Detection of Early Neuron Degeneration and Accompanying Microglial Responses in the Retina of a Rat Model of Glaucoma

Abstract

purpose. To characterize the early reaction of retinal ganglion cells (RGCs) in a rat model of glaucoma using in vivo imaging and to examine the involvement of retinal microglia in glaucomatous neuropathy.

methods. Glaucoma was induced in adult female Sprague-Dawley rats by cauterizing two episcleral veins, which resulted in a 1.6-fold increase in intraocular pressure (IOP). Retinal ganglion cells were retrogradely labeled with the fluorescent dye, 4-[didecylaminostyryl]-N-methyl-pyridinium-iodide (4-Di-10ASP) and monitored in vivo after elevation of IOP using fluorescence microscopy imaging. The number of RGCs was quantified on retinal flatmounts. Dying RGCs were surrounded by activated microglia that became visible after taking up the fluorescent debris. Immunocytochemistry was conducted to characterize further the ganglion cells and microglia.

results. Cauterizing two of the four episcleral veins resulted in a consistent increase of IOP to 25.3 ± 2.0 mm Hg, as measured with a handheld tonometer. IOP remained high for at least 3 months in glaucomatous eyes. The earliest sign of RGC death was detected in anesthetized animals 20 hours after induction of glaucoma. RGCs continued to decrease in number over time, with 40% of RGCs having degenerated after 2.5 months. Fundoscopic examination of the optic nerve head revealed cupping 2 months after induction of glaucoma. In addition, microglia were detected on retinal flatmounts as early as 72 hours after induction. Activated microglia and RGCs were also identified immunocytochemically, with an antibody against ionized calcium-binding adaptor molecule (Iba)-1 and an antibody specific to the 200-kDa subunit of the neurofilament protein, respectively.

conclusions. Occlusion of episcleral veins is a reproducible method that mimics human glaucoma, with chronically elevated IOP-induced RGC loss. This study shows that in vivo imaging permits the detection of ganglion cells in the living animal in the early stages of the disease and highlights the importance of in vivo imaging in understanding ophthalmic disorders such as glaucoma. Secondly, activation of intraretinal microglia coincides with degeneration of RGCs in glaucoma.

Glaucomatous optic neuropathy is a chronic disease accompanied by visual field loss, cupping of the optic nerve head, and irreversible loss of retinal ganglion cells (RGCs)—one of the most characteristic features of this disease. Increased intraocular pressure (IOP) is still considered to be one of the major risk factors in glaucoma, although visual field loss may continue, despite successful lowering of IOP. It has become clear that, in addition to pressure control, neuroprotective measures may be relevant in the treatment of glaucoma. 

The clinical features of glaucoma are well described, but the mechanisms resulting in optic nerve damage and RGC death remain to be elucidated. ¹ In glaucoma, RGCs apparently die by apoptosis, also termed programmed cell death. Apoptosis has been observed in animal models ^{2 3} of the disease and in humans. ⁴ This form of cell death resembles the mechanism by which neurons are eliminated during early development of the central nervous system (CNS) and the retina. In addition to the primary consequences of elevated IOP, secondary degeneration occurs, caused by a deficiency of growth factors ⁵ and excitotoxic mechanisms. ⁶ Thus, glaucoma can also be considered a neurodegenerative disease, especially with regard to RGCs. The further characterization of the mechanisms involved in glaucomatous neuropathy thus requires reliable and inexpensive animal models of the disease. 

Various investigators have used rodents to develop glaucoma models, including the injection of hypertonic saline into aqueous humor collecting veins, ⁷ cauterizing two or three episcleral veins, ⁸ blocking aqueous outflow pathways by photo-coagulation, or injecting india ink into the anterior chamber of the eye. ⁹ Using these models, it is possible to evaluate the extent of RGC and optic nerve damage at various time points after induction of glaucoma. 

Retrograde labeling of RGCs from the superior colliculus (SC) with fluorescent dyes enables the simultaneous visualization and quantification of RGCs and microglia in retinal flatmounts. Microglia, activated during the process of neuronal degeneration phagocytose the dying neurons and take up the fluorescent RGC debris, becoming fluorescent in turn. ¹⁰  

However, no noninvasive method exists to date that enables the investigator to follow the course of glaucomatous neuropathy in the same animal over time. ¹¹ In this study, we used the technique of in vivo imaging ¹² to monitor the early onset of RGC death after inducing glaucoma by cauterizing and occluding episcleral veins in the adult rat. 

The role of nonneuronal cells in the glaucomatous retina is still a matter for debate, ¹³ but it is a well-established fact that astrocytes and microglia serve to protect the integrity of the nervous system during development and disease. Microglia belong to the mononuclear phagocyte system, and form the resident macrophages in the brain, spinal cord, and retina ¹⁴ and are involved in many neurodegenerative diseases. Their activation in human glaucomatous optic nerves ¹⁵ and retinas ¹⁶ and in the retinas of a rat model of glaucoma ¹² indicate that these cells are also involved in glaucomatous pathophysiology. 

The present study was conducted to determine the earliest time point of RGC death after the elevation of IOP and to evaluate the role of microglia in glaucomatous RGC death. 

Methods

All experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Female Sprague-Dawley rats weighing 180 to 250 g, were housed in a standard animal room in a 12-hour light–dark cycle with food and water provided ad libitum. All surgical procedures on the rats were performed with animals under general anesthesia induced by a mixture of 50 mg/kg ketamine (Ceva-Sanofi, Düsseldorf, Germany) and 2 mg/kg xylazine (Ceva-Sanofi), given intraperitoneally. After ocular surgery a topical antibiotic containing gentamicin (Gentamytrex; Dr. Mann Pharma, Berlin, Germany) was applied. All surgical manipulations were unilateral, with the contralateral eye serving as the corresponding control. 

Induction of Glaucoma and IOP Measurement

IOP was elevated by cauterizing two episcleral veins of the eye, as has been described previously. ^{2 8} This procedure resulted in blockage of the venous outflow, which represents 50% of the total area available for episcleral venous return. ⁸ The IOP was elevated to 1.6 times the control levels immediately after cauterization (see Fig. 1 ). 

IOP was measured while rats were lightly anesthetized by ether inhalation, because it has been previously demonstrated that anesthetics cause a reduction in IOP. ^{17 18} The eyes were additionally anesthetized with a drop of topical 0.5% proparacaine (URSA-Pharm, Saarbrücken, Germany). All measurements were performed once between 9 AM and 12 PM, immediately after surgery and every 7 days thereafter with a handheld tonometer (Tono-Pen XL; Mentor, Norwell, MA). On each eye measured, 10 tonometer readings were taken directly from the display of the instrument, recorded, and averaged. “Off” (or outlier) readings and instrument-generated averages were ignored. Animals in which IOP returned to normal were excluded from the study. 

For examination and photography of the optic discs of the rats in vivo, they were anesthetized and their pupils dilated. The optic nerve head was photographed after moistening the cornea with proparacaine, and planar images were obtained by placing a glass slide on the cornea. 

Retrograde Labeling and Quantification of RGCs

RGCs were retrogradely labeled with the fluorescent dye 4-Di-10ASP (4-[didecylaminostyryl]-N-methyl-pyridinium-iodide) purchased from Molecular Probes (Eugene, OR), as previously described. ¹⁹ This technique resulted in the exclusive labeling of RGCs in a uniform manner across the entire retina, thus providing a method to study their reaction to glaucoma in the living animal and enabling their quantification on retinal flatmounts. ¹⁹  

Briefly, the SC contralateral to the experimental eye was surgically exposed in anesthetized animals, as described earlier. A few solid crystals of 4-Di-10ASP were inserted into the superficial layers of the SC after which 100 μL Freund’s adjuvant (Sigma, St. Louis, MO) was pipetted onto the crystals. The cortical cavity was filled with gel foam (Pharmacia & Upjohn, Kalamazoo, MI) and the skin wound was sutured, followed by the application of an antibiotic. The animals were returned to their cages, and their health and behavior were monitored regularly. 

Induction of glaucoma followed retrograde labeling of RGCs after 7 days, which is the time required for the dye to be taken up by the axon terminals of the RGCs in the SC and be transported retrogradely to their somas in the retina. This dye does not leak from the RGCs and persists in the cells for at least 10 months. ¹⁰  

Animals were killed 3, 5, and 10 weeks after induction of glaucoma and flatmounted retinas were fixed in 4% paraformaldehyde overnight at 4°C and assayed for RGC density. Cells were visualized under fluorescence microscopy (Axiophot; Carl Zeiss, Oberkochen, Germany) using the excitation filter of 515 nm and the band-pass emission filter of 563 nm, which is optimal for this dye. Five areas per retinal quadrant, each corresponding to an area of 0.098 mm², at five different eccentricities (uniform central to peripheral distribution) of the retinal radius were counted at a final magnification of ×400. The optic disc served as the point of reference for these measurements (see Fig. 4 , insert). In each retina, the total number of ganglion cells counted was divided by the area analyzed to determine the ganglion cell density (RGCs per square millimeter) for that retina. The average density of RGCs was compared between glaucomatous and nonglaucomatous retinas, to determine changes in the number of cells after episcleral vein occlusion. 

Microglia could also be visualized, in that they take up the fluorescent dye 4-Di-10ASP during the process of ganglion cell death. Microglia were identified based on their highly branched nature and small size (see the Results section for additional details). 

Statistical analysis was performed with the Student’s t-test and results were considered significant at the 95% level of confidence or higher. All data for ganglion cell quantification are expressed as the mean ± SD. 

Immunocytochemistry

To further characterize and identify the microglia seen on the retinal flatmounts, double-staining immunocytochemistry was performed. The ionized calcium-binding adaptor molecule (Iba)-1 protein was used to identify activated microglia, whereas RGCs were identified using an antibody specific to the 200-kDa subunit of the neurofilament protein. For this, retinal flatmounts were prepared on nitrocellulose filter, fixed overnight in 4% paraformaldehyde and then embedded in Tissue-Tek (Sakura Finetek, Torrance, CA). Frozen sections (12 μm) were cut and collected on gelatinized slides and stored at −20°C. The sections were fixed in cold acetone for 10 minutes, washed three times for 5 minutes each in phosphate-buffered saline (PBS) and blocked with 10% fetal calf serum (FCS) for 30 minutes. The polyclonal rabbit anti-rat Iba1 (gift of Yoshinori Imai, National Institute of Neuroscience, Kodaira, Tokyo, Japan) antibody was diluted in FCS (dilution 1:100) and the sections incubated overnight at 4°C. After the slides were rinsed three times for 5 minutes each in PBS, they were incubated with an anti-rabbit Cy2 antibody (dilution 1:200 in FCS; Dianova, Hamburg, Germany) for 30 minutes at room temperature (RT) and washed three times for 5 minutes each in PBS. The entire procedure was then repeated, beginning with application of the monoclonal anti-neurofilament 200 antibody (dilution 1:400 in FCS; Sigma) overnight at 4°C. After rinses (three times for 5 minutes each) in PBS, the sections were incubated for 30 minutes at RT with an anti-mouse tetramethylrhodamine isothiocyanate (TRITC)–conjugated secondary antibody (dilution 1:300 in FCS; Sigma). The sections were rinsed again in PBS (three times for 5 minutes each). Finally, the slides were coverslipped with antifade mounting medium (Mowiol; Hoechst, Frankfurt, Germany) and viewed with the appropriate filter on a microscope equipped with epifluorescence (Axiophot; Zeiss). Control samples were treated without the primary antibodies. 

In Vivo Imaging of RGCs

To visualize labeled RGCs in the living animal, a noninvasive imaging technique was used ¹² that enables the study of the effect of raised IOP on a mixed group of ganglion cells within a specific area of the retina over time (see Fig. 2 ). To this end, the stage of a conventional microscope (Axiovert; Zeiss) was removed to provide space for a platform that could be raised and lowered and on which the animal was placed. A head-holder fixed to the platform enabled the head of the animal to be moved along three axes so that different areas of the retina could be visualized. With a lens with a magnification of 2.5× or 5×, it was possible to view the retinal surface and clearly discern 4-Di-10ASP–labeled RGC somas. In vivo imaging was performed with animals under general anesthesia induced by a mixture of 50 mg/kg ketamine and 2 mg/kg xylazine, administered intraperitoneally. Anesthesia was maintained for the entire procedure. The corneal surface was treated with 0.5% proparacaine HCl and mydriatic drops (Mydriaticum; Pharma Stulln, Stulln, Germany). The corneal surface was kept moist, and a glass slide was placed on the cornea to obtain planar images. Applanation of the cornea did not have any observable influence on IOP or on optic nerve head morphology in control animals. Long-term corneal thickness also remained unaffected by the procedure. Retinas were observed and documented photographically every 2 hours after elevation of IOP to determine the earliest time point of RGC death. A in-built camera makes it possible to photographically document the retina with exposure times of 100 to 150 ms with a high-speed (3200 ASA), black-and-white film (T-MAX; Eastman Kodak, Rochester, NY). The retinal areas under examination were sketched with blood vessels used as landmarks, which made it possible to relocate RGCs at the various time points studied. Preliminary studies have shown that exposure to mercury vapor light during in vivo documentation of RGCs has no adverse effect on these cells. ¹²  

Results

Ten animals were used for the in vivo study, and IOP was elevated 1 week after retrograde labeling. IOP was measured using a handheld tonometer (Tono-Pen; Mentor), as described earlier. Elevation of IOP was noted immediately after the episcleral veins were cauterized. The IOP was compared between control and experimental eyes. Mean (±SD) control pressure was 15.4 ± 1.4 mm Hg, whereas the IOP was 1.6 times higher in the cauterized eye (mean, 25.3 ± 2.08 mm Hg). The control IOP ranged between 13 and 16 mm Hg in different animals, and the IOP of cauterized eyes ranged between 22 and 30 mm Hg. These readings were consistently recorded for longer than 8 weeks (Fig. 1) and correspond to values obtained by other groups. ^{20 21} Mean IOP in rats appeared to be nearly identical with that recorded in awake humans, ^{22 23} rabbits, ²⁴ and anesthetized monkeys. ²⁵  

To determine the earliest time point at which RGCs disappear after elevation of IOP the in vivo imaging method ¹² was used (Fig. 2) . The earliest signs of cell disappearance were detectable around 20 hours after elevation of IOP (Fig. 3a 3b) . 

Twenty animals were used in the study, which was performed to quantify RGC loss in retinal flatmounts (Fig. 4) . Animals were killed 3, 5, and 10 weeks after elevation of IOP, and RGCs were quantified and compared with counts in control retinas (n = 4), which had an average (±SD) of 1924 ± 31 RGCs/mm² prelabeled cells. These data are in agreement with previously published data on labeling RGCs quantified with the use of 4-Di-10ASP. ²⁶  

Three weeks after elevation of IOP 1657 ± 78 RGCs/mm² remained (n = 4), which corresponds to 86% of the control number. Five weeks after elevation of IOP 1400 ± 90 RGCs/mm² remained, corresponding to 73% of the control number (n = 4) . Further reduction was seen 2.5 months after elevation of IOP 1141 ± 130 (n = 5). RGCs corresponding to 59% of the control remained. The difference between the number of cells in treated specimens and the number in control specimens was statistically significant at each time point of quantification (Student’s t-test; P < 0.001). 

The fluorescent dye persisted within the RGCs and did not leak from the cells. ¹⁰ Microglia, which are activated by neuronal death in retrogradely labeled rat and mouse retinas, ¹⁰ can be visualized as well, as they take up the fluorescent dye released by the dying neurons, which they phagocytose. Hence, retinas of glaucomatous eyes were prepared as flatmounts at 1, 2, 3, 7, and 14 days and 3 and 5 weeks after elevation of IOP. Three days after elevation of IOP, a population of nonganglionic cells appeared to contain the fluorescent dye in the ganglion cell layer (GCL) and inner plexiform layer (IPL). These cells had small, elongated, irregularly shaped perikarya and dendritic extensions within the GCL. Their dendritic territories were much smaller than those of ganglion cells (Fig. 5) . The fluorescent dye was located in the cytoplasm and branches, whereas the cell nuclei were unlabeled. Microglia were primarily observed in the peripheral retina, located in the vicinity of RGCs and blood vessels at this time point. Preliminary data indicate that approximately 83/mm² labeled microglia were present in the retina 2 weeks after induction of glaucoma. Further quantification of microglia is part of an ongoing study. Microglia are probably involved in the degeneration of RGCs, because they were closely associated with them (Fig. 5) . Microglia were no longer visible at 2.5 months after elevation of IOP. 

The microglial nature of the labeled cells (Fig. 5) was confirmed by conducting immunocytochemistry on retinal frozen sections with the Iba-1 antibody, which recognizes activated microglia in the CNS and retina. Iba1-positive cells were detected 1 week after elevation of IOP (Fig. 6) . These strongly stained, elongated cells with highly branched processes were observed in the RGC layer, suggesting that they were microglia. Double labeling with the neurofilament antibody was conducted on the same sections to localize RGCs and their axons in retinal sections. The Iba-1–positive microglia were present in close approximation to neurofilament-positive RGCs (Fig. 6) . 

Discussion

Nonhuman primates ³ and other models of glaucoma, such as the dog ²⁷ and rabbit, ²⁸ are expensive and therefore are limited in their usefulness in the study of glaucomatous neuropathy. The rat model is inexpensive and easy to handle, and the literature on the rat retina and retinal neuropathy is extensive. The model in which IOP is elevated by cauterizing two to three episcleral veins in the rat ^{8 20} is excellent for induction of primary open-angle glaucoma in a reproducible fashion. Because features characteristic of human glaucoma, such as increased IOP and progressive RGC death can be observed in this model, it is a valuable tool to gain a deeper insight into the cellular mechanisms of glaucoma, which are yet to be fully understood. 

Elevation of IOP may be due to a combination of vascular congestion and reduction of aqueous humor outflow. Obstructing venous return from the anterior chamber decreases aqueous drainage through the canal of Schlemm, thus increasing IOP. ¹¹ Intraocular pressure remained consistently elevated in cauterized eyes for at least 3 months, compared with the contralateral control eyes, and resulted in cupping of the optic nerve and loss of RGCs. 

The use of microscopy to monitor changes in the nervous system over time began with the introduction of fluorescent dyes that label living neurons. Because RGCs are the only retinal neurons that project axons to the SC, retrograde labeling with a fluorescent dye is an excellent technique to selectively label RGCs in the retina. ^{10 29} Quantification of RGCs lost as a consequence of elevated IOP was performed at different experimental end points. A 1.6-fold elevation in IOP resulted in progressive loss of RGCs in a linear fashion. These data on ganglion cell loss are in agreement with results obtained by other groups. ^{20 30} Loss of RGCs has been documented as early as 1 ^{2 12} and 2 weeks ^{11 20} after elevation of IOP. However, this and other methods of histologic evaluation on the extent of glaucomatous neuropathy were conducted at various terminal time points, which means that animals had to be killed at different time intervals. As observed by Bayer et al. ¹¹ research on glaucoma has not provided a noninvasive procedure to monitor the progression of neuropathy from the onset of the disease over time in the same experimental animal. Quantification of RGCs on flatmounts provides valuable information on the rate and number of dying cells but is not useful in the early stages of glaucoma, when it is reasonable to assume that only a few cells are dying. The use of retinal sections has the disadvantage that the number of events is low in any given section, and retinal location is usually difficult to reconstruct from sections. 

When faced with the challenge of studying the cellular events of a disease such as glaucoma in their natural environment, “static” and “dynamic” strategies can be adopted. The static approach involves collecting data from a timed series of animals, whereas the technically more difficult dynamic approach is to follow the fate of a group of cells or of one particular cell over time. ³¹ This allows the investigator to obtain many data points from one animal and capture the key dynamic events directly. 

The optic nerve injury model, although not a model for glaucoma, enables the extent of the primary damage to RGCs to be quantified and secondary degeneration to be demonstrated and assessed. ⁵ The proponents of the crush model suggest that a similar mechanism may underlie the spread of damage seen in glaucoma. This model has been used extensively to study the pathophysiological degradation of the damaged RGC and its neighboring cells. It is known that a large number of RGCs die on the sixth day after axotomy, and RGC death was observed as early as 2 hours after axotomy by the in vivo imaging technique. ¹² This technique is also useful to monitor RGC death after elevation of IOP. Groups of ganglion cells in different regions were monitored immediately after induction of glaucoma. The first signs of ganglion cell degeneration were evident 20 hours after induction. A static analysis of the glaucomatous retina does not enable the detection of individual dying ganglion cells scattered across the retina. Knowing when individual cells die may allow the debilitating effects of glaucoma and other optic neuropathies to be reversed in various animal models of these diseases, by making an attempt to rescue neurons that are still intact. This observation is important when one considers that the time frame for successful pharmacologic intervention is crucial, especially considering that neuroprotection is a concept that is gaining considerable importance in glaucoma therapy. 

The pathophysiologic mechanisms underlying ganglion cell loss in glaucoma remain unclear. Glial cells such as microglia and astrocytes are postulated to participate in the pathologic course of neuronal damage after mechanical and ischemic injury. ³² The normal adult CNS contains roughly as many microglia as neurons ¹⁴ and based on this numerical relationship alone it could be argued that microglia are extremely important in the healthy and diseased states of a neuron. Hence, these cells may also be involved in ganglion cell dysfunction, as seen in glaucoma. 

In the present study, occasional labeled microglia were visualized in vivo in addition to RGCs, but were not photographically documented because of the resolution of this system. In retinal flatmounts, microglia were clearly visualized beginning at 3 days after elevation of IOP. Mittag et al. ³³ observed that RGC loss in a similar rat model was focal—that is, patches devoid of labeled RGCs were evident in all quadrants of the retina after 12 weeks of elevated IOP. The investigators postulated that these patches develop when degenerating RGCs are phagocytosed by microglia. 

Microglia are highly reactive and mobile cells that represent the intraretinal phagocytic cells that play a role in the development of the retina and CNS, as well in various disease states. ²⁹ One of the major functions of the microglia in the developing CNS is the continuous removal of cell debris resulting from naturally occurring cell death. ¹³ In the Royal College of Surgeons (RCS) rat, a model of retinitis pigmentosa, microglia are known to migrate from the ganglion cell layer to the photoreceptor layer, where they are responsible for the clearance of degenerating photoreceptor outer segments and, at later stages, the entire cell. ²⁹ In response to retrograde degeneration of ganglion cells in the retina, microglia have been shown to proliferate ²⁹ and remain in the tissue for long periods. Retinal microglia respond to optic nerve axotomy with hypertrophy, enhanced enzymatic activity, and transient increase in number. ³⁴ Microglia are also activated in an animal model of diabetic retinopathy, as visualized by isolectin B4 binding. ³⁵ They are believed to be sensitive to K⁺ conductivity and can produce aggressive oxygen radicals and induce release of glutamate ³⁶ —substances postulated to participate in the possible pathophysiology of glaucoma. ^{6 37 38}  

Ito et al. ³⁹ demonstrated that Iba1 mRNA is specifically expressed by microglia in cultured brain cells and in the brain of rats. The Iba1 protein is specifically localized to ramified microglia and is not found in neurons, astrocytes, or oligodendroglia. Iba1 is specifically upregulated during the activation of microglia. ⁴⁰ It is involved in the Rho family of small guanosine triphosphatase (GTPase), Rac, and calcium-signaling pathways and may be necessary for cell mobility and phagocytosis. ⁴¹ In the middle cerebral artery (MCA) model of ischemia, Iba1 immunoreactive cells were detected in the ischemic core, and double labeling revealed that these cells are microglia. ⁴² Iba1-positive microglia were found in the RGC layer 3 days after elevation of IOP, indicating that microglia are activated during glaucomatous neuronal degeneration and migrate to the site of RGC death. These data are in accord with our proposal that microglia are activated by an internal or external insult to the tissue, such as increased IOP. 

Wang et al. ¹³ showed that microglia, Müller cells, and probably astrocytes respond rapidly within hours of elevation of IOP in the retina, as detected by upregulation of OX-42 and glial fibrillary acidic protein (GFAP) immunocytochemistry staining. The increased reactivity of the glial cells lasted 2 months and correlated well with the neuronal loss in the ganglion cell layer. In an another study, ⁴³ the same group showed that the number of microglia increased moderately in both the dorsal and ventral lateral geniculate nuclei. This increase advanced with time, but dropped at 3 and 4 months after induction of glaucoma. This finding corresponds with the observations made in the present study, in which activated microglia were no longer detectable 2.5 months after elevation of IOP. Wang et al. ⁴³ postulate that the microglia response in their study was an early event after the onset of glaucoma. Our results are in agreement with these data. Microglial activation occurred well before and during significant ganglion cell downregulation and thus corresponds to the period of active ganglion cell death after glaucoma. 

The partial vein cauterization model is reproducible, shows characteristics typical of glaucoma, and allows the early events after elevation of IOP to be studied. The method of in vivo imaging has potential in glaucoma research, especially as far as monitoring the reaction of ganglion cells to various neuroprotective agents is concerned, which is the goal of future studies. A detailed understanding and characterization of the role played by microglia and other nonneuronal cells is vital to understanding the cascade of glaucomatous neuronal degeneration. 

 Supported by Grant Fö. 01KS960410 from the Federal Ministry of Education and Research (RN) and the Interdisciplinary Clinical Research Center (IZKF), Muenster.

 Submitted for publication October 31, 2001; revised April 15, 2002; accepted April 26, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Comparison of IOP between experimental eyes, in which two episcleral veins were cauterized, and the contralateral control eyes. IOP was measured in animals under light ether anesthesia. IOP in cauterized eyes was 1.6 times higher than in the corresponding control eyes. Each data point and error bar represent the mean ± SD IOP in seven animals at each time point of measurement.

Figure 4.

 

RGC loss 3, 5, and 10 weeks after elevation of IOP. The number of RGCs was compared with the control, and the difference was statistically significant (P < 0.001). Inset: Diagram of a retinal flatmount spread on a nitrocellulose filter. To quantify RGCs, the retina was divided into concentric zones with the optic nerve head serving as the reference. Ganglion cells were counted in five fields, each with an area of 0.098 mm². Each bar corresponds to the ocular grid and indicates that 20 fields were counted per retina.

Figure 2.

 

(a) Experimental setup for in vivo imaging. An anesthetized rat was placed on the stage, which could be raised and lowered. The stage replaced the original object table of the microscope. The head of the rat was held in position with a head holder (★) that could be moved along three axes. (b) A nonflexible contact lens of −3.0 to −8.0 D (arrow) was directly fixed to the objective (ml) to extend its focal length and enable fundoscopic examination of the labeled retina. (c) Schematic representation of the modified optics. A glass slide was placed gently on the corneal surface, allowing applanation and enlargement of the retinal field to be visualized.

Figure 3.

 

In vivo imaging. (a) Retrogradely labeled RGCs in a retina immediately after elevation of IOP. (b) The same area 20 hours after induction of glaucoma. Arrowheads: cells that had degenerated 20 hours after elevation of IOP. Magnification, ×25.

Figure 5.

 

(a) Example of a microglial cell (arrowhead) in close contact with a ganglion cell. Microglia appear ramified with dye distributed within the cytoplasm and branches. The cell nuclei remain unlabeled. (b) Retinal flatmount showing labeled RGCs and microglia (arrowheads) 3 days after elevation of IOP. Magnification: ×40.

Figure 6.

 

Neurofilament-positive RGCs (red) and Iba1-positive, activated microglia (green) in the rat retina, 1 week after induction of glaucoma. GCL, ganglion cell layer; IPL, inner plexiform layer. Magnification, ×40.

The authors thank Marliese Wagner and Susanne von der Heide for help with the photograph documentation.