|
|
||||||||
1From the Departments of Biomedical Engineering and 4Neurobiology and Physiology, Northwestern University, Evanston, Illinois; the 2Department of Ophthalmology, University of Alabama School of Medicine, Birmingham, Alabama; and the 3High Resolution Imaging Facility, University of Alabama at Birmingham, Birmingham, Alabama.
| Abstract |
|---|
|
|
|---|
METHODS. Four eyes from human donors aged 27, 41, 76, and 78 years were preserved within 4 hours of death. Full-thickness tissue blocks from the macula were prepared for TEM or for QFDE.
RESULTS. Ultrastructure seen by conventional TEM was revealed in greater detail by QFDE. Cholesterol-containing particles (mean diameter, 80 nm) formed a thin densely packed layer external to the basal lamina of the retinal pigment epithelium (RPE) only in older eyes. The mesh size of the RPE basal lamina was smaller than the particles, and it appeared larger in older eyes. QFDE also revealed less decorated collagen fibrils in older eyes.
CONCLUSIONS. The data suggest that the predilection of a extremely thin sublayer of inner Bruchs membrane for accumulating lipid particles may eventually lead to a confluent lipid wall capable of isolating the retina from its blood supply. If these lipids originate in the retinal pigment epithelium, then they are unlikely to have passed through the basal lamina in this form. The age-related increase in lipid particles corresponds with an age-related increase in hydraulic resistance determined in excised Bruchs membrane/choroid by others. QFDE will be useful for future modeling studies of Bruchs membrane transport and to identify those moieties responsible for deleterious age-related transport changes in Bruchs membrane.
Unlike age-related accumulation of cholesterol in other connective tissues (cornea, sclera, and arterial intima),16 17 18 19 20 that which occurs in Bruchs membrane is apparently unique in its potential impact on fluid and nutrient exchange. According to the lipid-barrier hypothesis, this accumulation renders Bruchs membrane increasingly hydrophobic, impeding translocation of hydrophilic substances to and from the choriocapillaris.7 21 22 Impaired movement of fluids, either derived from RPE or from choroidal neovascularization, may contribute to the formation of RPE detachments in patients with ARM.22 It may also prevent adequate replenishment of photoreceptors with plasma-derived retinoid derivatives, thus contributing to age- and ARM-related photoreceptor dysfunction.23
The lipid-barrier hypothesis would be strengthened by a morphologic demonstration of this barrier in human Bruchs membrane. Empirical studies24 25 suggest a two-phase decline in the hydraulic conductivity of isolated Bruchs membrane with a marked decline between birth and 40 years of age and a slower decline after that age. However, the structural basis for this decline has not been established. The decline in hydraulic conductivity in younger eyes has been attributed to increased collagen cross-linking,24 but no ultrastructural evidence supporting this idea is yet available. Throughout adulthood, Bruchs membrane accumulates solid particles containing esterified and unesterified cholesterol, roughly 100 nm in diameter, that form a discrete sublayer external to the RPE basal lamina in eyes of many persons older than 60 years.2
Determining how esterified cholesterol-containing particles interact with each other and extracellular matrix components to form a hydrophobic barrier and reconciling the different time courses of changes in the hydraulic conductivity and lipid content of Bruchs membrane would be facilitated by a method that preserves ultrastructure of lipids and extracellular matrix. The ultrastructure of Bruchs membrane has been examined extensively by conventional thin-section transmission electron microscopy (TEM), but the dehydration and embedding steps of tissue processing optimally preserve neither the small lipid particles nor the ultrastructure of the extracellular matrix.26 27 Bruchs membrane has also been examined using the osmium-tannic acid-paraphenylenediamine method to preserve lipid particles, but this method does not improve visualization of extracellular matrix components.2
Therefore, we examined macular Bruchs membrane of normal human eyes by quick-freeze/deep-etch (QFDE) microscopy, a method that provides a quasi-three-dimensional view of connective tissue ultrastructure and lipid particles in exquisite detail.
| Methods |
|---|
|
|
|---|
Maculas, including retina, choroid, and sclera, were dissected into 2 x 2-mm2 blocks and stored in 0.1 M phosphate buffer. One block adjacent to the fovea from each eye was prepared for light microscopic and conventional thin-section TEM examination. These tissue blocks were postfixed for 1 hour at 25°C in 1% osmium and 0.125% potassium ferricyanide, dehydrated through ethanol and propylene oxide, and embedded in resin (PolyBed 812; Polysciences, Warrington, PA). Blocks were then sectioned at 1 µm in the vertical or oblique planes and stained with 1% toluidine blue and sodium borohydrate. The 1-µm sections were evaluated for disease. Thin sections from each tissue block were cut and poststained with uranyl acetate and lead citrate for thin-section TEM examination.
The other macular tissue blocks were processed for QFDE. For each specimen, the retina was lifted from the underlying RPE. The specimen was placed on an aluminum tab and slam frozen (Gentleman Jim Slam Freezing Apparatus28 29 ) RPE side down, onto a copper block cooled to liquid nitrogen temperature (-195°C). The frozen specimen was then stored in liquid nitrogen until further use.
After transfer into a freeze-fracture apparatus (model CFE-40; Cressington Scientific Instruments, Watford, UK), the specimen was fractured obliquely through RPE cells, etched for 25 minutes at -95°C, and rotary shadowed with a platinum-carbon mixture at a 20° angle and then strengthened with carbon evaporated from a 90° angle overhead. In accordance with the method of Ou et al.,30 the replicated specimen was coated with collodion in 2% amyl acetate and placed in household bleach for digestion overnight. The bleach-cleaned replica was picked up on either nickel or copper mesh hexagonal grids and washed in amyl acetate to remove the collodion.
Thin sections and replicas were viewed on two transmission electron microscopes (model 1200 EX II and JEM-100 CX TEMSCAN microscopes, respectively; JEOL, Tokyo, Japan). An oblique fracture plane passing nearly parallel to the retina, as shown in Figure 1 , gave the best view of Bruchs membrane ultrastructure.
|
| Results |
|---|
|
|
|---|
Oblique sections of Bruchs membrane, as viewed by TEM, are shown for a 27-year-old eye (Fig. 2A) , a 41-year-old eye (Fig. 3A) , a 76-year-old eye (Fig. 3C) , and a 78-year-old eye (Fig. 2C) . In the younger eyes, the basal infoldings of the RPE, the RPE basal lamina and the inner collagenous layer (ICL) are apparent (Figs. 2A 3A) . Collagen fibrils and an unidentified banded material were apparent within the ICL. The ends of collagen fibrils could be seen closely apposed to, perhaps embedded in, the RPE basal lamina (Fig. 3A) . Consistent with previous TEM descriptions, the most prominent differences between the older (Figs. 2C 3C) and younger (Figs. 2A 3A) eyes were disorganization or disappearance of the RPE basal infoldings, increased electron density of the ICL, a paucity of collagen fibrils immediately adjacent to the RPE basal lamina, and the presence of numerous round electron-lucent profiles. These profiles were distributed along fibrils in the ICL and in a band between the ICL and the RPE basal lamina in the older donors (Figs. 2C 3C) . This band, two to four particles thick in cross-sections of many older eyes,2 looks wide in Figures 2C and 3C , because of the oblique plane of section. These profiles were previously identified as cholesterol-containing particles, because they are solid when viewed with lipid-preserving ultrastructural techniques, they are extractable with lipid solvents, and they become more numerous with age along with histochemically detectable and directly assayed esterified and unesterified cholesterol.2
|
|
More lipid particles appear embedded in the basal lamina in the thin-section TEM images (Figs. 2C 3C) than in the QFDE images (Figs. 2D 3D) . The apparent discrepancy between the two imaging modalities can be reconciled by considering Figure 2A , in which RPE basal infoldings also appear to be embedded in the basal lamina. This underscores that sections imaged by thin-section TEM are sufficiently thick to permit the superimposition of two very thin sublayers. Therefore it is likely that Figures 2C and 3C represent superimposition of the RPE basal lamina and the adjacent layer of Bruchs membrane rather than numerous particles embedded in basal lamina.
The main difference between the two older eyes examined was the number of these lipid particles. At the interface between the basal lamina and the ICL of the 76-year-old donor eye was a wall of tightly packed lipid particles appreciable by both thin-section TEM (Fig. 3C) and QFDE (Fig. 3D) . This wall was not apparent in the 78-year-old eye (Figs. 2C 2D) . Figure 4 demonstrates the high density of these embedded lipid rich particles showing a region where the fracture plane jumped from the basal lamina down to the ICL (the bottom of the groove) and then back up to the basal lamina.
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
Our most prominent finding is the markedly improved visualization of cholesterol-containing particles accumulated in a discrete sublayer external to the RPE basal lamina, a region we call the lipid wall. Previous work has shown an age-related increase of esterified cholesterol by filipin histochemistry and the presence of solid lipid particles by a lipid-preserving ultrastructural method in this region.2 Notably, the present study showed the packing of lipid particles to be so dense that only a small fraction of the volume of this layer appeared to be available for transport. Marshall et al.22 determined that the hydraulic resistance (the inverse of the hydraulic conductivity) of Bruchs membrane was closely correlated to lipid accumulation. Our data support the possibility that the age-related increase of the hydraulic resistance of Bruch s membrane22 is closely related to formation of the lipid wall. A detailed transport analysis has yet to be conducted, but transport of water and hydrophilic moieties across Bruchs membrane would be expected to be considerably compromised by the barrier shown in Figures 3C 3D and 4 .
In this investigation, QFDE revealed previously unappreciated details of lipid particle substructure, showing both exterior and interior components (Fig. 5) . Lipid particles with external shells were also seen in QFDE studies of cholesterol accumulation in the aortic intima of rabbit after a bolus injection of lipoproteins.33 34 However, those particles were rich in unesterified cholesterol and contained little esterified cholesterol,35 36 37 38 which is abundant in Bruchs membrane lipid particles. In our study, the shell on the particles exterior may represent a small component of unesterified cholesterol solubilized in phospholipid.26 39 Alternatively, this morphology could signify surface and core components typical of lipoprotein particles.40 41 The significance of lipid particle morphology revealed by QFDE will become clearer when these particles are isolated and characterized.
The source of the cholesterol-containing particles found in Bruchs membrane is unknown. A high proportion of esterified cholesterol in relation to total cholesterol is consistent with direct infusion of plasma low-density lipoproteins (LDLs) into Bruchs membrane, a process that occurs in other connective tissues.2 However, preliminary analysis indicates that the distribution of cholesteryl esters in Bruchs membrane does not resemble that of plasma LDL.42 Further, human RPE has potential for biosynthesis of apolipoprotein B (apoB), the principal protein of very-low-density lipoprotein (and its metabolic product, LDL) and intestinal chylomicrons, and apoB immunoreactivity can be found in normal Bruchs membrane.43 These data raise the possibility that the RPE assembles large apoB-containing lipoproteins for basal release into Bruchs membrane. Definitive identification of the source of esterified cholesterol in Bruchs membrane clearly requires further study.
Nevertheless, the details of where these lipid rich particles are found in relation to the RPE and its basal lamina, as shown in our study, are important for evaluating possible mechanisms by which cholesterol arrives in Bruchs membrane. The size of the lipid-containing particles (roughly 80100 nm) is much greater than the typical interfibrillar spacing in the basal lamina (Figs. 2B 2D 3B 3D 7) , and few particles were detected within the basal lamina. Assuming the basal lamina to be continuous and to be incapable of deforming sufficiently to allow the passage of these large particles, this suggests that if these particles originate in the RPE, they are unlikely to have passed through the basal lamina in this form. One possibility is that very small lipid particles pass through the basal lamina and coalesce to form the large particles that constitute the lipid wall. Another possibility is that the particles deform as they squeeze through the basal lamina of the RPE. Chylomicrons, lipoprotein particles more than 75 nm in diameter formed by intestinal enterocytes, are known to pass through a basal lamina en route to lymphatic channels.44 A third possibility is that these particles are derived from the choriocapillaris and do not pass through the RPE basal lamina, although, as mentioned, recent biochemical analysis has cast doubt on the hypothesis that these particles are exclusively blood derived.42 43
Although quantitative data are needed from more eyes to confirm our observations, our findings allow us to evaluate the potential contribution of age-related differences in extracellular matrix components to the observed age-related changes in hydraulic conductivity of Bruchs membrane. Previous examination of the age-related decrease in hydraulic conductivity of Bruchs membrane has emphasized fibrillar collagen in the ICL of Bruchs membrane.22 However, preliminary hydrodynamic analyses indicate that the collagen fibrils of this layer would present negligible hydrodynamic flow resistance.45 In the present study, we saw that labeling of individual fibers in this region was decreased in the older eyes (Fig. 8) , a finding that is not consistent with an age-related decrease in hydraulic conductivity. Our QFDE images (Figs. 2B 2D 3B 2D 4 6) indicate that the weave of the RPE basal lamina and staining of collagen fibrils in this region should be considered contributors to resistance as well, because the interfibrillar spaces are much smaller than those seen in the ICL. However, the RPE basal lamina does not appear any more tightly woven in older adults, and it is possibly even less tight than it is in the eyes of the younger adults we examined (Figs. 6 7) . Thus, this finding is also inconsistent with an age-related decrease in hydraulic conductivity.
It is logical to hypothesize that transport of water and hydrophilic moieties across Bruchs membrane in the older eyes examined in this study would be compromised by the lipid wall. The relationship of lipid accumulation in Bruchs membrane to the formation of basal linear deposits and drusen and the progression to neovascularization remains to be determined. However, the degeneration and death of RPE and photoreceptors in the overlying retina are aspects of ARM pathobiology that could be consequences of the development of the physical barrier described in this study.23 The precise physicochemical circumstances that allow lipid nucleation and aggregation confined to this narrow layer in the extracellular matrix must be investigated to elucidate how a lipid barrier could contribute to the pathogenesis of late ARM. A detailed engineering analysis of tissue transport properties based on high quality morphologic information afforded by QFDE has recently been achieved for human cornea.27 Our current results indicate that a similar approach to understanding transport across Bruchs membrane is now warranted.
| Footnotes |
|---|
Submitted for publication May 24, 2002; revised November 12, 2002; accepted December 6, 2002.
Commercial relationships policy: N.
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C.
1734 solely to indicate this fact.
Corresponding author: Mark Johnson, Department of Biomedical Engineering, Northwestern University, 2145 Sheridan Road, Evanston, IL 60208; m-johnson2{at}northwestern.edu.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J.-D. Huang, C. A. Curcio, and M. Johnson Morphometric Analysis of Lipoprotein-like Particle Accumulation in Aging Human Macular Bruch's Membrane Invest. Ophthalmol. Vis. Sci., June 1, 2008; 49(6): 2721 - 2727. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Rudolf, M. E. Clark, M. F. Chimento, C.-M. Li, N. E. Medeiros, and C. A. Curcio Prevalence and Morphology of Druse Types in the Macula and Periphery of Eyes with Age-Related Maculopathy Invest. Ophthalmol. Vis. Sci., March 1, 2008; 49(3): 1200 - 1209. [Abstract] [Full Text] [PDF] |
||||
![]() |
N Dashti, G McGwin, C Owsley, and C A Curcio Plasma apolipoproteins and risk for age related maculopathy Br. J. Ophthalmol., August 1, 2006; 90(8): 1028 - 1033. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-M. Li, M. E. Clark, M. F. Chimento, and C. A. Curcio Apolipoprotein localization in isolated drusen and retinal apolipoprotein gene expression. Invest. Ophthalmol. Vis. Sci., July 1, 2006; 47(7): 3119 - 3128. [Abstract] [Full Text] [PDF] |
||||
![]() |
B Y Ishida, K G Duncan, K R Bailey, J P Kane, and D M Schwartz High density lipoprotein mediated lipid efflux from retinal pigment epithelial cells in culture Br. J. Ophthalmol., May 1, 2006; 90(5): 616 - 620. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. C. Geiger, C. M. Waters, D. W. Kamp, and M. R. Glucksberg KGF Prevents Oxygen-Mediated Damage in ARPE-19 Cells Invest. Ophthalmol. Vis. Sci., September 1, 2005; 46(9): 3435 - 3442. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-M. Li, B. H. Chung, J. B. Presley, G. Malek, X. Zhang, N. Dashti, L. Li, J. Chen, K. Bradley, H. S. Kruth, et al. Lipoprotein-like Particles and Cholesteryl Esters in Human Bruch's Membrane: Initial Characterization Invest. Ophthalmol. Vis. Sci., July 1, 2005; 46(7): 2576 - 2586. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-M. Li, J. B. Presley, X. Zhang, N. Dashti, B. H. Chung, N. E. Medeiros, C. Guidry, and C. A. Curcio Retina expresses microsomal triglyceride transfer protein: implications for age-related maculopathy J. Lipid Res., April 1, 2005; 46(4): 628 - 640. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Itaya, V. Gullapalli, I. K. Sugino, M. Tamai, and M. A. Zarbin Iris Pigment Epithelium Attachment to Aged Submacular Human Bruch's Membrane Invest. Ophthalmol. Vis. Sci., December 1, 2004; 45(12): 4520 - 4528. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. H. Tezel, L. V. Del Priore, and H. J. Kaplan Reengineering of Aged Bruch's Membrane to Enhance Retinal Pigment Epithelium Repopulation Invest. Ophthalmol. Vis. Sci., September 1, 2004; 45(9): 3337 - 3348. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Gordiyenko, M. Campos, J. W. Lee, R. N. Fariss, J. Sztein, and I. R. Rodriguez RPE Cells Internalize Low-Density Lipoprotein (LDL) and Oxidized LDL (oxLDL) in Large Quantities In Vitro and In Vivo Invest. Ophthalmol. Vis. Sci., August 1, 2004; 45(8): 2822 - 2829. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Y. Ishida, K. R. Bailey, K. G. Duncan, R. J. Chalkley, A. L. Burlingame, J. P. Kane, and D. M. Schwartz Regulated expression of apolipoprotein E by human retinal pigment epithelial cells J. Lipid Res., February 1, 2004; 45(2): 263 - 271. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |