December 2004
Volume 45, Issue 12
Free
Retina  |   December 2004
Integration between Abutting Retinas: Role of Glial Structures and Associated Molecules at the Interface
Author Affiliations
  • Yiqin Zhang
    From the Departments of Ophthalmology and
    Schepens Eye Research Institute, Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts; and the
  • Agnieszka K. Kardaszewska
    From the Departments of Ophthalmology and
    T. Krwawicz Department of Ophthalmology and First Eye Hospital, Lublin University School of Medicine, Lublin, Poland.
  • Theo van Veen
    From the Departments of Ophthalmology and
  • Uwe Rauch
    Experimental Pathology, Lund University, Lund, Sweden; the
  • Maria-Thereza R. Perez
    From the Departments of Ophthalmology and
Investigative Ophthalmology & Visual Science December 2004, Vol.45, 4440-4449. doi:10.1167/iovs.04-0165
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Yiqin Zhang, Agnieszka K. Kardaszewska, Theo van Veen, Uwe Rauch, Maria-Thereza R. Perez; Integration between Abutting Retinas: Role of Glial Structures and Associated Molecules at the Interface. Invest. Ophthalmol. Vis. Sci. 2004;45(12):4440-4449. doi: 10.1167/iovs.04-0165.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. Integration between subretinal grafts and the host retina is limited in part by the presence of a barrier at the graft–host interface. This study was conducted to identify factors that may contribute to this barrier, by examining the distribution of glial structures and associated molecules in different setups of overlapping retinal pieces.

methods. Neuroretinal tissue derived from mice that express green fluorescent protein (GFP) was fragmented and transplanted into the subretinal space of adult rd1 mice. In an in vitro system, two retinal pieces, derived from GFP and rd1 mice, respectively, were placed overlapping each other and forming either laminar–laminar pairs or fragment–laminar pairs. The glia-associated markers analyzed included glial fibrillary acidic protein (GFAP), cellular retinaldehyde-binding protein (CRALBP), and two molecules known to inhibit neurite outgrowth: CD44 and neurocan. Bridging fibers and migrated cells were visualized with GFP fluorescence and retinal cell markers.

results. A thick CRALBP-immunolabeled band was observed in the interface in cultured laminar–laminar pairs, whereas a thinner band was seen in cultured fragment–laminar pairs and in transplants. Accumulation of CD44 and neurocan was also observed in the interface between abutting retinal pieces in all setups. GFP+ bridging fibers and GFP+ cells (some of which coexpressed neuronal markers) were observed within the abutting rd1 retina in some areas. However, such integration occurred exclusively where CRALBP, CD44, and neurocan immunolabeling appeared disrupted in the interface, but coincided with high GFAP expression within the rd1 retina.

conclusions. The results demonstrate that, on the one hand, an accumulation of glial-associated inhibitory molecules in the interface correlates with limited integration between overlapping retinal pieces. On the other hand, glial reactivity within the rd1 retina does not appear to be incompatible with integration.

Subretinally transplanted retinal cells (e.g., amacrine and bipolar cells) send neurites into the inner plexiform layer (IPL) of the host retina, despite a graft–host age mismatch. 1 2 This indicates not only that the grafted retinal cells retain a high degree of plasticity but also that the environment of the host retina is relatively permissive to the incoming graft neurites and that at least some guidance cues persist in the adult retina. However, although graft–host integration occurs, it is not extensive, and evidence of substantial functional improvement after transplantation is scarce. 1 2 3 4 5 6 7 8 9 10 11 12 Poor graft–host integration is observed, regardless of the age of the donor tissue (fetal or postnatal) or of the morphologic organization of the graft (laminar sheets, fragmented pieces, or dissociated cells). 
Using a recently developed in vitro system, 13 we found that extensive bridging of neuronal fibers occurs between two abutting retinas when the retinal pieces are placed side by side, but not if they overlap each other (which simulates the in vivo situation of subretinal transplantation). In the latter, integration was observed only in places where the surfaces of both retinal pieces were disrupted at the interface, which conforms to findings in previous in vivo studies. 5 12 14 After neuronal damage, an activation of glial cells occurs that involves structural and molecular changes, some of which have been reported to be inhibitory to neurite outgrowth. 15 16 17 18 19 20 To examine whether reactive glial processes and accumulation of inhibitory molecules at the interface may contribute to the barring of fibers crossing through the retinal surfaces, we examined the distribution of certain glia-associated molecules in different setups: abutting retinas co-cultured to simulate subretinal transplantation of laminar or fragmented tissue and subretinal transplantation. 
The proteins analyzed included the well-established glial cell markers, cellular retinaldehyde-binding protein (CRALBP) and glial fibrillary acidic protein (GFAP), as well as CD44 and neurocan. CRALBP binds 11-cis-retinol, which can be used by cones to regenerate photopigment. 21 22 23 In the normal retina, CRALBP expression is found in all cytoplasmic compartments of Müller cells but also in association with astrocytes during the first two postnatal weeks. 24 CRALBP immunolabeling is particularly visible in association with the apical processes of Müller cells, at the level of the outer limiting membrane (OLM). 21 CRALBP is used extensively to disclose retinal glial cell profiles, and previous studies have indicated that its expression is altered when the retina is challenged. 25 GFAP is expressed almost exclusively in astrocytes in the normal retina, but its levels are rapidly upregulated also in Müller cells in response to retinal injury and is therefore often used as a marker of glial cell activation. 26 27 CD44 is a cell surface glycoprotein involved in cell–matrix adhesion, motility, growth factor signaling, and metastasis. 28 29 Neurocan is a major constituent of chondroitin sulfate proteoglycans in the developing central nervous system (CNS) and is involved in the regulation of cell migration, axonal growth, and axonal path finding. 30 CD44 and neurocan are well-characterized hyaluronan-binding proteins, shown to inhibit neurite and axon outgrowth. 28 29 30 31 32 33 They are both expressed in the adult normal retina: CD44 is found in the Müller cell apical processes, whereas neurocan is expressed at the level of the interphotoreceptor matrix (IPM) and in the nerve fiber layer. 34 35 36 37 However, after injury or photoreceptor degeneration, the expression of these proteins is upregulated in glial cells, 34 35 37 38 39 similar to observations in other areas of the CNS. 19 20 40 41 42 43 44  
Materials and Methods
Animals
GFP mice (harboring a transgene consisting of enhanced-green fluorescent protein [EGFP] cDNA under the control of a chicken β-actin promoter and cytomegalovirus enhancer), 45 at the ages of postnatal day (P)0 or P5 and rd1 mice (C3H/HeA, rd/rd) 60 days old (P60) were used as donors and recipients, respectively, of retinal transplants and to compose the abutting retinal pairs in culture. In the rd1 retina, a nearly complete rod photoreceptor cell loss is noted as early as the third postnatal week. 46 47 The experiments were conducted with the approval of the local animal experimentation ethics committee. Animals were handled according to the guidelines on care and use of experimental animals set by the Government Committee on Animal Experimentation at the University of Lund and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 
Tissue Culture Preparation
The culture procedure has been described previously. 13 Briefly, after the superior and nasal cornea was marked, the eyes were enucleated under sterile conditions and transferred to a dish containing serum-free medium (R16; Invitrogen-Gibco, Gaithersburg, MD). 48 Each retina was dissected from the RPE and from the optic nerve head and was cut in fresh medium into four pieces along the superior–inferior and the nasal–temporal axes, in such a way that each quadrant could be identified. Retinal pieces from GFP and rd1 mice were placed flat and overlapping each other, forming laminar–laminar pairs (n = 6). In these pairs, the inner (vitreal) surface of the piece derived from P0 GFP mice faced the outer surface of the piece derived from P60 rd1 mice, thus simulating subretinal transplantation of a laminar sheet of donor tissue. In a second setup, fragmented pieces of retinas derived from the P0 GFP mice were placed abutting the outer surface of the piece derived from P60 rd1 mice (fragment–laminar pairs, n = 6), simulating subretinal transplantation of fragmented donor tissue. The retinal pairs were mounted on a cellulose filter attached to a polyamide grid, with the GFP-mouse–derived piece closest to the filter in laminar–laminar pairs and with the rd1-mouse–derived piece closest to the filter in fragment–laminar pairs. The pairs were incubated for 12 days in 1.6 mL of R16 serum-free medium at 37°C, and the medium was changed every 2 days. In the illustrations of cultured laminar–laminar pairs shown in the present study, the images were flipped vertically to fit with the orientation of transplant-recipient retinas. 
Transplantation
The surgical procedure used for transplantation in this study has been described previously. 2 12 49 Briefly, eyes of P0 or P5 GFP mice were enucleated. The neural retinas were dissected, transferred to fresh cold 4°C Ames’ medium (Sigma-Aldrich, St. Louis, MO) and cut into small pieces. The retinal pieces did not include the RPE or the optic nerve head region. The P60 rd1 recipient mice were anesthetized by an intraperitoneal injection of xylazine (Rompun, 100 mg/kg; Bayer AG, Göteborg, Sweden) and ketamine (Ketalar, 100 mg/kg; Parke-Davis, NJ) and locally anesthetized with 1% amethocaine hydrochloride. Donor retinal tissue (0.7 μL total volume) was drawn into a plastic (polyethylene) pipette tip (GELoader Tip; Eppendorf, Hamburg, Germany) connected to a precision microsyringe (Hamilton, Reno, NV) and injected through the sclera into the superior subretinal space of the recipients. Transplant-recipients were anesthetized with carbon dioxide and decapitated at different survival times: Animals receiving P0 donor tissue (n = 8) were killed 7 days after transplantation, and animals receiving P5 donor tissue were killed 7 (n = 5) or 21 (n = 4) days after transplantation. No immunosuppression was used. 
Tissue Preparation and Analysis
The transplant-recipient eyes were quickly enucleated, and the anterior segment, lens, and vitreous body were removed. The eyecups containing transplants and the explants were fixed in 4% paraformaldehyde in Sörensen’s buffer (0.1 M; pH 7.2) and kept at room temperature for 2 hours. The tissue was subsequently rinsed, cryoprotected in Sörensen’s buffer containing increasing concentrations of sucrose, embedded in an albumin-gelatin medium (30 g egg albumin, 3 g gelatin, 100 mL distilled water), and frozen. Sections were obtained on a cryostat (12 μm), collected on gelatin/chrome alum-coated glass slides, air-dried, and stored at −20°C until further processing. Some sections were stained with hematoxylin and eosin. The other sections were processed for immunohistochemistry using antibodies against well-established retinal neuronal and glial cell markers and against CD44 and neurocan. The primary antibody specifications and sources are listed in Table 1 . Sections were preincubated at room temperature for 90 minutes with 0.1 M PBS containing 0.25% Triton X-100, 1% BSA, and 5% normal serum, followed by incubation for 12 to 24 hours at 4°C with primary antibodies. After they were rinsed, sections were incubated for 90 minutes with Texas red sulfonyl chloride or Alexa Fluor 350 (Molecular Probes, Eugene, OR) conjugated to donkey anti-mouse, donkey anti-sheep, donkey anti-rat, donkey anti-rabbit, or goat anti-rat (1:100; Jackson ImmunoResearch, West Grove, PA, and Molecular Probes). After the staining was complete, the sections were rinsed, coverslipped with buffered glycerols containing the anti-fading agent phenylenediamine, and viewed with a light microscope equipped for fluorescence microscopy. 
Quantification
The number of GFP+ bridging fibers and of GFP+ cell bodies within the rd1 host retinas was determined in transplant-recipient animals killed 7 and 21 days after transplantation. Only specimen in which there was no indication that the transplants had been inadvertently placed intraretinally or intravitreally were used for quantification. Sections containing a subretinal transplant were screened under the microscope at 72-μm intervals. Results are expressed as the number of cells and fibers per 100 μm (mean ± SD). Statistical analysis was performed with the unpaired Student’s t-test and one-way analysis of variance (ANOVA). P < 0.05 was considered statistically significant. 
Results
Histologic Appearance
In cultured laminar–laminar pairs, both retinal pieces (derived from GFP and rd1 mice), displayed a stratified structure. All retinal layers in GFP-mouse–derived pieces (GFP pieces) as well as the inner retinal layers in rd1-mouse–derived pieces (rd1 pieces) were well preserved. The morphologic organization of both pieces was comparable to that of intact tissue, except for a slight overall thinning of the retinas and poor development of photoreceptor outer segments in the GFP pieces. Good adhesion between the two pieces was observed. Nevertheless, the inner surface of the GFP piece and the outer surface of the rd1 piece were clearly identifiable, creating, in most cases, a well-defined border between the abutting pieces. In some of the cultured pairs, structural disruption was observed at the edges of both retinal pieces, and a clear border was not identifiable in the overlapping regions (Fig. 1A)
In fragment–laminar pairs, the retinal pieces derived from GFP mice displayed a disorganized architecture, and specific layers were not distinguishable, except in association with rosettes, in which photoreceptor cells were recognized. Cells belonging to inner retinal layers were noted between the rosettes. The lamination of the pieces derived from rd1 mice was preserved, but the outer margins were jagged in some places (Fig. 1B)
After subretinal transplantation to the adult rd1 mice, the GFP pieces showed a morphology analogous to that of fragmented pieces in culture, except that typical rosettes (with photoreceptors surrounded by inner retinal cells) were more often present in transplants (Fig. 1E) . The same was observed irrespective of the age of the donor tissue (P0 or P5). The host rd1 retina also demonstrated a histologic arrangement similar to that observed in the equivalent pieces in fragment–laminar cultures (Fig. 1E)
Integration between GFP and rd1 Retinas
Both in vitro and in vivo, retinas derived from GFP mice exhibited a high level of GFP expression. This enabled easy localization of not only the border between the abutting pieces, but also of GFP-fluorescent (GFP+) cells and fibers within the rd1 piece. In cultured laminar–laminar pairs, GFP+ cells and processes were observed in the rd1 pieces, only near the edge of the abutting GFP piece (Figs. 1C 2D 3D 3E 4A) . In fragment–laminar pairs and after transplantation, GFP+ cells and processes were found within the rd1 piece, both at the edges and in a few places in the center of the overlapping area (Figs. 1D 1F 1G 1H 2H 2J 3F 3G 3H 3I 3J 3K 4B 4C 4D 4E 4F 4G 4H) . Most GFP+ cells observed within the rd1 retina were clustered in restricted areas, next to the interface or in the inner nuclear layer (INL), even at the longer survival time. Occasionally, a few GFP+ cells displaying a microglia-like morphology were found in the inner plexiform and ganglion cell layers (Fig. 4B)
In the areas where integration was observed, the border between the abutting retinal pieces was not always well defined, partially disguising the presence of GFP+ bridging fibers. Nevertheless, in a few cases GFP+ cells were noted within the rd1 retina, extending processes toward the graft and the IPL of the host retina (Fig. 1G)
There was a large variation in the number of GFP+ cells and fibers within the rd1 retina between different specimens and different sections of the same specimen in all setups. No significant difference was found between specimens analyzed 7 (cells, 7.3 ± 9.3; fibers, 3.7 ± 3.2) and 21 (cells, 8.5 ± 1.8; fibers, 6.9 ± 1.7) days after transplantation. 
CRALBP Expression
Immunolabeling of CRALBP disclosed the structural profile of glial cells in intact and manipulated retinas, including cultures and transplants. In intact retinas obtained from rd1 congenic control mice at P60, CRALBP immunoreactivity was observed in Müller cell bodies in the middle of the INL and in their radial processes spanning all layers of the retina. Labeling was also observed in association with the Müller cell apical processes, and in the inner retina, in association with their end feet (not shown). In intact rd1 mouse retinas at P60, the same distribution was observed. However, the density of immunolabeling appeared increased over the apical and vitreal Müller cell processes compared with normal retinas (not shown). In the cultured rd1 retinas and in the host rd1 retinas in transplantation, the Müller cell profile, as revealed by CRALBP immunostaining, was comparable to that of intact rd1 retinas, except that discontinuities were observed in the apical surface in some areas (Fig. 2) . The distribution of CRALBP in the GFP pieces of laminar–laminar pairs was comparable to that of intact retinas of the same age (not shown). As expected, the CRALBP-labeled structures appeared disorganized in retinas derived from GFP-mice, in the fragment–laminar pairs and in the transplants (Figs. 2E 2G 2I)
In the center of the overlapping areas in laminar–laminar pairs, CRALBP-labeled processes appeared in a wide band at the interface of the two abutting retinas (Fig. 2A) . Most of these processes also expressed GFP (Fig. 2B) and seemed to correspond mostly to Müller cell end feet (and possibly also some astrocytes) located at the inner margin of the GFP piece. In a few sections, processes labeled for CRALBP but not for GFP were discernible at the outer surface of the rd1 pieces (Fig. 2B) . At the edges of the abutting retinal piece in laminar–laminar pairs, CRALBP accumulation was either not observed at the interface or was discontinuous (Figs. 2C 2D) . In fragment–laminar pairs and in transplants, CRALBP labeling was also observed at the interface, but was restricted to a thinner band than in laminar–laminar pairs. These CRALBP-labeled processes were situated along the apical surface of the rd1 pieces and did not express GFP (Figs. 2E 2F 2G 2H 2I 2J) . CRALBP labeling appeared mostly continuous at this level (Fig. 2E) , but gaps were observed not only in the edges but also in a few places in the center of the overlapping areas (Figs. 2G 2I)
A direct correlation was noted between the presence of GFP+ cells (and of bridging fibers) within the abutting rd1 pieces and CRALBP labeling at the interface. In all setups, integration was observed only when and where CRALBP labeling appeared discontinuous or was absent at the interface (Figs. 2D 2H 2J)
CD44 and Neurocan Expression
In intact retinas obtained from rd1 congenic control mice at P60, weak CD44 immunolabeling appeared diffusely distributed above the OLM (not shown). In intact retinas of adult rd1 mice, a more localized and intense labeling was observed at the same level (not shown). Accumulation of labeling was also noted in cultured pairs in the interface between the abutting pieces (Figs. 3A 3D 3F) and in the nonabutting areas in the apical surface of the rd1 retinal piece (not shown). In vivo, labeling was observed in the graft–host interface (Figs. 3H 3J 3K) and along the OLM in the rest of the host rd1 retina (not shown). In laminar–laminar pairs, neurocan immunolabeling was observed in the IPL of the GFP piece (Figs. 3B 3C 4E) . Weak neurocan immunoreactivity was detected in intact adult rd1 retinas and in their congenic controls in the outer surface of the retina and in the nerve fiber layer (not shown). In contrast, strong and distinct neurocan immunostaining was noted in the interface of the abutting pieces in cultured pairs (Figs. 3B 3E 3G) and in the graft–host interface (Fig. 3I) . Unlike CD44, distinct neurocan expression was not visible in the outer surface of rd1 pieces in the nonabutting areas in cultured pairs, and only weak neurocan immunolabeling was observed in the host rd1 retina, in areas away from the graft (not shown). 
Double labeling of CD44 and neurocan disclosed the colocalization of CD44 and neurocan immunoreactivities in the interface of the abutting retinas in all setups (Figs. 3A 3B 3C) . Similar to observations with CRALBP, CD44 and neurocan immunolabeling appeared as continuous bands between the abutting pieces. However, irregular or no labeling was observed in regions where the retinal structure was disrupted. In laminar–laminar pairs, CD44 and neurocan immunolabeling were normally not seen in areas where the edge of one of the pieces came in contact with the other piece. In fragment–laminar pairs and in transplants, immunolabeling was absent in some places along the interface. GFP+ cells (located within the rd1 abutting piece) and bridging fibers were observed only in the areas where CD44 and neurocan labeling were absent at the interface in laminar–laminar pairs (Figs. 3D 3E) , in fragment–laminar pairs (Figs. 3F 3G) , and in transplants (Figs. 3H 3I 3J 3K)
Immunohistochemical Identification of Migrated Cells
By using markers for specific retinal cell types (Table 1) , GFP+ cells were identified in intact retinas of GFP mice as photoreceptors and glial cells (high GFP expression) and as amacrine, horizontal, and bipolar cells (lower GFP expression). Likewise, in abutting retinas both in vitro and in vivo, some GFP+ cells in the rd1 piece also coexpressed glial or neuronal markers. Figures 4A 4B 4C 4D show examples of cultured laminar–laminar and fragment–laminar pairs in which GFP+ cells within the rd1 piece also expressed amacrine cell markers, such as nitric oxide synthase (NOS) and calretinin (Figs. 4A 4B 4C) and the horizontal and amacrine cell marker, calbindin (Fig. 4D) . It was possible in some cases to follow the processes of GFP+ cells within the rd1 piece projecting to the corresponding synaptic layers. GFP+ cells in the rd1 piece did not coexpress rhodopsin, suggesting that rod photoreceptors may not migrate into the abutting rd1 retinal piece. 
GFAP Expression
GFAP labeling was observed not only in astrocytes but also in Müller cells in cultured retinas and in the graft and host retinas after transplantation (Figs. 4E 4F 4G 4H) . A few GFP+ cells and fibers in the abutting rd1 piece coexpressed GFAP. Further, colocalization of GFAP and GFP revealed that integration between the abutting pieces coincided with areas of particularly high GFAP immunoreactivity. Both, in laminar–laminar and fragment–laminar cultured pairs, GFP+ cells and fibers were found within the rd1 retinal piece in regions where intensely GFAP-labeled Müller cell processes were observed (Fig. 4F) . Likewise, in transplantation, integration occurred in areas where distinct GFAP signal was observed in the Müller cell radial processes and end feet and in the astrocytes of the host retina (Fig. 4H)
Discussion
The presence of a barrier at the interface has been assumed to be one of the reasons for limited graft–host integration, but a more systematic analysis of the factors present at the interface has not been performed. In the present study (1) integration was particularly limited in a laminar–laminar configuration (it occurred only in the edges); (2) integration was observed in fragment–laminar pairs in multiple places, similar to observations after subretinal transplantation of fragmented retina; (3) the presence of a glial boundary and accumulation of at least two inhibitory molecules at the interface correlated with a lack of integration; (4) in places where integration occurred (through breaks of the boundary at the interface), not only fibers but also cells invaded the abutting retina; and (5) migration of cells was accompanied by increased glial reactivity within the abutting retina. 
Association of Lack of Integration with the Glial Boundary in the Interface
In the interface in fragment–laminar pairs and in transplantation, CRALBP immunostaining was observed along the outer surface of the rd1 retinal piece. The density of CRALBP immunostaining was increased in the retinal surfaces of intact rd1 retinas, of cultured GFP and rd1 retinas, and of rd1 host retinas compared with intact normal retinas, probably reflecting an activation of Müller cells. This conforms to previous observations on glial cell activation with outgrowth of Müller cell processes after retinal detachment or photoreceptor cell loss, for example. 25 26 34 53 54 There are also indications that a, so-called, glial seal forms in the outer retina, as photoreceptors are lost. 54 55 56 The precise correlation observed in the present study between a lack of integration and the presence of CRALBP and of at least two molecules inhibitory to neurite extension suggests that the Müller cell apical processes and the molecular environment at the interface are likely to play a role in limiting integration in the fragment–laminar configuration. 
Numerous neurite growth-inhibitory molecules have been found to be upregulated on the surface of nonpermissive glial structures after injury to the CNS, as well as in the graft–host interface in brain cell transplantation, constituting a molecular barrier preventing graft-derived axons from entering the host. 15 16 17 18 19 20 43 44 57 58 59 In the spinal cord, especially chondroitin sulfate proteoglycans appear to contribute to the nonpermissive properties of the glial scar tissue, since it has been observed that treatment with chondroitinase, enzyme-digesting chondroitin sulfate chains, promotes regeneration of axons and even functional recovery. 60 An analysis of changes in the distribution and protein levels of five chondroitin sulfate proteoglycans in spinal cord scar tissue has also shown a robust increase in two of them, NG2 and neurocan, suggesting that particularly these proteoglycans play a role in preventing axon regeneration. 44 Neurocan is synthesized and released during development, predominantly by neurons. 30 61 However, the site of neurocan synthesis switches to reactive glia after injury or degeneration, leading to an upregulation around the glial cells, as shown in adult brain, spinal cord, and retina. 19 20 37 38 42 43 44 62 63 Further, it has been shown that the upregulated neurocan is expressed in a sharp border defining the region of neuronal sprouting. 42 CD44 is normally synthesized by glial cells in the CNS and is upregulated under pathologic conditions on the surface of reactive glial cells in the brain 41 and in the retina. 34 35 39 Both CD44 and neurocan bind hyaluronan, which has been detected in the mouse IPM (Rayborn ME, et al. IOVS 2002;43:ARVO E-Abstract 2862). CD44 has also been shown to bind to molecules modified with chondroitin sulfate proteoglycans, 64 found in the IPM as well. 37 65 66 67 Further, there is strong evidence that both CD44 and neurocan can inhibit retinal axon and neurite outgrowth. 31 32 33 It is thus reasonable to assume that these molecules could be associated with nonpermissive, reactive gliosis at the retinal outer surface and that they could contribute to limitation of the integration between abutting retinas in fragment–laminar pairs and in transplantation of fragmented tissue. 
In the laminar–laminar configuration, CRALBP immunostaining was observed not only in the outer surface of the rd1 retinal piece, but also in association with glial elements in the vitreal side of the GFP retinal piece. Previous studies have shown that during development, Müller cell end feet appear to contribute to orienting retinal ganglion cell axons to grow along the inner retinal surface, whereas their dendrites are directed toward the IPL. 68 69 Further, as observed previously 36 37 and in the present study, neurocan is accumulated in the IPL and at the level of the nerve fiber layer during development, where it is likely to contribute to directing dendritic and axonal outgrowth. 32 33 In line with this, neuronal fibers in laminar–laminar pairs ran along the border without crossing it, even when the two overlapping retinas were placed with the inner retinal surfaces against each other. 13 Thus, in the laminar–laminar configuration, integration appears to be prevented not only at the level of the outer surface of the rd1 retinal piece but also by the environment in the innermost layers of the GFP piece. 
It should be noted also that although both CD44 and neurocan were expressed in the interface of abutting retinas, the distribution of the two was somewhat different. CD44 expression was increased in rd1 intact retinas (compared with normal intact retinas) and no further increase was observed in culture or in transplantation. Further, CD44 was detected in the outer retinal surface of the entire rd1 retina, including overlapping and nonoverlapping areas. In contrast, accumulation of neurocan was not observed in the outer surface of the rd1 retinal piece in the nonoverlapping areas, which suggests that the factors triggering CD44 upregulation are primarily associated with the rd1 retinal degeneration, whereas neurocan accumulation (by increased synthesis and/or decreased degradation) results from the superimposing of the retinal pieces. Neurocan is a secreted and therefore diffusible molecule, so the accumulation observed in laminar–laminar pairs could be of GFP retinal origin, because this tissue was derived from a developmental stage during which, as mentioned earlier, neurocan deposition has been observed in the inner retina. In contrast, accumulation was observed in the interface also in fragment–laminar pairs and between graft and host, suggesting that neurocan may be produced by cells within the rd1 retina. It thus appears that molecular changes are induced not only as a result of glial activation (due to degeneration/manipulation), but also by the unnatural situation of the overlapping retinas, which also influences integration negatively. 
Effect of the Disruption of the Glial Boundary on Cell Migration into the Abutting Retina
As previously shown 70 and confirmed in the present study, in the few places where integration is observed, not only GFP+ crossing fibers but also GFP+ cells were observed within the abutting rd1 retinal piece. During subretinal transplantation, the outer surface of the host retina may be mechanically damaged, and graft cells may be inadvertently placed within the host retina. Disruption of the outer retinal surface was observed in a few places in the cultured rd1 pieces, which may have occurred during detachment of the retina from the RPE and the handling of the tissue for culture. However, the rd1 and GFP pieces were carefully placed abutting each other, suggesting that the GFP+ cells encountered within the rd1 piece had indeed migrated into the latter. 
GFP+ fibers and cells were present within the abutting rd1 retinal piece, coincidentally with areas where thickening of glial radial processes and accumulation of GFAP were more pronounced. There is evidence showing that axonal elongation can occur in the presence of reactive gliosis. Some strongly suggest that glial cell hypertrophy and GFAP upregulation in fact promote axonal growth in the cerebrum and spinal cord. 71 72 73 Consistent with this suggestion is the observation that sprouting neurites, both in photoreceptor degeneration and in retinal detachment, are in close association with intensely GFAP-labeled Müller cells. 74 75 It conforms also with the observations in previous transplantation studies that graft-derived neuronal fibers can project to the degenerated host retina despite reactive gliosis and that the projections can persist for several months. 5 12 14  
It is not possible, though, to determine at this point how the two phenomena (neuronal migration and higher glial reactivity) correlate in the system examined in this study: whether increased Müller cell reactivity is induced as a result of a disruption of the OLM, which in turn allows cells to enter the rd1 retina, or whether a break in the OLM allows cells to enter, which locally induces increased glial reactivity. In a recent study, 76 integration of subretinally injected dissociated retinal cells was shown to be significantly favored if cells were transplanted to GFAP-vimentin–deficient mice. It is thus possible that the glial boundary at the outer surface of the host GFAP-vimentin–deficient retina is weakened, which in itself allows more cells to cross the interface. If so, it can be speculated that there are at least two different gliotic components involved: one inhibitory, at the retinal surface, and another within the retina itself, which is not necessarily nonpermissive. That also implies that hypertrophy and GFAP expression are not sufficient or even appropriate indicators of a nonpermissive environment. 
It should be noted also that many of the migrated GFP+ cells in the present study in the INL of the host rd1 retina expressed specific neuronal cell markers and that a few appeared to extend processes to the corresponding synaptic layers. During normal development, the radial glial cells of the retina (the Müller cells), similar to radial glia in the brain, guide not only neurite extension, but also the migration of neuronal cells across the retina. 77 In previous studies, neuronal fibers originating within the retinal graft projected to synaptic layers within the adult host retina. 1 2 12 Thus, the adult retina seems to retain the ability to direct not only the extension of the neurites, but also the position of the migrated cells, and radial Müller glial cells (even when activated, or perhaps through activation) may play a role in these processes by acting as scaffolds for the migrating cells, even in the adult. In the current study, however, these processes (crossing of fibers and cells through the interface) took place only after barriers at the interface were disrupted. 
Conclusions and Considerations
As shown in this study, the microenvironment at the interface between abutting retinas is not favorable for neurite bridging or migration of cells. It may be possible to achieve better integration by manipulating the astroglial environment of the host retina, as previously proposed. 74 However, whether this would also favor integration of grafts delivered as laminar sheets must be determined. As shown here, integration was particularly limited in the laminar configuration. At the outer surface of the rd1 retina, it consisted of Müller cell apical processes, which either lack growth-promoting properties, and/or where inhibitory molecules are produced. At the inner surface of the abutting piece (corresponding to the graft), it consisted of the Müller cell end feet, the ILM, and the basal lamina, which appear to favor axonal (but not dendritic) growth and which favor growth along (but not across) the surface. These observations suggest that for extensive integration to occur, it may be necessary to manipulate not only the host retina, but also somehow to eliminate the inner layers of the donor tissue before transplantation of laminar sheets. 
 
Table 1.
 
Primary Antibodies Used in Immunohistochemistry
Table 1.
 
Primary Antibodies Used in Immunohistochemistry
Antigen Localization Antiserum Dilution Source
Neuronal nitric oxide synthase Amacrine cells, (bipolar cells) 50 Sheep anti-nNOS 1:4000 I. G. Charles, P. C. Emson, Cambridge, UK
Protein kinase C Rod bipolar cells 51 Rabbit anti-PKC 1:1000 Chemicon, Temecula, CA
Calbindin-D (28 kDa) Horizontal cells, Amacrine cells 51 Mouse anti-calbindin 1:200 Sigma-Aldrich
Calretinin Amacrine cells 51 Mouse anti-calretinin 1:2000 Chemicon
Rhodopsin Rod photoreceptor cells 52 Mouse anti-Rho-1D4 1:400 R. S. Molday, Vancouver, BC, Canada
GFAP Müller cells, astrocytes 27 Rabbit anti-GFAP 1:1500 Dako, Glostrup, Denmark
Neurocan Extracellular matrix (neurons, glial cells) 36 37 Rabbit anti-NC2* 1:2000 U. Rauch, Lund, Sweden
CRALBP Müller cells, astrocytes 21 24 Rabbit anti-CRALBP 1:5000 J. Saari, Seattle, WA
CD44 glycoprotein Müller cells 34 35 39 Rat anti-CD44, † 1:100 BD Biosciences, San Diego, CA
Figure 1.
 
(A, B) Hematoxylin and eosin (HE) staining. Cross sections illustrating the morphology of abutting retinal pieces in a laminar pair (A) and in a fragment–laminar pair (B). A clear border is visible in (A) in the center of the overlapping areas (arrowheads) but not at the edges ( Image not available ). (B) The outer surface of the rd1 retinal piece appears disrupted in some places ( Image not available ). (C–D) GFP fluorescence. GFP fibers and cells are visible within the rd1 retinal piece at the edge of the overlapping area in a laminar pair (C). They were present in multiple regions, including center and edges in a fragment–laminar pair (D). (E) HE staining of a cross section illustrating the morphology of a fragment subretinal transplant in an animal receiving transplanted P0 donor tissue and killed after 12 days (P0+12 d): Typical rosettes with photoreceptors surrounded by inner retinal cells were present. (F) GFP+ cells and fibers were visible within the host rd1 retina in the P0+12 d samples. (G) Example of GFP+ cell within the host rd1 retina extending neurites toward the transplant and the host IPL in animals that received P5 donor tissue and were killed after 7 days (P5+7 d). (H) GFP+ cells and fibers were seen in a few areas within the host rd1 retina also in P5+21 d samples. GFP, retinal piece derived from GFP mice; rd, retinal piece derived from rd1 mice; GFP(t), transplant derived from a GFP mouse.
Figure 1.
 
(A, B) Hematoxylin and eosin (HE) staining. Cross sections illustrating the morphology of abutting retinal pieces in a laminar pair (A) and in a fragment–laminar pair (B). A clear border is visible in (A) in the center of the overlapping areas (arrowheads) but not at the edges ( Image not available ). (B) The outer surface of the rd1 retinal piece appears disrupted in some places ( Image not available ). (C–D) GFP fluorescence. GFP fibers and cells are visible within the rd1 retinal piece at the edge of the overlapping area in a laminar pair (C). They were present in multiple regions, including center and edges in a fragment–laminar pair (D). (E) HE staining of a cross section illustrating the morphology of a fragment subretinal transplant in an animal receiving transplanted P0 donor tissue and killed after 12 days (P0+12 d): Typical rosettes with photoreceptors surrounded by inner retinal cells were present. (F) GFP+ cells and fibers were visible within the host rd1 retina in the P0+12 d samples. (G) Example of GFP+ cell within the host rd1 retina extending neurites toward the transplant and the host IPL in animals that received P5 donor tissue and were killed after 7 days (P5+7 d). (H) GFP+ cells and fibers were seen in a few areas within the host rd1 retina also in P5+21 d samples. GFP, retinal piece derived from GFP mice; rd, retinal piece derived from rd1 mice; GFP(t), transplant derived from a GFP mouse.
Figure 2.
 
(A–H) Double labeling of GFP (green) and CRALBP (red). Glial structures in the interface of two pieces were preserved in the center of the overlapping area in a laminar pair (A, arrowheads) where integration was not seen (B). Most CRALBP-labeled processes at the interface also expressed GFP, although some processes labeled for CRALBP but not GFP were seen at the outer margin of the rd1 piece (B, arrows). When the glial structure in the interface was disrupted at the edges (C, arrow), integration was visible (D, arrow). Glial structures of GFP pieces in fragment–laminar pairs appeared disorganized (E). Yet, as long as they were preserved in rd1 pieces (E, arrowheads), integration was not visible (F). When the structures of both pieces were disrupted (in the central part in fragment–laminar pairs; G, arrow), integration occurred in the corresponding area (H, arrow). (I, J) Double labeling of GFP (green) and CRALBP (red) in a transplant. Glial structures in the graft–host interface were disrupted (I, arrows), and integration was noted in the corresponding area (J, arrows).
Figure 2.
 
(A–H) Double labeling of GFP (green) and CRALBP (red). Glial structures in the interface of two pieces were preserved in the center of the overlapping area in a laminar pair (A, arrowheads) where integration was not seen (B). Most CRALBP-labeled processes at the interface also expressed GFP, although some processes labeled for CRALBP but not GFP were seen at the outer margin of the rd1 piece (B, arrows). When the glial structure in the interface was disrupted at the edges (C, arrow), integration was visible (D, arrow). Glial structures of GFP pieces in fragment–laminar pairs appeared disorganized (E). Yet, as long as they were preserved in rd1 pieces (E, arrowheads), integration was not visible (F). When the structures of both pieces were disrupted (in the central part in fragment–laminar pairs; G, arrow), integration occurred in the corresponding area (H, arrow). (I, J) Double labeling of GFP (green) and CRALBP (red) in a transplant. Glial structures in the graft–host interface were disrupted (I, arrows), and integration was noted in the corresponding area (J, arrows).
Figure 3.
 
(A–I) Triple labeling of GFP (green), CD44 (blue), and neurocan (red) in laminar pairs (A–E), fragment–laminar pairs (F, G), and transplants (H, I). (J, K) Double labeling of GFP (green) and CD44 (red) in transplants. (H, I) Animals that received transplants of P0 donor tissue, killed after 12 days; (J, K) animals that received transplants of P5 donor tissue, killed after (J) 7 and (K) 21 days. CD44 (A) and neurocan (B) expression appeared in a band and colocalized in the interface between the abutting retinas (C; D–G, arrowheads). Neurocan immunolabeling was also visible in the GFP piece in the IPL (B, C, E). Integration occurred only in areas where CD44/neurocan expression was not observed in the interface (D–K, arrows).
Figure 3.
 
(A–I) Triple labeling of GFP (green), CD44 (blue), and neurocan (red) in laminar pairs (A–E), fragment–laminar pairs (F, G), and transplants (H, I). (J, K) Double labeling of GFP (green) and CD44 (red) in transplants. (H, I) Animals that received transplants of P0 donor tissue, killed after 12 days; (J, K) animals that received transplants of P5 donor tissue, killed after (J) 7 and (K) 21 days. CD44 (A) and neurocan (B) expression appeared in a band and colocalized in the interface between the abutting retinas (C; D–G, arrowheads). Neurocan immunolabeling was also visible in the GFP piece in the IPL (B, C, E). Integration occurred only in areas where CD44/neurocan expression was not observed in the interface (D–K, arrows).
Figure 4.
 
(A-D) Double labeling of GFP (green) and neuronal markers (red), illustrating the presence of GFP+ cells and fibers within the INL of the rd1 retina that coexpressed NOS in a laminar pair (A) and in a fragment–laminar pair (B, large arrow). A few GFP+ cells are also noted in the GCL (small arrows), and some with a microglia-like morphology in the IPL (arrowheads). (C, D) GFP+ cells in the INL of the rd1 piece that coexpressed calbindin and calretinin, respectively, in fragment–laminar pairs (arrows). (E–H) Double labeling of GFP (green) and GFAP (red) in fragment–laminar pairs (E, F) and transplants (G, H) showing that integration (E–H, arrows) coincides with regions where hypertrophic Müller cells (E, F, arrows) and astrocytes (G-H, ★) appeared to be intensely GFAP positive. In transplants, the overall GFAP labeling in the host retina was lower in areas away from the graft (not shown).
Figure 4.
 
(A-D) Double labeling of GFP (green) and neuronal markers (red), illustrating the presence of GFP+ cells and fibers within the INL of the rd1 retina that coexpressed NOS in a laminar pair (A) and in a fragment–laminar pair (B, large arrow). A few GFP+ cells are also noted in the GCL (small arrows), and some with a microglia-like morphology in the IPL (arrowheads). (C, D) GFP+ cells in the INL of the rd1 piece that coexpressed calbindin and calretinin, respectively, in fragment–laminar pairs (arrows). (E–H) Double labeling of GFP (green) and GFAP (red) in fragment–laminar pairs (E, F) and transplants (G, H) showing that integration (E–H, arrows) coincides with regions where hypertrophic Müller cells (E, F, arrows) and astrocytes (G-H, ★) appeared to be intensely GFAP positive. In transplants, the overall GFAP labeling in the host retina was lower in areas away from the graft (not shown).
The authors thank A. Romeo Caffé and Berndt Ehinger for helpful discussions, Birgitta Klefbohm for animal care, and Hodan Abdalle for assistance with culture media. 
Ghosh F, Bruun A, Ehinger B. Graft-host connections in long-term full-thickness embryonic rabbit retinal transplants. Invest Ophthalmol Vis Sci. 1999;40:126–132. [PubMed]
Zhang Y, Sharma RK, Ehinger B, Perez MTR. Nitric oxide-producing cells project from retinal grafts to the inner plexiform layer of the host retina. Invest Ophthalmol Vis Sci. 1999;40:3062–3066. [PubMed]
Silverman MS, Hughes SE, Valentino TL, Liu Y. Photoreceptor transplantation: anatomic, electrophysiologic, and behavioral evidence for the functional reconstruction of retinas lacking photoreceptors. Exp Neurol. 1992;115:87–94. [CrossRef] [PubMed]
Adolph AR, Zucker CL, Ehinger B, Bergström A. Function and structure in retinal transplants. J Neural Transplant Plast. 1994;5:147–161. [CrossRef] [PubMed]
Gouras P, Du J, Kjeldbye H, Yamamoto S, Zack DJ. Long-term photoreceptor transplants in dystrophic and normal mouse retina. Invest Ophthalmol Vis Sci. 1994;35:3145–3153. [PubMed]
Seiler MJ, Aramant RB. Transplantation of embryonic retinal donor cells labelled with BrdU or carrying a genetic marker to adult retina. Exp Brain Res. 1995;105:59–66. [PubMed]
Sharma RK, Perez MTR, Ehinger B. Immunocytochemical localisation of neuronal nitric oxide synthase in developing and transplanted rabbit retinas. Histochem Cell Biol. 1997;107:449–458. [CrossRef] [PubMed]
Kwan AS, Wang S, Lund RD. Photoreceptor layer reconstruction in a rodent model of retinal degeneration. Exp Neurol. 1999;159:21–33. [CrossRef] [PubMed]
Radner W, Sadda SR, Humayun MS, et al. Light-driven retinal ganglion cell responses in blind rd mice after neural retinal transplantation. Invest Ophthalmol Vis Sci. 2001;42:1057–1065. [PubMed]
Woch G, Aramant RB, Seiler MJ, Sagdullaev BT, McCall MA. Retinal transplants restore visually evoked responses in rats with photoreceptor degeneration. Invest Ophthalmol Vis Sci. 2001;42:1669–1676. [PubMed]
Radtke ND, Seiler MJ, Aramant RB, Petry HM, Pidwell DJ. Transplantation of intact sheets of fetal neural retina with its retinal pigment epithelium in retinitis pigmentosa patients. Am J Ophthalmol. 2002;133:544–550. [CrossRef] [PubMed]
Zhang Y, Arnér K, Ehinger B, Perez MTR. Limitation of anatomical integration between subretinal transplants and the host retina. Invest Ophthalmol Vis Sci. 2003;44:324–331. [CrossRef] [PubMed]
Zhang Y, Caffé AR, Azadi S, van Veen T, Ehinger B, Perez MTR. Neuronal integration in an abutting-retinas culture system. Invest Ophthalmol Vis Sci. 2003.4936–4946.
Gouras P, Tanabe T. Survival and integration of neural retinal transplants in rd mice. Graefes Arch Clin Exp Ophthalmol. 2003;241:403–409. [CrossRef] [PubMed]
Snow DM, Lemmon V, Carrino DA, Caplan AI, Silver J. Sulfated proteoglycans in astroglial barriers inhibit neurite outgrowth in vitro. Exp Neurol. 1990;109:111–130. [CrossRef] [PubMed]
McKeon RJ, Schreiber RC, Rudge JS, Silver J. Reduction of neurite outgrowth in a model of glial scarring following CNS injury is correlated with the expression of inhibitory molecules on reactive astrocytes. J Neurosci. 1991;11:3398–3411. [PubMed]
Davies SJ, Fitch MT, Memberg SP, Hall AK, Raisman G, Silver J. Regeneration of adult axons in white matter tracts of the central nervous system. Nature. 1997;390:680–683. [PubMed]
Nieto-Sampedro M. Neurite outgrowth inhibitors in gliotic tissue. Adv Exp Med Biol. 1999;468:207–224. [PubMed]
Asher RA, Morgenstern DA, Moon LD, Fawcett JW. Chondroitin sulphate proteoglycans: inhibitory components of the glial scar. Prog Brain Res. 2001;132:611–619. [PubMed]
Jones LL, Margolis RU, Tuszynski MH. The chondroitin sulfate proteoglycans neurocan, brevican, phosphacan, and versican are differentially regulated following spinal cord injury. Exp Neurol. 2003;182:399–411. [CrossRef] [PubMed]
Bunt-Milam AH, Saari JC. Immunocytochemical localization of two retinoid-binding proteins in vertebrate retina. J Cell Biol. 1983;97:703–712. [CrossRef] [PubMed]
Mata NL, Radu RA, Clemmons RC, Travis GH. Isomerization and oxidation of vitamin a in cone-dominant retinas: a novel pathway for visual-pigment regeneration in daylight. Neuron. 2002;36:69–80. [CrossRef] [PubMed]
Kennedy BN, Li C, Ortego J, Coca-Prados M, Sarthy VP, Crabb JW. CRALBP transcriptional regulation in ciliary epithelial, retinal Müller and retinal pigment epithelial cells. Exp Eye Res. 2003;76:257–260. [CrossRef] [PubMed]
Johnson PT, Geller SF, Lewis GP, Reese BE. Cellular retinaldehyde binding protein in developing retinal astrocytes. Exp Eye Res. 1997;64:759–766. [CrossRef] [PubMed]
Sheedlo HJ, Jaynes D, Bolan AL, Turner JE. Müllerian glia in dystrophic rodent retinas: an immunocytochemical analysis. Brain Res Dev Brain Res. 1995;85:171–180. [CrossRef] [PubMed]
Bringmann A, Reichenbach A. Role of Müller cells in retinal degenerations. Front Biosci. 2001;6:E72–E92. [PubMed]
Sarthy V, Ripps H. The Retina Müller Cell: Structure and Function. 2001; Kluwer Academic/Plenum Press New York.
Sherman LS, Struve JN, Rangwala R, Wallingford NM, Tuohy TM, Kuntz C. Hyaluronate-based extracellular matrix: keeping glia in their place. Glia. 2002;38:93–102. [CrossRef] [PubMed]
Ponta H, Sherman L, Herrlich PA. CD44: From adhesion molecules to signalling regulators. Nat Rev Mol Cell Biol. 2003;4:33–45. [CrossRef] [PubMed]
Rauch U, Feng K, Zhou XH. Neurocan: a brain chondroitin sulfate proteoglycan. Cell Mol Life Sci. 2001;58:1842–1856. [CrossRef] [PubMed]
Sretavan DW, Feng L, Pure E, Reichardt LF. Embryonic neurons of the developing optic chiasm express L1 and CD44, cell surface molecules with opposing effects on retinal axon growth. Neuron. 1994;12:957–975. [CrossRef] [PubMed]
Li H, Leung TC, Hoffman S, Balsamo J, Lilien J. Coordinate regulation of cadherin and integrin function by the chondroitin sulfate proteoglycan neurocan. J Cell Biol. 2000;149:1275–1288. [CrossRef] [PubMed]
Inatani M, Honjo M, Otori Y, et al. Inhibitory effects of neurocan and phosphacan on neurite outgrowth from retinal ganglion cells in culture. Invest Ophthalmol Vis Sci. 2001;42:1930–1938. [PubMed]
Chaitin MH, Ankrum MT, Wortham HS. Distribution of CD44 in the retina during development and the rds degeneration. Dev Brain Res. 1996;94:92–98. [CrossRef]
Chaitin MH, Brun-Zinkernagel AM. Immunolocalization of CD44 in the dystrophic rat retina. Exp Eye Res. 1998;67:283–292. [CrossRef] [PubMed]
Inatani M, Tanihara H, Oohira A, Honjo M, Honda Y. Identification of a nervous tissue-specific chondroitin sulfate proteoglycan, neurocan, in developing rat retina. Invest Ophthalmol Vis Sci. 1999;40:2350–2359. [PubMed]
Zhang Y, Rauch U, Perez MTR. Accumulation of neurocan, a brain chondroitin sulfate proteoglycan, in association with the retinal vasculature in RCS rats. Invest Ophthalmol Vis Sci. 2003;44:1252–1261. [CrossRef] [PubMed]
Inatani M, Tanihara H, Oohira A, Honjo M, Kido N, Honda Y. Upregulated expression of neurocan, a nervous tissue specific proteoglycan, in transient retinal ischemia. Invest Ophthalmol Vis Sci. 2000;41:2748–2754. [PubMed]
Krishnamoorthy R, Agarwal N, Chaitin MH. Upregulation of CD44 expression in the retina during the rds degeneration. Brain Res Mol Brain Res. 2000;77:125–130. [CrossRef] [PubMed]
Asher R, Bignami A. Hyaluronate binding and CD44 expression in human glioblastoma cells and astrocytes. Exp Cell Res. 1992;203:80–90. [CrossRef] [PubMed]
Girgrah N, Letarte M, Becker LE, Cruz TF, Theriault E, Moscarello MA. Localization of the CD44 glycoprotein to fibrous astrocytes in normal white matter and to reactive astrocytes in active lesions in multiple sclerosis. J Neuropathol Exp Neurol. 1991;50:779–792. [CrossRef] [PubMed]
Haas CA, Rauch U, Thon N, Merten T, Deller T. Entorhinal cortex lesion in adult rats induces the expression of the neuronal chondroitin sulfate proteoglycan neurocan in reactive astrocytes. J Neurosci. 1999;19:9953–9963. [PubMed]
Morgenstern DA, Asher RA, Fawcett JW. Chondroitin sulphate proteoglycans in the CNS injury response. Prog Brain Res. 2002;137:313–332. [PubMed]
Tang X, Davies JE, Davies SJ. Changes in distribution, cell associations, and protein expression levels of NG2, neurocan, phosphacan, brevican, versican V2, and tenascin-C during acute to chronic maturation of spinal cord scar tissue. J Neurosci Res. 2003;71:427–444. [CrossRef] [PubMed]
Okabe M, Ikawa M, Kominami K, Nakanishi T, Nishimune Y. “Green mice” as a source of ubiquitous green cells. FEBS Lett. 1997;407:313–319. [CrossRef] [PubMed]
Farber DB, Flannery JG, Bowes-Rickman C. The rd mouse story: seventy years of research on an animal model of inherited retinal degeneration. Prog Retin Eye Res. 1994;13:31–64. [CrossRef]
LaVail MM, Matthes MT, Yasumura D, Steinberg RH. Variability in rate of cone degeneration in the retinal degeneration (rd/rd) mouse. Exp Eye Res. 1997;65:45–50. [CrossRef] [PubMed]
Caffé AR, Ahuja P, Holmqvist B, et al. Mouse retina explants after long-term culture in serum free medium. J Chem Neuroanat. 2001;22:263–273. [PubMed]
Juliusson B, Bergström A, van Veen T, Ehinger B. Cellular organization in retinal transplants using cell suspensions or fragments of embryonic retinal tissue. Cell Transplant. 1993;2:411–418. [PubMed]
Perez MTR, Larsson B, Alm P, Andersson KE, Ehinger B. Localisation of neuronal nitric oxide synthase-immunoreactivity in rat and rabbit retinas. Exp Brain Res. 1995;104:207–217. [PubMed]
Haverkamp S, Wässle H. Immunocytochemical analysis of the mouse retina. J Comp Neurol. 2000;424:1–23. [CrossRef] [PubMed]
Molday RS, MacKenzie D. Monoclonal antibodies to rhodopsin: characterization, cross-reactivity, and application as structural probes. Biochemistry. 1983;22:653–660. [CrossRef] [PubMed]
Lewis GP, Matsumoto B, Fisher SK. Changes in the organization and expression of cytoskeletal proteins during retinal degeneration induced by retinal detachment. Invest Ophthalmol Vis Sci. 1995;36:2404–2416. [PubMed]
Jones BW, Watt CB, Frederick JM, et al. Retinal remodeling triggered by photoreceptor degenerations. J Comp Neurol. 2003;464:1–16. [CrossRef] [PubMed]
Milam AH, Li ZY, Cideciyan AV, Jacobson SG. Clinicopathologic effects of the Q64ter rhodopsin mutation in retinitis pigmentosa. Invest Ophthalmol Vis Sci. 1996;37:753–765. [PubMed]
Marc RE, Jones BW, Watt CB, Strettoi E. Neural remodeling in retinal degeneration. Prog Retin Eye Res. 2003;22:607–655. [CrossRef] [PubMed]
Reier PJ, Houle JD. The glial scar: its bearing on axonal elongation and transplantation approaches to CNS repair. Adv Neurol. 1988;47:87–138. [PubMed]
Krüger S, Sievers J, Hansen C, Sadler M, Berry M. Three morphologically distinct types of interface develop between adult host and fetal brain transplants: implications for scar formation in the adult central nervous system. J Comp Neurol. 1986;249:103–116. [CrossRef] [PubMed]
Fawcett JW, Asher RA. The glial scar and central nervous system repair. Brain Res Bull. 1999;49:377–391. [CrossRef] [PubMed]
Bradbury EJ, Moon LD, Popat RJ, et al. Chondroitinase ABC promotes functional recovery after spinal cord injury. Nature. 2002;416:636–640. [CrossRef] [PubMed]
Engel M, Maurel P, Margolis RU, Margolis RK. Chondroitin sulfate proteoglycans in the developing central nervous system. I. Cellular sites of synthesis of neurocan and phosphacan. J Comp Neurol. 1996;366:34–43. [CrossRef] [PubMed]
McKeon RJ, Jurynec MJ, Buck CR. The chondroitin sulfate proteoglycans neurocan and phosphacan are expressed by reactive astrocytes in the chronic CNS glial scar. J Neurosci. 1999;19:10778–10788. [PubMed]
Asher RA, Morgenstern DA, Fidler PS, et al. Neurocan is upregulated in injured brain and in cytokine-treated astrocytes. J Neurosci. 2000;20:2427–2438. [PubMed]
Kawashima H, Hirose M, Hirose J, Nagakubo D, Plaas AH, Miyasaka M. Binding of a large chondroitin sulfate/dermatan sulfate proteoglycan, versican, to L-selectin, P-selectin, and CD44. J Biol Chem. 2000;275:35448–35456. [CrossRef] [PubMed]
Porrello K, Yasumura D, LaVail MM. Immunogold localization of chondroitin 6-sulfate in the interphotoreceptor matrix of normal and RCS rats. Invest Ophthalmol Vis Sci. 1989;30:638–651. [PubMed]
Hollyfield JG, Rayborn ME, Midura RJ, Shadrach KG, Acharya S. Chondroitin sulfate proteoglycan core proteins in the interphotoreceptor matrix: a comparative study using biochemical and immunohistochemical analysis. Exp Eye Res. 1999;69:311–322. [CrossRef] [PubMed]
Inatani M, Tanihara H. Proteoglycans in retina. Prog Retin Eye Res. 2002;21:429–447. [CrossRef] [PubMed]
Bauch H, Stier H, Schlosshauer B. Axonal versus dendritic outgrowth is differentially affected by radial glia in discrete layers of the retina. J Neurosci. 1998;18:1774–1785. [PubMed]
Steinbach K, Schlosshauer B. Regulatory cell interactions between retinal ganglion cells and radial glia during axonal and dendritic outgrowth. Microsc Res Tech. 2000;48:12–24. [CrossRef] [PubMed]
Lu B, Kwan T, Kurimoto Y, Shatos M, Lund RD, Young MJ. Transplantation of EGF-responsive neurospheres from GFP transgenic mice into the eyes of rd mice. Brain Res. 2002;943:292–300. [CrossRef] [PubMed]
Kawaja MD, Gage FH. Reactive astrocytes are substrates for the growth of adult CNS axons in the presence of elevated levels of nerve growth factor. Neuron. 1991;7:1019–1030. [CrossRef] [PubMed]
Li Y, Raisman G. Sprouts from cut corticospinal axons persist in the presence of astrocytic scarring in long-term lesions of the adult rat spinal cord. Exp Neurol. 1995;134:102–111. [CrossRef] [PubMed]
Ajtai BM, Kalman M. Axon growth failure following corpus callosum lesions precedes glial reaction in perinatal rats. Anat Embryol (Berl). 2000;202:313–321. [CrossRef] [PubMed]
Lewis GP, Linberg KA, Fisher SK. Neurite outgrowth from bipolar and horizontal cells after experimental retinal detachment. Invest Ophthalmol Vis Sci. 1998;39:424–434. [PubMed]
Fariss RN, Li ZY, Milam AH. Abnormalities in rod photoreceptors, amacrine cells, and horizontal cells in human retinas with retinitis pigmentosa. Am J Ophthalmol. 2000;129:215–223. [CrossRef] [PubMed]
Kinouchi R, Takeda M, Yang L, et al. Robust neural integration from retinal transplants in mice deficient in GFAP and vimentin. Nat Neurosci. 2003;6:863–868. [CrossRef] [PubMed]
Willbold E, Layer PG. Müller glia cells and their possible roles during retina differentiation in vivo and in vitro. Histol Histopathol. 1998;13:531–552. [PubMed]
Figure 1.
 
(A, B) Hematoxylin and eosin (HE) staining. Cross sections illustrating the morphology of abutting retinal pieces in a laminar pair (A) and in a fragment–laminar pair (B). A clear border is visible in (A) in the center of the overlapping areas (arrowheads) but not at the edges ( Image not available ). (B) The outer surface of the rd1 retinal piece appears disrupted in some places ( Image not available ). (C–D) GFP fluorescence. GFP fibers and cells are visible within the rd1 retinal piece at the edge of the overlapping area in a laminar pair (C). They were present in multiple regions, including center and edges in a fragment–laminar pair (D). (E) HE staining of a cross section illustrating the morphology of a fragment subretinal transplant in an animal receiving transplanted P0 donor tissue and killed after 12 days (P0+12 d): Typical rosettes with photoreceptors surrounded by inner retinal cells were present. (F) GFP+ cells and fibers were visible within the host rd1 retina in the P0+12 d samples. (G) Example of GFP+ cell within the host rd1 retina extending neurites toward the transplant and the host IPL in animals that received P5 donor tissue and were killed after 7 days (P5+7 d). (H) GFP+ cells and fibers were seen in a few areas within the host rd1 retina also in P5+21 d samples. GFP, retinal piece derived from GFP mice; rd, retinal piece derived from rd1 mice; GFP(t), transplant derived from a GFP mouse.
Figure 1.
 
(A, B) Hematoxylin and eosin (HE) staining. Cross sections illustrating the morphology of abutting retinal pieces in a laminar pair (A) and in a fragment–laminar pair (B). A clear border is visible in (A) in the center of the overlapping areas (arrowheads) but not at the edges ( Image not available ). (B) The outer surface of the rd1 retinal piece appears disrupted in some places ( Image not available ). (C–D) GFP fluorescence. GFP fibers and cells are visible within the rd1 retinal piece at the edge of the overlapping area in a laminar pair (C). They were present in multiple regions, including center and edges in a fragment–laminar pair (D). (E) HE staining of a cross section illustrating the morphology of a fragment subretinal transplant in an animal receiving transplanted P0 donor tissue and killed after 12 days (P0+12 d): Typical rosettes with photoreceptors surrounded by inner retinal cells were present. (F) GFP+ cells and fibers were visible within the host rd1 retina in the P0+12 d samples. (G) Example of GFP+ cell within the host rd1 retina extending neurites toward the transplant and the host IPL in animals that received P5 donor tissue and were killed after 7 days (P5+7 d). (H) GFP+ cells and fibers were seen in a few areas within the host rd1 retina also in P5+21 d samples. GFP, retinal piece derived from GFP mice; rd, retinal piece derived from rd1 mice; GFP(t), transplant derived from a GFP mouse.
Figure 2.
 
(A–H) Double labeling of GFP (green) and CRALBP (red). Glial structures in the interface of two pieces were preserved in the center of the overlapping area in a laminar pair (A, arrowheads) where integration was not seen (B). Most CRALBP-labeled processes at the interface also expressed GFP, although some processes labeled for CRALBP but not GFP were seen at the outer margin of the rd1 piece (B, arrows). When the glial structure in the interface was disrupted at the edges (C, arrow), integration was visible (D, arrow). Glial structures of GFP pieces in fragment–laminar pairs appeared disorganized (E). Yet, as long as they were preserved in rd1 pieces (E, arrowheads), integration was not visible (F). When the structures of both pieces were disrupted (in the central part in fragment–laminar pairs; G, arrow), integration occurred in the corresponding area (H, arrow). (I, J) Double labeling of GFP (green) and CRALBP (red) in a transplant. Glial structures in the graft–host interface were disrupted (I, arrows), and integration was noted in the corresponding area (J, arrows).
Figure 2.
 
(A–H) Double labeling of GFP (green) and CRALBP (red). Glial structures in the interface of two pieces were preserved in the center of the overlapping area in a laminar pair (A, arrowheads) where integration was not seen (B). Most CRALBP-labeled processes at the interface also expressed GFP, although some processes labeled for CRALBP but not GFP were seen at the outer margin of the rd1 piece (B, arrows). When the glial structure in the interface was disrupted at the edges (C, arrow), integration was visible (D, arrow). Glial structures of GFP pieces in fragment–laminar pairs appeared disorganized (E). Yet, as long as they were preserved in rd1 pieces (E, arrowheads), integration was not visible (F). When the structures of both pieces were disrupted (in the central part in fragment–laminar pairs; G, arrow), integration occurred in the corresponding area (H, arrow). (I, J) Double labeling of GFP (green) and CRALBP (red) in a transplant. Glial structures in the graft–host interface were disrupted (I, arrows), and integration was noted in the corresponding area (J, arrows).
Figure 3.
 
(A–I) Triple labeling of GFP (green), CD44 (blue), and neurocan (red) in laminar pairs (A–E), fragment–laminar pairs (F, G), and transplants (H, I). (J, K) Double labeling of GFP (green) and CD44 (red) in transplants. (H, I) Animals that received transplants of P0 donor tissue, killed after 12 days; (J, K) animals that received transplants of P5 donor tissue, killed after (J) 7 and (K) 21 days. CD44 (A) and neurocan (B) expression appeared in a band and colocalized in the interface between the abutting retinas (C; D–G, arrowheads). Neurocan immunolabeling was also visible in the GFP piece in the IPL (B, C, E). Integration occurred only in areas where CD44/neurocan expression was not observed in the interface (D–K, arrows).
Figure 3.
 
(A–I) Triple labeling of GFP (green), CD44 (blue), and neurocan (red) in laminar pairs (A–E), fragment–laminar pairs (F, G), and transplants (H, I). (J, K) Double labeling of GFP (green) and CD44 (red) in transplants. (H, I) Animals that received transplants of P0 donor tissue, killed after 12 days; (J, K) animals that received transplants of P5 donor tissue, killed after (J) 7 and (K) 21 days. CD44 (A) and neurocan (B) expression appeared in a band and colocalized in the interface between the abutting retinas (C; D–G, arrowheads). Neurocan immunolabeling was also visible in the GFP piece in the IPL (B, C, E). Integration occurred only in areas where CD44/neurocan expression was not observed in the interface (D–K, arrows).
Figure 4.
 
(A-D) Double labeling of GFP (green) and neuronal markers (red), illustrating the presence of GFP+ cells and fibers within the INL of the rd1 retina that coexpressed NOS in a laminar pair (A) and in a fragment–laminar pair (B, large arrow). A few GFP+ cells are also noted in the GCL (small arrows), and some with a microglia-like morphology in the IPL (arrowheads). (C, D) GFP+ cells in the INL of the rd1 piece that coexpressed calbindin and calretinin, respectively, in fragment–laminar pairs (arrows). (E–H) Double labeling of GFP (green) and GFAP (red) in fragment–laminar pairs (E, F) and transplants (G, H) showing that integration (E–H, arrows) coincides with regions where hypertrophic Müller cells (E, F, arrows) and astrocytes (G-H, ★) appeared to be intensely GFAP positive. In transplants, the overall GFAP labeling in the host retina was lower in areas away from the graft (not shown).
Figure 4.
 
(A-D) Double labeling of GFP (green) and neuronal markers (red), illustrating the presence of GFP+ cells and fibers within the INL of the rd1 retina that coexpressed NOS in a laminar pair (A) and in a fragment–laminar pair (B, large arrow). A few GFP+ cells are also noted in the GCL (small arrows), and some with a microglia-like morphology in the IPL (arrowheads). (C, D) GFP+ cells in the INL of the rd1 piece that coexpressed calbindin and calretinin, respectively, in fragment–laminar pairs (arrows). (E–H) Double labeling of GFP (green) and GFAP (red) in fragment–laminar pairs (E, F) and transplants (G, H) showing that integration (E–H, arrows) coincides with regions where hypertrophic Müller cells (E, F, arrows) and astrocytes (G-H, ★) appeared to be intensely GFAP positive. In transplants, the overall GFAP labeling in the host retina was lower in areas away from the graft (not shown).
Table 1.
 
Primary Antibodies Used in Immunohistochemistry
Table 1.
 
Primary Antibodies Used in Immunohistochemistry
Antigen Localization Antiserum Dilution Source
Neuronal nitric oxide synthase Amacrine cells, (bipolar cells) 50 Sheep anti-nNOS 1:4000 I. G. Charles, P. C. Emson, Cambridge, UK
Protein kinase C Rod bipolar cells 51 Rabbit anti-PKC 1:1000 Chemicon, Temecula, CA
Calbindin-D (28 kDa) Horizontal cells, Amacrine cells 51 Mouse anti-calbindin 1:200 Sigma-Aldrich
Calretinin Amacrine cells 51 Mouse anti-calretinin 1:2000 Chemicon
Rhodopsin Rod photoreceptor cells 52 Mouse anti-Rho-1D4 1:400 R. S. Molday, Vancouver, BC, Canada
GFAP Müller cells, astrocytes 27 Rabbit anti-GFAP 1:1500 Dako, Glostrup, Denmark
Neurocan Extracellular matrix (neurons, glial cells) 36 37 Rabbit anti-NC2* 1:2000 U. Rauch, Lund, Sweden
CRALBP Müller cells, astrocytes 21 24 Rabbit anti-CRALBP 1:5000 J. Saari, Seattle, WA
CD44 glycoprotein Müller cells 34 35 39 Rat anti-CD44, † 1:100 BD Biosciences, San Diego, CA
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×