Abstract
The availability of the human genome sequence together with sequenced genomes of several model organisms provides an unprecedented opportunity to utilize comparative genomic approaches for the discovery of genes that contribute to human disease. We have used transgenic flies to establish an experimental paradigm for the discovery of genes that might be involved in the development of glaucoma, a prevalent disease affecting a large segment of the population. Inherited mutations in the trabecular meshwork-inducible glucocorticoid response protein/myocilin (TIGR/MYOC) are associated with juvenile glaucoma and some cases of adult primary open angle glaucoma. The interrelationships between TIGR/MYOC and the development of glaucoma, however, are not understood. We show that overexpression of human TIGR/MYOC in the eyes of Drosophila melanogaster results in distortion of ommatidia accompanied by fluid discharge. High-density oligonucleotide microarrays identified altered expression of 50 transcripts in response to TIGR/MYOC overexpression, including homologs of aquaporin-4 and cytochrome-P450, previously associated with glaucoma, and several proteins of unknown function. We found that expression of Swiss Cheese, a neurodegenerative protein, increased 34-fold and that its human ortholog, neuropathy target esterase, is also upregulated in response to adenovirus-mediated overexpression of TIGR/MYOC in perfused postmortem human eyes. Our observations establish the Drosophila eye as an advantageous system for the discovery of genes that are associated with glaucoma.
THE availability of the human genome sequence together with sequenced genomes of several model organisms provides an unprecedented opportunity to utilize comparative genomic approaches for the discovery of genes that contribute to human disease. Drosophila melanogaster is a particularly powerful genetic model organism and is becoming increasingly appreciated as a model for human disease (Warricket al. 1998; Feany and Bender 2000; Wittmannet al. 2001; Muqit and Feany 2002). However, this model is useful only if evidence can be provided that demonstrates that human orthologs of genes discovered in Drosophila, in fact, contribute to the actual disease. On the basis of this criterion we have used transgenic flies in conjunction with human postmortem organ culture to establish an experimental paradigm for the discovery of genes that might be relevant for the development of glaucoma, a prevalent disease affecting a large segment of the population.
Glaucoma is the second leading cause of blindness affecting ∼66 million people worldwide (Quigley 1996). Glaucoma is often accompanied by ocular hypertension caused by impaired outflow facility of aqueous humor through the trabecular meshwork in the anterior segment of the eye (Shieldset al. 1996). Some individuals develop ocular hypertension when they are treated with corticosteroids. This observation led to the discovery of the trabecular meshwork-inducible glucocorticoid response protein (TIGR; Ortegoet al. 1997, Polanskyet al. 1997; Stoneet al. 1997; Nguyenet al. 1998) or myocilin (MYOC; Kubotaet al. 1997). The expression of TIGR/MYOC is upregulated dramatically by glucocorticoids (Nguyenet al. 1998; Clarket al. 2001) and to a lesser extent by prolonged elevated intraocular pressure (Borráset al. 2002), mechanical stress (Tammet al. 1999), and oxidative insult (Polanskyet al. 1997). Studies of families with glaucoma showed that mutations in TIGR/MYOC protein were associated with autosomal juvenile glaucoma and some cases of adult primary open angle glaucoma (Adamet al. 1997; Stoilovaet al. 1997; Stoneet al. 1997; Alwardet al. 1998; Richardset al. 1998; Wiggset al. 1998). Consequently, considerable effort has focused on determining the role of TIGR/MYOC in the pathogenesis of glaucoma. Central questions are whether altered expression of TIGR/MYOC is the cause or the consequence of ocular hypertension, whether overexpression of wild-type or mutant forms of TIGR/MYOC alone would be sufficient to trigger ocular hypertension, how upregulation of TIGR/MYOC affects the transcriptome, whether TIGR/MYOC interacts directly with other proteins in the trabecular meshwork, and whether modulation of TIGR/MYOC expression might be of therapeutic benefit. Progress in addressing these questions has been restricted by lack of an animal model for glaucoma.
Previously, D. melanogaster has been used as a model for a variety of neurodegenerative diseases (Muqit and Feany 2002), including Huntington's disease (Warricket al. 1998), Parkinson's disease (Feany and Bender 2000), and neurodegenerative tauopathy (Wittmannet al. 2001). Although the compound eye of Drosophila shows little structural similarity to the human eye, recent successes in targeted expression of human neurodegenerative disease genes in flies motivated us to investigate whether tissue-specific overexpression of TIGR/MYOC might establish the Drosophila eye as a model in which genomic approaches could be applied to determine the mechanisms that underlie the transcriptional response relevant to glaucoma. Here we report that targeted overexpression of human TIGR/MYOC in the fly eye results in fluid discharge from the eye. Furthermore, transcriptional profiling identifies ensembles of up- and downregulated genes in flies that overexpress TIGR/MYOC, several of which have human homologs previously associated with glaucoma. Finally, we show that upregulation of the neurodegenerative Drosophila swiss cheese gene product (Kretzschmaret al. 1997) is paralleled by increased expression of its human ortholog, neuropathy target esterase (NTE; Lushet al. 1998; Glynn 1999), in postmortem human eyes in which TIGR/MYOC is delivered to the trabecular meshwork via an adenovirus vector (Borráset al. 1998). Our studies establish Drosophila as the only in vivo model system available for studies of the effects of TIGR/MYOC on the transcriptional response to intraocular pressure and illustrate the power of parallel studies between Drosophila as a candidate gene discovery system and perfused postmortem human eyes as a validation system for the identification of genes that are modulated under conditions of elevated pressure and may, ultimately, contribute to the development of glaucoma.
MATERIALS AND METHODS
Generation and maintenance of transgenic flies: A cDNA insert encoding the complete coding region of human TIGR/ MYOC was excised from the pMC2 vector and cloned into the KpnI and NotI sites of the pUAST transformation vector. The sequence of the cloned insert was confirmed and the construct was designated pUAST-TIGR. P-element transformation was performed in the Samarkand w1118 strain (Rubin and Spradling 1982) with 400 μg/ml pUAST-TIGR DNA and 100 μg/ml of the wing clipped helper plasmid pπ25.7wc (Karess and Rubin 1984). Two independent transformant lines with different pUAST-TIGR insertion sites were derived, one homozygous viable and the other homozygous lethal. We crossed the homozygous viable line to flies carrying gmr-GAL4 (Drosophila Stock Center, Bloomington, IN) to drive expression of TIGR/MYOC in the eye. The GAL4 driver line, gmr-GAL4, carries the transposon P{GAL4-ninaE.GMR}, which drives GAL4 expression in all eye cells posterior to the furrow (Elliset al. 1993; Freeman 1996).
TIGR/MYOC antibody: The DNA sequence encoding the first 148 amino acids of TIGR/MYOC was amplified from genomic DNA with primers 5′-GGTATTGAGGGTCGCAGGA CAGCTCAGCTCAGCTCAGGAAG-3′ and 5′-AGAGGAGAGT TAGAGCCTCACAGCCTTGCTACCTCCTG-3′. Conditions for PCR amplification were 1 min at 94°, 2 min at 58°, and 3 min at 72° for 30 cycles followed by 10 min of incubation at 72°. The amplification product was cloned into the pET-32 Xa/LIC vector (Novagen, Cambridge, MA) and its sequence was verified. Following purification of the resulting thioredoxin fusion protein, rabbit antibodies were produced and tested against a preparation of perfused postmortem human eyes overexpressing TIGR/MYOC. The antisera showed specific immunoreactivity at 5000-fold dilution. Expression of TIGR/MYOC in transgenic flies was analyzed by Western blots with a 2500-fold dilution of antiserum. Bound antibody was visualized with a horseradish peroxidase-labeled goat-anti-rabbit secondary antibody using 3′3-diaminobenzidine as the chromogen (Vector Laboratories, Burlingame, CA). Migration distances were calibrated against Kaleidoscope prestained low-molecular-weight markers (Bio-Rad, Hercules, CA).
Transcriptional profiling: For gene profiling studies, RNA was extracted from heads of adult flies; harvested 3–5 days posteclosion; reared under controlled conditions of temperature (25°), humidity (70%), and light cycle (12 hr/12 hr); and deprived of food for several hours prior to RNA extraction. Biotinylated cRNA probes were prepared for hybridizations to high-density oligonucleotide microarrays (Affymetrix) and visualized with a streptavidin-phycoerythrin conjugate, as described in the Affymetrix GeneChip Expression Analysis 2000 technical manual, using internal references for quantification.
The average difference (AD) values were obtained for each gene (Affymetrix GeneChip Expression Analysis 2000 technical manual) and analyzed by one-way ANOVA according to the model Y =μ+ G + E, where G is the effect of genotype (pUAST-TIGR and gmr-GAL4 parents and the pUAST-TIGR; gmr-GAL4 F1 hybrid) and E is the error variance. All genes for which the main effect of G was significant at P < 0.05 were further tested to determine whether the mean expression of the F1 hybrid was significantly different from the average of the two parental strains. Analyses were performed using Proc GLM and Proc Means using SAS statistical software. Candidate genes for which expression in the F1 was significantly different from the mean of the parents at P < 0.01 were further filtered by requiring that there were no more than four absent calls, that AD values in the F1 were at least 600 units, and that the difference in expression between the F1 and the midparent was twofold or greater. Original microarray data sets are accessible at http://www.genetics.org/supplemental/.
Expression of TIGR/MYOC in perfused postmortem human eyes: Anterior segment organ cultures were prepared from postmortem human eyes within 30–40 hr of death (Johnson and Tschumper 1987; Borráset al. 2002). Ages of donors were between 75 and 86 years; none had been diagnosed with glaucoma. Eyes were bisected at the equator; lens, iris, and vitreous were removed; and anterior segments were mounted onto a perfusion chamber. High-glucose, serum-free medium was perfused at a constant flow of 3–4 μl/min using a Harvard microinfusion pump. Following 24 hr of perfusion to allow stabilization, one eye was injected through the cornea with 20 μl of AdhTIG3 (2 × 1010 particles) while the contralateral eye was injected with the same particle number of Ad5.CMV-Null. After additional 48- to 72-hr perfusion, anterior segments were immersed in RNAlater (Ambion, Austin, TX), trabecular meshworks were dissected, and RNA was extracted using a QIAshredder column and an RNeasy kit (QIAGEN, Valencia, CA). AdhTIG3 contains the full coding region of the TIGR/MYOC gene driven by the cytomegalovirus (CMV) promoter (Caballero and Borrás 2001) and the Ad5.CMV-Null contains no foreign cDNA (QBIOgene).
For mRNA quantitation of TIGR/MYOC and NTE, reverse transcriptions were performed using random primers (Ambion). Multiplex amplifications for each gene and ribosomal RNA were performed in triplicate with a predetermined linear cycle range. Ribosomal RNA was amplified at a nonsaturated concentration with a QuantumRNA 18S primer-competimer set (Ambion). Primers for TIGR/MYOC were 5′-CTGGAG GAAGAGAAGAAGCGACTAA-3′ and 5′-CTGTGTCATAAGC AAAGTTGACGGTA-3′. Primers for NTE were 5′-GGAAGAG GGACAAAGTGCTCTTCTA-3′ and 5′-TAGCTGAGGTGCTG ACCATTCT-3′. Amplified products were run in a 2% ethidium bromide agarose gel, band intensities were captured with a ChemiDoc system and LabWorks software (UVP, Upland, CA), and values were transferred to an Excel spreadsheet for calculation of means and standard errors.
RESULTS
Overexpression of TIGR/MYOC in Drosophila eyes results in periodic fluid discharge: To drive overexpression of TIGR/MYOC in the Drosophila eye we cloned a full-length cDNA encoding TIGR/MYOC into the pUAST transformation vector and generated transgenic pUAST-TIGR flies. In situ hybridization with a probe corresponding to the white gene to larval polytene salivary gland chromosomes showed that the transgene had inserted in cytological location 58E. We used transactivation through the binary GAL4-UAS system (Brand and Perrimon 1993) to drive expression of TIGR/MYOC by crossing the eye-specific gmr-GAL4 driver line (Elliset al. 1993) with the pUAST-TIGR parental line and analyzing the offspring. Overexpression of TIGR/MYOC in the F1 was confirmed by immunoblotting with an antibody against TIGR/MYOC. Whereas no specific immunoreactivity is observed in the parental strains, an intense immunoreactive band appears in the F1 with a molecular size of ∼62 kD (Nguyenet al. 1998; Figure 1). Higher molecular weight aggregates, previously observed in human eyes, were not detected. The molecular size of the band, slightly greater than the 55- to 57-kD band observed in human eyes, and its diffuse appearance suggest that the protein expressed in the Drosophila eye has undergone extensive glycosylation (Nguyenet al. 1998; Russellet al. 2001).
Whereas the eyes of the parental strains appear normal (Figure 2A), flies carrying both the gmr-GAL4 and the pUAST-TIGR constructs show periodic discharge of liquid from their eyes (Figure 2B). We observed droplets emerging from focal regions of the eye as well as large wet areas that dried to a crusty residue (Figure 2C). Flies at different stages of ocular fluid discharge were observed, ranging from eyes that appeared normal to eyes with extensive liquid discharge. Precise quantification of this phenotype is difficult, since fluid discharge is rapid and transient, and transferring flies from their vial onto a microscope stage easily brushes the liquid or residues off. Attempts to quantify this phenotype would, therefore, lead to gross underestimates and not faithfully represent the occurrence of fluid discharge. Thus, the assessment of the phenotype is at present only qualitative. However, flies discharging liquid from their eyes are readily observed in every batch examined.
—Overexpression of TIGR/MYOC in flies carrying the gmr-GAL4 and pUAST-TIGR transgenes. Each lane contains a homogenate of five fly heads. Antibodies against TIGR/ MYOC visualize intense expression of TIGR/MYOC in F1 offspring (lane 2; arrowhead), but not in the parental gmr-GAL4 (lane 1) or pUAST-TIGR (lane 3) lines.
Our observations suggest that either fluid extrusion occurs periodically in all flies or only a subpopulation displays this phenotype. To resolve this question we examined histological sections through heads from both parental strains and their progeny. Parental pUAST-TIGR flies show a normal geometric array of ommatidia (Figure 3A). Ommatidia of all flies carrying both the gmr-GAL4 and the pUAST-TIGR transgenes, however, are distorted with convex boundaries (Figure 3B). Such distorted ommatidia are observed in all eyes of TIGR/ MYOC-overexpressing flies, but never in controls. Since we observe in each section 50 fly heads in many, random orientations, we can rule out artifacts due to different planes of sectioning. In some cases the surface of the eye contains a dimpled area, underneath which ommatidia appear to have regained their normal shape (Figure 3C). We interpret such areas as regions where recent fluid discharge has occurred. These observations and our finding that all F1 flies observed histologically showed morphological abnormalities favor our interpretation that fluid discharge occurs periodically in all flies in which TIGR/MYOC is overexpressed in the eye.
—Fluid discharge from eyes of transgenic flies. (A) The normal appearance of eyes from a fly of the pUAST-TIGR parental line. Flies carrying both the gmr-GAL4 and the pUAST-TIGR transgenes show focal discharge of liquid from the eye (black arrows in B) and wet areas that appear hazy under the microscope (white arrow in B). As the liquid dries, a residue, which is readily removed upon handling of the flies or through grooming, forms on the eye (C, arrow).
To ascertain that the F1 phenotype did not reflect a positional effect of the pUAST-TIGR insertion site rather than the action of TIGR/MYOC protein, we evaluated an additional independent transformant line. Offspring from crosses with the gmr-GAL4 line contained flies that displayed the same ocular fluid discharge. We conclude that in the Drosophila system overexpression of TIGR/MYOC alone is sufficient to produce the ocular fluid discharge phenotype.
Modification of the transcriptome in Drosophila eyes that overexpress TIGR/MYOC: We investigated the genome-wide transcriptional response to TIGR/MYOC overexpression in transgenic flies to identify candidate genes with homologs whose expression may change in human eyes that overexpress TIGR/MYOC. We compared expression profiles of the parental gmr-GAL4 and pUAST-TIGR lines and the TIGR/MYOC-expressing hybrid. Duplicate RNA samples were extracted from heads of equal numbers of male and female flies and hybridized to high-density oligonucleotide expression arrays (Affymetrix). Data for each gene were analyzed by one-way ANOVA and ∼700 genes with significant variation were identified. To take into account the different genetic backgrounds of the parental lines, we tested for each of these genes whether the mean level of expression in the F1 was significantly different from the average of the two parental lines and retained genes with a contrast P value <0.01. Since our microarray data could nonetheless be influenced by genetic background effects in this mixed parental genotype transgenic model and to reduce the likelihood of false positives, we further imposed the constraints that the mean expression in the F1 be 600 units greater and be twofold or more different from the midparent. The 20 upregulated and 30 downregulated transcripts satisfying these criteria are listed in Table 1. Original microarray data sets are accessible at http://www.genetics.org/supplemental/.
Overexpression of TIGR/MYOC induces protein kinase C 98E and a serine/threonine kinase, both of unknown function. In addition, there is upregulation of a protein tyrosine kinase and a transmembrane receptor protein tyrosine phosphatase, suggesting induction of post-translational modifications of multiple target proteins. Of special interest is the dramatic 34-fold upregulation of the expression of a neurodegenerative protein, Swiss Cheese (Kretzschmaret al. 1997). Eye-specific gene products, including rhodopsin, bride-of-sevenless, and the inaF ion channel also show increased expression. Tissue damage (Figure 3B) is reflected in increased levels of proteolytic enzymes and reduction in transcripts that encode cytoskeletal and structural proteins, including actin and dynein heavy chain, and several proteins involved in nucleic acid and protein synthesis. Two downregulated transcripts that encode a water transporter-like protein homologous to aquaporin-4 and a cytochrome P450 isoenzyme are of special interest, as aquaporin-4 may play a role in glaucoma (Hanet al. 1998) and cytochrome P450 has been linked to families with the disease (Stoilovet al. 1997; Bejjaniet al. 1998; Kakiuchi-Matsumotoet al. 2001). Two G protein-linked orphan receptors together with 20 other gene products that encode proteins of unknown function also display altered expression as a consequence of TIGR/MYOC overexpression.
—Histology of ommatidia in transgenic flies. Formalin-fixed and paraffin-embedded 10-μm sections through 50 randomly oriented heads were stained with hematoxylin and eosin. The ommatidia of flies of the parental pUAST-TIGR line appear normal with a well-defined array of ommatidia with straight boundaries (A). Flies that carry both the gmr-GAL4 and the pUAST-TIGR transgenes have ommatidia that appear inflated with rounded boundaries (B). In some cases, presumably following fluid extrusion, the eye has a dimpled area underneath which ommatidia appear normal (C, between arrowheads). Bar, 50 μm.
Upregulation of the human ortholog of the Drosophila swiss cheese gene product in the trabecular meshwork of postmortem human eyes that express TIGR/MYOC: The restricted group of gene products modulated by TIGR/MYOC overexpression includes proteins previously implicated in glaucoma. We hypothesized that other proteins—either of unknown function or of known function but not previously implicated in glaucoma—that show altered expression in the Drosophila eye as a consequence of TIGR/MYOC expression might also contribute to the pathogenesis of glaucoma. To validate this hypothesis we focused on the swiss cheese transcript (Kretzschmaret al. 1997), which is 34-fold upregulated in the TIGR/MYOC-expressing Drosophila eye and which has a human ortholog, NTE (Lushet al. 1998; Glynn 1999). To test whether induction of TIGR/ MYOC in human eyes would result in expression of NTE, we introduced TIGR/MYOC in perfused postmortem human eyes via an adenovirus vector (Borráset al. 1998). This perfusion system, in which one eye serves as the experimental subject and the contralateral as a control, is not confounded by genetic background differences as both eyes are obtained from the same individual; it is essentially the equivalent of the GAL4-UAS binary expression system in Drosophila. To eliminate potential gene inductions due to the presence of the adenovirus core we treated the control eye with an empty adenovirus. Following single injections and 48-hr perfusion, trabecular meshworks were dissected and expression of TIGR/MYOC and NTE was assessed by relative quantitative PCR (Ambion). We used 18S ribosomal RNA as a control and examined its expression at a concentration similar to the gene under study by the use of a primer/competimer mix.
Data averaged from triplicate measurements of three separate sets of eyes show that increased expression of TIGR/MYOC is accompanied by upregulation of NTE. The 18S ribosomal RNA is not affected by TIGR/MYOC expression (Figure 4). Thus, regulation of transcription of NTE in human eyes exposed to TIGR/MYOC is affected in a similar manner as transcription of its Drosophila homolog in transgenic flies. These observations support our assessment that targeted overexpression of glaucoma-associated proteins in transgenic flies can serve as an advantageous system for the discovery of genes that are modulated in response to elevated pressure in the eye and may ultimately contribute to the development of glaucoma.
DISCUSSION
TIGR/MYOC-induced fluid discharge in the Drosophila eye: Research on the pathogenesis of glaucoma has been hampered by the lack of a suitable animal model. Only a few genes that are linked to glaucoma, including TIGR/MYOC (Adamet al. 1997; Stoilovaet al. 1997; Stoneet al. 1997; Alwardet al. 1998; Richardset al. 1998; Wiggset al. 1998), cytochrome P450 (CYP1B1; Stoilovet al. 1997; Bejjaniet al. 1998; Kakiuchi-Matsumotoet al. 2001), and most recently optineurin (Rezaieet al. 2002), were discovered as a result of genetic linkage studies. We have generated flies that overexpress TIGR/MYOC in their eyes. These flies have a striking phenotype: The ommatidia appear engorged and disrupted (Figure 3B) and fluid is periodically extruded from their eyes (Figure 2). This discharge appears to temporarily relieve the intraommatidial pressure allowing ommatidia to resume a near normal organization (Figure 3C).
TIGR/MYOC was discovered as a secreted protein overexpressed in response to dexamethasone in the human trabecular meshwork (Polanskyet al. 1997). TIGR/MYOC overexpression induces elevated intraocular pressure and glaucoma in ∼40% of individuals (Johnson 1997). Conversely, mutations in this gene, including a truncated form of the protein, are also linked to elevated pressure in patients with the disease (Rozsaet al. 1998). In humans, indirect evidence that TIGR/MYOC protects from or provokes elevated pressure has been reported (Fautschet al. 2000; Borráset al. 2002).
Changes in the transcriptome of flies overexpressing human TIGR/MYOC in their eyes
—Upregulation of the swiss cheese human ortholog NTE transcript in perfused postmortem anterior segments. One eye of each pair (n = 3) was injected with AdhTIG3 (treated) while each contralateral eye was injected with empty virus Ad5.CMV-Null (AdNull; control). Determinations of expression for each eye were performed in triplicate. (A) The ratio of expression values of the TIGR/MYOC (
) and 18S (▪) rRNA transcripts in the AdhTIG3-treated over AdNull control eyes (n = 9). (B) The ratio of expression values of the NTE (
) and 18S (▪) rRNA transcripts in the AdhTIG3-treated over AdNull control eyes (n = 9). Multiplex TIGR/MYOC and 18S amplifications were performed at 24, 27, and 25 cycles with 18S primer/competimer ratios of 1:15, 1:12, and 1:9, respectively. Multiplex NTE and 18S amplifications were performed at 36, 34, and 34 cycles. Representative gels for each of the transcripts are included. Treated eyes exhibiting an increase in TIGR/MYOC expression of 2.7 ± 0.3-fold showed an induction of NTE of 1.5 ± 0.15-fold. The 18S RNA expression did not change (1.2 ± 0.08-fold and 1.0 ± 0.08-fold, respectively).
Although we cannot measure intraocular pressure directly in the Drosophila eye, the most parsimonious explanation for the observed TIGR/MYOC-induced phenotype is an increase in ocular pressure, which can be relieved by extrusion of fluid. This could be due simply to the accumulation of extracellular material in the ommatidia. However, even if the cause for the phenotype observed in Drosophila is different from that responsible for the increased intraocular pressure observed in humans, it remains intriguing as it mimics some of its characteristics, such as the enlargement of eye structures and effect on fluid outflow.
The expression of the distinct phenotype in flies induced by overexpression of a glaucoma-linked gene may also result from the up- or downregulation of other genes and may help reveal hidden mechanisms relevant to the physiology of aqueous humor outflow and the development of glaucoma in humans. Thus, similar to the transcriptional alterations observed in the Drosophila eye, overexpression of TIGR/MYOC could modulate the expression of other genes to induce discharge of extra fluid in the human trabecular meshwork that could facilitate the flow of aqueous humor through Schlemm's canal.
Transcriptional profiling in the Drosophila eye as a discovery tool for genes that contribute to glaucoma: Our transcriptional profiling studies applied stringent criteria for the identification of up- or downregulated gene products as a consequence of TIGR/MYOC expression. Whereas our analysis limits the likelihood of false positives, it also prevents detection of subtle effects that could have important consequences. Furthermore, post-translational modifications can critically affect the phenotype, but would go undetected.
Nonetheless, we were able to identify 50 modulated transcripts, including several gene products previously implicated in glaucoma. Linkage studies of Saudi Arabian (Bejjaniet al. 1998), Japanese (Kakiuchi-Matsumotoet al. 2001), Slovenian (Plasilovaet al. 1999), and Indian (Panickeret al. 2002) populations have linked cytochrome P450 (CYP1B1) to congenital glaucoma. A role for aquaporin-4 in glaucoma has also been postulated. Aquaporin-4 is expressed in the ciliary body and inhibited by protein kinase C-mediated phosphorylation (Hanet al. 1998). It is of interest that we observed dramatic upregulation of protein kinase C 98E concomitant with substantial downregulation of the Drosophila aquaporin-4 homolog (Table 1). Inhibition of contraction of the iris through suppression of tyrosine kinase activity and the effects of protein kinase C on ciliary muscle contractility may also influence ocular outflow facility (Yousufzai and Abdel-Latif 1998; Wiederholtet al. 2000). Many studies have emphasized the importance of the integrity of the cytoskeleton for the function of trabecular meshwork cells (Clarket al. 1994; Epsteinet al. 1999). Glucocorticoid-induced alterations in the cytoskeleton may contribute to steroid-induced ocular hypertension (Clarket al. 1994). This is the first study in which multiple genes independently implicated in previous studies of glaucoma are shown to be coregulated as a consequence of overexpression of TIGR/MYOC and illustrates the power of our transcription profiling approach in Drosophila.
Whereas homologs of a subset of the up- or downregulated genes we identified in TIGR/MYOC-overexpressing flies may contribute to increased intraocular pressure in the human eye, clearly not all of them are relevant to human disease genes or have human homologs (e.g., salivary gland protein 8, Table 1). Ultimate validation of transcription profiling in Drosophila as a model for human disease requires demonstrating that candidate disease genes identified in Drosophila indeed have relevant human orthologs. Therefore, we performed parallel studies between Drosophila as a candidate gene discovery system and perfused postmortem human eyes as a validation system. We focused on the neurodegeneration gene swiss cheese (Kretzschmaret al. 1997), which has a well-defined human ortholog, NTE (Lushet al. 1998; Glynn 1999), not previously implicated in glaucoma. Swiss Cheese and NTE are involved in glia-neuron interactions (Kretzschmaret al. 1997; Lushet al. 1998; Glynn 1999). Introduction of TIGR/MYOC in perfused postmortem human eyes augments expression of NTE similar to the increase in swiss cheese in transgenic flies. The 50% increased expression of NTE is less dramatic than the 34-fold upregulation of its Drosophila counterpart, likely due to relative differences in efficiency of expression of TIGR/MYOC in transgenic flies and adenovirus transfected human eyes. Nevertheless, regulation of transcription of NTE in human eyes exposed to TIGR/MYOC is affected in a manner parallel to that of transcription of its Drosophila homolog in transgenic flies. Our observations underscore the value of Drosophila as a model system for the study of human disease and show that our approach can lead to the discovery of new susceptibility genes for glaucoma.
Acknowledgments
We thank J. Mahaffey for providing facilities for the generation of transgenic flies, E. Johannes for assistance with microscopy, M. Mattmüller for assistance with histology, and R. Wang and L. Rowlette for technical assistance. This work was supported by grants from the Glaucoma Research Foundation, the Research to Prevent Blindness Foundation (RPB), and the W. M. Keck Foundation and by National Institutes of Health grants GM-59469 (to R.R.H.A.), EY11906 and EY13126 (to T.B.), and GM-45146 and GM-45344 (to T.F.C.M.). T. Borrás is a Jules and Doris Stein RPB Professor Awardee.
Footnotes
-
Communicating editor: K. V. Anderson
- Received August 20, 2002.
- Accepted November 4, 2002.
- Copyright © 2003 by the Genetics Society of America
