Retinal Pigment Epithelial Cells From Dystrophic Rats
Form Normal Tight Junctions In Vitro
Chih-Wei Chang,* Dennis M. Defoe,^ and Ruth B. Caldwell*
Purpose. In the genetically defective Royal College of Surgeons (RCS) rat model for retinal
degeneration, a breakdown occurs in the retinal pigment epithelial (RPE) cell tight junctions
just as the photoreceptors begin to degenerate. These experiments sought to determine the
impact of the RPE genetic defect on this alteration in the RPE cell tight junctions.
Methods. Retinal pigment epithelial cell cultures prepared from RCS and control rats were
treated with hormonally defined medium (HDM), base medium conditioned by RCS or
control retinas, or unconditioned base medium. The tight junctions formed by these cultures
were assayed functionally by measuring transepithelial electrical resistance and permeability.
Junction structure was evaluated by immunolocalization of the tight junction protein zonula
occludens 1 and of the junction-associated actin microfilaments.
Results. Retinal pigment epithelial cultures from dystrophic rats formed structurally and functionally normal tight junctions when maintained in hormonally defined medium. The junctions remained stable when the medium bathing the apical surface was switched to base
medium preconditioned by normal retinas. In contrast, cultures treated with medium preconditioned by degenerating dystrophic retinas or with unconditioned medium exhibited a breakdown in their tight junctions.
Conclusions. Retinal pigment epithelial cells isolated from dystrophic RCS rats can form tight
junctions normally in vitro. Normal, but not dystrophic, retinas release factors that support
RPE tight junctions. Therefore, the junctional abnormality seen in dystrophic rat RPE cells
in vivo is probably caused by the loss of trophic factors normally provided by the healthy
neural retina rather than by a direct effect of the genetic defect on the tight junctions. Invest
Ophthalmol Vis Sci. 1997;38:188-195.
From the * Department of Cellular Biology and Anatomy and the Department of
Ophthalmology, The Vascular Biology Center, The Medical College of Georgia,
Augusta, and the -[Department of Anatomy and Cell Biology, East Tennessee State
University, College of Medicine, Johnson City, Tennessee.
Supported in part by National Institutes of Health/National Eye Institute grants
EY04618 (RBC) and EY07036 (DMD), by unrestricted grants from Research to
Prevent Blindness (RBC, DMD), and by the Royal Order of Eagles (RBC).
The experiments reported in this study were completed in partial fulfillment of the
requirements for the doctoral degree for C-WC.
Received for publcation March 12, 1996; revised August 19, 1996; accepted August
20, 1996.
Proprietary interest category: N.
Reprint requests: Ruth B. Caldwell, Department of Cellular Biology and Anatomy,
The Medical College of Geoigia, Augusta, CA 30912-3400.
permeable to the low-molecular-weight extracellular
tracer, lanthanum nitrate. 9 The tight junction breakdown progresses as the degeneration continues; at 2
months of age and later, permeability increases are
widespread. 910 The tight junctions also deteriorate
structurally, as indicated by freeze-fracture analyses
showing increased fragmentation and decreased numbers of tight junction strands.11'12 These tight junction
alterations have been presumed to occur secondarily
to the photoreceptor degeneration, 9 but their specific
relationship to the RPE cell genetic defect and the
dystrophic process has not been determined. To elucidate this relationship, we have assayed the function
and stability of tight junctions formed by dystrophic
rat RPE cells using cell culture methodologies. Experiments were designed to test two alternative hypotheses. The first is that structural breakdown and increased permeability of RPE cell tight junctions in
RCS rats are a direct result of the genetic defect—
188
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
Copyright © Association for Research in Vision and Ophthalmology
J. he Royal College of Surgeons (RCS) rat is a wellestablished model for inherited retinal degeneration
secondary to dysfunction of the retinal pigment epithelium (RPE).1'2 In this model, the photoreceptors
begin to degenerate during the third postnatal week
because of a genetic defect in RPE cell phagocytosis
of the photoreceptor outer segments. 3 " 8 At the same
time, the RPE cell tight junctions become abnormally
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
Retinal Pigment Epithelial Tight Junctions of RCS Rat
i.e., the RPE tight junctions are intrinsically defective.
The second is that the RPE tight junctions are normal
but become altered because of changes that occur in
the surrounding environment secondary to the dystrophic process.
In this study, we demonstrate that RPE cells from
RCS rats form tight junctions in vitro that are structurally and functionally indistinguishable from those
formed by normal RPE cells. We also show that medium conditioned by the healthy neural retinas of normal rats contains factors that maintain RPE cell tight
junctions, whereas unconditioned medium or medium conditioned by degenerating dystrophic retinas
fails to support RPE tight junctions. These observations support the hypothesis that the RPE cell tight
junction alterations in RCS rats occur secondarily to
the photoreceptor degeneration and are not caused
by direct effects of the mutation on the tightjunctions.
189
tures, conditioned medium was added to the apical
compartment of the Transwell, and hormonally defined medium was added to the basal compartment.
The volume of medium in the apical (200 //I) and
basal compartments (800 /A) was adjusted to avoid
hydrostatic pressure differences. Medium in the apical
compartment was changed every other day; that in the
basal compartment was changed every day.
Lactate Dehydrogenase Activity
Lactate dehydrogenase (LDH) was assayed using a kinetic
ultraviolet light method. Samples (100 fi\) were mixed
with 2.5 ml of phosphate buffer containing 16.2 mM
pyruvate and 0.194 mM nicotinamide-adenine dinucleotide at 25°C. Generation of NAD+ was read at 340 nm
after 30 seconds and was monitored at 1-minute intervals
for 3 minutes after the initial reading. One unit of LDH
activity is defined as the amount of enzyme that catalyzes
the formation of 1 mM/1 of NAD + /minute.
MATERIALS AND METHODS
Transepithelial Electrical Resistance
Cell Culture
An Epithelial Voltohmmeter (World Precision Instruments, Sarasota, FL) with STXII electrodes was used to
measure RPE cell transepithelial electrical resistance
(TER). Values for TER for the filters alone, without cells,
were measured as background and subtracted from values
for RPE cultures; TER was expressed as ohm • cm2.
Retinal pigment epithelial cell cultures were prepared
from 6- to 8-day-old rat retinas as described previously.13 The rats were treated in accordance with the
ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Retinas were obtained from
breeding colonies of pink-eyed and black-eyed dystrophic rats, congenic controls for RCS and Long Evans
rats. Retinal pigment epithelial cells were dissociated
into single cells and small clusters and were plated
into 24-well Transwells (0.4 fiM pore size; COSTAR,
Cambridge, MA) that had been coated with matrigel
(1:1 ratio in NCTC-135 medium; Collaborative Research, Bedford, MA). They were seeded at the density
of 2 X 105 cells/cm2 and were maintained in hormonally denned medium (HDM) prepared using NCTC135 base medium supplemented with growth factors,
proteinase inhibitors, and antibiotics as described.13
In some experiments, cells were plated initially in plastic dishes, amplified by culture for 3 to 5 days in
DMEM with 10% fetal bovine serum, and harvested
for the preparation of Transwells as described.
Conditioned Medium Preparation
Conditioned medium was prepared using the method
described by Rizzolo and Li.14 Briefly, neural retinas
from 1-month-old dystrophic or control rats were
placed in NCTC base medium that contained 50 mM
vitamin C, 1 mM glutathione, 1% bovine serum albumin, 20 mM Hepes, and 2% penicillin-steptomycin
(two retinas per milliliter). After incubation for 6
hours at 37°C under 95% O2 and 5% CO2, retinas were
discarded and the conditioned media were collected,
filtered, and stored at — 70°C. For treatment of cul-
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
Permeability Assay
Sodium fluorescein (376 Da) and horseradish peroxidase (40 kDa) were used as permeability tracers.
Tracers (25 /xg SF/ml and 50 fxg HRP/ml) were
constituted in HDM, added to the apical compartments alone or in combination, and incubated with
the cultures at 37°C with 5% CO2 for consecutive
intervals of 2 to 60 minutes. The amount of tracer
that crossed the monolayer to reach the basal compartments was determined for each time point. Sodium fluorescein was read directly in a CytoFluor II
Microplate Fluorescence Reader (Millipore, Marlborough, MA). Horseradish peroxidase concentration was determined by the reaction of 20-/xl samples
with 150 /il of freshly made substrate [o-phenylenediamine 400 /ig/ml in 0.05 M citric acid, 0.1 M phosphate, and 0.012% H 2 O 2 , pH 5]. The peroxidase
reaction was terminated by the addition of 50 fj,\
0.25 M H2SO4, and the product was read in a 7520
Microplate reader (Cambridge Technology, Watertown, MA). The tracer concentration for each sample was calculated by comparison with standard
curves prepared using tracer samples from time 0.
Immunocytochemistry
Retinal pigment epithelial cultures were fixed with 4%
formaldehyde in phosphate-buffered saline (PBS) for
10 minutes at room temperature. For ZO-1 labeling,
190
Investigative Ophthalmology 8c Visual Science, January 1997, Vol. 38, No. 1
60
I
10
20
30
Incubation Time (min)
Legend:
—°— Normal RPE
—-••— RCS RPE
20
40
Incubation Time (mln)
l. Retinal pigment epithelial cells from dystrophic retinas form tight junctions normally in vitro. Retinal pigment epithelial cultures were prepared from normal (open circles)
or dystrophic (filled circles) retinas, maintained in hormonally defined medium, and assayed
for transepithelial electrical resistance (A) and permeability to sodium fluorescein (B) and
horseradish peroxidase (C). Data are plotted as mean ± standard deviation.
FIGURE
cell cultures were incubated with antibody (monoclonal rat anti-ZO-1, diluted 1:100; Chemicon, Temecula, CA) overnight at 4°C. After they were washed
with PBS, cultures were incubated for 2 hours at 37°C
with secondary antibody (1:64, fluorescein isothiocyanate goat anti-rat; Sigma, St. Louis, MO). For actin
labeling, cultures were incubated with rhodaminephalloidin (1:50 in PBS; Molecular Probes, Eugene,
OR) for 1 hour at room temperature. All labeled cultures were washed with PBS five times, and the filters
with cells were separated from the Transwells. The
cultures were examined and photographed using an
Olympus BH-2 fluorescence microscope (Olympus,
Woodberry, NY).
RESULTS
Tight Junctions of Retinal Pigment Epithelial
Cells From Dystrophic Retinas Form Normally
In Vitro
To test whether the increased tight junction permeability observed in dystrophic rat RPE cells in vivo could be
a direct effect of the genetic RPE cell defect, cultures of
normal and dystrophic RPE cells were analyzed in parallel. Tight junction formation was examined by measuring TER and by permeability assay. One day after cell
plating, TER was measured, and this continued for 6
days. After the last TER reading, permeability was assayed. Data showed that TER levels comparable to those
in normal RPE cells developed in RPE cells from the
mutant retina (Fig. 1A). Permeability assay showed that
mutant RPE cells, like normal RPE cells, could restrict
the passage of sodium fluorescein (Fig. IB) or horseradish peroxidase (Fig. 1C). We used first-passage cultures
for this analysis so the cultures could be prepared simul-
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
taneously using the same batches of media, tracers, and
other reagents. This was done because our initial studies
showed that cultures prepared from freshly isolated RPE
cells exhibited slight differences in maximum TER levels
reached because of slight variations in cell attachment
and survival after trypsinization. In other experiments
using different batches of fresh RPE isolates, absolute
TER levels, ranging between 200 and 300 ohm-cm 2 ,
were higher than those shown in Figure 1. In fresh RPE
isolates from RCS retinas, TER levels were sometimes a
little higher, sometimes a little lower, than in controls.
No significant differences were detected between the
two strains.
Dystrophic RPE cells also were normal morphologically as indicated by the localization of junction-associated
proteins. Phalloidin labeling showed actin concentrated
apicolaterally in a circumferential microfilament bundle
that highlighted the junctional complex zone (Fig. 2).
No differences between the dystrophic and control rat
strains were evident in terms of their junctional morphology, cell shape, or cell size, nor did immunolabeling for
ZO-1 show any differences between dystrophic and control RPE cells (data not shown).
Medium Conditioned by Normal Retinas
Supports the Maintenance of Retinal Pigment
Epithelial Tight Junctions In Vitro
Next we tested whether the RPE tight junctions in the
dystrophic retina become altered because of changes in
the cellular milieu that occur secondarily to retinal degeneration. We reasoned that the tight junctions might be
affected adversely, either by negative factors released by
the degenerating retina or by the loss of positive factors
normally released by the healthy retina. To test these
possibilities, media preconditioned by normal or dystro-
Retinal Pigment Epithelial Tight Junctions of RCS Rat
FIGURE 2. Cellular morphology of dystrophic retinal pigment
epithelial (RPE) cells appears normal in vitro. Cultures prepared using RPE cells isolated from normal (A) or dystrophic (B) retinas were maintained in hormonally defined
medium for 6 days and were stained with phalloidin. Actin
microfilaments are distributed peripherally at the region of
the cell-cell junctional complex. Magnification, X275.
phic retinas were used to treat RPE cultures that had
reached stable levels of TER during culture in HDM.
Freshly isolated RPE cells were used for these experiments. Data were normalized to the starting level of TER
developed in HDM to allow comparison of cultures prepared from cells isolated at different times. This was necessary because small litter sizes of the inbred rat strains
limited the number of retinas available for cell isolations
and culture preparation.
The pattern of the RPE cultures' responses to the
different media treatments was highly consistent between
different batches of primary RPE cultures. Results showed
that when HDM in the apical compartments was replaced
by medium preconditioned by normal retinas, the cultures prepared from either normal (Fig. 3A) or dystrophic (Fig. 3B) rats could maintain their TER levels for
up to 5 days (analysis of variance [ANOVA] showed no
significant decline in TER measures between successive
days of treatment). In contrast, when the apical compartment HDM was replaced with medium preconditioned
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
191
by dystrophic retinas, TER declined steadily, dropping
60% during the 5-day test period (Fig. 3C; P < 0.001,
ANOVA). Retinal pigment epithelial cultures maintained
in unconditioned base medium without additional supplements showed similar TER decreases (Fig. 3D; P <
0.001, ANOVA). Results of permeability assays using the
low-molecular-weight tracer sodium fluorescein confirmed die RPE cell tight junction instability shown by
these TER assays. As TER decreased in die cultures
treated widi unconditioned base medium, their permeability increased, as indicated by increased flux of sodium
fluorescein across the monolayer (data not shown).
Immunolabeling analysis of cellular morphology in
the same cultures used in diese experiments showed that
cells were organized as a continuous monolayer of tighdy
packed hexagonal cells as described. ZO-1 was located at
die cell-cell contact zone, oudining die cells' polygonal
shape (Fig. 4). No differences in the morphology or viability of die RPE cells were detected under die different
medium treatments, nor were any obvious differences in
junction morphology detected. Actin labeling also
showed similar morphology in all groups (data not
shown). Apparendy, die functional assays of tightjunction
formation are much more sensitive to tightjunction alterations tiian diese immunolocalization approaches.
The actin-labeled cultures were analyzed using
morphometry to determine whether cell density was
affected by the conditioned medium treatments. Cell
counts in multiple, randomly selected microscope
fields showed that cell density was slightly lower in the
cultures treated with medium conditioned by normal
retinas (mean ± SD = 1.8 ± 0.1 X 105 cells/cm2) than
in the cultures treated with medium conditioned by
the dystrophic retinas (2.4 ± 0.2 X 10r' cells/cm2).
Results of the above experiments with conditioned medium suggest that the healthy retina releases
soluble factors needed to support stable tight junctions. These factors may be altered or missing in the
media conditioned by die degenerate dystrophic retina. Alternatively, dying cells in the degenerating retina may release factors that are damaging to the RPE
cells and their tightjunctions. To evaluate this possibility further, the extent of any cell lysis that might have
occurred during the preparation of conditioned medium was examined by assaying LDH activity. This
analysis showed that LDH released into the conditioned medium during the 6-hour incubation period
at 37°C was significantly lower in the dystrophic retinas
than in the normal control retinas (132 ± 44 U/l
versus 302 ±115 U/l, respectively). Thus, it is unlikely
that lytic factors released by the dystrophic retina during preparation of the conditioned medium are responsible for die dght junction changes observed in
our experiments. Furthermore, when the total
amount of LDH in the retinas before culture was assayed in samples subjected to three cycles of freezing
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
192
I
125
125
100
100
75
75
O
o
£
RCS RPE + NCM
Normal RPE + NCM
50
£
oc
50
oc
111
IS
25
25
1
2
3
1
4
2
3
4
Days of CM Treatment
Days of CM Treatment
125
125
Normal RPE + RCSCM
100
75
£
Days of CM Treatment
50
Normal RPE + UCM
RCS RPE + UCM
Days of UCM Treatment
FIGURE 3. Medium conditioned by normal retinas supports the stability of normal or dystrophic retinal pigment epithelial cell tightjunctions in vitro. Retinal pigment epithelial cultures
from normal (A,C,D) or dysuophic (B,D) rats were allowed to develop stable transepithelial
electrical resistance levels in hormonally defined medium. Then the hormonally defined
medium in the apical compartment was substituted with medium preconditioned by normal
(NCM; A,B) or dystrophic retinas (RCSCM; C) or with unconditioned base medium (UCM;
D), and transepithelial electrical resistance was measured for 5 successive days. Data are
plotted as mean ± standard deviation.
and thawing, the amount in the healthy retina was
DISCUSSION
found to be more than twice that in the dystrophic
retina (2249 U/l versus 1004 U/l). The reduction in
the total amount of LDH activity in the dystrophic
samples, compared with the control samples, probably
results from the fact that cell density is reduced substantially in the 1-month-old dystrophic retina because
many of the photoreceptors have degenerated.2 Because the percentage of LDH activity in the retinaconditioned medium relative to the total retinal LDH
activity was the same in both strains (13%), it can be
presumed that equivalent amounts of cell lysis occurred in both media preparations. Thus, it is unlikely
that lytic factors released by the degenerate retina are
responsible for the tight junction instability shown in
these experiments.
Increased permeability of the blood-retinal barrier is
an early complication in most retinal diseases and is
thought to contribute to the development of proliferative retinopathy. The dystrophic RCS rat provides an
excellent animal model for analysis of this disease progression because its RPE cell tightjunctions become
leakyjust as the photoreceptors begin to degenerate.9
As the RPE cell breakdown progresses, the retinal capillaries also become permeable and proliferate within
the subretinal space and in vitreoretinal membranes.10~1215"21 Numerous pathologic changes that
may contribute to this disease progression22"26 have
been described within RPE and glial cells of the dystrophic rat retina, but the primary stimulus for the RPE
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
Retinal Pigment Epithelial Tight Junctions of RCS Rat
193
are not yet fully mature and in which a genetic defect
in tight junction stability has not yet manifested itself.
Technical difficulties in obtaining sufficient quantities
of healthy, intact RPE cells from older animals precluded testing this premise specifically, but comparison of our TER data with results obtained in previous
studies of mature bovine retinas suggests that the tight
junctions of rat RPE cells are already functionally mature at 6 to 8 days. The TER values we obtained ranged
from 100 to 300 ohm • cm2, whereas TER values for
freshly isolated RPE-choroid preparations from adult
bovine retinas ranged from 110 to 220 ohm • cm2.27
Retinal pigment epithelial cells from 6- to 8-day-old
rat retinas are also functionally mature in terms of
their phagocytic activity. The genetic defect in phagocytosis is clearly evident in culture, even though
phagocytosis does not begin in vivo until the photoreceptors mature during the third postnatal week.2'45
Thus, it seems likely that the genetic defect in RCS
rats does not directly impair the assembly or stability
of the RPE tight junctions.
FIGURE 4. Cellular morphology of retinal pigment epithelial
(RPE) cultures treated with medium preconditioned by normal or dystrophic retinas appears normal in vitro. Normal
RPE cell cultures treated with normal (A) or dystrophic (B)
retina-conditioned medium, as described in Figure 3, were
immunoreacted with an antibody against ZO-1. In both
treatments, ZO-1 was distributed peripherally at the zone of
the cell-cell junctions. Magnification, X275.
cell permeability breakdown has not been investigated. The experiments reported here were designed
to address this issue by testing whether the RPE cell
permeability increase in RCS rats results from the inability of the mutant cells to form stable tight junctions
or whether they occur secondarily to retinal degeneration .
Our analyses of TER, permeability, and junction
morphology in RPE cultures prepared from 6- to 8day-old normal and dystrophic rat retinas showed that
the dystrophic RPE cells can form tight junctions that
are functionally and structurally indistinguishable
from those formed by normal RPE cells. In interpreting these data, it should be kept in mind that the RPE
cell tight junction breakdown first becomes evident in
vivo during the third postnatal week.9 Thus, the results
obtained here might represent an intermediate phase
of development in which the normal RPE junctions
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
After demonstrating that tight junctions of dystrophic RPE cells could form normally in vitro, we tested
whether the tight junction alterations seen in vivo
could be caused by changes in the cellular milieu of
the RPE monolayer that occur secondary to the photoreceptor degeneration. Our data showed that medium
conditioned by normal retinas supported the high
TER levels developed by normal and dystrophic RPE
cultures in HDM. In contrast, neither unconditioned
base medium nor medium preconditioned by dystrophic retinas was able to maintain the stability of the
tight junctions. Moreover, cell density determinations
showed that the decrease in TER in these cultures was
not caused by a loss of RPE cells. In fact, cell density
was slightly higher in the cultures treated with medium conditioned by dystrophic retinas than in those
treated with medium conditioned by normal retinas.
The cause of this cell density increase is unknown.
One possibility is that the medium conditioned by the
dystrophic retinas induced cell proliferation, leading
to increased numbers of smaller sized cells. Another
possibility is that it resulted from greater shrinkage of
the cultures during fixation secondary to weakening
of their junctions. Whatever the reason for the cell
density difference, the beneficial effects of the medium conditioned by the normal retina indicate that
this medium contains trophic factors that support
tight junction stability. This finding is consistent with
previous analyses using RPE cells isolated from the
developing chick retina showing that soluble retinaderived factors promote the development of RPE tight
junctions in vitro.14
The negative effects of the medium preconditioned by dystrophic retinas on RPE tight junction
stability probably reflect the fact that trophic factors
194
Investigative Ophthalmology & Visual Science, January 1997, Vol. 38, No. 1
released by the healthy cells in normal retinas are
decreased or lost as the photoreceptors degenerate in
the dystrophic retina. A possible alternative explanation is that lytic factors that impair tightjunction stability are released by dying cells in the degenerating dystrophic retina. We think we can rule out the latter
explanation based on the results of the LDH assay.
First, the amount of LDH in the medium preconditioned by the dystrophic retinas was substantially lower
than that in the medium conditioned by healthy retinas. This indicates that the amount of cell lysis in the
1-month-old dystrophic retina was even less than that
in the healthy, age-matched control preparation,
which had a beneficial effect on tightjunction stability. Second, the total amount of LDH in the freezethawed dystrophic retinas was only half that found in
the normal retinas, indicating a 50% decrease in living
cells probably because of photoreceptor degeneration
in dystrophic rats, which is extensive at 1 month (for
review, see reference 2). Because the dystrophic retina
at 1 month has fewer total living cells than the normal
retina, it probably contains reduced levels of trophic
factors normally released by healthy cells. This could
explain why the dystrophic retina-conditioned medium failed to support the stability of the tight junctions developed by normal RPE cultures. This idea
has been supported by recent experiments testing the
effects of conditioned medium supplemented with the
hormones and growth factors normally included in
HDM (Chang and Caldwell, unpublished data, 1995).
The data showed similar effects of the normal and
dystrophic retina-conditioned media preparations on
tight junction stability, suggesting that the negative
effect of the dystrophic retina-conditioned medium
reflects the absence of a positive factor rather than
the presence of a negative factor.
The LDH analysis also suggests that cell lysis did
not contribute to the beneficial effects of the medium
preconditioned by normal retinas. Absolute levels of
LDH activity in the conditioned medium were very
low (<1 U/ml), indicating that very little cell lysis
occurred during medium preparation in either rat
strain. Release of cytoplasmic material from retinal
cells damaged during the preparation of the conditioned medium was unlikely to contribute to the positive effects of the normal retina-conditioned medium
on RPE cells. Presumably, the putative trophic factor (s) was released by intact cells in the normal retina.
It should be noted that the gradual decline in
tightjunction stability observed in the RPE cultures
treated with unconditioned medium or medium conditioned by dystrophic retinas was not associated with
any obvious alterations in tightjunction structure, as
indicated by immunolocalization of ZO-1 or actin. Further analysis with more sensitive morphologic methods, such as electron microscopy, will be necessary
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
to determine the structural basis for the alterations
revealed by our functional assays. Biochemical analyses using Western blot or Northern blot procedures
may help to reveal quantitative differences in expression of junctional proteins.
In conclusion, these investigations are the first to
show that genetically defective RPE cells from dystrophic rats have the potential to form tight junctions
normally in vitro and that the stability of these junctions can be supported by soluble factors from normal
retinas, but not by those from dystrophic retinas. Further work is necessary to identify factors from normal
or dystrophic retinas that affect tightjunction stability
and to rule out conclusively the presence of negative
factors in medium conditioned by the dystrophic retina. Retinal cells are known to produce a number of
factors that modulate RPE cell growth, including basic
fibroblast growth factor (bFGF), vascular endothelial
growth factor, transforming growth factor-beta, and
others.28"32 These factors have been found to interact
with each other in modulating the rate of RPE cell
growth, and they also may affect other RPE cell responses, including tightjunction formation and function. Interestingly, a deficiency in bFGF receptors has
been reported in RPE cells of dystrophic rats.33 The
specific effects of this deficiency on RPE cell functions
have not been examined. Analyses using conditioned
medium approaches, together with bFGF and other
growth factors and their neutralizing antibodies, will
help to identify the factors involved in stabilizing or
destabilizing RPE cell tight junctions. Our data show
that RPE cells from normal and dystrophic rats develop comparable tight junctions when maintained in
a denned medium containing 0.1 ng/ml bFGF. We
have not determined whether the presence of bFGF or
other media components is required for tightjunction
formation or stability in vitro. However, preliminary
studies with neutralizing antibodies against bFGF suggest that it may have a modest role in stabilizing tight
junctions of normal RPE cells. Additional experiments
examining the potential interaction of bFGF with
other growth factors should help to clarify the role of
this factor in tight junction formation and to show
whether a bFGF receptor deficiency in dystrophic rats
has any effect on tightjunction stability.
Key Words
blood-retinal barrier, cell adhesion, retinal cell culture, retinal degeneration, retinal pigment epithelium, tightjunction
References
1. Bourne MC, Campbell DA, Tansley K. Hereditary degeneration of the rat retina. Br J Ophthalmol. 1938;
22:613-623.
2. LaVail MM. The retinal pigment epithelium in mice
Retinal Pigment Epithelial Tight Junctions of RCS Rat
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
and rats with inherited retinal dystrophy. In: Zinn KM,
Marmor MF, eds. The Retinal Pigment Epithelium. Cambridge: Harvard University Press; 1979; 357-380.
Mullen RJ, LaVail MM. Inherited retinal dystrophy: A
primary defect in pigment epithelium determined
with experimental rat chimeras. Science. 1976;
192:799-801.
Chaitin MH, Hall MO. Defective ingestion of rod
outer segments by cultured dystrophic rat pigment
epithelial cells. Invest Ophthalmol Vis Sci. 1983;24:812820.
Hall MO, Abrams T. Kinetic studies of rod outer segment binding and ingestion by cultured rat RPE cells.
ExpEyeRes. 1987; 45:907-922.
Tso MOM, Zhang C, Abler AS, et al. Apoptosis leads
to photoreceptor degeneration in inherited retinal
dystrophy of RCS rats. Invest Ophthalmol Vis Sci.
1994; 35:2693-2699.
Herron WL, Riegel BW, Meyers DE, Rubin ML. Retinal dystrophy in the rat: A pigment epithelial disease.
Invest Ophthalmol. 1969;8:595-604.
Bok D, Hall MO. The role of the pigment epithelium
in the etiology of inherited retinal dystrophy in the
rat. / Cell Biol. 1971;49:664-682.
Caldwell RB, McLaughlin BJ. Permeability of retinal
pigment epithelial cell junctions in the dystrophic rat
retina. ExpEyeRes. 1983;36:415-427.
Bok D. Unpublished observation cited in: Zinn KM,
Marmor MF, eds. The Retinal Pigment Epithelium. Cambridge: Harvard University Press; 1979:372.
Caldwell RB, McLaughlin BJ, Boykins LG. Intramembrane changes in retinal pigment epithelial cell junctions of the dystrophic rat retina. Invest Ophthalmol Vis
Sci. 1982; 23:305-318.
Caldwell RB, McLaughlin BJ. A quantitative study of
intramembrane changes during junctional breakdown in the dystrophic rat retinal pigment epithelium. Exp Cell Res. 1984; 150:104-117.
Chang C-W, Roque RS, Defoe DM, Caldwell RB. An
improved technique for isolation and culture of pigment epithelial cells from rat retina. Curr Eye Res.
1991;10:1081-1086.
Rizzolo LJ, Li Z. Diffusible retinal factors stimulate
the barrier properties of junctional complexes in the
retinal pigment epithelium. J Cell Sci. 1993; 106:859867.
Essner E, Pino RM, Griewski RA. Permeability of retinal capillaries in rats with inherited retinal degeneration. Invest Ophthalmol Vis Sci. 1979; 18:859-862.
Essner E, Pino RM, Griewski RA. Breakdown of blood
retinal barrier in RCS rats with inherited retinal degeneration. Lab Invest. 1980;43:418-426.
Weber ML, Mancini MA, Frank RN. Retinovitreal neovascularization in the Royal College of Surgeons Rat.
Curr Eye Res. 1989;8:61-74.
Caldwell RB, Slapnick SM, Roque RS. Retinal pigment
epithelial-associated extracellular matrix changes accompany retinal vascular proliferation and retinovitreal membranes in a new model for proliferative retinopathy: The dystrophic rat. In: LaVail MM, Hollyfield
Downloaded From: http://iovs.arvojournals.org/ on 07/28/2017
195
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
JG, Anderson EH, eds. Inherited and Environmentally
Induced Retinal Degenerations. New York: Alan R Liss;
1989:393-407.
Caldwell RB. Extracellular matrix alterations precede
vascularization of the retinal pigment epithelium in
dystrophic rats. Curr Eye Res. 1989;8:907-921.
Caldwell RB, Roque RS, Solomon SW. Increased vascular density and vitreo-retinal membranes accompany vascularization of the pigment epithelium in the
dystrophic rat retina. Curr Eye Res. 1989;8:923-937.
Fitzgerald MEC, Slapnick SM, Caldwell RB. Alterations in lectin binding accompany increased permeability in die dystrophic rat model for proliferative
retinopathy. In: LaVail MM, Anderson RE, Hollyfield
JG, eds. Inherited and Environmentally Induced Retinal
Degenerations. New York: Alan R Liss; 1989:409-425.
Caldwell RB, McLaughlin BJ. Redistribution of Na-KATPase in the dystrophic rat retinal pigment epithelium. JNeurocytol. 1983; 13:895-910.
Caldwell RB. Filipin and digitonin studies of cell membrane changes during junction breakdown in the dystrophic rat retinal pigment epithelium. Curr Eye Res.
1987;6:515-526.
Roque RS, Caldwell RB. Muller cell changes precede
vascularization of the pigment epithelium in the dystrophic rat retina. Glia. 1990;3:464-475.
Roque RS, Caldwell RB. Pigment epithelial cell
changes precede vascular transformations in the dystrophic rat retina. ExpEyeRes. 1991;53:787-798.
Roque RS, Caldwell RB. Microglial cells invade the
subretinal space as photoreceptors degenerate in
Royal College of Surgeons rats. Invest Ophthalmol Vis
Sci. 1996; 37:196-203.
Edelman JL, Miller SS. Epinephrine stimulates fluid
adsorption across bovine retinal pigment epithelium.
Invest Ophthalmol Vis Sci. 1991;32:3033-3040.
Schweigerer L, Malerstein B, Neufeld G, Gospodarowicz D. Basic fibroblast growth factor is synthesized
in cultured retinal pigment epithelial cells. Biochem
Biophys Res Comm. 1987; 143:934-940.
Mascarelli F, Raulais D, Counis MF, Courtois Y. Characterization of acidic and basic fibroblast growth factors in brain, retina and vitreous of the chick embryo.
Biochem Biophys Res Comm. 1987; 146:478-486.
Sternfeld MD, Robertson JE, Shipley GD, Tsai J, RosenbaumJT. Cultured human retinal pigment epithelial cells express basic fibroblast growth factor and its
receptor. Curr Eye Res. 1989;8:1029-1037.
Leschey KH, Hackett SF, Singer JH, Campochiaro PA.
Growth factor responsiveness of hiunan pigment epithelial cells. Invest Ophthalmol Vis Sci. 1990; 31:839-846.
Guerrin M, Moukadiri H, Chollet P, et al. Vasculotropin/vascular endothelial growth factor is an autocrine growth factor for human retinal pigment epithelial cells cultured in vitro./CellPhysiol. 1995; 164:385394.
Malcaze F, Mascarelli K, Bugar K, Fuhrmann G, Courtois Y, Hicks D. Fibroblast growth factor receptor deficiency in dystrophic retinal pigmented epithelium.
/ Cell Physiol. 1993; 154:631-642.