Facilitation of Artificial Retinal Detachment for Macular Translocation Surgery Tested in Rabbit

Abstract

purpose. For macular translocation surgery, the native attached retina has to be detached either locally or completely. Although different surgical techniques are used, there is a general search for supporting procedures that facilitate and accelerate the retinal detachment.

methods. Pars plana vitrectomies were performed in pigmented rabbits. In all experimental groups, a local retinal detachment was created by infusing the test solution with a thin glass micropipette attached to a glass syringe. In control animals a standard balanced salt solution was used at room temperature, in combination with a standard vitrectomy light source. In two test groups, a calcium- and magnesium-free solution was used for the vitrectomy, under illumination by a standard light source in group I (solution at room temperature) and group II (solution heated up to body temperature). In group III the rabbits were dark-adapted for half an hour, and then, during surgery, a red filter was used in front of the light source (standard balanced salt solution at room temperature). After the rabbits were killed at the end of surgery, the adherence of the retinal pigment epithelium (RPE) to the neural retina in the detachment area was quantified microscopically, and the morphologic integrity of the detached retinal tissue was examined by light and electron microscopy. No electrophysiology was performed.

results. In all four groups, it was possible to detach the retina. The maximum adherence of the RPE cells to the neural retina was observed in the control group. Virtually no decrease in adherence was found in test group II (36°C solution without calcium and magnesium), whereas a significant decrease was seen in both group I (calcium- and magnesium-free solution at room temperature) and group III (dark adaptation–red light technique; standard balanced salt solution at room temperature). In none of the experimental groups was any obvious damage of the retinal structure observed, even after exposure to the test solutions for 60 minutes.

conclusions. Both dark adaptation (red illumination) and the use of a calcium chloride– and magnesium chloride–free solution (at room temperature) can facilitate retinal detachment in macular translocation surgery. Both techniques are proposed as a gentle support for the operation, because they protect an intact RPE cell layer and do not cause retinal damage at the ultrastructural level.

Age-related macular degeneration (AMD) is the most frequent cause of blindness in people aged 50 years and more. ¹ The development of subfoveolar neovascularization is one of the dominant reasons for the loss of vision in AMD. Because of the poor prognosis regarding visual function after laser treatment, ^{2 3 4} new operative techniques have been developed to treat this vision-threatening disease. The surgical excision of subfoveal membranes, which represent approximately 75% of all choroidal neovascularizations in AMD, does not result in a convincing improvement of visual function. ^{5 6} As another approach, the translocation of the macula onto a healthy tissue was considered. ^{7 8} Machemer and Steinhorst ^{9 10} were the first to perform the technical procedure of rotating the retina and translocating the fovea. Macular translocation may constitute a suitable therapy for AMD. Promising results have been reported, ¹¹ and a critical review has recently been published. ¹² The main problem with the procedure lies in the artificial separation of the neural retina from the retinal pigment epithelium (RPE)—a procedure that is necessary to translocate the macula but is very difficult and time consuming. 

The mechanisms of retinal adhesion are complex and multifactorial, because a variety of physical and physiological forces impinge on the subretinal space and its matrix. ¹³ However, it has been demonstrated that retinal adhesion may be modulated by a variety of conditions, including changes in local osmolarity, pH, and other ionic concentrations. ¹³ In particular, a reduction of calcium and magnesium in the subretinal space decreases retinal adhesion, even in vivo ^{14 15} —an effect that is reversible within a few minutes. ¹⁶ Temperature has also been shown to affect the adhesive retinal force. In enucleated eyes it is possible to reduce retinal adhesion by raising the eyes to body temperature. This effect is increased by incubating the tissue in a calcium- and magnesium-free solution at 37°C. ¹⁷ Furthermore, the effect of light exposure on retinal adhesion has been investigated in rabbits. A significant difference in the retinal adhesive forces was found between light- and dark-adapted nondetached retinas. ¹⁸  

Thus, there are several promising approaches to reducing retinal adhesion in macular translocation surgery. However, only a few of the mentioned procedures have been applied in vivo ^{14 15} ; most have been tested in vitro, ^{16 17 18 19} although retinal adhesion is known to change rapidly after enucleation. ¹³ Furthermore, the cleavage zone between photoreceptors and RPE was examined, ^{14 18 19} but none of the studies was devoted to examining any possible damaging side effects on the (inner) retinal tissue. Such side effects, however, would greatly limit or even prevent the application of a technique to human retinal surgery. Thus, we decided to perform a series of experiments in an in vivo model, the rabbit, to assess both retinal adhesion and structural maintenance of the retina by using a calcium- and magnesium-free solution at low and high temperatures (with standard illumination) and a standard balanced salt solution after dark adaptation and surgery under dim red light. The degree of the RPE’s adherence to the neural retina was taken as an index for the force of adhesion. ¹⁷ As a main point, the structural integrity of the detached retinas was studied by light and electron microscopy throughout the thickness of the tissue up to the inner limiting membrane. 

Materials and Methods

Retinal Surgery

All experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Pigmented 2- to 4-kg rabbits (New Zealand) of both sexes were used. Animals were kept for 12 hours at daylight. A mixture of ketamine hydrochloride (1 ml/kg) and xylazine hydrochloride (1 ml) was used to anesthetize the animals. The pupil of the right eye was dilated by a topical application of 1% tropicamide and 5% phenylephrine hydrochloride. The eye was subsequently protruded and immobilized. Pars plana sclerotomies were performed under an operating microscope with a 20-gauge needle in the upper nasal and upper temporal quadrants 3 mm behind the limbus. A 4-mm infusion cannula was inserted through one sclerotomy. A circumscript vitrectomy was performed in the area of the planned retinal detachment through the other sclerotomy site, by using a vitrectomy system (Ocutome; Alcon Surgical, Irvine, TX). Thin glass micropipettes (40-μm inner diameter), attached to a 250-μl glass syringe (Hamilton, Reno, NV) and introduced through the sclerotomy, were used to raise a subretinal bleb and to create a retinal detachment. When the tip of the pipette was positioned close to the inner limiting membrane, a stream of the solution was administered to the retinal surface that produced a small retinal hole and a subretinal bleb. The rate of subretinal fluid injection was approximately 300μ l/min. The sclerotomies and the overlying conjunctiva were then closed with 6-0 Dexon (Dexon GmbH, Spangenberg, Germany) sutures. 

In this study, we used balanced salt solutions BSS plus I and II (Alcon Pharma, Fort Worth, TX) as a standard intraocular solution and BSS plus I as a calcium chloride– and magnesium chloride–free intraocular solution for infusion (vitrectomy) and bleb formation (Table 1) . The latter solution was chosen to guarantee applicability in human patients, standardization, and comparability with the results of others. ²⁰ In earlier animal experiments, Ca²⁺-containing and Ca²⁺-free Hanks’ solutions were used. ¹⁴ There are no notable differences between Alcon BSS plus and Hanks’ solution in ionic composition, osmolarity, and pH. Random charges of BSS plus I and II and BSS plus I (Alcon) were measured for their pH and osmolarity (Osmomat 030; Gonotec, Berlin, Germany); no significant differences were found between the two solutions (pH 7.27 versus 7.25; 313.8 mOsm versus 298.5 mOsm, respectively; the latter difference of 15 mOsm, due to the omission of CaCl₂, MgCl₂, and glucose, is not sufficient to account for the observed changes in adhesion ¹³ ). 

The calcium- and magnesium-free solutions were used either at room temperature (21°C –23°C) or at 36°C, respectively. In the latter case, the bottles and tubes were insulated to maintain the temperature of the solution until it entered the eye. In one series of experiments, the animals were dark adapted, and the surgery was performed under a dim red light. For this purpose, a filter (diameter, 60.0 mm; thickness, 3.0 mm; λ_c = 665 ± 6 nm; model RG 665; Schott Glas, Mainz, Germany) was placed in front of the light source. Despite reduced visibility of the retina under these conditions, the surgery was feasible. 

For assessment of retinal adhesion, both vitrectomy and retinal detachment were performed within approximately 5 to 10 minutes. To study the structural maintenance of the retinas, the tissue was examined after 60 minutes’ exposure to the test solution (continuous perfusion of the eyes in situ), because longer times are required for artificial retinal detachment in humans than in rabbits. We created the retinal detachment in regions B and C as defined by Endo et al. ¹⁷ The area of detachment was labeled by producing four endodiathermic points at its margins. After the end of the experiments, the animals were killed and the eyes enucleated. Table 2 summarizes the experimental conditions for the controls and the three experimental groups. 

Preparation of Retinal Tissue

Immediately after enucleation, the sclera was circumferentially dissected 2 mm posterior to the limbus. Four radial-relaxing incisions toward the optic nerve were made in the sclera, and the sclera was peeled away from the periphery to the optic nerve. Subsequently, the choroid and the RPE were peeled off. The detached retinal area (the bleb within the borders of the four endodiathermic points) was excised. Its size (approximately 3.5 mm in diameter) was almost equal throughout the series of experiments. It was then placed on slides (with its photoreceptor surface up) and either used for quantification of the adherent RPE or immersion-fixed for electron microscopy. 

Quantification of the Adherent RPE

The unstained detached retinal tissue was studied by a conventional light microscope (Jenaval; Carl Zeiss, Oberkochen, Germany). With a camera CF 8/4 (KAPPA optoelectronics GmbH, Gleichen, Germany) in combination with a software program (AnalySIS; SIS Software Imaging; Münster, Germany) images were obtained at the level of the photoreceptor outer segments. When a ×3.5 lens was used, almost the total bleb area was visualized at once. A computer and the image analysis software (AnalySIS; SIS) were used to calculate the amount of adherent RPE on the detached retina in relation to the area of the detached retina. A measuring raster was displayed on the image of the retinal surface. Within the individual squares of a raster (each corresponding to 300 × 300 μm at the retinal surface) along both diagonals (i.e., 12 squares for each bleb), the natively black RPE areas were binarized. Thereafter, the percentage of the RPE cell-containing (black) area was calculated in comparison with the total retinal surface area. This percentage was used as an indirect measure of retinal adhesion. 

This assay was introduced by Endo et al. ¹⁷ in an in vitro study. It should be noted that in comparison with their data (80% and more of the total retinal area with adherent RPE), ¹⁷ we found much lower levels (generally <10% as a mean; cf. Fig. 2 ). This discrepancy is probably due to methodical differences: Whereas Endo et al. peeled the retina from the underlying tissue using jeweler’s forceps (i.e., exerted a force that may have caused a disruption anywhere between the retina and the fixed sclera, in principle), we injected a solution between the retina and the RPE. Under the latter condition, by definition, only local RPE adherences to the retina were possible. Furthermore, Endo et al. reported that their peeling rate was approximately 120 mm/min, whereas in our case the detachment rate was lower by at least one order of magnitude (the 3–4-mm diameter bleb was created within approximately 20 seconds). 

Because the collection of the assayed areas was independent of any bias of the examiner, this part of the experiments was not performed as a blind study (i.e., the examiner was aware of the treatment group studied). The data (percentages of the RPE cell-containing area, from each 12 assayed squares of 40 animals studied) were analyzed by computer (SPSS software; SPSS Science, Chicago, IL) for possible correlations with the four experimental groups or with the sex of the animals. Although there was no dependence on the sex of the animals (data not shown), the Mann–Whitney test results revealed significant differences between the experimental groups (see Fig. 2 ). 

Electron Microscopy

The detached retina was fixed in a buffered mixture of 2.5% glutaraldehyde and 3% paraformaldehyde (PFA) for 3 hours and afterward in 4% PFA overnight, rinsed, and post-fixed in 1% osmium tetroxide for 2 hours. The tissue was then rinsed, dehydrated in ethanol, and stained overnight in 70% ethanol saturated with uranylacetate. After further dehydration in absolute ethanol and propylene oxide, the samples were embedded in Araldite (Ciba–Geigy, Basel, Switzerland) or Durcupan (Fluka, Neu-Ulm, Germany) and sectioned on an ultramicrotome (Ultracut; Leica, Heidelberg, Germany). Most of the ultrathin sections were stained with lead citrate. They were studied using an omega electron microscope (Leo EM 900 or EM 912; Zeiss). Some semithin sections were cut, stained with cresyl violet, and studied by light microscope. 

Another series of fixed retinas, as described, were used for scanning electron microscopy. After fixation (1 hour in 1% osmium tetroxide buffered in 1% phosphate buffer [PB] at 4°C), the tissue was rinsed three times for each 10 minutes in 1% PB and dehydrated in ethanol (up to the final three times in 100%). Thereafter, critical-point drying (CPD 030; Bal-Tec, Balzers, Lichtenstein) was performed. The tissue was then placed on stubs by means of self-adhering carbon tabs and sputtered with Au (thickness, 20 nm) by using argon plasma (MED 020; Bal-Tec) for 40 seconds at 40 mA, with a probe–target distance of approximately 50 mm. The tissue was studied by microscope (Leo 1430; Zeiss). 

All morphologic work was organized as a blind study (i.e., the examiner was masked to the treatment group). 

Results

Retina–RPE Adhesion

Under all experimental conditions tested, subretinal blebs and retinal detachment were created easily within a few minutes. There were, however, apparent differences in the amount of adherent RPE. Figure 1 shows samples of neural retinas (with photoreceptor outer segments up) after retinal detachment had been induced under different conditions. The lower magnifications (Fig. 1 , left column) provide a view of larger retinal areas (1.7 × 2.2 mm), whereas the higher magnifications (Fig. 1 , right column) show arrangements of individual adherent RPE cells. It is apparent that at room temperature, less RPE was adherent when the retinal detachments were performed with the calcium- and magnesium-free solution (Figs. 1C 1D) than when performed with a standard intraocular solution (Figs. 1A 1B) . By contrast, if the calcium- and magnesium-free solution was used at body temperature, the adherence of RPE to the neural retina was similar to that under standard conditions (Figs. 1E 1F) . When the retinal detachment was created with the standard solution in dark adapted animals, virtually no RPE was adherent to the neural retina (Figs. 1G 1H) . Similar results were obtained when scanning electron micrographs were studied (compare with Figs. 4A 4B 4C ). In some cases, large sheets of continuous RPE adhered to the light-adapted retina after detachment in standard solution (Fig. 4A) . This was never observed when the calcium- and magnesium-free solution was used or when the animals had been dark adapted. 

Figure 2 shows the statistical analysis of these observations as a box plot diagram. A highly significant reduction in RPE adherence was achieved by using the calcium- and magnesium-free solution at room temperature (P < 0.01) and, particularly, by dark adaptation of the surgically treated eyes (P < 0.001). 

Morphologic Integrity of the Detached Retina

Because the use of a calcium- and magnesium-free solution may exert damaging effects on the neural retina, the structural maintenance of the tissue was studied after retinal detachment in this and in the control solution. To our surprise, there was considerable intraretinal swelling immediately after the detaching procedure (5–10 minutes), even if the standard solution was used. The degree of swelling was usually stronger under control conditions (Fig. 3B ) than with the calcium- and magnesium-free solution (Fig. 3A) . A transmission electron microscopic analysis of the outer segments of the photoreceptor cells revealed no apparent difference between the two conditions (Figs. 3C 3D) . Scanning electron microscopy (Figs. 4A 4B 4C ) also failed to demonstrate visible effects of the experimental conditions on the maintenance of photoreceptor segments. 

Because macular translocation surgery is often time consuming, the retinas were further investigated by electron microscope after a 1-hour exposure to the test solution (Figs. 5 6 7) . After this time, ultrastructural alterations of the tissue were observed under all conditions studied, although the most severe lesions were found in the retinas detached under standard conditions (Fig. 5) . The lesions included extracellular and perinuclear edema (empty spaces, indicated by asterisks in Figs. 5 6 7 ), swelling of mitochondria (Figs. 5 6 7 ; arrows), and distortion of photoreceptor outer segments (Figs. 5E 6E ; double arrows), the latter of which was rarely observed in the dark-adapted retinas (Fig. 7E) . It can be safely concluded that neither dark adaptation nor the use of a calcium- and magnesium-free solution caused retinal damage exceeding that observed under standard conditions; rather, these two conditions improved the maintenance of the retinal ultrastructure. 

Discussion

If the macula is to be translocated onto an area of healthy RPE, as described by Machemer and Steinhorst ^{9 10} in patients with AMD, strong mechanical forces are necessary to separate the neural retina from the RPE. Diminishing the adhesive forces between retina and RPE is highly desirable to reduce the duration of surgery (i.e., the accumulation of intraoperative risks) as well as to minimize the damage of both tissues caused by mechanical stress (i.e., the postoperative risks due to holes in the RPE sheath and/or to photoreceptor cell damage, for instance). It should be kept in mind, however, that a procedure suitable to solving these problems may cause intraretinal damage, which then may worsen the long-term prognosis of the surgery. The present study was designed to provide data that allow for clear-cut conclusions regarding the question of which of the methods tested is preferable for use in human retinal surgery. Although there are interspecies differences in retinal adhesion, ^{13 21} the principal mechanisms have been shown to be similar in rabbits and primates, ^{13 21} and the rabbit thus seems to be an applicable model for this purpose. In this study, no electrophysiology was performed; rather, our conclusions are based on ultrastructural examination of the tissue. 

Calcium- and Magnesium-Free Solutions: Reduced Retinal Adhesion and No Additional Damage to the Tissue

Our data show that a calcium- and magnesium-free solution, if used at room temperature, reduces retinal adhesion. This is in accordance with earlier data obtained from human eyes after death ^{19 21} and from living rabbit eyes, ^{14 21} in which it has been shown that only approximately 30% of normal retinal adhesive force remains in a low-calcium environment. ¹⁴ Thus, there is ample evidence for the conclusion that the absence of calcium in the subretinal space helps to prevent RPE adhesion to the overlying retina. The reason for calcium’s effect(s) on retinal adhesion is still unclear. The calcium concentration may affect the interphotoreceptor matrix, subretinal transport processes, and subcellular components of the RPE. ^{13 14 22}  

The use of a calcium- and magnesium-free solution, in addition to decreasing the duration of surgery ²³ and preventing the codetachment of RPE sheets together with the retina, also preserves the outer segments of the photoreceptors (Fig. 6E versus Fig. 5E ). Moreover, it caused no specific structural damage of the retina proper (Figs. 3A versus Fig. 3B ), even when the solution was applied over a time of 1 hour (Fig. 5 versus Fig. 6 ). Machemer and Steinhorst ⁹ observed a loss of some outer segments of photoreceptors when they used a standard balanced salt solution. We also found some degree of damage to the outer segments of the photoreceptors as well as to various intraretinal structures (Figs. 3B 3D 5) when the balanced salt solution was used. In comparison to these alterations, the damaging effects of the calcium- and magnesium-free solution (Figs. 3A 3C 6) were less rather than more pronounced. Similar observations have been published, in preliminary form, by other investigators. ^{24 25}  

Body Temperature: Reversed Effect of Calcium- and Magnesium-Free Solutions

Warming to body temperature of the calcium chloride- and magnesium chloride-free solution seemed to be a method for further weakening retinal adhesive forces, in that an adhesion-reducing effect of elevated temperatures had been reported. ¹⁷ In our experiments, we obtained contradictory results (Figs. 1E 1F 2) . The reasons for this discrepancy are unclear, although significantly different experimental conditions were used. The study of Endo et al. ¹⁷ was performed on enucleated eyes. In our in vivo experiments, the warming up of the calcium- and magnesium-free solution produced results that were not superior to those obtained with the standard balanced salt solution. 

Dark Adaptation: Retinal Detachment Facilitated and Reduced Retinal Damage

We found a markedly reduced loss of RPE cells when surgery was performed in dark-adapted animals under a dim red light (656 nm). This is in accordance with the data of Owczarek et al. ¹⁸ who found that a 20% stronger force was necessary to detach a retina in light-adapted eyes than in dark-adapted ones. They noted that in dark-adapted eyes the cleavage occurred primarily in the outer segments, whereas in light-adapted eyes the RPE was more extensively involved. ¹⁸ The normal function of photoreceptor cells depends on multiple interactions between the photoreceptors and the RPE cells. These interactions are partly mediated by the interphotoreceptor matrix (IPM) surrounding the photoreceptor segments and the RPE microvilli. In the rat retina, the IPM has been shown to undergo a major shift between light and dark adaptation. ²⁶ In light adapted rats, there was a strong concentration of densely stained IPM in the apical outer segment zone, at the interface to the RPE. This may be one reason for the strong adherence of RPE to the neural retina under standard illumination conditions. The underlying mechanism of the light-evoked IPM changes could be a light-induced release of calcium by the photoreceptors into the interphotoreceptor space. ^{27 28} This may also be one of the reasons that RPE cell loss is decreased in the calcium- and magnesium-free solution. During dark adaptation, the distribution of the IPM is changed. The interstitial zone of the IPM is more concentrated, and the basal outer segment concentration is lost. ²⁶ This could be the explanation for the decrease of RPE cell loss in dark-adapted animals. 

Although further research is necessary to reveal the exact mechanism(s) of the effects of dark adaptation onto retinal adhesion, it can be stated that among the procedures tested in our study, dark adaptation is a very effective means to reduce retinal adhesion (Fig. 2) and causes reduced damage to the retina, particularly to the photoreceptor outer segments (Fig. 7) . Thus, both the dark adaptation technique and the use of calcium- and magnesium-free solutions are proposed as useful tools for creating the artificial retinal detachment in macular translocation surgery. 

 Supported by the Retina Research Foundation, Houston, Texas; the Bundesministerium für Bildung, Forschung, und Technologie (BMBF); the Interdisciplinary Center for Clinical Research at the University of Leipzig (01KS9504, Project C1 and C5); and the Deutsche Forschungsgemeinschaft, Graduiertenkolleg“ Intercell” (GRK 250/2-99).

 Submitted for publication June 1, 2000; revised October 2 and December 7, 2000; accepted January 12, 2001.

 Commercial relationships policy: N.

 

Table 1.

 

Solutions Used for Artificial Retinal Detachment

Contents	NaCl	KCl	CaCl2	MgCl2	Na2PO3	NaHCO3	Glucose	GSSG*
BSS plus I and II (standard solution)	122.2	5.08	1.04	0.98	3.0	25.0	5.11	0.30
BSS plus I (Ca2+-, Mg2+-free)	122.2	5.08	—	—	3.0	25.0	—	—

 Data are expressed in millimolar. BSS plus is produced by Alcon Surgical (Houston, TX).

*  Glutathione (oxidized).

Table 2.

 

Experimental Conditions for Artificial Retinal Detachment

Group	Total Subjects	Number of Animals							Dark Adaptation	Solution	Temperature (°C)
				Adherence Assay	TEM (5 min)	SEM (5 min)	TEM (1 hr)
Controls	15	11	2	1	1	No	Standard*	21–23
Test group I	16	12	2	1	1	No	Ca2+/Mg2+-free, †	21–23
Test group II	9	9	—	—	—	No	Ca2+/Mg2+-free, †	36
Test group III	11	8	1	1	1	Yes	Standard*	21–23

 BSS plus is produced by Alcon Surgical (Houston, TX). TEM, transmission electron microscopy; SEM, scanning electron microscopy.

*  BSS plus I and II.

†  BSS plus I.

Figure 1.

 

Qualitative assessment of retina–RPE adhesion. Artificially detached unstained retinas (photoreceptor surface up) shown at lower (left column: retinal areas, 1.7 × 2.2 mm) and higher magnification (right column: retinal area, each 0.5 × 0.6 mm). The following experimental conditions were used: (A, B) control group: standard intraocular solution containing calcium and magnesium, room temperature, light adaptation; (C, D) test group I: intraocular solution without calcium and magnesium, room temperature, light adaptation; (E, F) test group II: intraocular solution without calcium and magnesium, 36°C, light adaptation; (G, H) test group III: standard intraocular solution, room temperature, dark adaptation. Adherent RPE cells appear as black dots.

Figure 2.

 

Quantitative assessment of retina–RPE adhesion. Test groups are as described in Figure 1 . *Significant differences compared with group I (**P < 0.1, ***P < 0.01).

Figure 3.

 

Morphologic maintenance of the detached retinas immediately after detachment surgery (after 5–10 minutes). (A, B) semithin sections; (C, D) transmission electron micrographs. (A, C) Standard intraocular solution without calcium and magnesium; (B, D) control conditions (standard intraocular solution containing calcium and magnesium). Although intraretinal swelling was visible, even at the light microscopic level (A, B; arrows), the outer segments of the photoreceptor cells were rather well maintained at the electron microscopic level (C, D).

Figure 4.

 

Scanning electron microscopy of the sclerad surface of retinas detached in standard intraocular solution containing calcium and magnesium, at room temperature with light adaptation (A); calcium- and magnesium-free solution, room temperature, light adaptation (B); and standard intraocular solution, room temperature, dark adaptation (C). Whereas large plates of adherent RPE (A, right) were regularly found under standard conditions, such plates were never observed when the calcium- and magnesium-free solution was used or when the animals were dark adapted. Scale bar, 30 μm.

Figure 5.

 

Ultrastructure of the detached retina after a 1-hour exposure to a standard intraocular solution containing calcium and magnesium, at room temperature with light adaptation: (A) vitread retinal surface and nerve fiber layer; (B) inner plexiform layer; (C) outer nuclear layer; (D) photoreceptor inner segments; (E) photoreceptor outer segments. There was considerable extracellular and perinuclear edema throughout the retina proper (A, B, and C; ). Within the subretinal space, most of the mitochondria in the inner segments of photoreceptors were swollen (D, arrows), and many outer segments were severely distorted (E, double arrows). Scale bars, 2 μm.

Figure 6.

 

Ultrastructure of the detached retina after a 1-hour exposure to an intraocular solution without calcium and magnesium, at room temperature with light adaptation: (A) vitread retinal surface and nerve fiber layer; (B) inner plexiform layer; (C) outer nuclear layer; (D) photoreceptor inner segments; (E) photoreceptor outer segments. Extracellular and perinuclear edema was found in the retina (A and C; ), but most of the photoreceptor inner segments contained apparently intact mitochondria (D), and the distortion of the outer segments was less pronounced (E) than with standard solution. The arrows in (B) and (D) point to mitochondria in various stages of swelling. The double arrows in (E) show distorted disks within photoreceptor outer segments. Scale bars, 2 μm.

Figure 7.

 

Ultrastructure of the detached retina after a 1-hour exposure to a standard intraocular solution at room temperature. The dark-adapted animal underwent surgery under dim red light: (A) vitread retinal surface and nerve fiber layer; (B) inner plexiform layer; (C) outer nuclear layer; (D) photoreceptor inner segments; (E) photoreceptor outer segments. There was not much extracellular edema (C, ), but several mitochondria were swollen in both the inner plexiform layer (B) and the photoreceptor inner segments (D). The outer segments of most photoreceptors were rather well maintained (E). The arrows in (B) and (D) point to mitochondria in various stages of swelling. Scale bars, 2 μm.