Coordinated Changes in Classes of Ribosomal Protein Gene Expression Is Associated with Light-Induced Retinal Degeneration

Abstract

purpose. To identify genes with altered expression levels in the degenerating retina in a light-induced retinal degeneration (LIRD) model.

methods. Adult Sprague-Dawley rats were exposed to intense green light for 4 hours. After this treatment, the retinas were excised, RNA was extracted, and a cDNA library was prepared. The cDNA library was differentially cross-screened with probes representing 0-hour and 4-hour light-exposed rat retina. Transcripts with altered expression levels were sequenced and expression was confirmed by Northern blot analysis. Gene-specific primers were designed and used to examine the expression levels of other genes involved in protein synthesis. Promoter sequences of the ribosomal-binding protein (Rbp) genes were analyzed for transcription-binding sites.

results. Of the 10,000 clones that were initially screened, 41 exhibited altered expression levels. Six of these corresponded to five known Rbp genes. Six additional Rbp genes were also examined. In total, 9 of 11 Rbp genes exhibited an increase in expression levels in response to a 4-hour light exposure. In contrast, the transcript levels of elongation factor 1α1 and 18S rRNA did not increase. The most abundant transcription factor–binding sites conserved in the promoter regions of all Rbp genes examined in this study include AP-1, Oct-1, V-myb, USF, Pax-4, and the FOX family of transcription factors.

conclusions. The results indicate that light-induced retinal degeneration (LIRD) is associated with increased expression of specific Rbp genes. These Rbp genes may be involved in mediating visual cell loss in LIRD through a translational or an extraribosomal mechanism.

Light-induced retinal degeneration (LIRD) in rats leads to photoreceptor cell loss. ^{1 2} In the case of human retinal disorders such as retinitis pigmentosa (RP), the primary genetic lesion in different clinical forms vary, but the eventual outcome is always photoreceptor cell loss leading to blindness. In both human disease and LIRD, active cell death is believed to underlie the eventual loss of visual cells. ³ However, in contrast to RP, LIRD in rodents involves synchronous damage to numerous photoreceptors in a short period. The extent of visual cell loss in LIRD varies with age, genetics, diet, and prior light-exposure history of the experimental animal. ^{2 4 5} The precise cellular events that lead to the pathologic changes that underlie the process of retinal degeneration is not completely understood. 

The mechanism that leads to light-induced visual cell loss involves an oxidative-stress–mediated apoptotic pathway ^{1 4 6 7 8} that is initiated by the bleaching of rhodopsin. ¹ Pretreatment of animals with antioxidants such as ascorbate and dimethylthiourea (DMTU) has been found to ameliorate the effects of intense light treatment. ^{2 8 9 10 11 12 13} Apoptosis is an active process that requires altered gene and protein expression. ¹⁴ We and other research groups have already demonstrated that the retina responds to intense light exposure through alterations in the transcript levels of specific expressed genes. ^{15 16 17} These genes include clusterin, hemeoxygenase1, IRBP, c-fos, c-jun, the caspases, and RPE65. A systematic analysis of a series of knockout mice has revealed that the effects of LIRD are p53 independent ¹⁸ and appear to be either AP-1 or transducin dependent. ¹⁹ The observations of fragmented DNA and elevated levels of caspase expression in retinal tissue after intense green or blue light exposure suggest that a classic caspase cell death pathway is involved in the LIRD process. ^{16 20 21 22 23 24 25} In contrast, several studies have reported induction of retinal degeneration by a caspase-independent mechanism, suggesting that there may be more than one pathway that leads to LIRD. ^{26 27} For the most part, little is known about the precise molecular events that occur in the retina between the initial light-induced bleaching of rhodopsin and the final loss of photoreceptor cells in LIRD. 

In this study, we used a differential screening approach to analyze a cDNA library representative of intense-light–exposed rat retinas. The goal was to identify LIRD-associated genes that may define early events in LIRD. In our initial screening of 10,000 clones, we identified 41 that were representative of intense light-inducible or -repressible genes. Twelve of these genes (29% of all clones identified) are involved in the process of protein synthesis and six of these (15% of all clones identified) correspond to known ribosomal binding protein (Rbp) genes. Ribosomal proteins associate with either the large or small subunit of ribosomes and help regulate the process of protein synthesis. The eukaryotic ribosome is composed of four ribosomal RNA (rRNA) molecules and >80 Rbps. ^{28 29} Each Rbp gene appears to represent an evolutionarily conserved sequence. ³⁰ A general concept is that changes in the specific types of Rbps that interact with the ribosome may alter the mechanism by which the translational machinery functions. In addition to their roles in ribosome function, several Rbps are thought to be multifunctional. Several ribosomal proteins have been associated with DNA damage repair, cell cycle regulation, cell differentiation, oogenesis, development, and cell death. ^{30 31 32 33 34 35 36 37 38 39 40 41 42} Several of the genes examined in the present study, Rps3, ⁴⁴ Rps3a, ³¹ and Rpl7 ³⁶ have been associated with a role in apoptosis in nonretinal models of active cell death. Alterations in Rbp gene expression have been associated with various human diseases, including several different cancer states as well as age-related cataracts. ^{31 43}  

Our differential analysis of a light-exposed retina–derived cDNA library provides new evidence that select Rbp genes alter their levels of expression over the course of LIRD. In addition, changes in expression were found for genes coding for large and small ribosomal subunit protein genes, suggesting that alterations in both subunits may be important. Moreover, because of the multifunctional nature of some Rbp genes, it is uncertain whether it is the direct role that they have in protein synthesis or their alternate roles that have an integral impact on a retinal degenerative phenotype. Aspects of all these issues are discussed in light of our data. 

Materials and Methods

Animals

Weanling male albino Sprague-Dawley rats were obtained from Harlan, Inc. (Indianapolis, IN) and maintained in darkness until 60 days of age. The rats were then exposed to intense visible light for up to 16 hours. Light exposures were started at 9 AM and performed in green Plexiglas chambers (no. 2092; Dayton Plastics, Dayton, OH) transmitting 490 to 580 nm light ¹ with an illuminance of 1200 lux. Some rats were given the synthetic antioxidant dimethylthiourea (DMTU) at a dose of 500 mg/kg intraperitoneally, 24 hours before and again just before light treatment. Rats were killed in carbon dioxide–saturated chambers immediately after exposure to light and retinas were excised, flash frozen on dry ice, and stored at −70°C. For each treatment point in a single light-exposure profile (for example, 0, 4, 8, and 16 hours) retinas were pooled from at least five animals. The treatment profile was repeated with additional groups of animals three times. In all cases, animals were cared for in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

RNA Isolation and cDNA Library Synthesis

Total RNA from all samples was isolated (RNAzol; Tel-Test, Inc., Friendswood, TX) according to the manufacturer’s protocol. RNA was dissolved in diethyl pyrocarbonate (DEPC)-treated water and kept frozen at −70°C until use. Isolated RNA was run on a 1% agarose gel and visualized after ethidium bromide staining over UV light, to verify the integrity of the 28S and 18S rRNA in the sample. Poly(A⁺) mRNA was isolated from total RNA extracted from retinas taken from animals exposed to 4 hours of light (Oligotex mRNA Mini kit; Qiagen, Mississauga, Ontario, Canada). The poly(A⁺) mRNA was quantified and used for synthesis of a cDNA library. The cDNA library was constructed with a kit (ZAP cDNA Synthesis kit and a UNIZAP-XR/Gigapack III Gold Cloning kit) according to the manufacturer’s protocols (Stratagene, La Jolla, CA). The quality of the cDNA synthesized was confirmed by gel electrophoresis and autoradiography. Size fractionation of the cDNA was performed to ensure that most of the clones generated would be at least 500 bp in length. 

Probe Synthesis

Total cDNA was synthesized from purified mRNA, from the unexposed and 4-hour light-exposed retina, using a standard cDNA synthesis kit (Invitrogen-Gibco, Gaithersburg, MD). cDNA was radiolabeled with [α³²P] dCTP, dGTP, and dATP by random oligo-nucleotide labeling. Clones representing Rpl7, Rpl12, Rpl19, Rpl30, and Rps16, which were used as probes, were isolated from the cDNA library. Clones representing rat Rpl5, Rps2, Rps3, Rps3a, and Rps10 were purchased from Research Genetics (Gaithersburg, MD). A clone representing human Rps28 was isolated in a separate series of experiments (data not shown). Insert DNA from each clone was amplified with T7 and T3 primers that flanked the insertion site and was radiolabeled with [α³²P] dCTP using a standard random oligo-nucleotide labeling reaction. DNA fragments for elongation factor (EF) 1α1, 18S rRNA, and actin were synthesized using RT-PCR and the primers listed in Table 1 . RT PCR was performed with a kit according to the protocol provided (One Step RT-PCR; Invitrogen, Carlsbad, CA). 

cDNA Library Screening

Duplicate plaque lifts of the plated cDNA library were made as previously described. ⁴⁵ Plaque lifts were prehybridized (Hybrisol II; Chemicon International, Temecula, CA) for 4 hours at 65°C. For each pair of duplicate lifts, one was hybridized with radiolabeled cDNA probe representing the unexposed retina in the solution at 65°C overnight and the second was hybridized with the radiolabeled cDNA probe representing the 4-hour light-exposed retina under the same conditions. The plaque lifts were then washed twice at 65°C in 2× SSC for 15 minutes, once in 2× SSC and 0.1% SDS for 30 minutes, and once in 0.1× SSC and 0.1% SDS for 10 minutes. Autoradiography was performed at −70°C (X-OMAT film; Kodak, Rochester, NY) between two intensifying screens. Visual inspection of the developed autoradiographs led to the identification of clones showing differential hybridization patterns with the two cDNA probes in question. These clones were isolated and transferred into 500 mL of SM buffer with 10% chloroform and stored at 4°C. ⁴⁶ The cDNA inserts from the differentially selected clones were PCR amplified using T7 and T3 primers and arrayed onto nylon membranes (Genescreen Plus; Dupont-NEN, Boston, MA). Arrays were screened as described for the plaque lifts as a secondary screen. 

Sequence Analysis of Differentially Expressed Clones

PCR amplified cDNA inserts from the differentially selected clones were purified (QIAquick Gel Extraction kit; Qiagen). A DNA sequencing kit (Prism; Applied Biosystems [ABI], Warrington, UK) and either SK or T7 primers were used for single-pass automated sequencing, according to the manufacturer’s protocol. Sequencing reactions were electrophoresed on an automated sequencer (model 377; ABI). DNA sequences were analyzed with the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/; provided in the public domain by the National Center for Biotechnology Information [NCBI], Bethesda, MD). 

Northern Blot Analysis

Total RNA was electrophoresed in 1.2% agarose gels in a standard formaldehyde running buffer system. Northern transfer onto nylon membrane (Genescreen Plus; Dupont, NEN) was performed by passive blotting. Prehybridization and hybridization with radiolabeled DNA probes were performed as described earlier. In all cases, the expression profile of each gene examined was verified on Northern blot analysis made from a different set of mRNA samples extracted from independent sets of animals subjected to the same treatment profile as the animals represented on the initial blot screened. Densitometry of distinct bands was performed on each autoradiograph using the histogram function of image-analysis software (Photoshop, ver. 5; Adobe Systems, Mountain View, CA). A normalized value was calculated as follows: net intensity of the band minus the net intensity of the background divided by the net intensity of the RNA loading control minus the net intensity of the corresponding background. To ensure that data sets derived from independent animal sets and Northern analyses were comparable, the normalized value for retinas without any light exposure was set at 1. All statistical calculations were made on computer (Excel 97; Microsoft, Redmond, WA) and all averages were calculated using three independently determined values (n = 3). 

In Silico Promoter Analysis

Full rat gene sequences were obtained directly from GenBank (http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). They included Rpl7: XM_216318, chromosome 5 NW_043815; Rpl30: NM_022699, chromosome 7 NW_047778; Rps3a: NM_017153, chromosome 2 NW_043532; Rpl5: NM_031099, chromosome 14 NW_047419; Rps2: NM_031838, chromosome 10 NW_042648; Rps10: NM_031109, chromosome 20 NW_043430; Rpl19: NM_031103; chromosome 10 NW_042672; Rps3: XM_214997, chromosome 1 NW_043387; Rps16: XM_341815, chromosome 1 NW_043361; Rpl12: XM_216039, chromosome 3 NW_043713; and Rps28: chromosome 7 NW_047773). For all genes, the transcription start site (TSS) was considered to be the 5′ end of the gene (+1) and a 4.5-kbp region upstream of the TSS was retrieved. The 4.5-kbp upstream region of each gene was analyzed for known transcription factor (TF) binding motifs, using the MATCH program (http://www.bioinfo.de/isb/gcb01/poster /index.html) provided by Biobase Biological Databases (Wolfenbuttel, Germany; http://www.gene-regulation.com/). The following search parameters were used: (1) group of matrices, vertebrates; and (2) high-quality matrices only, with the cutoff selection for matrix group set to minimize the sum of both the false positives and false negatives. Results obtained were transferred to disk (Excel 97; Microsoft) for further analysis with respect to quantifying the number of TF DNA-binding motifs present. 

Results

Differentially Expressed Genes Involved in Protein Metabolism

Ten thousand clones were examined in the initial plaque lift screens. Secondary screening followed by sequencing and sequence analysis identified 31 clones with identities to known rat genes, 7 novel rat clones with sequence similarities to genes defined in other species, and 3 novel uncharacterized rat genes. In total, 41 putative 4-hour light-inducible, differentially expressed genes were identified (Table 2) . Five of the clones represented known retinal enriched genes such as opsin, rod transducin and S-antigen (arrestin), which we would expect to be differentially expressed in our system. The clones identified could be placed into several large functional gene categories based on sequence similarity to known genes (Table 2 ; Fig. 1 ). The largest group of genes falls into those involved in protein metabolism. Within this group, 12 clones corresponded to transcribed sequences that had relevance to the translational machinery. Of these 12 clones, 1 corresponded to EF 2, 5 corresponded to the external transcribed spacer sequence of 18S ribosomal RNA, and 6 represented five different Rbp genes (Rpl19, Rpl12, Rpl30, Rps16, and Rpl7). 

18S rRNA Levels during LIRD

Analysis of 18S rRNA levels by ethidium bromide staining is a well-accepted RNA loading control for Northern blot analysis of total RNA. ⁴⁷ 18S rRNA, however, is a component of the protein synthesis machinery. Given the observation that 12 of the 41 clones isolated are related to protein synthesis led us to question the appropriateness of using 18S rRNA as the loading control. Typically, we load approximately 5 μm of total RNA per lane on an agarose gel and after electrophoresis we invariably found that ethidium bromide staining of the samples gave a uniform 18S rRNA profile (Fig. 2 , top). We also generated a DNA fragment based on the known ribosomal 18S rRNA nucleotide sequence and used it to screen the Northern blot of the gel shown in the top blot (Fig. 2 , middle). The resultant autoradiograph shows uniform labeling of the 18S ribosomal band. The Northern blot was stripped and rescreened with a probe for β-actin (Fig. 2 , bottom). The resultant autoradiograph also showed evidence of a uniform level of β-actin hybridization to the blot. Ethidium bromide staining of 18S ribosomal RNA, radioactive probing for 18S RNA, or radioactive probing for β-actin are therefore all usable markers to control for uniform loading of total RNA samples when examining LIRD. Because of the functional nature of the genes we examined in the present study, we decided to use β-actin as the loading control for subsequent experiments, as it does not have a known role in the protein synthesis machinery. 

Effect of Light-Induced Oxidative Stress on Expression of Rpl7

Two independently isolated clones from our cDNA library screen were found to represent the Rbp gene Rpl7 (Table 2) . Given that only 41 clones were isolated from a screening of 10,000 clones, the isolation of two clones representing the same gene is significant and suggests that Rpl7 may be highly expressed after a 4-hour exposure to light. To test this, we used the clone for Rpl7 as a probe to screen a light-treatment Northern profile (Figs. 3A 3B) . Rpl7 levels were relatively low in unexposed retina and showed a marked increase after a 4-hour exposure to light, followed by a decrease to barely detectable levels after 8 hours of light. To test whether the induction of the Rpl7 transcript level was a result of general exposure to light, or represented a light-induced oxidative stress mediated response, we examined the effect of DMTU, a synthetic antioxidant, on Rpl7 transcript levels after 4 hours of light treatment. Pretreatment of rats with DMTU before exposure to light prevented the increase in Rpl7 levels after 4 hours of exposure to intense light (Figs. 3C 3D) . This demonstrates that the increase in Rpl7 levels without antioxidant pretreatment is a light-induced, oxidative-stress–mediated response. 

Expression of Other Ribosomal Proteins in Response to Intense Light

Clones for five different Rbp genes were identified in our initial screening (Rpl7, Rpl12, Rpl16, Rpl19, and Rpl30). Four of these genes were shown to be differentially expressed by Northern blot analysis (Figs. 4 5A) , indicating that our differential screen was successful. To obtain a broader perspective on the nature of Rbp mRNA expression over the course of LIRD, additional clones for other Rbp genes (Rps2, Rps3, Rps3a, Rpl5, Rps10, and Rps28) were examined by Northern blot analysis (Figs. 4 5A) . Of the 11 ribosomal proteins examined, 9 showed relatively high levels in one or more of the exposure times used in the course of LIRD. There was a change in ribosomal protein gene expression that is specific for the large ribosomal subunits (Rpl7, Rpl16, Rpl19, Rpl30, and Rpl5) as well as the small ribosomal subunits (Rps10, Rps2, Rps3, and Rps3a), suggesting that changes occur in genes relevant to both subunits. 

In the unexposed retina, three of the nine genes (Rps3, Rps16, and Rpl19) were expressed at modest levels (Fig. 4C) . After a 4-hour exposure to light Rpl7, Rpl30, Rps3a, Rps10, Rps2, Rpl5, Rps16, Rpl19, and Rps3 transcript levels were all elevated, suggesting a clear induction of gene expression (Fig. 4) . Careful inspection of the profiles after increasing durations of exposure to light revealed four distinct profiles or trends of expression (Figs. 4 5A) . In one case, the Rbp genes in question showed a marked increase in transcript levels after 4 hours of light with marked decrease after 8 hours, and very low levels after 16 hours (Fig. 4A) . In the second distinct profile, there was a marked increase in transcript levels after a 4-hour exposure to light followed by a modest decrease in levels throughout the profile examined (Fig. 4B) . In the third scenario, the marked increase in transcript levels after 4 hours of light is followed by little change in transcript levels during the remainder of the profile examined (Fig. 4C) . 

In contrast, two Rbp genes (Rps28 and Rpl12) did not appear to be highly expressed in the retina before or after exposure to intense light (Fig. 5A) . In both cases levels were low enough that the differential nature of expression of the two genes could not be determined. We also examined the expression profile of EF 1α1, another component of the protein synthesis machinery, over the course of LIRD (Figs. 5B 5C) . In contrast to the profiles in the Rbp genes, EF 1α1 transcript levels showed an immediate decrease after a very brief exposure to light (2 hours), supporting the notion that not all components of the translation machinery respond the same way during LIRD. 

In Silico Promoter Analysis

The promoter region of the Rbp genes under study was examined to identify abundant TF binding sites that may play a role in regulating Rbp gene expression in the degenerating retina. A crude analysis of these regions revealed that the promoters of the 11 Rbp genes under study are defined by 106 different TF DNA-binding motifs. Seventeen of these sites are conserved in every gene examined in this study (Table 3) . These DNA-binding sites include those corresponding to the following DNA-binding factors: FoxD3, AP-1, v-Myb, Pax-4, USF, Oct-1, HNF-3b, CDP CR1, HNF-4, FoxJ2, Evi-1, Lmo2 complex, c-Rel, GATA-1, Comp-1, Nkx2-5, and Hand1/E47. An explanation of TF abbreviations is given in Table 3 . The six most abundant categories of TF binding sites found in the promoter regions of the 11 genes in question are FOX family binding factors as a group and AP-1, Oct-1, v-Myb, USF, and Pax-4. These six categories of DNA-binding factor motifs constitute 41% of all motifs found collectively in these genes within 4.5 kB upstream of the TSS. 

Foxd3, a member of the Fox family of TFs, has binding sites present in the promoter region of all 11 genes. Other members of the Fox family of TFs (Fox-J2, Freac-7, Freac-3, Freac-2, HFH-3, HFH-8, HFH-1, and HNF-3β) are also present in the promoter regions of some of the genes surveyed. Overall, the DNA-binding motifs for members of the Fox family of genes are the most abundant motifs in the 11 genes surveyed. A visual inspection of the TF-binding sites of each gene revealed that the Fox motifs tend to cluster in two distinct patterns. The promoter regions of all the Rbp genes examined are characterized by areas in which there are a series of overlapping Fox motifs embedded into the sequence (Fig. 6A) . In the case of 7 of the 11 genes examined (Rps3a, Rps10, Rps16, Rpl19, Rps3, Rpl12, and Rps28) specific regions of Fox motifs took on a more structured arrangement with a minimum of three Fox motif loci that were spaced from 8 to 30 nucleotides away from each other (Fig. 6B) . Both patterns of motif clustering would encourage the binding of at least one Fox TF to the DNA at those sites and suggests that Fox-related TFs may play a major role in the regulation of ribosomal protein gene expression. An example of these clustering patterns is shown for Rps10 and Rpl19 (Fig. 6) . 

Discussion

The molecular mechanism that underlies LIRD is not well understood. We are attempting to define a gene expression signature that is representative of the process. In albino rats that have been reared in the dark, over 80% of the photoreceptor cells are lost in response to 4 hours of intense light followed by a 2-week dark recovery period. ⁶ The 4-hour exposure to light is not sufficient to cause an immediate morphologic effect on the retina, and photoreceptor-enriched genes such as rhodopsin and IRBP are still at relatively high normal levels directly after exposure to light. ^{15 21} Oxidative-stress–inducible genes, such as hemeoxygenase-1 or clusterin, are not induced and DNA fragmentation is not noticeable directly after 4 hours of exposure to light. ^{11 15 48} The observations therefore suggest that 4 hours of light, at the intensity we used, is sufficient to direct the cells toward an apoptotic fate but does not immediately force them to manifest an active cell death phenotype. In our study, we found that a large number of genes isolated from a differential screen of a 4-hour light-exposed rat retinal cDNA library represented distinct components of the protein synthesis machinery. The apoptotic process is thought to be an active one, requiring gene and protein expression for the manifestation of active cell death. ⁴⁹ In general, pretreatment of cells with RNA and protein synthesis inhibitors blocks cell death, suggesting that both transcription and translation are necessary for the apoptotic process to occur. ¹⁴  

In our initial study examining Rpl7 expression during LIRD, we found a very sharp induction of transcript levels after 4 hours of light followed by a sharp decrease in levels after 8 hours. Pretreatment of animals, with the antioxidant DMTU, before a 4-hour exposure to light successfully suppressed the observed Rpl7 induction (Fig. 3) . Our conclusion is that the sharp induction in Rpl7 transcript expression is due to the effects of the intense light, which is known to generate an oxidative stress environment, and not simply from light alone. An analysis of 10 other Rbp genes established that not all Rbp genes were highly expressed after exposure to intense light. Both retinal Rps28 and Rpl12 are not highly expressed in untreated animals or in light-treated animals. In contrast to the profiles of Rbp gene expression, 18S ribosomal RNA levels are high and unchanged throughout the LIRD profile and EF 1α1 transcript levels show an immediate decrease from normal levels after 2 and 4 hours of light. Like the Rbps, both 18S ribosomal RNA and EF 1α1 are components of the protein synthesis machinery. Hence, the alteration in transcript levels of the nine Rbp genes that showed an induction represents a specific response to the effects of intense light. 

The induction of Rbp expression after 4 hours of light treatment could be associated with their role in translation. An involvement of this machinery in apoptosis is not unexpected, as the apoptotic process necessitates new gene and protein expression. Eventually, however, the protein synthesis machinery also becomes a target as apoptosis proceeds; thus, the induction of Rbp transcript levels after light treatment could be a response to the cell’s increased demand for protein synthesis as the apoptotic program is initiated. Inhibition of protein synthesis could eventually be mediated by caspase-directed inactivation of a number of different EFs central to the initiation of translation. ^{50 51} Alternatively, as part of the ribosomal machinery, ribosomal proteins could be involved in regulating which transcripts are translated into proteins. Therefore, variation in Rbp expression could alter the balance of anti- and proapoptotic factors that are translated and have a direct role in defining a cell death fate. ⁵²  

An increased demand for translation usually results in an increased efficiency of translation rather than an increase in the number of ribosomes. ³⁵ Therefore, if ribosomal protein gene expression is induced solely in response to an increased demand for translation, the induction of EF 1α1 protein and 18S rRNA would also be expected, because ribosome assembly occurs in a stoichiometric manner. ²⁹ Our observations suggest that changes in a subset of retinal Rbp genes by exposure to intense light is a specific alteration in expression and not a general phenomenon that affects all genes of the protein synthesis machinery. By far the largest change we detect is after a 4-hour exposure, an exposure that does not immediately result in signs of extensive oxidative stress or signs of photoreceptor cell dysfunction. ¹⁵  

Changes in Rbp gene expression have not been previously associated with retinal degeneration; however, they have been reported in various cancerous states, ^{2 8 32 35} abnormal blood cell differentiation, ³⁹ and Turner syndrome. ⁵³ In addition, select changes in specific large ribosomal subunit protein genes (Rpl21, Rpl15, RPL3a, and Rpl7a) have recently been associated with age-related cataracts, ⁴³ suggesting that pathways mediated by Rbps may be involved in maintaining lens transparency. An intriguing observation is the finding that some Rbps may be multifunctional in nature. Several ribosomal proteins have been associated with DNA damage repair, cell cycle regulation, cell differentiation, oogenesis, development, and cell death. ^{30 31 32 33 34 35 36 37 38 39 40 41 42} Currently, there is some question as to whether these genes initially coded for proteins explicitly involved in ribosome function, or whether they represented preexisting genes, responsible for other cellular functions, and were evolutionarily recruited into a role in protein synthesis. ^{30 41}  

Of the genes we studied, the strongest candidates for participating in extraribosomal roles during retinal degeneration were Rpl7, Rps3a, and Rps3. Extraribosomal functions have been documented for all three proteins in other model systems for studying active cell death. Rps3 has been shown to have an oxidative DNA damage repair capacity and therefore may provide a means of survival for a cell placed in an oxidative stress environment. ⁴⁴ In contrast, Rps3a expression has been associated with the induction of apoptosis in NIH 3T3-derived cells. ³¹ Rpl7 has a basic leucine zipper motif and can function as a coactivator of nuclear receptors. ⁵⁴ Constitutively high Rpl7 levels in Jurkat cells is associated with increased cell sensitivity to apoptotic stimuli. ³⁶ A function of Rpl3, Rpl3a, or Rpl7 in apoptosis, cell death susceptibility, or cell death sensitivity could all be significant in the process of retinal degeneration. However the association of the six additional Rbps, which do not have any known independent apoptotic functions, with the LIRD process implicates the ribosome as a possible mediator of cell death. 

To examine how gene expression may be regulated, common regulatory elements can be identified by examining the promoter region of each gene. ⁵⁵ Analysis of the promoter region revealed that all 11 of the Rbp genes in question are characterized by a very similar set of TF-binding motifs. The most abundant of these putative TF binding sites are for AP-1, Oct-1, v-Myb, Pax-4, USF, and Fox binding factors. The activities of both AP-1 and Oct-1 TFs have been documented during LIRD in mice. ⁵⁶ AP-1 DNA binding increases and Oct-1 DNA binding decreases after exposure to intense light. ⁵⁷ Oct-1 is a ubiquitous TF of the POU-homeo family of proteins, that is involved in regulating housekeeping genes such as small nuclear RNA genes. ⁵⁸ The myb gene family consists of three members: A-myb, B-myb, and C-myb. V-myb arises as a truncated form of the C-myb protein. The three members of the myb gene family have high homology in their DNA-binding domains, and bind DNA with overlapping sequence specificities. ⁵⁹ A-myb is believed to be involved in the proliferation and differentiation of neurogenic cells and is expressed in neural tube, hindbrain, olfactory epithelium, and neural retina during mouse embryogenesis. ⁶⁰ V-myb induces proliferation of chicken neuroretina cells in vitro. ⁶¹  

The presence of USF and Pax-4 has not been examined in the retina. Hence, their function in retinal cell loss is unknown. The abundance of binding sites for these TFs in the Rbp genes examined suggests that they may be involved in regulating these genes. The upstream stimulatory TFs 1 and 2 (USF) are ubiquitous helix-loop-helix TFs. USFs are expressed in the brain, and recent research suggests that the activate of USFs may be induced by calcium influx into neurons. ⁶² The paired-homeobox TF Pax- 4 has been studied most extensively in pancreatic beta cells and is involved in islet development in the pancreas. ⁶³ Pax-4 is thought to have an effect on gene expression by inhibiting Pax-6 DNA binding, either through competition for DNA-binding sites or protein–protein interactions. ⁶³ Although Pax-4 has not been identified in the retina, both Pax-6 and -2 are involved in retinal development and specific eye disorders. ^{65 66} The predominant class of TF DNA-binding motifs found in all Rbp genes we examined corresponded to sites recognized by TFs belonging to the FOX family of genes. Of the 30 members of the FOX protein family, ⁶⁷ three were found consistently throughout the promoter regions of the ribosomal-binding protein genes examined. These include: FoxD3, FoxJ2, and HNF-3B (FoxA2). Many of these FOX TF DNA-binding sites appear in clusters that are separated by <30 bp within a given putative Rbp promoter region or as complicated overlapping networks of FOX TF DNA-binding sites (Fig. 6) . In either case, the repetitive nature of these motifs would encourage interaction of Fox-related TFs should they be expressed. ⁶⁸ Mutations in several members of the Fox TF gene family (FoxC1, FoxE3, FoxL2, and FoxC2) are known to underlie anterior eye-related disorders. ^{69 70 71 72 73} Mutations in Fox genes have not been associated with human retinal degenerative disease, although FoxG1, FoxD1, and Foxn-4 are all known to be involved in retinal development. ^{74 75} The most abundant DNA-binding motif found in the genes under study corresponds to the site recognized by FoxD3, a TF involved in establishing the neural crest lineage during development. ⁷⁶ FoxD3 expression, however, has not been reported in retinal tissue. 

The current report is the first that associates changes in the expression of specific Rbp genes with the process of retinal degeneration. Analysis of the putative promoter regions of each of the Rbp genes examined led to the identification of TF-binding factors that may be relevant in defining the process of LIRD but that have not been studied in the retina before, thus opening new avenues for study. Although we have identified several TF motifs that are present in all the Rbp genes examined in this study, we have yet to explore the essential differences in the promoter regions of these genes that mediates four different profiles of Rbp expression (Figs. 4 5) . We believe that the induction of Rpl7, Rpl16, Rpl19, Rpl30, Rpl5, Rps10, Rps2, Rps3, and Rps3a during LIRD represents a specific response to exposure to intense light and represents a gene expression signature that marks the progression of LIRD. Further work is needed to determine whether changes in the expression of these genes affect translation and whether this in turn has an effect on the LIRD phenotype, or whether it is the alternate multifunctional nature of some of these genes that mediates LIRD. Moreover, whether or not such changes in gene expression precede or occur during visual cell death in retinal degenerative disorders such as RP remains to be elucidated. 

²  Contributed equally to this work and therefore should be considered equivalent senior authors.

 Supported by the National Sciences and Engineering Research Council (PW, MP, JS, TE, RG); RP Foundation Fighting Blindness (PW); Alberta Ingenuity (RG); National Eye Institute Grant EY01959 (DTO); and Mary Petticrew, Springfield, OH (DTO).

 Submitted for publication March 31, 2004; revised July 16, 2004; accepted July 23, 2004.

 Disclosure: R. Grewal, None; J. Stepczynski, None; R. Kelln, None; T. Erickson, None; R. Darrow, None; L. Barsalou, None; M. Patterson, None; D.T. Organisciak, None; P. Wong, None

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

 

Table 1.

 

Gene-Specific Primers

Gene	Forward Primer	Reverse Primer
Rat EF 1α1	5′-CCGGCCACCTGATCTACAAATGT-3′	5′-GGGGCCATCTTCCAGCTTCTTAC-3′
Rat 18S rRNA	5′-CGCGGTTCTATTTTGTTGGT-3′	5′-AGTCGGCATCGTTTATGGTC-3′
Rat β-actin	5′-AGCCATGTACGTAGCCATCC-3′	5′-CTCTCAGCTGTGGTGGTGAA-3′

Table 2.

 

Summary of Identified Clones in an LIRD cDNA Library Screen

Clone Name	Clone Number	DbEST ID	GenBank Accession Number	Search of NR Database Best Match	S	GenBank Accession Number	Function: Descriptor 1	Function: Descriptor 2
IGA	1084	19544765	CF273483	Otx2 (homeobox domain)	R	XM_224009	DNA-transcription	Transcription factor
IGA	1102	19544766	CF273484	Enhancer of split homologue (R-esp1)	R	L14462	DNA-transcription	Transcription factor
IGA	130C	19544767	CF273485	Sequence-specific single-stranded-DNA-binding	R	AF121893	DNA structure	DNA binding
IGA	2269A	19544768	CF273486	* Nucleosome assembly protein 1-like 1 (Nap1/1)	M	NM_015781	DNA structure	DNA binding
IGA	716	19544769	CF273487	Cytochrome c oxidase Via	R	X72757	Energy Metabolism	Mitochondria
IGA	25	19544770	CF273488	ATP synthase subunit 8 and 6 gene	R	AF115771.1	Energy Metabolism	Mitochondria
IGA	1157	19544771	CF273489	Cytochrome c oxidase subunits I-III, and ATPase subunit 6	R	M27315	Energy Metabolism	Mitochondria
IGA	130B	19544773	CF273491	Coupling factor 6 of mitochondrial ATP synthase complex	R	X54510	Energy Metabolism	Mitochondria
IGA	1137	19544772	CF273490	Triosephosphate isomerase	R	L36250	Energy Metabolism	Pentose phosphate shunt
IGA	325	19544774	CF273492	* External transcribed spacer B2 element-18S (ETS)	M	X56974	Protein-translation	ETS
IGA	959	19544776	CF273494	* External transcribed spacer B2 element-18S (ETS)	M	X56974	Protein-translation	ETS
IGA	1530	19544779	CF273497	* External transcribed spacer B2 element-18S (ETS)	M	X56974	Protein-translation	ETS
IGA	1551	19544780	CF273498	* External transcribed spacer B2 element-18S (ETS)	M	X56974	Protein-translation	ETS
IGA	736A	19544783	CF273501	External transcribed spacer B2 element-18S (ETS)	R	X56974	Protein-translation	ETS
IGA	569	19544775	CF273493	Ribosomal protein L30	R	NM_022699	Protein-translation	Ribosomal protein
IGA	986	19544777	CF273495	Ribosomal protein L19	R	J02650.1	Protein-translation	Ribosomal protein
IGA	1287	19544778	CF273496	Ribosomal protein S16	R	X17665	Protein-translation	Ribosomal protein
IGA	736B	19544784	CF273502	Ribosomal protein L7	R	M17422	Protein-translation	Ribosomal protein
IGA	830B	19544785	CF273503	Ribosomal protein L12	R	X53504	Protein-translation	Ribosomal protein
IGA	M8	19544786	CF273504	Ribosomal protein L7	R	M17422	Protein-translation	Ribosomal protein
IGA	1755A	19544782	CF273500	EF2	R	Y07504.1	Protein-translation	Translation inhibitor
IGA	1362A	19544781	CF273499	Similar to Proteasome inhibitor PI31 subunit	R	XM_230718	Protein-degradation
IGA	12	19544787	CF273505	Retina S-antigen (arrestin)	R	X51781	Signaling pathway	Vision
IGA	1005	19544789	CF273507	* Guanine nucleotide binding protein, α transducin	M	NM_008140	Signaling pathway	Vision
IGA	1A	19544793	CF273511	* Guanine nucleotide binding protein, α transducin	M	NM_008140	Signaling pathway	Vision
IGA	2A	19544794	CF273512	* Opsin gene	M	BC031766	Signaling pathway	Vision
IGA	3A	19544795	CF273513	Opsin gene	R	U22180	Signaling pathway	Vision
IGA	767	19544788	CF273506	Guanine nucleotide binding protein G-O, α subunit	R	M17526	Signaling pathway
IGA	1114	19544790	CF273508	* PTK7 protein tyrosine kinase 7 (Ptk7)	M	NM_2821	Signaling pathway
IGA	1095A1	19544791	CF273509	Neonatal glycine receptor	R	X57281	Signaling pathway
IGA	1095A2	19544792	CF273510	* Similar to mitogen-activated protein kinase 8 interacting protein 2	M	NM_021921	Signaling pathway
IGA	M7	19544796	CF273514	GTP-binding protein (G-alpha-O)	R	M17526	Signaling pathway
IGA	219	19544797	CF273515	Alpha-A-crystallin	R	U47922	Stress-chaperone	Protein folding
IGA	246	19544798	CF273516	BetaA2-crystallin (Cryba2	R	NM_173140	Stress-chaperone	Protein folding
IGA	1014	19544800	CF273518	Beta-B3-crystallin	R	X05899	Stress-chaperone	Protein folding
IGA	1212A	19544801	CF273519	Alpha-A-crystallin	R	U47922	Stress-chaperone	Protein folding
IGA	830A	19544802	CF273520	* BetaB1-crystallin (Crybb1)	M	AF106853	Stress-chaperone	Protein folding
IGA	748	19544799	CF273517	Gamma-b3-glutathione-S-transferase	R	J02744	Stress-antioxidant	Stress
IGA	17	19544803	CF273521	* RIKEN cDNA 4930583H14 gene	M	BC049556	Unknown
IGA	577	19544804	CF273522	* RIKEN cDNA D430030K24 gene	M	NM_178732	Unknown
IGA	705	19544805	CF273523	Brain ID transcript (BC1 RNA)	R	U25484	Unknown

 S, species; R, Rattus norvegicus; M, Mus musculus,

*  , Similar to.

Figure 1.

 

Clones identified by general classification in LIRD.

Figure 2.

 

Northern blot analysis of 18S ribosomal RNA and actin levels over the course of LIRD (0-, 4-, 8-, and 16-hour exposure to intense green light). Top: a formaldehyde gel, stained with ethidium bromide, shows the light exposure Northern blot. Middle and bottom: autoradiographs of the Northern blot probed with 18S and actin, respectively.

Figure 3.

 

Northern blot analysis of Rpl7 over the course of LIRD (0-, 4-, 8-, and 16-hour exposure to intense green light; A) and after DMTU treatment and exposure to light (C). The actin profile is provided as an RNA loading control. Representative results are shown. (B, D) Average and standard deviations for three independent experiments. The general order of the data depicted in (C) has been changed in (D) to illustrate that DMTU treatment to the retina alone (no exposure to light) did not cause any significant changes in Rpl7 mRNA levels. DMTU pretreatment of animals subjected to a 4-hour exposure to light significantly suppressed the induction of Rpl7 expression that was present without the antioxidant pretreatment, implying that the induction in Rpl7 levels is dependent on light-induced oxidative stress.

Figure 4.

 

Northern blot analysis of eight additional intense light-inducible ribosomal-binding protein genes over the course of LIRD. Three distinct profiles were evident: (A) Rpl7, Rpl30, and Rps3a; (B) Rps10, Rps2, and Rpl5; and (C) Rps16, Rpl19, and Rps3. The main differences in the profiles lie in the relative rate at which levels decreased after the initial increase observed after a 4-hour exposure to light. For each gene examined graphs depicting the average mRNA levels in three independent determinations are also shown.

Figure 5.

 

Northern blot analysis of components of the translational machinery that do not markedly increase their levels of expression over the course of LIRD (A) Rpl12, Rps28 and (B) EF 1α1. Whereas Rpl12 and Rps28 did not appear to be highly expressed at any point in the expression profile examined, EF 1α1 showed a decrease in mRNA levels after immediate exposure to intense light. For EF 1α1 a graph depicting the average mRNA levels in three independent determinations is shown (C).

Table 3.

 

Distribution of Transcription Factor Binding Sites That Are Conserved in the Ribosomal Binding Protein Genes Examined in the Study

TF Motif	Rpl7	Rp30	Rps3a	Rpl5	Rps2	Rps10	Rpl3	Rps19	Rps6	Rpl12	Rps28	Total
FoxD3	9	16	29	10	17	37	18	28	37	18	55	274
AP-1	9	10	18	8	22	18	14	12	15	18	6	150
v-Myb	7	20	9	16	17	11	12	11	12	12	11	138
Pax-4	19	12	13	9	18	8	9	14	12	14	9	137
USF	9	10	20	12	8	12	16	17	7	7	11	129
Oct-1	10	18	6	14	4	17	15	5	10	10	11	120
HNF-3B	3	7	19	7	2	15	12	12	11	8	18	114
CDP CR1	9	14	10	6	8	7	7	12	8	11	5	97
HNF-4	7	3	12	8	6	10	12	8	9	8	12	95
Fox-J2	1	3	14	4	1	10	10	7	2	3	12	67
Evi-1	3	13	6	8	1	8	10	4	3	1	7	64
Lmo2 complex	3	2	1	3	6	22	2	2	2	3	6	52
c-Rel	8	6	5	3	5	5	1	4	3	7	4	51
GATA-1	1	5	5	5	3	4	6	6	3	2	6	46
Comp-1	6	4	1	7	5	2	3	3	4	3	2	40
Nkx2-5	2	2	5	2	6	2	6	2	5	2	5	39
Hand 1/E47	3	2	1	3	1	2	1	1	2	1	2	19

 Data are the number of transcription factor binding sites in the ribosomal binding proteins. FOXD3, forkhead box class D3; AP-1, Activator protein 1; v-Myb, Oncogene of avian myelloblastosis virus; PAX-4, Paired box 4; USF, Upstream stimulatory factor; Oct-1, Octamer binding protein 1; HNF-3B, Hepatocyte nuclear factor 3 beta; CDP CR1, CCAAT displacement protein cut repeat 1; HNF-4, Hepatocyte nuclear factor 4; FOXJ2, Forkhead box class J2; Evi-1, Ecotropic virus integration site 1; LMO2 complex, Lim domain protein 2; c-Rel, Cellular homolog of the Onc sequence in the avian retrovirus reticuloendotheliosis virus strain T; GATA-1, A/G GATA A/G; COMP-1, Cartilage oligomeric matrix protein 1; Nkx2-5, NK2 homeobox protein 2-5; HAND 1/E47, Heart autonomic nervous system neural crest derivatives 1/E47.

Figure 6.

 

Two different patterns of Fox TF motif clustering were found in the promoter regions of the Rbp genes studied. Shown are representative examples of the two different patterns detected.