Molecular Vision 2001; 7:192-201 <http://www.molvis.org/molvis/v7/a27/> Received 24 July 2001 | Accepted 11 August 2001 | Published 15 August 2001

Download Reprint

Changes in Thy1 gene expression associated with damaged retinal ganglion cells

Cassandra L. Schlamp,¹ Elaine C. Johnson,² Yan Li,¹ John C. Morrison,² Robert W. Nickells¹
 
 

¹Department of Ophthalmology and Visual Sciences, University of Wisconsin, Madison, WI; ²Casey Eye Institute, Oregon Health Sciences University, Portland, OR

Correspondence to: Dr. Robert W. Nickells, Department of Ophthalmology and Visual Science, Room 6640, Medical Science Center, University of Wisconsin, 1300 University Avenue, Madison, WI, 53706; Phone: (608) 265-6037; FAX: (608) 262-0479; email: nickells@facstaff.wisc.edu

Abstract

Purpose: The temporal series of molecular events that occur in dying retinal ganglion cells is poorly understood. We have examined the change in expression of a normally-expressed ganglion cell marker gene, Thy1, relative to the kinetics of cell loss caused by acute and chronic damaging stimuli.

Methods: For acute experiments, mice were subjected to optic nerve crush or intravitreal injections of N-methyl-D-aspartate (NMDA) to induce ganglion cell death. RNase protection analysis was used to quantify Thy1 mRNA levels from total retina RNA and in situ hybridization was used to monitor the pattern of Thy1 positive cells. Changes in Thy1 expression were compared to the time course of cell loss induced by each treatment. To induce elevated intraocular pressure (IOP), the episcleral veins of rats were injected with hypertonic saline, which scleroses Schlemm's Canal and the trabecular meshwork. Elevated IOP was monitored every day for 35 days after which the animals were sacrificed and the retinas harvested for quantitative RT-PCR or fixed for in situ hybridization studies. Evaluation of glaucomatous damage caused by elevated IOP was determined from histological sections of the optic nerves of all rat eyes.

Results: After optic nerve crush, Thy1 mRNA levels decreased within 24 h, although the number of expressing cells did not decline until 7 days. Both measures showed a loss of Thy1 well in advance of cell loss, which was detected by 2 weeks after surgery. This change in expression was not dependent on execution of the cell death program since a similar decrease was detected in Bax^-/- ganglion cells, which are resistant to cell death induced by optic nerve crush. Thy1 mRNA levels and the number of expressing cells also decreased within 6 h after NMDA injection, in advance of cell loss, which was detected by 24 h. Similarly, elevated intraocular pressure was associated with a decrease in mRNA and expressing cells in a pressure-dependent manner. In moderately hypertensive rat eyes, the number of cells expressing Thy1 decreased before significant cell loss in the retina. Virtually no Thy1-expressing cells were detected in eyes with severe disease.

Conclusions: Thy1 mRNA abundance and expressing cells, decreased in advance of detectable ganglion cell loss caused by three different modalities of damage. This change is independent of the committed step of cell death.

Introduction

A complete understanding of neuronal cell death may have an important impact on the development of new treatments for a variety of chronic and acute neurological diseases. Retinal ganglion cells have historically been a model cell-type for study both in vitro and in vivo, since they represent an accessible population of neurons of the central nervous system [1] and have been examined in numerous experimental paradigms. Ganglion cells are also the primary retinal cell-type affected by several blinding diseases including glaucoma, tumor-associated optic nerve compression, and ischemic optic neuropathy.

The mechanism of ganglion cell death has been studied under a variety of conditions. Several experimental stimuli activate their death including direct damage to the optic nerve, exposure to elevated concentrations of excitatory amino acids, or moderate to severe elevations in intraocular pressure (IOP). These cells also undergo programmed cell death where half the population dies during normal retinal development [2,3]. In each of these circumstances, cell death exhibits morphological and biochemical characteristics of apoptosis [4-13]. Recent studies have focused on the genes that control ganglion cell death with particular emphasis on the genes that are known to regulate this process in other cell-types [14]. Ganglion cell survival is influenced by genes such as p53 [15], Bcl2 [16,17], and Bax [18-20]. Kinetic analysis of p53 [15,21] and Bax expression [19] indicate that both these genes are upregulated in ganglion cells during periods of cell death. In addition, retinal ganglion cells in Bax-deficient mice show nearly complete attenuation of developmental cell loss [18] and are significantly resistant to the damaging effects of optic nerve crush [20]. These studies have provided some insight into the sequence of events in the ganglion cell apoptotic program.

Reports on other neuronal cell-types have suggested that prior to the upregulation of apoptotic regulatory genes, metabolic processes, such as protein synthesis and glucose uptake, are decreased [22]. Similar decreases in gene expression may also occur. Retinal ganglion cells, for example, exhibit a dramatic decrease in the expression of some genes, including BclX [23], before detectable cell loss. It is not clear if this decrease in gene expression is a consequence of the activation of the cell death program or if it is an independent event. To better understand the phenomenon, we characterized the expression of the Thy1 gene and compared it to the rate of ganglion cell loss in response to 3 independent damaging stimuli. The Thy1 gene codes for a 25 kDa cell surface glycoprotein [24] of unknown function. It is predominantly expressed by ganglion cells in the retina [25,26], making it a useful marker for evaluating gene expression in these cells. Previous studies have shown that Thy1 message and/or protein were depleted after damage to the optic nerve and retina [26-28], an observation that has been attributed to the loss of ganglion cells in affected retinas. In other experiments, however, the loss of Thy1 expression appeared to precede cell death [11]. In this study, changes in Thy1 mRNA abundance were compared to the number of cells expressing Thy1 and the rate of cell loss in retinas exposed to two acute damaging stimuli in mice and chronic elevated IOP in a rat model of experimental glaucoma. In each experimental paradigm, Thy1 mRNA levels decreased prior to significant cell loss. In addition, Thy1 expression also decreased in ganglion cells resistant to cell death, suggesting that this downregulation precedes a committed step in the apoptotic program of these cells.

Methods

Experimental animals

The rats and mice used in this study were handled in accordance with the Association for Research in Vision and Ophthalmology Resolution for the use of animals in research. Optic nerve crush and intravitreal injections of 160 nmoles of NMDA were conducted as described previously [11]. In each procedure, only one eye of each mouse was treated, leaving the other eye as a control for that mouse. To control for the injection procedure, balanced saline solution (BSS) without NMDA was injected. In the optic nerve crush experiments, animals were sacrificed and the eyes harvested at 1, 3, 7, 14, and 21 days after crush. For NMDA-injections, the animals were sacrificed and the eyes harvested at 1, 6, 12, 24, and 48 h after injection. The analysis of Thy1 gene expression relative to cell loss was conducted in CB6F1 mice (an inbred F1 hybrid originally from the cross of a BALB/c female with a C57BL/6J male). The analysis of Thy1 expression in a Bax^-/- background was conducted on mice with a predominantly C57BL/6J background [29].

Elevated IOP was induced in Brown Norway rats (Rattus norvegicus) by the introduction of 50 ml a hypertonic saline solution (3 M NaCl) into a single episcleral vein that results in the sclerosis of Schlemm's Canal and the trabecular meshwork as described previously [30]. In all cases, animals were maintained under constant light conditions to reduce circadian-rhythm induced changes in IOP [31]. Pressures were measured daily on awake animals using a calibrated hand-held Tonopen (Mentor; Norwell MA) for a period of 35 days after surgery. All the measurements of IOP reported here are actual uncorrected Tonopen readings [30]. At the end of this time, the animals were sacrificed. In all eyes the optic nerve was harvested and sectioned to determine the extent of nerve damage sustained using a 1-5 grading scale described previously [30,31].

RNase protection analysis

The abundance of Thy1 mRNA was determined by RNase protection analysis (RPA). Each experiment was conducted using total RNA isolated from four pooled retinas from either control or experimental eyes at each time point. The methods used for RNA isolation and RPA have been described previously [23,32]. For each sample, the amount of total RNA was quantified by spectrophotometric and denaturing gel electrophoretic analysis and an equal amount of RNA was input into each reaction. The probe used for Thy1 analysis was a 305 bp DNA fragment corresponding to the third exon of the murine Thy1 gene [11]. Quantification of the amount of radioactivity in a protected probe was conducted by excising the fragment from a denaturing polyacrylamide gel and counting it in a Beckman LS 6500 Liquid Scintillation Counter (Beckman Instruments, Inc., Fullerton, CA). The amount of Thy1 mRNA for each sample of pooled experimental retinas was calculated as a percentage of Thy1 mRNA in the pooled control retinas of the same mice. Three separate RPA experiments were used to collect data for the kinetic analysis of Thy1 mRNA loss and two separate experiments were used to monitor the loss of Thy1 mRNA in Bax^+/+ and Bax^-/- animals.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

Retinas were quickly dissected from enucleated eyes and frozen at -70 °C. RNA was extracted by the method of Chomczynski [33], quantified spectrophotometrically and checked for quality by gel electrophoresis. For reverse transcription (RT) of total RNA, 150 ng retinal RNA was incubated in 20 ml of RT transcription buffer (50 mM Tris HCl, pH 8.3, 75 mM KCl, 3 mM MgCl₂, 8 mM dithiothreitol, 0.1 mM each deoxyribonucleoside triphosphates, 0.8 U/ml RNAsin RNase inhibitor, and 8 ng/ml oligo-dT_12-18), and incubated at 37 °C for 2 h with 8 U/ml M-MLV reverse transcriptase (BRL; Rockville, MD). As a control for DNA contamination of retinal RNA, identical 150 ng aliquots from each sample were pretreated with 20 ug/ml RNase A and 14 U/ml RNase T1 for 45 min at 37 °C prior to RT. A standard RNA curve (38-600 ng retinal RNA) was utilized to check the linearity of the transcription reaction with the amount of added RNA. For uniformity in transcription, the RT reaction was performed on dilution curve RNAs and all sample RNAs simultaneously.

For PCR cDNA amplification of each of the sample and standard curve RT reaction products, a 2 ml aliquot of a 1:3 RT dilution was added to 18 ml of polymerase-free PCR reaction mixture overlaid with mineral oil. Following denaturation and "hotstart" (4 min at 94 °C), AmpliTaq DNA Polymerase (0.625 units/5 ml; Perkin Elmer; Branchburg, NJ) was added, followed by "touchdown" (20 cycles of annealing, extension and denaturation with the annealing temperature decreasing from 69 °C by 1 degree each 2 cycles), and 8 cycles of denaturation (15 s at 94 °C), annealing (1 min at 55 °C) and polymerase extension (2 min at 72 °C). A combination of MgCl₂ (4 mM), deoxyribonucleotides (1 mM each), and primer pairs (8 pmol for Thy1 and 3 pmol for GAPDH) were found to result in the linear production of reaction products. For uniformity of amplification, all samples and standards were amplified in one procedure.

The primers used for each cDNA were designed using Primer Designer 3 software for Windows (Sci-ed Software, State Line, PA) using sequences for rat mRNAs available in the NIH National Center for Biotechnology Information databases. For Thy1, the forward primer was 5' CAG GAC GGA GCT ATT GGC ACC AT and the reverse primer was 5' ACG GCA GTC CAG TCG AAG GTT CT, to yield a product of 138 bp. For GAPDH, the forward primer was 5' CAT CAA GAA GGT GGT GAA GCA GG and the reverse primer was 5' CCA CCA CCC TGT TGC TGT AGC CA to yield a product of 206 bp.

PCR product gel electrophoresis, photography and semi-quantitative analysis

PCR products from all samples and standards were separated on a single ethidium bromide-3% agarose gel and exposed to UV light. Digital images were obtained and the density of the product bands determined using LabWorks Image Acquisition and Analysis software (Upland, CA). Thy1 message density was normalized relative to GAPDH density, which did not change quantitatively in any of the samples and served as a control for both mRNA quantification and RT-PCR cDNA yield. RT-PCR data were collected from samples originating from single eyes. Data were then pooled into groups that were defined by the mean IOP of each eye over the 35 day period. Each group was based on data obtained from a minimum of 8 different eyes.

In situ hybridization

The same Thy1 probe used for RPA experiments was used for in situ hybridization. After sacrificing the animals (rats or mice), the eyes were enucleated, punctured with a number 65 scalpel blade, and fixed in 4% (W/V) paraformaldehyde in 100 mM NaPO₄ buffer (pH 7.2) for 1 h at 22 °C, followed by 0.4% paraformaldehyde in phosphate buffer overnight at 4 °C. The retinas were then dissected from the fixed globes and analyzed using a whole-mount protocol and digoxigenin-labeled RNA probes (Roche Biochemicals, Indianapolis, IN) as described previously [23,32]. For mouse studies, hypbridization was performed on the entire superior half of each retina. For rat studies, two pieces of the superior retina, each measuring approximately 2 x 2 mm, were excised from each eye and used for hybridization. Thy1 mRNA in the retina was visualized using anti-digoxigenin antibodies conjugated to alkaline phosphatase. Stained retinas were embedded in JB-4 glycolmethacrylate (Polysciences, Inc., Warrington, PA) and sectioned at a thickness of 4 mm. Sections were adhered to slides and counter-stained with 4,6-diamidino-2-phenylindole (DAPI; Roche Molecular Biochemicals, Indianapolis, IN). A region of 400 mm of retina adjacent to the optic nerve (in mice) or consisting of the entire block for the rat eyes, was sequentially photographed under fluorescence microscopy and Nomarski interference microscopy. Quantification of Thy1-positive cells was determined first by counting all of the DAPI-stained nuclei in the ganglion cell layer that were larger than 2 mm in diameter and exhibiting obvious nucleoli. The number of Thy1-positive cells was determined by counting the number of alkaline phosphatase-labeled cells in the ganglion cell layer in the Nomarski photographs of the same section. For mice, a minimum of two sections of each eye were photographed and scored by two masked observers. Each point of counted data shown represents the mean (± standard deviation) of 6-8 mice. Rats were evaluated in a similar manner, except that each eye was scored separately from a minimum of 11 sections of each retina.

Results

Thy1 mRNA levels rapidly decrease in response to acute stimuli that cause ganglion cell death

RPA analysis of total retina RNA indicated that Thy1 mRNA levels decreased within 24 h after optic nerve crush and steadily declined to nearly undetectable levels by 2 weeks (Figure 1). Intravitreal injection of 160 nmoles of NMDA initially induced an increase in Thy1 mRNA levels, which then rapidly decreased below control eye levels by 6 h after injection. By 48 h, only traces of Thy1 mRNA were detected in NMDA-injected eyes. The transient increase in Thy1 message was not observed in BSS-injected eyes (Figure 1).

The decrease in Thy1 mRNA levels in acutely-stressed retinal ganglion cells precedes cell death

In situ hybridization analysis was used to monitor the pattern of Thy1 expressing cells in control and experimental eyes. Figure 2 shows a series of micrographs taken from representative specimens at different time points after optic nerve crush. Numerous positively-stained cells were detected in the ganglion cell layer as late as 3 days after surgery (data not shown), but few positive cells were evident by 7 days and no positive cells were found by 14 days (Figure 2D,F). In comparison, the corresponding DAPI-stained images of the same sections showed no cell loss at 7 days and minimal cell loss at 14 days (Figure 2C,E). Figure 3 shows in situ hybridization results following NMDA injection. Similar to the optic nerve crush results, the number of Thy1 positive cells rapidly decreased after injection of the excitotoxin. Very few stained cells were detected after 6 h, although DAPI-staining indicated that the majority of cells were still present in the experimental retinas at this time point (Figure 3C,D). By 24 h after injection, Thy1 positive cells were scarce (Figure 3F).

The relative amount of Thy1 mRNA determined by RPA analysis was compared to the DAPI-counts and Thy1-positive cell counts obtained from similarly treated mice for both optic nerve crush (Figure 4A) and NMDA-injection (Figure 4B). Within 24 h after optic nerve crush, Thy1 mRNA levels had dropped rapidly. This decrease was not evident by in situ hybridization, which showed the same number of expressing cells as control retinas. By 7 days, both the mRNA level and the number of Thy1-positive cells had dramatically decreased to approximately 20% of the control fellow eyes and by 14 days Thy1 mRNA could only be detected using RPA. In contrast with the loss of Thy1 expression, the total number of DAPI-stained cells present in the ganglion cell layer was relatively unchanged 7 days after crush (91.4 ± 3.3%, percent of control retinas, mean ± SD) and only reduced by approximately 28% (72.25 ± 19.9%) by 14 days. The injection of NMDA stimulated a transient increase in Thy1 mRNA accumulation, but not an increase in the total number of Thy1 expressing cells. By 6 h, both the number of positive cells and the Thy1 mRNA levels had decreased while the total number of cells present had not changed. By 48 h, the retinas exhibited approximately 43% cell loss (57.2 ± 16.9% cells remaining relative to control fellow eyes), but more than 80% loss of the Thy1 mRNA or expressing cells compared to the fellow control eyes.

Analysis of Thy1 expression in response to optic nerve crush in Bax^+/+ and Bax^-/- mice

Optic nerve crush was also performed on mice deficient for the Bax gene. Bax^-/- mice have twice the number of retinal ganglion cells than wild-type littermates [18], and do not exhibit any cell loss up to 4 weeks after optic nerve crush [20]. The change in Thy1 expression 1 week after optic nerve crush was monitored by in situ hybridization and RPA in Bax^-/- animals and wild-type littermates (Figure 5). Wild-type mice showed a dramatic reduction in the number of Thy1 expressing cells (Figure 5A,B), although 87.4 ± 13.3% (mean ± standard deviation; n = 4) of the cells were still remaining in the ganglion cell layer. A similar decrease in the number of Thy1 expressing cells was observed in Bax^-/- mice at the same time-point (Figure 5C,D), although these animals still contained 93.3 ± 11.5% (n = 4) of cells in the ganglion cell layer. RPA analysis confirmed the loss of Thy1 mRNA in both groups of mice (Figure 5E). Quantitative analysis of two experiments indicated that Bax^-/- mice contained approximately 2.5 times more Thy1 mRNA per mg of total RNA relative to wild-type mice reflecting the increased number of ganglion cells in these animals. One week after optic nerve crush, both groups of mice exhibited an average of 50% and 45% loss of Thy1 mRNA levels in the wild-type and knock-out mice, respectively.

Analysis of Thy1 expression in rats with ocular hypertension and experimental glaucoma

RT-PCR analysis was used to quantify Thy1 mRNA in rat retinas subjected to elevated IOP. Figure 6 shows an ethidium bromide-stained gel of representative samples from a typical experiment and a summary table of all the RT-PCR data. The samples have been grouped according to the mean elevated IOP sustained by each eye during the 35 day period of the experiment. An increase in optic nerve damage was correlated with increased elevated IOP in all groups, consistent with previous observations of this model of experimental glaucoma [30]. Similarly, IOPs of 30 or greater were associated with a decrease in Thy1 mRNA levels, with statistically significant decreases observed in eyes exposed to mean IOPs greater than 34 mm Hg. Similar results were obtained from the in situ hybridization analysis of the superior retinas of hypertensive rat eyes (Figure 7 and Table 1). Control rat eyes and eyes with no discernable optic nerve disease exhibited a mean of 48.5 ± 6.1, (SD) cells/400 mm of ganglion cell layer. Of these cells, a mean of 68.1 ± 5.3% (SD) were positive for Thy1 mRNA (Figure 7A). Hypertensive eyes used in this study could be grouped into two categories. The first of these contained retinas with minimal optic nerve damage and no significant loss of retinal cells (47.5 ± 4.0 cells/400 mm, p > 0.10, Mann Whitney Test). These eyes, on average, exhibited fewer Thy1 expressing cells, although this decrease was not statistically significant (57.5 ± 13.1%, p > 0.10). More striking, however, was the observation that 3 of the 5 eyes in this group exhibited a qualitative difference in expression compared to normal eyes. This difference consisted of both strongly and weakly expressing cells (Figure 7B). In our quantitative analysis of these data, however, both these kinds of cells were counted as expressing Thy1. The second category of eyes included specimens with extensive optic nerve disease, and a significant decrease in both total cell number (25.5 ± 6.9 cells/400 mm, p = 0.05) and the number of Thy1-positive cells (3.9 ± 3.1%, p = 0.05, Figure 7C). A summary table of the in situ hybridization experiments is shown in Table 1.

Discussion

A comparative analysis of Thy1 mRNA levels by RPA and Thy1 expressing cells by in situ hybridization showed that stimuli that activate ganglion cell death caused a rapid decrease in the expression of this gene. After optic nerve crush, the decrease of message probably occurred uniformly in most of the ganglion cells since mRNA levels decreased before a reduction in the number of expressing cells was detected. Similarly, after NMDA injection, the transient increase in mRNA did not coincide with an increase in the number of expressing cells indicating that these cells likely increased already active gene expression. After this increase, however, mRNA levels and the number of expressing cells decreased with similar kinetics suggesting that mRNA loss could be attributed to the rapid shut-down of expression in individual cells. Thy1 expression also decreased prior to cell death in a rat model of experimental glaucoma. These changes were apparently not uniform, however, since rats with mild to moderate damage showed qualitative changes in Thy1 expression with neighboring cells exhibiting either very high or very low levels of Thy1 in in situ hybridization experiments. Collectively, the loss of Thy1 expression preceded detectable ganglion cell loss in all experimental conditions examined.

Several studies have indicated that the majority of Thy1 expression in the adult mammalian retina is attributable to ganglion cells [25,26,34,35]. However, is the phenomenon we have observed in this study restricted to these cells? Panning experiments with antiThy1 antibodies [36], optic nerve transection studies [26], and transgenic mouse lines made using the Thy1 promoter [35], have suggested that other retinal cells also express this gene. The most likely candidates for these "other" cell types are amacrine and Müller cells. Two lines of evidence suggest that our results are predominantly ganglion cell-specific. First, most of the data indicating non-ganglion cell expression were obtained from early post-natal retinas before completion of programmed cell death of the ganglion cell layer. Similar results have not been reported for adult retinas. Second, Dabin and colleagues [34] found that Müller cell expression was linked to neuronal damage. Although damage is an existing condition in our experiments, we found (i) nearly complete loss of Thy1 mRNA levels, (ii) minimal indication of expression in cells in the inner nuclear layer, where the cells bodies of Müller cells reside, (iii) no subsequent increase that could be attributable to stress-induced activation of expression in Müller cells, and (iv) an equal loss of both the number of Thy1 positive cells and, eventually, actual cell loss under conditions in which only ganglion cells are expected to die (optic nerve crush). Based on these arguments, it is likely that the loss of Thy1 mRNA we observed are representative of changes in ganglion cell gene expression. It is possible, however, that the residual Thy1 expression that we detect using RNase protection assays or RT-PCR is due to continued expression of this gene in unaffected non-ganglion cells or in surviving ganglion cells themselves.

Thy1 as an indicator of stressed ganglion cells

One of the greatest difficulties faced by researchers studying the biology of ganglion cell death is finding a suitable functional assay for these cells. This is particularly relevant in studies designed to test the efficacy of neuroprotective drugs in vivo [37]. At present, the most commonly accepted measure of effect is to count "surviving" ganglion cells, or axons in the optic nerve, at a given time point in the experiment. This measure is insufficient to determine how healthy these cells are or whether they retain the potential to function properly.

Several independent studies indicate that ganglion cells undergo early functional deficits and atrophy prior to dying, particularly in a glaucomatous eye. Detailed morphometric analyses showed significant degeneration and shrinkage in the dendritic arbors of dye-injected parasol ganglion cells in a primate model of experimental glaucoma [38]. These changes in the arbor were followed by shrinkage of both the cell soma and axon diameter. Similarly, Morgan and co-workers [39] documented a decrease in overall cell size that preceded ganglion cell death in experimental glaucoma. Li et al. [20] also found decreases in cell size in Bax knock-out mice after optic nerve crush, indicating that this phenomenon can be independent of cell death. Together, the early morphological changes and our results showing early decreases in the expression of a marker gene indicate that ganglion cells experience important changes that precede cell death. Weber and colleagues [38] suggested that this period of change could represent a window of opportunity in which sick ganglion cells may be rescued from the subsequent steps leading to cell death. This hypothesis can be tested using the expression of Thy1 as a molecular marker. Decreasing IOP in the rat model of experimental glaucoma, for example, may rescue ganglion cells, which in turn would reactivate Thy1 expression. Additionally, Thy1 expression could serve as a sensitive indicator of effect of neuroprotective agents, eliminating the need to wait for ganglion cell death to occur.

The mechanism of Thy1 down-regulation and the effects on other genes

The mechanism that controls the down-regulation of Thy1 is not known, but it may affect other genes as well. Previous studies by this laboratory and others have shown that several "normal" genes, including the anti-apoptotic gene BclX [23], are down-regulated in ganglion cells exposed to damaging stimuli. It is possible that damaged ganglion cells undergo a shut-down of many normally expressed genes as an early step in apoptosis, which is consistent with reports that neurons exposed to excitotoxins exhibit a rapid decrease in several metabolic functions including protein synthesis and glucose transport [22]. It is reasonable to predict that a dying cell would shut down non-essential functions as a prelude to executing its cell death program. At minimum, such a mechanism would help the cell preserve existing energy stores for this ATP-dependent process, which includes subsequent upregulation of a variety of genes essential to the cell death pathway. At present, the importance of the down-regulation phenomenon to the execution of the cell death program is not clear, but it may play a critical role in the complex changes that signal this program.

A principal mechanism for down-regulation of expression is a reduction in transcription. Assuming that mRNA turnover rates remain constant, a decrease in transcription would result in the depletion of transcript pools and eventually a reduction in newly synthesized protein. For example, NMDA injection causes a loss of Thy1 mRNA levels within hours, but Thy1 protein immunoreactivity decreases over 8 days after similar treatment in rats [28]. Widespread transcriptional silencing could also be achieved by a simple mechanism such as altering the conformation of active chromatin [40,41]. Currently, the number of normal genes examined in ganglion cells is too small to support the argument that a global shutdown of gene expression is part of the apoptotic pathway in these cells. Further evaluation of this phenomenon, however, may provide important insight into the sequence of events occurring in dying ganglion cells.

Acknowledgements

This work was supported in part by grants from the National Eye Institute (R29 EY12223 to RWN and R01 EY10145 to JCM); the American Health Assistance Foundation; the Retina Research Foundation; the Glaucoma Research Foundation and Research to Prevent Blindness, Inc. Dr. Morrison is also a recipient of a Research to Prevent Blindness Lew R. Wasserman Merit Award. The authors would like to thank Dr. Stanley Korsmeyer for generously providing Bax^-/+ mice for these experiments and Mr. Keith Poulsen for genotyping the animals.

References

1. Dowling JE. The retina: an approachable part of the brain. Cambridge (MA): Harvard University Press; 1987.

2. Rodriguez-Tebar A, Jeffrey PL, Thoenen H, Barde YA. The survival of chick retinal ganglion cells in response to brain-derived neurotrophic factor depends on their embryonic age. Dev Biol 1989; 136:296-303.

3. Henderson CE. Programmed cell death in the developing nervous system. Neuron 1996; 17:579-85.

4. Ilschner SU, Waring P. Fragmentation of DNA in the retina of chicken embryos coincides with retinal ganglion cell death. Biochem Biophys Res Commun 1992; 183:1056-61.

5. 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; 36:774-86.

6. 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.

7. Garcia-Valenzuela E, Gorczyca W, Darzynkiewicz Z, Sharma SC. Apoptosis in adult retinal ganglion cells after axotomy. J Neurobiol 1994; 25:431-8.

8. Berkelaar M, Clarke DB, Wang YC, Bray GM, Aguayo AJ. Axotomy results in delayed death and apoptosis of retinal ganglion cells in adult rats. J Neurosci 1994; 14:4368-74.

9. Kerrigan LA, Zack DJ, Quigley HA, Smith SD, Pease ME. TUNEL-positive ganglion cells in human primary open-angle glaucoma. Arch Ophthalmol 1997; 115:1031-5.

10. Levin LA, Louhab A. Apoptosis of retinal ganglion cells in anterior ischemic optic neuropathy. Arch Ophthalmol 1996; 114:488-91.

11. Li Y, Schlamp CL, Nickells RW. Experimental induction of retinal ganglion cell death in adult mice. Invest Ophthalmol Vis Sci 1999; 40:1004-8.

12. Lam TT, Abler AS, Kwong JM, Tso MO. N-methyl-D-aspartate (NMDA)--induced apoptosis in rat retina. Invest Ophthalmol Vis Sci 1999; 40:2391-7.

13. Mittag TW, Danais J, Pohorenec G, Yuan HM, Burakgazi E, Chalmers-Redman R, Podos SM, Tatton WG. Retinal damage after 3 to 4 months of elevated intraocular pressure in a rat glaucoma model. Invest Ophthalmol Vis Sci 2000; 41:3451-9.

14. Nickells RW. Apoptosis of retinal ganglion cells in glaucoma: an update of the molecular pathways involved in cell death. Surv Ophthalmol 1999; 43:S151-61.

15. Rosenbaum DM, Rosenbaum PS, Gupta H, Singh M, Aggarwal A, Hall DH, Roth S, Kessler JA. The role of the p53 protein in the selective vulnerability of the inner retina to transient ischemia. Invest Ophthalmol Vis Sci 1998; 39:2132-9.

16. Martinou JC, Dubois-Dauphin M, Staple JK, Rodriguez I, Frankowski H, Missotten M, Albertini P, Talabot D, Catsicas S, Pietra C, et al. Overexpression of BCL-2 in transgenic mice protects neurons from naturally occurring cell death and experimental ischemia. Neuron 1994; 13:1017-30.

17. Bonfanti L, Strettoi E, Chierzi S, Cenni MC, Liu XH, Martinou J-C, Maffei L, Rabacchi SA. Protection of retinal ganglion cells from natural and axotomy-induced cell death in neonatal transgenic mice overexpressing bcl-2. J Neurosci 1996; 16:4186-94.

18. Mosinger Ogilvie J, Deckwerth TL, Knudson CM, Korsmeyer SJ. Suppression of developmental retinal cell death but not photoreceptor degeneration in Bax-deficient mice. Invest Ophthalmol Vis Sci 1998; 39:1713-20.

19. Isenmann S, Wahl C, Krajewski S, Reed JC, Bahr M. Up-regulation of Bax protein in degenerating retinal ganglion cells precedes apoptotic cell death after optic nerve lesion in the rat. Eur J Neurosci 1997; 9:1763-72.

20. Li Y, Schlamp CL, Poulsen KP, Nickells RW. Bax-dependent and independent pathways of retinal ganglion cell death induced by different damaging stimuli. Exp Eye Res 2000; 71:209-13.

21. Joo CK, Choi JS, Ko HW, Park KY, Sohn S, Chun MH, Oh YJ, Gwag BJ. Necrosis and apoptosis after retinal ischemia: involvement of NMDA-mediated excitotoxicity and p53. Invest Ophthalmol Vis Sci 1999; 40:713-20.

22. Miller TM, Moulder KL, Knudson CM, Creedon DJ, Deshmukh M, Korsmeyer SJ, Johnson EM Jr. Bax deletion further orders the cell death pathway in cerebellar granule cells and suggests a caspase-independent pathway to cell death. J Cell Biol 1997; 139:205-17.

23. Levin LA, Schlamp CL, Spieldoch RL, Geszvain KM, Nickells RW. Identification of bcl-2 family of genes in the rat retina. Invest Ophthalmol Vis Sci 1997; 38:2545-53.

24. Williams AF, Gagnon J. Neuronal cell Thy-1 glycoprotein: homology with immunoglobulin. Science 1982; 216:696-703.

25. Barnstable CJ, Drager UC. Thy-1 antigen: a ganglion cell specific marker in the rodent retina. Neuroscience 1984; 11:847-55.

26. Perry VH, Morris RJ, Raisman G. Is Thy-1 expressed only by ganglion cells and their axons in the retina and optic nerve? J Neurocytol 1984; 13:809-24.

27. Osborne NN, Larsen AK. Antigens associated with specific retinal cells are affected by ischaemia caused by raised intraocular pressure: effect of glutamate antagonists. Neurochem Int 1996; 29:263-70.

28. Nash MS, Osborne NN. Assessment of Thy-1 mRNA levels as an index of retinal ganglion cell damage. Invest Ophthalmol Vis Sci 1999; 40:1293-8.

29. Knudson CM, Tung KS, Tourtellotte WG, Brown GA, Korsmeyer SJ. Bax-deficient mice with lymphoid hyperplasia and male germ cell death. Science 1995; 270:96-9.

30. Morrison JC, Moore CG, Deppmeier LM, Gold BG, Meshul CK, Johnson EC. A rat model of chronic pressure-induced optic nerve damage. Exp Eye Res 1997; 64:85-96.

31. Jia L, Cepurna WO, Johnson EC, Morrison JC. Patterns of intraocular pressure elevation after aqueous humor outflow obstruction in rats. Invest Ophthalmol Vis Sci 2000; 41:1380-5.

32. Schlamp CL, Nickells RW. Light and dark cause a shift in the spatial expression of a neuropeptide-processing enzyme in the rat retina. J Neurosci 1996; 16:2164-71.

33. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162:156-9.

34. Dabin I, Barnstable CJ. Rat retinal Muller cells express Thy-1 following neuronal cell death. Glia 1995; 14:23-32.

35. Feng G, Mellor RH, Bernstein M, Keller-Peck C, Nguyen QT, Wallace M, Nerbonne JM, Lichtman JW, Sanes JR. Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron 2000; 28:41-51.

36. Barres BA, Silverstein BE, Corey DP, Chun LL. Immunological, morphological, and electrophysiological variation among retinal ganglion cells purified by panning. Neuron 1988; 1:791-803.

37. Neufeld AH, Sawada A, Becker B. Inhibition of nitric-oxide synthase 2 by aminoguanidine provides neuroprotection of retinal ganglion cells in a rat model of chronic glaucoma. Proc Natl Acad Sci U S A 1999; 96:9944-8.

38. Weber AJ, Kaufman PL, Hubbard WC. Morphology of single ganglion cells in the glaucomatous primate retina. Invest Ophthalmol Vis Sci 1998; 39:2304-20.

39. Morgan JE, Uchida H, Caprioli J. Retinal ganglion cell death in experimental glaucoma. Br J Ophthalmol 2000; 84:303-10.

40. Pazin MJ, Kadonaga JT. What's up and down with histone deacetylation and transcription? Cell 1997; 89:325-8.

41. Wolffe AP, Pruss D. Targeting chromatin disruption: transcription regulators that acetylate histones. Cell 1996; 84:817-9.