Download Analysis of immune deviation elicited by antigens injected

Analysis of Immune Deviation Elicited by Antigens
Injected into the Subretinal Space
Hartmut Wenkel andj. Wayne Streilein
determine whether the subretinal space supports the induction of deviant immune
responses to cell-associated and soluble antigens and to elucidate factors influencing the immunologic properties of the subretinal space.
PURPOSE. TO
P815 mastocytoma cells were used as cell-associated antigens and were inoculated into
the anterior chamber (AC), the vitreous cavity (VC), the subretinal space, and subconjunctivally in
H2-compatible, but minor H-incompatible, BALB/c mice. Ovalbumin, as a soluble antigen, was
similarly injected into eyes, after which recipient animals were immunized with ovalbumin and
complete Freund's adjuvant. Delayed-type hypersensitivity (DTH) was assessed by ear challenge. To
alter the conditions in the subretinal space, the outer blood-retinal barrier was disrupted by
compromising retinal pigment epithelial (RPE) cells with a systemic injection of sodium iodate.
Immune privilege in the AC was abolished by mild corneal cauterization.
METHODS.
Antigen-specific DTH did not develop in mice in which alloantigenic tumor cells or
ovalbumin was injected into the AC, the VC, or the subretinal space, and the mice's spleens
contained lymphocytes capable of suppressing DTH when adoptively transferred into naive mice.
When RPE cells were compromised with sodium iodate, tumor cells or ovalbumin injected into the
subretinal space or the VC did not induce immune deviation, although the AC of these eyes still
promoted AC-associated immune deviation. By contrast, when immune privilege in the AC was
abolished by corneal cauterization, neither tumor cells nor ovalbumin injected into the subretinal
space or the VC of eyes elicited immune deviation.
RESULTS.
The subretinal space supports immune deviation for histoincompatible tumor cells
and soluble protein antigens by actively suppressing antigen-specific DTH. Acute loss of immune
privilege in the subretinal space and the VC does not cause loss of privilege in the AC, but abolition
of immune privilege in the AC eliminates the capacity of the subretinal space and the VC to support
immune deviation to antigens injected locally. (Invest Ophthalmol Vis Set. 1998;39:1823-1834)
CONCLUSIONS.
A
deviant immune response to cell-associated alloantigens and soluble antigens injected into the anterior
chamber (AC) has been described in detail in the literature.1"7 The main characteristics of anterior chamber-associated immune deviation (ACAID) are impaired delayed-type
hypersensitivity (DTH), T cells that suppress DTH, and the
inability to produce complement-fixing antibodies.1 Antigenspecific immune deviation can be adoptively transferred into
naive animals by an intravenous injection of splenocytes derived from mice that have had antigen injected into the AC.
Splenic regulatory T cells that are responsible for immune
deviation bear Thy 1.2 and CD8. A minority of these cells also
express CD4 surface marker.3
So far, only a small number of stvidies have been reported
that involve immune deviation elicited by antigens injected
into the posterior part of the eye. 89 Jiang and Streilein showed
From the Schepens Eye Research Institute, Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts.
Supported in part by Grant WE 1462/3-1 from Deutsche Forschungsgemeinschaft and by Grant EY 09595 from the United States
Public Health Service, Washington, DC.
Submitted for publication January 22, 1998; revised May 5, 1998;
accepted June 4, 1998.
Proprietary interest category: N.
Reprint requests: J. Wayne Streilein, The Schepens Eye Research
Institute, 20 Staniford Street, Boston, MA 02114.
Investigative Ophthalmology & Visual Science, September 1998, Vol. 39, No. 10
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
in 1991 that immune privilege is extended to allogeneic tumor
cells injected into the vitreous cavity (VC).10 Because an important goal of our research is to transplant retinal tissues into
the subretinal space, our interests were focused on immune
reactions after inoculation of antigens into the subretinal
space. Experiments reporting long-term survival of grafts and
suppression of DTH after transplantation of neonatal ocular
tissue into the subretinal space 1112 suggest that the subretinal
space elicits immune deviation, resembling ACAID. However,
the transplanted tissues in those experiments were eye derived
and therefore probably represent immune-privileged tissues,
which may considerably alter the observed immune responses.
Anatomically, the subretinal space is a potential space
located between the retinal pigment epithelial (RPE) layer and
the photoreceptor outer segments. This space becomes visible
in retinal detachment when the photoreceptors lose contact
with the RPE cell layer. Because there is free fluid exchange
between the photoreceptor layer and the subretinal space, the
functional borders of the subretinal space are defined by the
RPE layer and by the outer limiting membrane of the retina.
The RPE lining the outer border of the subretinal space forms
the outer blood-retinal barrier and allows exchange of molecules and cells only through the RPE layer itself.13 The inner
barrier is formed by close intercellular contacts (zonula adherens) among Miiller cell processes with a pore size of approximately 300 nm to 360 nm, permitting the exchange of smaller
1823
1824
Wenkel and Streilein
FA
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
10VS, September 1998, Vol. 39, No. 10
IOVS, September 1998, Vol. 39, No. 10
molecules for the nutrition of the outer part of the neurosensory retina. Normally, the space is filled with the so-called
interphotoreceptor matrix, a complex mixture of soluble and
nonsoluble substances, including hyaluronic acid, chondroitin
sulfate, interphotoreceptor retinol-binding protein, and transforming growth factor-/3. However, the exact composition is
still unknown.13 Importantly, this space is supposed to be
devoid of bone marrow- derived cells. Although the main function of the interphotoreceptor matrix is the support and maintenance of photoreceptor segments, the space may also have
immunologic functions, because the subretinal space has a
strategic location between the retina and the choroid. Many
immunologic and infectious diseases of the posterior part of
the eye take place at this site.
Our experimental goal was to determine whether a deviant immune response, similar to ACAID,2'6 could be induced
after injection of cell-associated or soluble antigens into the
subretinal space. In addition, we sought to identify parameters
essential for the capacity of the subretinal space to induce
antigen-specific immune deviation. Sodium iodate was used to
disrupt the outer blood-retinal barrier by compromising RPE
cells,14 and mild corneal cauterization was used to abolish
ACAID and immune privilege in the anterior chamber.'5
MATERIALS AND METHODS
Animals
Adult female BALB/c mice aged 6 to 8 weeks were obtained
from the animal facilities at the Schepens Eye Research Institute or from Jackson Laboratories (Bar Harbor, ME). Mice were
maintained in a common room of a vivarium. Inoculations,
injections, clinical examinations, and enucleations were conducted in animals under anesthesia induced by intraperitoneal
injection of 0.075 mg/kg ketamine (Ketalar; Parke Davis, Paramus, NJ), and 0.006 mg/kg xylazine (Rompun; Phoenix Pharmaceutical, St. Joseph, MO). All experimental procedures conformed to the ARVO Statement for the Use of Animals in
Ophthalmic and Vision Research. Five mice comprised each
group, and experiments were repeated at least twice.
Intraocular Injections
For AC, VC, and subconjunctival injections, a 0.3-mm penetrating wound was made with a 30-gauge needle in the peripheral
portion of the cornea, 1 mm posterior to the limbus, or in the
fornix of the conjunctiva, respectively. For injections in the
subretinal space, the temporal conjunctiva was opened parallel
to the limbus, and the eye was rotated to expose the posterior
part of the sclera. A 0.3-mm tangential sclerotomy was made
with a 30-gauge needle, and a retinal bleb was created with 0.5
Subretinal Space-Associated Immune Deviation 1825
/xl Hank's balanced salt solution. Tumor cells or ovalbumin
were injected into the wound.
Tumor Cells and Injections
P815 mastocytoma cells (DBA/2 origin) were grown as previously described.4 For experiments, 2 X 105 P815 cells in a
volume of 2 /xl were slowly injected into the ocular wound
using a glass micropipette connected to a 10-/LXI syringe. Fifty
micrograms ovalbumin in Hank's balanced salt solution was
used as a soluble antigen and was injected in a volume of 2 /xl
into the eyes of experimental animals.
Sodium Iodate Injections and Horseradish
Peroxidase Method
Sodium iodate, a substance especially toxic to RPE cells' 4
was used to disrupt the outer blood-retinal barrier. Experimental animals received a single injection of 40 mg/kg
sodium iodate (0.08 M at pH 7.4) in a tail vein 30 or 2 days
before antigen injection. For evaluation of outer bloodretinal barrier breakdown through extravasation of horseradish peroxidase (HRP), control animals (no tumor cell
injection) received an intravenous (IV) injection of 200
mg/kg HRP type II 15 minutes before they were killed at
different time points (6, 12, and 24 hours and 2, 3, 7, 14, and
21 days). Eyes were enucleated and fixed for 30 minutes in
ice-cold 10% buffered formalin. Tissue was then processed
for frozen sections and 5-/xm sections were cut. Sections
were reacted with 3,3-diaminobenzidine and hydrogen peroxidase and evaluated by means of light microscopy."'
Cauterization of the Cornea
To alter immune privilege in die AC, the left cornea of experimental animals was cauterized as previously described.15
Briefly, the midperiphery of the cornea was cauterized at five
points by means of electrocautery, until a white discoloration
was seen. Seven days after cauterization, antigens were inserted into the AC, VC, or subretinal space.
Clinical Evaluation of Tumor Growth of P815
Cells
The anterior segment of eyes after tumor cell inoculation
was examined clinically with a slit lamp (Topcon; Tokyo
Optical, Tokyo, Japan), and the posterior segment was examined with a dissecting microscope through a coverslip.
All eyes were examined at 3, 7, and 14 days after injection of
tumor cells.
Histopathologic Evaluation of Tumor Growth of
P815 Cells
For histologic evaluation, tumor-bearing control eyes of
each group were enucleated at different time points (3, 7,
FIGURE 1. Photomicrographs of sections of the retina of mice enucleated 15 minutes after systemic injection of horseradish
peroxidase to evaluate the integrity of the blood-retinal barrier. After sections were reacted with 3,3-diaminobenzidine and
hydrogen peroxidase, they were evaluated for extravascular detection of horseradish peroxidase. (A) Retina of mouse eye
enucleated 12 hours after systemic injection of sodium iodate. Extravascular dark granular staining of the subretinal space and the
photoreceptor segments show leakage of horseradish peroxidase through the outer blood-retinal barrier. (B) Retina of mouse eye
enucleated 48 hours after systemic injection of sodium iodate. Positive peroxidase staining is visible extravascularly throughout the
entire retina. (C) Retina of mouse eye enucleated 21 days after systemic injection of sodium iodate. Positive staining of endogenous
peroxidase is seen only in intravascular erythrocytes of retinal and choroidal vessels. No extravascular leakage of horseradish
peroxidase into the retina was detected.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
1826
Wenkel and Streilein
IOVS, September 1998, Vol. 39, No. 10
and 14 days), fixed with 10% buffered formalin, processed
for routine histopathology and stained with hematoxylin-
eosin.
Assay for DTH
Delayed-type hypersensitivity for cell-associated antigens was
evaluated 14 days after inoculation of P815 tumor cells. For
soluble antigen, 7 days after intraocular inoculation of ovalbumin, animals were immunized subcutaneously with 100 /Ltg
ovalbumin and complete Freund's adjuvant. Ear-swelling analysis was performed 7 days later. Delayed-type hypersensitivity
was measured based on ear swelling, as previously described.12
Briefly, 5 X 105 P815 tumor cells (irradiated with 2000 R) or
200 jttg ovalbumin in 10 /u.1 Hanks' balanced salt solution were
injected into the left ear pinnae of the mice. The right ear
served as untreated control. Both ear pinnae were measured
immediately before injection and 24 hours later with an engineer's micrometer (Mitutoyo, Tokyo, Japan). The measurements were performed in triplicate. Results were expressed as
specific ear swelling = (24-hour measurement — 0-hour measurement in the experimental ear — 24 hour measurement —
0-hour measurement in the negative control ear) X 10~3 mm.
A two-tailed Student's £-test was used, with significance assumed at P < 0.05. For graphic illustrations, the mean of
negative control measurements was subtracted from single
values in each experiment.
Adoptive Transfer Assay for Suppression of DTH
For adoptive transfer studies for cell-associated antigens,
splenocytes obtained from ocular (P815 cell) tumor-bearing
mice were injected IV (6 X 107 spleen cells) into naive
BALB/c mice, as previously described.10 In negative and
positive control animals, the same amount of spleen cells
of naive BALB/c mice was infused IV. To control for antigen-specificity of adoptive transfer, a control group of
mice received an injection of a nonrelevant antigen (ovalbumin) into the subretinal space; spleen cells from these
mice were later transferred into naive mice. Within 2 hours
all experimental mice, the specificity control animals, and
the positive control animals received a subconjunctival injection of 2 X 105 P815 cells. Negative control mice received no subconjunctival tumor cell injection. Delayedtype hypersensitivity was assayed 10 days after transfer of
spleen cells. For soluble antigens, naive mice received an
IV injection of splenocytes (6 X 107 cells) derived from
animals that had received 50 jag ovalbumin intraocularly 7
days before adoptive transfer. Twenty-four hours later, all
experimental mice and the positive control mice were immunized with 100 /xg ovalbumin and CFA subcutaneously.
Negative control mice were not immunized. Ten days later
DTH was assayed.
60 .
Scon
AC
SR
Naive
2. Delayed-type hypersensitivity measurement after injection of P815 cells into different ocular sites. Ear-swelling analysis
24 hours after ear challenge with P815 cells in animals with inoculation of P815 cells into the anterior chamber (AC) or the
subretinal space (SR) 14 days earlier. As a positive control, animals had received P815 cells subconjunctivally (Scon). Naive animals
that underwent ear challenge only served as negative control. *Significantly lower mean response than in positive control animals
(P < 0.05).
FIGURE
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Subretinal Space-Associated Immune Deviation
IOVS, September 1998, Vol. 39, No. 10
1827
80 -
Naive
AC
SR
Control
OVA
FIGURE 3.
Adoptive transfer of impaired delayed-type hypersensitivity response induced by P815 cells in allogeneic BALB/c mice.
Naive animals had received spleen cells intravenously from mice bearing P815 tumor cells in the anterior chamber (AC) or the
subretinal space (SR) for 12 days. Control groups received spleen cells from naive animals. Specificity, control group (OVA) mice
received spleen cells harvested from animals after ovalbumin injection into the subretinal space. One hour later, all recipients of
spleen cells received 2 X 105 P815 cells into the subconjunctival space. Negative control animals (Control) received no
subconjunctival tumor cell injection. Ten days later, delayed-type hypersensitivity was evaluated by ear-swelling analysis. *Significantly lower mean response than positive control (Naive; P < 0.05).
RESULTS
Alteration of Outer Blood-Retinal Barrier
After IV injection of sodium iodate, breakdown of the outer
blood-retinal barrier was detected by extravascular HRP
within 12 hours (Fig. 1A). During the first 24 hours after
inoculation, leakage of HRP was confined to the subretinal
space and the photoreceptor region. After 2 days, peroxidase
staining was detected throughout the entire retina (Fig. IB).
There was no leakage of HRP out of retinal vessels or through
vessels of the ciliary body or iris. Slowly, the barrier re-formed,
so that by day 14, intravenous injected HRP only weakly
stained the retina. By day 21, no extravascular HRP was detected in the retina. (Fig. 1C). Sodium iodate caused no gross
morphologic changes in the RPE, examined by light microscopy, and no macrophages or lymphocytes were observed in
the subretinal space. In addition, no choroiditis was detected
after sodium iodate injection.
Immune Deviation to Cell-Associated Antigens in
the Subretinal Space
Mice bearing P815 cells in the AC, the subretinal or the subconjunctival space were evaluated for DTH by ear challenge
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
assay 14 days after tumor cell inoculation. Results shown in
Figure 2 indicate that after inoculation of tumor cells into AC or
subretinal space, respectively, no DTH was detectable. By
contrast, recipients of subconjunctivally injected tumor cells
displayed intense DTH. Moreover, spleen cells adoptively
transferred from animals bearing intraocular tumors into naive
recipients suppressed DTH when the latter mice received a
subconjunctival immunization with P815 cells (Fig. 3). In contrast, control animals that had received spleen cells from naive
donors or from donors that received a nonrelevant antigen
(ovalbumin) intraocularly showed intense DTH after subconjunctival immunization with P815 cells.
Growth Patterns of Tumor Cells Injected
Intracamerally
P815 cells that have been injected into AC and VC develop into
a progressively growing intraocular neoplasm.410 We compared the growth potential of P815 cells injected into the
subretinal space with cells similarly injected into the AC. We
observed that tumor cells inoculated into both sites grew
progressively in eyes of all mice through day 14, at which time
the eyes were enucleated (data not shown). As visualized by
indirect ophthalmoscopy, early tumor growth in the subretinal
1828
Wenkel and Streilein
Positive
IOVS, September 1998, Vol. 39, No. 10
Naive
AC
VC
SR
4. Delayed-type hypersensitivity measurements after inoculation of ovalbumin (OVA) in different ocular sites. Ear-swelling
analysis was performed 24 hours after ear challenge with ovalbumin in animals that had had subcutaneous immunization with
ovalbumin and complete Freund's adjuvant 7 days earlier. One week before immunization, ovalbumin was injected into the
anterior chamber (AC), the vitreous cavity (VC), or the subretinal space (SR). Negative control animals (Naive) were not
immunized. Positive control mice were immunized only. *Significantly lower mean response than in positive control animals (P <
0.05).
FIGURE
space (days 3-7) proceeded mainly horizontally (underneath
the retina). Later, the cells penetrated into the choroid and into
the vitreous cavity. By contrast, tumor cells injected into the
subconjunctival space were rejected promptly. After an initial
growth that was detectable by day 7, the subconjunctival
masses quickly disappeared.
Immune Deviation to Soluble Antigens in the
Subretinal Space
After injection of ovalbumin in the AC, the VC, the subretinal space, or the subconjunctival space, animals were immunized with ovalbumin and CFA at day 7. Evaluation by ear
challenge analysis was performed 10 days later. No DTH was
detectable after intraocular injection (AC, VC, or subretinal
space) of ovalbumin (Fig. 4). By contrast, mice immunized
with ovalbumin-CFA without prior intraocular injection of
ovalbumin displayed intense DTH. Moreover, spleen cells
adoptively transferred from animals after intraocular injection of ovalbumin suppressed DTH when the latter mice
received a subcutaneous immunization with ovalbumin and
CFA (Fig. 5). In contrast, control animals that had received
spleen cells from naive mice showed intense DTH after
subcutaneous immunization.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Effect of Sodium Iodate on Immune Privilege and
Deviation in the Subretinal Space
As shown earlier, IV injection of sodium iodate transiently
compromised the outer blood-retinal barrier in mice by
affecting RPE cell function. In the next experiments, we
evaluated whether antigens injected into the subretinal
space of these mice could support induction of immune
deviation. Ovalbumin or P815 cells were injected into the
subretinal space of eyes of mice treated two days earlier
with sodium iodate. Recipients of P815 cells underwent ear
challenge 10 days later with irradiated P815 cells. Recipients
of ovalbumin were immunized subsequently with ovalbumin
plus CFA, and ear challenge was conducted as before. Neither ovalbumin nor P815 cells injected into eyes compromised by sodium iodate acquired immune deviation (Figs. 6,
7). Instead, these mice showed intense, antigen-specific
DTH. Moreover, tumor cells injected into the subretinal
space after sodium iodate treatment failed to establish progressively growing tumor masses, and were eventually rejected (data not shown). When ovalbumin was injected into
eyes of mice treated with sodium iodate 30 days earlier, the
mice mounted strong DTH responses (Fig. 7); that is, they
displayed no immune deviation.
Subretinal Space—Associated Immune Deviation
IOVS, September 1998, Vol. 39, No. 10
Naive
AC
SR
1829
Control
FIGURE 5-
Adoptive transfer of impaired delayed-type hypersensitivity response to ovalbumin. Naive animals received spleen cells
intravenously from mice that had undergone anterior chamber (AC) or subretinal (SR) injection of ovalbumin 7 days earlier. Control
groups received spleen cells from naive animals. One day later, all recipients of spleen cells were immunized with ovalbumin and
complete Freund's adjuvant subcutaneously. Negative control animals (Control) were not immunized. Ten days later, delayed-type
hypersensitivity was evaluated by ear-swelling analysis. 'Significantly lower mean response than in positive control (Naive) animals
(/> < 0.05).
Effect of Sodium Iodate on Immune Privilege in
the Anterior Chamber and the Vitreous Cavity
The previous experiments indicate that immune privilege in
the subretinal space was lost for at least 30 days after sodium
iodate treatment of mice. Even though we found no histologic evidence of compromise of the other vascular barriers
in the eye (retinal vessels, vessels of iris and ciliary body),
we wondered whether loss of immune privilege in the
subretinal space would interfere with immune privilege in
the anterior segment or the VC. Accordingly, groups of mice
that had been pretreated with sodium iodate received
inoculations of P815 cells or ovalbumin into the AC or
the VC (only tested for ovalbumin), respectively. Subsequently, ovalbumin recipients were immunized with ovalbumin in CFA. Ear challenge of mice after injection of
ovalbumin into the VC resulted in vigorous DTH (Fig. 7).
When ears of mice after AC injection were challenged
with P815 cells or ovalbumin, they displayed impaired DTH
(Figs. 6, 7, respectively). Moreover, P815 cells injected into
the AC of eyes of sodium iodate-treated mice grew progressively, as they do in eyes of normal mice.16 These results
indicate that transient loss of immune privilege in the subretinal space correlated with loss of immune privilege in the
VC, but did not simultaneously disrupt immune privilege in
the AC.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Effect of Loss of Immune Privilege in the Anterior
Chamber on Immune Privilege and Deviation in
the Subretinal Space and the Vitreous Cavity
Immune privilege in the AC can be abrogated by a variety of
experimental manipulations, including light cauterization of
the central corneal surface. Injection of antigens into the AC of
eyes with corneas cauterized 1 week previously does not
induce ACAID1517 (Figs. 8, 9). To determine whether loss of
privilege in the AC has a disruptive effect on immune privilege
in the subretinal space or the VC, corneal surfaces of eyes of
BALB/c mice received light cautery. One week later, ovalbumin or P815 cells were injected into the subretinal space or the
VC (only performed for ovalbumin). Subsequently, the ovalbumin recipients were immunized with ovalbumin and CFA.
When ears of these mice were challenged with ovalbumin and
P815 cells, intense DTH was observed (Fig. 8, 9). Moreover,
growth of tumor cells in the subretinal space of eyes with
cauterized corneas was severely curtailed, compared with the
unrestricted growth of tumors in the subretinal space of normal mice.
DISCUSSION
Our experiments show that the subretinal space is an immunologically privileged site for cell-associated and soluble anti-
1830
Wenkel and Streilein
IOVS, September 1998, Vol. 39, No. 10
80 -
0-
Scon
SR
SR/SI
AC
AC/SI
Naive
FIGURE 6. Delayed-type hypersensitivity measurement after intraocular injection of P815 cells with or without pretreatment with
sodium iodate. Delayed-type hypersensitivity was assessed by ear-swelling analysis 24 hours after ear challenge with P815 cells.
Before ear challenge, animals were injected with P815 cells into subretinal (SR) space or anterior chamber (AC). Groups of animals
had received systemic administration of sodium iodate (AC/SI; SR/SI) 2 days earlier. As a positive control, animals received P815
cells subconjunctivally (Scon). Naive animals with ear challenge only served as a negative control. *Significantly lower mean
response than in positive control animals (JP < 0.05).
gens. After injection into the subretinal space, allogeneic tumor cells grow progressively, and DTH is suppressed,
compared with the effect of tumor cells injected into the
subconjunctival space. The characteristics discovered so far
resemble the findings in the AC: prolonged survival of allogeneic cells, suppressed DTH expression (immune deviation),
and acquisition of splenocytes that can adoptively suppress
DTH in naive animals.' "7 Although we did not analyze further
the regulatory cells responsible for immune deviation in the
subretinal space, we assume that the relevant splenic cells bear
Thy 1.2 and CD8, as reported for the AC.3 However, there are
certainly immune differences between the AC and the subretinal space. For example, we know that aqueous humor drains
directly into the blood, which explains why the spleen serves
as the focus of ACAID-inducing signals from the AC. We do not
know how the signals that generate immune deviation after
antigen injection into the subretinal space or the VC escape
from the eye. It is possible that the antigen-bearing signals
responsible for immune privilege from the subretinal space
escape through the choroid and therefore travel through
lymph vessels to cervical lymph nodes. Egan et al.18 reported
T-cell expansion in the draining lymph nodes after intravitreal
injection of antigen. In addition, they found greater T-cell
expansion within the periarteriolar lymphoid splenic sheaths
of VC injected mice, compared with that in mice immunized
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
through a conventional site. Whether these findings indicate
differences between immune deviation in the AC and in the
posterior part of the eye must await further investigations.
Another issue is the antigen-specific signal within the
blood that confers immune deviation in naive mice. For the AC,
this signal proved to be F4/80+ monocytes that seem to arise
from the stroma of iris, ciliary body, and trabecular meshwork.l9 The potential equivalents in the posterior part of the
eye may be resident dendritic cells in the choroid, or microglia
present within the neuroretina. Our current experiments explore these possibilities.
Our present findings support the results of other investigators who found prolonged survival of neuroretinal or RPE
grafts in the subretinal space. 1112 We are mindful of the likelihood that neuronal retinal tissue and RPE (used in previous
experiments) may themselves represent immune-privileged tissues, and conclusions about immune privilege in the subretinal
space based on experiments using immune-privileged tissues
as grafts must therefore be circumspect. Sertoli cells are a case
in point. These cells normally reside in an immune-privileged
site, the testis. However, Sertoli cells show prolonged survival
when implanted in non-immune-privileged sites, such as beneath the kidney capsule.20 In the present experiments we
used P815 cells, which are non-immune-privileged cells that
IOVS, September 1998, Vol. 39, No. 10
Subretinal Space-Associated Immune Deviation
1831
60
(/)
0)
C
o
20
10
Positive
AC/SI
SR/SI(30d)
Naive
FIGURE 7.
Delayed-type hypersensitivity measurement after intraocular injection of ovalbumin with or without pretreatment with
sodium iodate. Ear-swelling analysis was performed 24 hours after ear challenge with ovalbumin in animals immunized subcutaneously with ovalbumin and complete Freund's adjuvant 7 days earlier. One week before immunization, ovalbumin was injected
into the anterior chamber (AC), the vitreous cavity (VC), or the subretinal space (SR). Groups of animals had received systemic
administration of sodium iodate (SR/SI; VC/SI; AC/SO 2 days earlier. One group of animals had received a single systemic injection
of sodium iodate 30 days before subretinal inoculation of ovalbumin (SR/SI[30 d]). Negative control animals (Naive) were not
immunized. Positive control mice were immunized only. 'Significantly lower mean response than in positive control animals (P <
0.05).
are readily rejected at conventional sites.16 Nonetheless, P815
cells showed prolonged survival in the subretinal space.
Based on our experimental results and the findings of
others, we conclude that the subretinal space is an immuneprivileged site. However, the site seems unable to confer indefinite survival on foreign tissue grafts placed within it,
whether the experiments are conducted in animals or in human eyes.">2' Especially because retinal transplantation will
eventually be considered for diseased human eyes, the privileged nature of this site may be open to question. Therefore, it
seemed important to determine the extent to which immune
privilege and immune deviation are lost when the properties of
the subretinal space are altered experimentally. As a case in
point, in the human disease age-related macular degeneration,
the rejection rates of allogeneic human RPE grafts are reported
to be lower in nonexudative than in neovascular age-related
macular degeneration.21 To evaluate the immune consequences of altering immune privilege in the subretinal space,
two techniques were used in our study: The RPE layer was
compromised with a prior IV injection of sodium iodate, which
led to a breakdown of the outer blood-retinal barrier, whereas
immune privilege in the AC was broken by corneal thermal
injury.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
To test the influence of a disrupted outer blood-retinal
barrier, we injected sodium iodate IV before inoculation of
antigen. Sodium iodate is known to be almost exclusively toxic
to RPE cells, resulting in a rapid loss of barrier function.1'* In
general, effects on the systemic immune system of experimental animals treated with sodium are thought to be trivial.22 For
example, experimental allergic encephalomyelitis develops unimpeded in animals pretreated with sodium iodate.23 However,
sodium iodate may have an effect on S-antigen-induced experimental autoimmune uveitis. High doses of sodium iodate decreased the rate of experimental autoimmune uveitis developing in a susceptible rat strain,23 whereas low-dose sodium
iodate in a primarily unsusceptible rat strain promoted the
development of experimental autoimmune uveitis.2'1 These
dose-dependent effects may be caused by different degrees of
RPE damage and blood-retinal barrier breakdown, which
could alter the course of experimental autoimmune uveitis.
Whether the impairment of the RPE cells by sodium iodate has
a direct effect on immune mechanisms in the eye can only be
speculated.
In our animals the outer blood-retinal barrier was disrupted for approximately 14 days after sodium iodate injection,
after which repair began, and the barrier was eventually re-
1832
Wenkel and Streilein
IOVS, September 1998, Vol. 39, No. 10
70 •
Scon
SR
SR/Cau
AC
AC/Cau
Naive
8. Delayed-type hypersensitivity measurement after intraocular injection of P815 cells with or without previous corneal
cauterization. Delayed-type hypersensitivity was assessed by ear-swelling analysis 24 hours after ear challenge with P815 cells.
Before ear challenge, animals were injected with P815 cells into the subretinal (SR) space or anterior chamber (AC). Corresponding
groups of animals had received corneal cauterization 7 days before anterior chamber (AC/Cau) or subretinal (SR/Cau) injection of
P815 tumor cells. As a positive control, animals received P815 cells subconjunctivally (Scon). Naive animals that had ear challenge
only served as a negative control. *Significantly higher mean response than corresponding group of animals without previous
corneal cauterization (P < 0.05).
FIGURE
established. We found that injection of sodium iodate a few
days before antigen inoculation into the subretinal space abolished immune privilege and abrogated immune deviation to
cell-associated and soluble antigens in the subretinal space and
the VC. Experiments performed with ovalbumin in mice pretreated with sodium iodate 30 days earlier showed persistently
impaired immune deviation, even though the blood-retinal
barrier is believed to re-form within 3 weeks. This suggests that
loss of immune privilege after injection of sodium iodate is not
simply caused by breakdown of the blood-retinal barrier. One
possible explanation for sustained loss of immune privilege
after sodium iodate may be the entry of plasma proteins into
the subretinal space. Serum proteinases could lower the concentration of transforming growth factor-/3 (or other immunosuppressive factors) in the subretinal space, thwarting the
induction of immune deviation. Transforming growth factor-j3
is a major factor known to be important in immune deviation
in the AC,25 and it is also present in die vitreous body.26
Re-establishment of this subretinal microenvironment after
breakdown of the outer blood-retinal barrier could require a
prolonged period.
Another possible explanation for sustained loss of immune
privilege in the subretinal space after sodium iodate injection
could be prolonged impairment of RPE cell function. Retinal
pigment epithelial cells are probably involved in creating the
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
appropriate microenvironment in the subretinal space through
secretion of relevant cytokines and growth factors. However,
until now very little has been known about the factors that
promote immune deviation in the posterior part of the eye or
their cellular origin.
Therefore, a functionally intact RPE layer with an undisturbed outer blood-retinal barrier seems to be critical in the
maintenance of the local microenvironment required for induction of immune deviation in the posterior part of the eye.
Breakdown of this barrier may consequently cause the failure
of allografts placed in the subretinal space.
It is interesting that short-term abrogation of the outer
blood-retinal barrier had little or no effect on immune privilege and deviation in the AC. Although tumor cells did not
establish successful tumors in subretinal spaces with disrupted
outer blood-retinal barriers, they grew well in the AC. In
addition, immune deviation to minor histoincompatibility antigens and to soluble ovalbumin was readily induced through
the AC of eyes of sodium iodate-treated mice. It has been
reported that ACAID cannot be induced by injecting exogenous antigens into die AC of eyes of transgenic mice with
selective expression of interferon-7 in photoreceptor cells.27
This report clearly indicates that alterations in the cytokine
milieu of the subretinal space can interfere with ACAID in the
AC. In these transgenic mice, expression of interferon-y is
Subretinal Space-Associated Immune Deviation
IOVS, September 1998, Vol. 39, No. 10
1833
70 -
Positive
SR
SR/Cau
VC
VC/Cau
AC
AC/Cau
Naive
9- Delayed-type hypersensitivity measurement after intraocular injection of ovalbumin with or without previous corneal
cauterization. Ear-swelling analysis was performed 24 hours after ear challenge with ovalbumin in animals immunized subcutaneously with ovalbumin and complete Freund's adjuvant 7 days earlier. One week before immunization, ovalbumin was injected
into the anterior chamber (AC), the vitreous cavity (VC), or the subretinal space (SR). Corresponding groups of animals had
received corneal cauterization 7 days before anterior chamber (AC/Cau), intravitreal (VC/Cau), or subretinal (AR/Cau) injection of
ovalbumin. Negative control animals (Naive) were not immunized. Positive control mice were immunized only. 'Significantly
higher mean response than in corresponding group of animals without previous corneal cauterization (P < 0.05).
FIGURE
constitutive and constant, once the rhodopsin gene is upregulated. Therefore, we suspect that retention of immune privilege in the AC in the face of loss of the outer blood-retinal
barrier may relate to the transient nature of the disruption
caused in our experiment. Perhaps, a sustained loss of this
barrier would lead eventually to loss of privilege in the vitreous
cavity and the AC, but we have not yet achieved a stable
disruption of the outer blood-retinal barrier to test this possibility.
The potential for communication among the subretinal
space, the VC, and the AC is relevant to ocular immune
privilege, because our experiments indicate that abrogation
of immune privilege in the AC (by corneal cauterization) can
lead to loss of privilege and immune deviation in the subretinal space and the VC. Our experiments gave no hint of
why loss of privilege in the AC should spread to the subretinal space, but we consider two possible explanations worthy of inquiry. First, after corneal cauterization, cells of the
iris and ciliary body in that eye lose their capacity to secrete
immunosuppressive factors.15 It is possible that modulators
constitutively released from these cells reach the posterior
compartment, even the subretinal space, and contribute to
the normal immunosuppressive microenvironment. Thus, a
resultant deficit of immunosuppressive factors from iris and
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
ciliary body cells could rob the subretinal space of its immunomodulatory properties and thereby its immune privilege. Second, cautery of the cornea may alter indirectly the
properties of the outflow tract through which antigenic
information must pass to induce immune deviation. If the
only route of escape for antigens from the subretinal space
and the VC that induces immune deviation is anteriorly
through the trabecular meshwork, and if this pathway is
altered after corneal cauterization, antigenic signals from the
posterior part of the eye probably will not promote immune
deviation. Experiments to test these possibilities are currently under way.
References
1. Streilein JW. Immune regulation and the eye: a dangerous compromise. FASEBJ. 1987; 1:199-208.
2. Streilein JW, Niederkorn JY, ShadduckJA. Systemic immune unresponsiveness induced in adult mice by anterior chamber presentation of minor histocompatibility antigens. / Exp Meet. 1980; 152:
1121-1125.
3. Streilein JW, Niederkorn JY. Characterization of the suppressor
cell(s) responsible for anterior chamber-associated immune deviation (ACA1D) induced in BALB/c mice by P815 cells. JT Immunol.
1985;134:1381-1387.
1834
Wenkel and Streilein
4. Ksander B, Streilein JW. Immune privilege to MHC-disparate tumor
grafts in the anterior chamber of the eye. Transplantation. 1989;
47:661-667.
5. Streilein JW, Okamoto S, Hani Y, Kosiewicz M, Ksander B. Bloodborne signals that induce anterior chamber-associated immune
deviation after intracameral injection of antigen. Invest Ophthalmol Vis Sci. 1997;38:2245-2254.
6. Mizuno K, Clark AF, Streilein JW. Anterior chamber-associated
immune deviation induced by soluble antigens. Invest Ophthalmol
Vis Sci. 1989;30:l 112-1119.
7. Wilbanks GA, Streilein JW. Characterization of suppressor cells in
anterior chamber-associated immune deviation (ACAID) induced
by soluble antigen. Evidence of two functionally and phenotypically distinct T-suppressor cell populations. Immunology. 1990;
71:383-389.
8. Tsumura K, Watanabe M, Sakata H. Cytotoxic T lymphocyte activity after intravitreal inoculation of herpes simplex vims type I in
mice. Ophthalmic Res. 1995;27:32-36.
9. Atherton SS, Pesicka GA, Streilein JW. Retinitis and deviant immune responses following intravitreal inoculation of HSV-1. Invest
Ophthalmol Vis Sci. 1987;28:859-866.
10. Jiang LQ, Streilein JW. Immune privilege extended to allogeneic
tumor cells in the vitreous cavity. Invest Ophthalmol Vis Sci.
1991 ;32:224 -228.
11. Jiang LQ, Jorquera M, Streilein JW, Ishioka M. Unconventional
rejection of neural retinal allografts implanted into the immunologically privileged site of the eye. Transplantation. 1995;59:
1201-1207.
12. Jiang LQ, Jorquera M, Streilein JW. Immunologic consequences of
intraocular implantation of retinal pigment epithelial allografts.
Exp Eye Res. 1994;58:719-728.
13. Hewitt AT, Adler R. The retinal pigment epithelium and interphotoreceptor matrix: structure and specialized functions. In: Ogden
TE, ed. Retina. Basic Science and Inherited Retinal Disease. Vol
1. St Louis: Mosby; 1994:58-71.
14. Anstadt B, Blair NP, Rusin M, Cunha-Vaz JG, Tso MOM. Alteration
of the blood-retinal barrier by sodium iodate: kinetic vitreous
fluorophotometry and horseradish peroxidase tracer studies. Exp
Eye Res. 1982;35:653-662.
15. Streilein JW, Bradley D, Sano Y, Sonoda Y. Immunosuppressive
properties of tissue obtained from eyes with experimentally manipulated corneas. Invest Ophthalmol Vis Sci. 1996;37:4l3-424.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
IOVS, September 1998, Vol. 39, No. 10
16. Niederkorn J, Streilein JW, Shadduck JA. Deviant immune responses to allogeneic tumors injected intracamerally and subcutaneously in mice. Invest Ophthalmol Vis Sci. 1980;20:355-363.
17. Williamson JSP, DiMarco S, Streilein JW. Immunobiology of Langerhans cells on the ocular surface, I: Langerhans cells within the
central cornea interfere with induction of anterior chamber associated immune deviation. Invest Ophthalmol Vis Sci. 1987;28:
1527-1532.
18. Egan RM, Yorkey C, Black R, Loh WK, Stevens JL, Woodard JG.
Peptide-specific T cell clonal expansion in vivo following immunization in the eye, an immune-privileged site. / Immunol. 1996;
157:2262-2271.
19- Wilbanks GA, Mammolenti M, Streilein JW. Studies on the induction of anterior chamber-associated immune deviation (ACAID), II:
Eye-derived cells participate in generating blood-borne signals that
induce ACAID./Immunol. 1991;l46:3018-3024.
20. Bellgrau D, Gold D, Selawry H, Moore J, Franzusoff A, Duke RC. A
role for CD95 ligand in preventing graft rejection. Nature. 1995;
377:630-632.
21. Algevere PV, Bergllin 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:l49-158.
22. Webster SH, Rice ME, Highman B, Oettingen von WF. The toxicology of potassium and sodium iodate: acute toxicity in mice.
J Pharmacol Exp Ther. 1957;120:171-178.
23. Konda BR, Pararajasegaram G, Wu G-S, Stanforth D, Rao NA. Role
of retinal pigment epithelium in the development of experimental
autoimmune uveitis. Invest Ophthalmol Vis Sci. 1994;35:4O-47.
24. Tanaka T. Participation of sodium iodate in the induction of experimental autoimmune uveoretinitis (EAU). Ada Soc Ophthalmol
Jpn. 1991;95:1077-1084.
25. Cousins SW, McCabe MM, Danielpour D, Streilein JW. Identification of transforming growth factor-beta as an immunosuppressive
factor in aqueous humor. Invest Ophthalmol Vis Sci. 1991 ;32:
2201-2211.
26. Yoshitoshi T, Shichi H. Immunosuppressive factors in porcine
vitreous body. Curr Eye Res. 1991;12:ll4l-ll49.
27. Geiger K, Sarvetnick N. Local production of IFN-gamma abrogates
the intraocular immune privilege in transgenic mice and prevents
the induction of ACAID. / Immunol. 1994; 153:5239-5246.