An epithelial blood-aqueous barrier to horseradish peroxidase in the ciliary processes of the vervet monkey (Cercopithecus aethiops)

A contemporary concept of the blood-aqueous barrier

This review traces the evolution of the concept of the blood-aqueous barrier (BAB) during the past 20 years. The Classical model simply stipulated that the tight junctions of the iris vasculature and ciliary epithelium excluded plasma proteins from the aqueous humor (AH). It failed to reconcile the presence of AH protein levels equal to 1% of that found in plasma. Moreover, models of barrier kinetics assumed that the processes of AH secretion and plasma protein entry were directly linked. Thus, elevations of AH protein levels could only be explained by a pathological breakdown of the BAB. Over the last 20 years it has been shown that the plasma proteins in normal AH by-pass the posterior chamber entirely. Instead, these proteins diffuse from the capillaries of ciliary body stroma, into the iris stroma and then into the anterior chamber. This creates a reservoir of plasma-proteins in the iris stroma that is not derived from the iris vessels. This reservoir is prevented from diffusing posteriorly by tight junctions in the posterior iris epithelium. The one-way valve created by the pupil resting on the anterior lens capsule, combined with the continuous, forward flow of AH through the pupil, prevents protein reflux into the posterior chamber. Importantly, in the new paradigm, secretion of AH and the entry of plasma proteins into AH, are semi-independent events. This opens the possibility that AH protein levels could increase in the absence of breakdown of the BAB. Clinical consequences of this new paradigm of the BAB are discussed.

Intercellular junctions in rabbit eye ora serrata

Summary The aim of this study was to describe and localize the intercellular junctions in the ora serrata region of albino and pigmented rabbit eyes. Eyes of albino and pigmented rabbits were fixed and processed for transmission electron microscopy. Light and electron microscope examination was carried out on semithin and ultrathin sections. The ora serrata region showed adherens, gap and tight junctions in the retinal and ciliary margins of albino and pigmented rabbit eyes. In the retinal margin, zonulae adherens between Müller cells and photoreceptors are associated with tight junctions. In the ciliary margin, epithelial cells are joined by adherens, gap and tight junctions localized between apical and apicolateral cell membranes. Tight junctions appear as zonulae occludens in the non-pigmented apicolateral cell membranes and as tight focal junctions between pigmented and non-pigmented apical cell membranes. Between the ciliary and retinal margins there are adherens and tight focal junctions which attach pigmented apical cell membranes to adjacent cells. There were no differences in the distribution of intercellular junctions between albino and pigmented rabbits.

Transport of glucose across the blood-tissue barriers

In specialized parts of the body, free exchange of substances between blood and tissue cells is hindered by the presence of a barrier cell layer(s). Specialized milieu of the compartments provided by these "blood-tissue barriers" seems to be important for specific functions of the tissue cells guarded by the barriers. In blood-tissue barriers, such as the blood-brain barrier, blood-cerebrospinal fluid barrier, blood-nerve barrier, blood-retinal barrier, blood-aqueous barrier, blood-perilymph barrier, and placental barrier, endothelial or epithelial cells sealed by tight junctions, or a syncytial cell layer(s), serve as a structural basis of the barrier. A selective transport system localized in the cells of the barrier provides substances needed by the cells inside the barrier. GLUT1, an isoform of facilitated-diffusion glucose transporters, is abundant in cells of the barrier. GLUT1 is concentrated at the critical plasma membranes of cells of the barriers and thereby constitutes the major machinery for the transport of glucose across these barriers where transport occurs by a transcellular mechanism. In the barrier composed of double-epithelial layers, such as the epithelium of the ciliary body in the case of the blood-aqueous barrier, gap junctions appear to play an important role in addition to GLUT1 for the transfer of glucose across the barrier.

Improved ultrastructural preservation of rat ciliary body after high pressure freezing and freeze substitution: a perspective view based upon comparison with tissue processed according to a conventional protocol or by osmium tetroxide/microwave fixation

Conventional fixation of the delicate, highly folded rat ciliary body and its iridial extension, as well as of vitreal structures, is associated with the induction of a number of artifacts, thus limiting the reliability of morphological interpretations. Improved ultrastructural preservation may be achieved by microwave heating in combination with osmium tetroxide fixation. This protocol, although simple and cheap, yields results, particularly with respect to the extracellular matrix compartment between inner and outer ciliary epithelial cells, which are not greatly inferior to those obtained by implementing the sophisticated high pressure freezing and freeze substitution technique. The latter affords good to very good ultrastructural preservation of epithelium and stromal components, such as blood vessels, neural elements, smooth muscle cells, fibrocytes, and free cells, up to a depth of 50-100 microns from the tissue surface. Its superiority over osmium tetroxide/microwave fixation is revealed in the cytoplasmic, intraorganellar, and vitreal matrix compartments, which incur no obvious losses.

Characterization of bovine ciliary pigmented epithelial cells in monolayer culture: an ultrastructural, enzyme histochemical and immunohistochemical study

Monolayer cultures of the pigmented epithelial (PE) cells derived from two regions of the pars plicata of bovine eyes were established and grown up to the third passage. After this passage, the cultures became senescent. During the first three passages, the PE cells lost their pigment granules but developed a distinct cellular polarity by forming junctional complexes at their apical cell portions and depositing basement membrane like material on their basal side. The junctional complexes were shown to be impermeable for horseradish peroxidase, suggesting that they contained tight junctions. Histochemically, the monolayer cells stained for carbonic anhydrase (CA) and Na+/K(+)-ATPase, enzymes involved in active fluid secretion. Staining for CA and Na+/K(+)-ATPase as well as for acid phosphatase and immunostaining for vimentin and actin of the cultured PE cells were comparable with that of PE cells in vivo. Therefore, PE monolayer cultures are considered to be a suitable model for experimental studies in vitro.

Oxygen delivery to the anterior chamber of the eye--a novel function of the anterior iris surface

By means of pO2-needle electrodes the oxygen tension in the anterior chamber of rabbit and cynomolgus monkey eyes was measured and correlated with the topography. It can be clearly shown that the pO2 in front of the pupil is substantially lower than in front of the anterior iris surface. It is evident that the O2-supply of the anterior chamber is provided by the iris vasculature. This result was confirmed by measurements after partial or total iridectomy. A brief recapitulation of the morphology of the iris showed that the abundance of iris vasculature is well understandable in view of these pO2 measurements.

Synthesis and migration of 3H-fucose-labeled glycoproteins in the ciliary epithelium of the eye: effects of microtubule-disrupting drugs

3H-fucose was injected intravenously or intravitreously into albino rats. After time intervals of 10, 40, and 50 min, 1, 1.5, and 4 hr, 1, 3, and 7 days, and 1, 2, and 4 weeks after injection, the animals were sacrificed by intracardiac perfusion with glutaraldehyde. Samples of the ciliary body were prepared for light and electron microscope radioautography. Light microscope autoradiographs showed that the cells of both the inner and outer layers of ciliary epithelium actively incorporated 3H-fucose label in a reaction that peaked in intensity at 4 hr after injection, and then progressively declined. Electron microscope radioautographs revealed that, at early time intervals, most of the label was localized to the Golgi apparatus. With time, the plasma membrane of both cell types became increasingly labeled, and accounted for 60-70% of the total silver grains at 4 hr after injection. Adjacent to the basal cell surface of the inner layer cells, the fibers of the zonula became increasingly labeled from 1.5 hr onwards, providing strong evidence that these cells secrete glycoproteins to the zonula. When vinblastine was administered 30 min before 3H-fucose injection, followed by sacrifice 1.5 hr later, a much larger proportion of label remained localized to the Golgi apparatus than in controls, and the plasma membrane and zonula were much less labeled. These results suggest that, as documented in other cell types, microtubules may play a role in the intracellular transport of membrane and secretory glycoproteins in these cells.

Effects of water-soluble marihuana-derived material (MDM) on the rabbit ciliary body: light and electron microscopy

The response of the ciliary processes of the rabbit eye to water-soluble marihuana-derived material (MDM) has been examined with light and electron microscopy. Following intravenous injection of MDM, the processes undergo considerable swelling within 1 hour followed by thrombus formation in the capillaries and extravasation of red cells. Later phases include the formation of cysts between the non-pigmented and pigmented cell layers of the ciliary epithelium. The ciliary process edema coincides with the initial hypertensive phase seen after intravenous MDM, while the hematogenous response coincides with the fall in intraocular pressure. Following intravitreal injection of MDM, a similar pattern of structural changes occurs that accompanies a fall in intraocular pressure that lasts for several days; because the physiological response occurs over a longer time course (14-20 hours) relative to intravenous administration where the intraocular pressure changes occur rapidly, the ciliary process swelling phase does not result in an increase in intraocular pressure. The physiologic changes in the eye caused by MDM appear to be related to the induction of a general inflammatory response in the ciliary processes, with a primary effect on the vascular system.

Quantitative analysis of tight junctions during ciliary epithelium development

The tight junctions of the ciliary epithelium constitute the main morphological counterpart of the blood-aqueous barrier. Although well-known in the adult, they have only been poorly studied during the morphogenesis of the ciliary processes. We thus analysed them during this period in order to detect whether any changes appear in their general pattern. In the rat, results show that during ciliary body development, the junctional depth and the number of superimposed junctional fibrils decrease, the P-face intramembranous particle density within and outside the junctional domains increase, and particular structures, namely the 'complex strands', are numerous at the early stages and become rare in the adult. Thus, the tight junctions appear as non-stable structures during ciliary-body development and these modifications in their morphology may correspond to changes in barrier properties during the development of this tissue.

The structure of tight junctions in the ciliary epithelium

The tight junctions of the ciliary epithelium act as a barrier preventing the passage of blood borne macromolecules into the posterior chamber. The use of the freeze-fracture technique has led to a good knowledge of their morphological pattern in various species. However, in order to attempt a correlation of the morphology of the tight junctions with their physiological properties, their intimate substructure must be considered. As in other glutaraldehyde-fixed epithelia, the tight junctions appear as networks (variable in their apico-basal thickness) of more or less discontinuous P-face ridges and as complementary E-face furrows in which some particles or short bars are found. The significance of the discontinuities of ridges has been analysed. The continuity of the junctional fibrils was evident as assessed both by quantitative measurements as well as morphological examination of complementary fracture faces. In addition, the absence of loss of junctional material showed that the integrity of the junction was preserved during the freeze-fracture process, even in conditions where an increase in "transfer" of junctional elements was experimentally induced. Most of all, a "pore system" due to visualizable gaps in the fibrils is not tenable for the ciliary epithelium. Furthermore, the analysis of transition steps at the level of membrane "fusion" showed that the tight junctions of the ciliary epithelium must now be considered as formed by two slightly offset fibrils, one per adjacent plasma membrane.

Freeze-fractured replica of Schlemm's canal and the trabecular meshwork in the primate

Using a freeze-fracture technique, the fine structure of the Schlemm's canal and trabecular meshwork in the monkey was studied with special reference to intercellular junctions. Gap junctions were observed between the endothelial cells of Schlemm's canal and between the trabecular meshwork cells. Maculae occludentes were rarely observed between the endothelial cells of the inner wall of Schlemm's canal. The trabecular meshwork cells appeared to act in an electrotonic and metabolic syncytium.

Nature of blood-labyrinth barrier in experimental conditions

The blood-labyrinth barrier is concept that has evolved based on marked difference in chemical composition between perilymph and blood. Studies reported here have been designed to manipulate physiologic, metabolic, and pharmacologic conditions in experimental animals in order to determine the characteristics of this regulatory mechanisms. Tracer studies of uptake of sodium, calcium, and albumin from blood into perilymph showed that these substances penetrate into inner ear fluids quite slowly. Injections of ototoxic substances (kanamycin, furosemide) show limited transport of these agents into perilymph. Administration of an osmotic agent (urea) resulted in a parallel but delayed elevation of perilymph concentration. The possible role of a alteration of blood-labyrinth barrier in inner ear disorders has been discussed.

Increased vascular permeability in the rabbit iris induced by prostaglandin E1. An electron microscopic study using lanthanum as a tracer in vivo

Prostaglandin E1 (PGE1) (25 micrograms) was applied topically to the eyes of albino rabbits, and lanthanum was used as an electron microscopic tracer to study possible vascular permeability changes in the iris of these eyes. Lanthanum was injected directly into the blood circulation 15 min after application of the PG, and the eyes were enucleated 30 to 60 min later. PGE1 was found to induce increased vascular permeability. Heavy deposits of the tracer were found in the walls of iridial venules, especially between the endothelium and its basement lamina. The highest concentration of lanthanum was found in close relation to the interendothelial clefts. Occasionally, gaps between neighbouring endothelial cells were observed. In control eyes, not treated with PG, no extravasation of the tracer was found. The study indicates that the iridial blood vessels contribute to the production of the plasmoid aqueous humour induced by PGE1 in the rabbit eye.

The tight junctions of rabbit choroid plexus and ciliary body epithelia. A comparative study using the double replica freeze-fracture technique

The spatial arrangement of tight junctions in choroid plexus and ciliary body rabbit epithelia has been determined by studying freeze-fracture complementary replicas. In the choroid plexus epithelium, the interruptions of the junctional P-face fibrils were measured to be 14% in their total length. In the ciliary body epithelium, where the fibrils were found to be more fragmented than in the choroid plexus, the P-face fibril interruption accounted for 12% of the total length of the zonulae occludentes sealing the non-pigmented cells and 30% in the focal linear tight junctions connecting the non-pigmented cells at their apices. In both epithelia, the interruptions of the ridges are precisely complemented by particles or short bars of similar length found in the E-face furrows. Consequently, it is possible to conclude that the junctional fibrils are continuous in these two epithelia. For the zonulae occludentes, this continuity appears to be inconsistent with the 'leaky' properties of these epithelia shown by some physiological investigations.

Endothelial cell junctions in the ciliary body microvasculature. A freeze-fracture study in the rabbit

Endothelial cells of the microvasculature (arterioles, fenestrated capillaries, and small venules) of the rabbit ciliary body have been examined in freeze-fracture replicas in ultrahigh vacuum and at very low temperatures, with special regard to their junctional boundaries. The junctional complexes of fenestrated capillaries and venular endothelium are discontinuous and do not seal the interendothelial clefts. They cannot be compared with zonulae occludentes or tight junctions, while the junctions in the arteriolar endothelium are continuous and well organized, suggesting a very low degree of paracellular permeability. It would be of interest to know if these structural differences at the junctional level in each segment of the microvasculature correspond to different paracellular permeability properties.

Effect of prostaglandin E2 on the permeability of the iris vessels to horseradish peroxidase in the rabbit

Prostaglandin E2 (0.5 microgram in 10 microliter 10% ethanol) was introduced into the anterior chambers of rabbit eyes. Using the horseradish peroxidase method, it was shown under the electron microscope that the endothelial barrier of the iris vessels broke down. The peroxidase penetrated as far as the basis of the posterior epithelial cells, however, without entering their lateral intercellular spaces. The question of whether the effect of prostaglandins on the barrier was a direct effect or at least a partially indirect one, i.e., a haemodynamic action, is discussed.

The permeability of the human ciliary and iridial epithelium to horseradish peroxidase. An in vitro study

The permeability of the human ciliary epithelium to horseradish peroxidase (PO) has been studied in vitro with the electron microscope. Ciliary body and iris specimens were obtained from freshly enucleated eyes. PO was applied at the stromal side of the epithelium, and was left for 120 min. The movement of PO through the intercellular spaces of the human ciliary epithelium was blocked apically in the lateral intercellular spaces of the non-pigmented epithelial cells, indicating that these cells are girdled by zonulae occludentes. In the iridial epithelium, the same distribution pattern of peroxidase reaction products (PORP) was found, i.e. the progression of PO was blocked apically in the lateral intercellular spaces of the posterior epithelial cells. The study indicates that the human ciliary and iridial epithelium contains a system of zonulae occludentes, which represents a diffusion barrier to high molecular, water soluble substances. This is consistent with previous studies in several species of animals.

Movement of horseradish peroxidase in the cornea, sclera and the anterior uvea

The anterior chambers of rabbit, monkey and human eyes were perfused with horseradish peroxidase. The perfusion was started immediately after enucleation of the monkey and human eyes, while the rabbit eyes were perfused in situ after the animals had been killed. Comparative results were obtained after 1 h or perfusion and were mainly based on frozen sections. Intensive staining was found in the Descemet's membrane, the sclera and the iridial stroma, considerably more than in the corneal stroma and in the ciliary body base. In the vervet monkey and human eyes only traces of peroxidase reaction products were seen in the stroma of the ciliary body base. The movement of the protein tracer from the anterior chamber into the corneal stroma is probably dependent upon vesicular transport across the corneal endothelium. The Discemet's membrane was interpreted to be more porous than the corneal stroma. A transcorneal and a corneo-scleral movement appeared to be more efficient than the uveo-scleral one, at least in the vervet monkey and the human eyes.

Ultrastructure of the blood-aqueous barrier in normal condition and after paracentesis A freeze-fracture study in the rabbit

There is in the ciliary epithelium a barrier which prevents the passage of plasma proteins into the posterior chamber of the eye. The anatomical site of this blood-aqueous barrier is at the zonulae occludentes (tight junctions) located between the lateral plasma-membranes of the nonpigmented epithelial cells. Using the freeze-fracture technique we have studied these junctions in the pars plicata. As in many epithelia, they are formed by an anastomosing network of ridges on the fracture PF face and by a complementary set of furrows on the fracture EF face of plasma membranes. Features of these junctions are the presence of frequent parallel and closely associated double or triple ridges, and a few small gap junctions, associated with tight junctional elements of the zonulae occludentes. We also show that the number of junctional strands (ridges and furrows from apex to base) varies from a few strands to about 15, according to their position around the nonpigmented epithelial cells. These observations suggest that the rabbit's ciliary epithelium is to be classified among the 'tight' or 'very tight' epithelia (Claude and Goodenough, 1973). After paracentesis, we observed no breakdown of the zonulae occludentes. This supports the hypothesis that the site of leakage of the plasma proteins is elsewhere than ciliary epithelium.

Blood-aqueous barrier in newborn and young rabbits. An electron microscopic study

Distribution of intravenously injected horseradish peroxidase (HRP) in the ciliary body of newborn and young (2 weeks old) rabbits was studied with the electron microscope. The reaction product was localized in all animals in the lumen of the blood vessels, in the perivascular area, in the surrounding connective tissue stroma and also in the epithelial cell layers. The peroxidase penetrated through the intercellular junctions of the pigmented cells to the intercellular space between the apical membranes of the two epithelial cell layers. In young rabbits peroxidase penetrated between the apical intercellular clefts of the non-pigmented cells in small quantities only, whereas in the newborn animals the clefts were filled with peroxidase. This reflects differences in the blood-aqueous barrier mechanism between the two studied groups and is evidently due to developmental stages of the eyes.

Blood-aqueous barrier can be circumvented by lowering intraocular pressure

Rhesus monkeys were injected intravenously with hypertonic urea (9 ml/kg body weight of 30% urea in 10% invert sugar) and the intraocular pressure was measured with an applamatic tonometer. When this pressure reached its minimum (20% of the normal value) horseradish peroxidase (molecular weight 40,000; radius of an equivalent hydrodynamic sphere about 2.5 nm; 0.5 g/kg body weight), was injected intravenously. Twenty minutes following peroxidase administration, either aqueous humor was sampled from the anterior chamber for biochemical determination of peroxidase activity, or one eyeball was enucleated and processed for light and electron microscopic localization of the enzymatic tracer. This experiment showed that: (1) therapeutic doses of hypertonic urea do not cause a breakdown of either the blood-retina or the blood-aqueous barriers; (2) as intraocular pressure decreases, peroxidase-containing blood flows back from the episcleral veins into the Schlemm canal; (3) macromolecules up to the dimensions of horseradish peroxidase leak through the intercellular clefts of the endothelium of the Schlemm canal, permeate the juxtacanalicular connective tissue and trabecular meshwork, and finally enter the anterior chamber. Thus, blood-borne substances can circumvent the blood-aqueous barrier when intraocular pressure is decreased, and administration of a hypertonic agent may represent a simple pharmacological device to cause penetration into the ocular chambers by drugs that are normally excluded from the interior of the eye.

Structural alterations in the ciliary process and the blood-aqueous barrier of the monkey after systemic urea injections

The structural alterations induced in the monkey ciliary epithelium by intravascular injection of urea solutions were studied with the light and electron microscopes. Arterial injection of urea resulted in destruction of the nonpigmented epithelium and the consequent breakdown of the blood-aqueous barrier. The intravenous injection did not significantly affect the structural integrity of the ciliary epithelium. The intercellular zonulae occludens were not altered and the intercellular pathway from blood to posterior chamber remained closed to horseradish peroxidase. Intercellular uptake in large cytoplasmic vacuoles appeared to account for some late transport to the basal end of the nonpigmented epithelium. There was no comparable transport of peroxidase in vesicles or vacuoles through the nonpigmented epithelium in animals not subjected to intravenous urea treatment. Compared to the arterial route, intravenous administration of urea does not appear to pose a serious threat to the integrity of the ciliary epithelium.

Electron microscopic studies on the blood-aqueous barrier of prostaglandin-treated rabbit eyes. I. Iridial and ciliary processes

Prostaglandins E1 and E2 were applied topically to rabbit eyes. Structures related to the blood-aqueous barrier in the iridial and ciliary processes, as well as the permeability of the ciliary epithelium to the protein tracer horseradish peroxidase, were studied with the electron microscope. Marked morphological changes, including dilations of the intercellular spaces and separation of the two epithelial cell layers, were found in the epithelium of the iridial processes. Only minor structural changes were found in the epithelium of the ciliary processes. Leakage of peroxidase through the intercellular spaces of the epithelium was demonstrated in the iridial processes and in the anterior parts of the ciliary processes. In the ciliary vessels of the same regions, opening of interendothelial gaps, platelet aggregations, microthrombi, and haemorrhages were found. In a previous in vitro study on the effects of prostaglandins on the movement of peroxidase in the ciliary epithelium, no structural changes of the epithelium were found, and the epithelial diffusion barrier to peroxidase was found to be intact. It is assumed that the breakdown of this barrier in vivo is secondary to vascular changes.

In vitro studies on peroxidase movement in the epithelium of prostaglandin-treated rabbit ciliary bodies

The movement of horseradish peroxidase (HRP) in the epithelium of isolated rabbit iris/ciliary body preparations, has been studied with the electron microscope. HRP was applied at the stromal side of the epithelium, and was left for 60 and 120 min. The distribution pattern of HRP found in the epithelium of the iridial and ciliary processes is consistent with in vivo studies, i.e. the progression of HRP is blocked at the site of the zonula occludens of the superficial epithelium. The HRP distribution pattern found in the iris epithelium indicates that also the superficial epithelial cells of this epithelium are girdled by zonulae occludentes. Specimens treated with prostaglandins E1, E2, and F2alpha, showed no change in the epithelial distribution pattern of HRP, and the occluding zonules were found to be intact.

Freeze-fracture analysis of junctional complexes in human ciliary epithelia

The junctions connecting the epithelial cells of the human ciliary body were studied on replicas of freeze-fractured preparations. Gap junctions are present at the lateral surfaces of the cells of the pigmented epithelium. Gap junctions, isolated linear strands and desmosomes are found between the apical surfaces of pigmented and non-pigmented epithelial cells. Gap junctions, zonulae occludentes and desmosomes join the lateral surfaces of the non-pigmented epithelial cells. These findings suggest that the blood-aqueous barrier is located at the level of the zonulae occludentes of the non-pigmented epithelial cells. This is consistent with previous studies using thin sections and tracer substances. The function of the many gap junctions connecting the cells of the two epithelial layers is obscure since their presence is not consistent with electrophysiological data from this region.

Effect of prostaglandin on the blood aqueous barrier in the rabbit ciliary process

Distinct structural changes occur in the rabbit ciliary epithelium following intravitreal injection of prostaglandin E-1 (PGE-1). Up to four hours after PGE-1 administration, alteration of the pigmented epithelium was characterized by dilated intercellular spaces and the disruption of many intercellular junctions. The nonpigmented epithelium demonstrates a spectrum of morphologic variation from only some thinning of cytoplasmic processes to area of severe distortion. In these regions, marked thinning of the nonpigmented cells occurs in association with an absence of apical tight junctions. This alteration of the nonpigmented epithelium and its tight junctions allows for the leakage of proteins into the posterior chamber which is consistent with the breakdown in the blood-aqueous barrier. The temporal sequence of these changes would suggest a differential susceptibility of the pigmented and nonpigmented layers with the pigmented layers being affected earliest and the nonpigmented epithelium altered subsequently. The recovery of this epithelial change was rapid and complete and demonstrated the transient effects of PG on the ciliary epithelium with recovery of the blood-aqueous function by 8 hours after injection.

Selective destruction of the pigmented epithelium in the ciliary body of the eye

Perfusion of the internal carotid artery with hypertonic solution selectively destroys most of the pigmented epithelium of the ciliary body of the monkey eye, converts fenestrated endothelium of capillaries to continuous endothelium, and transiently breaks down the blood-aqueous barrier. The nonpigmented epithelium regenerates and the intraocular pressure returns to normal even though the pigmented epithelium is permanently destroyed.