Ultrastructural distribution of collagen types I-VI in aging human retinal vessels.

Retinal vessels from freshly enucleated human eyes were classified into three stages of the hyalinisation process. The distribution of collagen types I-VI within the vessel walls was studied ultrastructurally by immunogold labelling combined with the tissue preparation techniques of cryoultramicrotomy and London resin white embedding. Collagen types I, III, IV, V, and VI were found in large vessels, types I, IV, and V plus small amounts of III and VI in small vessels, and types I, IV, and V in capillaries. Hyalinised vessel walls consisted mainly of types I, IV, and VI collagen.

Ultrastructural localization of extracellular matrix components in human retinal vessels and Bruch's membrane.

We have used the electron microscopic immunogold technique to localize precisely various extracellular matrix components in human retinal vessels and Bruch's membrane. Collagen types IV and V, laminin, and heparan sulfate proteoglycan core protein were localized in basement membranes of retinal capillaries. In addition to the capillary basement membrane components, collagen types I and III and fibronectin were found in the basement membranes of retinal arterioles and venules. The basement membranes of choriocapillaries and retinal pigment epithelial cells in Bruch's membrane also showed a similar distribution of the retinal capillary basement membrane components. Both the inner and outer collagenous layers of Bruch's membrane contained collagen types I and III along with fibronectin, whereas collagen type VI was mostly limited to the central elastic lamina. The precise localization and distribution of various extracellular matrix components in human retinal vessels and Bruch's membrane may be important for understanding their normal function as well as their alteration in disease.

Immunogold localization of basement membrane molecules in rat retinal capillaries.

Vascular basement membrane contains laminin, fibronectin, proteoglycan and collagens. These molecules have been identified in various tissues by immunolabeling methods and biochemical analyses. We have previously localized laminin, fibronectin and type IV collagen to the basement membrane of rat retinal vessels at the ultrastructural level using an immunoperoxidase method. In this study, we use an immunogold method to re-examine the distribution of these molecules and also to study the localization of heparan sulfate proteoglycan and types I, III and V collagen in the retinal capillary basement membrane. Gold labeling for laminin, type IV collagen and proteoglycan were found diffusely on the basement membrane of the endothelium and pericyte, while that for fibronectin and type V collagen was spotty and variable and that for types I and III collagen was negligible. The segment of basement membrane between the endothelial cell and pericyte appeared less reactive to anti-laminin and anti-type IV collagen than the membrane between the pericyte and perivascular neuroretina. The immunogold method may be useful in quantitative studies of thickened basement membranes under abnormal conditions.

Localization of gamma glutamyl transpeptidase in bovine retina.

Gamma-glutamyl transpeptidase (GGTP) is a membrane bound enzyme which has an important role in regulation of glutathione and glutamate in the retina. We have used histochemical and colorimetric enzyme assays to localize GGTP in the bovine retina and choroid. Our results demonstrate that (i) GGTP is present in retinal microvessels but not choroidal microvessels. (ii) Retinal microvascular endothelium loses the ability to express GGTP in cultured cells. (iii) GGTP is present in Muller cells. (iv) Isolated and purified rod outer segments contain high levels of GGTP. (v) Retinal pigment epithelial cells (RPE) in vivo and in culture contain GGTP. The findings of this study lend support to the concept that GGTP may be a biochemical marker for cellular systems which are part of specialized diffusion barriers.

[Retinal vascular beta-adrenergic receptors in man]. / Récepteurs bêta-adrénergiques vasculaires rétiniens chez l'homme.

Beta-blockers are widely used in the primary open angle glaucoma treatment. The molecular mechanism by which these drugs reduce intraocular pressure is essentially based on the blockade of beta-adrenergic receptors localized on the ciliary processes. Vascular effects of beta-blockers, which are difficult to exhibit clinically, have been recently reported for some drugs with intrinsic sympathomimetic activity. In this study, we attempted to investigate the presence of beta-adrenergic specific binding sites on the human retinal vessels, by means of an in vitro autoradiographic technique. These receptors are localized both on arteries and veins; displacement studies indicated that they are mainly of beta-2 subtype.

Immunocytochemical localization of laminin, type IV collagen and fibronectin in rat retinal vessels.

Laminin, type IV collagen and fibronectin have been identified as major components of the basement membrane (basal lamina) in various tissues. These antigens have also been identified in retinal vessels by light microscopic immunofluorescence but their precise location could not be determined at this level of resolution. In this study, we examined the localization of these constituents at the ultrastructural level using the protein A-immunoperoxidase technique. The basal lamina of all retinal capillaries, arterioles and venules was immunostained after exposure to antisera against laminin, type IV collagen and fibronectin. Staining was localized to the lamina densa, which appeared as a single or double layer. Immunostaining for fibronectin showed the weakest activity. The reaction was also seen in discrete patches between endothelial cells and pericytes. The inner limiting membrane of the retina was also reactive for laminin and type IV collagen but not for fibronectin. The results indicate that laminin, type IV collagen and fibronectin are components of the basal lamina in all types of retinal vessels. The presence of fibronectin at the endothelial-pericyte interface suggests that this protein may promote adhesion between cells and thus help to maintain the integrity of the vessel wall.

Topographical specificity in isolated retinal capillary basement membranes: a high-resolution scanning electron microscope analysis.

Numerous investigations have demonstrated that basement membranes (BMs) are composed of type IV collagen, laminin, heparan sulfate proteoglycan, nidogen, and possibly fibronectin. The precise proportion and supramolecular organization of these molecules within BMs is unclear, but is believed to be tissue-specific. In an effort to provide morphological evidence for BM specificity, we studied isolated bovine retinal capillary BMs by high-resolution SEM. Cryofractured specimens demonstrated that surfaces of BM leaflets and pericytic matrix (PCM) within the retinal capillary BM complex are composed of 20- to 100-nm granules and beaded fibrils arranged in patterns which are specific for each cell type. Subendothelial BMs and the subjacent PCM are composed of 20- to 30-nm granules loosely arranged and marked by numerous pits, features that are consistent with their TEM morphology and known susceptibility to proteolytic attack. These BMs also frequently exhibit large openings or fenestrations. These compare favorably with their fragmented image by TEM and probably represent BM discontinuities necessary for direct contact of pericytes and endothelial cells. Muller cell BMs are also composed of granules though they are much larger (40-100 nm) and more densely packed then those of subendothelial BMs. Moreover, they frequently contain interstitial collagen fibrils which could account for the tube-like structural rigidity exhibited by acellular retinal vessel BMs in SEM views. Data in the current study provide morphological evidence for direct contact of pericytes and endothelial cells in vivo and support the view that tissue specificity of BMs may be more exquisite than previously believed, extending even to surface topography of BM leaflets within capillary BM complexes.

Adrenergic receptors in bovine retinal microvessels: presence of alpha 2- and beta- but not alpha 1-receptors.

In bovine retinal microvessels, alpha 1, alpha 2- and beta-adrenergic receptors were characterized by binding assay, using [3H]prazosin, [3H]para-aminoclonidine and [125I]iodocyanopindolol as radioligands, respectively. The microvessels were purified from bovine eyes by differential centrifugation through a high concentration of bovine serum albumin followed by use of a glass bead filtration technique. In the preparation, specific binding sites for [3H]para-aminoclonidine and [125I]iodocyanopindolol were observed, whereas [3H]prazosin binding was not detected. The [3H]para-aminoclonidine binding sites localized to the microvessels were characterized by high affinity and saturability (KD: 173 +/- 9 pM; Bmax: 394 +/- 11 fmol/mg protein) as well as the [125I]iodocyanopindolol binding sites (KD: 20 +/- 3 pM; Bmax: 43 +/- 4 fmol/mg protein). Furthermore, the specificity of both binding sites was pharmacologically evaluated by measuring the inhibitory effects of various adrenergic reagents on binding. The existence of alpha 2- and beta-adrenergic receptors which were characterized by high affinity, saturability and stereospecificity, leads to the hypothesis that the retinal microcirculation is under neuronal control.

In vitro production of glycosaminoglycans by retinal microvessel cells and lens epithelium.

Analysis of bovine lens capsules and retinal microvessel basement membranes by cellulose acetate electrophoresis, together with treatment by specific enzymes or by nitrous acid, reveals that heparan sulfate is the only demonstrable glycosaminoglycan (GAG), consistent with the observations of most other investigators on basement membrane preparations from several different tissues. When bovine retinal microvessel pericytes or endothelial cells or bovine lens epithelial cells are grown in culture in the presence of [35S]-sulfate or [3H]-glucosamine, however, both cellulose acetate electrophoresis and gel filtration chromatography reveal that the cultured pericytes synthesize primarily chondroitin sulfate, whereas the lens epithelial cells and microvascular endothelial cells produce a mixture of GAGs consisting of approximately 60% heparan sulfate and 40% chondroitin sulfate. The chondroitin sulfate chains have an Mr of 70,000, but the Mr of the heparan sulfate chains is 10,000, based on gel filtration chromatography on Sepharose CL-6B. Hence, cell culture conditions may produce phenotypic modulation of the biosynthetic capacities of these cells for GAGs. However, the difference between the in vitro and in vivo findings may in part be explained if many of the chondroitin sulfate GAGs synthesized in vitro are cell associated and are not incorporated into the basement membrane.

Collagens of the retinal microvascular basement membrane and of retinal microvascular cells in vitro.

We have analysed the collagens present in vascular basement membranes isolated from bovine retinal and cerebral microvessels and bovine renal glomeruli, and from the non-vascular basement membrane of bovine lens capsule. These are compared with the collagens produced by cultured bovine retinal microvascular pericytes and lens epithelial cells, and by canine retinal microvascular endothelial cells, in vitro. Biochemical and immunocytochemical analyses indicate that all of the vascular basement membrane preparations have an identical collagenous composition, consisting of the same polypeptides present in lens capsule (primarily type IV collagen), together with other polypeptides that are identified as type I, and a small amount of type III collagen. Identification of the latter is based on two-dimensional gel electrophoresis in the presence and absence of a reducing agent. Immunocytochemical studies, however, demonstrate type I, type IV and some type V collagen in the basement membranes of the isolated microvessels. The cultured microvascular cells produce predominantly type I collagen molecules, but they also produce other collagen peptides that appear to be type IV, and, at least in some experiments, small amounts of type III collagen. The biochemical identification of collagens type I and IV is confirmed by immunocytochemistry. However, results with anti-type I collagen and procollagen antibodies in cultured pericytes vary with antibodies from different sources. The quantities of the type IV peptides produced by the cultured cells also vary in different experiments.

Retinal microvessel extracellular matrix: an immunofluorescent study.

The vasculature of the retina functions within a sheath of extracellular matrix (ECM). Unfortunately, little is known about the biochemical composition of this matrix. Abnormalities in the ECM of the retinal microvasculature are important in diabetic retinopathy as well as vasculopathies associated with connective tissue disorders. The ECM of unfixed frozen human retinal blood vessels was examined by indirect immunofluorescence using antibodies raised against collagen types I, II, III, IV, and V as well as the structural glycoproteins laminin and fibronectin. Antisera against collagen types I and IV as well as laminin and fibronectin stained a broad spectrum of retinal vessels, from large thick-walled vessels down to microvessels less than 10 micron in diameter. In contrast, antibodies against types III and V collagen were seen to stain primarily the walls of the larger vessels. Antibodies against type II collagen did not react with retinal vessels. Preincubation with the appropriate antigen or preimmune serum eliminated staining of the vessels by the antisera.

Angioarchitecture of intraorbital part of human optic nerve.

The distribution of blood vessels in the intraorbital part of the human optic nerve was examined after injecting silicone rubber into the ophthalmic artery of eyes obtained after autopsy. Particular attention was focused on the possible existence of an end artery which might play an important role in ischemic optic neuropathy. In the anterior part of the optic nerve, the vascular architecture consisted of axial and peripheral vascular systems. The axial system was constituted of intraneural branches of the central retinal artery. The peripheral system was mostly branches derived from the intravaginal part of the central retinal artery. Few branches of the posterior ciliary arteries extending to the optic nerve were observed in this part. In the posterior part of the optic nerve, the vascular architecture was formed mostly by the peripheral system which consisted of the extravaginal and recurrent intravaginal branches of the central retinal artery and branches of the ophthalmic artery. These branches forming the peripheral system anastomosed with each other on the surface of the optic nerve. Within the anterior part of the optic nerve, intraneural branches of the central retinal artery and centripetal branches of the pial plexus anastomosed and formed a fine vascular network. The density of the intraneural vascular meshwork was sparse in the lamina cribrosa and the posterior part of the optic nerve. On the surface of the optic nerve and in its inner part, blood vessels anastomosed and formed a network: no end arteries appeared to exist.

Unidirectional vesicular transport mechanism in retinal vessels.

When horseradish peroxidase (HRP) is introduced into the bloodstream, it is retained in the lumen of the retinal vessels (blood-retina barrier). In this paper, we report that when the same tracer is injected into the vitreous body, it penetrates the lumen of retinal vessels by transcellular vesicular transport. This unidirectional movement of macromolecules out of the eye is not inhibited by ouabain, fluoroacetate, or low temperatures.

Actin filaments in retinal pericytes and endothelial cells.

The contractile protein actin was identified within retinal capillary pericytes and endothelial cells. In capillary cross-sections, circumferential pericyte processes showed numerous parallel bundles of actin filaments forming a cap over the adjacent endothelial cells. The latter contained few actin filaments but a considerable number of intermediate filaments. Pericytes and endothelial cells communicated via gaps in the adluminal basal lamina. These special anatomic relationships are interpreted as evidence for a contractile role of retinal capillary pericytes and against a contractile role of retinal capillary endothelial cells

Contractile proteins in retinal endothelium and other non-muscle tissues of the eye.

The present study is the first investigation to demonstrate, by employing the combined approach of immunologically and electron microscope methods, the presence of actin-like contractile proteins in the mammalian retina, the corneal epithelium and endothelium, the iris, and the ciliary body, and to confirm their presence in lens epithelium. This is also the first report to demonstrate by these methods the presence of microfilaments and intermediate filaments in retinal vascular endothelium. Since we have shown that actin filaments are especially abundant in immature retinal endothelial cells, the question of their function arises, and we have discussed their possible relevance to the closure of immature retinal vessels when exposed to hyperoxia.

Reticulum fibres in relation to retinal vessels.

Argyrophilic perivascular and intervascular fibres in the mammalian retina are shown by specific antireticulin immunofluorescence to consist of reticulin. The possible significance of these findings is briefly discussed.

Angiomatosis retinae. An ultrastructural study and lipid analysis.

A nonfamilial case of agiomatosis retinae (retinal hemangioblastoma) was studied by electron microscopy. In addition to the three major types of cells previously identified within the tumor (endothelial cells, pericytes, heavily lipidized stromal cells), fibrous astrocytes in different stages of lipidization were also found. The endothelial cells were fenestrated, providing the basis for the extravasated exudate that is characteristic of the tumor. The pericytes were completely surrounded by casement membranes and displayed no significant lipidization; in a cellular plaque of vasular tissue at the base of the lesion, however, some of the multilaminar pericytes showed evidence of early smooth muscle differentiation. The stromal cells contained abundant lipid vacuoles and a few organelles, and exhibited granular degeneration of cytoplasmic filaments between the lipid vacuoles. There was spotty basement membrane formation where the stromal cells abutted on the vascular elements. No interconversion could be demonstrated among the endothelial cells, pericytes, and stromal cells. A source for the stromal cells was discovered in the early lipisization of fibrous astrocytes. Analysis of the extracted lipid from the tumor by means of infrared spectroscopy, lipid chromatography, and x-ray diffraction disclosed that the lipid was mostly cholestrol stearate, a plasma lipid. It is suggested that in the retinal lesions the leaky (fenestrated) capillaries of the tumor allowed the passive imbibition of plasma lipid by the fibrous astrocytes, leading to their gradual transformation into the fully lipidized stromal cells.

Glycogen concentration changes in retina, vitreous body and other eye tissues caused by disturbances of blood circulation.

The glycogen content in the individual eye tissues is strongly correlated to blood supply. Our investigations on the retina of bovines, which have not been fully evaluated, show that the time interval between interruption of blood supply and preparation of the retina is of special importance. Pressure ischemia affects a decrease in glycogen content in the retina and vitreous of rabbits, which is, however, less distinct in the vitreous. Decrease of glycogen with ischemia also takes place in the cornea and, to a lesser degree, in iris and choroid. In contrast, there is no decrease in the glycogen content of the lens. Changes in glycogen content of the rabbit retina after ligation of the A. carotis communis is less distinct than with pressure ischemia. In the vitreous, changes in glycogen content could not be observed. Values measured in both tissues of the ligated eye decrease with additional pressure ischemia.