• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of taosLink to Publisher's site
Trans Am Ophthalmol Soc. Dec 2004; 102: 123–138.
PMCID: PMC1280093




To determine whether cultured fetal human retinal pigment epithelium (RPE) cells can attach and differentiate on submacular Bruch’s membrane from donors over age 55.


Differential debridements of Bruch’s membrane were performed to expose three different surfaces: the RPE basement membrane, the superficial inner collagenous layer (ICL) directly below the RPE basement membrane, and the deeper ICL. Approximately 3,146 cells/mm2 were seeded onto these Bruch’s membrane explants and cultured for 1 or 7 days. Explants were bisected and examined histologically or analyzed with scanning electron microscopy. Nuclear density counts were performed on stained sections. Morphology and cell density were compared to those of cells seeded onto bovine corneal endothelial cell–extracellular matrix (BCE-ECM).


Compared to cells seeded onto BCE-ECM at similar density, cell coverage and cellular morphology were poor at both time points. Unlike cells on BCE-ECM, cell density remained the same or decreased with time. In general, cell morphology on all surfaces worsened by day 7 compared to day 1. Although cells were more pigmented on RPE basement membrane and deep ICL at day 7, poor cellular morphology indicated the remaining cells were not well differentiated. An explant from a donor with large soft drusen showed the poorest resurfacing at day 7 in organ culture.


These data indicate that aged submacular human Bruch’s membrane does not support transplanted RPE survival and differentiation. The formation of localized RPE defects, cell death, and worsening cellular morphology on aged Bruch’s membrane suggest that modification of Bruch’s membrane may be necessary in patients with age-related macular degeneration receiving RPE transplants to prevent graft failure.


Choroidal new vessel (CNV) excision has been proposed as a treatment for choroidal neovascularization1 and, in patients with age-related macular degeneration (AMD), is associated with iatrogenic retinal pigment epithelium (RPE) defects due to the intimate association of RPE cells and the CNVs.25 Combined RPE transplantation and CNV excision has been attempted in eyes with AMD, but it has not led to significant visual improvement in most patients.68 In contrast, RPE transplantation in animal models of retinal degeneration has been proved to rescue photoreceptors and preserve visual acuity.912 Possible causes of RPE transplant failure in human patients include immune rejection, which can be overcome with immune suppressive therapy, and inability of transplanted RPE cells to survive and differentiate on aged submacular Bruch’s membrane. An important distinction between humans with AMD and laboratory animals in which RPE transplantation has been successful is the age-related modification of Bruch’s membrane in human eyes, which may have a significant effect on RPE graft survival.

With normal aging, human Bruch’s membrane, especially in the submacular region, undergoes numerous changes (eg, increased thickness, deposition of extracellular matrix and lipids, cross-linking of protein, nonenzymatic formation of advanced glycation end products).1315 The impact of these changes on RPE survival on Bruch’s membrane has not been elucidated thoroughly. In vitro experiments, however, indicate that some of these changes can adversely affect cell attachment and survival.1618 In addition to age-related changes in Bruch’s membrane, CNV excision can disrupt Bruch’s membrane with exposure of the inner collagenous layer and its lipids and, occasionally, exposure of the elastic layer.2,4,5,19 Thus, it is important to know whether aged submacular Bruch’s membrane supports RPE graft survival and differentiation, independent of immune rejection.

Because aging changes occur more prominently in submacular Bruch’s membrane and because AMD-related changes predominate in the macular region, we sought to examine the ability of submacular Bruch’s membrane to support initial and long-term survival of RPE cells in order to improve results of RPE transplantation in humans with AMD. We chose to address this issue by comparing the ability of cultured fetal human RPE cells to attach and grow on an “ideal” surface (ie, bovine corneal endothelial cell–extracellular matrix [BCE-ECM])20 with their ability to grow on different sublaminae of aged submacular Bruch’s membrane, which are the surfaces that transplanted RPE cells will encounter in AMD eyes that have undergone CNV excision. We chose to work with cultured fetal cells (vs adult RPE cells) for two reasons. First, a cultured fetal RPE cell line is fairly homogenous (vs cultured adult RPE cultures),21,22 and we did not want variability in cell behavior to alter cell survival and differentiation on a given surface. Second, cultured human fetal RPE cells are robust and can adhere to various sublaminae of Bruch’s membrane (vs adult RPE cells, which do not adhere well to the inner collagenous layer of aged submacular Bruch’s membrane).23 We examined cultured, passaged human fetal RPE cell survival on (1) aged submacular human RPE basement membrane, (2) the lipid-rich superficial inner collagenous layer (ICL), and (3) the deeper ICL. Despite the ability of cultured fetal RPE cells to initially resurface all three layers, we demonstrate decreasing cell survival and worsening morphology with time on all three surfaces.


Human Donor Tissue

Fetal eyes were obtained from the Central Laboratory for Human Embryology (University of Washington, Seattle, Washington). Eyes used for Bruch’s membrane explants were received from numerous eye banks placed through the National Disease Research Interchange (Philadelphia, Pennsylvania) or from the North Carolina Eye Bank, a Vision Share member (Apex, North Carolina). Donor criteria for tissue acceptance included (1) no history of chemotherapy or radiation to the head, (2) up to 8 hours from death to enucleation, (3) up to 48 hours from death to experimentation, and (4) intact Bruch’s membrane under the macula as visualized through a dissecting microscope. This research followed the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of the New Jersey Medical School, University of Medicine and Dentistry of New Jersey.

RPE Isolation and Culture

Human fetal RPE cells were isolated using 0.8 mg/mL collagenase type IV (Sigma, St Louis, Missouri), as previously described.23 Isolated RPE sheets were plated on BCE-ECM–coated dishes and cultured in supplemented Dulbecco’s modified Eagle’s medium (DMEM).23 A single fetal RPE cell line (gestation age, 17.4 weeks) of passage 2 to 4 was used in all the attachment studies to avoid complicating factors arising from differences in cell lines.

Bruch’s Membrane Explant Preparation

Globes were immersed briefly in povidone-iodine solution followed by two 10-minute washes in DMEM containing 2.5 mg/mL amphotericin B and prepared as previously described.23,24 To reproduce different surfaces to which RPE cells must attach in patients following CNV excision, debridements were created to expose the RPE basement membrane, the superficial ICL directly below the RPE basement membrane, or deeper layers of the ICL. A modification of a previously published method was used to create differential debridements of Bruch’s membrane from older donors.24 Differential debridements were created by wiping with a moistened surgical sponge to expose the desired Bruch’s membrane sublamina. Specimens (27 Bruch’s membrane explant specimens; donor age, 65.6 ± 10 years) were prepared prior to beginning the RPE attachment studies to determine reproducibility of debridements. These specimens were analyzed with scanning electron microscopy (SEM). Identification of the depth of debridements was based on the surface morphology of the explant with SEM24 and further confirmed with light microscopic analysis.

Attachment Studies

Differential debridements were performed on submacular Bruch’s membrane explants as described above. Each submacular explant of a donor pair was prepared to expose a different layer of Bruch’s membrane (eg, RPE basement membrane in one eye and superficial ICL in the fellow eye). Explants were situated within 7-mm-diameter trephines (Storz Ophthalmics, St Louis, Missouri), seeded with cultured fetal human RPE (3,146 viable cells/mm2), and incubated for 1 day (RPE basement membrane, n = 7 [74.86 ± 9.3 years]; superficial ICL, n = 7 [66.3 ± 6.8 years]; deep ICL, n = 6 [66.5 ± 8.1 years]) or 7 days (RPE basement membrane, n = 6 [77.0 ± 10.3 years]; superficial ICL, n = 6 [81.7 ± 6.8 years]; deep ICL, n = 6 [77.2 ± 9.0 years]). Debridement depth was confirmed in toluidine blue or MSS (basic fuchsin and toluidine blue, Polysciences, Inc, Warrington, Pennsylvania) stained sections and by high-magnification SEM of uncovered areas. Cells were also seeded onto BCE-ECM plates at the same density and served as positive controls.

Two additional specimens were prepared from donor eyes with large, soft drusen. One donor was previously diagnosed as having AMD (7-day organ culture specimen), and the other donor was not. The latter donor showed drusen in one eye only. Submacular RPE cells in these eyes were loosely attached and could be peeled off as a sheet. Cells that remained after peeling were gently brushed off. The explants were seeded with cells and followed for 1 day (donor age, 69 years) or 7 days (donor age, 74 years) in organ culture.

All explants were bisected for histologic evaluation and SEM following fixation in 4% paraformaldehyde.

Light Microscopic Analysis

Following rinsing in phosphate-buffered saline, explant halves were dehydrated in a graded series of ethanol and infiltrated and embedded in LR White (EMS, Hatfield, Pennsylvania). Four to six sections of 2-μm thickness were mounted on Probe-on Plus (Fisher Scientific, Pittsburgh, Pennsylvania) slides and dried overnight. Nonadjacent slides were stained with 0.03% toluidine blue or a combination of toluidine blue and basic fuchsin (MSS stain). Slides were screened initially to include only those explants with intact Bruch’s membrane and choroid. Nuclear density counts rather than cell counts were performed to compare cell attachment because of the difficulty in discerning cell boundaries in very flat cells. The observers were blinded to the conditions of the experiment. Every fourth or fifth slide was selected for analysis, and nuclear counts from five slides were performed on one section per slide from the central 3- to 4-mm area of the section (ie, submacular Bruch’s membrane). In areas of multilayers, only cells attached to Bruch’s membrane were counted. Linear measurements of Bruch’s membrane in the analyzed areas were obtained by digital image acquisition and measurement with the freehand line tool using NIH Image J. In some specimens, one fourth of the explant was fixed in half-strength modified Karnovsky fixative (2% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4) to confirm morphological observations following embedment in JB4. For statistical analysis, normal distribution of the data was confirmed before using analysis of variance. Comparisons between two groups were performed by Student’s t test, and if there were more than two groups, the Tukey-Kramer honestly significant difference test was used. A P value of <.05 was considered statistically significant.

Scanning Electron Microscopic Analysis

Explant halves were further fixed in half-strength modified Karnovsky fixative. Following rinsing in phosphate-buffered saline, explants were dehydrated in a graded series of ethanol, critical point dried, and sputter-coated with 20 nm gold/palladium. Samples were examined in a JEOL 35U (JEOL, Peabody, Massachusetts) and were analyzed for debridement surface, cell coverage, and cellular morphology.


Day 1

Bovine Corneal Endothelial Cell–Extracellular Matrix

Cells on BCE-ECM were large, elongate, and very flat. Nuclei were distributed unevenly and were oval or rectangular in shape. Cell thickness appeared uniform even though some cells were extremely elongated (Figure 1A). Cisternae were infrequent.

Figure 1Figure 1
Cultured fetal RPE on bovine corneal endothelial cell–extracellular matrix after 1 day in culture. A, Light micrograph. Nuclei are unevenly distributed on the surface in a monolayer of uniform thickness (toluidine blue stain; bar = 20 ...

By SEM, cells appeared fairly uniform in shape and showed smooth, flattened surfaces. Short cell extensions over adjacent cells were common (Figure 1B).

RPE Basement Membrane

Cells were present on Bruch’s membrane predominantly as a monolayer and, unlike cells on BCE-ECM, had highly variable morphology. Bruch’s membrane was totally or almost totally resurfaced except for a few small defects in coverage by RPE cells. Cell shape and thickness were highly variable between explants. Cellular morphology ranged from round, to flattened or extremely flattened, to spindle-shaped, or rectangular. Cellular morphology within a single explant could be variable or predominantly one cell shape and/or thickness (Figure 2A). Defects in RPE coverage of Bruch’s membrane could be seen even in explants exhibiting high RPE cell density in some areas. The ability of the cells to resurface the explant at this time did not appear to depend on Bruch’s membrane donor age. The explant showing the most compact cells was that of a 90-year-old donor. In the areas where the cell density was high, cells were often on top of basal laminar deposit (Figure 2B). In areas showing defects in RPE cell coverage, basal laminar deposit was not evident (Figure 2C). Subcellular or intercellular spaces (cisternae) were present in all explants to varying degrees (Figure 2A, arrowhead). Condensed, darkly staining nuclei were present on all specimens (Figures 2A and and2B,2B, large arrows).

Figure 2Figure 2Figure 2Figure 2Figure 2
Cultured fetal RPE seeded onto RPE basement membrane of aged submacular Bruch’s membrane after 1 day in culture. A, Light micrograph of cells on a 79-year-old explant. A monolayer of small, flattened cells is present on the RPE basement membrane ...

Scanning electron microscopy of cells seeded onto RPE basement membrane confirmed light microscopic observations and showed variable cell size and shape and, except for one explant (donor age, 90 years), total or almost total Bruch’s membrane coverage with small RPE defects (<1 cell diameter in size). Compared with cells seeded onto BCE-ECM, cells could be similar, larger or smaller, or rounder or flatter, and in some explants a mixture of cell sizes was present (Figure 2D). In general, cell extensions onto neighboring cells were more numerous and often larger and thicker than those found on cells seeded on BCE-ECM (Figure 2D, inset). Most explants showed more supernumerary cells than those seeded on BCE-ECM. The supernumerary cells were not uniformly distributed on the explants. The oldest explant (donor age, 90 years) with incomplete coverage showed areas resurfaced by compact cells with adjacent areas of defects. Cells around the RPE defect were thickened (Figure 2E).

Superficial Inner Collagenous Layer

Cells seeded onto the superficial ICL showed a higher degree of variability between and within explants than those seeded onto RPE basement membrane. Four of seven explants showed incomplete coverage by RPE cells with variable morphology. In these explants, many single cells or cell patches were present, comprising rounded and sometimes enlarged or spindle-shaped cells (Figure 3A), whereas in other areas cells were flat and elongate. In addition, some areas were resurfaced with a continuous monolayer of cells, but cell morphology was highly variable (Figure 3B). In explants showing almost complete coverage, RPE defects were more numerous, of variable size, and often larger than those found on RPE basement membrane. The cells covering the explants were variable in morphology: flat, extremely flat, or spindle-shaped (Figure 3B). Densely staining cells and cells with misshapen and/or densely staining nuclei were observed in five of seven explants.

Figure 3Figure 3Figure 3Figure 3
Cultured fetal RPE seeded onto the superficial inner collagenous layer (ICL) of aged submacular Bruch’s membrane after 1 day in culture. A, Light micrograph of cells on a 62-year-old explant showing incomplete resurfacing. Cells incompletely ...

Scanning electron microscopy confirmed the presence of smooth, flattened cells of highly variable size and shape in the three of four explants with defects in RPE coverage of Bruch’s membrane (Figure 3C). The fourth explant was incompletely resurfaced with enlarged, thick cells. Long cell extensions were sometimes observed with lamellipodia and filopodia extending over several cells. One explant showing almost complete coverage by light microscopy (Figure 3B) was covered with cell debris, which obscured the underlying cells. The remaining two explants showed almost complete Bruch’s membrane coverage by flattened cells with small defects in the resurfacing (Figure 3D). Varying amounts of supernumerary cells or cell debris were present on the surface of attached cells or directly on the superficial ICL in Bruch’s membrane explants.

Deep Inner Collagenous Layer

Cell resurfacing and morphology were similar to that of cells seeded onto RPE basement membrane in five of six explants, except that defects in Bruch’s membrane coverage were more common and often larger. Cells seeded onto the deep ICL were predominantly in a monolayer with highly variable morphology (Figure 4A). As with cells on the RPE basement membrane, cell density within a single explant was highly variable. The sixth explant was incompletely resurfaced by enlarged cells with enlarged nuclei (Figure 4B).

Figure 4Figure 4Figure 4Figure 4
Cultured fetal RPE seeded onto the deep inner collagenous layer (ICL) of aged submacular Bruch’s membrane after 1 day in culture. A, Light micrograph of cells on a 79-year-old explant. Spindle-shaped cells resurfaced this area of the explant ...

Similar to cells seeded onto RPE basement membrane, SEM showed that predominantly a monolayer of cells of variable cell size and shape resurfaced the deep ICL. As with cells on RPE basement membrane, cell thickness and shape were variable between Bruch’s membrane explants (Figure 4C). Some multilayering was evident in three of six explants in localized areas, with enlarged or elongate cells present on top of the monolayer. Cell extensions were common. Small to large RPE defects were evident in five of the six samples (Figure 4D).

Age-Related Macular Degeneration Eye

This donor eye showed large, soft drusen in one eye only. Light microscopy (Figures 5A and and5B)5B) showed both eyes resurfaced by a monolayer of large, flattened cells with large nuclei. No cisternae were present.

Figure 5Figure 5Figure 5Figure 5
Cultured fetal RPE seeded onto a donor pair with and without drusen (donor age, 69 years) after 1 day in culture. A and B, Light micrographs of cells on submacular Bruch’s membrane explant. Cells resurfacing both explants (A, eye with age-related ...

Scanning electron microscopy of the eye with drusen showed areas of Bruch’s membrane lacking RPE cells in the center of the explant with complete coverage by RPE cells in the subperimacular area by large, flattened polymorphic cells (Figure 5C). The areas lacking cells were not as large as the defects seen in 90-year-old donor explant with a similar debridement (Figure 2E). The fellow eye, which did not have drusen, was resurfaced by cells of similar size and morphology with infrequent small defects (Figure 5D).

Day 7

Bovine Corneal Endothelial Cell–Extracellular Matrix

Cells uniformly resurfaced the culture dish with predominantly a monolayer of flattened cells (Figure 6A). Flattened nuclei were distributed fairly uniformly. Some localized bilayers were seen with cells exhibiting very flat nuclei in the top layer (Figure 6B). Cell cytoplasm was stained uniformly. Cisternae were abundant and occasionally large. The majority of cells were unpigmented. Occasional large, rounded pigmented cells were seen. Cells were much flatter than they were at day 1.

Figure 6Figure 6Figure 6
Cultured fetal RPE on bovine corneal endothelial cell–extracellular matrix (BCE-ECM) after 7 days in culture. A, Light micrograph. RPE cells resurfaced BCE-ECM with small, flattened cells of uniform thickness and regularly spaced nuclei. Cisternae ...

Scanning electron microscopy showed a monolayer of small, hexagonal-shaped cells on BCE-ECM. Larger, flat cells were present within the monolayer interspersed among the hexagonal cells (Figure 6C).

RPE Basement Membrane

Cells were present on RPE basement membrane in incomplete monolayers with some multilayering in four of six explants. In three explants, cells were highly variable in morphology. Many RPE cells in the monolayer showed variable loss of cytoplasm with a remaining (sometimes condensed) nucleus. Many cells were spindle-shaped and flattened to varying degrees, and other cells were very elongate, extending over subcellular cisternae, cell fragments, or cell ghosts (Figure 7A). Cisternae were common. Cells were more pigmented than at day 1, but even in areas with highly pigmented cells, cells were irregularly shaped. The fourth explant with incomplete resurfacing showed a monolayer of unpigmented cells that were healthier (more intact) and often on residual basal laminar deposits (Figure 7B). Cells on bare RPE basement membrane without deposits were of more variable morphology, ranging from similar to those on basal laminar deposit to elongate and spindle-shaped. Few small drusen were located on the surface with overlying cells (Figure 7C). The unresurfaced areas appeared to be free of basal laminar deposit. Of the two explants showing the best resurfacing, both explants had cells that were predominantly attached onto basal laminar deposit as a monolayer or a monolayer with localized bilayers. When a bilayer was present on basal laminar deposit, the cells directly on top of the deposit were often large and rounded with lightly staining cytoplasm (Figure 7D). Cells on top of this layer were either similarly enlarged or very flat with flat nuclei. Some cells rested on cell fragments and ghosts. Cells directly on basal laminar deposit were flat with flat nuclei, most closely resembling cells on BCE-ECM (Figure 7E). Most RPE cells on these resurfaced explants were not pigmented.

Figure 7Figure 7Figure 7Figure 7Figure 7Figure 7Figure 7Figure 7
Cultured fetal RPE seeded onto RPE basement membrane of aged submacular Bruch’s membrane after 7 days in culture. A, Light micrograph of cells on a 68-year-old explant showing incomplete resurfacing. In this explant, cells sparsely covering the ...

Scanning electron microscopy revealed that the RPE basement membrane was resurfaced predominantly by a monolayer of RPE cells that appeared more uniform than at day 1 owing to the lack of numerous cell extensions (frequently seen at day 1) and the uniform thickness of adjacent cells. Cell shape within an explant could transition from extremely flat and smooth polygons to thick and elongate cells, or small and compact cells (but larger than those on BCE-ECM). Large defects several cell diameters wide were present in four of seven explants (Figure 7F). Cells bordering the RPE defects were either thickened and elongated, forming a distinct border around the defect, or extended very flat processes into the defect. The explant showing the best resurfacing (cells on top of basal laminar deposit) showed no defects in Bruch’s membrane coverage with cells of highly variable thickness and shape (Figure 7G). This explant did not have the very large, flattened cells seen in the explants with RPE defects. The second explant with cells on basal laminar deposit showed the smallest cells on RPE basement membrane (Figure 7H). Supernumerary cells were not as common as at day 1 on all explants.

Superficial Inner Collagenous Layer

Cells resurfacing the superficial ICL had variable morphology, sometimes with large RPE defects (two explants). Four of six explants were almost completely resurfaced with small RPE defects. In explants with large defects plus one explant that was almost completely resurfaced, cells were in a monolayer of variably flattened to extremely flattened cells. Many cells were spindle-shaped (more than seen on RPE basement membrane). Some cells were poorly attached (lifted off Bruch’s membrane), and some were fragmented (Figure 8A). Some areas were resurfaced by flat cells that appeared to be lying on top of cell debris. RPE cells were nonpigmented or variably pigmented, but they never were as pigmented as on RPE basement membrane. Cisternae were abundant. Of the remaining two explants, one was resurfaced with large cells with large nuclei (similar to that seen in Figure 4B). The remaining explant was resurfaced with a monolayer and bilayer of cells that were similar to cells seeded onto basal laminar deposit at day 7 in some areas (Figure 7D and and7E),7E), whereas other areas were resurfaced with spindle-shaped cells. Cells were not pigmented in these explants.

Figure 8Figure 8
Cultured fetal RPE seeded onto the superficial inner collegenous layer (ICL) of aged submacular Bruch’s membrane after 7 days in culture. A, Light micrograph of cells on an 84-year-old explant. Cells typically resurfacing the superficial ICL ...

Scanning electron microscopy of RPE cells growing on the superficial ICL showed sparse to incomplete coverage in two of six explants, with single cells or patches consisting of flat cells. The remaining four explants showed almost complete coverage by large flat cells with small defects (Figure 8B). Some multilayering was seen in two of these explants, with large, flat, elongated cells on top of the monolayer. Varying amounts of cell debris were found on three of six explants.

Deep Inner Collagenous Layer

Five of the six explants were resurfaced predominantly by a monolayer of very flat, spindle-shaped cells with abundant cisternae (Figure 9A). Cells were variably pigmented. Many cells appeared to be fragmented. The remaining explant was resurfaced by healthier-appearing cells (more intact) that were of variable morphology but mostly unpigmented. This explant also contained areas where the cells were extremely flat and spindle-shaped.

Figure 9Figure 9
Cultured fetal RPE seeded onto the deep inner collagenous layer (ICL) of aged submacular Bruch’s membrane after 7 days in culture. A, Light micrograph of cells on a 78-year-old explant. Cells on the deep ICL tended to be flat and spindle-shaped, ...

Scanning electron microscopy of cells on deep ICL shows large defects in Bruch’s membrane in only one explant. Resurfacing in all explants tended to be by large, polymorphic, and flattened cells (Figure 9B). Two of the explants had varying amounts of cell debris.

Age-Related Macular Degeneration Eye

Cells resurfacing the central portion of the explant (ie, submacular Bruch’s membrane) were present on RPE basement membrane and the superficial ICL as patches (sometimes clumps) or single cells that were predominantly flattened and spindle-shaped or rounded (Figure 10A). Many cisternae were present under and between cells. Large RPE defects were present. A few clumps contained enlarged, pigmented cells, but most of the cells were unpigmented. In the subperimacular area, the explant was resurfaced with RPE cells that were of more uniform morphology; cells covered the surface with a monolayer and, occasionally, a localized multilayer of flattened, spindle-shaped cells (Figure 10B).

Figure 10Figure 10Figure 10Figure 10Figure 10
Cultured fetal RPE seeded onto an explant with age-related macular degeneration (AMD) (donor age, 74 years) after 7 days in culture. A, Light micrograph of cells resurfacing AMD eye in the submacular area. Cells resurfacing the AMD explant showed sparse ...

Scanning electron microscopy showed large defects in the center of the explant with resurfacing in the subperimacular area by a complete monolayer of cells (Figures 10C and 10D). RPE defects showed distinct borders with few lamellipodia or filopodia extending onto the bare surface (Figure 10B). High magnification revealed that the surface was superficial ICL in the areas not resurfaced.

Summary of Morphological Observations

All donor eyes showed basal linear deposit throughout the ICL and in the intercapillary pillars. These studies showed no correlation between attached RPE cell density and the extent of Bruch’s membrane deposits as visualized by toluidine blue staining. There did not appear to be a clear correlation between cellular morphology and the age of the explant. Cells seeded onto the superficial ICL appeared to have the poorest cellular morphology and the least amount of Bruch’s membrane surface coverage at day 1. By day 7, cellular morphology was poor with apparent cell death on all three surfaces and cell defects in almost all explants, regardless of the surface. Cells seeded on basal laminar deposit or on top of other cells (which were sometimes dead [ghosts] or fragmented) had better cellular morphology than those directly in contact with Bruch’s membrane at day 7. Flattened, spindle-shaped cells were the most common cellular morphology on all three surfaces.

Nuclear Density Measurements

At day 1, the nuclear density was similar on all three surfaces of Bruch’s membrane examined as well as on BCE-ECM (Figure 11). By day 7 in culture, there was a statistically significant increase in the nuclear density of RPE cells growing on BCE-ECM (P = .007). However, the RPE cell nuclear density declined on Bruch’s membrane on all three surfaces. Only the decrease on the deep ICL was statistically significant (P = .004).

Figure 11
Nuclear density of cultured fetal RPE seeded on aged submacular human Bruch’s membrane sublaminae or bovine corneal endothelial cell–extra-cellular matrix (BCE-ECM) at day 1 and day 7. The number of RPE nuclei was counted, and the length ...


The lack of significant visual improvement in human patients undergoing RPE transplantation has been attributed to immune rejection, inability of cells to resurface an aged and/or damaged Bruch’s membrane, or atrophy of choriocapillaris, all causing death of the RPE cell graft. Irreversible loss of photoreceptors prior to or during transplantation may also contribute to lack of improvement in vision.25

Aging and AMD are characterized by numerous changes in Bruch’s membrane that may have an adverse effect on survival and function of both in situ and transplanted RPE cells.26 One way to assess the effect of aged submacular Bruch’s membrane on transplanted RPE cell survival is to grow RPE cells on Bruch’s membrane explants from human donor eyes. Several studies have used this explant model to examine the ability of RPE to attach to and resurface Bruch’s membrane.2733 Limitations of this system include absence of overlying photoreceptors and underlying choroidal circulation that may influence RPE behavior in the subretinal space following transplantation. Similarly, culture conditions, including levels of serum and growth factors, do not reflect conditions in the subretinal space. Nevertheless, the human Bruch’s membrane explant system allows the study of RPE interaction with Bruch’s membrane independent of other factors (with the possible exception that surviving choroidal cells in the explant may affect the RPE).

Using cultured adult human RPE and peripheral human Bruch’s membrane, Tezel and coworkers31 showed that the attachment rate was highest on RPE basement membrane and was lower on the outer layers of Bruch’s membrane. Unlike the findings from the present study, cultured adult RPE resurfaced peripheral Bruch’s membrane almost completely by 14 days when native RPE basement membrane was present.32 Because light microscopic examination of the tissue was not performed and because the appearance of the cell surface by SEM does not correlate with the actual condition of the cells (as observed by light microscopy in the present study), it is difficult to know whether the morphology of cells declined on RPE basement membrane at day 14 or day 21 in the previous study. Whereas the current study demonstrates resurfacing of the explant at day 7, there is no increase in nuclear density with time. Additionally, there are several differences between the explant system used in the current study versus previous studies:31,32 (1) cultured fetal RPE cells were used in our study (vs primary or passaged cultured adult RPE); (2) submacular Bruch’s membrane specimens were used in this study (vs peripheral Bruch’s membrane specimens); (3) RPE cells were removed mechanically in this study (vs RPE removal with ammonium hydroxide); (4) RPE cell density at initial seeding (3,146 cells/mm2) was high in this study (vs low [530.59 cells/mm2]); (5) serum-containing medium was used in this study (vs serum-free medium for a 24-hour incubation); and (6) live choroidal and vascular endothelial cells were present in the explant system used in this study (vs no live cells in explant due to ammonium hydroxide removal of RPE cells). Still, the results from previous and current experiments might mean that submacular and peripheral Bruch’s membrane differs in the ability to support RPE survival following initial attachment. Additionally, in the present study, the existence of large RPE defects in the center of the submacular Bruch’s membrane explant of the eye with AMD, where the drusen were present, may indicate that submacular Bruch’s membrane versus subperimacular Bruch’s membrane also differ in their ability to support RPE cell survival and differentiation.

Our published studies show that the initial attachment of cultured human fetal RPE onto submacular RPE basement membrane was higher than attachment onto the deep ICL 1 hour after seeding.29 However, the results of the current study indicate that although initial attachment may be lower on the deep ICL, by 24 hours after seeding, RPE nuclear density and cellular morphology on the RPE basement membrane and the deep ICL are similar (29.22 ± 5.4 nuclei/mm and 30.09 ± 3.1 nuclei/mm, respectively). Whereas RPE cells attached onto the superficial ICL showed a slightly lower nuclear density (27.93 ± 4.4 nuclei/mm), the presence of cell debris, rounded cells, and incompletely or sparsely resurfaced explants (four of seven explants) indicates that early resurfacing of the superficial ICL may not be as effective as on RPE basement membrane or the deep ICL. Because the ICL immediately below the RPE basement membrane (ie, the superficial ICL) has been shown to accumulate dense lipid deposits with aging and can appear as a spherical carpet of cholesterol-rich particles,34,35 the presence of this material may lead to less effective initial RPE attachment onto this layer. Additionally, the ICL external to this layer (ie, the deep ICL) can accumulate particulate material with aging,34,35 and the collagen fibers themselves can become thickened (V. K. Gullapalli, MD, and H. Wang, MD, unpublished data, 2003). Perhaps these changes underlie the poor survival of cells on the deep ICL at 7 days. Thus, the differences in cell behavior within explants and between donors may reflect differences in composition or accumulation of aging changes in Bruch’s membrane.

By using cultured fetal RPE that grows robustly in culture, the explant system eliminates RPE-related factors such as cell variability seen in cultured aged adult RPE that might lead to poor growth. In contrast to the nearly homogenous appearance of the cells on BCE-ECM, defects in RPE coverage can be present on all three of the surfaces of aged submacular Bruch’s membrane explants studied. The presence of RPE defects on explants with viable surrounding cells indicates that the defects are directly related to RPE–Bruch’s membrane interactions. Additionally, published studies using a similar organ explant culture system show that the Bruch’s membrane explant can support in situ aged RPE survival, migration, and proliferation.24 Thus, it is reasonable to attribute poor RPE survival observed in the present study to the Bruch’s membrane surface on which the cells grew. We noted that cell morphology appeared to be best if the RPE cells were attached onto basal laminar deposit rather than directly to RPE basement membrane or the ICL.

In the AMD donor explants, the areas not covered by RPE cells were mostly limited to the submacular area, where the drusen were located (Figure 10). The cells surrounding the defects in RPE coverage of Bruch’s membrane were similar in appearance to aged adult RPE cells resurfacing mechanically debrided aged submacular Bruch’s membrane explants.24 The cells at the edge of the RPE defects were either thickened and oriented parallel to the defect perimeter or extended very flat lamellipodia into the uncovered region of Bruch’s membrane. The inability of fetal RPE to attach and survive on or resurface the defects (the latter having been created from cells either not attaching or dying after attachment) indicates that RPE resurfacing of AMD eyes via transplantation or wound healing will be impaired. The inability of even cultured fetal RPE to resurface portions of aged submacular Bruch’s membrane implies that even if the aged adult RPE cells were robust (as the fetal cells are), they would not be able to resurface iatrogenic RPE defects created in patients with AMD after CNV excision.


The inability of aged submacular human Bruch’s membrane to support transplanted RPE survival in organ culture is consistent with the histologic findings from a patient who received a transplant of uncultured aged adult RPE.7 In this immune suppressed patient, transplanted RPE did not fully resurface the localized RPE defect, and the RPE cells remaining in the subretinal space were not directly in contact with Bruch’s membrane but were resting on the residual portions of the CNV. Within the time course of the current organ culture study (7 days), we observed RPE cell death; RPE defects formed with surrounding RPE unable to resurface the defects, and remaining RPE appeared relatively undifferentiated compared to their behavior on BCE-ECM. Resurfacing of two donor eyes with AMD was the poorest we observed. These results indicate that modification of Bruch’s membrane may be necessary to support RPE survival and differentiation in the context of RPE transplantation in AMD patients undergoing CNV excision.


Dr Susan G. Elner

Dr Zarbin and co-workers present their study evaluating the ability of fetal human RPE cells to attach and differentiate on aged Bruch’s membrane explants from donors aged 55–75 years. Mechanical debridement was used to expose varied layers of Bruch’s membrane to simulate the substrate that may result following surgical excision of choroidal neovascular membranes and upon which RPE cells might be transplanted. The RPE nuclear density, cell morphology, and cell coverage were evaluated at one- and seven-day time points. Bovine corneal endothelial-extracellular matrix was used as the positive control substrate. In this study, the authors demonstrated that transplanted fetal RPE could resurface debrided RPE-basement membrane or the inner collagenous layer of Bruch’s membrane, but that the transplanted fetal RPE failed to grow by Day 7, with worsening cell morphology and decreased nuclear counts compared to control. The authors conclude that aged Bruch’s membrane does not support transplanted RPE survival.

The design of this study, however, still leaves some question whether it is the “aged” Bruch’s membrane that is solely responsible for the failure of the RPE to survive and thrive. Inclusion of Bruch’s membrane from young donors as a control in this study rather than bovine corneal endothelial cell-extracellular matrix as control would strengthen the authors’ conclusions. The issue of young versus aged Bruch’s membrane has previously been investigated by Ho and co-workers,1 who studied RPE attachment to Bruch’s membrane obtained from young (average age 32 years) compared to older (average age 76 years) donors. Higher RPE reattachment rates after 6 hours were found on younger Bruch’s membrane compared to older donors (64+2.5% vs. 52.4+3.6%). Subsequent studies by Del Priore and co-workers2 and Tezel and co-workers3 failed to demonstrate a significant difference in RPE attachment rates to intact Bruch’s membrane between young and old donors. A significant reduction in RPE attachment/resurfacing to older Bruch’s membrane that underwent debridement to deeper collagenous and elastin layers, however, was seen. This later finding correlates with the findings of poor RPE resurfacing on debrided Bruch’s membrane as reported by Dr Zarbin in this study.

With age, many changes occur in Bruch’s membrane, including the deposition of peroxidized lipids, esterified cholesterol, triglycerides and phospholipids.4 Increased advanced glycation end products (AGEs), collagen type I, matrix metalloproteinases -2 and -95 have also been noted with age. However, collagen type IV, laminin, and fibronectin have all been noted to decrease in the RPE basement membrane with aging, particularly in areas of drusen formation.4 The question remains: what are the changes in aged Bruch’s membrane that reduce RPE attachment and survival? Is it the deposition of extraneous materials or the reduction of possible important RPE attachment factors that reduces RPE resurfacing and survival?

This study continues the ongoing work of Dr Zarbin and colleagues toward advancing our knowledge of RPE-Bruch’s membrane interactions with the anticipated application of this knowledge eventually to successful RPE transplantation in age-related macular degeneration. I would like to thank the authors for providing their manuscript with ample time for review and congratulate them for their ongoing research, which may eventually benefit many patients afflicted with age-related macular degeneration.


1. Ho TC, Del Priore LV. Reattachment of cultured human retinal pigment epithelium to extracellular matrix and human Bruch’s membrane. Invest Ophthalmol Vis Sci. 1997;38:1110–1118. [PubMed]
2. Del Priore LV, Tezel TH. Reattachment of human retinal pigment epithelium to layers of human Bruch’s membrane. Arch Ophthalmol. 1998;116:335–341. [PubMed]
3. Tezel TH, Kaplan HJ, Del Priore LV. Fate of human retinal pigment epithelial cells seeded onto layers of human Bruch’s membrane. Invest Opthalmol Vis Sci. 1999;40:467–476. [PubMed]
4. Pauleikhoff P, Wojtecki S, Muller D, et al. Adhesion molecules and lipid accumulation in Bruch’s membrane with age. Invest Ophthalmol Vis Sci. 1999;40:S920.
5. Guo L, Hussain AA, Limb GA, et al. Age-dependent variation in metalloproteinase activity of isolated human Bruch’s membrane and choroid. Invest Ophthalmol Vis Sci. 1999;40:2676–2682. [PubMed]

Dr Marco A. Zarbin

I thank Dr Elner for her careful evaluation of our work. We agree that it is essential to carry out comparative experiments using young versus old submacular Bruch’s membrane to determine whether aging changes in Bruch’s membrane or some other property of Bruch’s membrane underlies poor cultured fetal human RPE survival on aged submacular human Bruch’s membrane. It has been difficult for us to obtain young tissue that is in good condition, but we are undertaking these studies currently. I would like to emphasize that our studies focus not on RPE cell attachment, but on RPE survival and differentiation following initial attachment. The conditions of our experiments are such that there is 100 percent coverage of Bruch’s membrane by cultured fetal RPE cells 24 hours after seeding the cells in organ culture, even on aged submacular human Bruch’s membrane. Over time (e.g., seven or 14 days after seeding), however, these cells die, which leads to a decline in RPE nuclear density (in contrast to results seen on bovine corneal endothelial cell extracellular matrix-coated dishes). We are now trying to delineate (and reverse) the sequence of events in which initial successful attachment to Bruch’s membrane is followed by the unexpected adverse outcome of RPE cell oncosis.


This work was supported by grants from the Foundation Fighting Blindness, Owings Mills, Maryland; Research to Prevent Blindness, Inc, New York, New York (unrestricted grant); the Foundation of the University of Medicine and Dentistry of New Jersey, Newark, New Jersey; the Lions Eye Research Foundation of New Jersey, Newark, New Jersey; the Eye Institute of New Jersey, Newark, New Jersey; the William G. and Helen C. Hoffman Foundation, Summit, New Jersey; the National Eye Institute, National Institutes of Health (grant NEI 09750); and the Janice Mitchell Vassar and Ashby John Mitchell Fellowship; and by a gift from Mrs Georgaynne Holst-Knudsen.


1. Thomas MA, Kaplan HJ. Surgical removal of subfoveal neovascularization in the presumed ocular histoplasmosis syndrome. Am J Ophthalmol. 1991;111:1–7. [PubMed]
2. Nasir MA, Sugino I, Zarbin MA. Decreased choriocapillaris perfusion following surgical excision of choroidal neovascular membranes in age-related macular degeneration. Br J Ophthalmol. 1997;81:481–489. [PMC free article] [PubMed]
3. Castellarin AA, Nasir MA, Sugino IK, et al. Progressive presumed choriocapillaris atrophy after surgery for age-related macular degeneration. Retina. 1998;18:143–149. [PubMed]
4. Hsu JK, Thomas MA, Ibanez H, et al. Clinicopathologic studies of an eye after submacular membranectomy for choroidal neovascularization. Retina. 1995;15:43–52. [PubMed]
5. Rosa RH, Thomas MA, Green WR. Clinicopathologic correlation of submacular membranectomy with retention of good vision in a patient with age-related macular degeneration. Arch Ophthalmol. 1996;114:480–487. [PubMed]
6. Algvere PV, Berglin L, Gouras P, et al. Transplantation of fetal retinal pigment epithelium in age-related macular degeneration with subfoveal neovascularization. Graefes Arch Clin Exp Ophthalmol. 1994;232:707–716. [PubMed]
7. Del Priore LV, Kaplan HJ, Tezel TH, et al. Retinal pigment epithelial cell transplantation after subfoveal membranectomy in age-related macular degeneration: clinicopathologic correlation. Am J Ophthalmol. 2001;131:472–480. [PubMed]
8. Binder S, Stolba U, Krebs I, et al. Transplantation of autologous retinal pigment epithelium in eyes with foveal neovascularization resulting from age-related macular degeneration: a pilot study. Am J Ophthalmol. 2002;133:215–225. [PubMed]
9. Li LX, Turner JE. Inherited retinal dystrophy in the RCS rat: prevention of photoreceptor degeneration by pigment epithelial cell transplantation. Exp Eye Res. 1988;47:911–917. [PubMed]
10. Coffey PJ, Girman S, Wang SM, et al. Long-term preservation of cortically dependent visual function in RCS rats by transplantation. Nat Neurosci. 2002;5:53–56. [PubMed]
11. Castillo BV, Jr, del Cerro M, White RM, et al. Efficacy of nonfetal human RPE for photoreceptor rescue: a study in dystrophic RCS rats. Exp Neurol. 1997;146:1–9. [PubMed]
12. Lund RD, Adamson P, Sauve Y, et al. Subretinal transplantation of genetically modified human cell lines attenuates loss of visual function in dystrophic rats. Proc Natl Acad Sci U S A. 2001;98:9942–9947. [PMC free article] [PubMed]
13. Guymer R, Luthert P, Bird A. Changes in Bruch’s membrane and related structures with age. Prog Retin Eye Res. 1999;18:59–90. [PubMed]
14. Marshall J, Hussain AA, Starita C, et al. Aging and Bruch’s membrane. In: Marmor MF, Wolfensberger TJ, eds. The Retinal Pigment Epithelium. New York: Oxford University Press; 1998:669–692.
15. Abdelsalam A, Del Priore L, Zarbin MA. Drusen in age-related macular degeneration: pathogenesis, natural course, and laser photocoagulation-induced regression. Surv Ophthalmol. 1999;44:1–29. [PubMed]
16. Krishnamurti U, Rondeau E, Sraer JD, et al. Alterations in human glomerular epithelial cells interacting with nonenzymatically glycosylated matrix. J Biol Chem. 1997;272:27966–27970. [PubMed]
17. Anderson SS, Kim Y, Tsilibary EC. Effects of matrix glycation on mesangial cell adhesion, spreading and proliferation. Kidney Int. 1994;46:1359–1367. [PubMed]
18. Paul RG, Bailey AJ. The effect of advanced glycation end-product formation upon cell-matrix interactions. Int J Biochem Cell Biol. 1999;31:653–660. [PubMed]
19. Grossniklaus HE, Green WR. Histopathologic and ultra-structural findings of surgically excised choroidal neovascularization. Submacular Surgery Trials Research Group. Arch Ophthalmol. 1998;116:745–749. [PubMed]
20. Song MK, Lui GM. Propagation of fetal human RPE cells: preservation of original culture morphology after serial passage. J Cell Physiol. 1990;143:196–203. [PubMed]
21. Burke JM, Skumatz CM, Irving PE, et al. Phenotypic heterogeneity of retinal pigment epithelial cells in vitro and in situ. Exp Eye Res. 1996;62:63–73. [PubMed]
22. McKay BS, Burke JM. Separation of phenotypically distinct subpopulations of cultured human retinal pigment epithelial cells. Exp Cell Res. 1994;213:85–92. [PubMed]
23. Zarbin MA. Analysis of retinal pigment epithelium integrin expression and adhesion to aged submacular human Bruch’s membrane. Trans Am Ophthalmol Soc. 2003;101:499–520. [PMC free article] [PubMed]
24. Wang H, Ninomiya Y, Sugino IK, et al. Retinal pigment epithelium wound healing on human Bruch’s membrane explants. Invest Ophthalmol Vis Sci. 2003;44:2199–2210. [PubMed]
25. Zarbin MA, Sugino IK, Castellarin A. RPE transplantation for age-related macular degeneration. In: Fine S, Berger J, Maguire M, eds. Age-related Macular Degeneration. Philadelphia: Mosby-Year Book Inc; 1999:363–382.
26. Zarbin MA. Current concepts in the pathogenesis of age-related macular degeneration. Arch Ophthalmol. 2004;122:598–614. [PubMed]
27. Ho TC, Del Priore LV. Reattachment of cultured human retinal pigment epithelium to extracellular matrix and human Bruch’s membrane. Invest Ophthalmol Vis Sci. 1997;38:1110–1118. [PubMed]
28. Tsukahara I, Ninomiya S, Castellarin A, et al. Early attachment of uncultured retinal pigment epithelium from aged donors onto Bruch’s membrane explants. Exp Eye Res. 2002;74:255–266. [PubMed]
29. Castellarin AA, Sugino IK, Vargas JA, et al. In vitro transplantation of fetal human retinal pigment epithelial cells onto human cadaver Bruch’s membrane. Exp Eye Res. 1998;66:49–67. [PubMed]
30. Del Priore LV, Tezel TH. Reattachment rate of human retinal pigment epithelium to layers of human Bruch’s membrane. Arch Ophthalmol. 1998;116:335–341. [PubMed]
31. Tezel TH, Kaplan HJ, Del Priore LV. Fate of human retinal pigment epithelial cells seeded onto layers of human Bruch’s membrane. Invest Ophthalmol Vis Sci. 1999;40:467–476. [PubMed]
32. Tezel TH, Del Priore LV. Repopulation of different layers of host human Bruch’s membrane by retinal pigment epithelial cell grafts. Invest Ophthalmol Vis Sci. 1999;40:767–774. [PubMed]
33. Del Priore LV, Geng L, Tezel TH, et al. Extracellular matrix ligands promote RPE attachment to inner Bruch’s membrane. Curr Eye Res. 2002;25:79–89. [PubMed]
34. Curcio CA, Millican CL, Bailey T, et al. Accumulation of cholesterol with age in human Bruch’s membrane. Invest Ophthalmol Vis Sci. 2001;42:265–274. [PubMed]
35. Ruberti JW, Curcio CA, Millican CL, et al. Quick-freeze/deep-etch visualization of age-related lipid accumulation in Bruch’s membrane. Invest Ophthalmol Vis Sci. 2003;44:1753–1759. [PubMed]

Articles from Transactions of the American Ophthalmological Society are provided here courtesy of American Ophthalmological Society
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...