/− (Laser)
2 wk
5 wk
Spheres
Polymer-Progenitor Composite
54/− (None)
77/⫹⫹ (Laser)
58/⫹⫹⫹ (None)
80/⫹⫹⫹ (Scraped)
61/⫹⫹ (None)
81/⫹⫹ (Laser)
62/⫹ (None)
79/⫹ (Laser)
63/− (None)
78/⫹⫹⫹ (Scraped)
59/− (None)
82/− (Scraped)
Abbreviations: GFP⫹, green fluorescent protein–positive; −, no cells; ⫹, few cells; ⫹⫹, moderate number of cells; ⫹⫹⫹, many cells.
Table 3. Pigs With Surviving Cells and Integration and Migration Into the Retina of GFPⴙ Cells After Grafting
Integration
Suspension
Spheres
Polymer-progenitor
composite
Pig No.
Migration
Into the Retina
RPE or Retina
Subretinal Space or Vitreous
64
55
73
74
56
75
57
109
77
61
81
62
79
58
80
78
−
−
⫹⫹⫹
⫹⫹⫹
⫹
⫹⫹⫹
−
−
−
−
−
−
−
−
−
−
None
RPE
RPE, retina
RPE, retina
RPE, retina
RPE, retina
None
RPE
RPE
None
None
None
None
None
None
None
Subretinal space
None
Vitreous
Subretinal space
None
Vitreous
Subretinal space
None
Subretinal space, vitreous
Subretinal space, vitreous
Subretinal space, vitreous
Subretinal space
Subretinal space
Cells only in composite
Cells only in composite
Cells only in composite
Abbreviations: GFP⫹, green fluorescent protein–positive; RPE, retinal pigment epithelium; −, none; ⫹, few cells found in a restricted area of the retina; ⫹⫹⫹,
many cells in all layers of the retina.
CELL SUSPENSION: AFTER GRAFTING
To validate the surgical method used for subretinal transplantation, 1 pig that had received RPCs as a single-cell
suspension was killed 30 minutes after grafting. At this
time, a large retinal detachment was evident (Figure 2A).
Rounded GFP⫹ cells were found floating freely in the
subretinal space or attached to the RPE or photoreceptor outer segments (Figure 2B and C). The GFP⫹ cells
coexpressed nestin, and some were in a proliferative stage,
as shown by PCNA immunostaining. Some GFP⫹ cells
coexpressed GFAP or MAP-2.
Three pigs were killed 1 week after transplantation,
2 of which had been treated with retinal burns at the
time of surgery, and surviving RPCs were found in all
3 animals. In the laser-treated pigs, GFP⫹ cells integrated into the RPE (Figure 3A and D) and all layers
of the retina (Figure 3B-E) and were also found in or
close to the nerve fiber layer (Figure 3F). The cells
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1388
had matured and exhibited a range of phenotypes, the
majority appearing morphologically neuronlike. In the
untreated pig, GFP⫹ cells were mainly found in the
RPE layer, although occasional GFP⫹ fibers were seen
in the retina as well (data not shown). Immunohistochemistry showed expression of nestin (Figure 3F) and
GFAP (Figure 3G) by donor cells, but not vimentin
(Figure 3H). Autofluorescence throughout the entire
retina was a common finding in all the laser-treated animals (Figure 3A-C).
Two pigs were killed 2 weeks after transplantation,
1 of which had been treated with laser burns, and surviving donor cells were found in both animals. In the
untreated pig, GFP⫹ cells had integrated into the RPE
layer, although, in contrast to host RPE cells, the
donor cells did not express the RPE-specific markers
RPE65 or cytokeratin ( Figure 4 A and B). In the
laser-treated animal, GFP⫹ cells were found in the
retina at or near the laser wounds. Despite the morWWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
HTX-Eosin
B
GFP
C
PCNA
E
GFAP
F
MAP-2
Polymer
A
D
Nestin
Figure 1. Polymer-progenitor composite cultured for 14 days (scale bars, 90 µm). A, Hematoxylin (HTX)-eosin staining of the composite shows blue nuclei of the
mouse cells covering the surface of the polymer. The vertical line on the left side indicates the thickness of the polymer. B, Visualization of the composite by
fluorescence microscopy shows green fluorescent protein (GFP)–positive (GFP⫹) profiles on the polymer surface. C, Immunohistochemistry of the composite
using antibodies against the proliferating cell nuclear antigen (PCNA). Double-labeling of PCNA (red) and GFP shows that most cells are in a proliferative stage
(yellow-orange). D-F, Subpopulations of GFP⫹ cells also express nestin (D), as well as glial fibrillary acidic protein (GFAP) (E) and microtubule-associated protein
2 (MAP-2) (F).
GFP
A
B
C
RPE
ONL
ONL
Figure 2. Retinal progenitor cell suspension 30 minutes after grafting to the pig eye. A, A large area of retinal detachment is evident (hematoxylin-eosin, original
magnification ⫻8). B, Under the detachment, cells are found floating within the subretinal space or adhering to the retinal photoreceptors (ONL indicates outer
nuclear layer) (hematoxylin-eosin; scale bar, 45 µm). C, Fluorescent microscopy shows that the grafted cells express green fluorescent protein and in some cases
adhere to the photoreceptors (arrowheads) or retinal pigment epithelium (RPE) (arrows) (scale bar, 45 µm).
phologic evidence of donor cell differentiation, immunohistochemical studies were negative for markers of
neuronal phenotype.
Six animals were killed 5 weeks after transplantation, 5 with laser treatment and 1 without. Survival was
markedly decreased at this point, with surviving cells being
found in only 2 pigs (1 laser treated, 1 untreated). The
surviving cells were very few and immunohistochemical analysis was not undertaken.
SPHERES: AFTER GRAFTING
Two animals injected with RPCs as spherical aggregates
were killed 1 week after grafting, 1 of which had been
treated with laser and 1 of which was untreated. The
treated pig displayed surviving GFP⫹ cells, and these were
located in the subretinal space and the vitreous
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1389
(Figure 5A and B). These cells were found either as
spheres or as single cells, a few of which had extended
processes at this point. A small number of cells had also
integrated into the retina. As seen in the pigs grafted with
cell suspension, only nestin and GFAP colocalized with
the GFP⫹ profiles (Figure 5C).
Two animals were killed 2 weeks after grafting, 1 laser treated and 1 untreated. In the laser-treated pig, single
cells or layers of cells were seen covering the inner (vitreal) surface of the retina. These cells were immunoreactive for nestin and GFAP. In the untreated pig, spheres
of GFP⫹ cells were found in the vitreous. These cells were
also nestin and GFAP positive.
Two pigs were killed 5 weeks after grafting, 1 laser
treated and 1 untreated. Both animals displayed foci where
a few GFP⫹ cells had been incorporated into the RPE.
Immunohistochemical analysis was not carried out.
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
A
RPE
B
RPE
Subretinal Space
C
ONL
INL
IPL
D
RPE
E
INL
IPL
NF
F
RPE
Nestin
G
RPE
GFAP
H
RPE
Vimentin
Figure 3. Integration and differentiation patterns 1 week after grafting as cell suspension (scale bars, 60 µm). Panels F through H show immunolabeling with
different antibodies (scale bars, 45 µm). A, Integration of green fluorescent protein (GFP)–positive (GFP⫹) donor cells into the retinal pigment epithelium (RPE).
B, In this image retinal progenitor cells are not integrated into the RPE, but form a layer covering the photoreceptors. Green fluorescent protein–positive donor
cells are also found in the outer retinal layers, sending radially oriented processes into the inner retina (arrows). C, All retinal layers harbor GFP⫹ cells in this
section. Processes are found reaching both radially and horizontally. Autofluorescent bodies (yellow) are also seen within the retina. ONL indicates outer nuclear
layer; INL, inner nuclear layer; and IPL, inner plexiform layer. D, In this section, murine retinal progenitor cells have integrated into the RPE and retina. Radially
oriented GFP⫹ fibers are present (arrows). In the middle of the image a donor cell, probably situated in the INL, sends processes into the ONL. E, This image
shows GFP⫹ cells in the nerve fiber layer (NF), but also a well-developed cell (arrow) in the INL. F, A layer of GFP⫹ donor cells that coexpress nestin is found
covering the photoreceptors (arrowheads). Some cells send processes into the retina (arrow). G, Up-regulation of glial fibrillary acidic protein (GFAP) is found in
the host retina. This up-regulation corresponded to the location of the subretinal bleb induced at the time of transplantation and thus the area of transient focal
detachment. Some of the GFAP-positive cells coexpress GFP, consistent with donor cells (arrowheads). H, Vimentin is also up-regulated in the host corresponding
to the area of local detachment, but no coexpression of vimentin and GFP was found.
BIODEGRADABLE
POLYMER-PROGENITOR COMPOSITE
Two animals transplanted with biodegradable polymerprogenitor composite were killed 1 week after grafting,
1 that had been treated with subretinal scraping and 1
untreated. Surviving GFP⫹ cells were found in both pigs.
Grafted cells were generally confined to the polymer structure and exhibited neuronallike and gliallike phenotypes, often with long slender processes extending across
the width of the scaffold (Figure 6). Again, some of the
GFP⫹ cells colabeled with nestin or GFAP.
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1390
Two weeks after transplantation, 2 pigs grafted with the
polymer-progenitor composite were killed; 1 had received subretinal scraping before transplantation and 1 was
untreated. The treated animal contained GFP⫹ cells that
exhibited both gliallike and neuronallike morphologic features, as seen in the 1-week pigs. In this case the retina was
severely disrupted, likely related to complications at the time
of surgery, and further analysis was not performed.
Two polymer-grafted animals were killed 5 weeks after grafting. In both cases the scaffolds showed clear signs
of biodegradation; however, no surviving cells were found
in these specimens.
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
COMMENT
This study is the first, to our knowledge, to demonstrate
morphologic integration of transplanted progenitor cells
into the neural retina and RPE of a large mammal. This
integration occurred in developmentally mature hosts (3
months) during a relatively short period (within 1 week)
and despite substantial genetic disparity between donor
(mouse) and recipient (pig). This study also demonstrates the feasibility of using biodegradable polymerprogenitor composite for cell delivery in a large, humanlike eye, although no cells escaped the biopolymer scaffold
after transplantation into the pig eye.
All findings were obtained in the absence of active immunosuppression, a fact that has direct bearing on the
decrease in donor cell survival seen at 5 weeks. Preliminary data indicate that a substantial host immune response, initiated by preformed antibodies and directed
against the foreign donor cells, is responsible for this decreased survival.6 Previous reports on the transplantation of central nervous system progenitor cells to the retina
include studies in the rat,7-10 mouse,4 and Brazilian opossum.11 The work in the pig presented herein extends the
existing literature to a large animal model where the biological and surgical challenges more closely approximate those faced in the human retina.
The surgical approach used in this study, namely, pars
plana vitrectomy followed by creation of a retinotomy
and the raising of a focal retinal bleb, proved to be a reliable method for accessing the subretinal space. This was
demonstrated by results from the pig killed 30 minutes
after grafting, as well as the results from later times. All
retinal blebs resolved and the retina was reattached by
the end of the first week. In response to the transplant
procedure, the host retina up-regulated GFAP and vimentin, consistent with an activation of Müller cells and
astrocytes; however, there was no further evidence of the
remodeling events that can occur in the retina as a complication of retinal detachment.2,12-14 Specifically, gliosis
and subretinal fibrosis were not observed at later points.
Of the 3 methods of delivery used in the present study,
dissociated cell suspension was the simplest way to deliver these cells and resulted in maximal intraretinal migration. In addition, a drawback to the sphere delivery
method was the necessity to extend the retinotomy to accommodate the larger-gauge cannula needed for delivery. This resulted in an increased tendency toward reflux of GFP⫹ spheres into the vitreous cavity. Consistent
with this observation, the histologic data from animals
with surviving grafts showed GFP⫹ cells present in the
vitreous of 3 (60%) of 5 animals when grafted as spheres,
compared with 2 (25%) of 8 for suspension and 0 (0%)
of 3 for polymer-progenitor composite. Cells grafted on
the biodegradable polymer were invariably found in the
subretinal space (100% [3/3]) and remained within the
polymer scaffold for the duration of the study without
migration into the host retina or RPE.
The mouse and the pig are distantly related species, and
therefore the risk of immune rejection is high for grafts
between these animals. Furthermore, while there is growing literature exploring the factors underlying the rejec(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1391
A
RPE
Cytokeratin
Subretinal Space
ONL
INL
IPL
B
RPE65
Figure 4. Immunolabeling of a noninjured retina for the retinal pigment
epithelium (RPE) markers cytokeratin (A) and RPE65 (B) (scale bars, 180
µm). No double-labeling of the green fluorescent protein–positive cells
integrated into the RPE layer was observed. ONL indicates outer nuclear
layer; INL, inner nuclear layer; and IPL, inner plexiform layer.
tion of porcine cells transplanted to the mouse,15-17 the risk
of rejection is at least as great when the pig, with its more
sophisticated immune system, serves as the recipient. Nevertheless, the absence of available porcine retinal progenitors together with a number of mitigating considerations
led to the decision to go forward with this study. These
included the absence of passenger leukocytes in the grafts,
the lack of detectable major histocompatibility complex
class I or class II expression by murine central nervous system progenitor cells,18,19 and the relative immune privilege associated with the retina as a host site.20,21
No evidence of migration into the host was observed
30 minutes after transplantation, whereas at the 1-week
and later points GFP⫹ cells had been integrated into the
RPE and all layers of the neural retina, including the nerve
fiber layer. There was variability in the degree of integration between animals, and this appeared to relate to
treatment condition. Retinal integration of GFP⫹ cells
was observed only in animals that received grafts as a dissociated cell suspension. In addition, injuring the retina
with laser burns at the time of transplantation appeared
to promote retinal and RPE integration. Further support for this association comes from the observation that
integration of GFP⫹ cells was centered on areas of laser
injury. This tendency toward wound tropism is in accord with a number of other studies showing that developmental immaturity,8,11 lesions, or genetic dystrophies9,22-24 strongly enhance the migration and integration
of grafted progenitor cells. A corollary of this tropism concept is that, in the setting of disease, grafted progenitor
cells should home to sites where they are needed and avoid
interfering with healthy areas of the retina. Compared
with previous studies of progenitor cell integration in the
retina, the significance of the current study centers on
the fact that the porcine eye is much closer to that of huWWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
RPE
RPE
ONL
INL
IPL
GCL
A
B
Vitreous
C
GFAP
Figure 5. Retinal progenitor spheres after transplantation to a pig that was laser treated. A, A green fluorescent protein–positive sphere in the vitreous (scale bar,
60 µm). B, Green fluorescent protein–positive cells in the vitreous, in the subretinal space, and a few in the retina. Donor cells were found as either spheres or
single cells, a few of which had matured to cells with processes (scale bar, 180 µm). RPE indicates retinal pigment epithelium; ONL, outer nuclear layer; INL, inner
nuclear layer; IPL, inner plexiform layer; and GCL, ganglion cell layer. C, As seen in the pigs grafted with cell suspension, green fluorescent protein–positive cells
double-labeled with glial fibrillary acidic protein (GFAP) (arrowheads). Also evident in this image is up-regulation of GFAP in the host retina (scale bar, 45 µm).
Figure 6. Polymer-progenitor composite within the subretinal space. The
dashed line represents the border of the polymer. The green fluorescent
protein–positive donor cells have a more differentiated appearance and have
extended long slender processes that span the cross section of the polymer
(scale bar, 90 µm).
mans. The vitreous cavity is much larger than that of rodents, while the retina is thicker and contains many more
cones. Thus, the results obtained in this large animal
model are of particular relevance in the context of the
potential clinical applications.
The integration of progenitor cells into the RPE is an
interesting finding, although the grafted cells did not express either of the 2 RPE-related genes examined. Many
studies have shown that transplanted RPE cells tend to
adhere to each other in the subretinal space rather than
integrating into the host RPE monolayer or binding extensively to Bruch membrane.25-28 In addition, integration into the RPE layer has not been seen in many previous studies of central nervous system progenitor
transplantation to the retina.4,8,9 Integration of progenitor cells into the RPE layer was first described by
Warfvinge and coworkers7 by using an immortalized rat
neural precursor line in the rat. In the present study, the
GFP⫹ mouse RPCs were integrated into the pig RPE layer
in a manner reminiscent of the earlier rat study, despite
the fact that these cells do not exhibit this behavior when
grafted to mice.4 Although many questions remain to be
answered, particularly with respect to the long-term functional capabilities of grafted cells, the phenomenon of RPE
integration raises the possibility of using progenitor cells
in diseases with RPE defects, such as age-related macular degeneration.
Apart from the 1 case of RPE integration mentioned already, cells grafted as spheres showed no evidence of mi(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1392
gration into the host tissue. Interestingly, the GFP⫹ cells
within the scaffolds survived and extended processes with
a radial orientation. On the basis of these results, there is
little rationale for transplanting cells as spheres in the context of this model; however, the polymer-progenitor composite grafts may provide a means of transplanting cells
to the subretinal space in an organized manner. This may
be of interest in situations where high densities of donor
cells are required to reconstruct the outer nuclear layer,
while a widespread migration into the inner retina is not
necessary. In addition, the polymer component of the graft
can be modified to provide a vehicle for sustained delivery of a range of therapeutic agents.29,30
In culture, and after transplantation to the retina of
mice, RPCs from the GFP mouse can be induced to differentiate into glia and neurons, including presumptive
photoreceptors.4 After transplantation of these cells to
the pig, a similar yet less clear-cut picture emerges. Morphologically, the cells appeared to differentiate into both
glial and neuronal phenotypes. However, marker studies only partially confirmed this impression.
Cells prepared as single-cell dissociates, as well as cells
grown as spheres, displayed immunoreactivity to the proliferative marker PCNA, the immature cell marker nestin,
the glial marker GFAP, and the neuronal marker MAP-2.
After grafting, there was evidence of expression of GFAP
and, to a limited extent, nestin and MAP-2. However, markers specific for retinal neurons were not seen. This implies
that the proliferative capacity has ceased and a downregulation of nestin and MAP-2 has occurred over time.
Incomplete expression of markers by progenitor cells
after transplantation has been previously reported,8 although the basis of this phenomenon remains to be elucidated. The cells in this case started from a developmentally immature state, as evidenced by widespread
nestin expression. The persistence of nestin in GFP⫹ cells
at later points suggests that many cells remained developmentally immature after transplantation. Insufficient
instructional cues may have been available in the porcine retinal microenvironment to provide adequate phenotypic direction to the grafted cells, thereby resulting
in incomplete differentiation. Yet another possibility is
phenotypic restriction of the donor cells, either in culture or as an artifact of the transplantation process. DeWWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.
spite the unanswered questions, this study shows that sufficient cues are present in the mature retina of the pig to
allow for a remarkable degree of cross-species integration, even when progenitor cells from a rodent are used.
In conclusion, this study demonstrates that the phenomenon of progenitor cell integration seen in the neural retina and RPE of rodents can be replicated in a large
animal model. While the degree of integration and differentiation achieved were somewhat less extensive than
those seen in mice, rats, and the Brazilian opossum, many
options remain available for further development of the
pig model, including the development of an allograft
model through the isolation of porcine retinal progenitors,31 together with the possibility of grafting into a porcine model of retinal degeneration such as the rhodopsin, transgenic pig.32 In addition, a major advantage of
the pig is that it allows for much more detailed functional analysis than is possible in rodents. Finally, the
large eye of the pig provides a convenient proving ground
for bioengineering strategies, such as cell delivery on biodegradable polymer scaffolds as well as controlled drug
delivery, both of which are likely to play an increasingly
important role in retinal reconstruction research.
Submitted for Publication: November 2, 2004; final revision received January 26, 2005; accepted February 2,
2005.
Correspondence: Karin Warfvinge, Department of Ophthalmology, Lund University Hospital, S-221 84 Lund,
Sweden.
Financial Disclosure: None.
Funding/Support: This study was supported by the
Richard D. and Gail Siegal Foundation, New York, NY
(Drs Lavik, Langer, and Young); the Crown Princess Margareta’s Committee for the Blind, Stockholm, Sweden (Dr
Warfvinge); the Swedish Association of the Visually Impaired, Stockholm (Dr Warfvinge); the Swedish Science
Council, Stockholm (Medicine) (Dr Warfvinge); Second ONCE International Award for New Technologies
for the Blind, Madrid, Spain (Dr Warfvinge); the Danish
Eye Foundation, Copenhagen (Drs Kiilgaard and Scherfig); grant NS44060 from the National Institutes of Health,
Bethesda, Md; the CHOC Foundation, Guilds, and Padrinos, Orange, Calif (Dr Klassen); and the Minda de Gunzburg Center for Retinal Transplantation, Schepens Eye
Research Institute, Boston, Mass (Dr Young).
REFERENCES
1. Gouras P, Tanabe T. Survival and integration of neural retinal transplants in rd
mice. Graefes Arch Clin Exp Ophthalmol. 2003;241:403-409.
2. Wojciechowski AB, Englund U, Lundberg C, Warfvinge K. The migratory capacity of brain-derived precursor cells and the host glial response, after subretinal
transplantation to normal adult rats. Glia. 2004;47:58-67.
3. Klassen H, Sakaguchi DS, Young MJ. Stem cells and retinal repair. Prog Retin
Eye Res. 2004;23:149-181.
4. Shatos MA, Mizumoto K, Mizumoto H, et al. Multipotent stem cells from the brain
and retina of green mice. J Reg Med. 2001;2:13-15.
5. Lavik EB, Klassen H, Warfvinge K, Langer R, Young MJ. Fabrication of degradable polymer scaffolds to direct the integration and differentiation of retinal
progenitors. Biomaterials. 2005;26:3187-3196.
6. Kiilgaard JF, Scherfig E, Klassen H, Warfvinge K, Prause JU, Young MJ. Transplantation of xenogeneic retinal stem cells to pig subretinal space [abstract]. In-
(REPRINTED) ARCH OPHTHALMOL / VOL 123, OCT 2005
1393
vest Ophthalmol Vis Sci. 2003;44:e-abstract 487.
7. Warfvinge K, Kamme C, Englund U, Wictorin K. Retinal integration of grafts of
brain-derived precursor cell lines implanted subretinally into adult, normal rats.
Exp Neurol. 2001;169:1-12.
8. Takahashi M, Palmer TD, Takahashi J, Gage F. Widespread integration and survival of adult-derived neural progenitor cells in the developing optic retina. Mol
Cell Neurosci. 1998;12:340-348.
9. Young MJ, Ray J, Whiteley SJO, Klassen H, Gage FH. Neural differentiation and morphological integration of hippocampal progenitor cells transplanted to the retina of
immature and mature dystrophic rats. Mol Cell Neurosci. 2000;16:197-205.
10. Mizumoto H, Mizumoto K, Whiteley SJO, et al. Transplantation of human neural
progenitor cells to the vitreous cavity of the Royal College of Surgeons rat. Cell
Transplant. 2001;10:223-233.
11. Van Hoffelen SJ, Young MJ, Shatos MA, Sakaguchi DS. Incorporation of murine
brain progenitor cells into the developing mammalian retina. Invest Ophthalmol
Vis Sci. 2003;44:426-434.
12. Lewis GP, Fisher SK. Müller cell outgrowth after retinal detachment: association
with cone photoreceptors. Invest Ophthalmol Vis Sci. 2000;41:1542-1545.
13. Lewis GP, Charteris DG, Sethi CS, et al. The ability of rapid retinal reattachment
to stop or reverse the cellular and molecular events initiated by detachment. Invest Ophthalmol Vis Sci. 2002;43:2412-2420.
14. Fisher SK, Lewis GP. Müller cell and neuronal remodeling in retinal detachment
and reattachment and their potential consequences for visual recovery: a review
and reconsideration of recent data. Vision Res. 2003;43:887-897.
15. Wennberg L, Song Z, Bennet W, et al. Importance of the Gal alpha1-3 Gal antigen in discordant islet xenotransplantation: immunosuppression, which inhibits porcine islet xenograft rejection in ordinary mice, is equally effective in Galknockout mice. Transplantation. 2004;77:1275-1280.
16. Costa C, Pizzolato MC, Shen Y, Wang Y, Fodor WL. CD86 blockade in genetically modified porcine cells delays xenograft rejection by inhibiting T-cell and
NK-cell activation. Cell Transplant. 2004;13:75-87.
17. McKenzie IF, Li YQ, Xing PX, et al. CD46 protects pig islets from antibody but
not cell-mediated destruction in the mouse. Xenotransplantation. 2003;10:
615-621.
18. Klassen H, Schwartz MR, Bailey AH, Young MJ. Surface markers expressed by
multipotent human and mouse neural progenitor cells include tetraspanins and
non-protein epitopes. Neurosci Lett. 2001;312:180-182.
19. Hori J, Ng TF, Shatos M, et al. Neural progenitor cells lack immunogenicity and
resist destruction as allografts. Stem Cells. 2003;21:405-416.
20. Streilein JW. Ocular immune privilege and the Faustian dilemma: the Proctor
Lecture. Invest Ophthalmol Vis Sci. 1996;37:1940-1950.
21. Streilein JW, Ksander BR, Taylor AW. Immune deviation in relation to ocular immune privilege. J Immunol. 1997;158:3557-3560.
22. Brustle O, Jones KN, Learish RD, et al. Embryonic stem cell-derived glial precursors: a source of myelinating transplants. Science. 1999;285:754-756.
23. Aboody KS, Brown A, Rainov NG, et al. Neural stem cells display extensive tropism for pathology in adult brain: evidence from intracranial gliomas. Proc Natl
Acad Sci U S A. 2000;97:12846-12851.
24. Otani A, Kinder K, Ewalt K, et al. Bone marrow–derived stem cells target retinal
astrocytes and can promote or inhibit retinal angiogenesis. Nat Med. 2002;
8:1004-1010.
25. Algvere PV, Berglin L, Gouras P, Sheng Y, Kopp ED. Transplantation of RPE in
age-related macular degeneration: observations in disciform lesions and dry RPE
atrophy. Graefes Arch Clin Exp Ophthalmol. 1997;235:149-158.
26. Del Priore LV, Tezel TH, Kaplan HJ. Survival of allogeneic porcine retinal pigment epithelial sheets after subretinal transplantation. Invest Ophthalmol Vis Sci.
2004;45:985-992.
27. Wang H, Yagi F, Cheewatrakoolpong N, Sugino IK, Zarbin MA. Short-term study
of retinal pigment epithelium sheet transplants onto Bruch’s membrane. Exp Eye
Res. 2004;78:53-65.
28. Klassen H, Whiteley SJ, Young MJ, Lund RD. Graft location affects functional
rescue following RPE cell transplantation in the RCS rat. Exp Neurol. 2001;
169:114-121.
29. Langer R, Brem H, Tapper D. Biocompatibility of polymeric delivery systems for
macromolecules. J Biomed Mater Res. 1981;15:267-277.
30. Murray J, Brown L, Langer R. Controlled release of microquantities of
macromolecules. Cancer Drug Deliv. 1984;1:119-123.
31. Shatos MA, Klassen H, Scherfig E, et al. Isolation, characterization and expansion of porcine retinal progenitor cells [abstract]. Invest Ophthalmol Vis Sci. 2003;
44:e-abstract 1694.
32. Petters RM, Alexander CA, Wells KD, et al. Genetically engineered large animal
model for studying cone photoreceptor survival and degeneration in retinitis
pigmentosa. Nat Biotechnol. 1997;15:965-970.
WWW.ARCHOPHTHALMOL.COM
Downloaded from www.archophthalmol.com at Danmarks Natur og Laegevidenskabelige Bibliotek, on February 13, 2006
©2005 American Medical Association. All rights reserved.