Retinal Ganglion Cell Death After Different ... - Semantic Scholar

2 downloads 0 Views 2MB Size Report
Retinal ganglion cell death also can be induced by retinal ischemia. Recent studies of cell numbers and thickness of the retinal layers have documented.
Retinal Ganglion Cell Death After Different Transient Periods of Pressure-Induced Ischemia and Survival Intervals A Quantitative In Vivo Study Inmaculada Selles-Navarro* Maria P. Villegas-Perez,* Mercedes Salvador-Silva* Jose M. Ruiz-Gomez^ and Manuel Vidal-Sanz*%

Purpose. To quantitate in vivo retinal ganglion cell (RGC) survival after transient periods of pressure-induced ischemia of the rat retina and after different survival intervals. Methods. In adult rats, RGCs were labeled with fluorogold applied to their main targets in the brain. Seven days later, in several groups of rats, the left retinas were subjected to transient periods of ischemia of 30, 45, 60, 75, 90, 105, or 120 minutes, respectively, by increasing the intraocular pressure (IOP) of the left eye above systolic values. Five, 7, 14, and 30 days later, the rats were killed, and their retinas were prepared as wholemounts for examination under fluorescence microscopy to estimate RGC survival. Results. The authors found that periods of ischemia of 30 and 45 minutes do not induce RGC death; longer periods of transient ischemia induce the death of a proportion of RGCs, and the proportion increases with the duration of the ischemia; RGC death, which can be observed as early as 5 days after ischemia, continues during the 30-day study period; and periods of ischemia that last 90 minutes or more cause the death of approximately 95% of the RGC population 30 days later. Conclusions. Increases of the IOP above systolic levels for periods of 60 minutes or more result in RGC loss in the rat retina. Both the duration of the initial transient period of ischemia and the duration of the survival period influence the proportion of RGC death. Invest Ophthalmol Vis Sci. 1996;37:2002-2014.

JL revious studies have documented that in the adult rat, optic nerve section causes retinal ganglion cell (RGC) death proportionally related to the distance of the site of injury to the eye and the time after axotomy.1'2 Retinal ganglion cell death does not occur immediately after optic nerve section; it starts 5 days later3'4 and proceeds at different speeds, depending on the proximity of the axonal injury to the eye.2 Retinal ganglion cell death after axotomy shows apoptotic From the Labaratorio de Oftalmologin Experimental, Facultnd de. Medirina, and the fDepartamento de Matemdtica Aplkada y Estadislica, Facultad de. Malematicas, Universidad de Murcia; and the %lnstituto Cajal, Consejo Superior Investigaciones Cientificas, Madrid, Spain. Supported by research grants from the Regional Government of Murcia (P1B 93/98, 94/15) and the Spanish Ministry of Health (FIS 95/1720). Submitted for publication January 31, 1996; revised May 13, 1996; accepted May 14, 1996. Proprietary interest category: N. Reprint requests: Manuel Vidal-Sanz, Uiboratorio de Oftalmologia Experimental, Facultad de Medicina, 30100 Espinardo, Murcia, Spain.

2002

characteristics3'5 b and may be prevented with the administration of the neurotrophins, brain-derived neurotrophic factor47 or NT-4.4 Retinal ganglion cell death also can be induced by retinal ischemia. Recent studies of cell numbers and thickness of the retinal layers have documented a decrease in RGC numbers after interruption of the retinal blood flow in rats8"12 and other mammals.13"15 Retinal ganglion cell degeneration after ischemia may have a different morphology,9>lh and it can be prevented in part by the use of neurotrophic factors17 and flunarizine.12 However, until now, all the investigations of RGC survival after ischemia lacked a method to distinguish RGCs from other types of cells in the RGC layer, such as displaced amacrine cells,18 or they distinguished them on the basis of their morphology. Because the displaced amacrine cells are abundant in the rat,18 and because the neuronal phe-

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10 Copyright © Association for Research in Vision and Ophthalmology

Retinal Ganglion Cell Survival After Ischemia

notype may change after injury, cell counts in earlier studies may have been influenced by the presence of these neurons. In addition, the pattern and time course of RGC loss that follows pressure-induced ischemia of the retina remains unknown. In the current study, the fate of RGCs after transient periods of ischemia is examined. To accomplish this, we have prelabeled RGCs using previously developed techniques,2'119 induced transient periods of ischemia of different duration, and estimated RGC survival at different survival intervals. A detailed quantification of the amount of RGC death and its time course after retinal ischemia may allow future comparisons of the characteristics of RGC death induced by different injuries, such as axotomy and ischemia, and thus may facilitate an understanding of glaucoma. It also may facilitate quantitative studies to investigate the neuroprotective effects of several molecules that may influence RGC survival after ischemia.l7 Brief accounts of this work have been presented in abstract form.20'21

2003

exposed, and, after the pia overlying the superior colliculi and dorsolateral geniculate nuclei was removed, a small pledget of gelatine sponge (Spongostan Film, Ferronsan, Denmark) soaked in a 0.9 NaCl solution containing 3% fluorogold (FG) and 10% dimethyl sulfoxide was laid over the entire surface of the superior colliculi and dorsal lateral geniculate nuclei to label RGCs by retrograde axonal transport. Fluorogold, a soluble fluorescent tracer whose characteristics as a retrograde tracer have been described,2'1 has an intense fluorescence that persists within RGC soma for periods of at least 3 weeks without apparent fading or leakage'1 (see Fluorogold Labeling in Control Retinas) . Previous observations in this laboratory'1 indicate that 7 days after FG application in the brain, the densities of RGCs retrogradely labeled with FG are similar to those obtained with other well-characterized, retrogradely transported fluorescent and nonfluorescent tracers219 similarly applied to the main retino-recipient target regions in the brain.

Retrograde Labeling of Retinal Ganglion Cells

Transient Periods of Pressure-Induced Ischemia of the Retina Seven days after FG application, transient periods of retinal blood flow interruption were induced in the left eye by increasing the intraocular pressure (IOP) above systolic arterial levels. For all the rats, the right eye, which had not been manipulated, served as a normal counterpart or control. To increase the IOP, two 6-0 silk sutures were placed on the bulbar conjunctiva of the left eye just below and above the corneoscleral limbus and were used to pull tangentially in opposite directions until the blood flow to the retina was interrupted completely. Sutures were tied to a metal frame designed for these experiments to maintain the IOP above systolic arterial levels throughout the transient ischemic period. Interruption of the blood flow to the retina was assessed constantly by examining the fundus of the eye through the operating microscope, and the sutures were retightened when needed to maintain interruption of the intraretinal blood flow. To permit microscopic observation of the eye fundus and, thus, of retinal blood flow, the pupil was dilated with a drop of 1% tropicamide (Cusi Laboratories, Barcelona, Spain), and the corneal surface was covered with a drop of 2% hydroxipropilmethylcellulose (Gonioftal 4000; Cusi Laboratories) and a coverslip. During the periods of retinal ischemia, interruption of the blood flow to the iris also was observed.

To identify RGCs, we applied the fluorescent tracer Fluoro-gold (Fluorochrome, Engelwood, CO) to the main RGC targets in the brain, using previously described methods1"2'19 that involve the application of the tracer to the superior colliculi and dorsolateral geniculate nuclei, the main targets of rat RGCs.23'23 In brief, the rats were anesthetized, each midbrain was

The animals were divided into several groups according to duration of the transient period of complete retinal ischemia (30, 45, 60, 75, 90, 105, or 120 minutes) and length of the surviving interval after ischemia (5, 7, 14, or 30 days). At the end of the transient ischemic period, the conjunctival sutures were released slowly, and this was

METHODS Animals Adult female Sprague-Dawley rats (each weighing 180 to 200 g) were obtained from the breeding colony of the University of Murcia. The animals were fed ad libitum and were maintained in cages in temperaturecontrolled rooms with a 12-hour light-12-hour dark cycle (light period from 8 AM to 8 PM). The light intensity in the cages ranged from 8 to 24 lux. Experiments were carried out in accordance with the European Union guidelines and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. All the experimental manipulations were carried under general anesthesia (7% chloral hydrate dissolved in saline, administered intraperitoneally at a dose of 0.42 mg/g). Maintenance of anesthesia was obtained by repeated injections of smaller amounts of the same anesthetic. The animals were killed with an overdose of the same anesthetic. During recovery from anesthesia, the rats were placed in their cages, and a steroid-antibiotic ointment (Fludronef, Iquinosa, Madrid) was applied on the cornea prevent corneal desiccation.

2004

Investigative Ophthalmology 8c Visual Science, September 1996, Vol. 37, No. 10

followed by the complete restoration of the blood flow to the retina and the iris. Animals in which restoration of the retinal blood flow was associated with retinal or vitreous hemorrhages were eliminated from the study. Regular findings at the termination of the ischemic period were mild edema of the conjunctiva and cornea. Because body temperature may influence ischemia-induced RGC death,25 we monitored the body temperatures of eight rats anesthetized and subjected to transient retinal ischemia with a rectal probe. Mean (±SEM) rectal temperatures in these eight animals was 36.5°C ± 0.06°C; 35.4°C ± 0.18°C; 34.6°C ± 0.2°C; 34°C ± 0.2°C; 33.9°C ± 0.3°C; 33.8°C ± 0.3°C, and 34°C ± 0.4°C, respectively, when measured before and 5, 30, 60, 90, 105, and 120 minutes after the instauration of retinal ischemia. Room temperature was maintained at approximately 23°C. Thus, body temperature fluctuations in these animals ranged from 35.4°C right after the initiation of retinal ischemia to 33.8°C two hours later. Because the temperature within the eye was not monitored, we ignored whether these small changes in body temperature could have affected RGC survival in the current experiments. Effects of the Increase of Intraocular Pressure in Retinal Blood Flow In 14 additional rats, we examined the effect of conjunctival traction on IOP and retinal blood flow. For this purpose, a cannula was introduced into the femoral artery, and a needle was introduced into the vitreous chamber of the eye through the pars plana. Both were connected to pressure transducers to monitor simultaneously arterial and intraocular pressures. Two sutures were placed in the conjunctiva and were pulled tangentially as described above, and this was observed to cause an increase in the IOP. The eye fundus also was examined, and the central artery of the retina was observed to pulse when the IOP was increased above diastolic but below systolic arterial levels. Interruption of the retinal blood supply was achieved when the IOP was raised above arterial systolic levels (data not shown). Two additional rats were used to corroborate that the increase in the IOP above systolic level interrupted all blood supply to the retina, that is, to all retinal and choroidal vessels. For this purpose, in these animals, the IOP was raised and maintained above systolic level while they were perfused first with saline and later with a 10:1 china ink solution in 10% paraformaldehyde. In one animal, bodi retinas were dissected as wholemounts, whereas in the other animal, 30-jum thick paraffin-embedded radial sections from both eyecups were obtained in the microtome. Wholemounts and sections were examined microscopically for the presence of vessels filled with china ink. Although china

a

I. Micrographs (X457) from representative regions of 10 /zm thick paraffin-embedded radial sections of the right (a) and left (b) retinas of a rat in which the intraocular pressure of the left eye was raised and maintained above systolic levels while perfused with saline and a china ink solution in 10% paraformaldehyde. Although china ink fills retinal and choroidal vessels {arrows) of the control right eye {top, a), it was absent from the retina and choroid of the left ischemic eye {bottom, b). Vitreal surface at the top of the micrographs. Sections were stained with hematoxylin and eosin. FIGURE

ink was observed to fill the retinal vessels in the wholemount preparation and in the retinal and choroidal vessels in the cross-sections of the right control eyes, it was absent from the retina and choroid of left ischemic eyes (Fig. 1). These results confirm that in our experiments, the elevation of the IOP above systolic levels interrupts the retinal and choroidal circulations, thus inducing complete ischemia of the entire retina. Tissue Processing Five, seven, 14, or 30 days after the induction of retinal ischemia, the rats were anesthetized, and each eye fundus was microscopically examined in a search for abnormalities. Each rat with a normal eye fundus was killed, perfused through the ascending aorta first with saline and then with 4% paraformaldehyde in 0.1 M phosphate buffer, and both eyes were enucleated. Both retinas were dissected, flattened by four radial

Retinal Ganglion Cell Survival After Ischemia

2005

temporal, inferotemporal, superonasal, and inferonasal) at 0.875, 1.925, and 2.975 mm, respectively, from the optic disc. The number of labeled cells in the 12 photographs was divided by the area of the region and pooled to calculate mean densities of labeled neurons/mm2 for each retina. Cell counts were conducted by the same investigator in a masked fashion; the identity of the retinas that led to the micrographs was unknown until cell counts from different groups were complete. For statistical analysis, data from experimental and control groups after different periods of ischemia and at different time intervals were compared using the Student's /-test or the nonparametric KruskalWallis analysis of variance, followed by the MannWhitney test. RESULTS Using quantitative anatomic techniques, the current studies were conducted to investigate the effects of transient periods of blood flow deprivation and different survival intervals after ischemia on RGC death. To address this, RGCs were labeled retrogradely with FG from their main targets in the brain. Densities of FGFIGURE 2. Fluorescent micrographs (X228) from representa-

2500

tive regions in similar quadrants and distance from the optic disc in flatmounted retinas showing fluorogold-labeled retinal neurons on the right (a) and left (b) retina of one rat 14 days after 90 minutes of transient ischemia in the left eye. Retinal ganglion cells were labeled by applying fluorogold to the superior colliculus and dorsal lateral geniculate nuclei 7 days before transient ischemia of the left eye.

2000 -

cuts (the deepest in the superior pole and the others in the inferior, temporal, and nasal poles), fixed for an additional 30 minutes, and mounted vitreal side up on gelatine-coated slides in a solution of 50% glycerol in 0.1 M sodium carbonate buffer (pH 9) containing 0.04% p-phenylenediamine.*' The retinas were examined through a fluorescence microscope (Axiophot; Zeiss, Oberkochen, Germany) equipped with an ultraviolet filter that permits the visualization of FG fluorescence (excitation filter BP 365/11, suppression filter LP397). Estimation of Retinal Ganglion Cell Densities The mean densities of FG-labeled RGCs in the ganglion cell layer of every retina were estimated following already described methods.I>!M Briefly, labeled RGCs were counted from printed fluorescent micrographs of 12 standard areas of each retina. Each rectangular area measured 0.36 X 0.24 mm2, and there were three areas in each retinal quadrant (supero-

1500 -

1000 -

500 -

30'

45'

60'

75'

90'

105' 120'

Minutes of ischemia

FIGURE3. Bar histogram representing mean densities (cells/ mm11 ± SEM) of fluorogold-labeled retinal ganglion cells for the experimental (left) retinas 5, 7, and 14 days after induction of different periods of transient ischemia of the left eye. For each period of transient ischemia, * indicates densities that were significandy different from the densities obtained 5 days after ischemia (P < 0.05, Mann-Whitney test), and f indicates densities that were significantly different from the densities obtained 7 days after ischemia (P < 0.05, Mann-Whitney test).

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10

2006

l. Densities of Fluorogold-Labeled Cells 5 Days After Ischemia

TABLE

Minutes of Ischemia

Right Retina (control)

Left Retina (experimental)

30

2311 2211 2658 2360 ± 159

2297 2010 2217 2174 ± 85 92 2502 1844 1971 2105 ± 201 97 1168 1353 1967 1197 1464 2153 1550 ± 169* 67 1489 1229 1499 1179 1230 1325 ± 69 57 1052 883 1234 1351 1889 1281 ± 171 52 521 732 778 652 671 ± 56 27

Mean ± SEM % survival 45 Mean ± SEM % survival 60

Mean ± SEM % survival 75

Mean ± SEM % survival

90

Mean ± SEM % survival 120

Mean ± SEM % survival

2470 1711 1971 2161 ± 230 2244 1945 2517 2531 2174 2405 2302 ± 92 2154 2290 2190 2473 2557 2332 ± 79 2134 2477 2184 2599 2854 2449 ± 133 2610 2433 2361 2339 2435 ± 61

* First transient period of ischemia that appeared statistically different from values in the contralateral control retinas (MannWhitney test, P < 0.005).

labeled RGCs were estimated after different transient periods of pressure-induced ischemia of the retina. We found that densities of FG-labeled RGCs declined for transient periods of ischemia of 60 minutes or more. In addition, this study addresses the effects of the survival interval on RGC loss after transient periods of ischemia. To address this, densities of retrogradely labeled RGCs in retinas subjected to similar periods of ischemia were compared at different survival intervals. We found that the numbers of FG-labeled RGCs also declined progressively as the survival interval progressed. Fluorogold Labeling in Control Retinas In control retinas, the only cells that appeared to be labeled with FG were RGCs. These were identified by

the typical punctate fluorescence present in the somata as well as within the initial segments of primary dendritic processes (Fig. 2A). Although the intensity of the fluorescence of these cells did not seem to decrease with time after FG-application, the densities of FG-labeled RGCs diminished with time. When the densities of FG-labeled RGCs (cells/mm2) in the control (right) retinas of the experimental animals were pooled at each survival period, these were 2335 ± 49 (mean ± SEM; n = 26), 2268 ± 40 (mean ± SEM; n

2. Densities of Fluorogold-Labeled Cells 7 Days After Ischemia

TABLE

Minutes of Ischemia

Right Retina (control)

Left Retina (experimental)

30

2530 2223 2321 2329 2368 ± 65

2035 2184 2215 2506 2235 ± 98 94 1935 1688 2103 2407 2033 ± 151 97 1385 803 1544 2424 1539 ± 335* 63 699 1314 622 1042 919 ± 160 41 598 1041 616 1365 928 910 ± 142 43 1117 1566 472 623 944 ± 249 40 936 421 129 314 450 ± 173 22

Mean ± SEM % survival 45

Mean ± SEM % survival 60

Mean ± SEM % survival

75

Mean ± SEM % survival

1889 1818 2227 2483 2104 ± 155 2443 2209 2485 2608 2436 ± 83 2219 2360 2370 2094 2260 ± 65

90

2306 2126 2034 2608

Mean ± SEM % survival 105

2113 ± 60

Mean ± SEM % survival 120

Mean ± SEM % survival

2588 2190 2381 2260 2354 ± 87 2000 2284 2045 1997 2081 ± 68

* First transient period of ischemia that appeared statistically different from values in the contralateral control retinas (MannWhitney test, P< 0.03).

2007

Retinal Ganglion Cell Survival After Ischemia

= 29), 2221 ± 37 (mean ± SEM; n = 42), and 1859 ± 35 (mean ± SEM; n = 41), respectively, for the groups of animals analyzed at 5, 7, 14, and 30 days after the induction of ischemia in the left eye. These densities were very close in the groups of animals analyzed after 5, 7, or 14 days of survival. However, the densities obtained after 30 days of survival were approximately 10% smaller that those obtained after shorter survival times. There were no statistically significant differences between the densities obtained after survival periods of 5, 7, and 14 days. However, there was a significant difference between the densities of RGCs in the retinas analyzed at 5, 7, or 14 days and those obtained after 30 days of survival (Student's Hest, P < 0.05). Because FG was applied to the main retino-recipient target regions in the brain 7 days before the induction of transient periods of ischemia, we conclude from these data that persistence of FG within the cell somata of the entire RGC population is limited to 21 days after tracer application.

3. Densities of Fluorogold-Labeled Cells 14 Days After Ischemia

TABLE

Minutes of Ischemia

Right Retina (control)

Left Retina (experimental)

30

2136 2351 2257 2236 2703 2517 2281 2119 2325 ±

70

1702 1875 2635 2263 2164 2127 ±

62

1977 1937 1941 2340 2329 2298 2225 2330 2172 ±

2084 2474 2071 2255 2370 1947 2368 2350 2243 :t: 36 96 1575 1843 2377 2203 1785 1956:!: 146 92 1724 532 603 762 565 2168 1265

65

Mean ± S E M % survival

45

Mean ± S E M % survival

60

Fluorogold Labeling in Experimental Ischemic Retinas In the experimental retinas, typical FGfluorescencealso was found in the cytoplasm of the somata and sometimes in the proximal dendrites of RGCs (Fig. 2B). Retinas subjected to 30 or 45 minutes of transient ischemia only showed FG-labeled RGCs. However, in retinas subjected to 60 minutes or more of transient ischemia, there were diminutions in the densities of FG-labeled RGCs (Fig. 3, Tables 1 to 4). In addition, in these retinas, we observed other cells intensely labeled with FG intermingled with the RGCs, which had small somata with fine tortuous processes that extended into the ganglion cell layer. These cells, clearly distinguishable from RGCs, had the morphology of microglia. Retinal microglial cells also appeared fluorescently labeled in other studies when the RGCs were labeled with fluorescent tracers and then died as a result of injury. The neuronal tracers used in these studies were FG4 or other persistent fluorescent tracers of the carbocyanine family.227"29 It has been proposed that retinal microglial cells phagocytose the debris of degenerating RGCs30 and thus become fluorescently labeled.27"29 In the current study, we have not determined the frequency or distribution of these microglial cells.

Mean ± S E M % survival

75

Mean ± S E M % survival

90

Mean ± S E M % survival

105

Mean ± S E M % survival

120

Influence of the Duration of the Transient Period of Ischemia on Retinal Ganglion Cell Survival Because we quantified the densities of FG-labeled RGCs for all control and ischemic retinas, the comparison of densities between the right and left retinas at each survival period allowed us to examine the effects of the duration of the transient period of ischemia on RGC survival.

Mean ± S E M % survival

1835 2285 2215 2402 2188 2093 2169 ± 2187 1540 2523 2223 2430 2181 ± 1913 2468 2333 2191 2226 ± 2124 2296 2131 2468 2522 2293 2305 ±

1339 1119 ± 215*

51 429 356 210 495 509 355 79

391 ±

45

18 393 495 321 236 297 172

348 ±

44

16 173 585 249 170 119

294 ±

98

3 493 47 209 218 1306

205 67

413 ±

188

18

* First transient period of ischemia that appeared statistically different from values in the contralateral control retinas (MannWhitney test, P< 0.02).

2008 TABLE 4.

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10

Densities of Fluorogold-Labeled Cells 30 Days After Ischemia

Retinal Ganglion Cell Density 5 Days After Retinal Ischemia. Five days after the induction of retinal ischemia—the shortest time period analyzed in these Right Retina Left Retina Minutes of studies—the mean densities of FG-labeled RGCs in (control) (experimental) Ischemia the left experimental retinas differed substantially depending on the duration of the ischemic period (Fig. 1484 1623 30 2048 2198 3, Table 1). Furthermore, not only after 5 days of 1945 2085 survival but for all survival periods, the amount of 1716 1799 cell death depended on the duration of the ischemic 1785 1802 period (see below). Five days after ischemia, the mean 1908 1902 densities of FG-labeled RGCs after periods of ischemia 1636 1819 of 30 or 45 minutes were 2174 ± 85 (n = 3) and 2105 1808 ± 61 1869 ± 87 Mean ± S E M ± 201 (n = 3) cells/mm2 (mean ± SEM), respectively. % survival 98 These densities were similar and not significantly dif1969 45 1868 1995 1631 ferent from the densities obtained in contralateral in2167 2209 tact retinas (2360 ± 159 and 2161 ± 230 cells/mm2, 1807 1490 respectively). 2084 1735 The mean density of FG-labeled RGCs obtained 2153 2162 after 60 minutes of ischemia was 1550 ±169 cells/ Mean ± S E M 1842 ± 120 2012 ± 61 mm2 and represents approximately 67% of the cells 93 % survival 1711 844 60 labeled in the contralateral retinas. Individual densi1814 521 ties obtained in these retinas were statistically different 1664 770 from the densities obtained in their contralateral con1730 1312 trol retinas using the Mann-Whitney test (2303 ± 92 1399 1638 2 ; n = 6; P < 0.005). Thus, 60 minutes of cells/mm 2057 736 ischemia cause the death of approximately 33% of the 2283 198 RGC population by 5 days after the insult. Mean ± S E M 826 ± 159* 1842 ± 91 % survival 45 The mean densities (cells/mm2) of FG-labeled 651 75 1872 RGCs in the retinas that suffered longer periods of isch1907 332 emia of 75, 90, and 120 minutes were, respectively: 1325 1644 150 ± 169, n = 5; 1281 ± 171, n = 5; and 671 ± 56, n = 1860 416 4. These densities were statistically different from the 1508 579 densities obtained in their contralateral retinas (P < Mean ± S E M 1758 ± 78 426 ± 89 % survival 24 0.015; Mann-Whitney test). Five days after transient 120 1523 90 ischemic periods of 60 minutes or more, there was a 2003 110 significant decrease in the numbers of FG-labeled RGCs. 59 2098 Retinal Ganglion Cell Densities 7 Days After Retinal 33 1465 Ischemia. The mean densities of FG-labeled RGCs 7 1871 41 days after ischemia were 2235 ± 98 (n = 4), 2033 ± Mean ± S E M 73 ± 18 1792 ± 127 4 % survival 151 (n = 4), 1539 ± 335 (n = 4), 919 ± 160 (n = 4), 1867 37 105 910 ± 142 (n = 5), 944 ± 249 (n = 4), and 450 ± 1598 123 173 (n = 4) cells/mm2 (Fig. 3, Table 2), respectively, 1898 212 for ischemic periods of 30, 45, 60, 75, 90,105, and 120 11 1470 minutes. The comparison of cell densities between 354 2081 experimental and control retinas revealed that there 147 ± 62 Mean ± S E M 1782 ±: 10 8 % survival were statistically significant differences when the isch40 1615 120 emic period was of 60 or more minutes (P < 0.03; 115 1865 Mann-Whitney test). Therefore, for the retinas ana109 1830 lyzed at 7 days, 60 minutes or more of ischemia caused 190 1996 significant RGC loss. In this group of animals, the 61 1914 amount of cell death also increased with the duration 83 2289 of the ischemic periods but was not proportional to 100 ± 21 1918 ±: 91 Mean ± S E M 5 % survival it. Thus, although periods of ischemia of 60 minutes produced the loss of approximately 37% of the RGCs * First transient period of ischemia that appeared statistically and 75 minutes of ischemia produced the loss of 59% different from values in the contralateral control retinas (MannWhitney test, P < 0.005). of the RGCs, periods of ischemia of 90, 105, and 120

Retinal Ganglion Cell Survival After Ischemia

2009

minutes produced similar proportions of cell death of approximately 57%, 60%, and 78%, respectively. Retinal Ganglion Cell Density 14 Days After Retinal

Ischemia. Fourteen days after transient periods of pressure-induced ischemia, the mean densities of FG-labeled neurons in the left experimental retinas also differed substantially, depending on the duration of the ischemic period (Fig. 3, Table 3). As was observed in the groups of animals analyzed 5 and 7 days after ischemia, the densities obtained 14 days after periods of ischemia of 30 or 45 minutes were similar to the densities obtained in the contralateral intact retinas; there was no RGC loss. In addition, as observed in the group of animals analyzed 7 days after the insult, periods of ischemia of 60 minutes caused significant cell death (P < 0.0005; Mann-Whitney test). Thus, it appears that a transient period of ischemia lasting between 45 and 60 minutes is the threshold for RGC death. Finally, the amount of cell death increased with the time of ischemia, and the effects of ischemia lasting 75 or more minutes were similar. Transient ischemic periods of 75, 90, 105, and 120 minutes caused 82% to 87% of RGC deaths. Retinal Ganglion Cell Densities 30 Days After Retinal

Ischemia. The mean densities of FG-labeled neurons in the experimental left retinas of this group of animals differed substantially, depending on the duration of the ischemic period (Fig. 3, Table 4). For transient periods of ischemia of 30 or 45 minutes duration, the mean densities of FG-labeled RGCs were 1808 ± 87 (n = 7) and 1842 ± 120 (n = 6) cells/mm2, respectively. These values were similar to the densities observed in the contralateral control retinas, indicating that these periods of transient ischemia do not induce RGC death (Table 1). After transient periods of 60 minutes of ischemia, however, the mean densities of FG-labeled RGCs had decreased to 826 ± 159 (n = 7) cells/mm2, corresponding to 45% of the contralateral RGC population. The comparison of cell densities between experimental and control retinas revealed that there were statistically significant differences when the ischemic period was of 60 or more minutes (P< 0.004, Mann-Whitney test). There were greater reductions in the densities of the retinas that suffered 75 and 90 minutes of ischemia, with mean densities of surviving RGCs of 426 ± 89 (n = 5) and 73 ± 18 (n = 5) cells/ mm2, respectively, corresponding to 24% and 4% of the contralateral retinas. Finally, the retinas that suffered transient periods of ischemia of 105 or 120 minutes had mean densities of FG-labeled RGCs of 147 ± 62 (n = 5) and 100 ± 21 (n = 6) cells/mm2, representing 8% and 5%, respectively, of their control contralateral retinas and indicating cell death of 92% and 95% of the original RGC population. These densities were not statistically different from those obtained in the group that suffered 90 minutes of ischemia.

45'

60'

75'

90'

105'

120'

Minutes of ischemia

FIGURE 4. Percentages representing proportions of fluorogold-labeled retinal ganglion cells in the left experimental retinas compared with the right retinas of rats analyzed after different periods of transient ischemia and different survival intervals. Cell densities decreased progressively with time for 60 or more minutes of transient ischemia.

Influence of the Duration of the Reperfusion Period on Retinal Ganglion Cell Densities The densities of FG-labeled RGCs in the experimental left retinas that underwent similar periods of transient ischemia were compared at different survival time intervals to determine whether the survival interval also had an effect on RGC survival. In particular, we analyzed statistically the data obtained in groups of retinas subjected to transient periods of ischemia that induce RGC death—60, 75, 90, 105, and 120 minutes. The percentages of RGC survival in the experimental left retinas after periods of ischemia of 60 minutes were 67%, 63%, 51%, and 45%, respectively, for survival periods of 5, 7, 14, and 30 days. There were no significant differences between the densities obtained at 5, 7, or 14 days after ischemia (Figs. 3, 4). For transient periods of ischemia of 75 minutes, the percentages of RGC survival in experimental retinas versus controls were 57%, 41%, 18%, and 24% for retinas analyzed 5, 7, 14, or 30 days, respectively, after ischemia. Statistical analysis indicated significant differences between the densities obtained after survival periods of 5 and 14 days (P < 0.003, Mann-Whitney test) and after survival periods of 7 and 14 days (P < 0.005, Mann-Whitney test). These results indicate that in this group of retinas subjected to 75 minutes of ischemia, RGC death was already present by day 5 and continued at least between days 5 and 14 after ischemia. Similarly, the percentage of RGCs surviving in experimental retinas after 90 minutes of ischemia decreased with increasing survival intervals. After 90 minutes of ischemia, there were significant differences between the densities of RGCs obtained after 5 and

2010

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10

14 days of survival (P < 0.005, Mann-Whitney test) and after 7 and 14 days of survival (P < 0.005, MannWhitney test). For the group of retinas that underwent transient periods of ischemia of 105 minutes, there were significant differences between the densities of RGCs obtained after 7 and 14 days of survival (P < 0.03, Mann-Whitney test). For the group of retinas that underwent transient periods of ischemia of 120 minutes, there were significant differences between the densities of RGCs obtained after 5 and 14 days of survival (P< 0.005, Mann-Whitney test). Altogether, these results indicate that RGC death induced by transient ischemia is observed as soon as 5 days after injury but also progresses during the following days; thus, the proportion of cell death depends not only of the duration of the blood flow deprivation, but also of the duration of the survival interval (Fig. 4). DISCUSSION The experimental strategy used in this study allows quantitative estimations of the neuronal loss that follows transient periods of ischemia within a well-defined part of the central nervous system of the adult mammal, the retina. The quantification of RGC loss has been possible by the use of the fluorescent tracer, FG, which, when applied to the main retino-recipient targets in the brain, labels the entire population of RGCs. Retinal ischemia was induced by increasing the intraocular pressure by means of conjunctival traction. Both techniques, RGC labeling and intraocular pressure increase by traction, were combined for the first time in these studies to investigate RGC loss after ischemia. Use of Fluorogold to Identify Retinal Ganglion Cells Prelabeling RGCs before ocular injury is important to distinguish them from other retinal neurons present in the ganglion cell layer, such as displaced amacrine cells.18'31 Retinal ganglion cells and displaced amacrine cells are not easily distinguishable on the basis only of morphologic appearance or cell size.1'2 Furthermore, injured RGCs may change their phenotype, and this may make even more difficult the identification of these neurons.32"35 The use of neuroanatomic tracers applied to the main retino-recipient targets in the brain of the rat to mark RGCs by retrograde transport allows in vivo quantitative studies of RGC survival after injury in wholemounts1"4"7'27"29'36 or radial sections28 of the retina. In this study, we have estimated FG-labeled RGC densities in each retina from the pictures taken from 12 standard areas located within the central retina, between 1 and 3 mm from the optic disc. Because the pictures were taken from the central 3 mm of the

retina, we did not examine cell densities in more peripheral regions of the retina, where RGC densities are smaller37'38 and where the effects of transient ischemia of the retina have been reported to be less severe.10 Fluorogold Labeling in Control Retinas In the control retinas, FG fluorescence was found in the cytoplasm of the somata and sometimes proximal dendrites of RGCs. The intensity of the fluorescence of the RGCs did not seem to vary throughout the study. However, the quantification of labeled cells documented that there was a 10% decline in the densities of FG-labeled RGCs in the retinas examined 37 days after FG application. The difference between densities of FG-labeled RGCs 12,14, or 21 days after FG application and 37 days after FG application was statistically significant and revealed that FG does not persist within the cytoplasm of the entire RGC population for periods of time exceeding 3 weeks after its application to the superior colliculi and dorsolateral geniculate nuclei. The decrease in the density of RGCs labeled with FG 21 days or more after its application in the brain is probably not caused by neurotoxicity of the tracer and death of the labeled cells. Seven days after FG injection into the lateral rectus muscle, the entire population of abducens motoneurons in the brain stem appears labeled with FG.39 However, when abducens nuclei were examined for longer periods of time after FG injection (up to 3 months), there was a diminution with time in the number of labeled abducens motoneurons, indicating that FG does not persist for long periods of time. If long periods persisted after FG injection, when the numbers of labeled motoneurons diminished, we administered a second injection of FG into the lateral rectus muscle. The numbers of labeled cells are similar and not statistically significant from the numbers obtained 7 days after the first application. This indicates that FG is not neurotoxic and that FG does not persist in the cytoplasm of neurons for long periods of time.39 Furthermore, FG labeling of cells other than RGCs in control retinas or cells other than the abducens motoneurons in the abducens nuclei was not observed, suggesting that the disappearance of FG is not caused by transcellular passage of the tracer. Thus, it is possible that with time FG loses its fluorescence or is metabolized, as has been suggested in other studies.40 Use of Increments of the Intraocular Pressure to Induce Retinal Ischemia Several strategies that increase the intraocular pressure have been used previously to induce transient ischemia of the retina, and most of these require intraocular puncture.9'12'15'11''41"50 We raised the intraocular

Retinal Ganglion Cell Survival After Ischemia

2011

pressure by pulling in opposite directions from two silk sutures anchored on the bulbar conjunctiva, next to the corneoscleral limbus. This method has several advantages: It allows direct visualization of the eye fundus through the microscope during the ischemic period to assess the intraretinal blood flow; it is easy to implement and is reproducible, as shown by the consistency of the results obtained in the different groups analyzed in the current study; and finally, this method is not invasive and does not require intraocular puncture of the anterior or posterior chamber of the eye, diminishing the risk for infection, direct lesions to the intraocular structures, or bleeding. Intraocular puncture may lead to nonspecific inflammatory neuroprotective effects on injured RGCs7 and thus could affect RGC survival after ischemia.

13.52,53 j n m o n k e y S j for example, the time period that induces irreversible damage to the retina appears to be even longer than that observed in the adult rat. Electrophysiological experiments have indicated that the occlusion of the central retinal artery in squirrel monkeys induces within 90 seconds cessation of spontaneous activity of RGCs, as well as a loss of optic tract discharges in response to photic stimulation of the retina.54 Complete recovery of these electrophysiological signals was observed after periods of ischemia of less than 120 minutes.54 Therefore, it appears that in monkeys, the threshold lies between 90 and 120 minutes. A similar situation may occur in humans, in whom it is classically accepted that the critical period for irreversible retinal damage after central retinal artery occlusion is approximately 3 hours.51

This method, however, also has several disadvantages: the retinal ischemia induced in these experiments is complete because this method interrupts the blood flow in the central artery of the retina as well as in the choroid, and this is not usually seen clinically in occlusion of the central artery of the retina.51 Furthermore, increases of intraocular pressure induce a more severe loss of retinal neurons than other methods used to induce transient ischemia of the retina, such as ligature of the ophthalmic vessels or ligature of both carotid arteries. 4 ' Finally, we cannot exclude the possibility that in addition to ischemia-induced injury of the retina, other types of injury over the RGCs or their axons might by inflicted by the elevation of the intraocular pressure above systolic levels.

Between 60 and 90 minutes of ischemia, RGC death progresses dramatically. Fourteen days after ischemia, the proportion of surviving RGCs is approximately 15%, and it is similar for all the periods of ischemia of 75 or more minutes. However, 30 days after ischemia, RGC survival is higher after 75 minutes of ischemia than after 90 or more minutes, when only approximately 5% of the RGC population survive. Thus, it appears that there is a critical period for inducing the death of the majority of the population of RGCs; in this study, it was between 75 and 90 minutes. It is interesting that the same proportion of RGCs, approximately 5%, survive the longer periods of transient ischemia (90, 105, and 120 minutes). We have observed that a similar small percentage of the RGC population survives for long periods of time up to 21 months after intraorbital optic nerve transection.^ It is possible that a small population of RGCs may survive different types of injury. In the current study, however, we were unable to examine whether this proportion of cells survived longer than 1 month.

Effects of the Duration of the Transient Ischemia on Retinal Ganglion Cell Death One of the main objectives of the current study was to quantify the effects of transient periods of ischemia on the survival of RGCs. We induced periods of ischemia of 30, 45, 60, 75, 90, 105, and 120 minutes and analyzed the animals 5, 7, 14, or 30 days later. We found that periods of ischemia of 30 or 45 minutes, irrespective of the interval of survival, do not produce significant cell death. Because we did not find FG-labeled microglial cells in the retinas of these animals, and because the results of these experiments were consistent for all groups of animals analyzed at different survival intervals, it is unlikely that our method failed to detect cell death induced by these periods of ischemia. We have observed that 60 minutes of ischemia induces significant RGC loss at all the survival intervals. Thus, there is a threshold or critical period of ischemia that produces irreversible damage to the rat RGCs, which in the current study was situated between 45 and 60 minutes. This observation is in agreement with previous studies also indicating a remarkable tolerance of the inner retinal layers to ischemia.8"10 "'

Effects of Survival Interval After Transient Ischemia on Retinal Ganglion Cell Death By 5 days after transient ischemia of 60 or more minutes, an important percentage of the RGC population is already lost. For example, 5 days after 90 or 120 minutes of transient ischemia, nearly 50% or 70%, respectively, of the RGC population is dead. In preliminary experiments, we have observed RGC loss as soon as 3 days after ischemia (data not shown), although technical reasons prevented us from quantifying the loss of RGCs at this time period. The early abrupt cell loss that follows transient periods of intraocular pressure increased above systolic levels may be an early response to ischemia. Previous studies have indicated RGC death as early as 3 and 24 hours, and 3 days after transient ischemia9 with the characteristics of both apoptosis and necrosis.9 Previous in vivo and in vitro studies55 suggest that large RGCs are more susceptible

2012

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10

to glutamate neurotoxicity, one of the excitotoxic mechanisms implicated in the ischemic injury. Our study did not address the question of which type of RGC survive better after retinal ischemia. To examine the role of large survival intervals on RGC loss after transient pressure-induced ischemia of the retina, RGC densities obtained for similar periods of transient ischemia were compared at different survival intervals. The principal finding was that the proportion of RGCs that were lost depended not only on the duration of the transient ischemia but also on the survival interval; the longer the survival interval, the more extensive the loss of RGCs. Thus, a novel finding of these studies is that RGC loss continues between day 5 and 30 after the ischemic insult. We have found that for similar periods of transient ischemia, RGC loss is smaller for shorter survival intervals. This suggests that after ischemia of the retina, a substantial proportion of RGCs undergo delayed death. It is possible that damaging mechanisms induced at the time of transient ischemia may have persisted for the entire survival interval of our experiments. Hence, in addition to the acute effects of transient ischemia,916 other mechanisms may have played a role in the RGC death observed in the current experiments. Other non-RGC retinal neurons, which we have not investigated, also are affected by complete transient ischemia of the retina.91049 For example, previous studies indicate ischemia-induced death of retinal neurons, with smaller thresholds for inner nuclear layers than for outer nuclear layers.910"5'49 It is conceivable that the loss of inner and outer nuclear layer retinal neurons might induce transneuronal degeneration of RGCs. It is also possible that the increase of intraocular pressure used in this study caused not only an interruption of retinal blood flow but an obstruction of the axonal transport at the optic nerve head, thus mimicking the effects of an a axotomy. Recent experiments that have induced chronic increases of the IOP to reproduce experimental glaucoma have indicated impairment of axonal transport in RGCs at the level of the optic nerve head and delayed apoptotic RGC death. 6 ' 6 Retinal ganglion cell death after intraorbital ON section first appears by day 53A and continues for the life span of the animal.2 It will be interesting to determine whether the delayed RGC death observed in the current study is caused by axonal compression from increased IOP. A new set of experiments to elucidate this possibility is required and may shed light on the understanding of glaucoma. Although great effort has been directed toward the identification of putative neuroprotective substances after ischemia,"12'17'47"57'58 few of these substances have been assessed quantitatively in vivo. The procedure used here, to study the survival of a homog-

enous population of central nervous system prelabeled neurons, could provide the basis for further detailed quantitative analysis on the effects of such putative neuroprotective substances in mammals after transient ischemia. In summary, in the adult rat retina, irrespective of the survival interval, transient ischemia of 45 minutes or less does not induce RGC loss; periods of ischemia between 60 and 90 minutes induce the loss of a proportion of the original RGC population that is related to the duration of the transient period of ischemia; and periods of ischemia between 90 and 120 minutes induce the loss of approximately 75% to 95% of the original RGC population. A small proportion of RGCs does not appear to be affected by these periods of ischemia of the retina. Our results also revealed that transient ischemia induced-RGC death was already present 5 days later but continued for the duration of the study. In conclusion, our results demonstrate that the duration of the initial transient period of ischemia and the duration of the survival interval influence RGC death. Key Words adult rat retina, fluorogold, intraocular pressure, neuronal degeneration, retinal ganglion cell death, transient ischemia Acknowledgments The authors thank Dr. Marcelino Aviles for his advice. References 1. Villegas-Perez MP, Vidal-Sanz M, Bray GH, Aguayo A. Influences of peripheral nerve grafts on the survival and regrowth of axotomiced retinal ganglion cells in adult rats. /Neurosd. 1988;8:265-280. 2. Villegas-Perez MP, Vidal-Sanz M, Raminsky M, Bray GH, Aguayo AJ. Rapid and protected phases of retinal ganglion cell loss follow axotomy in the nerve of adulr rats. / Neurobiol. 1993;24:23-36. 3. Berkelaar M, Clarke DB, Wang Y-C, Bray GM, Aguayo AJ. Axotomy results in delayed death and apoptosis of retinal ganglion cells in adults rats. / Neurosd. 1994; 14:4368-4374. 4. Peinado-Ramon P, Salvador M, Villegas-Perez MP, Vidal-Sanz M. Effects of axotomy and intraocular administration of NT-4, NT-3, and BDNF on the survival of adult rat retinal ganglion cells: A quantitative in vivo study. Invest Ophthalmol Vis Sd. 1996; 37:489-500. 5. Garcia-Valenzuela E, Gorczyca W, Darzynkiewicz Z, Sharma SC. Apoptosis in adult rat retinal ganglion cells after axotomy. /Neurobiol. 1994;25:431-438. 6. 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 Sd. 1995; 36:774-786. 7. Mansour-Robaey S, Clarke DB, Wang Y-C, Bray GM, Aguayo AJ. Effects of ocular injury and the administration of brain-derived neurotrophic factor (BDNF) on

Retinal Ganglion Cell Survival After Ischemia

8.

9.

10. 11.

12.

13. 14. 15.

16.

17.

18. 19.

20.

21.

22. 23.

24.

the survival and regrowth of axotomized retinal ganglion cells. ProcNatlAcadSci USA. 1994;91:1632-1636. Biichi ER, Suivaizdis I, FuJ. Pressure-induced retinal ischemia in rats: An experimental model for quantitative study. Opthalmobgica. 1991;203:138-147. Biichi ER. Cell death in the rat retin after a pressureinduced ischaemia-reperfusion insult: An electron microscopic study: I: Ganglion cell layer and inner nuclear layer. Exp Eye Res. 1992a; 55:605-613. Hughes WF. Quantitation of ischemic damage in the rat retina. Exp Eye Res. 1991; 53:573-582. Szabo ME, Droy-Lefaix MT, Doly M, Carre C, Braquet P. Ischemia and reperfusion-induced histologic changes in the rat retina. Demonstration of a free radical mediated mechanism. Invest Ophthlamol Vis Sci. 1991;32:1471-1478. Takahasi K, Lam TT, Edward DP, Buchi ER, Tso MOM. Protective effects of flunarizine on ischemic injury in the rat retina. Arch Ophthalmol. 1992; 110:862-870. Kroll AJ. Experimental central retinal artery occlusion. Arch Ophthalmol. 1968; 79:453-469. Johnson NF, Foulds WS. The effects of total acute ischaemia on the structure of the rabbit retina. Exp Eye Res. 1978;27:45-59. Louzada-Junior P, DiasJJ, Santos WF, LachatJJ, Bradford HF, Coutinho-Netto J. Glutamate release in experimental ischaemia of the retina: An approach using microdialysis. / Neurochem. 1992; 59:358—363. Biichi ER. Cell death in rat reina after pressure-induced ischaemia-reperfusion insult: Electron microscopic study: II: Outer nuclear layer. JpnJ Ophthalmol. 1992b; 36:62-68. Unoki K, LaVail M. Protection of the rat retina from ischemic injury by brain-derived neurotrophic factor, ciliar neurotrophic factor, and basic fibroblast growth factor. Invest Ophthalmol Vis Sci. 1994; 35:907-915. Perry VH. Evidence for an amacrine cell system in the ganglion cell layer of the rat retina. Neurosdence. 1981;6:931-944. Vidal-Sanz M, Villegas-Perez MP, Bray GM, Aguayo AJ. Persistent retrograde labeling of adult rat retinal ganglion cells with the carbocyanine dye dil. Exp Neurol. 1988; 102:92-101. Vidal-Sanz M, Selles I, Salvador M, Villegas-Perez MP. Quantitative studies on retinal ganglion cell death after transient periods of ischemia. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1994;35:S2087. Vidal-Sanz M, Selles I, Salvador M, Villegas-Perez MP. Short and long term effects of transient periods of ischemia on retinal ganglion cell death a quantitative in vivo study. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1995;36:4295. Linden R, Perry VH. Massive retinotectal projection in rats. Brain Res. 1983;272:145-149. Martin PR. The proyection of different retinal ganglion cell classes to the dorsal lateral geniculate nucleus in the hooded rat. Exp Brain Res. 1986; 62:7788. Schumed LC, Fallon JH. Fluoro-gold: A newfluores-

2013

25.

26.

27.

28.

29.

30. 31. 32.

33.

34.

35.

36. 37. 38.

39.

40.

cent retrograde axonal tracer with numerous unique propierties. Brain Res. 1986;377:147-154. Faberowski N, Stefansson E, Davidson RC. Local hypothermia protects the retina from ischemia. Invest Ophthalmol Vis Sci. 1989; 30:2309-2313. Dodd J, Solter D, Jessell TM. Monoclonal antibodies against carbohydrate differentiation antigens identify subsets of primary sensory neurones. Nature. 1984; 311:469-472. Thanos S. The relationship of microglial cells to dying neurons during natural neuronal cell death and axotomy induced degeneration of the rat retina. Eur J Neurosci. 1991; 3:1189-1207. Thanos S, Mey J, Thield HJ. Specific transcellular staining of microglia-suppressing factors retards axotomy-induced neuronal degradation and enhances axonal regeneration 'in vivo' and 'in vitro.' J Neurosci. 1992; 13:455-466. Thanos S, Mey J, Wild M. Treatment of the adult retina with microglia-suppressing factors retards axotomy-induced neuronal degradation and enhances axonal regeneration 'in vivo' and 'in vitro.'J Neurosci. 1993; 13:455-466. Perry VH, Hendrerson Z, Linden R. Postnatal changes in retinal ganglion cell and optic axon populations in the pigmented rat. / Comp Neurol. 1983;219:356-368. Perry VH, Cowey A. The effects of unilateral cortical and tectal lesions on retinal ganglion cells in rats. Exp Brain Res. 1979; 35:85-95. Vidal-Sanz M, Bray GM, Villegas-Perez MP, Aguayo AJ. Axonal regeneration and synapse formation in the superior colliculus by retinal ganglion cells in the adult rats. J Neurosci. 1987;8:265-280. Thanos S. Alterations in the morphology of ganglion cell dendrites in the adult retina after optic transection and grafting of peripheral nerve segments. Cell Tissue Res. 1988; 254:599-609. McKerracher L, Vidal-Sanz M, Aguayo AJ. Slow transport rates of cytoskeletal proteins change during regeneration of axotomized retinal neurons in adults rats. J Neurosci. 1990a; 10:641-648. McKerracher L, Vidal-Sanz M, Essagian C, Aguayo AJ. Selective impairments of slow axonal transport after optic nerve injury in adult rats. / Neurosci. 1990b; 10:2834-2841. Vidal-Sanz M, Villegas-Perez MP, Bray GM, Aguayo AJ. Use of peripheral nerve grafts to study regeneration after CNS injury. Neuroprotocol. 1993;3:29-33. Fukuda Y. A three group classification of rat retinal ganglion cells: Histological and physiological studies. Brain Res. 1977; 119:327-344. Schober W, Gruscha H. Die Ganglionzelle der retina der Albinoaratte: Eine qualitative und quantitative studie. ZMikroskAnatForsch. 1977;91:39997-414. In German. Gomez Ramirez AM, Villegas-Perez MP, Salvador M, Vidal Sanz M. Use the fluorescent tracer fluorogold to identify the motoneuron population of the abducens nucleus: A quantitative in vivo study. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1995;36:S687. Crews LL, Wigston DJ. Dependence of motoneurons

2014

41.

42. 43.

44.

45.

46. 47. 48.

49.

Investigative Ophthalmology & Visual Science, September 1996, Vol. 37, No. 10

on their target muscle during natal development of the mouse. J Neurosd. 1990; 10:1643-1653. Reinecke RD, Kuwara T, Cogan DG, Weiss DR. Retinal vascular patterns: V: Experimental ischemia of the cat eye. Arch Ophthalmol. 1962;67:110-115. Anderson DR, Davis B. Sensitives of ocular tissues to acute pressure-induced ischemia. Arch Ophthalmol. 1975;93:267-274. Melamed S, Ben-Sira I, Ben-Shaul Y. Acutely elevated intraocular pressure on rabbit retinal pigment epithelium and photorecepters. Glaucoma. 1981JanFeb:59-66. Neetens A, Delaunois AL, Hendrata Y, Rompaey JV. Experimental animal study of the effects of intraocular pressure and systemic blood pressure on optic pathway action potentials. Exp Eye Res. 1981;32:575581. Siliprandi R, Bucci MG, Canella R, Carmignoto G. Flash and pattern electroretinograms during and after acute intraocular pressure elevation in cats. Invest Ophthalmol Vis Sd. 1988; 4:558-565. Yoon YH, Marmor MF. Dextromethorphan protects retina against ischemic injury 'in vivo.' Arch Ophthalmol. 1989; 107:409-441. Tsukahara Y, Blair NP, Spellman Eappen DC, et al. Ketamina suppresses ischemic injury in the rabit retina. Invest Ophthalmol Vis Sd. 1992;33:1822-1825. Nayak MS, Kita M, Marmor MF. Protection of rabbit retina from injury by superoxide dismutase an catalase. Invest Ophthalmol Vis Sd. 1993; 34:2018-2022. Osborne NN, Larsen A, Barnett NL. Influence of excitatory amino acids and ischemia on rat retinal cho-

50.

51. 52.

53.

54.

55. 56. 57. 58.

line acetyltransferase-containing cells. Invest Ophthalmol Vis Sd. 1995;36:1692-l700. Veriac S, Tissie G, Bonne C. Oxygen free radicals adversely affect the regulation of vascular tone by nitric oxide in the rabbit retina under high intraocular pressure. Exp Eye Res. 1993;56:85-88. Brown GC. Retinal arterial obstructive disease. In: Schachat AA, Murphy RB, eds. Ryan's Retina. St. Louis: CVMosby; 1994; 1361-1377. Faberowski N, Stefansson E, Davidson RC. Local hypothermia protects the retina from ischemia. A quantitative study in the rat. Invest Ophthalmol Vis Sd. 1989; 30:2309-2313. Hayreh SS, Weingeist TA. Experimental occlusion of the central artery of the retina: IV: retinal tolerance time to acute ischaemia. Br J Ophthalmol. 1980b; 64:818-825. Hamasaki I, Kroll AJ. Experimental central retinal artery occlusion: An electrophysiological study. Arch Ophthalmol. 1968; 80:243-248. Dreyer EB, Pan ZH, Storm S, Lipton SA. Greater sensitivity of larger retinal ganglion cell death. Neuroreport. 1994;5:629-631. Garcia-Valenzuela E, Shareef S, Walsh J, Sharma SC. Programmed cell death of retinal ganglion cells during experimental glaucoma. Exp Eye Res. 1995; 61:33-44. Zhu CZ, Auer ZN. Centrally administered insulin and IGF-1 in transient forebrain ischaemia in fasted rats. NeurolRes. 1994; 16:116-120. Blair NP, Shaw WE, Dunn R, et al. Limitation of retinal injury by vitreoperfusion initiated after onset of ischemia. Arch Ophthalmol. 1991; 109:113-118.