Cell Transplantation, Vol. 12, pp. 717731, 2003 0963-6897/03 $20.00 + .

00
Printed in the USA. All rights reserved. Copyright 2003 Cognizant Comm. Corp.
www.cognizantcommunication.com

Generation of 3D Retina-Like Structures From a Human Retinal Cell Line

in a NASA Bioreactor
Kamla Dutt,* Sandra Harris-Hooker,* Debra Ellerson, Dione Layne, Ravindra Kumar,*
and Richard Hunt

Departments of *Pathology, Medicine, and Microbiology, Biochemistry & Immunology, Morehouse School of Medicine,
Atlanta, GA 30310-1495
Department of Microbiology, University of South Carolina, Medical School, Columbia, SC

Replacement of damaged cells is a promising approach for treatment of age-related macular degeneration
(AMD) and retinitis pigmentosa (RP); however, availability of donor tissue for transplantation remains a major
obstacle. Key factors for successful engineering of a tissue include the identification of a neural cell line
that is: homogeneous but can be expanded to give rise to multiple cells types; is nontumorigenic, yet capable
of secreting neurotrophic factors; and is able to form three-dimensional (3D), differentiated structures. The
goal of this study was to test the feasibility of tissue engineering from a multipotential human retinal cell
line using a NASA-developed bioreactor. A multipotential human retinal precursor cell line was used to
generate 3D structures. In addition, retinal pigment epithelium (RPE) cells were cocultured with neural cells
to determine if 3D retinal structures could be generated in the bioreactor with cells grown on laminin-coated
cytodex 3 beads. Cell growth, morphology, and differentiation were monitored by light and scanning electron
microscopy, Western blot analysis, and analysis of glucose use and lactate production. The neuronal retinal
precursor cell line cultured in a bioreactor gave rise to most retinal cell types seen in monolayer culture.
They formed composite structures with cell-covered beads associated with one another in a tissue-like array.
The beginning of layering and/or separation of cell types was observed. The neuronal cell types previously
seen in monolayer cultures were also seen in the bioreactor. Some of the retinal cells differentiate into
photoreceptors in the bioreactor with well-developed outer segment-like structures, a process that is critical
for retinal function. Moreover, the neuronal cells that were generated resembled their in vivo phenotype
more closely than those grown under other conditions. Outer segments were almost never seen in the mono-
layer cultures, even in the presence of photoreceptor-inducing growth factors such as basic fibroblast growth
factor (bFGF) and transforming growth factor (TGF-). Muller cells were occasionally seen when retinal,
RPE cells were cocultured with retinal cells in the bioreactor. These have never been seen in this retinal cell
line before. Cells grown in the bioreactor expressed several proteins specific for the retinal cell types: opsin,
protein kinase C-, dopamine receptor D4, tyrosine hydroxylase, and calbindin.

Key words: Tissue engineering; Bioreactor; Human retinal precursors; Neurons; Photoreceptors

INTRODUCTION for replacement of degenerated retinal tissue, but previ-

Retinitis pigmentosa (RP) and age-related macular
degeneration (AMD) are the leading causes of blindness
worldwide (4,22). RP results primarily from photorecep-
tor degeneration and/or retinal pigment epithelium (RPE)
damage (32,35,39,41). Retinal transplantation to restore
visual function together with the delivery of corrective
genes or growth factors to photoreceptors has been sug-
gested as a potential therapeutic approach. In addition,
electrical stimulation of the inner retina and grafting of
various tissues have been reported as possible therapies
(5,8,17,20,21,23,24). Grafting is one promising approach
ous experimental attempts have met with only limited
success (1,14,17,20), often due to the lack of available
tissue (9,28). In animal models, graft material used to
restore damaged photoreceptors has varied from normal
embryonic rat retina (3,26,47,48,52) and adult rat retina
(44), to cotransplantation of retina RPE sheets (26,35,44,
47). Others have performed cross-species grafts in a rab-
bit model and have reported survival of up to 11 months
when full thickness embryonic retinas were transplanted
(17). Modified cells have also been employed to rescue
degenerating retinal tissues (21,30,46).
However, the frequency of success of transplants re-

Accepted April 29, 2003.

Address correspondence to Kamla Dutt, Ph.D., Department of Pathology, Morehouse School of Medicine, 720 Westview Drive, S.W., Atlanta, GA
30310-1495. Tel: (404) 752-1769; Fax: (404) 752-1108; E-mail: duttk@msm.edu

717
718
mains variable and low. Even if unlimited quantities of
fetal retinal tissues were available, they pose a variety of
problems. These include cell heterogeneity, possible vi-
ral contamination, and the capacity for renewal. The
ability to differentiate and secrete neurotrophic factors
necessary for survival and optimal functioning is equally
problematic. Additionally, ethical issues in using human
fetal tissue loom high. Transplant of photoreceptors (28)
from cadavers, though very promising, will pose similar
problems and the availability of human donor tissues
will remain a major obstacle. Fetal tissue, though ex-
tremely effective in alleviating symptoms of Parkinsons
disease (51), may not be appropriate for retinal trans-
plants and no data are currently available on the suitabil-
ity of fetal retinal human transplants to generate adult
functional retina.
Tissue engineering provides a viable approach for an
adequate source of differentiated cells, and some success
has been achieved in generating various tissues (skin,
cartilage, bone, and liver) by this method (31). This ap-
proach, as a strategy to achieve organ replacement and
tissue repair for a variety of therapeutic needs, has been
quite promising in studies of Parkinsons disease, hemo-
philia, damaged skin, cartilage, blood vessels, etc. (31).
However, the success of a strategy for engineering neu-
ral tissue depends on the availability of: neural precursor
cells that have the capacity to expand and give rise to
various cell types; cells that remain neurotrophic and
retain an identifiable marker and are amenable to modi-
fication; and substrates that permit the generation of
three-dimensional (3D) structures.
The cell line used in these studies is the multipoten-
tial human retinal (progenitor) cell line developed and
characterized by us previously (11,12,15). The cell line
grows as a monolayer in contact-inhibited manner. A small
percentage of cells in this cell line acquire a neuron-like
phenotype on their own. Additionally, the cell line can
be induced to differentiate in response to differentiation-
inducing agents such as dibutyryl adenosine 3-5 cyclic
monophosphate (cAMP), 12-O-tetradecanoyl phorbol-
13-acetate (TPA), etc. cAMP promotes differentiation
into multipolar neurons, whereas, TPA induces differen-
tiation towards a bipolar phenotype (photoreceptors,
bipolars). Basic fibroblast growth factor (bFGF) and
transforming growth factor- (TGF-) promote differ-
entiation towards a photoreceptor phenotype (15).
In this study, the multipotential human retinal cell
line was cultured in a NASA-developed, horizontally
rotating bioreactor (HRB). This system was initially de-
veloped to simulate some aspects of microgravity; how-
ever, it has also proven valuable in the growth of both
normal and cancerous cells (2,10,16,25,34,36,42,43). The
system allows interaction between similar and dissimilar
cell types and promotes spatial interaction between cell
types (18,19,45). Growth of anchorage-dependent cells
DUTT ET AL.
in three dimensions with multiple layers within the frame-
work provided is a major aspect of this system. Various
culture systems have been compared for growth, mono-
layers, static matrix cultures, roller bottles, cell suspen-
sion, and hollow tubes; of these, only HRB provides low
shear, high density, and three-dimensional growth and
tissue assembly (18,29,42). It is reported here that HRBs
could prove extremely valuable in the generation of 3D
structures from precursor retinal cells and could also
allow coculture of various cell types.
MATERIALS AND METHODS
The human retinal progenitor cell line KGLDMSM
was used (1113,15) at 2030 passages. D407 human
retinal pigment epithelial cells (7) were used at 6070
passages. Cytodex 3 (175 m) beads were obtained from
Pharmacia (Piscataway, NJ). Laminin (10 g/ml) (Col-
laborative Research) was used to coat the cytodex 3 beads.
In some experiments cytodex 3 beads were coated with
polylysine (Sigma Chemical, St. Louis, MO) for 1 h and
then coated with laminin after three extensive washes in
distilled water and phosphate-buffered saline (PBS).
Rotating-Wall Vessel and Control Cultures
The cells were cultured in Dulbeccos modified Eagle
medium/nutrient mixture F-12 Ham (1:1) (GIBCO, Bethes-
da, MD) and supplemented with 10% fetal bovine serum
(Hyclone Laboratories, Logan, UT), 10% serum supple-
ment (JRH Biosciences, Lenexa, KS), 2 mM L-gluta-
mine, 0.075% sodium bicarbonate, penicillin 100 U/ml,
and streptomycin 100 g/ml (GIBCO). Cells were seeded
in the above medium as previously described (11
13,15) in a 55-ml bioreactor rotating-wall culture vessel
(Synthecon, Inc., Friendswood, TX). Cytodex 3 micro-
carrier beads (175 m) (Pharmacia) were introduced
into the vessels at a density of 5 mg/ml and cells were
seeded at a density of 0.51 105 to 1 106 cells/ml
(510 cells/microcarrier). In coculture, the ratio of RPE
to retinal cells was 1:3. Six to nine 106 transformed hu-
man retinal cells (11,12) were cocultured with 3 106
D407, the human retinal pigment epithelial cell line (7).
All cultures were monitored daily for pH, dissolved CO2
and O2, and glucose utilization and were refed accord-
ingly. The feeding of stationary control cultures was
matched to that required by the bioreactor vessel cul-
tures. Control stirrer cultures and adherent monolayer
cultures were incubated under the same conditions. Ves-
sel rotation was initiated at 45 rpm and increased to 10
rpm as required to maintain cells in suspension.
Determination of Cell Growth
To determine cell growth and viability, samples were
drawn from the bioreactor and stirrer cultures prior to
feeding. Cells were detached from the beads using 0.25%
trypsin (w/v) and 0.1% EDTA (w/v). The cell/bead mix-
RETINAL TISSUE ENGINEERING
ture was centrifuged at 200300 rpm to remove the
beads. A viable cell number was determined by trypan
blue exclusion and cell counts were determined by use
of a hemocytometer and Coulter counter. Three sets of
counts from a single sample drawn on each day were
averaged and the experiments were repeated four times.
No standard error bars (Fig. 2) are drawn as the three
set of counts were from the same sample and as such
there was very little variability.
Determination of Glucose Utilization
and Lactate Production
Media (samples) removed from cultures were put into
tubes and placed on ice. The samples were presented to
the probe of the NOVA biomedical analyzer; after the
sample was positioned, the analyzer was activated. The
pump turns, pulling the sample through the probe, S-
line, sample preheater, and flow cell. The probe auto-
matically withdraws and the results of the analysis are
displayed on the computer screen.
Glucose utilization was determined as follows. The
glucose utilization graph represented the amount of glu-
cose at day 0 minus glucose at a specific day (for exam-
ple day 10) divided by the number of cells. Therefore,
the result at day 10 for cells in the spinner flask was:
375 mg/dl glucose (day 0) 125 mg/dl glucose (day 10)
100 (# cells) = 2.5. Result at day 10 for cells in the bio-
reactor was: 375 mg/dl glucose (day 0) 275 mg/dl glu-
cose (day 10) 40 (# cells) = 2.5.
Immunocytochemistry
Cell aggregates at various time points were fixed
with either acetone at 20C for 20 min or 2% ice-cold
paraformaldehyde in PBS for 2060 min. To block non-
specific binding to antibodies, cell aggregates were
preincubated with nonspecific antisera for 2 h. Cell ag-
gregates were incubated with primary antibodies against
neuron-specific enolase (NSE), glial fibrillary acidic
protein (GFAP), and neurofilament (NF) protein (200
kDa). These three antibodies were from Boehringer
Mannheim (Indianapolis, IN). Opsin antibody (Rho 4D,
ID4) was generously shared by L. and R. Molday (De-
partment of Biochemistry, British Columbia, Vancou-
ver, BC) (38) and another opsin antibody was shared by
P. Hargrave (Univ. of Florida, Gainesville). Incubation
with the primary antibody was for 14 h followed by
reaction with the appropriate fluorescent secondary anti-
bodies. In each experiment, appropriate controls were
included, omitting primary antibody in one sample and
omitting secondary antibody in the second set.
Scanning Electron Microscopy
Aggregates from stirrer or bioreactor cultures were
fixed in 2.5% glutaraldehyde (Polyscience, Warrington,
PA), 3% paraformaldehyde (EM Science, Fort Washing-
719
ton, PA) in 0.1 M cacodylate buffer, pH 7.4, for 2 h.
Samples were then rinsed twice in the same buffer and
fixed in 1% osmium tetroxide in 0.1 M cacodylate buffer,
pH 7.4, for 1 h. After rinsing several times with water,
the samples were dehydrated in a graded ethanol series
and kept in 100% ethanol for critical-point drying using
a Samdri-780A (Tousimis Research, Rockville, MD),
after which they were mounted on stubs and sputter-
coated with gold-palladium for 2 min at 15 mA. The
samples were then examined using a Joel Scanning Mi-
croscope JSM 820.
Western Blot Analysis
At appropriate times, cells attached to microcarrier
beads from bioreactor culture and monolayer culture
were washed three times in cold PBS. Cells were lysed
in lysis buffer (50 mM Tris, pH7.4, 150 mM NaCl, 1
mM DTT, 0.5% Triton X-100, 0.2 mM AEBSF, 0.5 mM
benzamidine, 2 g/ml aprotinin, 2 g/ml chymostatin,
0.5 mM leupetin, 5 mg/ml pepstatin A, and 0.25 mg/
ml soybean tyrosine inhibitor). The resulting supernatant
was used for Western blot analysis. Whole-cell extracts
containing 50 g proteins were separated by electropho-
resis on 10% SDS polyacrylamide gels and electropho-
retically transferred to a polyvinylidene difluoride (PVDF)
membrane (Immobilon P; Millipore Corp., CA) in Tris-
glycine buffer containing methanol and SDS. The non-
specific binding sites were blocked by immersing the
membrane in 5% (w/v) nonfat dry milk in TBS-T (25
mM Tris-HCl, pH 7.6, 150 mM NaCl, and 0.05%
Tween-20) for 2 h at room temperature on an orbital
shaker. Membranes were rinsed with two changes of
washing buffer. The membranes were reacted with spe-
cific primary antibodies to determine the presence of
specific retinal proteins. The antibody for opsin detec-
tion was ID4 (against the carboxy terminal of rhodopsin)
and was generously shared by R. Molday and L. Mol-
day. Antibodies for tyrosine hydroxylase, D4 receptor
protein, calbindin, and protein kinase C- (PKC-) were
procured from Chemicon Inc. (Tamecula, CA) and the
antibody for neuron specific enolase was from Zymed
Corporation (San Francisco, CA). After reaction of the
membranes with appropriate primary antibodies, the
membranes were reacted with horseradish peroxidase-
conjugated appropriate secondary antibodies. Detection
was done with chemiluminescent substrates and x-ray
film exposure.
RESULTS
Description of Retinal Cell Line Used in These Studies
A human retinal cell line was established from a first
trimester, spontaneously aborted fetus by SV-40T DNA
transfection (11,12,15). The cell line has been cloned
and subcloned and grows as a monolayer in a contact-
inhibited manner. It is nontumorigenic and serum depen-
720 DUTT ET AL.

dent. The doubling time varies from 16 to 24 h. The
culture is composed of a monolayer of flat precursor
cells overlaid with neurons (Fig. 1, arrows point to neu-
ronal cells and arrowheads point to precursors). A small
percentage of cells in this line differentiate in serum-
containing medium to give rise to the neuronal pheno-
types in the retina. The phenotype of all the neurons
present in adult retina is replicated in these cultures
(11,12). Neurons with large cell bodies and multiple
processes are possible ganglion cells. Neurons with
small cell bodies and two long, thin neuritic processes
are possibly bipolar cells. Rods with one short and one
long neuritic process, cones with one short neuritic pro-
cess, and cells resembling amacrine and horizontal cell
phenotype are also seen. Immunocytochemistry con-
firms the phenotype of most of the neurons identified
morphologically as already reported (12,13). Differenti-
ation can be induced in the cell line by exposure to

Figure 1. Phase contrast micrographs of the retinal cell line used in this study. Human retinal cell
line (SV-40T transformed) in 22nd passage (A) and 30th passage (B). The cells grow in a mono-
layer with a flat layer of precursor cells (arrowheads) overlaid with neurons (arrows). The neuronal
cells have small cell bodies, and thin neuritic (arrowheads) processes. Original magnification 400.
RETINAL TISSUE ENGINEERING 721

cAMP, which promoted all the precursors to differenti-

ate into multipolar neurons under serum-free conditions.
Exposure to TPA promoted a bipolar, rod photoreceptor
phenotype (12), while exposure to bFGF and TGF- en-
hanced the photoreceptor-like phenotype (15). Although
this cell line is multipotential, expression of a glial phe-
notype or glial antigens has never been seen, suggesting
that there is a limit to the multipotential nature of these
cells. The RPE cell line used in this study has been de-
scribed previously (7).

Growth Kinetics and Metabolic Studies of Cells

Retinal cells were grown in the bioreactor attached
to the laminin-coated cytodex beads. Parallel stirrer and
monolayer cultures were started at the same cell density.
All three culture sets (bioreactor, stirrer, and monolayer)
were fed daily. The monolayer cultures were terminated,
as they were overgrown by day 3. Much higher cell den-
sities were reached in stirrer versus HRB cultures (Fig.
2A). However, pH levels, glucose consumption (Fig.
2B), and the production of lactate levels (Fig. 2C) were
similar in both stirrer and HRB cultures, pointing to op-
timal O2 content and access to nutrients.

Glucose Utilization
The glucose utilization graph represents the amount
of glucose at day 0 minus glucose at a specific day (e.g.,
day 10) divided by the number of cells. Therefore, the
result at day 10 for cells that were in the spinner flask
as well as in the bioreactor was 2.5.
The viability of cells between day 1 and day 18 in
the HRB cultures was greater than 90% and in the stirrer
cultures between 80% and 90% as judged by trypan ex-
clusion. Results were very similar in four different sets
of experiments.

Phenotype of the Cells Grown in the HRB

Observation of the morphology of the retinal cells
grown in the HRB from days 1 to 13 indicated that
some beads were totally covered with cells while some
remained naked. As the time in culture progressed, cells
growing on separate beads began to associate to form
bridges, thus enlarging the 3D aggregate structures. As
the tissue mass grew, it covered the beads. Growth
between the beads resulted in tissue-like structures and
the elimination of beads from the tissue mass. Cell
aggregates of 20100 beads were formed, reaching the
dimension of several hundred microns visible to the
naked eye.
Our data on cell density versus growth and differenti-
ation are similar to those on the monolayer culture. At
lower cell densities (1 105 cells/ml) on day 3, a higher
number of cells acquired a neuronal phenotype. In Fig-
Figure 2. Growth kinetics and status of retinal cells grown in
the bioreactor () and stirrer cultures (). Much higher cell
densities were reached in stirrer cultures compared with biore-
actor cultures (A). Glucose utilization (B) and lactate produc-
tion (C) were similar and corresponded to the cell numbers in
each culture vessel.
ure 3A, all the cells show a neuronal phenotype with a
large cell body and multiple processes. One cell shown
in Figure 3B is possibly a photoreceptor with two axons
(arrows), and a well-rounded cell with a single axon (ar-
rowhead) is a possible ganglion cell. Several neurons in
722 DUTT ET AL.

Figure 3. Scanning electron micrograph of retinal cells grown in the bioreactor at a lower cell density of 1 105 cells/ml (day 3).
(A) Precursor cells dividing and giving rise to neurons, which are mostly ganglionic (large cell bodies and multiple processes) in
phenotype (arrows). (B) A possible photoreceptor cell traverses the whole length of the bead (arrow). A cell with large cell body
and single process (arrowhead), possible ganglion cell. (C) Several cells resembling amacrine and horizontal cells (arrows). (D)
Possible photoreceptor with single neuritic process (arrows) and several multipolar neurons (arrowheads). (E) Several differentiating
neurons. (F) One neuron from (E) enlarged (arrow); note the axon with branching processes.

Figure 4. Scanning electron micrograph of retinal cells grown in the bioreactor at a higher cell
density (1 106 cells/ml) on laminin (10 g/ml)-coated cytodex beads on day 12. (A, B) A large
tissue mass (arrows). This tissue aggregate was made up of between 40 and 50 beads. (B) As the
tissue mass grows, beads appear to be ejected (arrowheads). (C) Tissue-like structure growing
between beads (arrows). (D) Tissue-like multilayered structure growing without beads (arrows).
RETINAL TISSUE ENGINEERING 723

the process of differentiation are shown in Figure 3C
(arrows) while Figure 3DF depicts multipolar neurons
(arrowheads) and differentiated neurons that are possible
photoreceptors (arrows, note bipolar polarized appear-
ance). At higher densities (1 106 cells/ml), the cells with
neuronal phenotype are proportionately reduced. Retinal
cells grown in the bioreactor aggregated to form multi-
ple cell-coated beads. In addition, cellular aggregates
without beads formed 3D structures in suspension (Figs.
46). Initially, the tissue-like mass grew attached to the
beads but as the size increased the cells separated from
the beads while continuing to grow into a tissue-like
structure. Figure 4A shows an aggregate of cells in a
large tissue mass (arrows). As these aggregates grew,
the beads appeared to be ejected (Fig. 4B, C, arrow-
head). Figure 4D shows tissue growing without beads.
In four separate experiments, separation of cells and the
onset of possible layering was seen. Cells with the phe-
notype of rods and cones were always at the periphery
of the beads projecting from the cellular aggregates (Fig.
5C). These cells appeared to have rudimentary outer
segments. Figure 5C and D shows large multilayer struc-
tures formed around individual beads between days 14
and 18 in the HRB. In Figure 5A and B, arrows indicate
structures developed around the beads (area in brackets)
as the latter were being pushed out. Figure 5C and D
represents multilayered tissue built around the beads
with photoreceptors aligned at the periphery (arrows).
However, we cannot rule out that layering might have
been just cells separating. In Figure 6AG cells identi-
fied as photoreceptors are shown aligned at the periph-
ery of tissue aggregates. In the HRB, both neurons that
resemble rod and cone photoreceptors and well-devel-
oped structures that resemble rod and cone outer seg-
ments are seen as well. Figure 6A shows a photoreceptor
with a rudimentary outer segment while Figure 6B illus-
trates photoreceptors emerging from a tissue mass in the
process of developing an outer segment (arrowheads) and
fully developed outer segments (arrows). Further indica-
tions of the formation of photoreceptors include the for-
mation of connecting cilium-like structures (Fig. 6C),
together with cells with a rod-like (Fig. 6D) and cone-

Figure 5. Scanning electron micrograph of a multilayered tissue mass being formed between days 14 and 18. (A, B) Multilayered
structures seem to be formed around the beads. Arrowheads point to cell separation and/or possible formation of multilayered
structures from which beads appear to have been ejected. (C) In multilayered structure, cells with photoreceptor phenotype are
aligned (arrows). See also the layers demarcated by lines. (D) High magnification micrograph of photoreceptors from (C) with
rudimentary outer segments (arrows).
724 DUTT ET AL.

Figure 6. Scanning electron micrographs of photoreceptor-like cells in tissue mass between days 8 and 14 (high-density 1 106
cells/ml). (A) Photoreceptors with outer segment being formed. (B) Photoreceptors in various stages of development (line with
arrowheads). Arrow points to fully developed photoreceptors with outer segments. (C) Connecting cilium between inner and outer
segments. (D) Possible rod photoreceptors. (E, F) Possible cones. (G) Possible double cones.

like (Fig. 6E, F) phenotype. Figure 7A and B illustrates
neurons with extensive axonal processes extending the
length of a whole bead (arrows). These cells possibly
may be ganglions (large cell body and multiple pro-
cesses). Figure 7C shows a possible bipolar cell (arrow-
head) synapsing on a cone photoreceptor (arrow). In
Figure 7D, possible ganglions with a large cell body and
multiple processes are shown. The cell in Figure 7E re-
sembles an amacrine cell. A group of photoreceptor-like
cells aligned at the periphery of tissue-like structures is
seen in Figure 7F (arrows).
The stirrer culture resembled the HRB culture in the
formation of large bead aggregates and the formation of
tissue-like masses, but no differentiated neuronal struc-
tures could be seen in the aggregates; instead, mostly
precursor cells were observed (Fig. 8AC). Figure 8
represents the high-density stirrer culture (control) cor-
responding to the large tissue masses seen (Figs. 67 in
the bioreactor), but the cells are relatively undifferenti-
ated large cell bodies (arrows) with extensive collagen-
like fibers of uniform size. The fibrils were smaller than
neuritic processes and were usually seen at the point of
attachment of cells to beads (arrowheads in Fig. 8C).
Very few cells showed differentiated neuritic-like pro-
cesses.
Characterization of Retinal Cells Cocultured With RPE
Conditioned medium from RPE cells enhanced the
neuronal phenotype in retinal cells grown in monolayer
culture (15), and when retinal RPE cells were cocultured
in the HRB, some layering was seen. Frequently, cells
sorted out into different layers, and occasionally cells
that resembled Muller cells were observed (Fig. 9C). In
coculture, an enhanced photoreceptor phenotype was
noted. When retinal cells were cocultured with RPE,
mixing of neuronal and RPE cells was initially seen
(Fig. 9A). Occasionally beads showed either neurons or
RPE cells, suggesting that the cells had sorted out. Fig-
ure 9B shows many RPE cells that are very rounded
phenotype (arrow). The arrowhead in this figure indi-
RETINAL TISSUE ENGINEERING 725

Figure 7. Scanning electron micrographs of various neuronal cell types seen in high-density culture on days 814. (A) Many
neurons with extensive axonal processing (arrows). (B) Axons traversing the whole length of cytodex beads (175 m) (arrows).
(C) A bipolar-like cell synapsing on cone-like neuron, bipolar (arrowhead), cone (arrow). (D) A ganglion-like neuron with large
cell body and multiple processes (arrows). (E) A cell with amacrine phenotype. (F) A cluster of possible rods and cones (arrows).

cates a bead covered mostly by neurons with neuritic
processes, possibly photoreceptors. Figure 9D shows
mostly neurons. Figure 9C shows sorted RPE cells that
are very rounded (arrows). The arrowhead shows a large
Muller cell traversing the whole bead with branching
end feet.
Immunophenotyping
Retinal cells grown in the bioreactor alone or be-
tween beads as tissue masses stained positively for the
neuronal marker NSE enolase, as is the case with mono-
layer-grown cells (11,12). Cells resembling photorecep-
tors were positive for ID4, an antibody against rhodop-
sin. This was confirmed by Western blotting (Fig. 11A).
All cells were positive for NSE (Fig. 10A, B). In Figure
10E, a photoreceptor labeled with another opsin anti-
body, A2B5, is shown (provided by Dr. Hargrave, note
polarized labeling for opsin). Figure 10C and D depicts
cells double labeled for NSE and opsin secondary anti-
bodies labeled with rhodamine and fluorescein, respec-
tively.
Western Blot Analysis
To determine the differentiated state of structures
generated in the bioreactor, expression of several other
retinal-specific proteins was determined by Western blot
analysis. Figure 11A depicts opsin expression. Figure
11B-1 shows expression of neuron-specific enolase in
retinal cells cultured alone or cocultured with RPE in
the bioreactor. Tyrosine hydroxylase, a rate-limiting en-
zyme in catecholamine synthesis, was not expressed in
monolayer cultures but upregulated in cells grown in the
bioreactor, specifically on day 5 (Fig. 11B-2). Tyrosine
hydroxylase levels were upregulated in a transient man-
ner on day 5 and dropped considerably by day 10. D4
receptor expressed in photoreceptors was significantly
upregulated in retinal and RPE cells cocultured in the
bioreactor (Fig. 11B-5). Calbindin, a protein expressed
in cone bipolars, was slightly upregulated in retinal cells
grown in the bioreactor with increasing days of culture
(Fig. 11B-4). PKC- was also slightly upregulated in
the bioreactor (Fig. 11 B-3). Protein PKC- has been
implicated in phototransduction, synaptic plasticity, and
in modulating responses of glutamate receptors and do-
pamine release.
DISCUSSION
A variety of experimental protocols are being tested
to correct or postpone the degenerative process that oc-
curs in AMD and RP (5,14,17,26,47,48); of these, tissue
transplantation may be a viable approach. Tissues used
in transplantation include retinal precursors, immortal-
726
Figure 8. Scanning electron micrograph of high-density reti-
nal cell cultures grown in the stirrer and bioreactor on day
814. (A, B) Large tissue masses growing without well-differ-
entiated neuronal phenotype. (C) High magnification picture
of one aggregate (arrows). All cell types have undifferentiated
phenotype: large round cells (arrows) with extensive collagen
fibrils (arrowheads). Extensive collagen-like fibers are never
seen in cells grown in the bioreactor.
ized cell lines, vibrotome-generated sheets of cells, and
photoreceptors isolated from human cadavers (28).
While these strategies hold promise, only limited suc-
cess has been achieved in some models. Even if all prob-
lems associated with tissue transplants were resolved,
DUTT ET AL.
other issues would still impact the success of this ap-
proach; tissue engineering provides a viable alternative.
Success has been achieved in obtaining material for
transplantation to the central nervous system, liver, he-
mopoietic tissue, cartilage, and bone (31). In the case of
neuronal tissues, success of tissue engineering depends
on the availability of homogeneous multipotential neural
precursors and their ability to differentiate. Also impor-
tant are the scaffolds or substrates or the tissue culture
system used to generate 3D structures, together with their
immune acceptance. It is also clear that cellular interac-
tions within multiple cell types generated from precur-
sors and their integration into host tissues are necessary
factors for success.
A standard tissue culture system confines cell-to-cell
interaction along defined surface areas and allows only
2D interactions. Stirrer cultures, though promoting 3D
growth, propel cells into suspension under conditions of
high shear (18,29,42,43). The perfusion gel system pro-
motes growth and interaction of cells, and allows 3D
assembly but is hard to evaluate and analyze. Aggregate
rotary cultures suffer from problems of diffusion-related
necrosis (18,19,45). Horizontally rotating bioreactors,
developed by NASA to simulate microgravity, have
proven very valuable in the growth of a variety of nor-
mal and transformed cell lines. The mechanism by which
the HRB induces differentiation and expression of tis-
sue-specific proteins has been postulated to be mediated
by the provision of unique cell culture hemodynamics,
characterized by 3D growth in low shear environment
(16,29,45,54). When the cultures are initiated in the
HRB, the rotation necessary to keep particles suspended
has been calculated to expose cells to a shear stress of
approximately 0.2 dyn/cm2 on the surface of 150-m-
diameter microcarrier beads that are covered with an av-
erage of 15 cells (29). The HRB induces shear forces
between 0.5 and 1 dyn/cm2 (2,16,18,29). This is in con-
trast with stirrer cultures where shear forces are as high
as 240 dyn/cm2, based on the design of the vessel and
rotation speed (16,29); as the beads aggregate, the speed
of rotation must be increased to keep the particles sus-
pended. However, HRBs produce 3D structures in a
low-shear and low-turbulence environment in which
suspended similar and dissimilar cells can interact to
produce 3D structures resembling the in vivo state.
The HRB has proved very useful for the growth of a
variety of tissues. These include the cultivation of carti-
lage on polymer scaffolding for prosthetic use (10,16),
growth of a variety of tumor cells (2,19), tissue model-
ing in bladder cells (43) and salivary glands (19,34), and
the coculture of cells as a model for tumor metastasis
(19). This method has been previously used to generate
3D cultures of diverse cells from bladder (43), ovarian
(2), cancerous and noncancerous (45), prostate (54), and
RETINAL TISSUE ENGINEERING 727

Figure 9. Scanning electron micrographs of retinal cells and D407 RPE cells cocultured in the
bioreactor on days 814. (A) Aggregate of several beads. Most beads have RPE cells (arrows) and
neuronal cells (arrowheads). (B) Arrowhead shows a bead with neurons aligned at the periphery
and arrows show a bead with mostly polarized RPE cells. (C) Note: RPE cells are aligned in a
row with a very rounded appearance (arrows). (C) Arrowhead points to a large Muller cell travers-
ing the whole bead with well-differentiated branching end feet. (D) This bead has mostly neuronal
cells (arrowheads pointing to neuritic processes).

cartilage (10) cells. More recently, pancreatic islet cell
cultures generated in the HRB are being used in phase
III trials in humans (49,50). The vessel also induces
expression of specific proteins in diverse tissues (49);
for example, pancreatic islet cells prepared in the HRB
maintain production and regulation of insulin secretion
(49,50). The implantation of pancreatic islets has main-
tained normoglycemia for 18 months in diabetic patients
(50). Furthermore, glioblastoma cells grown in an HRB
were much more differentiated than similar cells grown
in monolayer cultures (36). Murine PC12 cells produce
organoid structures in the HRB (33) and human colon
cultures produce signet cells, which secrete mucin. This
cannot be achieved in monolayer cultures. Kaysen et al.
(29) reported alteration in gene expression in the HRB.
In this study, we explored the utility of the HRB in
the propagation of differentiated retina cells. This appa-
ratus allows 3D aggregation under conditions of low
shear and controlled access to O2 and nutrients. This en-
vironment permits cell-to-cell and cell-to-ECM interac-
tions.
Although precursors grown in monolayers are capa-
ble of generating multiple cell types, monolayer cultures
do not allow 3D assembly and cell-to-cell interaction to
the same degree. On the other hand, HRB culturing does
appear to favor the differentiation of neural precursors,
with the generation of unique, 3D neural tissues consist-
ing of primitive proliferating cells on the outside and
cells immunoreactive for neuronal and glial markers on
the inside (36). In this study, it has been shown that
human retinal precursor cells can grow in the HRB into
large cell masses. Moreover, they retain the capacity for
multidirectional differentiation. The beginning of some
layering has also been observed. The system is also
amenable to the coculture of different cell types such as
RPE and endothelial cells (unpublished data). In this
study, we have shown that retinal cells grown in the
HRB express neural cell markers.
We have shown in this study that retinal cells grown
in the bioreactor express several neural cell markers
(opsin, PKC-, calbindin, tyrosine hydroxylase, and D4
receptor). These cells, in addition to expressing the pho-
toreceptor marker protein opsin, are expressing several
retinal-specific proteins (6,27,37,40,53). Tyrosine hy-
728 DUTT ET AL.

Figure 10. Immunophenotyping of the retinal cells grown in the bioreactor on days 8 and 14. (A, B) Cells attached to the bead
are labeled for neuron-specific enolase antibody (NSE). The secondary antibody is fluorescence conjugated. All cells are positive
for neuronal marker (NSE). (C, D) Cells double labeled for neuronal marker NSE (secondary antibody rhodamine conjugated) and
rhodopsin antibody ID4 (secondary antibody fluorescence conjugated). All cells are double labeled for both antibodies. (E) A single
photoreceptor labeled for opsin antibody. The arrowhead shows the polarized distribution of opsin.

droxylase, a rate-limiting enzyme in catecholamine syn-
thesis (27) is not expressed in monolayers but is expressed
in HRB cultured cells. Similarly, calbindin, a calcium
binding protein (37) localized on several retinal cell
types including a subset of cone bipolars, is expressed
in both monolayer and bioreactor cultured cells. PKC-
, a protein implicated in phototransduction (53) and
synaptic plasticity, is also expressed in bioreactor cul-
tured cells. Last but not least, D4 receptor (6,40), which
mediates dopamine-coupled cascade of light responses
to adenyl cyclase and is localized to photoreceptor layer,
is also expressed in the bioreactor culture.
For the first time, photoreceptor-like cells have been
seen with well-developed outer segment-like structures,
and the presence of opsin in these cells has been con-
firmed by immunolabeling and Western blot analysis.
The possibility of developing photoreceptors with outer
segments/inner segments and other retinal cell types
from a multipotential cell line could prove extremely
critical to blinding eye disease. When retinal cells were
cocultured with endothelial cells, the beginning of capil-
lary formation was observed as early as 48 h (unpub-
lished data). A variety of models are currently being tested
for the feasibility of retinal tissue transplant. Our model
offers several attractive features: 1) homogeneous hu-
man retinal precursors are used; 2) the cells remain
multipotential in the HRB; 3) some success in generat-
ing neurons that appear to be mature photoreceptors has
been achieved; 4) the system could prove useful to study
neovascularization in diseases such as diabetic retinopa-
thy and AMD. The system is also amenable to the study
of cell-to-cell interaction seen in retinal development
and survival and in the functioning in the adult eye.
The model that has been described needs to be tested
RETINAL TISSUE ENGINEERING 729

Figure 11. (A, B) Western Blot analysis of the expression of opsin in photoreceptor cells in retinal RPE cocultures after 8 days in
the bioreactor (A) and expression of other retinal proteins (B16). The cells were lysed and proteins separated on 10% acrylamide
gels and processed for Western blot analysis using Novex gel buffers and protocols. (A) Lane 1 immunoreactive 3330-kDa bands
corresponding to rhodopsin. Lane 2 is a protein ladder. (B) The expression of several retinal cell markers; neuron-specific enolase
(NSE), tyrosine hydroxylase (TH), protein kinase C- (PKC-), calbindin, and D4 receptor protein expression in retinal cells grown
alone and as coculture with RPE at days 110. Note significant upregulation in expression of tyrosine hydroxylase (B-2) and D4
receptor protein (B-5) in retinal RPE cells grown in the bioreactor. A slight upregulation of NSE was seen in cells grown in the
bioreactor (B-1) whereas PKC- and calbindin expression did not change significantly from levels expressed in monolayer culture
of retinal cells.
730
for many factors including the orientation of the cell types
generated and the ability of individual cell types to rec-
ognize each other and associate appropriately. The pre-
liminary study, however, is promising. The use of cyto-
dex beads could pose problems in transplant studies.
Eventually, biodegradable scaffolds will have to be
identified that create structures suitable for transplanta-
tion. Engineered structures could prove valuable for the
study of cell-to-cell interaction and the identification of
paracrine/autocrine factors implicated in retinal genera-
tion and regeneration.
AKNOWLEDGMENTS: The authors thank Mrs. Deborah B.
Jones and Renarder Pressley for excellent secretarial assis-
tance, and Ms. Suzanne Alexander and Dr. G. Sanford for crit-
ical reading of the manuscript. We thank Drs. R. and L. Mol-
day (The University of British Columbia, Vancouver) for the
opsin antibody, Rho-ID4, and Dr. P. Hargrave (University of
Florida, Gainesville, Florida) for another opsin antibody,
A2B5. We acknowledge Mr. L. Brako for scanning electron
micrographs and Mr. Patrick Abramson for all aspects of
graphics. The authors also acknowledge the advice provided
by Dr. Clarence Sams, NASA Space and Gravitational Biology
Center, and Dr. Jackie Duke, University of Texas Health Sci-
ence Center. These studies were supported by NASA grants
NCC 9-112 (K.D.), NAG 9-969 (K.D.), and NIH grant
EY10516 (R.C.H.).
REFERENCES
1. Aramant, R.; Seiler, M.; Turner, J. E. Donor age influ-
ences on the success of retinal grafts to adult rat retina.
Invest. Ophthalmol. Vis. Sci. 29:498503; 1988.
2. Becker, J. L.; Prewett, T. L.; Spaulding, G. F.; Goodwin,
T. J. Three-dimensional growth and differentiation of
ovarian tumor cell line in high aspect rotating-wall vessel:
Morphologic and embryologic considerations. J. Cell.
Biochem. 51:283289; 1993.
3. Bergstrom, A.; Ehinger, B.; Wilke, K.; Zucker, C. L.;
Adolph, A. R.; Aramant, R.; Seiler, M. Transplantation of
embryonic retina to the subretinal space in rabbits. Exp.
Eye Res. 55:2937; 1992.
4. Berson, E. L. Retinitis pigmentosa: The Friedenwald Lec-
ture. Invest. Ophthalmol. Vis. Sci. 34:16591676; 1993.
5. Bok, D. Retinal transplantation and gene therapy: Present
realities and future possibilities. Invest. Ophthalmol. Vis.
Sci. 34:473476; 1993.
6. Cohen, A. I.; Todd, R. D.; Harmon, S.; OMalley, K. L.
Photoreceptors of mouse retinas possess D4 receptors cou-
pled to adenylate cyclase. Proc. Natl. Acad. Sci. USA 89:
1209312097; 1992.
7. Davis, A. A.; Bernstein, S.; Bok, D.; Turner, J.; Nachtigal,
M.; Hunt, R. C. A human retinal pigment epithelial cell
line that retains epithelial characteristics after prolonged
culture. Invest. Ophthalmol. Vis. Sci. 36:955964; 1995.
8. del Cerro, M.; Gash, D. M.; Rao, G. N.; Notter, M. F.;
Wiegand, S. J.; Gupta, M. Intraocular retinal transplants.
Invest. Ophthalmol. Vis. Sci. 26:11821185; 1985.
9. DiLoreto, D. A., Jr.; del Cerro, C.; Lazar, E. S.; Cox, C.;
del Cerro, M. Storage of human fetal retina in optisol prior
to subretinal transplantation. Cell Transplant. 5(5):553
561; 1996.
10. Duke, P. J.; Daane, E. L.; Montufar-Solis, D. Studies of
chondrogenesis in rotating systems. J. Cell. Biochem. 51:
274282; 1993.
DUTT ET AL.
11. Dutt, K.; Scott, M.; Wang, M.; Semple, B.; Sharma, G. P.;
Srinivasan, A. Establishment of a human retinal cell line
by transfection of SV40 T antigen gene with potential to
undergo neuronal differentiation. DNA Cell Biol. 13:909
921; 1994.
12. Dutt, K.; Ezeonu, I.; Scott, M.; Semple, E.; Srinivasan, A.
Proto-oncogene expression in cAMP and TPA-mediated
neuronal differentiation in a human retinal cell line KGL-
DMSM. Curr. Eye Res. 15:477485; 1996.
13. Dutt, K.; Harris-Hooker, S.; Sanford, G.; Hunt, R.; Lane,
D.; Ellerson, D. Three-dimensional growth of retinal and
endothelial cell co-cultures in the horizontally rotating bi-
oreactor. Invest. Ophthalmol. Vis. Sci. 40:S614; 1999.
14. Ehinger, B.; Bergstrom, A.; Seiler, M.; Armant, R. B.;
Zuker, C. L.; Gustavii, B.; Adolph, A. R. Ultrastructure of
human retinal cell transplants with long survival times in
rats. Exp. Eye Res. 53:447460; 1991.
15. Ezeonu, I.; Derrickson, B.; Dutt, K. Cell fate decisions in
a human retinal precursor cell line: Basic fibroblast
growth factor and transforming growth factor-alpha-medi-
ated differentiation. DNA Cell Biol. 19:527537; 2000.
16. Freed, L. E.; Vunjak-Novakovic, G.; Langer, R. Cultiva-
tion of cellpolymer cartilage implants in bioreactors. J.
Cell. Biochem. 51:257264; 1993.
17. Ghosh, F.; Bruun, A.; Ehinger, B. Grafthost connec-
tions in long-term full-thickness embryonic rabbit retinal
transplants. Invest. Ophthalmol. Vis. Sci. 40:126132;
1999.
18. Goodwin, T. J.; Prewett, T. L.; Wolf, D. A.; Spaulding,
G. F. Reduced shear stress: A major component in the
ability of mammalian tissues to form three-dimensional
assemblies in simulated microgravity. J. Cell. Biochem.
51:301311; 1993.
19. Goodwin, T. J.; Schroeder, W. L.; Wolf, D. A.; Moyer,
M. P. Rotating-wall vessel coculture of small intestine as
a prelude to tissue modeling: Aspects of simulated micro-
gravity application. Proc. Soc. Exp. Biol. Med. 202:181
192; 1993.
20. Gouras, P.; Du, J.; Kjeldbye, H.; Yamamoto, S.; Zack,
D. J. Long-term photoreceptor transplants in dystrophic
and normal mouse retina. Invest. Ophthalmol. Vis. Sci.
35:31453153; 1994.
21. Gouras, P.; Lopez, R.; Du, J.; Gelanze, M.; Kwun, R.;
Brittis, M.; Kjeldbye, H. Transplantation of retinal cells.
Neuroophthalmology 10:165176; 1990.
22. Heckenlively, J. R.; Bouchman, J.; Friedman, J. Diagnosis
and classification of retinitis pigmentosa. In: Hecken-
lively, J. R., ed. Retinitis pigmentosa. Philadelphia: J. B.
Lippincott; 1988:21.
23. Humayun, M. S.; Prince, M.; de Juan, E., Jr.; Barron, Y.;
Moskowitz, M.; Klock, I. B.; Milam, A. H. Morphometric
analysis of the extramacular retina from postmortem eyes
with retinitis pigmentosa. Invest. Ophthalmol. Vis. Sci.
40:143148; 1999.
24. Humayun, M. S.; De Juan, Jr., E.; Greenberg, R. J.; Dag-
nelie, G.; Rader, R. S.; Katona, S. Electrical stimulation
of the retina in patients with photoreceptor loss. Invest.
Ophthalmol. Vis. Sci. 38:S39; 1997.
25. Ingram, M.; Techy, G. B.; Saroufeem, R.; Yazan, O.; Nar-
ayan, K. S.; Goodwin, T. J.; Spaulding, G. F. Three-
dimensional growth patterns of various human tumor cell
lines in simulated microgravity of a NASA bioreactor. In
Vitro Cell. Dev. Biol. 33:459466; 1997.
26. Jiang, L. Q.; del Cerro, M. Reciprocal retinal transplanta-
tion: A tool for the study of an inherited retinal degenera-
tion. Exp. Neurol. 115:325334; 1992.
RETINAL TISSUE ENGINEERING
27. Kagami, H.; Sakai, H.; Uryu, K.; Kaneda, T.; Sakanaka,
M. Development of tyrosine hydroxylase-like immunore-
active structures in the chick retina: Three-dimensional
analysis. J. Comp. Neurol. 308:356370; 1991.
28. Kaplan, H. J.; Tezel, T. H.; Berger, A. S.; Wolf, M. L.;
Del Priore, L. V. Human photoreceptor transplantation in
retinitis pigmentosa. A safety study. Arch. Ophthalmol.
115:11681172; 1997.
29. Kaysen, J. H.; Campbell, W. C.; Majewski, R. R.; Goda,
F. O.; Navar, G. L.; Lewis, F. C.; Goodwin, T. J.; Ham-
mond, T .G. Select de novo gene and protein expression
during renal epithelial cell culture in rotating wall vessels
is shear stress dependent. J. Membr. Biol. 168:7789;
1999.
30. Konobu, T.; Sessler, F.; Luo, L. Y.; Lehmann, J. The hNT
human neuronal cell line survives and migrates into rat
retina. Cell Transplant. 7:549558; 1998.
31. Langer, R.; Vancanti, J. P. Tissue engineering. Science
260:920926; 1993.
32. LaVail, M. M.; Mullen, R. J. Role of the pigment epithelium
in inherited retinal degeneration analyzed with experimental
mouse chimeras. Exp. Eye Res. 23:227245; 1976.
33. Lelkes, P. I.; Galvan, D. L.; Hayman, G. T.; Goodwin,
T. J.; Chatman, D. Y.; Cherian, S.; Garcia, R. M.; Uns-
worth B. R. Simulated microgravity conditions enhance
differentiation of cultured PC12 cells towards the neuro-
endocrine phenotype. In Vitro Cell. Dev. Biol. 34:316
325; 1998.
34. Lewis, M. L.; Moriarity, D. M.; Campbell, P. S. Use of
microgravity bioreactors for development of an in vitro rat
salivary gland cell culture model. J. Cell. Biochem. 51:
265273; 1993.
35. Litchfield, T. M.; Whiteley, S. J. O.; Lund, R. D. Trans-
plantation of retinal pigment epithelial, photoreceptor and
other cells as treatment for retinal degeneration. Exp. Eye
Res. 64:655666; 1997.
36. Low, H. P.; Savarese, T. M.; Schwartz, W.J. Neural pre-
cursor cells form rudimentary tissue-like structures in a
rotating-wall vessel bioreactor. In Vitro Cell. Dev. Biol.
37:141147; 2001.
37. Massey, S. C.; Mills, S. L. A calbindin-immunoreactive
cone bipolar cell type in the rabbit retina. J. Comp. Neu-
rol. 366:1533; 1996.
38. Molday, R. S.; MacKenzie, D. Monoclonal antibodies to
rhodopsin: Characterization, cross-reactivity, and applica-
tion as structural probes. Biochemistry 22:653660; 1983.
39. Newsome, D. A. Retinitis pigmentosa, Ushers syndrome,
and other pigmentary retinopathies. In: Newsome, D. A.,
ed. Retinal dystrophies and degenerations. New York: Ra-
ven Press; 1988:161194.
40. Nir, I.; Harrison, J. M.; Haque R.; Low, M. J.; Grandy,
D. K.; Rubinstein, M.; Iuvone, P. M. Dysfunctional light-
evoked regulation of cAMP in photorecetors and abnor-
mal retinal adaptation in mice lacking dopamine D4 recep-
tors. J. Neurosci. 15:20632073; 2002.
731
41. Nishida, K.; Yamanishi, K.; Yamada, K.; Dota, A.; Kawa-
saki, S.; Quantock, A. J.; Kinoshita, S. Epithelial hyper-
proliferation and transglutaminase I gene expression in
Stevens-Johnson syndrome conjunctiva. Am. J. Pathol.
154:331336; 1999.
42. Pellis, N. R.; Goodwin, T. J.; Risin, D.; McIntyre, B. W.;
Pizzini, R. P.; Cooper, D.; Baker, T. L.; Spaulding, G. F.
Changes in gravity inhibit lymphocyte locomotion
through type I collagen. In Vitro Cell. Dev. Biol. 33:398
405; 1997.
43. Prewett, T. L.; Goodwin, T. J.; Spaulding, G. G. Three-
dimensional modeling of T-24 human bladder carcinoma
cell line: A new simulated microgravity vessel. J. Tissue
Culture Methods 15; 1993.
44. Schuschereba, S. T.; Silverman, M. S. Retinal cell and
photoreceptor transplantation between adult New Zealand
red rabbit retinas. Exp. Neurol. 115:9599; 1992.
45. Schwarz, R. P.; Goodwin, T. J.; Wolf, D. A. Cell culture
for three-dimensional modeling in rotation-wall vessels:
An application of simulated microgravity. J. Tissue Cul-
ture Methods 14:5157; 1992.
46. Seigel, G. M.; Takahashi, M.; Adamus, G.; McDaniel, T.
Intraocular transplantation of ElA-immortalized retinal
precursor cells. Cell Transplant. 7:559566; 1998.
47. Seiler, M. J.; Aramant, R. B.; Bergstrom, A. Co-transplan-
tation of embryonic retina and retinal pigment epithelial
cells to rabbit retina. Curr. Eye Res. 14:199207; 1995.
48. Seiler, M.; Aramant, R. B.; Ehinger, B.; Adolph, A. R.
Transplantation of embryonic retina to adult retina in rab-
bits. Exp. Eye Res. 51:225228; 1990.
49. Soon-Shiong, P.; Heintz, R. E.; Merideth, N.; Yao, Q. X.;
Yao, Z.; Zheng T.; Murphy, M.; Moloney, M. K.;
Schmeni, M.; Harris, M.; Mendez, R.; Mendez, R.; San-
ford, P. A. Insulin Independence in a type I diabetic pa-
tient after encapsulated islet transplantation. Lancet 343:
950951; 1994.
50. Soon-Shiong, P.; Feldman, E.; Nelson, R.; Heintz, R.;
Yao, Q.; Yao, Z.; Zheng, T.; Merideth, N.; Skjak-Brack,
G.; Espevik, T.; Smidsrod, O.; Sanford. P. Long-term re-
versal of diabetes by the injection of immunoprotected is-
lets. Proc. Natl. Acad. Sci. USA 90:58435847; 1993.
51. Tabbal, S.; Fahn, S.; Frucht, S. Fetal tissue transplantation
[correction of transplantation] in Parkinsons disease.
Curr. Opin. Neurol. 11:341349; 1998.
52. Turner, J. E.; Blair, J. R. Newborn rat retinal cells trans-
planted into a retinal lesion site in adult host eyes. Brain
Res. 391:91104; 1986.
53. Wood, J. P.; McCord, R. J.; Osborne, N. N. Retinal pro-
tein kinase C. Neurochem. Int. 30:119136; 1997.
54. Zhau, H. E.; Goodwin, T. J.; Chang, S. M.; Baker, T. L.;
Chung, L. W. Establishment of a three-dimensional hu-
man prostate organoid coculture under microgravity-simu-
lated conditions: Evaluation of androgen-induced growth
and PSA expression. In Vitro Cell. Dev. Biol. 33:375
380; 1997.