Amniotic membrane as support for human ... - Wiley Online Library

ACTA OPHTHALMOLOGICA SCANDINAVICA 2003

Amniotic membrane as support for human retinal pigment epithelium (RPE) cell growth Carmen Capeańs,1,2 Antonio Pin˜eiro,2 Marı´ a Pardo,2 Catalina Sueiro-Lo´pez,3 Marı´ a Jose´ Blanco,1 Fernando Domı´ nguez4 and Manuel Sańchez-Salorio2 1

Servicio de Oftalmologı´ a, Complejo Hospitalario Universitario de Santiago, Santiago de Compostela, Spain 2 Instituto Gallego de Oftalmologı´ a, Santiago de Compostela, Spain 3 Departamento de Biologı´ a Fundamental, Facultad de Biologı´ a, Universidad de Santiago de Compostela, Santiago de Compostela, Spain 4 Departamento de Fisiologı´ a, Facultad de Medicina, Universidad de Santiago de Compostela, Santiago de Compostela, Spain

ABSTRACT. Purpose: The aim of this work was to culture human retinal pigment epithelium (hRPE) cells over human amniotic membrane (hAM). Human AM was studied for its viability as an adequate support for transplantation of an hRPE cell monolayer with preserved cell polarity to the subretinal space. Methods: Human AM was obtained from pregnant women during caesarean section. The hAM was sectioned and the pieces were fixed to culture dishes. Human RPE cells were cultured from adult corneal donors and were seeded over hAM. Phase-contrast photographs were obtained. Selected specimens were processed by transmission electronic microscopy (TEM). Results: The attachment and growth of hRPE cells over hAM was observed. Human RPE cells constituted tight colonies that maintained epithelial phenotype. Using TEM, we identified a monolayer of hRPE cells, with cuboidal to spheroidal morphology. These cells showed integration with the substrate and cellcell contacts were detected. Conclusion: Amniotic membrane may be a suitable substrate for hRPE growth. Further studies are required in order to determine the viability of hRPE on hAM in the subretinal space. Key words: amniotic membrane – cell culture – retinal pigment epithelium – retinal transplantation

Acta Ophthalmol. Scand. 2003: 81: 271–277 Copyright # Acta Ophthalmol Scand 2003. ISSN 1395-3907

Introduction Retinal cell transplantation is an experimental therapeutic approach to certain degenerative diseases of the retina (i.e. age-related macular degeneration (AMD) and retinitis pigmen-

tosa). Different procedures have been studied in experimental and clinical studies, with varying degrees of success: implantation of fetal or adult human retinal epithelial cells into subretinal space (Algvere et al. 1994), and transplantation of autologous iris pigment

epithelial (IPE) cells (Crafoord et al. 2002) and cultured photoreceptors (Kaplan et al. 1997). Retinal pigment epithelium (RPE) transplantation has been shown to rescue photoreceptor cells in dystrophic rats (Lin et al. 1996). Retinal pigment epithelium cell transplantation has been attempted in patients with non-exudative and exudative AMD by means of several techniques (cell suspension, patch transplants) (Algvere et al. 1999). The main disadvantage of cell suspension transplantation is the random organization in multilayers of cells in the subretinal space of the host (Crafoord et al. 1999). Other complications have been described, such as subretinal fibrosis, invasion of the retina by pigmented cells and retinal pigment epithelial cells in the vitreous cavity causing proliferative vitreoretinopathy (Liu et al. 1992). In order to preserve cell orientation, some researchers have tried to maintain cell monolayers and polarity by means of different supports. These include biological supports such as Descemet’s membrane (Thumann et al. 1997), lens capsule (Hartmann et al. 1999), Bruch’s membrane or blood cryoprecipitates (Farrokh-Siar et al. 1999). Other groups have studied the application of

271

ACTA OPHTHALMOLOGICA SCANDINAVICA 2003 synthetic supports, such as collagen substrates (Bhatt et al. 1994), biodegradable polymer films (Lu et al. 1998) or microspheres (Oganesani et al. 1996). The human amniotic membrane (hAM) is a thin and elastic tissue that forms the inner layer of the amniotic sack. Human AM, with its thick basement membrane and avascular stromal matrix, has been successfully used for surface reconstruction in a variety of ocular surface disorders (severe pterigyum, chemical burn, ocular cicatricial pemphygoid and Stevens–Johnson syndrome) (Tseng et al. 1997). In these cases, hAM works as an optimal biological support for conjunctival cell growth. Recent publications show that hAM constitutes an adequate substrate for in vitro growth and expansion of conjunctival epithelial progenitor cells (Grueterich & Tseng 2002; Meller et al. 2002). Shimazaki et al. (2002) reported short-term clinical results after transplantation of human limbal epithelium cultivated on amniotic membrane for the treatment of severe ocular disorders. Rosenfeld et al. (1999) used a rabbit model to demonstrate that subretinal implantation of hAM appears to be well tolerated without evidence of inflammation. In view of all these considerations, the purpose of the present work was to study the amniotic matrix as a feasible substrate for the attachment and growth of hRPE cells.

Material and Methods Human amniotic membrane (hAM)

Two hAMs were obtained from two pregnant women during caesarean section after obtaining informed consent. Under sterile conditions, the hAM was processed as described previously by Tseng et al. (1997) for hAM transplantation to ocular surface disease. Briefly, the hAM was washed with Dulbecco’s Modification of Eagle’s Minimal Essential Medium (DMEM, Gibco BRL, Gaithersburg, Maryland, USA), supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin (ICN Biomedicals, Costa Mesa, California, USA) and 2.5 mg/ml amphotericin B (SigmaAldrich Quı´ mica, Madrid, Spain). In the present study the hAM was not

272

frozen, but was freshly used in all experiments. Human retinal pigment epithelial (hRPE) cultures

Human RPE cultures were established from adult human corneal donors to the Eye Bank of the Complejo Hospitalario Universitario de Santiago. One eye of each corneal donor (42 and 30year-old men) was utilized for hRPE culture. The culture technique followed that of previous reports (Capeans et al. 1998). Briefly, under sterile conditions, the anterior segment, vitreous body and retina were discarded. The eyecup was filled with DMEM supplemented with 10% fetal calf serum (FCS) (Gibco BRL, Gaithersburg, Maryland, USA) and 0.25% trypsin-0.02% EDTA (Sigma-Aldrich Quı´ mica, Madrid, Spain). It was then incubated at 37 in 5% CO2 humidified atmosphere for 45 mins. After this, hRPE cells were removed by gentle pippetting. The hRPE cells obtained were transferred to 60 mm culture dishes (Nunc, Roskilde, Denmark) in a complete medium composed of 10% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine (ICN Biomedicals, CA, USA) and 2.5 mg/ml amphotericin B in DMEM. In their initial passages, the cells had a polygonal pigmented morphology and they grew forming monolayers. To check that the cells had an epithelial origin, they were stained for cytokeratin (AE1/AE3) as previously described (Leschey et al. 1990). The hRPE cells used throughout this work were taken between the second and fifth passages and were always cultured under subconfluence conditions.

Afterward, the hAM pieces were washed in PBS. Thus, the epithelial hAM layer could be easily and gently removed. Phase-contrast microscopy controls were carried out in order to ensure that the hAM was free of epithelium. The hAM pieces were then sectioned into 2 2 cm fragments using a surgical blade. These fragments were attached to glass slides by means of cyanocrylate adhesive (Histoacryl, B Braun, Tuttlingen, Germany) with the epithelial basement membrane facing up. The metallic guide was removed from a 25-gauge Abbocath-1-T (Abbott Ireland Ltd, Dublin, Ireland) and the device was used to connect the plastic tube to a Histoacryl bottle in order to obtain a single, small drop of cyanocrylate. The adhesive was applied carefully, keeping the hAM as tense as possible. Once the adhesive was dry, the glass slides were placed into empty 90 mm culture dishes (Nunc, Roskilde, Denmark). Human RPE cells from subconfluent second to fifth passages were trypsinized and counted. A total of 1.5 106 cells were seeded by gentle pippetting over each piece of hAM attached to the glass slides. In the control cases the same number of NIH3T3 cells were seeded over hAM. Each culture was properly identified. In order to reach cellhAM attachment, culture dishes were carefully filled with complete medium 30 mins after cell seeding. The culture media was changed every 3 days and daily phase-contrast controls were performed. Microphotographs were taken with a phase-contrast microscope during the culture period. Some of these images were enlarged in the photography laboratory in order to improve the identification of details.

NIH3T3 fibroblasts cultures

NIH3T3 fibroblasts were cultured in complete medium composed by 10% FCS, 100 U/ml penicillin and 100 mg/ml streptomycin in DMEM on plastic dishes (Nunc, Roskilde, Denmark). Preparation of hAM with hRPE cells

In order to remove the attached amniotic epithelial cells, huge pieces of hAM were incubated in dispase II (1.2 U/ml Dispase II; Roche Diagnostics, Molecular Biochemicals, Barcelona, Spain) in Ca2þ Mg2þ free phosphate-buffered saline (PBS) for 15 mins, at room temperature under sterile conditions.

Transmission electronic microscopy

Selected specimens were processed for transmission electronic microscopy (TEM) after 30 days in culture. Human AM with hRPE or NIH3T3 fibroblasts was fixed with 2% glutaraldehyde (Sigma-Aldrich Quı´ mica, Madrid, Spain) in 0.1 M phosphate buffer pH 7.3 for 2 hours. After fixation, the hAM sections were postfixed in 1% OsO4 in 0.1 M phosphate buffer for 15 mins at room temperature, dehydrated in a graded series of alcohols and embedded in Spurr. Semi-thin sections (1 mm thick) were cut on a Reichert-Jung

ACTA OPHTHALMOLOGICA SCANDINAVICA 2003 ultramicrotome (Ultracut-E) and then stained with toluidine blue and studied under a light microscope. Finally, ultra thin sections (50–80 Zm thick) were stained with uranyl acetate and lead citrate for 5 mins and observed and photographed with a Philips CM-12 Transmission Electron Microscope (TEM) (SEI, Electron Optics, Eindhoren, the Netherlands). Some TEM images were enlarged in the laboratory in order to improve the identification of details.

Results Morphology and viability studies

Under phase-contrast microscopy, freshly obtained hAM showed an epithelial surface constituted by a cobblestone epithelium. Isolated pigmented epithelial cells were found in all studied specimens (Fig. 1A). Dispasetreated hAM showed its epithelial

basement membrane face after gentle scraping (Fig. 1B). In our study, the hAM thickness ranged between 90 and 190 mm. The hAM fixed by means of cyanoacrylate maintained its adherence to the glass slides for at least 4 weeks without detaching. Cell cultures over hAM

The human RPE cells cultured over dispase-treated hAM fragments were attached to the epithelial basement membrane side within 24 hours of seeding. After settling down, hRPE cells displayed their typical in vitro morphology. Human RPE cells were organized in tight colonies constituted by small, polygonal and uniform epithelial cells, after 3–5 days of culture (Fig. 1C, D). No differences were found in morphology, attachment and proliferation of hRPE cells from the two cultures utilized. Control NIH3T3 cells seeded over hAM took a similar period of time

until adhesion, showing a spindleshaped morphology and an irregular arrangement over hAM (data not shown). Transmission electronic microscopy

When hRPE cells were seeded over dispase-treated hAM, semi-thin sections demonstrated that, after 30 days of culture, the cells were organized in a tight monolayer of large cuboidal to round cells (Fig. 2). Consistent with this, the TEM study showed a tight monolayer of cuboidal to spheroidal hRPE cells growing over epithelium-free hAM (Figs 3 and 4). The hAM matrix was constituted by collagen spindles organized in different directions and embedded in an amorphous and clear ground substance (Fig. 3, black arrows). We found well-defined footplates (Fig. 3, FP), observed as elongations of hRPE cell basal membrane immersed in hAM. It was easy to find

Fig. 1. Phase-contrast microphotographs. (A) Cobblestone disposition of the amniotic epithelium in a freshly obtained hAM. Sporadic pigmented epithelial amniotic cells were found (white arrows). [Original magnification 40.] (B) After dispase treatment basement membrane was exposed. [Original magnification 40.] (C, D) After seeding and culture of hRPE cells over denuded hAM, tight colonies (black arrowheads) could be identified. In both cases, pictures were taken after 30 days in culture. [Original magnification: (C) 40; (D) 60.]

273


Fig. 2. Toluidine blue stained sections of: (A) hRPE and (B) NIH3T3 cells after 30 days in culture over hAM (AM). Black arrows point to seeded cells in both cases. Scale bars: (A) 208 mm; (B) 150 mm.

intercellular membranes running parallel from apical to basal poles (Fig. 3, black arrowheads). No desmosomes were detected; nevertheless, we found some areas characterized by apposition of electrodense membranes (Figs 3 and 4). Moreover, several focal contacts, separated by cleftlike spaces, were detected. Prominent and isolated buds (Fig. 3B) were identified at the apical side of hRPE cells. Cellular components such as mitochondrias and endoplasmic reticulum were present (Fig. 3, MI and R, respectively). Melanin granules (melanosomes), characteristic of these types of cells, were easily identified (Fig. 3, ME). Large-sized nuclei were balloon-shaped and were placed in the centre of the cuboidal cells (Figs 3 and 4, N). Chromatin was dispersed and sporadic nuclear pores were seen. Nevertheless, NIH3T3 cells, as seen under optical microscopy, were smaller than hRPE cells and showed an irregular arrangement (Fig. 5). By means of TEM, we proved that these cells grew in a multilayer disposition. Finally, no cellcell or cellhAM attachments could be seen in these cells.

Discussion Alteration in the RPE layer occurs in a broad spectrum of inherited and/or

274

acquired retinal degenerative disorders (loss of RPE cells results in photoreceptor dysfunction in non-exudative and exudative AMD). The surgical removal of RPE cells during excision of subretinal neovascularization leads to atrophy of the underlying choriocapillaris, limiting visual recovery. The spectrum of diseases that can potentially be treated by RPE transplantation has been expanded by findings that describe mutations in the RPE65 gene as the origin of autosomal recessive childhoodonset severe retinal dystrophy (Gu et al. 1997). The idea of replacing aged and/or diseased RPE with healthy RPE grafts has been extensively investigated in animal models and in humans (Algvere et al. 1994; Radtke et al. 2002; Tsukahara et al. 2002). Retinal pigment epithelium and iris pigment epithelial transplants have been shown to survive well in the subretinal space (Crafoord et al. 2002) and prevent photoreceptor degeneration to some extent (Lin et al. 1996). It seems that transplantation in the form of a patch is important in preventing cell disorganization, excessive mechanical stress or damage of the well oriented cells and in facilitating precise localization of the graft. Several support matrices have been utilized in different studies (Bhatt et al. 1994; Oganesani et al. 1996; Thumann et al. 1997; Lu et al. 1998; Farrokh-Siar et al. 1999;

Hartmann et al. 1999; Tezel & Del Priore 1999; Tsukahara et al. 2002). Most of these supports turned out either to be toxic to the cells or to disintegrate on exposure to liquid media. Human amniotic membrane consists of a single cell layer of epithelium bound to a continuous basement membrane, which is constituted of type IV collagen and laminine, and which interfaces with an avascular collagenous stroma composed of interstitial collagen and elastin. Human AM is an elastic and transparent tissue that is easy to obtain from pregnant women: there is normally a great amount of hAM available during a caesarean section. Procedures for handling and storing hAM have been well described by other groups (Tseng et al. 1997). It has been demonstrated that tracheal epithelium cultured on amniotic membrane showed attachment and differentiation (Noguchi et al. 1995). Recent findings indicate that human corneal epithelial cells can be successfully regenerated in vitro over amniotic membrane by a limbal epithelium cellsuspension culture system (Koizumi et al. 2002). In this paper, we propose the use of epithelium-free hAM as support for human RPE growth. Rosenfeld et al. (1999) showed that it is feasible to implant hAM into the rabbit subretinal space and showed how hAM can be well tolerated by the rabbit retina. Tsukahara et al. (2002) demonstrated that freshly harvested human RPE cells seeded over Bruch’s membrane explants show better attachment if the basement membrane is not debrided. These results suggest that RPE cell transplantation requires a basement membrane in order to ensure cell attachment. We have demonstrated that hRPE cells seeded over hAM showed adhesion to the hAM in about 24 hours. Moreover, hRPE cells maintain epithelial features (morphology, pigment) and can proliferate over epitheliumfree hAM. The hRPE cells growing over hAM were found to be highly organized. These cells constitute a tight monolayer with well defined cellcell and cell–substrate interactions. As a control for the hRPE cells, we used NIH3T3 transformed fibroblasts seeded over hAM. Conversely, these cells were randomly organized in multilayers without apparent cell–cell interactions.


Fig. 3. Transmission electron microphotograph (photomontage) of hRPE after 30 days of culture over epithelium-denuded hAM. Amniotic membrane matrix (AM) shows collagen spindles organized in different directions (black arrows) and embedded in an amorphous and clear fundamental substance. Human RPE cells were organized in a tight monolayer of large cuboidal to round cells. Note intercellular membranes running parallel from apical to basal poles (black arrowheads). Apposition of electrodense membranes is seen. We found well-defined footplates (FP), observed as elongations of hRPE cell basal membrane immersed in hAM. Isolated and prominent buds (B) were detected in the apical side. Cellular components mitochondria (MI) and endoplasmic reticulum (R) were found. Melanin granules (melanosomes), characteristic of this kind of cells, were easily identified (ME). Large-sized nuclei were balloon-shaped and were placed in the centre of the cuboidal cells (N). Chromatin was dispersed and sporadic nuclear pores were seen. AS: apical side; BS: basal side. Scale bar: 3 mm.

The development of cell polarity over hAM has been described previously by Noguchi et al. (1995), who worked with tracheal epithelium. Some authors have reported the existence of tight junctions between RPE cells after a period of culture over substrate (Hartmann et al. 1999). We believe these findings depend on the biological function assumed by cells over a new environment – note that the neurosensory retina, choriocapillaris and choroid are not present in in vitro systems. Although we found certain cell structural polarity in our model, we do not know whether hRPE, seeded over hAM or other supports, can organize itself into a physical and functional barrier. It will be necessary to carry out additional assays in order to study the degree of functional relationship between hRPE cells and hAM (i.e. rod outer segments degradation studies).

Several in vivo human studies have demonstrated that hAM is an optimal substrate for conjunctival growth in severe damaged ocular surfaces; it has been observed that hAM develops tight integration with superficial ocular tissues (conjunctiva, cornea), promoting re-epitelization. The mechanisms by which this phenomenon occurs are still unknown. However, it has been suggested that amniotic stroma can secrete growth factors or express adhesion molecules that can facilitate cell growth. We believe that these characteristics make hAM an adequate support to RPE growth and further transplantation. Thus, hAM could play an important role in maintaining the viability and integration of RPE cells into the subretinal space. To our knowledge, Rosenfeld et al. (1999) did not reduce the thickness of hAM and they showed that the pres-

ence of hAM did not induce visible signs of rejection or outer retinal disorganization. The hAM utilized in our work had variable thickness, ranging between approximately 90 mm and 190 mm, and we did not select hAM fragments according to thickness. Further experiments, carried out in animals, could require a diminution of hAM thickness in order to maintain the retinal architecture and promote RPE cell integration. We believe this could be reached by means of enzymatic digestion of the hAM matrix (i.e. collagenase). Despite the fact that the eye is immunologically privileged, immune rejection may be a barrier to successful retinal transplantation. Our work was developed in vitro. We believe that further experiments must focus on rejection of hAM implanted to subretinal space. For this, animal models

275


Fig. 4. (A) TEM microphotography shows an area of contact between hRPE cells after 30 days of culture over hAM. No hemidesmosomes could be detected, but several areas of membrane contact were found (arrowhead). N: nucleus. (B) TEM microphotography magnifies the membrane contacts shown in (A) (arrowheads). N: nucleus. (C) TEM microphotography shows multiple focal contacts between membranes of hRPE cells cultured over hAM (arrowheads). N: nucleus. (D) High magnification in TEM microphotography shows an intimate apposition between hRPE membranes and condensation of cytoplasmic and intermembranous spaces (black arrow). Scale bars: (A and C) 3 mm; (B and D) 1 mm.

would be needed to determine rejection levels (autologous versus allotransplants). One hypothetical idea arising from our work concerns the possibility

of amniotic membrane antigenic immunity. Our work suggests that epitheliumfree hAM might be an optimal sub-

strate for RPE transplantation to the subretinal space. Further experiments must be carried out to establish the degree of integration of this graft within the host retina (extensive animal studies) and analyse the molecular properties of RPE cells on this substrate.

Acknowledgements

Fig. 5. NIH3T3 cells as seen in optical microscopy were smaller than hRPE cells and showed an irregular arrangement. By means of TEM, we proved that these cells (F) grew in a multilayer disposition. No cellcell or cellhAM attachment could be seen. AM: hAM. Scale bar: 2 mm.

276

This work was supported by a grant by the Galician Education Department to Manuel Sańchez-Salorio (grant no. XUGA 90201B97). Authors’ thanks are due to Professor Toma´s Garcı´ a-Caballero (Departamento de Anatomı´ a Patolo´gica, Complejo Hospitalario Universitario de Santiago de Compostela) for his collaboration in immunocytochemical techniques for cytokeratin staining of hRPE cells. The authors also thank Professor Celina Rodicio Rodicio and Professor Isabel Rodrı´ guez-Moldes

ACTA OPHTHALMOLOGICA SCANDINAVICA 2003 (Departamento de Biologı´ a Fundamental, Universidad de Santiago de Compostela) for their collaboration in the selection and study of TEM images. Finally, the authors thank Miro Barreiro Pe´rez (Servicio de Microscopı´ a Electrońica, Universidad de Santiago de Compostela) for technical assistance and Andrew Day for help in the translation of the text.

References Algvere PV, Berglin L, Gouras P & Sheng Y (1994): Transplantation of fetal retinal pigment epithelium in age-related macular degeneration with subfoveal neovascularization. Graefes Arch Clin Exp Ophthalmol 232: 707–716. Algvere PV, Gouras P & Dafgard Kopp E (1999): Longterm outcome of RPE allografts in non-immunosuppressed patients with AMD. Eur J Ophthalmol 9: 217–230. Bhatt NS, Newsome DA, Fenech T, Hessburg TP, Diamond JG, Miceli MV, Kratz KE & Oliver PD (1994): Experimental transplantation of human retinal pigment epithelial cell on collagen substrates. Am J Ophthalmol 117: 214–221. Capeans C, Pin˜eiro A, Domı´ nguez F, Loidi L, Buceta M, Carneiro C, Garcı´ a-Caballero T & Sańchez-Salorio M (1998): A c-myc antisense oligonucleotide inhibits human retinal pigment epithelial cell proliferation. Exp Eye Res 66: 581–589. Crafoord S, Algvere PV, Seregard S & Kopp ED (1999): Longterm outcome of RPE allografts to the subretinal space of rabbits. Acta Ophthalmol Scand 77: 247–254. Crafoord S, Geng L, Seregard S & Algvere P (2002): Photoreceptor survival in transplantation of autologous iris pigment epithelial cells to the subretinal space. Acta Ophthalmol Scand 80: 387–394. Farrokh-Siar L, Rezai KA, Patel SC & Ernest JT (1999): Cryoprecipitate: an autologous substrate for human fetal retinal pigment epithelium. Curr Eye Res 19: 89–94. Grueterich M & Tseng SC (2002): Human limbal progenitor cells expanded on intact amniotic membrane ex vivo. Arch Ophthalmol 120: 783–790.

Gu SM, Thompson DA, Srikumari CRS et al. (1997): Mutations in RPE65 cause autosomal recessive childhood-onset severe retinal dystrophy. Nat Genet 17: 194–197. Hartmann U, Sistani F & Steinhorst UH (1999): Human and porcine anterior lens capsule as support for growing and grafting retinal pigment epithelium and iris pigment epithelium. Graefes Arch Clin Exp Ophthalmol 237: 940–945. Kaplan HJ, Tezel TH, Berger AS, Wolf ML & del Priore LV (1997): Human photoreceptor transplantation in retinitis pigmentosa. Arch Ophthalmol 115: 1168–1172. Koizumi N, Copper LJ, Fullwood NJ, Nakamura T, Inoki K, Tsuzuki M & Kinoshita S (2002): An evaluation of cultivated corneal limbal epithelial cells, using cellsuspension culture. Invest Ophthalmol Vis Sci 43: 2114–2121. Leschey KH, Hackett SF, Singer JH & Campochiaro PA (1990): Growth factor responsiveness of human retinal pigment epithelial cells. Invest Ophthalmol Vis Sci 31: 839–846. Lin N, Fan W, Sheedlo HJ, Aschenbrenner JE & Turner JE (1996): Photoreceptor repair in response to RPE transplants in RCS rats: outer segment regeneration. Curr Eye Res 15: 1069–1077. Liu Y, Silverman MS, Berger AS & Kaplan HJ (1992): Transplantation of confluent sheets of adult human RPE. Invest Ophthalmol Vis Sci 33: 2180. Lu L, Garcia CA & Mikos AG (1998): Retinal pigment epithelium cell culture on thin biodegradable poly (DL-lactic-co-glycolic acid) films. J Biomater Sci Polymer Edn 9: 1187–1205. Meller D, Dabul V & Tseng SC (2002): Expansion of conjunctival epithelial progenitor cells on amniotic membrane. Exp Eye Res 74: 537–545. Noguchi Y, Uchida Y, Endo T et al. (1995): The induction of cell differentiation and polarity of tracheal epithelium cultured on the amniotic membrane. Biochem Biophys Res Commun 210: 302–309. Oganesani A, Gabrielian K & Verp MS (1996): A new model of RPE transplantation with microspheres. Invest Ophthalmol Vis Sci 37: 115. Radtke ND, Seiler MJ, Aramant RB, Petry HM & Pidwell DJ (2002): Transplantation of intact sheets of fetal neural retina with

its retinal pigment epithelium in retinitis pigmentosa patients. Am J Ophthalmol 133: 544–550. Rosenfeld PJ, Merritt J, Hernandez E, Meller D, Rosa RH Jr & Tseng SCG (1999): Subretinal implantation of human amniotic membrane: a rabbit model for the replacement of Bruch’s membrane during submacular surgery. Invest Ophthalmol Vis Sci 40: 206. Shimazaki J, Aiba M, Goto E, Kato N, Shimmura S & Tsubota K (2002): Transplantation of human limbal epithelium cultivated on amniotic membrane for the treatment of severe ocular surface disorders. Ophthalmology 109: 1285–1290. Tezel TH & del Priore LV (1999): Repopulation of different layers of host human Bruch’s membrane by retinal pigment epithelial cell grafts. Invest Ophthalmol Vis Sci 40: 767–774. Thumann G, Schraermeyer U, Bartz-Schimdt KU & Heimann K (1997): Descemet’s membrane as membranous support in RPE/ IPE transplantation. Curr Eye Res 16: 1236–1238. Tseng SC, Prabahasawat P & Lee SH (1997): Amniotic membrane transplantation for conjunctival surface reconstruction. Am J Ophthalmol 124: 765–774. Tsukahara I, Ninomiya S, Castellarin A, Yagi F, Sugino IK & Zarbin MA (2002): Early attachment of uncultured retinal pigment epithelium from aged donors onto Bruch’s membrane explants. Exp Eye Res 74: 255–266.

Received on 11 September, 2002. Accepted on 13 February, 2003. Correspondence: Dr Antonio Pin˜eiro Ces Instituto Gallego de Oftalmologia Hospital de Conxo Ruá Ramoń Baltar s/n 15706 Santiago de Compostela Spain Tel: þ 34 981 534 501 Fax: þ 34 981 531 627 Email: [email protected]

277