A Method to Enhance Cell Survival on Bruch's Membrane in Eyes Affected by Age and Age-Related Macular Degeneration

Abstract

Purpose.: To determine whether conditioned medium (CM) derived from bovine corneal endothelial cells (BCECs) can support transplanted cells on aged and age-related macular degeneration (AMD) Bruch's membrane (BM).

Methods.: Retinal pigment epithelium (RPE) cells derived from human embryonic stem cells (hES-RPE) and cultured fetal and aged adult RPE were seeded onto the inner collagenous layer of submacular BM-choroid-sclera explants generated from aged and AMD human donor eyes. Paired explants were cultured in BCEC-CM or CM vehicle. To assess cell behavior after attachment to BM was established, explants were harvested after 21 days in culture. To assess whether sustained exposure to BCEC-CM was necessary for improved cell survival on BM, short exposure to BCEC-CM (3, 7, 14 days) was compared with 21-day exposure. Explants were harvested and evaluated by scanning electron and light microscopy. Extracellular matrix (ECM) deposition after exposure to BCEC-CM was evaluated following RPE cell removal after day 21 on tissue culture dishes or on BM.

Results.: BCEC-CM significantly enhanced hES-RPE, fetal RPE, and aged adult RPE survival on BM, regardless of submacular pathology. Although shorter BCEC-CM exposure times showed significant improvement in cell survival compared with culture in CM vehicle, longer BCEC-CM exposure times were more effective. BCEC-CM increased RPE ECM deposition on tissue culture plastic and on BM.

Conclusions.: The results of this study indicate that RPE survival is possible on AMD BM and offer a method that could be developed for enhancing transplanted cell survival on AMD BM. Increased ECM deposition may account for improved cell survival after culture in BCEC-CM.

Currently, no proven treatment options exist for patients with geographic atrophy, an advanced form of age-related macular degeneration (AMD). ¹ For selected patients with extensive drusen or geographic atrophy threatening the fovea, cell transplants might prevent central vision loss through replacement of dysfunctional or dead retinal pigment epithelium (RPE) cells. Antivascular endothelial growth factor therapy is the best treatment available for AMD-associated choroidal neovascularization (CNV), but randomized studies indicate that only 25% to 40% of treated patients experience at least moderate visual improvement. ^{2 –5} Thus, even today, a significant number of patients become blind despite the availability of pathway-based therapy for AMD-associated CNV. If cell transplants could prevent CNV development or rescue photoreceptors after CNV excision, these transplants also might have an impact on CNV-related blindness. ^6,7  

A major obstacle to the success of RPE transplants in AMD patients is the failure of transplanted RPE cells to survive and become functional in the diseased AMD eye. RPE transplantation in patients with AMD (atrophic and neovascular) typically has produced limited visual recovery regardless of the type of cell transplanted (e.g., autologous or allogeneic, adult or fetal RPE) or whether the cells are transplanted with or without choroid. ^{8 –10} In contrast, RPE transplantation in animal models of retinal degeneration has been proved to rescue photoreceptors and to preserve visual acuity. ^{11 –16} Although animal studies validate cell transplantation as a means of achieving photoreceptor rescue, important distinctions between humans with AMD and laboratory animals in which RPE transplantation has been successful include the age- and AMD-related modifications of the surface on which human RPE cells reside in situ (i.e., Bruch's membrane), which may have a significant effect on RPE graft survival. Evidence from human donor eye organ culture experiments indicates that healthy RPE cells cannot survive for an extended time on aged submacular Bruch's membrane, and the poorest survival is observed on AMD Bruch's membrane. ^17,18 These in vitro studies were performed on human submacular Bruch's membrane with no treatment to improve cell survival. Previous studies to improve cell survival on aged Bruch's membrane included adding extracellular matrix (ECM) ligands singly or in combination to “coat” Bruch's membrane, ¹⁹ detergent treatment to eliminate debris accumulated within Bruch's membrane followed by ECM ligand coating, ²⁰ and resurfacing Bruch's membrane with a cell-deposited matrix. ²¹ The first two methods showed limited improvement in attachment and early survival. Long-term survival was not demonstrated. The last method improved long-term cell survival more than 200%. However, from a therapeutic standpoint, resurfacing submacular Bruch's membrane with the cell-deposited ECM was problematic because of the inability to solubilize ECM components in a manner compatible with clinical application. ²¹ These studies demonstrate the need to develop a method to improve long-term cell transplant survival in AMD patients. 

Bovine corneal endothelial cells (BCECs) secrete an ECM that supports rapid attachment, growth, and differentiation of RPE cells. ²² During BCEC-ECM formation, in addition to basal secretion, BCECs secrete ECM components into the overlying medium, including collagens, ²³ proteoglycans, ²⁴ and entactin/nidogen. ²⁵ Secretion of ECM components into the overlying medium is most abundant in early-passage cells and exceeds basal ECM deposition in quantity. ²³ Given that soluble ECM can affect cell shape and metabolism and can stimulate the production of ECM molecules, ²⁶ the presence of these proteins suggests that conditioned medium (CM) harvested from BCEC cultures (BCEC-CM) could be a source of cell-supporting soluble proteins and, if effective, could lead to the development of an adjunct to cell-based therapy for AMD. Here, we characterize the behavior of RPE cells transplanted onto Bruch's membrane of aged and AMD donor eyes cultured in BCEC-CM or CM vehicle using a previously characterized human submacular Bruch's membrane bioassay. ^17,18,27,28  

Materials and Methods

Conditioned Medium Preparation

Cow eyes (age range, 6 months-3 years) were obtained from local slaughterhouses. Each globe was rinsed briefly in 70% ethanol, and the cornea was separated from the rest of the globe by making a circumferential cut anterior to the limbus. The cornea was rinsed quickly in PBS and positioned with the epithelial surface down on a sterile support placed on a Petri dish. The cup formed by the cornea was filled with 0.05% trypsin-0.02% EDTA (Invitrogen-Gibco, Life Technologies, Carlsbad, CA) and was placed in a 37°C, 10% CO₂ incubator for 30 to 60 minutes. BCECs were scraped off gently using a blunt metal spatula and were collected into a 15-mL tube containing Dulbecco's modified Eagle's medium (Cellgro, Manassas, VA) supplemented with 2 mM glutamine, 15% fetal bovine serum (FBS), 2.5 μg/mL amphotericin B, 50 μg/mL gentamicin, and 1 ng/mL bFGF (all from Invitrogen-Gibco) (hereafter referred to as RPE medium). ²² Cells were spun down, resuspended in RPE medium, seeded onto 60-mm dishes, and cultured at 37°C in 10% CO₂. Cultures were passaged at confluence. For BCEC-CM harvest, passage 2 or passage 4 cells were cultured in RPE medium with 10% FBS and 5% donor bovine serum (Invitrogen-Gibco) instead of 15% FBS until they were confluent. BCEC-CM was obtained by incubating confluent BCEC cultures for 72 hours in Madin-Darby Bovine Kidney Maintenance Medium (MDBK-MM; Sigma-Aldrich, St. Louis, MO) supplemented with 2.5 μg/mL amphotericin B and 50 μg/mL gentamicin. The vehicle, MDBK-MM (hereafter referred to as CM vehicle), is a serum- and protein-free, defined medium designed for maintaining high-density cultures of MDBK cells. After collection, BCEC-CM was centrifuged briefly to remove cellular debris, and the supernatant was stored at −80°C. Twelve batches of BCEC-CM were used. 

Cell Culture

RPE cells were isolated from fetal eyes (Advanced Bioscience Resources, Inc., Alameda, CA; gestational age, 18–22 weeks) or adult eyes (donor ages, 58, 71, 78 years) after incubation of RPE/choroid pieces in 0.8 mg/mL (fetal eyes) or 0.4 mg/mL collagenase type IV (Sigma-Aldrich) (adult eyes), as described previously. ^17,28,29 RPE cells were cultured in RPE medium on BCEC-ECM–coated tissue culture dishes prepared in this laboratory according to a previously described protocol. ²² After achieving confluence, primary fetal RPE cultures were passaged at a 1:6 split ratio onto BCEC-ECM–coated dishes using 0.25% trypsin-EDTA to harvest the cells. Subsequent cultures were passaged at a 1:4 split ratio. Adult RPE cells seeded onto Bruch's membrane were from day 11 to day 15 primary cultures; fetal RPE cells were harvested from cultures (passages 1–3) 3 to 7 days after seeding. 

Human embryonic stem cell-derived RPE cells (hES-RPE), provided by Irina Klimanskaya of Advanced Cell Technology (Worcester, MA), were established from a stem cell culture designated as MA09. ³⁰ Cells were maintained in MDBK-MM (Sigma-Aldrich) until removal from flasks and seeding onto Bruch's membrane explants. Cells of passage 32 (stem cells) and passage 2 or 3 (hES-RPE) were used and were removed from culture dishes with trypsin/EDTA after 50 to 85 days in culture. 

Bruch's Membrane Organ Culture

Adult donor eyes were received from the Lions Eye Institute for Transplant and Research (Tampa, FL) and eye banks placing donor eyes through their Web site (Ocular Research Biologics System, http://www.orbsproject.org), Midwest Eye-Banks (includes eye banks in Illinois, Michigan, and New Jersey), the San Diego Eye Bank (San Diego, CA), and eye banks placing tissue through the National Disease Research Interchange (Philadelphia, PA). Acceptance criteria for donor eyes included death to enucleation time no more than 7 hours, death to receipt time no more than 48 hours, no ventilator support before death, no chemotherapy within the last 6 months before death, no radiation to the head within the last 6 months before death, no recent head trauma, and no ocular history affecting the posterior segment except for AMD. These acceptance criteria have been found in previous studies to yield well-preserved explants. ^18,21 Posterior segments were examined through a dissecting microscope for submacular pathology and documented by photography. A previously published method was used to create inner collagenous layer (ICL) surfaces by mechanical debridement. ^{17,18,21,27,28,31} Six-millimeter-diameter corneal trephines (Bausch and Lomb, Rochester, NY) were used to create macula-centered Bruch's membrane explants. Explants were placed in wells of 96-well plates for cell seeding and organ culture. Cells were seeded at a seeding density of 3164 cells/mm², a seeding density that has been shown to yield a monolayer of cells on a 6-mm-diameter Bruch's membrane explant in organ culture one day after seeding. ²⁷ Explants were harvested at day 21, fixed in phosphate-buffered 2% paraformaldehyde and 2.5% glutaraldehyde, bisected, and processed for light microscopy (LM) or scanning electron microscopy (SEM). 

Scanning Electron Microscopy

Explant halves for SEM were postfixed in phosphate-buffered osmium tetroxide, dehydrated using a graded series of ethanol, critical point dried (Tousimis, Rockville, MD), and sputter-coated (Denton, Moorestown, NJ) according to standard SEM protocols. SEM image acquisition (JSM 6510; JEOL, Tokyo, Japan) was performed with routine photography at 30×, 50×, 200×, and 1000×. SEM evaluation of Bruch's membrane involved assessment of the cell surface morphology and, in areas not resurfaced by cells, the level of Bruch's membrane exposed by debridement. 

Light Microscopy

Bruch's membrane explant halves processed for histology were embedded in resin (LR White; Electron Microscopy Sciences, Hatfield, PA); four to six sections of 2-μm thickness were mounted on slides, dried overnight, and stained with 0.03% toluidine blue (Electron Microscopy Sciences). LM evaluation focused on RPE morphology (cell shape, density, pigmentation, polarization) and evaluation of Bruch's membrane and choroid. Nuclear density counts were performed to assess treatment success quantitatively comparing paired explants from fellow eyes. Nuclear density counts were performed by counting the number of RPE nuclei in intact cells in contact with Bruch's membrane in the central 3 mm of four to five nonconsecutive slides (approximately every fifth slide). ¹⁷ Linear measurements of Bruch's membrane in the analyzed area were obtained by digital image acquisition and measurement with the freehand line tool using ImageJ software (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html). Nuclear density was expressed as the number of nuclei per mm of Bruch's membrane. 

Statistical Analysis

Statistical differences between pairs were determined by Wilcoxon signed-rank tests. For comparisons between time points and comparisons between groups, the existence of significant differences was determined by Kruskal-Wallis one-way analysis of variance (ANOVA) on ranks. If significance was observed, all pairwise multiple-comparison procedures testing (Dunn's method) determined the significance between pairs of groups. Comparisons between two groups in unpaired studies were by Mann-Whitney rank sum tests. Comparisons between ages of two groups were by unpaired t-tests or between multiple groups by one-way ANOVA. Ages are indicated as mean age with SD. P < 0.05 was considered statistically significant. 

Extracellular Matrix Deposition

Fetal RPE (3164 cells/mm²) were seeded onto tissue culture-treated plastic (48-well plates) or Bruch's membrane and cultured in BCEC-CM or RPE medium. ECM on tissue culture plates was analyzed at days 7, 14, and 21 (n = 3). ECM on Bruch's membrane (six donor pairs; three pairs with extensive drusen; three pairs normal; mean donor age, 79.2 ± 3.17 years) was analyzed at day 21 only. Primary antibodies were mouse monoclonal collagen IV (1:500 dilution; Sigma-Aldrich), rabbit polyclonal laminin (1:25 dilution; Sigma-Aldrich), and mouse monoclonal fibronectin (1:50 dilution; Abcam, Cambridge, MA). Secondary antibodies were FITC-conjugated goat anti-mouse IgG (H+L) and tetramethyl rhodamine isothiocyanate conjugated goat anti-rabbit IgG (H+L) applied at 1:50 dilution (both from Jackson ImmunoResearch Laboratories, West Grove, PA). All antibodies were diluted in 2% normal goat serum, 0.3% Triton X-100 (both from Sigma-Aldrich) in PBS. 

Cell Culture.

Cells were removed from culture wells by incubation in 0.02 M NH₄OH for 5 minutes, followed by rinsing in PBS. The exposed ECM was fixed for 15 minutes in cold 4% paraformaldehyde followed by three PBS washes. Wells were then incubated for 45 minutes at room temperature in blocking solution (2% normal goat serum, 0.5% BSA in PBS). Primary antibodies were applied to culture wells and incubated for 2 hours at room temperature. After washing with PBS plus 0.3% triton, secondary antibodies were applied, and wells were incubated for 1 hour at room temperature. After washing with PBS plus triton, mounting medium (Vectashield; Vector Laboratories, Burlingame, CA) was added to the wells. Epifluorescence images for each protein at the same time point were photographed at the same exposure to determine relative differences in the amount of deposited protein using an inverted microscope equipped with the appropriate fluorescein and rhodamine filters (10× Neofluar objective; Axiovert, Carl Zeiss, Thornwood, NY). After immunostaining photography, ECM was stained with 0.1% Ponceau S (Sigma-Aldrich) for 10 minutes at room temperature and photographed using a 32× phase objective. 

Bruch's Membrane.

At day 21, live RPE cells were imaged by calcein staining (Live/Dead viability/toxicity assay; Molecular Probes, Eugene, OR) to determine surface coverage by cells. After a brief rinse in PBS, explants were incubated in 2 μM calcein for 1 hour at room temperature. Explants were rinsed briefly, then photographed at 2.5× magnification using a fluorescence microscope equipped with a fluorescein filter set (Axiophot; Carl Zeiss). Montages of calcein-imaged explants were created in an image editing program (Photoshop CS4; Adobe Systems, Mountain View, CA). After calcein imaging and RPE removal by 5-minute incubation in 0.02 M NH₄OH, explants were fixed in 4% paraformaldehyde for 1 hour at 4°C. After washing in PBS, explants were blocked at room temperature for 45 minutes. The explants were then bisected before immunostaining, and the surface of Bruch's membrane was immunostained for laminin and collagen IV (one half) and fibronectin and laminin (other half) or cut into thirds, with the third piece used for controls. Primary antibodies were applied to explants, which were then incubated overnight at 4°C. The following day, explants were rinsed with PBS, and secondary antibodies were applied. Explants were incubated for 2 hours at room temperature, followed by washing with PBS. Explants were stored and examined in mounting medium. Single images or z-stacks were acquired from the surface of Bruch's membrane using a 40× water-immersion lens on a confocal microscope (LSM510; Carl Zeiss). Laser lines and corresponding emission filters were 488-nm excitation with 505- to 530-nm bandpass filter for FITC and 543-nm excitation with 560- to 615-nm bandpass filter for rhodamine. After confocal microscopy evaluation, explants were processed for SEM. 

Results

Effect of BCEC-CM on Long-Term Cell Survival on Aged and AMD Bruch's Membrane

RPE derived from human embryonic stem cells (hES-RPE), fetal RPE, and aged adult RPE were seeded onto the inner collagenous layer of submacular Bruch's membrane of eyes of donors aged 62 and older (see Supplementary Table S1, for donor information) and were cultured in BCEC-CM or CM vehicle (representative images, Figs. 1 2 3 4–5). On explants cultured in CM vehicle, limited resurfacing was seen at day 21 with no or few surviving cells on most explants (hES-RPE, 3 of 6 total explants resurfaced with few cells and no cells on the remaining 3 of 6 explants; fetal RPE cells, 20 of 22 explants with no or few cells; adult RPE cells, 5 of 7 explants with no or few single cells) (Figs. 1A–C to 5A–C). When present, intact cells were large and flat, regardless of cell type (Figs. 2C, 5B). Cytoplasmic vacuoles were common. All explants cultured in BCEC-CM showed cells remaining at day 21, with many explant surfaces almost fully (>75%) covered by cells or fully resurfaced (hES-RPE, 3 of 6 explants completely or almost fully resurfaced; fetal RPE cells, 16 of 22 explants completely or almost fully resurfaced; adult RPE cells, 6 of 7 explants almost fully resurfaced). hES-RPE (Figs. 1E–G) were predominantly monolayered with highly variable morphology and were larger and flatter than fetal RPE cells (Figs. 2E–G to 4E–G). Fetal RPE cells showed focal areas of bilayers and multilayers, with the most extensive multilayering found in explants that underwent CNV removal before cell seeding (Fig. 3E). Resurfacing was extensive, and many cells were compact with abundant expression of well-developed surface apical processes regardless of submacular pathology (Figs. 2E, 3E, 4E, insets). Adult RPE cells, generally larger than fetal RPE cells, were predominantly monolayered with localized multilayered clumps of cells (Figs. 5E–G). Adult RPE cells exhibited abundant short apical processes (Fig. 5E). Because adult RPE cells were from primary cultures, the cells were generally more pigmented than hES-RPE or fetal RPE cells (Fig. 5G). 

Figure 1.

 

Paired submacular explants from a 74-year-old woman with soft drusen, seeded with hES-RPE. In CM vehicle, (A) postmortem clinical photograph shows soft drusen (arrow) in the macula. Inset is a higher magnification image of the area indicated by the arrow. The drusen are not easily visualized in this photomicrograph because of postmortem changes. (B, C) No intact cells are seen on the cultured explant. In BCEC-CM, (D) arrow points to a patch of confluent soft drusen in the macula of the fellow eye, shown in the high-magnification inset. (E) Cells almost fully resurface the explant with small defects in coverage. Cells are variable in size and shape. (inset) Cells are generally flat, with most exhibiting short processes on their surfaces. (F, G) Very flat, elongated cells resurface the explant in a monolayer. (G) Arrowhead points to cell-containing vesicles. CM vehicle nuclear density (ND), 0; BCEC-CM ND, 19.90 ± 0.35. Scale bars: 100 μm (E); 20 μm (E, inset); 50 μm (F); 20 μm (G). Toluidine blue staining.

Figure 2.

 

Paired explants from an 81-year-old man with no submacular pathology, seeded with fetal RPE cells. (A, D) No submacular pathology is seen in the postmortem clinical photographs. In CM vehicle, (B) cellular debris, but not intact cells, is seen on the surface of Bruch's membrane. Few remaining patches of RPE basement membrane (arrows, inset) are present. (C) Rare single cells are seen on the explant surface. Arrow points to a single, very flat cell. In BCEC-CM, (E) the explant is almost fully resurfaced with small defects in cell coverage (arrows). Patches of small, rounded cells are interspersed with localized areas in which cells are more variable in size and shape. Cells express abundant short apical processes on their surfaces (inset). (F, G) The explant is resurfaced by a monolayer and a bilayer of cells. (G) Arrowhead points to a cell overlying a cell on Bruch's membrane. CM vehicle ND, 0.51 ± 0.16; BCEC-CM ND, 26.8 ± 0.41. Scale bars: 100 μm (E); 20 μm (E, inset); 50 μm (F); 20 μm (G). Toluidine blue staining.

Figure 3.

 

Paired explants from a 75-year-old woman with large bilateral subfoveal CNV seeded with fetal RPE cells after mechanical CNV removal. (A, D) Arrows point to CNV in postmortem clinical photographs; insets show CNV after surgical dissection from Bruch's membrane. In CM vehicle, (A) the CNV measured approximately 4.8 × 3 mm. (B, C) No intact cells are seen on the explant surface. In BCEC-CM, (D) the CNV measured approximately 4 × 3.5 mm. (E) Fetal RPE cells fully resurface the explant, with some areas of thick multilayers (arrowhead). The cell surfaces are covered with apical processes (inset). (F) Cells resurfacing the explant are predominantly monolayered with localized areas in which thin or spindle-shaped cells overlay cells on Bruch's membrane. The cells resurfacing the explant are more variable in size and shape than those observed on explants from donors with geographic atrophy. (G) Fetal RPE cells are able to resurface small drusen (arrow points to druse on Bruch's membrane) and basal laminar deposits (asterisk). Arrowhead points to a cell with a darkly staining, irregularly shaped nucleus. CM vehicle ND, 0; BCEC-CM ND, 25.10 ± 0.30. Scale bars: 100 μm (E); 20 μm (E, inset); 50 μm (F); 20 μm (G). Toluidine blue staining.

Figure 4.

 

Paired explants from an 82-year-old woman with geographic atrophy, seeded with fetal RPE cells. The patient's clinical history noted AMD for 20 years. (A, D) Postmortem clinical photographs showing subfoveal geographic atrophy before RPE cell seeding. In CM vehicle, (B) only a few dead cells (arrows) and cellular debris are present on the explant surface. (C) No cells are present on Bruch's membrane surface. In BCEC-CM, (E) RPE cells fully resurface Bruch's membrane in the area of geographic atrophy with a few very small defects (arrows). Localized areas of multilayering are present. Cell surfaces show abundant apical processes (inset). (F) In this field, cells resurfacing the BCEC-CM explant are predominantly bilayered. Cells directly on Bruch's membrane are small and tightly packed; flat cells appear to overlie the cells in contact with Bruch's membrane. (G) Flattened cell processes overlying cells on top of Bruch's membrane are indicated by arrowheads. The cell processes contain vesicles. CM vehicle ND, 0; BCEC-CM ND, 19.61 ± 0.43. Scale bars: 100 μm (E); 20 μm (E, inset); 50 μm (F); 20 μm (G). Toluidine blue staining.

Figure 5.

 

Paired explants from an 80-year-old man with intermediate-size drusen in the CM vehicle-cultured eye and intermediate and large drusen in the BCEC-CM-cultured eye, seeded with cultured adult RPE cells (isolated from a 70-year-old donor). In CM vehicle, (A) two drusen (closely clustered) were present in the macula (arrow and arrowhead, high-magnification inset). The drusen are not easily visualized in these photomicrographs because of postmortem changes. (B) Very few large cells are observed on the explant surface (six cells in this image field). Arrow points to a pair of very large, flat cells. (C) No cells are present on the surface of Bruch's membrane. In BCEC-CM, (D) a cluster of drusen (arrow and high-magnification inset) can be seen in the macula of the fellow eye. (E) The explant is fully resurfaced by a monolayer of cells that are highly variable in size. Some of the cells within the monolayer do not have intact cell membranes (cells that appear white in the low-magnification image), and some cells died with remnants of cellular debris (arrows). The high-magnification inset shows that most of the cells are covered with short apical processes, including cells that are very large (fetal RPE cells of this size on submacular Bruch's membrane generally have smooth surfaces with no apical processes). One cell in the field exhibits surface blebs. (F, G) Cells resurfacing the explant are generally large and often pigmented. Localized areas of bilayering are present (F, arrowheads). CM vehicle ND, 0.14 ± 0.14; BCEC-CM ND, 12.0 ± 0.77. Scale bars: 100 μm (E); 20 μm (E, inset); 50 μm (F); 20 μm (G). Toluidine blue staining.

The nuclear density of hES-RPE grown in BCEC-CM (mean, 20.1 nuclei/mm Bruch's membrane) was significantly higher than that of cells grown in CM vehicle (mean, 5.1 nuclei/mm Bruch's membrane) (P = 0.031). Fetal RPE cell nuclear density after culture in BCEC-CM (mean, 20.6 nuclei/mm Bruch's membrane) was significantly higher than that of cells cultured in CM vehicle (mean, 2.2 nuclei/mm Bruch's membrane) (P < 0.001). The nuclear density of adult RPE cells cultured in BCEC-CM (mean, 10.0 nuclei/mm Bruch's membrane) was also significantly higher than the nuclear density of cells cultured in CM vehicle (mean, 1.2 nuclei/mm Bruch's membrane) (P = 0.016) (see Fig. 6A). ANOVA on ranks showed significant differences in the median values among the cell types cultured in BCEC-CM (P = 0.004). The nuclear densities of fetal RPE and hES-RPE cultured in BCEC-CM were significantly higher than the nuclear density of adult RPE cultured in BCEC-CM (P < 0.05). Fetal RPE and hES-RPE nuclear densities were not significantly different when cultured in BCEC-CM (P > 0.05). Nuclear densities of hES-RPE, fetal RPE, and cultured adult RPE were not significantly different when cultured in CM vehicle only (P = 0.060). There were no statistically significant differences in ages of donor eye explants between groups (P = 0.345: hES-RPE mean donor explant age, 80.2 ± 8.4 years; fetal RPE mean donor explant age, 80.2 ± 7.8 years; aged adult RPE mean donor explant age, 75.7 ± 3.6 years). We assessed whether RPE survival on age-matched AMD versus non-AMD Bruch's membrane was similar in the presence of BCEC-CM (Fig. 6B). Explants seeded with fetal RPE on aged Bruch's membrane from eyes without significant AMD changes (including donor eyes with submacular focal RPE hyperplasia and few small [<100 μm] drusen) were compared with explants seeded on AMD Bruch's membrane. AMD donor eyes included those with CNV (removed before cell seeding with no subsequent debridement), geographic atrophy, and extensive large (≥125 μm) drusen. After culture in BCEC-CM, non-AMD donor eye mean nuclear density was 19.6 nuclei/mm Bruch's membrane, and AMD donor eye mean nuclear density was 21.3 nuclei/mm Bruch's membrane. After culture in CM vehicle, non-AMD donor eye mean nuclear density was 2.4 nuclei/mm Bruch's membrane, and AMD donor eye mean nuclear density was 2.0 nuclei/mm Bruch's membrane. Differences in fetal RPE nuclear density on Bruch's membrane in the presence of BCEC-CM versus CM vehicle were statistically significant for both non-AMD and AMD donors (non-AMD, P = 0.004; AMD, P < 0.001). For a given culture medium, the nuclear densities of fetal RPE on non-AMD versus AMD explants were not significantly different (culture in BCEC-CM, P = 0.548; culture in CM vehicle, P = 0.231). Ages of the two groups were not statistically different (P = 0.226: aged, non-AMD mean donor age, 77.8 ± 5.0 years; AMD mean donor age, 82.2 ± 9.3 years). 

Figure 6.

 

Nuclear densities of cells seeded on aged submacular Bruch's membrane explants after 21-day culture in CM vehicle or BCEC-CM (paired explants from the same donor). (A) Nuclear density comparison of RPE cells derived from hES-RPE (n = 6), cultured human fetal RPE (fRPE, n = 22), and cultured human adult RPE (donor ages 58, 71, 78 years; n = 7). Within each group, significant differences were observed between cells cultured in CM vehicle and cells cultured in BCEC-CM. The nuclear density of cells cultured in CM vehicle was not statistically different between groups. The nuclear densities of hES-RPE and fRPE were not significantly different from each other but were significantly higher than the nuclear density of adult RPE cells after culture in BCEC-CM. (B) Comparison of nuclear densities of fRPE on age-matched, non-AMD versus AMD Bruch's membrane at day 21. Explants seeded with fRPE on aged Bruch's membrane (n = 9) were compared with explants seeded on AMD submacular Bruch's membrane (n = 13). No significant differences were observed in the nuclear densities of fRPE on non-AMD versus AMD explants for a given medium, although the nuclear density was significantly higher in the presence of BCEC-CM versus CM vehicle. Nuclear density values are counts of nuclei of cells directly in contact with Bruch's membrane, expressed as mean nuclear density/mm Bruch's membrane. Bars indicate mean ± SE nuclear density. *P < 0.05; **P < 0.001.

RPE Cell Survival after Different BCEC-CM Culture Times

To determine whether RPE behavior on submacular Bruch's membrane explants depended on the time of exposure to BCEC-CM, explants seeded with fetal RPE and cultured for different periods of time in BCEC-CM were compared with fellow eye explants cultured for the entire incubation period (21 days) in BCEC-CM. One explant of the pair was cultured in BCEC-CM for 3, 7, or 14 days, followed by culturing in CM vehicle, to bring the total number of days in culture to 21 (Fig. 7). Fellow eye explants were cultured in BCEC-CM for the entire 21-day period. Fetal RPE nuclear densities on explants cultured for 3 days in BCEC-CM (mean nuclear density, 3.5 nuclei/mm Bruch's membrane) compared with 21 days in BCEC-CM (mean nuclear density, 21.8 nuclei/mm Bruch's membrane) were significantly different (P = 0.016). Nuclear densities on explants after 7-day BCEC-CM culture (mean nuclear density, 10.9 nuclei/mm Bruch's membrane) compared with 21 days (mean nuclear density, 28.0 nuclei/mm Bruch's membrane) were significantly different (P = 0.008). Nuclear densities after 14-day BCEC-CM culture (mean nuclear density, 17.0 nuclei/mm Bruch's membrane) compared with 21-day BCEC-CM culture (mean nuclear density, 27.9 nuclei/mm Bruch's membrane) were significantly different (P = 0.031). Nuclear densities of explants cultured for 14 days in BCEC-CM were significantly higher than those of explants cultured for 3 days in BCEC-CM (P < 0.05), whereas 3-day versus 7-day and 7-day versus 14-day nuclear densities were not significantly different (P > 0.05). There were no statistically significant differences between the control groups cultured for the entire 21-day period in BCEC-CM (P = 0.074). Ages between groups were not significantly different (P = 0.881; mean donor ages: 3-day cohort, 78.3 ± 7.6 years; 7-day cohort, 76.5 ± 6.0 years; 14-day cohort, 77.2 ± 7.1 years). 

Figure 7.

 

Nuclear densities of fetal RPE cells cultured in BCEC-CM for 3 (n = 7), 7 (n = 8), or 14 days (n = 6), followed by culture in CM vehicle for a total culturing period of 21 days. Submacular Bruch's membrane explants from fellow eyes were cultured in BCEC-CM for the entire 21-day period. Nuclear densities were significantly higher when cultured for the entire 21-day period in BCEC-CM compared with shorter periods of time in BCEC-CM. Nuclear density after 3-day culture was significantly lower than nuclear density after 14-day culture in BCEC-CM. Bars indicate mean ± SE nuclear density. *P < 0.05.

Comparison of RPE Nuclear Density after 21-Day Culture in Different Media and on Different Surfaces

Nuclear densities of fetal RPE after 21-day culture in BCEC-CM or CM vehicle on AMD (including late AMD) and aged, non-AMD explants (present study, mean donor age: CM vehicle, 80.2 ± 7.8 years [Fig. 6A]; BCEC-CM, 78.8 ± 7.3 years [Figs. 6A, 7]) were compared with 21-day fetal RPE nuclear densities on young explants (mean donor age, 44.8 ± 2.3 years) cultured in RPE medium, ²¹ with aged and early AMD Bruch's membrane explants (mean donor age, 73.6 ± 6.4 years) cultured in RPE medium, ^18,21 and with BCEC-ECM–resurfaced aged Bruch's membrane (mean donor age, 73.9 ± 7.4 years) cultured in RPE medium ²¹ (Fig. 8). Of the aged explants studied, mean donor ages of those cultured in BCEC-CM and CM vehicle (includes eyes with early as well as late AMD) were significantly higher than those of aged, including early AMD, explants cultured in RPE media (P < 0.05) but were not significantly different from those of explants resurfaced with BCEC-ECM (aged, non-AMD) (P > 0.05). The nuclear densities of fetal RPE on BCEC-ECM–resurfaced aged Bruch's membrane (mean nuclear density, 23.2 nuclei/mm Bruch's membrane) and unresurfaced young Bruch's membrane (mean nuclear density, 27.2 nuclei/mm Bruch's membrane) were significantly higher than that on aged and early AMD, untreated Bruch's membrane after culture in RPE medium (mean nuclear density, 11.2 nuclei/mm Bruch's membrane ²¹ ) (P < 0.05) and were not significantly different from that on explants cultured in BCEC-CM (mean nuclear density, 23.2 nuclei/mm Bruch's membrane) (P > 0.05). Nuclear densities of fetal RPE cells in these three conditions (i.e., BCEC-CM-cultured, BCEC-ECM–resurfaced, and young Bruch's membrane) were significantly higher than those on aged and AMD Bruch's membrane cultured in CM vehicle (mean nuclear density, 2.2 nuclei/mm Bruch's membrane) (P < 0.05). Nuclear densities of cells cultured in RPE medium on aged and early AMD Bruch's membrane were significantly higher than those of cells cultured in CM vehicle on aged and AMD Bruch's membrane (P < 0.05). 

Figure 8.

 

Comparison of fetal RPE cell nuclear density after 21-day culture in different media and on different surfaces. Nuclear densities of fetal RPE cells after culture in BCEC-CM on aged and AMD Bruch's membrane (n = 43), BCEC-ECM–resurfaced aged Bruch's membrane cultured in RPE medium (n = 11), and young Bruch's membrane cultured in RPE medium (n = 5) were not significantly different. Culture on Bruch's membrane from aged and early AMD donors in RPE medium (n = 33) resulted in significantly lower nuclear densities than those observed in BCEC-CM cultured, BCEC-ECM–resurfaced, and young Bruch's membrane explants. RPE nuclear density after culture in CM vehicle on aged and AMD Bruch's membrane (n = 22) was significantly lower than cultures in RPE medium on aged and early AMD explants. BCEC-CM explant nuclear densities are combined data from 21-day fetal RPE nuclear density counts (see Effects of BCEC-CM on Long-Term Cell Survival) (Fig. 6A) and 21-day BCEC-CM controls (see Cell Survival after Different BCEC-CM Culture Times) (Fig. 7). Data for BCEC-ECM resurfaced explants and young donor explants are from Sugino et al. ²¹ Data for fetal RPE cells on aged (including early AMD) Bruch's membrane explants were combined data from Sugino et al. ^18,21 (data were not significantly different; P = 0.745). Bars indicate mean ± SE nuclear density. *P < 0.05.

ECM Deposition after Culture in BCEC-CM

Providing a newly deposited ECM on the surface of Bruch's membrane can significantly improve cell survival in long-term organ culture, ²¹ and the resultant nuclear densities are similar to those observed after culture in BCEC-CM (Fig. 8). To determine whether incubation in BCEC-CM increases ECM deposition, which might account for the cell-preserving effect of BCEC-CM, we investigated whether RPE ECM deposition on Bruch's membrane increased after culture in BCEC-CM. Since little RPE cell survival on Bruch's membrane was seen in CM vehicle, we compared ECM deposition after culture in BCEC-CM with ECM deposition after culture in standard RPE culture medium, in which some degree of RPE resurfacing was more likely (Fig. 8). To assess ECM deposition on a nontoxic substrate, ECM deposition onto tissue culture dishes was examined at days 7, 14, and 21 (n = 3). 

Stained fibers were present on the surfaces of culture dishes in both media at all three time points. As the time in culture increased, the amount of ECM deposition increased. ECM visualized by Ponceau S staining showed deposition after BCEC-CM culture to be more extensive than deposition after RPE medium culture at all time points (Fig. 9). In BCEC-CM cultures, collagen IV and laminin were present as a thick coating with defects that became smaller with time in culture (Figs. 10 11–12). Collagen IV deposition appeared to be more extensive than laminin and fibronectin deposition at day 7. At days 14 and 21, collagen IV and laminin showed more uniform coating of the culture dish surface than at day 7. Collagen IV and laminin appeared to be extensively colocalized at all three time points (Figs. 10C, 11C, 12C). Fibronectin labeling was identified on a network of thin fibers at week 1, with diffuse labeling of the culture dish between fibers seen at days 14 and 21. Some colocalization of fibronectin with laminin was seen (Figs. 10F, 11F, 12F), but it was not as extensively colocalized as laminin was with collagen IV. In RPE medium cultures (Figs. 10G–L to 12G–L), collagen IV and laminin labeling at day 7 was not as extensive as that seen in BCEC-CM cultures with labeling seen as an open mesh, with localized areas coating the culture dish between the fibers of the mesh. Localized areas of surfaces coated with collagen IV and laminin were more extensive at days 14 and 21 than at day 7, but there was little, if any, increase in labeling between days 14 and 21. Faint fibronectin labeling was detected in RPE medium cultures at all time points, with sparse to moderate labeling seen at day 21. Similar to that observed in BCEC-CM cultures, collagen IV and laminin showed extensive colocalization. Fibronectin colocalization was difficult to determine because of the sparse labeling in some cultures, but it did seem to partially colocalize with laminin (Fig. 10L). Controls showed diffuse, faint fluorescence with rhodamine filters with secondary antibody only in the 3-week RPE medium cultures. No detectable fluorescence was found with FITC or rhodamine filters in preparations at other time points or in CM vehicle cultures at all time points (not shown). Controls in which the antibody was omitted showed no fluorescence with either filter set (not shown). 

Figure 9.

 

Fetal RPE ECM deposition onto tissue culture dishes after 7-, 14-, and 21-day culture in BCEC-CM or RPE medium. ECM is deposited to a higher degree when cells are cultured in BCEC-CM (A–C) over the 21-day period compared with that observed after culture in RPE medium (D–F). Increase in the numbers of thick fibers can be seen in BCEC-CM culture with time while thick fiber deposition seems to be less extensive at all time points after culture in RPE medium. ECM coating of the tissue culture plastic is evident by the disappearance of the culture plastic striations (barely discernable in BCEC-CM cultures at day 7) at day 14 and −21. In RPE medium, culture plastic striations can be seen at day 7 and −14 but not at day 21, indicating that some material coats the culture dish. Scale bar, 50 μm; 0.1% Ponceau S stain, phase contrast.

Figure 10.

 

Immunocytochemical labeling (epifluorescence) of collagen IV, laminin, and fibronectin deposition onto tissue culture dishes after 7-day culture in BCEC-CM or RPE medium. In BCEC-CM, collagen IV labeling (A) is visualized as a network of fibers with some thickened fibers and localized areas of continuous coating. Laminin labeling (B, E) is similar of that of collagen IV but not as extensive. Laminin appears to colocalize with some collagen IV fibers (C, collagen IV, laminin overlay). Fibronectin labeling (D) is an open network of fibers with some areas in which fibers appear to have heavier deposition. Localized nonfibrous coating of the tissue culture dish can be seen adjacent to fibers. Fibronectin-laminin overlay (F) shows some colocalization of label. In RPE medium, collagen IV (G) labeling is more extensive than laminin (H, K). Some colocalization of collagen and laminin is seen in the overlay (I). Very little fibronectin labeling (J) is present. Colocalization of fibronectin and laminin (L) was difficult to determine due to the paucity of fibronectin labeling. Labeling of all three ECM proteins is not as extensive as that seen after BCEC-CM culture (images for each protein were photographed at same exposures). Scale bar, 200 μm.

Figure 11.

 

Immunocytochemical labeling (epifluorescence) of collagen IV, laminin, and fibronectin deposition onto tissue culture dishes after 14-day culture in BCEC-CM or RPE medium. In BCEC-CM, collagen IV (A), laminin (B, E), and fibronectin (D) deposition is more extensive than that seen at day 7 (Fig. 10). All three proteins show extensive resurfacing of the tissue culture dish, with small defects in coverage. Collagen IV and laminin are highly colocalized (C), whereas fibronectin and laminin are partially colocalized (F). In RPE medium, collagen IV (G) and laminin (H, K) labeling are more extensive than at day 7 (Fig. 10) but are not as extensive as labeling seen after culture in BCEC-CM for the same time period. Very little fibronectin (J) is present. Images for each protein in the two conditions were photographed at the same exposure. Collagen IV and laminin are extensively colocalized (I), whereas fibronectin and laminin (L) are colocalized in part. (Intensity of fibronectin labeling has been increased for the overlay.) Scale bar, 200 μm.

Figure 12.

 

Immunocytochemical labeling (epifluorescence) of collagen IV, laminin, and fibronectin deposition onto tissue culture dishes after 21-day culture in BCEC-CM or RPE medium. In BCEC-CM, as in 14-day culture, collagen IV (A) and laminin (B, E) extensively resurface the culture dish and are highly colocalized (C). Fibronectin (D) does not appear to be as extensively deposited as collagen IV and laminin and is colocalized in part with laminin (F). In RPE medium, collagen IV (G) and laminin (H, K) appear to be deposited at levels similar to those seen at day 14 and are not as extensive as those deposited after culture in BCEC-CM. Both proteins appear to be colocalized (I). Very little fibronectin was detected (J), some of which appeared to colocalize with laminin (L). Images for each protein in the two conditions were photographed at the same exposure. Scale bar, 200 μm.

Calcein imaging confirmed the presence of RPE cells on Bruch's membrane explants cultured in both media for 21 days, with more extensive resurfacing and smaller, more compact cells on explants cultured in BCEC-CM (Figs. 13A, 13L). On explants cultured in BCEC-CM, SEM and confocal evaluation showed extensive ECM on the surface of the inner collagenous layer for all three markers (n = 6; mean donor age, 79.2 ± 3.17 years) (Figs. 13B–F, 13H–J). The extent and complexity of the ECM varied within and between BCEC-CM–treated explants, ranging from a complex mesh of thick and thin fibers to a fairly continuous ECM sheet with a rough surface (Figs. 13B–F, 13H–J). Collagen IV and laminin labeling were found as thick and thin fibers, with some localized thickening giving the labeling a punctate appearance. Both collagen IV and laminin formed a continuous sheet in localized areas, visualized between fibers. Collagen IV labeling was more variable than that of fibronectin and laminin and sometimes not as extensive. There was some colocalization of collagen IV with laminin. Fibronectin was found mainly in fibers and did not appear to colocalize with laminin. On explants cultured in RPE medium, no or sparse localized labeling was observed despite RPE presence on Bruch's membrane, as visualized by calcein imaging (Figs. 13L, 13M). When present, laminin labeling could be seen on short strands. If present, fibronectin was present in short (thin) strands; sometimes labeling was punctate. Collagen IV labeling generally was sparse or not present. ECM tended to be more prevalent outside the submacular area on the periphery of the explant. SEM evaluation of the explants showed predominantly bare ICL or a few strands on the surface of the bare ICL in explants cultured in RPE medium (Fig. 13M). Control explant pieces (omission of primary antibody or both primary and secondary antibodies) showed slight autofluorescence of the tissue (Figs. 13G, 13K) at the microscope settings used to obtain immunolabeled images. 

Figure 13.

 

ECM deposition under fetal RPE cells on Bruch's membrane from a 70-year-old donor (no submacular pathology) after 21-day culture in BCEC-CM or RPE medium. In BCEC-CM, (A) calcein imaging of cells on Bruch's membrane before removal with ammonium hydroxide. The explant is resurfaced almost completely with small, highly fluorescent cells. Small defects are present in the RPE cell layer. (B) SEM of the surface of Bruch's membrane revealed after cell removal in an area in which the ECM has been damaged (possibly at the time of cell removal or during confocal imaging manipulation), demonstrating the difference in surface morphology of the ICL compared with the newly deposited ECM. Arrows point to the folded edge of the ECM. (C) A network of open and fused fibers covers the surface of the ICL. The ECM forms a fairly continuous sheet in some areas. High-magnification inset shows the ECM surface details. Both collagen IV (D) and laminin (E) covered the explant with an extensive meshlike deposition. There was some colocalization of label (F, overlay). (G) Control (no primary antibody, overlay) imaged at similar pinhole settings as in D and E, with higher detector gain in both FITC and rhodamine channels. Very little fluorescence is seen in either channel, with some choroidal autofluorescence (FITC) seen in the upper left (arrow) of the image (tissue was not flat). (H–J) Fibronectin labeling (H) of ECM fibers was evident on the explant, whereas laminin labeling (I) similar to that seen in E was seen in fibers, between fibers, and as punctate labeling associated with fibers. Punctate laminin labeling shows up best in the overlay (J). Fibronectin and laminin did not appear to be colocalized to any significant degree (J, overlay). (K) Control (no primary or secondary antibodies, overlay) imaged at the same settings as in H to J. Only faint autofluorescence could be detected. In RPE medium, (L) calcein imaging of the explant shows cells resurfaced Bruch's membrane with several small RPE cell defects (arrows) in the submacular area. A large RPE defect is present on one edge (asterisk). Photographed at the same intensity settings as in A, the overall intensity of calcein imaging appears to be less except at the edge of the large defect. (M) SEM examination of the surface of this explant revealed no ECM deposition, confirming negative labeling (not shown) of all three markers by confocal examination. Collagen fibers are partially obscured by deposits. (A, L) Epifluorescence. (B, C, M) SEM. (D–K) Confocal compressed z-stacks. SEM scale bar, 10 μm; (C, inset), 5 μm; confocal scale bar, 50 μm.

Discussion

In previous studies examining fetal RPE attachment to aged and AMD submacular Bruch's membrane, we showed that RPE cells can attach, to a high degree, to the RPE basement membrane and to the inner collagenous layer, indicating that attachment to these layers may not be the limiting factor in cell transplant success. ^17,27 Immunochemistry studies showed that at days 3 and 7 after seeding, fetal RPE cells were present on Bruch's membrane and appeared to be fairly healthy based on the appearance of stained nuclei. Between days 7 and 14, a high degree of cell death had occurred, and additional cell death was observed at days 14 and 21. ¹⁸ At these later time points, abundant condensed and fragmented nuclei were present. These studies provide evidence that a method to ensure cell survival (vs. attachment alone) must be developed for RPE cell replacement therapy to be successful in AMD patients. 

Previously, we demonstrated that cell survival on aged submacular Bruch's membrane can be enhanced greatly by resurfacing Bruch's membrane with a cell-deposited ECM. ²¹ This resurfacing strategy was chosen because ECM deposited by BCECs supports rapid fetal RPE attachment and growth in cell culture. ^22,32 We showed that resurfacing the aged Bruch's membrane with BCEC-ECM improved long-term cell survival significantly (>200%). A limitation to the feasibility of using this ECM in clinical applications was the insolubility of the ECM and the resultant low yield of proteins with ECM harvest, both for transfer to Bruch's membrane and for quantitative analysis. ²¹ In the present study, we used BCEC-CM harvested from BCECs cultured under conditions similar to those that yield BCEC-ECM–coated culture dishes and BCEC-ECM deposition on Bruch's membrane. ²¹ The rationale for this choice was based on the reported abundance of potentially cell-supporting proteins secreted into the medium, such as ECM and ECM-associated molecules. ^{23 –25} In the present study, we demonstrated that cell survival is greatly enhanced when RPE cells are exposed to this BCEC-CM in long-term culture and, importantly, that cell survival is enhanced on the submacular Bruch's membrane of AMD eyes. We showed previously that RPE cell survival on AMD submacular Bruch's membrane explants was severely impaired after culture in RPE medium. ¹⁷ However, the nuclear density of fetal RPE cells on submacular human Bruch's membrane cultured in RPE medium was significantly higher than the nuclear density of fetal RPE cells cultured in CM vehicle (Fig. 8). This difference could be related to the significantly lower mean donor age of the explants cultured in RPE media and, therefore, possibly fewer aged and AMD changes to Bruch's membrane. However, culture on non-AMD and AMD explants was similarly poor in CM vehicle. It seems likely that the serum in the RPE medium aided in supporting cell survival to a slight degree but not to the degree seen in BCEC-CM. 

BCEC-CM may supply soluble matrix proteins for ECM deposition, stimulate ECM deposition, and/or stimulate the RPE cells in some other fashion, thus allowing better survival that could lead to increased ECM deposition on aged and AMD submacular Bruch's membrane. ^26,33 The presence of increased ECM deposition under the cells cultured in BCEC-CM compared with cells cultured in RPE medium may reveal a mechanism by which cell survival is enhanced, perhaps in the same manner that BCEC-ECM–resurfaced explants support long-term cell survival on submacular human Bruch's membrane. We do not know whether ECM deposition per se enabled cell survival in the same way as resurfacing Bruch's membrane with BCEC-ECM or whether ECM deposition was a reflection of long-term survival of the cells by another mechanism. When RPE cell survival is observed on explants cultured in RPE medium, the cells are not as differentiated as those cultured in BCEC-CM and have not deposited ECM to any degree. The degree of ECM deposition by RPE cells on culture dishes in RPE medium or in BCEC-CM was much greater than that observed on Bruch's membrane explants. This difference may arise because RPE cells appear to mature more slowly on Bruch's membrane, ¹⁷ and a certain degree of maturity must be reached before ECM deposition can occur. ^34,35 Differences in the amount and composition of ECM deposition may also be related to differences in cellular response to the underlying substrate. ^36,37 Because the antibodies used in this study were not specific to humans, we do not know whether the deposited proteins originated from the BCEC-CM (bovine proteins) or from protein synthesis by the RPE, or both. As mentioned previously, when cultured in standard RPE medium, fetal RPE cells can survive to a high degree up to 7 days in organ culture on submacular human Bruch's membrane. ^17,18 After day 7, survival was impaired, with decreasing presence of cells as the time in culture increased. The necessity of long-term presence of BCEC-CM to ensure cell survival was consistent with the notion that sustained cell stimulation is a factor in ensuring cell survival. RPE cell survival after culture in BCEC-CM as measured by nuclear density was statistically similar to that of cells cultured on BCEC-ECM and similar to that observed on young Bruch's membrane (both the latter were cultured in RPE medium) (Fig. 8). The nuclear densities observed in these studies were much lower than those of submacular in situ RPE cells even when comparing age-matched nuclear densities and were much lower than those of fetal RPE cells in culture. ¹⁷ The lower nuclear density on submacular Bruch's membrane after culture in BCEC-CM was due in part to the existence of defects in surface coverage on some explants. Many explants are almost fully resurfaced or are fully resurfaced after culture in BCEC-CM, and many cells appear to show some morphologic features of differentiation (e.g., apical processes, tight junctions). However, there is variability in cell size, with some cells fairly large, especially compared with the size of cultured fetal RPE cells. One source of the variability in cell behavior between explants might have been biological variability in the composition of BCEC-CM between batches because some batches appeared to be less effective than others, showing more and larger defects in surface coverage and larger and flatter cells. Another source of variability in cell behavior arose from differences in cell survival on localized areas of Bruch's membrane within explants because some explants showing excellent overall resurfacing also exhibited small localized defects in RPE cell coverage (e.g., Figs. 2, 4). Lastly, particularly in reference to the AMD cohort, variability in cell behavior on explants with CNV was likely related, at least in part, to the differences in the Bruch's membrane surface after CNV removal because no additional debridement was performed. The degree of damage to Bruch's membrane or the preservation/removal of deposits on Bruch's membrane after CNV removal was highly variable within and between donor eyes. 

In a previous study, ¹⁸ hES-RPE were shown to have the potential to survive on equatorial and submacular Bruch's membrane to a similar degree as fetal RPE cells after day 21 culture in RPE medium. On submacular Bruch's membrane, the survival was poor for both cell types, with hES-RPE survival impaired at earlier times in culture than fetal RPE cell survival. ¹⁸ hES-RPE were from frozen stock, which is a possible cause for the difference in initial cell survival because the fetal RPE cells were from fresh cultures. However, in the present study, hES-RPE were from fresh stock, and although the nuclear densities of hES-RPE cultured in BCEC-CM were similar to those of fetal RPE cells, hES-RPE in general were very flat and were not differentiated to the same degree as fetal RPE cells on submacular Bruch's membrane. These results imply that hES-RPE may take longer to acquire mature RPE cell features on Bruch's membrane than fetal RPE cells, consistent with the behavior observed in cell culture. ^17,30 Whether fetal RPE cells or hES-RPE can achieve size and differentiation features found in cell culture or in situ and whether the cells can perform RPE cell functions are subjects that must be considered for future study. 

Cultured aged adult RPE cells were generally larger than hES-RPE and fetal RPE cells in cell culture and on Bruch's membrane. On Bruch's membrane, most adult RPE cells showed well-developed apical processes even in very large flat cells, but their survival in general was not as good as that of fetal RPE cells or hES-RPE. The nuclear density of cultured adult RPE cells after BCEC-CM culture (10.0 ± 0.95 nuclei/mm Bruch's membrane) was not significantly different (P = 0.887) from the nuclear density of fetal RPE cells after RPE medium culture (11.2 ± 1.7 nuclei/mm Bruch's membrane; Fig. 8). These results indicate that though culture in BCEC-CM significantly enhances RPE cell survival on aged Bruch's membrane, aged adult RPE cells may not be the best choice for cell transplantation, especially when compared with the resurfacing achieved by fetal RPE cells and hES-RPE. 

No other technique known to us promotes such robust RPE cell survival on submacular AMD Bruch's membrane (including eyes with geographic atrophy and CNV). ^19,20 Identification of the critical components of this BCEC-CM and RPE cell function testing are the next steps in the development of a surgically usable adjunct to improve RPE cell survival and differentiation on submacular human AMD Bruch's membrane. 

Supplementary Materials

Table st1, XLS - Table st1, XLS 

Footnotes

 Supported by the Lincy Foundation, the Foundation Fighting Blindness, an unrestricted grant from Research to Prevent Blindness, the Eye Institute of New Jersey, the Janice Mitchell Vassar and Ashby John Mitchell Fellowship, the Joseph J. and Marguerite DiSepio Retina Research Fund, the Foundation of UMDNJ, and the New Jersey Lions Eye Research Foundation.

Footnotes

 Disclosure: I.K. Sugino, P; A. Rapista, None; Q. Sun, None; J. Wang, None; C.F. Nunes, None; N. Cheewatrakoolpong, None; M.A. Zarbin, P

The authors thank Carola Springer for technical assistance.