June 2004
Volume 45, Issue 6
Free
Anatomy and Pathology/Oncology  |   June 2004
Cytoarchitecture of Choroidal Capillary Endothelial Cells
Author Affiliations
  • Robyn H. Guymer
    From the Centre for Eye Research Australia, University of Melbourne, Victoria, Australia;
  • Alan C. Bird
    The Institute of Ophthalmology, University College, London, United Kingdom; and
  • Gregory S. Hageman
    The University of Iowa Center for Macular Degeneration, Department of Ophthalmology, University of Iowa, Iowa City, Iowa.
Investigative Ophthalmology & Visual Science June 2004, Vol.45, 1660-1666. doi:https://doi.org/10.1167/iovs.03-0913
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Robyn H. Guymer, Alan C. Bird, Gregory S. Hageman; Cytoarchitecture of Choroidal Capillary Endothelial Cells. Invest. Ophthalmol. Vis. Sci. 2004;45(6):1660-1666. https://doi.org/10.1167/iovs.03-0913.

      Download citation file:


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

      ×
  • Supplements
Abstract

purpose. To describe the cytoarchitecture of the choroidal capillary endothelial cells, especially as it relates to cellular processes that protrude through the basal lamina into Bruch’s membrane (BM).

methods. Human donor eyes and monkey and hamster eyes were examined by transmission electron microscopy and freeze-fracture replication. The number of endothelial cell processes and the characteristics of the processes and surrounding structures were determined in the maculae of human eyes and correlated with age-related changes in neighboring structures.

results. Endothelial cell processes were observed in eyes of all species examined and at all ages. They typically occurred at sites of focally thickened, nonfenestrated regions of the endothelial cells. The basal lamina adjacent to the processes was often hypertrophic and associated deposits of long-spacing collagen (LSC) were observed frequently. In humans, there was no correlation between the number of processes per 100 μm of the BM with age, sex, cause of death, postmortem time, RPE autofluorescence, or RPE residual body content. There was a weak linear association with the thickness of the BM.

conclusions. The finding of cellular processes of choroidal capillary endothelial cells penetrating their basal laminae is normal. These processes may serve to stabilize choroidal endothelial cells physically and play an important structural role in the maintenance of patency of the choriocapillaris. It is also possible that they have additional functions, as suggested for similar processes in other tissues. They are not necessarily the harbinger of choroidal neovascularization, although growth of new vessels may result from distortion of this normal attribute.

Choroidal capillary endothelial cell processes penetrating Bruch’s membrane have been reported in a variety of species. 1 2 3 4 5 These processes were illustrated clearly in humans in one of the first electron microscopic studies of Bruch’s membrane and the choriocapillaris, although there was no reference to them in the text. 1 They were subsequently described in rats as occasional pegs of cytoplasm and in the chick, in which they often extended to the RPE. They have subsequently been reported in the rabbit, hamster, and hedgehog. 2 3 5  
The first systematic examination of these structures in human eyes was conducted by Yamamoto and Yamashita 4 6 and Yamamoto et al. 7 in 47 eyes of human donors aged 0 to 91 years. The processes are described as pseudopodia of the choroidal capillary endothelium containing actin filaments, ribosomes, and endoplasmic reticulum-like organelles. Pseudopodia were found in every age group and were situated preferentially on the RPE aspect of the choroidal capillaries. There was no correlation between the frequency of pseudopodia and age, sex, race, postmortem time, location in the fundus, or age-related changes in Bruch’s membrane. Several donor eyes had ocular tumors, but it was concluded that these processes existed in normal eyes of all ages and that the lack of correlation with age implied that this process was unrelated to choroidal neovascularization, for which age is a major determinant. They suggested that the pseudopodia might be involved in the deposition of vesicular structures in Bruch’s membrane. Others have proposed that they may be important in cellular remodeling in response to changing functional demands of the surrounding local environment or facilitate transfer of metabolites between the intravascular spaces and the RPE. 2 5 8  
Cellular processes of endothelial cells penetrating basal lamina are not unique to the choroid, and terms such as cytoplasmic pegs, cytoplasmic processes, microvilli, pseudopodia, and footlike processes have been used to denote similar structures in different tissues; the term cell processes is used in this article. 9 10 11 12 13 14  
Because these structures have been identified in rats after photic injury and one of the initial steps of angiogenesis is dissolution of the basal lamina, it has been suggested that these structures may represent nascent choroidal neovascularization. 15 16 17 18 However, when the observations after laser photocoagulation were repeated and compared with controls, both groups of rats were found to have cellular processes of the choriocapillaris passing through their basal lamina toward the RPE, indicating that this is likely to be a normal feature of rat choriocapillaris. 19  
We sought to examine the distribution and substructure of these capillary endothelial cell processes more thoroughly, using electron microscopy (EM), and to document their frequency as a function of age in human eyes. We used predominantly human eye bank eyes, but we also used eyes obtained immediately after exenteration to ensure that the phenomenon was not due to postmortem artifact. We also examined monkeys and hamsters to establish their presence in other species that do not exhibit choroidal neovascularization and to describe their cytoarchitecture. 
Materials and Methods
Tissues
Sixty-five unpaired human donor eyes (donor ages, 7 to 87 years) were obtained from the Eye Banks of the Moorfields and United Kingdom Transplant Service (Bristol, UK) within 72 hours of death. No information regarding previous eye disease was available, but eyes with macroscopically visible macular disease (e.g., scar, vessel occlusion) were excluded from the study. The globes were hemisected by a circumferential incision through the pars plana. Macular discs, centered on the fovea, were obtained using an 8-mm trephine. A portion of each macular disc was fixed in 1% paraformaldehyde and 3% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4), dehydrated in acetone and propylene oxide, and embedded in Araldite or Epon resins for electron microscopy (EM). Ultrathin sections were stained with uranyl acetate and lead citrate. Other macular segments were prepared for frozen and paraffin sections to assess age-related change in the RPE and Bruch’s membrane, the methods used, and the results obtained were published by Okubo et al., 20 and were used to determine whether any correlation existed with the cellular processes. 
A further group of three eyes were collected from three female patients (patient 1, 58 years; patient 2, 45 years; and patient 3, 61 years) immediately after exenteration for meibomian gland carcinoma. Two of these eyes had light photocoagulation applied to the inferior macula 15 hours and 5 days before surgery as part of another research project. 20 21 These eyes were fixed in 3% glutaraldehyde and 1% formaldehyde in sodium cacodylate buffer within 15 minutes of exenteration and prepared for EM, in the same manner as the eye bank eyes. The sections used in the study were taken from the superior macula, at least 4 mm from a laser burn. 
Three eyes from adult macaque monkeys were enucleated after perfusion fixation in paraformaldehyde, subsequently fixed with 3% glutaraldehyde and 1% formaldehyde in sodium cacodylate buffer, and prepared for transmission EM. The exenterated human eyes and the monkey eyes were postfixed in osmic acid. 
Institutional ethics approval was obtained, and our study complied with the tenets of the Declaration of Helsinki, with written informed consent being obtained from all subjects or their next of kin. The animal experiments were performed in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 
Tissue Analysis
Sections were examined with the electron microscope (JEOL 1010; Welwyn Garden City, Buckinghamshire, UK). Ten contiguous lengths of 100 μm of Bruch’s membrane were examined, and the number of processes was counted, measured, and photographed. A process was counted only if it clearly transgressed the basal lamina of the choriocapillaris. The average number of processes per 100 μm of Bruch’s membrane, the average total length of processes per 100 μm of Bruch’s membrane, and the longest and widest process was recorded for each eye. The presence and location of long-spacing collagen (LSC) and focal thickening of the basal lamina associated with each process were also noted. 
The data from the 65 eye bank eyes were used to examine the association between number and length of the processes with numerous variables. The number of processes in donor eyes appeared to fall within a Poisson distribution with a mean of 1.5 processes per 100 μm. Pearson’s correlation coefficient was computed to examine the associations between the number and length of the processes in the 65 eye bank eyes with each of the following: Bruch’s membrane thickness, RPE autofluorescence, and RPE residual body content, as defined by Okubo et al. 20 Our null hypothesis was that there was no linear association between these variables. To determine whether the processes were a postmortem artifact we used a two-tailed test of significance to determine whether the number of processes per unit length differed between the donor eyes and those eyes enucleated and rapidly fixed. Our null hypothesis was that the number of processes in the freshly exenterated eyes would equal the number in postmortem donor eyes. Photographs were also scrutinized to seek a relationship between LSC location, focal thickening in the basal lamina, and the site of cellular processes. 
Freeze-Fracture Replication
Fifteen normally pigmented golden hamsters (Mesocricetus auratus) were used for this study. Animals were anesthetized by intraperitoneal injection of 0.85 mg pentobarbital sodium (Nembutal; Abbott Laboratories, Abbott Park, IL) per gram of body weight in accordance with the ARVO Statement and guidelines established by the University of Southern California and St. Louis University Animal Care Committees. Hamster and human tissues used for freeze-fracture were obtained after perfusion or immersion fixation in 2% formaldehyde (generated fresh from paraformaldehyde) and 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4), containing 0.025% calcium chloride. Tissues were stored at 4°C in 0.1 M sodium cacodylate buffer. They were later cut into 100- to 150-μm thick sections with a tissue chopper (TC-2; Sorvall, Newtown, CT). Tissues were passed through solutions of 10%, 20%, and 30% glycerol and 0.1 M cacodylate buffer, over 60 to 90 minutes, and then placed between gold alloy specimen discs and frozen immediately in Freon-22 (DuPont, Wilmington, DE). Discs were placed in a double-replica specimen holder that was precooled in liquid nitrogen and subsequently transferred to a freeze-fracture unit (model 301; Balzers, Hudson, NH). 
Fracturing, shadowing, and replication were accomplished at a vacuum of at least 10−6 mm Hg and at a specimen post temperature of −119°C. The postreplication temperature was raised to −20°C in some cases before the second coat of carbon was evaporated over the fractured specimen. Replicas were prepared with minimum etching and subsequently cleaned in 100% methanol for 1 hour, 5% sodium hypochlorite for 12 to 24 hours, 50% sulfuric acid for 1 to 4 hours, and three to four rinses of distilled water. Replicas were picked up on resin-coated copper grids (Parlodion; SPI Supplies, West Chester, PA) and observed in a transmission electron microscopes equipped with goniometer stages (models 100C or 1200EX; JEOL). 
Results
General Morphologic Observations
Freeze-fracture replication of the choriocapillaris in human (not shown) and hamster eyes (Fig. 1A 1B) reveal large expanses of densely packed, 75- to 85-nm diameter fenestrae (46.3 ± 2.8/μm2) interspersed between a network of branching and anastomosing nonfenestrated membrane. The network of nonfenestrated cell membrane corresponded to a system of cytoplasmic channels seen by TEM (Figs. 2 3) . Choroidal capillary endothelial cell processes were associated only with these nonfenestrated cytoplasmic channels. The apical portions of endothelial cell processes typically broke off during the replication process and were often visualized as stumps (Figs. 2A 2B 2C) . In some fortuitous fractures, the processes remained intact and were visualized extending into the pericapillary extracellular matrix or Bruch’s membrane (Fig. 2D)
Choroidal capillary endothelial cell processes arise from nonfenestrated, focally thickened regions of endothelial cells in hamster, monkey, and human eyes. These regions correspond to the cytoplasmic channels observed in freeze-fracture replicas (Figs. 2 3) . The channels were filled with bundles of unidirectionally oriented microtubules and often encircled the entire capillary (Fig. 4) . The processes were most abundant along the retinal and lateral aspects of the capillaries and were often branched (Fig. 4A) . They typically extended through the capillary basal lamina and into the pericapillary extracellular matrix or Bruch’s membrane. In the hamster, these processes often abutted, but did not penetrate, the basal lamina of the RPE (Fig. 5A) . In primates, the processes rarely extended through the elastic lamina of Bruch’s membrane (Fig. 5B) . In these processes 5- to 7-nm diameter filaments were abundant and typically terminated along dense plaques adjacent to the basal lamina (Fig. 4A) . These accumulations of filaments were even more apparent in regions in which the process passed through thickened portions of basal laminae. Marked focal thickenings and/or reduplication of the basal laminae associated with these processes were often observed in all three species (Fig. 6A) . Banded material (LSC) with a major periodicity of 100 to 110 nm was frequently associated with these thickened regions of basal lamina in humans (Fig. 6B)
Statistical Analysis
In the eye bank human eyes, the number of processes per 100 μm of Bruch’s membrane ranged between 0.7 and 4.1 with a mean of 1.6. They fell within a Poisson distribution, and the average total length of processes over 100 μm of Bruch’s membrane ranged between 0.12 and 1.5 μm. The longest individual process was 1.7 μm, and the widest process was 1.9 μm at its base. The number of processes and total length of processes per 100 μm had a clear linear relationship (Pearson’s coefficient = 0.9, P < 0.001), and for this reason further association was sought only for number of processes per unit length of Bruch’s membrane. There was no correlation between the number of processes and age, RPE autofluorescence, or RPE residual body content, as determined in sections from the same eyes by Okubo et al. 20 . These findings support our null hypothesis. There was a weak positive correlation between Bruch’s membrane thickness and the number of processes (correlation coefficient = 0.3, P = 0.016). 
The number of processes per 100 μm in the three eyes fixed immediately after exenteration had a mean of 5.7 per 100 μm in patient 1, 6.3 in patient 2, and 4.5 in patient 3. The processes ranged between 0.67 and 8.3 per 100 μm, with the average total length per 100 μm of 1.7 to 2.7 μm. There was a difference compared with the eye bank eyes with longer fixation times. The data from the eyes of patients 1 and 2 fell outside the number expected from the eye bank eyes (two-tailed test of significance: P = 0.037, P = 0.09, respectively), but the eye of patient 3 was not significantly different (two-tailed test of significance: P = 0.13). That all three sets of data were within the expected findings for eye bank eyes was highly improbable (P = 0.0004). The longest process was 1.0 μm, and the widest was 0.91 μm, which fell within the sizes measured in the eye bank eyes. These data strongly suggest that the processes are not merely postmortem artifact and indeed, to the contrary, suggest that they were even more common in the freshly exenterated eyes. 
The average number of processes per 100 μm in the three monkeys was 10.4, 3.8, and 8.4, and ranged between 2.2 and 13 per 100 μm. Two of the three monkeys had more processes per 100 μm than the eye bank and fresh human eyes. The average total length per 100 μm of Bruch’s membrane ranged between 1.4 and 3.4 μm. The longest process was 1.0 μm, and the widest was 1.0 μm, which was similar to the human eyes. 
Discussion
Function of Processes
Our results support the view that choroidal capillary endothelial cell processes protruding through the basal lamina represent a normal feature of the choriocapillaris. We do not discount, however, that these processes may be involved in choroidal neovascularization. The finding of greater densities in monkeys with Bruch’s membrane free of debris, compared with humans, suggests that the number of processes are not likely to be the result of pathologic changes in Bruch’s membrane. The presence of similar processes in human eyes fixed within 15 minutes of enucleation and in eyes of perfusion-fixed hamsters and monkeys shows that these processes are not postmortem artifacts. 
Several different functions have been postulated for similar structures in the various tissues in which they have been recognized. In the trabecular meshwork, the cells covering the trabecular surfaces have “cellular pegs” that protrude into the basal lamina. 9 As the cells of the juxtacanalicular meshwork are known to contain actin filaments in their cytoskeleton, it is possible that these processes play a role in regulating cell shape, thus influencing meshwork porosity in response to humoral and neuronal agents. 22 23  
In the kidney, the afferent and efferent glomerular arterioles and interlobular arteries have endothelial cell processes that penetrate the basal lamina and indent adjacent media cells. 19 These myoendothelial contacts may allow detection and propagation of mechanical and humoral stimuli and may play an autoregulatory role in the vascular beds. It has been postulated that the protruding part of the endothelial cells may be tension sensors that detect changes in transmural pressure, and initiate smooth muscle contraction when stretched by an increase in transmural pressure. 10 The afferent descending vasa recti (DVR) and efferent ascending vasa recti (AVR) in the renal medulla also have such processes. 11 The thin-walled, fenestrated AVR is described as being attached at its corners to surrounding structures by microvilli that project from the abluminal surface of endothelial cells of AVR and insert into the basal lamina of neighboring DVR and tubules. It is postulated that these microvilli are probably responsible for holding vessels open when interstitial fluid pressure rises above intravascular pressure by anchoring the wall of the AVR to surrounding structures, thus preventing collapse of the lumen. 11 12  
Mammalian heart muscle capillaries and small arteries and thigh muscle arterioles and precapillary sphincters have endothelial cells processes that make contact with smooth muscle, providing a system of myoendothelial junctions. 13 14 They may represent structures that stabilize the microvascular wall mechanically, as they are seen only in regions of the vessel without elastic interna or containing large areas of fenestrae. Others suggest that they may be a pathway for exchange of metabolites or facilitate transmission of hormonal substances. 23  
In a similar way, the endothelial processes of the choroidal capillary endothelial cell may stabilize physically the angioarchitecture of the inner choroid. Arising from the cytoplasmic channels that form a network around the cell, they could have a widespread influence on cell form. The consistent finding that the basal lamina is focally thickened around the base of processes, at least in the young, lends further credence to the concept that they play a structural role. The thickening of the basal lamina and microtubules would serve to spread the physical load on the cell membrane. The fact that they are described as a common feature of thin-walled, fenestrated endothelia makes this even more likely. However, if this were the case, it is somewhat surprising that focal thickening was not observed regularly in the eyes of older donors. 
Other functions for these processes have been postulated. The possibility that cellular processes from the choriocapillaris could be involved in phagocytosis may be of importance, given the large trafficking through Bruch’s membrane. This concept was introduced with respect to the choroid when cellular processes from the choroidal pericytes were described in the inner portion of Bruch’s membrane after photocoagulation of drusen. 24 It was suggested that they may play a role in the resolution of drusen after laser treatment, and this process was likened to the phagocytic activity of mesangial cells in the glomerulus. 24 Endothelial cells may also undertake this function. In the trabecular meshwork, endothelial cells are thought to phagocytose debris, and it has been suggested that the surface “microvilli” play a role in this process. 25 A similar role has also been ascribed to vascular endothelium in skin, muscle, aorta, heart, and kidney, but only in the newborn. A phagocytic role has been postulated for similar processes in the lung throughout life, and it is proposed that they are responsible for clearance of microorganisms and other material from the circulation. 26 27  
Relationship between Endothelial Processes and Neovascularization
The mechanisms that initiate and modulate the normal rate of basal lamina dissolution and cellular process formation, and the influences that precipitate neovascularization are unknown. As yet, there is no obvious morphologic difference to distinguish the initiation of the two processes. That the choriocapillaris may respond to its environment has been amply demonstrated both in vitro and in vivo. 28 29 30 31 Local control mechanisms must exist to prevent one from translating into the other, and these may be altered in disease. 31  
One of the initial steps in angiogenesis is the degradation of the basal lamina and adjacent extracellular matrix. 15 16 29 Metalloproteinases that cleave basal lamina collagens IV and V have been extracted from endothelial cells and are believed to be responsible for disruption of the basal lamina locally and facilitate outgrowth of capillary sprouts. 32 The process thereafter is likely to depend on the relative concentrations of various growth factors and the composition of the extracellular matrix. There is ample evidence that various factors with the potential to modify cell behavior exist in Bruch’s membrane, and Bruch’s membrane changes greatly with age. 33 34 35 36 These considerations suggest that choroidal neovascularization, as part of age-related macular disease, may occur as a distortion of a normal mechanism rather than representing a process unique to the aging eye. 
By analogy with other vascular systems, it is possible that the primary function of the cellular processes into Bruch’s membrane, together with the endothelial cytoskeleton, is to stabilize the cell physically, a phenomenon common to other fenestrated thin-walled capillaries. 11 This concept of a structural role is supported by the finding of microfilaments within the processes and the presence of localized thickening of basal lamina around the processes. This is not to deny that they may serve other functions, such as sampling the local environment, and may contribute to the normal clearance of debris from Bruch’s membrane, either by phagocytosis or by modifying the biophysical properties of Bruch’s membrane. However, none of our findings lend direct support to these subsidiary functions in Bruch’s membrane, although these functions have been identified in other tissues. If choroidal neovascularization in age-related maculopathy is a distortion of this normal phenomenon, its upregulation after laser photocoagulation could be associated with potential neovascular complications. 21  
 
Figure 1.
 
(A) The fracture plane, as viewed in Figures 1 and 2 . It passes through the luminal plasma membrane, exposing its endothelial (E)-face (a). It then skips across the endothelium to expose the protoplasmic (P)-face (b) of the adluminal plasma membrane (adjacent to Bruch’s membrane). The fracture plane within this membrane passes through numerous fenestrae and across one of the choriocapillaris processes (c), leaving a stumplike remnant of this structure ( Image not available ). (B, C) Freeze-fracture replicas depicting the P-face of the cell membrane of the retinal aspect of the hamster choriocapillaris. (B) Three endothelial cells. Note in (B) and (C) the large expanses of densely packed fenestrae interspersed within a network of branching and anastomosing nonfenestrated membrane (arrow). Magnification: (B) ×3,300; (C) ×33,000.
Figure 1.
 
(A) The fracture plane, as viewed in Figures 1 and 2 . It passes through the luminal plasma membrane, exposing its endothelial (E)-face (a). It then skips across the endothelium to expose the protoplasmic (P)-face (b) of the adluminal plasma membrane (adjacent to Bruch’s membrane). The fracture plane within this membrane passes through numerous fenestrae and across one of the choriocapillaris processes (c), leaving a stumplike remnant of this structure ( Image not available ). (B, C) Freeze-fracture replicas depicting the P-face of the cell membrane of the retinal aspect of the hamster choriocapillaris. (B) Three endothelial cells. Note in (B) and (C) the large expanses of densely packed fenestrae interspersed within a network of branching and anastomosing nonfenestrated membrane (arrow). Magnification: (B) ×3,300; (C) ×33,000.
Figure 2.
 
Freeze-fracture replicas depicting the protoplasmic (P)-face of the retinal aspect of the hamster choriocapillaris. (A) Choroidal capillary endothelial cell processes always arose from nonfenestrated, focally thickened regions of endothelial cells. The apical portions of endothelial cell processes typically broke off during the replication process and were often visualized as stumps (A, B, C, Image not available ). (D) Note that the processes remained intact and were visualized extending into Bruch’s membrane (BM; arrows). Magnification: (A) ×6,700; (B) ×10,000; (C) ×10,000; (D) ×13,000.
Figure 2.
 
Freeze-fracture replicas depicting the protoplasmic (P)-face of the retinal aspect of the hamster choriocapillaris. (A) Choroidal capillary endothelial cell processes always arose from nonfenestrated, focally thickened regions of endothelial cells. The apical portions of endothelial cell processes typically broke off during the replication process and were often visualized as stumps (A, B, C, Image not available ). (D) Note that the processes remained intact and were visualized extending into Bruch’s membrane (BM; arrows). Magnification: (A) ×6,700; (B) ×10,000; (C) ×10,000; (D) ×13,000.
Figure 3.
 
Transmission electron micrograph. (A, B) Human endothelial cell processes (arrows) protruding through the basal lamina into the outer collagenous zone (OCZ) of Bruch’s membrane (BM). (A) A female donor aged 22 years and (B) a 45-year-old female donor. (C, D) Monkey endothelial cell processes (arrows) protruding through the basal lamina (thickened in D, arrow) into Bruch’s membrane. CC, choriocapillaris; RPE, retinal pigment epithelium. Magnification: (A) ×40,000; (B) ×20,000; (C, D) ×24,000.
Figure 3.
 
Transmission electron micrograph. (A, B) Human endothelial cell processes (arrows) protruding through the basal lamina into the outer collagenous zone (OCZ) of Bruch’s membrane (BM). (A) A female donor aged 22 years and (B) a 45-year-old female donor. (C, D) Monkey endothelial cell processes (arrows) protruding through the basal lamina (thickened in D, arrow) into Bruch’s membrane. CC, choriocapillaris; RPE, retinal pigment epithelium. Magnification: (A) ×40,000; (B) ×20,000; (C, D) ×24,000.
Figure 4.
 
Transmission electron micrograph of human (A, B) and hamster (C, D) endothelial cell processes. These cytoplasmic channels were filled with microfilaments (B, D, arrows) that terminated in dense plaques (A, arrow) and bundles of unidirectionally oriented microtubules (C, arrow) and often encircled the entire capillary. The processes often branch as seen in (A). Magnification: (A, C, D) ×20,000; (B) ×40,000.
Figure 4.
 
Transmission electron micrograph of human (A, B) and hamster (C, D) endothelial cell processes. These cytoplasmic channels were filled with microfilaments (B, D, arrows) that terminated in dense plaques (A, arrow) and bundles of unidirectionally oriented microtubules (C, arrow) and often encircled the entire capillary. The processes often branch as seen in (A). Magnification: (A, C, D) ×20,000; (B) ×40,000.
Figure 5.
 
(A) Transmission electron micrographs of hamster (A) and human (B) endothelial cell processes. The processes extended through the capillary basal lamina and into BM. In the hamster, these processes often abutted, but did not penetrate, the basal lamina of the RPE (A, arrow). In primates, the processes rarely extended through the elastic lamina (EL) of the basement membrane; however, one is depicted in (B, arrow). Magnification: (A) ×6,700; (B) ×20,000.
Figure 5.
 
(A) Transmission electron micrographs of hamster (A) and human (B) endothelial cell processes. The processes extended through the capillary basal lamina and into BM. In the hamster, these processes often abutted, but did not penetrate, the basal lamina of the RPE (A, arrow). In primates, the processes rarely extended through the elastic lamina (EL) of the basement membrane; however, one is depicted in (B, arrow). Magnification: (A) ×6,700; (B) ×20,000.
Figure 6.
 
Transmission electron micrographs of human endothelial cell processes protruding through the basal lamina. (A) Eye of a 25-year-old male. Note the marked focal thickening of the basal lamina around the processes, which is thought to spread the physical load on the membrane (A, arrow). (B) Eye of an 80-year-old female donor. Note the LSC in close association with the cell processes (B, arrow).
Figure 6.
 
Transmission electron micrographs of human endothelial cell processes protruding through the basal lamina. (A) Eye of a 25-year-old male. Note the marked focal thickening of the basal lamina around the processes, which is thought to spread the physical load on the membrane (A, arrow). (B) Eye of an 80-year-old female donor. Note the LSC in close association with the cell processes (B, arrow).
The authors thank Robin Howes, and Bobbie Schneider for the preparation of tissue for EM; Akiko Okubo and Robert Rosa for providing the Bruch’s membrane data in human eyes to correlate with the numbers of endothelial cell processes; Jonathan Lund and Jennifer Levitt for the monkey eyes; patients JN, SW, and PC, who willingly donated their eyes for research after exenteration; Bridget Mulholland at Moorfields Eye Hospital, who coordinated the collection of this material; Catey Bunce for data analysis; and the Mid-America Transplant Services (St. Louis, MO), The Moorfields Eye Bank, and The UK Transplant Centre (Bristol, UK) for providing human donor eyes. 
Garron LK. The ultrastructure of the retinal pigment epithelium with observations on the choriocapillaris and Bruch’s membrane. Trans Am Ophthalmol Soc. 1963;61:545–588. [PubMed]
Leeson TS, Leeson CR. Choriocapillaris and lamina elastica (vitrea) of the rat eye. Br J Ophthalmol. 1967;51:599–615. [CrossRef] [PubMed]
Matsusaka T. Undescribed endothelial processes of the choriocapillaris extending to the retinal pigment epithelium of the chick. Br J Ophthalmol. 1968;52:887–892. [CrossRef] [PubMed]
Yamamoto T, Yamashita H. Pseudopodia of choriocapillary endothelium. Jpn J Ophthalmol. 1989;33:327–336. [PubMed]
Korte GE, Chase J. Additional evidence for remodelling of normal choriocapillaris. Exp Eye Res. 1989;49:299–303. [CrossRef] [PubMed]
Yamamoto T, Yamashita H. Pseudopodia of choriocapillary endothelium in ocular tissues. Jpn J Ophthalmol. 1990;34:181–187. [PubMed]
Yamamoto T, Fukuda S, Obata H, Yamashita H. Electron microscopic observation of pseudopodia from choriocapillary endothelium. Jpn J Ophthalmol. 1994;38:129–138. [PubMed]
Manche EE, Korte GE. Ultrastructural evidence of remodelling in the microvasculature of the normal rabbit and human eye. Acta Anat. 1990;138:89–96. [CrossRef] [PubMed]
McMenamin PG, Lee WR, Aitken DA. Age related changes in the human outflow apparatus. Ophthalmology. 1986;93:194–209. [PubMed]
Taugner R, Kirchheim H, Forssmann WG. Myoendothelial contacts in glomerular arterioles and in the renal interlobular arteries of rats, mouse and Tupaia belangeri. Cell Tissue Res. 1984;235:319–325. [PubMed]
Takahashi-Iwanaga H. The three dimensional cytoarchitecture of the interstitial tissue in the rat kidney. Cell Tissue Res. 1991;264:269–281. [CrossRef] [PubMed]
Michel CC. Renal medullary microcirculation: architecture and exchange. Microcirculation. 1995;2:125–139. [CrossRef] [PubMed]
Moore DH, Ruska H. The fine structure of capillaries and small arteries. J Biophys Biochem Cytol. 1957;3:457–465. [CrossRef] [PubMed]
Rhodin JAG. The ultrastructure of mammalian arterioles and precapillary sphincters. J Ultrastruct Res. 1967;18:181–223. [CrossRef] [PubMed]
Folkman J. Angiogenesis: initiation and control. Ann NY Acad Sci. 1982;401:212–227. [CrossRef] [PubMed]
Ausprunk DH, Folkman J. MIgration and proliferation of endothelial cells in preformed and newly formed blood vessels during tumour angiogenesis. Microvasc Res. 1997;14:53–65.
Killingsworth MC. Angiogenesis in early choroidal neovascularization secondary to age related macular degeneration. Graefes Arch Clin Exp Ophthalmol. 1995;233:313–323. [CrossRef] [PubMed]
Heriot WJ, Henkind P, Bellhorn RW, Burns MS. Choroidal neovascularization can digest Bruch’s membrane: a break is not essential. Ophthalmology. 1984;91:1603–1608. [CrossRef] [PubMed]
Orzalesi N, de Felice GP, Carta S. Budding of choriocapillary endothelial cells in the normal pigmented rat. Exp Eye Res. 1988;46:105–109. [CrossRef] [PubMed]
Okubo A, Rosa RH, Jr, Bunce CV, et al. The relationships of age changes in retinal pigment epithelium and Bruch’s membrane. Invest Ophthalmol Vis Sci. 1999;40:443–449. [PubMed]
Guymer RH, Hageman G, Bird AC. Influences of laser photocoagulation on choroidal capillary cytoarchitecture. Br J Ophthalmol. 2001;85:40–46. [CrossRef] [PubMed]
Tripathi RC, Tripathi BJ. Contractile protein alteration in trabecular endothelium in primary open angle glaucoma. Exp Eye Res. 1980;31:721–724. [CrossRef] [PubMed]
Gipson IK, Anderson RA. Actin filaments in cells of human trabecular meshwork and Schlemm’s canal. Invest Ophthalmol Vis Sci. 1979;18:547–561. [PubMed]
Duvall J, Tso MOM. Cellular mechanisms of resolution of drusen after laser coagulation: an experimental study. Arch Ophthalmol. 1985;103:694–703. [CrossRef] [PubMed]
Grierson I, Day J, Unger WG, Ahmed A. Phagocytosis of latex microspheres by bovine meshwork cells in culture. Graefes Arch Clin Exp Ophthalmol. 1986;224:536–544. [CrossRef] [PubMed]
Suter ER, Manjo G. Passage of lipids across vascular endothelium in newborn rats. J Cell Biol. 1965;27:166–177.
Gabbianni G, Majno G. Pathophysiology of small vessel permeability. Kaley G Altura BM eds. Microcirculation. 1980;3:143–163. University Park Press Baltimore.
Korte GE, Burns MS, Bellhorn RW. Epithelium-capillary interactions in the eye: the retinal pigment epithelium and the choriocapillaris. Int Rev Cytol. 1989;114:221–248. [PubMed]
Glaser BM. Extracellular modulating factors and the control of intraocular neovascularization. Arch Ophthalmol. 1988;106:603–607. [CrossRef] [PubMed]
Kalebic T, Garbisa S, Glaser B, Liotta LA. Basal lamina collagen: degradation by migrating endothelial cells. Science. 1983;221:281–283. [CrossRef] [PubMed]
Witmer AN, Vrensen GF, Van Noorden CJ, Schlingemann RO. Vascular endothelial growth factors and angiogenesis in eye disease. Prog Retin Eye Res. 2003;22:1–29. [CrossRef] [PubMed]
Mancini MA, Frank RN, Keirn RJ, Kennedy A, Khoury JK. Does the retinal pigment epithelium polarize the choriocapillaris?. Invest Ophthalmol Vis Sci. 1986;27:336–345. [PubMed]
van der Schaft TL, de Bruijn WC, Mooy CM, et al. Is basal laminar deposits unique for age related macular degeneration?. Arch Ophthalmol. 1991;109:420–425. [CrossRef] [PubMed]
Feeney-Burns L, Ellersieck MR. Age related changes in the ultrastructure of Bruch’s membrane. Am J Ophthalmol. 1985;100:686–697. [CrossRef] [PubMed]
Loffler KU, Lee WR. Basal linear deposit in the human macula. Graefes Arch Clin Exp Ophthalmol. 1986;224:493–501. [CrossRef] [PubMed]
Sarks SH. Ageing and degeneration in the macular region: a clinico-pathological study. Br J Ophthalmol. 1976;60:324–341. [CrossRef] [PubMed]
Figure 1.
 
(A) The fracture plane, as viewed in Figures 1 and 2 . It passes through the luminal plasma membrane, exposing its endothelial (E)-face (a). It then skips across the endothelium to expose the protoplasmic (P)-face (b) of the adluminal plasma membrane (adjacent to Bruch’s membrane). The fracture plane within this membrane passes through numerous fenestrae and across one of the choriocapillaris processes (c), leaving a stumplike remnant of this structure ( Image not available ). (B, C) Freeze-fracture replicas depicting the P-face of the cell membrane of the retinal aspect of the hamster choriocapillaris. (B) Three endothelial cells. Note in (B) and (C) the large expanses of densely packed fenestrae interspersed within a network of branching and anastomosing nonfenestrated membrane (arrow). Magnification: (B) ×3,300; (C) ×33,000.
Figure 1.
 
(A) The fracture plane, as viewed in Figures 1 and 2 . It passes through the luminal plasma membrane, exposing its endothelial (E)-face (a). It then skips across the endothelium to expose the protoplasmic (P)-face (b) of the adluminal plasma membrane (adjacent to Bruch’s membrane). The fracture plane within this membrane passes through numerous fenestrae and across one of the choriocapillaris processes (c), leaving a stumplike remnant of this structure ( Image not available ). (B, C) Freeze-fracture replicas depicting the P-face of the cell membrane of the retinal aspect of the hamster choriocapillaris. (B) Three endothelial cells. Note in (B) and (C) the large expanses of densely packed fenestrae interspersed within a network of branching and anastomosing nonfenestrated membrane (arrow). Magnification: (B) ×3,300; (C) ×33,000.
Figure 2.
 
Freeze-fracture replicas depicting the protoplasmic (P)-face of the retinal aspect of the hamster choriocapillaris. (A) Choroidal capillary endothelial cell processes always arose from nonfenestrated, focally thickened regions of endothelial cells. The apical portions of endothelial cell processes typically broke off during the replication process and were often visualized as stumps (A, B, C, Image not available ). (D) Note that the processes remained intact and were visualized extending into Bruch’s membrane (BM; arrows). Magnification: (A) ×6,700; (B) ×10,000; (C) ×10,000; (D) ×13,000.
Figure 2.
 
Freeze-fracture replicas depicting the protoplasmic (P)-face of the retinal aspect of the hamster choriocapillaris. (A) Choroidal capillary endothelial cell processes always arose from nonfenestrated, focally thickened regions of endothelial cells. The apical portions of endothelial cell processes typically broke off during the replication process and were often visualized as stumps (A, B, C, Image not available ). (D) Note that the processes remained intact and were visualized extending into Bruch’s membrane (BM; arrows). Magnification: (A) ×6,700; (B) ×10,000; (C) ×10,000; (D) ×13,000.
Figure 3.
 
Transmission electron micrograph. (A, B) Human endothelial cell processes (arrows) protruding through the basal lamina into the outer collagenous zone (OCZ) of Bruch’s membrane (BM). (A) A female donor aged 22 years and (B) a 45-year-old female donor. (C, D) Monkey endothelial cell processes (arrows) protruding through the basal lamina (thickened in D, arrow) into Bruch’s membrane. CC, choriocapillaris; RPE, retinal pigment epithelium. Magnification: (A) ×40,000; (B) ×20,000; (C, D) ×24,000.
Figure 3.
 
Transmission electron micrograph. (A, B) Human endothelial cell processes (arrows) protruding through the basal lamina into the outer collagenous zone (OCZ) of Bruch’s membrane (BM). (A) A female donor aged 22 years and (B) a 45-year-old female donor. (C, D) Monkey endothelial cell processes (arrows) protruding through the basal lamina (thickened in D, arrow) into Bruch’s membrane. CC, choriocapillaris; RPE, retinal pigment epithelium. Magnification: (A) ×40,000; (B) ×20,000; (C, D) ×24,000.
Figure 4.
 
Transmission electron micrograph of human (A, B) and hamster (C, D) endothelial cell processes. These cytoplasmic channels were filled with microfilaments (B, D, arrows) that terminated in dense plaques (A, arrow) and bundles of unidirectionally oriented microtubules (C, arrow) and often encircled the entire capillary. The processes often branch as seen in (A). Magnification: (A, C, D) ×20,000; (B) ×40,000.
Figure 4.
 
Transmission electron micrograph of human (A, B) and hamster (C, D) endothelial cell processes. These cytoplasmic channels were filled with microfilaments (B, D, arrows) that terminated in dense plaques (A, arrow) and bundles of unidirectionally oriented microtubules (C, arrow) and often encircled the entire capillary. The processes often branch as seen in (A). Magnification: (A, C, D) ×20,000; (B) ×40,000.
Figure 5.
 
(A) Transmission electron micrographs of hamster (A) and human (B) endothelial cell processes. The processes extended through the capillary basal lamina and into BM. In the hamster, these processes often abutted, but did not penetrate, the basal lamina of the RPE (A, arrow). In primates, the processes rarely extended through the elastic lamina (EL) of the basement membrane; however, one is depicted in (B, arrow). Magnification: (A) ×6,700; (B) ×20,000.
Figure 5.
 
(A) Transmission electron micrographs of hamster (A) and human (B) endothelial cell processes. The processes extended through the capillary basal lamina and into BM. In the hamster, these processes often abutted, but did not penetrate, the basal lamina of the RPE (A, arrow). In primates, the processes rarely extended through the elastic lamina (EL) of the basement membrane; however, one is depicted in (B, arrow). Magnification: (A) ×6,700; (B) ×20,000.
Figure 6.
 
Transmission electron micrographs of human endothelial cell processes protruding through the basal lamina. (A) Eye of a 25-year-old male. Note the marked focal thickening of the basal lamina around the processes, which is thought to spread the physical load on the membrane (A, arrow). (B) Eye of an 80-year-old female donor. Note the LSC in close association with the cell processes (B, arrow).
Figure 6.
 
Transmission electron micrographs of human endothelial cell processes protruding through the basal lamina. (A) Eye of a 25-year-old male. Note the marked focal thickening of the basal lamina around the processes, which is thought to spread the physical load on the membrane (A, arrow). (B) Eye of an 80-year-old female donor. Note the LSC in close association with the cell processes (B, arrow).
×
×

This PDF is available to Subscribers Only

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

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

×