| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Schepens Eye Research Institute, Department of Ophthalmology, Harvard Medical School, Boston, MA 02114
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The immunologic dimension of the neural retina transplant problem is complicated by the fact that the intraocular sites into which these grafts have been placed (anterior chamber, vitreous cavity, and subretinal space) are themselves immune-privileged sites (reviewed in Refs. 7, 8, 9). In this setting it is almost impossible to determine the extent to which a neural retinal graft is immunogenic (i.e., leads to allosensitization of the recipient) or antigenic (i.e., expresses Ags toward which an immune effector response can be directed). In fact, the neural retina itself may be an immune-privileged tissue, a status that would make it less vulnerable to immune rejection. In this regard, it has recently been reported that neonatal retinal pigment epithelium sheets display inherent immune privilege (10). That is, allografts of neonatal retinal pigment epithelium survive indefinitely when placed beneath the kidney capsule, a non-immune-privileged site.
To study whether the neuronal retina possesses the properties of immune privilege, we have conducted a series of experiments in which retinal grafts were placed beneath the kidney capsule. The fate of the grafts was assessed by histology, and the impact of the graft on the recipient was assessed by measuring the ability of graft recipients to display graft-specific delayed hypersensitivity (DH).3 We report here that allografts of neonatal (but not adult) neuronal retina survived for a prolonged interval beneath the kidney capsule, and that during this interval the recipients failed to display DH. When rejection did occur, it was rapid and temporally associated with the acquisition by the recipient of donor-specific DH. These results indicate that the neonatal neuronal retina (NNR) displays the properties of a partially immune-privileged tissue.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adult male BALB/c, C57BL/6, and C3H/HeN mice, aged 6–8 wk, were purchased from Taconic Farm (Germantown, NY) or from the animal colonies at The Schepens Eye Research Institute (Boston, MA). All experimental procedures concerning animals in this study were performed with approval of The Schepens Eye Research Institute animal care and use committee. Mice were maintained in a 12 h on/12 h off light/dark cycle.
Preparation of neuronal retina
The previously described (11, 12, 13) procedure for preparing NNR tissue for grafting was used with some modification. The eyes were enucleated from neonatal BALB/c and C57BL/6 mice and placed in ice-cold calcium-/magnesium-free HBSS. NNR was dissected from the back of the eye with two pairs of number 5 tweezers and transferred to NeuroBasal (Life Technologies, Gaithersburg, MD). The entire NNR was cut into quarters in preparation for transplantation. For preparation of adult neuronal retina, whole eye globes were enucleated, and the cornea was separated from the posterior segment by cutting along the limbus. The lens was removed, and the neuronal retina was carefully separated from the retinal pigment epithelium. The adult retina was cut into four equal quarters in preparation for transplantation.
Implantation of retinal tissue beneath the kidney capsule
Implantation of the adult or neonatal neuronal retinal tissue beneath the kidney capsule was performed as described previously (10, 14). Recipient mice received general anesthesia with a mixture of 150 mg/kg ketamine (Phoenix Pharmaceutical, St. Joseph, MO) and 6 mg/kg xylazine (Phoenix Pharmaceutical) before surgery. A 1.5-cm long opening was made in the back of the animals parallel to the spinal cord. Under a surgical microscope the peritoneum was opened, and the kidney was extruded. A small pouch was made in the kidney capsule using the number 5 tweezers, and the NNR was transferred into the pouch by a spatula. A small piece of Gelfoam (Pharmacia, Peapack, NJ) soaked with NeuroBasal acted as plug to prevent the NNR from protruding out through the capsule. The kidney was returned to its original position, and the wound was closed using a surgical stapler. In some experiments small segments of adult mouse skin, peeled from the inner surface of the ear pinna, were placed beneath the kidney capsule using a similar method.
Assay for DH
An ear-swelling assay was used to measure DH. At least five animals were used in each group. Irradiated (2000 rad) splenocytes (1 x 106 cells/10 µl) from C57BL/6 donors were injected into the right pinnae of recipient mice. Positive control mice were immunized s.c. with 1 x 107 C57BL/6 splenocytes 1 wk before ear challenge. Negative control mice received only ear pinnae challenge. Both ear pinnae were measured immediately before injection and 24 h later with a low pressure engineer’s micrometer (Mitsutoyo; MTI, Paramus, NJ). Ear swelling was expressed as follows: specific swelling = ((24-h numerical values of right ear - 0 h numerical values of right ear) - (24 h numerical values of left ear - 0 h numerical values of left ear)) x 10-3 mm. Ear-swelling responses at 24 h after ear injection are presented as the group mean ± SEM.
Local adoptive transfer assay of DH
This assay detects within lymphoid cell suspensions regulator cells that inhibit the expression of DH in vivo. Regulator cells were collected from the spleen and mesenteric, cervical, and axillary lymph nodes of mice that received C57BL/6 NNR grafts in the kidney subcapsular space 10 days previously. Lymph node cells from the cervical and axillary lymph nodes were pooled. Stimulator cells (peritoneal exudate cells (PEC)) were harvested from normal C57BL/6 or BALB/c mice that received 2.5 ml of thioglycolate (Sigma, St. Louis, MO) i.p. 3 days earlier. Responder cells were collected from the cervical lymph nodes of BALB/c mice that received 1 x 107 allogeneic C57BL/6 spleen cells injected s.c. 7 days previously. Both regulator cells and stimulator cells were exposed to x-irradiation (2000 rad). Regulator cells were added (1 x 106/10 µl) to cell mixtures containing stimulators (PEC, 1 x 106/10 µl) and responder lymphoid cells (1 x 106/10 µl). Mixtures of responders, stimulators, and regulators were then injected (10 µl/injection) into the ear pinnae of naive BALB/c mice. At lease five animals were used in each group. Ear swelling responses were measured after 24 h as described above.
Histology and immunohistochemistry of NNR grafts
The animals were sacrificed at selected times after grafting as following; adult retina graft, 1 h (n = 5) and 24 h (n = 5); and NNR grafts, 1, 2, and 3 days (n = 5 in each group) and 5, 12 and 20 days (n = 15 in each group). The experiment for the graft survival at 5, 12, and 20 days were repeated three times. The graft-bearing kidneys were fixed in 4% paraformaldehyde for 45 min. Some samples were embedded in methacrylate, and 3-µm sections were cut. Tissue was stained with H&E. Grafts were then evaluated by light microscopy. For some samples, after cryoprotection in 30% sucrose solution, the grafted kidney was embedded in OCT medium (Tissue-Tek, Torrance, CA), and 10-µm cryosections were prepared. A panel of Abs including F4/80 (1/100; Caltag, Burlingame, CA), MHC II (1/100; anti-I-Ab and anti-I-Ad; BD Bioscience, San Jose, CA), anti-Neurofilament-M (1/1000; Chemicon, Temecula, CA), anti-CD3 (BD Bioscience, San Jose, CA), and anti-interphotoreceptor retinoid binding protein (anti-IRBP; gift from Dr. A. Adler) were used. The secondary Abs (goat anti-rabbit, anti-mouse, or goat anti-rat) were labeled with Cy2 (green), or Cy3 (red; Jackson ImmunoResearch, West Grove, PA). Images of immunolabeled sections were obtained with confocal microscopy (TCM D4; Leica, Deerfield, IL), and images of H&E-stained sections were obtained with a light microscope equipped with a Spot camera (Diagnostic, Sterling Heights, MI).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first investigated whether full-thickness neuronal retinal
grafts can survive when implanted beneath the kidney capsule. Tissue
equivalent to one quarter of a whole retina from eyes of adult C57BL/6
(allogeneic) and BALB/c (syngeneic) mice was transplanted beneath the
kidney capsule of BALB/c mice. The graft-bearing kidneys were excised
after 1 and 24 h, and the grafts were examined by histology
(n = 5 at each time point). Adult neuronal retinal
grafts examined 1 h after implantation appeared to be
architecturally intact. Each of the distinct layers of the retina could
be distinguished, and there was little or no evidence of mechanically
induced damage (Fig. 1
A). The
opposite was true for adult grafts examined after 24 h in
residence beneath the murine kidney capsule. Whether syngeneic or
allogeneic, adult retina grafts after 24 h were barely
recognizable as retinal tissue. The laminar architecture was completely
obliterated (Fig. 1B). Since adult grafts disassembled
shortly after implantation beneath the kidney capsule and well before
any immune response could have been invoked, it was decided that adult
grafts were unsuitable to use for our studies.
|
, C and
D). At this stage the cells in the NNRs appeared relatively
undifferentiated. Gradually over the next 2 days (days 2 and 3
postimplantation) the undifferentiated appearance began to change, with
the emergence of structures that resembled rosettes. Rosettes are
circular structures composed of a ring of photoreceptor cells with a
central lumen lined by a refractile structure (external/outer limiting
membrane) into which photoreceptors outer segments project. When
implants were examined on day 5, the rosettes were well formed (Fig. 1, E and F), and the photoreceptors had become
organized as the inner layer of the rosettes, forming a central lumen.
Cells with the morphologic appearance of ganglion cells/inner nuclear
layer cells were observed in the layer outside of the rosettes. The
appearance of syngeneic and allogeneic NNRs remained virtually
identical when observed on day 12 postimplantation (Fig. 2, A and B).
Well-defined rosettes were heavily represented in these grafts, and
there was little, if any, evidence to suggest that these grafts were
infiltrated with peripheral blood-derived leukocytes.
|
C), a disorganized morphology was observed in allogeneic
NNR grafts (Fig. 2D) at this time. No rosettes were
detectable, and very few elements of the original neuronal retina were
identifiable. Instead, the graft was heavily infiltrated with
inflammatory cells. These results indicate that syngeneic and
allogeneic NNR grafts are able to establish themselves in the
heterotopic site beneath the kidney capsule and to proceed through
their genetically determined developmental program. Between 12 and 20
days, however, the allogeneic grafts (but not the syngeneic grafts)
undergo an inflammatory response that leads to their dissolution. Characterization of neural cells and inflammatory cells in NNR implants beneath the kidney capsule
We next turned to immunohistochemistry to determine the extent to
which donor-derived and neural cells appeared in NNR implants beneath
the kidney capsule, and the nature of the infiltrating host-derived
inflammatory cells. Abs to Neurofilament-M and IRBP identify cells of
neural origin and photoreceptor cells, respectively. In NNR allografts,
Ab F4/80 identifies infiltrating macrophages and dendritic cells, and
anti-I-Ab identifies donor-derived microglia,
whereas anti-I-Ad identifies infiltrating
recipient cells with Ag-presenting capability. As displayed in Fig. 3A, Neurofilament-M-staining
cells comprised the entirety of the NNR grafts in 5-day implants. An
oval structure with the central region stained positively for IRBP
(Fig. 3B) revealed the outer segments of photoreceptors in a
fully formed rosette. Although staining with F4/80 revealed the
presence of macrophages/microglia within allogeneic NNRs at 5 days
(Fig. 3C), these cells were not stained with either
anti-I-Ab or
anti-I-Ad (data not shown). In only a few
allografts, recipient-derived I-Ad-bearing cells
were observed at the graft rim and never within rosettes (Fig. 3D). For allografts examined at 12 days postimplantation,
scattered I-Ad cells were observed (Fig. 3E). It is important to point out that no donor-derived
I-A+ cells were ever observed in the center of
rosettes or anywhere else in these allografts, and no
CD3+ T cells were detected in 12-day implants.
These findings indicate that healthy donor neuronal retina tissue
develops within these heterotopic implants, including IRBP-synthesizing
photoreceptor cells, and that small numbers of recipient
I-A+ cells with the potential to carry
immunogenic signals to the recipient immune system are present in
allografts that appear to be healthy before the onset of graft
destruction.
|
The differential survival of syngeneic and allogeneic NNR implants
beneath the kidney capsule suggested that a destructive immune response
directed toward the latter might be responsible for its destruction. To
test the possibility that heterotopically grafted allogeneic NNR grafts
might induce graft-specific delayed hypersensitivity, allogeneic
(C57BL/6) spleen cells were injected into the ear pinnae of mice
bearing syngeneic or allogeneic NNR grafts beneath the kidney capsule
at 12 and 20 days after implantation. For positive controls, grafts of
syngeneic or allogeneic skin were implanted beneath the kidney capsule.
Ear-swelling responses were determined 24 and 48 h later as a
measure of the acquisition of graft-induced DH. In no instance did mice
bearing syngeneic grafts display positive ear-swelling responses (data
not shown). The results of a representative experiment (of three)
involving allogeneic implants are presented in Fig. 4. Whereas positive ear swelling was
displayed by mice bearing allogeneic skin grafts beneath the kidney
capsule on days 12 and 20 postimplantation, ear-swelling responses to
C57BL/6 spleen cells were only positive on day 20 postimplantation in
mice bearing allogeneic NNR grafts beneath the kidney capsule (Fig. 4B). The absence of detectable donor-specific alloimmune DH
response at 12 days, which correlates with healthy NNR allografts
beneath the kidney capsule, may well reflect the existence of immune
deviation. Alternatively, failed acquisition of donor-specific DH after
grafts have been in residence for 12 days could also reflect
immunologic ignorance. Regardless of whether deviation or ignorance is
responsible, the emergence of donor-specific DH at 20 days, which
correlates with inflamed, destroyed NNR allografts, indicates that the
non-responsive state is transient, and immune rejection intervenes to
destroy the graft.
|
Allogeneic NNR grafts placed beneath the kidney capsule failed to
induce donor- or retina-specific DH when assayed on day 12. The reasons
for this failure include 1) the recipient immune system might be
ignorant of the existence of the graft at this heterotopic site, and 2)
the graft has induced a deviant systemic immune response that lacks a
DH effector component. To test the latter possibility, we examined
secondary lymphoid organs of graft-bearing mice for the presence of
DH-regulating lymphoid cells. A local adoptive transfer assay was used
in which lymphoid cells (as responders) from BALB/c mice presensitized
to C57BL/6 alloantigens were added to C57BL/6 PEC (as stimulators). To
this mixture, lymphoid cells harvested from the spleen, mesenteric
lymph nodes, or cervical and axillary lymph nodes obtained from BALB/c
mice bearing C57BL/6 NNRs beneath the kidney capsule were added as
regulators. This tripartite mixture was then injected into the dermis
of the ear pinnae of naive BALB/c mice. Ear-swelling responses were
assessed 24 and 48 h later. Positive controls received mixtures
containing primed T cells, stimulators, and regulators from naive
BALB/c mice. The results of a representative experiment (of three) are
presented in Fig. 5. Whereas spleen cells
from graft-bearing mice had no effect on the local adoptive transfer
reaction, lymphoid cells harvested from the cervical/axillary nodes as
well as from mesenteric nodes significantly suppressed ear-swelling
responses. These findings indicate that lymph nodes of mice bearing
allogeneic NNR grafts beneath the kidney capsule at 12 days contain
lymphoid cells that can suppress the expression of donor-specific DH.
At one level, the presence of donor-specific regulator cells in these
lymph nodes argues against the presence of immunologic ignorance as the
explanation for why donor-specific DH is not detected at 12 days.
However, the mere presence of regulator cells gives no guarantee that
these cells are responsible for graft survival at 12 days. Experiments
to test this possibility are planned.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nonetheless, our results indicate that neonatal neural retina grafts were able to acclimate to the environment beneath the kidney capsule. Moreover, these immature implants showed signs that their genetically determined differentiation program was able to express itself. Neuronal cells staining positively for Neurofilament-M and IRBP-positive cells that formed circular and oval rosettes indicated that NNR tissues implanted beneath the kidney capsule differentiated, respectively, into ganglion cells and photoreceptors, the typical neural parenchymal cells of the retina. This evidence indicates that the microenvironment of the subcapsular sinus of the kidney is no more alien to differentiation of NNR grafts than is the anterior chamber or the subretinal space of the normal eye. Interestingly, the NNR grafts survived as xenografts when transplanted into the tectum of the neonatal rat brain. They were not only differentiated into a different laminar structure of the retina, but they also functioned to receive light and caused pupil reflex (18).
The histologic appearances of syngeneic and allogeneic NNR grafts placed beneath the kidney capsule were almost identical when examined through 12 days postimplantation. The healthy appearance of allogeneic NNR grafts contrasted sharply with the appearance of allogeneic grafts of skin beneath the kidney capsule, and the appearance of the latter was similar to what has been reported previously for allogeneic grafts of islets of Langerhans (19), and heart tissue. Allografts of the latter types are uniformly rejected between 7 and 10 days beneath the kidney capsule. For this reason we regard the acceptance beneath the kidney capsule (a non-immune-privileged site) of allogeneic NNR grafts at 12 days as evidence that NNR tissue is itself immune-privileged. However, over the next 7 days allogeneic, but not syngeneic, NNR grafts became surrounded by an inflammatory infiltrate, and rosettes disappeared, as did cells staining positively for IRBP. The destruction of allogeneic NNR grafts in this circumstance, compared with similar syngeneic NNR grafts, indicates that NNR as a tissue is only partially immune-privileged. As mentioned previously, allogeneic grafts of testis (16, 20), neonatal retinal pigment epithelium (10), and epithelium-deprived corneas (14) survive almost indefinitely beneath the kidney capsule, which is characteristic of tissues that express complete immune privilege. In the case of testis, neonatal retinal pigment epithelium, and epithelium-deprived cornea grafts, constitutive expression of CD95 ligand (CD95L) has been implicated in conferring on these tissues the immune-privileged status (10, 14, 16). While cells within the neuronal retina are known to express CD95L, we do not know whether the partial immune privilege displayed by NNR tissue correlates with CD95L expression.
We found that mice that reject allogeneic NNRs beneath the kidney capsule on day 20 also display donor-specific DH. This result supports our view that NNR allografts are eventually capable of activating recipient T cells specific for donor alloantigens. This is not too surprising, since even allogeneic NNR grafts placed in the anterior chamber of the eye elicit a systemic immune response directed at donor alloantigens (11, 12, 13). However, in the latter case the systemic immune response is deviant, and mice bearing allogeneic NNR grafts in the anterior chamber acquire donor-specific anterior chamber-associated immune deviation. Thus, depending upon the site of engraftment, NNR allografts can elicit two different types of immune response: conventional alloimmunity and deviant alloimmunity. Both types of sensitization require that donor Ags escape from the graft site and activate specific lymphoid cells in regional secondary lymphoid organs. We conclude, therefore, that donor Ag somehow escapes from NNR allografts placed beneath the kidney capsule. Moreover, if the presence of regulatory lymphoid cells in the draining lymph node corresponds to the induction of systemic immune deviation, then it may be true that the early survival of allogeneic NNR grafts beneath the kidney capsule is secured by suppressed alloimmunity, albeit transiently.
It is worth commenting further on the possible mechanisms by which allogeneic NNR grafts placed beneath the kidney capsule might sensitize their recipients. Ma and Streilein (13) reported recently that microglia within allogeneic NNR grafts placed in the anterior chamber or subretinal space quickly become activated, expressing class II MHC molecules as well as the carbohydrate determinant recognized by Griffonia simplicifolia lectin. These activated microglia predominantly occupy the central regions of rosettes, where they phagocytize rod outer segments in a fashion similar to retinal pigment epithelium. Eventually, intraocular NNR allografts became infiltrated with CD3+ T cells, and Ma and Streilein (13) proposed that rejection of allogeneic NNR grafts within the eye was accomplished by T cells attacking these donor microglia. Among allogeneic NNR grafts placed beneath the kidney capsule, we observed far fewer class II MHC+ microglia of donor origin than were ever observed in intraocular NNR grafts. Moreover, no donor microglia were found within rosettes of grafts beneath the kidney capsule. Instead, NNR allografts beneath the kidney capsule acquired recipient-derived, class II MHC-bearing APCs in abundance at 12 days, before any histologic evidence of a rejection process. We are considering the possibilities that allogeneic NNRs beneath the kidney capsule sensitize their hosts by either of two routes: 1) donor microglia migrate out of the graft and directly activate alloreactive T cells in the draining lymph node; or 2) recipient APCs infiltrate the graft, endocytose dead or dying graft cells and carry Ags derived from these cells to the draining lymph node for T cell activation. Experiments to test these interesting possibilities are now underway.
We were interested to find that lymph nodes of mice bearing allogeneic NNR grafts beneath the kidney capsule contain regulatory cells that suppress DH directed at alloantigens expressed on the grafts. Future experiments need to be performed to determine whether these regulators are T cells or APC, and whether the suppression is alloantigen specific. Our interest is piqued by the previously reported finding that Ags placed within the striatum of mouse brains also generate regulatory cells that appear in lymph nodes, rather than spleen (21). A simple explanation for why allogeneic NNR grafts beneath the kidney capsule and Ag injected into the brain give rise to lymph node-based regulators is that these tissues each enjoy a robust lymphatic drainage. By contrast, Ag placed in the eye generates regulators that are found primarily in the spleen, and the eye has minimal lymphatic drainage, but a robust pathway for delivery of ocular fluids i.v.
If future experiments confirm our proposal that NNR is partially immune-privileged, the prospect of using immature retinal tissue for transplantation is made brighter. Moreover, if our suspicion is correct that microglia within NNR grafts are responsible for calling the attention of the host immune response to alloantigens on the graft, then strategies to reduce, eliminate, or inactivate these cells within NNR tissue before implantation should have a salutary effect on retinal transplant survival.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. J. Wayne Streilein, The Schepens Eye Research Institute, Department of Ophthalmology, Harvard Medical School, 20 Staniford Street, Boston, MA 02114. E-mail address: waynes{at}vision.eri.harvard.edu
3 Abbreviations used in this paper: DH, delayed hypersensitivity; CD95L, CD95 ligand; IRBP, interphotoreceptor retinoid binding protein; NNR, neonatal neuronal retina; PEC, peritoneal exudate cells.
Received for publication July 9, 2002. Accepted for publication September 10, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. F. Ng, N. Kitaichi, and A. W. Taylor In Vitro Generated Autoimmune Regulatory T Cells Enhance Intravitreous Allogeneic Retinal Graft Survival Invest. Ophthalmol. Vis. Sci., November 1, 2007; 48(11): 5112 - 5117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. F. Ng, E. Lavik, H. Keino, A. W. Taylor, R. S. Langer, and M. J. Young Creating an Immune-Privileged Site Using Retinal Progenitor Cells and Biodegradable Polymers Stem Cells, June 1, 2007; 25(6): 1552 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Young, T. Borras, M. Walter, and R. Ritch Tissue Bioengineering: Potential Applications to Glaucoma Arch Ophthalmol, December 1, 2005; 123(12): 1725 - 1731. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
