Mammalian eyes mediate both image-forming and non-image-forming visual functions. The image-forming visual pathway is composed of photoreceptors (rods and cones), relay neurons (retinal ganglion cells projecting to the superior colliculus, scRGCs), and the neurons in visual cortex. The structure and function of this pathway have been studied extensively. Non-image-forming vision provides a measure of the ambient light for the purposes of synchronizing the biological clock with the surrounding light-dark cycle, controlling the pupil size, and other functions. However, until recently, we knew very little about the non-image-forming visual pathway, especially about photoreceptors.1 Recently, a subset of melanopsin-containing retinal ganglion cells (mcRGCs), projecting to the suprachiasmatic nucleus, has been identified and determined to be photosensitive. It has been suggested that these cells act as the photoreceptors in the non-image-forming visual pathway, in which melanopsin is considered to be the phototransducing photopigment.2–4
Glaucoma is a group of diseases which can cause degenerative changes and loss of retinal ganglion cells (RGCs) and can eventually lead to visual field defects and vision loss. Elevated intraocular pressure (IOP) is a primary risk factor for the initiation and progression of glaucomatous neuropathy.5–7
In the past, most studies on glaucoma have focused on the scRGCs, and have found that elevated IOP can cause scRGC loss. In clinical studies, glaucoma patients can have afferent papillary defects during the early stages of the disease,8,9 and a high prevalence of sleep disorders during later stages.10 These clues imply that there is some damage to the non-image-forming visual pathway in these glaucoma patients. We therefore hypothesize that mcRGCs, as well as other RGCs, can be damaged by elevated IOP in glaucoma patients.
To examine this hypothesis, we investigated the pathologic changes in mcRGCs in whole mount retinas from a Wistar rat glaucoma model. A recently published paper showed that chronic IOP elevation did not damage mcRGCs in Sprague-Dawley (SD) rats.11 However, our results showed a significant reduction in the number of mcRGCs after 12 weeks of IOP elevation. These results indicate that glaucomatous neural degeneration involves the non-image-forming visual pathway.
All experimental and animal care procedures complied with the the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Care and Use Committee at the China Academy of Sciences and the Capital Medical University.
Induction of elevated IOP
Twelve male adult Wistar rats (200–220 g) were reared in a temperature-controlled room with a 12-hour light-dark cycle in the Animal Laboratory of the Institute of Biophysics, Chinese Academy of Sciences. After the animals were anesthetized with an intramuscular injection of 0.5–0.6 ml of 20% chloral hydrate (500 mg/kg), surgery was performed on the right eye, essentially as described by Shareef et al.12,13 Two dorsal episcleral veins (vortex veins) located near the superior rectus muscle and one temporal episcleral vein (vortex vein) near the lateral rectus muscle were severed with a standard disposable ophthalmic cautery. After surgical isolation, the veins were lifted up and away from adjacent tissues with an iris repository before cauterization. Care was taken to avoid thermal damage and damage to the surrounding tissues with the surgical instruments. The contralateral control eye was sham-operated by similarly isolating the veins, but not cauterizing them. The eyes were flushed with saline, and treated with erythromycin ointment.
Measurement of IOP
IOP was measured in both eyes using the TonoLab rebound tonometer for rodents (Colonial Medical Supply, USA), according to the manufacturer's recommended procedures. Measurements were performed on each animal 1 hour after surgery and then at 1- or 2- week intervals, under anesthesia with a mixture of ketamine and diazepam. Before measurement, one drop of propacaine hydrochloride 0.5% (Alcaine; Alcon Ltd., Fort Worth, TX, USA) was applied to the eye to desensitize the cornea. Five values were obtained and averaged for each measurement.14 All IOP measurements, except for the measurement after surgery, were taken at approximately 10 am to avoid any effects due to diurnal variation.
The skin over the cranium was incised and the scalp exposed. Holes approximately 2 mm in diameter were drilled in the skull on both sides of the midline, 6.2 mm posterior to the bregma and 1.5 mm lateral to the midline, using a dentist's drill (Dremel, Racine, WI, USA). These positions correspond to the superior colliculi, as determined from a stereotaxic rat brain atlas. RGCs were labeled retrogradely by injection of 1.5 μl fluorescent tracer Fluorogold (Fluorochrome, Denver, CO, USA) 3% in saline, into the superior colliculi at two locations on each side, using a 10 μl Hamilton syringe (Reno, NV, USA) with a 28-gauge needle. Injections were performed by placing the rat's head in a stereotaxic apparatus, using coordinates determined from the rat brain atlas, while the animal was under chloral hydrate anesthesia. The overlying skin was sutured and antibiotic ointment was applied externally.15,16 Rats were sacrificed with an overdose of pentobarbital sodium (150 mg/kg) 5 days after the Fluorogold injections, after which they were perfused transcardially with 37°C 0.9% saline, and the eyes were quickly enucleated, after the superior pole of the limbus was marked.
Retinal processing and immunohistochemistry of mcRGCs
A slit was cut in the sclera close to the cornea. The eyes were then submerged in 0.01 mol/L phosphate-buffered saline (PBS). The front part (cornea, lens, and vitreous) of the eye was cut away and the retina was carefully isolated from the pigment epithelium. The retinas were fixed in fresh 4% paraformaldehyde in PBS for 30 minutes and then washed three times in PBS for 5 minutes each. The free-floating retinas were first blocked with 10% inactivated fetal bovine serum albumin for 1 hour at room temperature and were then incubated with a primary melanopsin antibody (polyclonal rabbit anti-melanopsin; Affinity Bioreagents, Golden, CO) at 1:500 dilution in PBS/0.3% Triton X-100/10% bovine serum albumin for 48 hours at 4°C. After three washes in PBS of 15 minutes each, the fluorescence-conjugated secondary antibody (Alexa Fluor 488 goat antibody to rabbit immunoglobulin G; Molecular Probes, Eugene, OR, USA) was applied to the sample as previously described, except that incubation was for 2 hours at 37°C. The free-floating retinas were washed again as described above, flat mounted onto glass slides and coverslips were applied using Vect Mount Permanent Mounting Medium. (Vector Laboratories, Burlingame, CA, USA).2
Counting of RGCs
mcRGCs were counted at 200× magnification using a fluorescence microscope (LEICA DM 400B, Germany). Cells were identified using individual and composite filter sets appropriate for Alexa Fluor 488. Each retina was divided into superior, inferior, nasal and temple quadrants and three visual fields were taken along the median line of each quadrant from the optic disc to the peripheral border of the retina at 1 mm intervals and were counted in a double-blind manner. The area of one visual field was 1.512 mm2, and the mcRGC density was calculated. For the scRGCs, 12 samples (3 samples per quadrant) were selected from the peripheral retina at 3 mm distance from the center of the optic disc. The sample retinas were scanned with a confocal microscope (LEICA TCS SP2) at 400× magnification, and the scRGC density of each sample was calculated by image analysis software (LEICA QWin). The mcRGC and scRGC densities in the complete retina were determined by averaging the data from each sample. The mcRGC survival rate was defined as (mcRGC density of experimental eye)/(mcRGC density of control eye) ×100%. The scRGC survival rate was defined as (scRGC density of experimental eye)/(scRGC density of control eye) ×100%.
All data were expressed by mean ± standard deviation (SD). Paired-sample t test was used to compare the IOP before and after surgery, and the differences in mean mcRGC density and mean scRGC density between the experimental eye and the control eye. P<0.05 was considered statistically significant.
Cauterization of three episcleral veins induced stable IOP elevation for 12 weeks in some rats. Of the 12 rats, seven rats received second surgery 1 week after the first surgery, due to a decline in the IOP. After the second operation, the IOPs of some rats were increased and maintained at a high level. Five rats had stable IOP elevation for 12 weeks and these were selected as the experimental rats. Pre-operational IOPs were about 10 mmHg and 1 hour after operation, the IOP of experimental eyes was dramatically elevated to approximately 33 mmHg, due to blockage of the aqueous outflow. The IOP remained at about 30 mmHg for the first 3 weeks and then decreased slowly. After surgery, the IOP at each point was significantly elevated compared with the IOP before surgery (P=0.001, Figure 1).
mcRGC in rat glaucoma model
Melanopsin is the photopigment of the intrinsically photosensitive RGCs projecting to the suprachiasmatic nucleus. mcRGCs can be immunohistochemically labeled using antibodies against melanopsin and the cell processes can be shown by examining stacked confocal images. In agreement with previous reports, mcRGCs were found to be present throughout the retina. The density of mcRGCs in normal Wistar rats was higher in the peripheral retina than in the retina near the optic disc. The density was also higher in the superior and temple quadrants than that in the inferior and nasal quadrants (Figure 2). The density of mcRGCs was significantly (t=-4.393, P=0.012) reduced in rats with chronic ocular hypertension ((20.62±1.52)/mm2) when compared with control rats ((26.20±2.10)/mm2) (Figure 3). The survival rate of mcRGCs was 78.69%. The morphology of the soma and dendrites showed no obvious changes in the IOP-elevated retina, compared with the control retina.
scRGC in rat glaucoma model
In the present study, the mean density of scRGCs in control retinas was (2317.41±29.96)/mm2, while the density in experimental retinas was (1815.82±24.25)/mm2. The scRGC survival rate was 78.35%. There was a significant difference in the density between the experimental eyes and control eyes (t=-10.340, P=0.000) (Figure 4).
We have demonstrated, using anti-melanopsin antibody immunohistochemical staining and retrograde labeling, that chronic IOP elevation induced by episcleral vein cauterization can cause significant loss of both mcRGCs and scRGCs in Wistar rats. The extent of cell loss of the two types of RGCs is similar.
Glaucoma is a slow, progressive neurodegenerative disease that results in the loss of RGCs, the neurons that project to the brain via the optic nerve. Previous studies demonstrated different degrees of RGC loss due to elevated IOP, depending on the ocular hypertension model and animal species used. In the present study, 78.35% of the scRGCs survived after 12 weeks of chronic hypertension induced by cauterization of three episcleral veins in rats. This result is consistent with previous reports.17–19
RGCs in the rat retina can be classified into several types, according to their soma size and dendritic field.20 Morphologically, mcRGCs are among the largest ganglion cells in the rat retina. Most mcRGCs reside in the ganglion cell layer, but a few are displaced to the inner nuclear layer; the average soma size is about 20–30 μm, and the dendritic profiles are large, spanning about 500 μm.2,3 Unlike scRGCs, mcRGCs project to the suprachiasmatic nucleus, and serve as the photosensitive cells of the non-image-forming visual pathway. Our results showed that the density of this special type of RGC was significantly reduced after 12 weeks of chronic ocular hypertension, and the severity of damage to the mcRGCs and scRGCs was similar, suggesting that not only the image-forming visual system, but also the non-image-forming system, was affected by chronic IOP elevation. The results also suggest that serious attention should be paid to the function of the non-image-forming system in glaucoma patients in the future.
Our findings were supported by a recent study on RGCs in DBA/2J mice,21 a mouse model of inherited pigment dispersion glaucoma with an increase in IOP and RGC degeneration. This study showed that the density of mcRGCs was significantly reduced, and no particular type of ganglion cell was especially vulnerable or resistant to degeneration. In another study, one month after optic nerve axotomy in C57/BL6 mice, the densities of both the melanopsin-positive ganglion cells and non-melanopsin ganglion cells were significantly reduced, but the melanopsin-positive ganglion cells survived significantly better than the non-melanopsin ganglion cells.22
mcRGCs are functionally different from other RGCs, but their anatomical features, such as distribution, soma size and dendritic morphology, are similar to some other types of RGCs. It is therefore plausible that this type of RGC may experience similar pathological changes to other RGCs after the induction of ocular hypertension.
Contrary to our results, Li et al11 showed that mcRGCs were not significantly affected after 12 weeks of IOP elevation in SD rats, even though scRGCs showed a great reduction. This apparent difference may be due to the different animal models used in the two studies. In rats, the limbal venous plexus is connected to Schlemmn's canal through multiple collecting channels and drains into radial aqueous-containing veins and then into the vortex veins.23 In our study, as well as those of Grozdanic et al,24 the episcleral veins we cauterized may have been aqueous-containing veins or vortex veins, thus, some additional biological effects unrelated to ocular hypertension, such as ocular ischemia, congestion, and abnormal production of cytokines may have occurred. This may be different from the model Li used,11 in which limbal veins and episcleral veins were photocoagulated. The species difference may also be a possible explanation for the different conclusions. In Wistar rats, mcRGCs are not evenly distributed in the four quadrants. We do not know if this is also true in SD rats. Because the mcRGCs are sparsely distributed in the central retina and in the inferior and nasal quadrants, only one or two cells, or possibly no cells at all, may be seen in a sample field of 400 μm × 400 μm. In order to minimize the selective bias, we counted all the cells in a 200× visual field (the area is 1.512 mm2), using a fluorescence microscope. Sampling the retina with a small sample field using a confocal microscope may have introduced some bias.
In our study, the soma size and dendritic morphology showed no obvious changes after 12 weeks of chronic ocular hypertension, which is consistent with the report of Li et al.11 We are unable to rule out the possibility that this may be explained by the unique properties of mcRGCs.
Our findings demonstrate that both mcRGCs and scRGCs were affected by chronic IOP elevation, suggesting that the non-image-forming visual pathway could be damaged in glaucoma. This study implies that attention should be paid not only to image-forming visual functions in glaucoma, but also to non-image-forming visual functions, such as control of circadian rhythm. Further studies are needed to investigate the mechanism whereby glaucoma can damage the non-image-forming visual functions.
The authors thank Dr. HE Shi-gang and the staff at the Institute of Biophysics, Chinese Academy of Science for their technical assistance in the study.
1. Moore RY, Speh JC, Card JP. The retinalhypothalamic tract originates from a distinct subset of RGCs. J Comp Neurol 1995; 352: 351–366.
2. Hatter S, Liao HW, Takao M, Berson DM, Yau KY. Melanopsin-containing retinal ganglion cells: architecture, projections, and intrinsic photosensitivity. Science 2002; 295: 1065–1070.
3. Berson DM, Dunn FA, Takao M. Phototransduction by retinal ganglion cells that set the circadian clock. Science 2002; 295: 1070–1073.
4. Lucas RJ, Hatter S, Takao M, Berson DM, Foster RG, Yau KY. Diminished pupillary light reflex at high irradiances in melanopsin-knockout mice. Science 2003; 299: 245–247.
5. Quigley HA, Green WR. The histology of human glaucoma cupping and optic nerve damage: clinicopathologic correlation in 21 eyes. Ophthalmology 1979; 86: 1803–1830.
6. Quigley HA, Dunkelberger GR, Green R. Chronic human glaucoma causing selectively greater loss of large optic nerve fibers. Ophthalmol 1988; 95: 357–365.
7. Quigley HA, Nickells RW, Kerrigan LA, Pease ME, Thibault DJ, Zack DJ. Retinal ganglion cell death in experimental glaucoma and after axotomy occurs by apoptosis. Invest Ophthalmol Vis Sci 1995; 36: 774–786.
8. Kaback MB, Burde RM, Becker B. Relative afferent pupillary defect in glaucoma. Am J Ophthalmol 1976; 81: 462–468.
9. Kohn AN, Moss AP, Podos SM. Relative afferent pupillary defects in glaucoma without characteristic field loss. Arch Ophthalmol 1979; 97: 294–296.
10. Onen SH, Mouriaux F, Berramdane L, Dascotte JC, Kulik JF, Rouland JF. High prevalence of sleep-disordered breathing in patients with primary open-angle glaucoma. Acta Ophthalmol Scan 2000; 78: 638–641.
11. Li RS, Chen BY, Tay DK, Chan HH, Pu ML, So KF. Melanopsin-expressing retinal ganglion cells are more injury-resistant in a chronic ocular hypertension model. Invest Ophthalmol Vis Sci 2006; 47: 2951–2958.
12. Shareef SR, Garcia-Valenzuela E, Salierno A, Walsh J, Sharma SC. Chronic ocular hypertension following episcleral venous occlusion in rats. Exp Eye Res 1995; 61: 379–382.
13. Mittag TW, Danias J, Pohorenec G, Yuan HM, Burakgazi E, Chalmers-Redman R, et al. Retinal damage after 3 to 4 months of elevated intraocular pressure
in a rat glaucoma model. Invest Ophthalmol Vis Sci 2000; 41: 3451–3459.
14. Wang WH, Millar JC, Pang IH, Wax MB, Clark AF. Noninvasive measurement of rodent intraocular pressure
with a rebound tonometer. Invest Ophthalmol Vis Sci 2005; 46: 4617–4621.
15. Wang N, Zeng M, Ruan Y, Wu H, Chen J, Fan Z, et al. Protection of retinal ganglion cells against glaucomatous neuropathy by neurotrophin-producing, genetically modified neural progenitor cells in a rat model. Chin Med J 2002; 115: 1394–1400.
16. Sellés-Navarro I, Villegas-Pérez MP, Salvador-Silva M, Ruiz-Gómez JM, Vidal-Sanz M. Retinal ganglion cell death after different transient periods of pressure-induced ischemia and survival intervals. A quantitative in vivo
study. Invest Ophthalmol Vis Sci 1996; 37: 2002–2014.
17. Ahmed FA, Chaudhary P, Sharma SC. Effects of increased intraocular pressure
on rat retinal ganglion cells. Int J Dev Neurosci 2001; 19: 209–218.
18. Laquis S, Chaudhary P, Sharma SC. The patterns of retinal ganglion cell death in hypertensive eyes. Brain Res 1998; 784: 100–104.
19. Sawada A, Neufeld AH. Confirmation of the rat model of chronic, moderately elevated intraocular pressure
. Exp Eye Res 1999; 69: 525–531.
20. Sun W, Li N, He S. Large-scale morophological survey of rat retinal ganglion cells. Visual Neurosci 2002; 19: 483–493.
21. Jakobs TC, Libby RT, Ben Y, John SW, Masland RH. Retinal ganglion cell degeneration is topological but not cell type specific in DBA/2J mice. J Cell Biol 2005; 171: 313–325.
22. Robinson GA, Madison RD. Axotomized mouse retinal ganglion cells containing melanopsin show enhanced survival, but not enhanced axon regrowth into a peripheral nerve graft. Vision Res 2004; 44: 2667–2674.
23. Morrison JC, Fraunfelder FW, Milne ST, Moore CG. Limbal microvasculature of the rat eye. Invest Ophthalmol Vis Sci 1995; 36: 751–756.
24. Grozdanic SD, Betts DM, Sakaguchi DS, Kwon YH, Kardon RH, Sonea IM. Temporary elevation of the intraocular pressure
by cauterization of vortex and episcleral veins in rats causes functional deficits in the retina and optic nerve. Exp Eye Res 2003; 77: 27–33.