|Molecular Vision 2005;
Received 26 May 2005 | Accepted 16 November 2005 | Published 23 November 2005
Global gene profiling reveals novel glucocorticoid induced changes in gene expression of human lens epithelial cells
Vanita Gupta,1,2 Anthony Galante,3 Patricia
Soteropoulos,3,4 Suqin Guo,5 B. J.
Departments of 1Biochemistry & Molecular Biology, 4Microbiology & Molecular Genetics, and 5Ophthalmology, and the 2Graduate School of Biomedical Sciences, UMDNJ-New Jersey Medical School, Newark, NJ; 3Center for Applied Genomics, Public Health Research Institute, Newark, NJ
Correspondence to: B. J. Wagner, Department of Biochemistry & Molecular Biology, UMDNJ-New Jersey Medical School, PO Box 1709, 185 South Orange Avenue, Newark, NJ, 07101-1709; Phone: (973) 972-5335; FAX: (973) 972-5594; email: firstname.lastname@example.org
Purpose: Prolonged use of glucocorticoids can lead to the formation of a cataract, however the mechanism is not known. We recently reported the presence of the functional glucocorticoid receptor in immortalized cultured mammalian lens epithelial cells (LECs), but the biological effect is not known. This study seeks to determine if freshly isolated human LECs respond to glucocorticoid treatment and to examine glucocorticoid induced changes in global gene expression in LECs.
Methods: Capsulorhexis specimens obtained in surgery from eyes with cataract were cultured. Primary lens cultures were transfected, in triplicate, with pGRE.Luc, which drives the expression of firefly luciferase, and treated with dexamethasone (Dex) or vehicle (Veh). RNA isolated from HLE B-3 cells, treated with Dex or Veh for 4 or 16 h in triplicate, was used to analyze global changes in gene expression by microarray hybridization. Data and cluster analyses were performed using Microarray Suite 5.0, GeneSpring 6.1, EASE, NetAffx, and SAM. Real Time PCR was used to confirm microarray data in RNA isolated from HLE B-3 cells in triplicate and a primary culture of human lens epithelial cells.
Results: Transfected primary cultures of human LECs treated with Dex demonstrated a glucocorticoid response with a greater than 4 fold increase in firefly luciferase activity over controls. Microarray data revealed that 136 genes were modulated with 4 h treatment with Dex. Of the 136 genes, 93 transcripts were upregulated and 43 were downregulated by greater than 1.5 fold. Eighty-six genes were modulated with 16 h Dex treatment. Of the 86 genes, 30 transcripts were upregulated and 56 were downregulated by greater than 1.5 fold. Microarray results were verified by Real Time PCR in both the HLE B-3 and primary cultures of lens epithelial cell.
Conclusions: The activation of a GRE reporter gene in primary cultures of human LECs demonstrates that the glucocorticoid receptor is functional in non-immortalized human lens cells. Microarray studies at 2 time periods demonstrate that glucocorticoids modulate gene expression in immortalized human LECs, reveal novel changes in gene expression, and confirm an endogenous genomic lens glucocorticoid response. This study demonstrates that primary cultures of lens epithelial cells and microarray technology can be used to determine pathways involved in a lens glucocorticoid response and lead to a better understanding of the formation of a steroid induced cataract.
Administration of glucocorticoids is an important therapeutic treatment for diseases such as rheumatoid arthritis, asthma, and various ocular diseases. It has been well established that a complication and side effect of prolonged corticosteroid therapy is the formation of posterior subcapsular cataract with the finding of nucleated epithelial cells in the posterior region of the lens [1-6]. The mechanism of cataract formation or of glucocorticoid action in the lens is not known.
Glucocorticoids (GC) are steroid hormones that play a role in numerous physiological processes, such as regulation of glucose, protein, and fat metabolism, and anti-inflammatory and immunosuppressive actions [7,8]. GCs exert their effects by a variety of different mechanisms. Classically, they exert their effects by binding to a specific intracellular receptor, the glucocorticoid receptor (GR), which acts as a ligand dependant transcription factor [9,10]. The ligand-receptor complex dimerizes, translocates to the nucleus, and binds to a cis acting element, the glucocorticoid response element (GRE), to modulate the expression of target genes.
Alternatively, GCs have been proposed to act on the lens indirectly through mechanisms involving oxidative stress and depletion of glutathione [11-15]. Another hypothesis involves a nonspecific action of GC through the covalent addition of steroids to lens proteins which results in destabilization of protein confirmation, oxidation, and cross linking of protein thiol groups [16,17]. This model may be related to studies demonstrating that the synthetic steroid dexamethasone bound α-crystalline nonspecifically in the bovine lens . A membrane steroid binding protein was also recently identified in bovine lens epithelial cells . Although this receptor is able to bind GC, its mRNA and protein sequence differ from the classical intracellular GR and a membrane steroid binding protein must act by non-genomic actions.
Previous studies provided evidence suggesting that the mammalian lens contained a classical intracellular GR. The bovine lens was reported to contain a GC binding protein that exhibited the characteristics of a receptor [20,21] and immunohistochemical and in situ hybridization studies demonstrated that the rat and human lens contained the GR [22,23]. Intracellular GRs are ubiquitously expressed [24,25], however, despite past evidence, the presence of the GR in the mammalian lens epithelium was questioned . We reported the unquestionable presence of GR mRNA and protein by PCR and western blotting in immortalized and freshly isolated human and mouse lens epithelial cells . Furthermore, we reported that the GR identified in the immortalized human (HLE B-3) and mouse (αTN4) lens epithelial cell cultures was able to activate transcription from a reporter vector containing a GRE element demonstrating a classical functional GR. The expression of the GR mRNA and protein in HLE B-3 cells is similar to that identified in freshly isolated human lens epithelial cells  suggesting the same GC induced GRE directed transcription would occur in a primary culture of human lens epithelial cells. However, immortalized cells differ from primary cultures. Proteome analysis and other studies revealed differences in protein expression between the HLE B-3 cell line and freshly isolated human lenses [28,29]. Immortalized cell lines are useful models to study, however, results need to be confirmed in primary cultures. It has yet to be determined if the GR identified in primary cultures of human lens epithelial cells (hLEC) is transcriptionally active.
The identification of the functional GR capable of inducing gene expression in immortalized human lens epithelial cells suggests that glucocorticoids are able to modulate the expression of target genes. Previous studies have failed to identify glucocorticoid target genes in lens epithelial cells . Studies have shown changes in protein expression without demonstrating changes in gene expression . Our own studies examining well known GC targets, such as IκBα and αB-crystallin, have been inconclusive (unpublished studies). GCs have a wide array of effects and play a role in a variety of cell functions that are cell type specific. The regulation of a well known target gene by GC in one cell type does not guarantee that it will be modulated in another cell type by GC treatment . It may be difficult to identify specific GC targets by looking at genes individually.
Oligonucleotide microarrays have served as useful tools to monitor global gene expression changes in lens epithelial cells [33,34]. In the present report, we have utilized oligonucleotide microarrays to compare the global gene expression profiles between HLE B-3 cells treated with dexamethasone (Dex), a synthetic glucocorticoid, or vehicle (Veh) at two different time periods that demonstrate early lens GC responses and identify possible mechanisms of GC action in lens cells. Real time PCR verified the microarray results. Functional clustering of the modulated genes revealed that GCs play a role in multiple biological processes and molecular pathways. Furthermore, in this study, we demonstrated that primary cultures of human lens epithelial cells contain a transcriptionally active GR and verified that gene changes identified by microarray occurred in primary lens cells as well. These data demonstrate changes in expression of specific gene targets due to GC treatment, identify a GC response in lens epithelial cells and for the first time identify a response in primary cultures of hLECs. These novel findings conclusively identify specific glucocorticoid targets, which provide the basis for further experimentation into the action of glucocorticoids in lens epithelial cells. The broad and large number of functional categories may be related to and provide an explanation for the historical difficulty in understanding a lens glucocorticoid response.
Tissue, cell culture, and treatment
Freshly isolated human lens epithelial cells were obtained from capsulorhexis specimens obtained from surgery or from eye bank donor eyes. Capsulorhexis specimens were placed into culture dishes. The epithelial layer (from donor lenses) was carefully separated from the fiber cells using a dissecting microscope and cultured in a dish. Cultured lens epithelial cells were maintained in phenol red free DMEM with 20% serum and passaged 1 time before treatment. Medium was replaced with medium containing a reduced amount of charcoal stripped serum 16 h before treatment. All procedures complied with the Declaration of Helsinki.
HLE B-3 cells (a gift from Dr. Usha Andley) were maintained in phenol red free MEM with 20% serum. Medium was replaced with medium containing a reduced amount of charcoal stripped serum 16 h before treatment.
Dexamethasone was purchased from Sigma (St. Louis, MO), dissolved in absolute ethanol and diluted in media according to manufacturer's protocols. Absolute ethanol served as a vehicle control. The final concentration of absolute ethanol in the samples never exceeded 0.1%.
Plasmid pGRE.Luc (Clonetech, Palo Alto, CA) contains three copies of the GRE enhancer element fused to the TATA-like promoter region from the HSV-TK promoter and drives the expression of the firefly luciferase reporter gene. Plasmid pRL-SV40 (Promega, Madison, WI) contains the SV40 early enhancer-promoter region driving expression of the Renilla luciferase reporter gene and serves to normalize transfection efficiencies. HLE B-3 cells were seeded (in triplicate) in a 24 well plate 24 h before transfection. Primary cultures of lens epithelial cells, from capsulorhexis specimens, were seeded in triplicate in 96 well plates. Cells were co-transfected with pRL-SV40 and pGRE.Luc using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) as previously reported. Transfection medium was changed after 5 h, and 16 h later cells were treated with 1 μM Dex or Veh for 24 h. The Dual Luciferase Reporter Assay System (Promega, Madison, WI) was used according to manufacturer's instructions to assay cell extracts. Luciferase activity was measured on a luminometer (Packard LumiCount, Packard Instrument Company, Downers Grove, IL).
Microarray RNA preparation, hybridization, and analysis
HLE B-3 cells (tenth passage) were seeded (in triplicate) in full serum medium and 24 h later cells were washed and medium was changed to 2% charcoal-stripped serum medium and cells were incubated for 16 h before treatment. Cells were treated with 1 μM Dex or Veh for 4 or 16 h in triplicate. All samples for a single time point were processed and analyzed at the same time.
RNA extraction was performed with RNAzol (Tel-Test, Friendswood, TX). RNAzol (1.6 ml) was added to each flask for 3 min before the homogenate was vigorously pipetted and transferred to 2 ml tubes on ice. Chloroform (160 μl) was added, shaken vigorously, and incubated on ice for 7 min. Samples were centrifuged at 12,000x g and 4 °C for 15 min and the aqueous layer was transferred to a clean tube. Isopropanol (70 μl) was added to each tube and was mixed gently by inversion before storage at -80 °C overnight. The following day the sample was centrifuged at 12,000x g for 15 min at 4 °C. The supernatant was removed and the pellet was washed with 80% ethanol (2 times) and the pellet was air dried and resuspended in 30 μl RNAse free water (Ambion, Austin, TX). Each sample was analyzed by spectrophotometry and agarose gel electrophoresis before RNA purification and clean up.
RNA was purified using the RNeasy kit (Qiagen, Valencia, CA) according to manufacturer's protocols. Concentration was determined spectrophotometrically and examined for quality on 1% agarose gels. 260/280 ratios were greater than 1.8.
Double-stranded cDNA was synthesized from total RNA using the Superscript Double Stranded cDNA Synthesis Kit (Invitrogen, Carlsbad, CA). First strand synthesis was carried out with 5 μg total RNA, 100 pM T7-(dT)24 primer and DEPC treated water. The sequence of the HPLC purified T7-(dT)24 primer was 5'-GGC CAG TGA ATT GTA ATA CGA CTC ACT ATA GGG AGG CGG TTT TTT TTT TTT TTT TTT TTT TTT-3' (Integrated DNA Technologies, INC, Coralville, IA). The RNA primer mixture was incubated at 70 °C for 10 min. First strand buffer and 10 mM DTT and 500 μM of each dNTP was added to the mixture and incubated at 42 °C for 2 min. Finally, 200 Units of SuperScript II Reverse Transcriptase was added and the RT reaction was carried out at 42 °C for 1 h.
Second strand cDNA synthesis was carried out by adding second strand buffer, 200 μM each dNTP, 10 Units DNA ligase, 40 Units DNA polymerase I, 2 Units RNase H and water to the first strand synthesis reaction. The mixture was incubated at 16 °C for 2 h. 10 Units T4 DNA polymerase was added to the reaction and incubated at 16 °C for 5 min. EDTA (10 μl of 0.5 M) was used to stop the reaction.
To clean up the double stranded cDNA, 1.5 ml Phase Lock Gel light tube (Eppendorf, Westbury, NY) was centrifuged for 30 s at 12,000x g to pellet the gel. Phenol:chloroform:isoamyl alcohol (25:24:1, saturated with Tris-HCL, pH 8.0, 1 mM EDTA) was added to the double stranded cDNA synthesis preparation and vortexed. The cDNA-phenol mixture was transferred to the Phase Lock Gel tube and centrifuged at 12,000x g at room temperature for 2 min. The aqueous layer was transferred to a fresh tube. NH4Ac (0.5 volumes of 7.5 M) and 2.5 volumes of cold absolute ethanol were added to the sample, vortexed, and stored at -20 °C overnight. The next day, the mixture was centrifuged at 12,000x g for 20 min at 5 °C and the supernatant was discarded. The pellet was washed with 80% ethanol and centrifuged at 12,000x g for 5 min at 5 °C. The wash was repeated and the pellet was dried and resuspended in RNase free water and analyzed on a 1% agarose gel.
cRNA synthesis was carried out through an in vitro transcription reaction using the ENZO Bioarray HighYield RNA Transcript Labeling Kit (Affymetrix, Santa Clara, CA) according to manufacturer's protocols. cDNA was added to a mixture containing 1X HY reaction buffer, 1X biotin labeled ribonucleotides, 1X DTT, 1X RNase Inhibitor mix, 1X T7 RNA Polymerase, DEPC treated water and incubated at 37 °C for 4-5 h with gentle mixing by pipetting every 45 min. Each reaction was divided in half and half was stored away at -80 °C. The other half was purified with the Qiagen RNeasy Mini Kit (Qiagen, Valencia, CA).
RLT buffer (without β-mercaptoethanol) was added to the sample and mixed. Absolute ethanol was added and the mixture was placed into a RNeasy mini spin column and centrifuged at 8000x g for 15 s. The flow through was reapplied to the column and centrifuged at 8000x g for 15 s. The column was transferred to a fresh collection tube and buffer RPE was added before centrifugation was repeated. Filtrate was discarded and the RPE wash step was repeated. The column was placed into a new collection tube and centrifuged for 2 min at maximum speed. The column was placed into a fresh collection tube and RNAse free water was added directly to the column membrane and incubated at room temperature for 1 min. cRNA was extracted by centrifugation at 8000x g for 1 min at room temperature.
To concentrate the cRNA, 0.5 volumes of 7.5 M NH4Ac and 2.5 volumes of cold absolute ethanol were added to the sample, mixed and stored at -20 °C overnight. The mixture was centrifuged at 12,000x g for 20 min at 5 °C and the supernatant was discarded. The pellet was washed with 80% ethanol and centrifuged at 12,000x g for 5 min at 5 °C. The wash was repeated and the pellet was dried in a Speedvac and resuspended in RNAse free water. cRNA concentration was determined by spectrophotometry and was analyzed by electrophoresis on a 1% agarose-formaldehyde-borate gel.
cRNA (15 μg) was fragmented and added to a hybridization mixture at the Center for Applied Genomics (The Public Health Research Institute, Newark, NJ.). Expression profiles were created using the HG-U133A GeneChip (Affymetrix, Santa Clara, CA), which contains 22,283 known human transcripts and ESTs coding for about 15,000 known genes. These transcripts are designed using 11-20 probe pairs consisting of 25-mer oligonucleotides. Hybridization was done overnight at 45 °C for 16 h using the GeneChip Hybridization Oven 640 (Affymetrix, Santa Clara, CA). Washing and staining (Streptavidin Phycoerythrin) was accomplished with the GeneChip Fluidics Station 400 (Affymetrix, Santa Clara, CA) using the EukGE-WS2v4 protocol. Images were acquired using the Affymetrix GeneArray scanner. Data was extracted using Affymetrix Microarray Suite 5.0.
Data analysis was performed using several different software packages including Microarray Suite 5.0 (Affymetrix, Santa Clara, CA) and GeneSpring 6.2 (Silicon Genetics, Redwood City, CA). Numeric data were extracted from DAT images and normalized using Microarray Suite. The method of normalization used was a scaling algorithm which involves multiplying the mean intensity of each chip (not including the upper and bottom 2%) by a factor which changes the mean intensity to 500 for every chip. By scaling each chip, a direct comparison could be made between all the chips. The data were entered into the GeneSpring for further analysis. Data were filtered based on both Detection Call and Signal Log Ratio in any of the comparisons between 4 h Dex, 4 h Veh, 16 h Dex, and 16 h Veh. Comparisons made between groups had to be uniform within groups. In the case of the Affymetrix Detection Call algorithm, data that were declared Absent in all samples compared were filtered out. Samples had to be declared Present in at least two of three chips within a treatment group to be kept within the set. An additional filter was made based on the Signal Log Ratio which is the log base 2.0 of the fold change. Any comparisons which had a conditional group declared Present had to have at least one group with a Signal Log Ratio greater than ±-0.58 (a fold change greater than ±-1.5). In addition, ANOVA (assuming unequal variances) was run on the filtered list. The Benjamin-Hobbson false discovery rate was applied as a multiple correction factor.
Filtered genes identified to be differentially expressed by 1.5 fold or greater in two of three chips were analyzed for functional gene clusters using the Expression Analysis Systematic Explorer (EASE) , NetAffx Analysis Center , Significance Analysis of Microarrays (SAM) , and GeneSpring . These programs are used to determine functional clusters by statistical representation of individual genes in specific categories relative to all genes in the same category on the array. EASE provides statistical methods for discovering enriched biological themes within gene lists and generating gene annotation tables. The NetAffx Analysis Center allows the correlation of the microarray results with the specific array design and with annotation tools. The Gene Ontology (GO) Mining Tool, used in the EASE and NetAffx Analysis Center, matches GeneChip probe sets to annotated genes within the biological process, molecular function, or cellular components to allow for biological interpretation of microarray results. GeneSpring uses data found publicly in genomics databases to build gene ontologies based on annotation information.
Real time polymerase chain reaction (RT-PCR)
Capsules containing the epithelial layer from a single pair of 46-year-old donor lenses were dissected from the fiber cells and placed in DMEM containing 20% serum. HLE B-3 cells (tenth passage) were seeded (in triplicate) in phenol red free MEM containing 20% serum. Sixteen h before treatment, medium was changed to phenol red free MEM with 2% charcoal stripped serum. Cultures were treated with 1 μM Dex or Veh for 4 h.
Total RNA was isolated using RNAzol (Tel-Test, Friendswood, TX) according to manufacturer's protocols. RNA concentration was determined by spectrophotometry. Isolated RNA was then aliquoted and stored at -80 °C.
RNA was reverse transcribed using Applied Biosystems reagents on a thermocycler (GeneAmp PCR system 9700; PE Applied Biosystems, Foster City, CA) according to the manufacturer's protocols. Briefly, 1 μg total RNA was mixed with 5 mM MgCl2, 1X PCR buffer, 4 mM each dNTP, RNase Inhibitor, Oligo dT, and MuLv reverse transcriptase. The reaction was incubated at 42 °C for 60 min, 95 °C for 5 min, and held at 4 °C.
For each sample, primers for actin were used to determine the quality of the RNA. The sequences of human specific primers used in this study, along with their corresponding GenBank accession numbers and product sizes are shown in Table 1. Primers were designed by using Primer3. The human specific primers were designed to Period 1 (Per1), Delta Sleep Inducing Peptide Like Immunoreactor (DSIP), Heat Shock Protein 70 (HSP70), Protein Kinase c-AMP dependant Regulatory type 1 Alpha (PRKAR1A), Coagulation Factor II (Thrombin) Receptor (F2R), Plasminogen Activator Inhibitor-1 (PAI-1), Growth Arrest and DNA-damage-inducible protein (GADD45), Serum Glucocorticoid Regulated Kinase (SGK), Pleckstrin Homology-Like Domain, Family A (PHLDA), Immediate Early Response 3 (IER3), Nerve Growth Factor (NGF), Sodium Channel, Non-voltage-Gated 1 Alpha (SCNN1A), Cyclin D1/BCL-1 (CCND1), and Cholecystokinin (CCK). The sequence of the primers for the Monocyte Chemotactic Protein 1 (MCP-1) and Dual Specificity Phosphatase 1 (DUSP1) were previously published [39,40].
PCR was performed using the LightCycler FastStart DNA Master SYBR Green I kit (Roche Applied Science, Indianapolis, IN), and the LightCycler 1.0 (Roche Applied Science, Indianapolis, IN). SYBR green fluoresces upon binding to the minor groove of dsDNA. Monitoring the fluorescence of the reaction in real time allows the amplification to be halted when the sample is undergoing exponential growth making quantification of small differences possible. The reaction was stopped during the log phase to allow for quantification of small differences.
The quantification of material labeled with SYBR green was analyzed by crossing point analysis, which represents the cycle number at which the sample begins exponential growth over the background noise. The data were presented in fluorescence versus cycle format in which all sample baselines were brought to a comparable level. The baselines were brought into a similar range by an arithmetic baseline adjustment in which the mean of the five lowest measured data points for each sample was subtracted from each data point. Next the exponential curve was transformed into a linear curve and a noise band was set that excluded background noise levels and established the lower limit of analysis in the exponential phase for all samples. The data were then presented in log-linear format. The number of Fit Points plotted on the exponential portion of the curve were increased to establish the upper limit of analysis and to include the maximum number of acquisition events in the crossing point assessment. The relative fold difference in exponential growth was defined as 2(a-b) where a and b represent the crossing points of the two samples being compared. Results were normalized to actin. Specificity of PCR products was determined by melting curve analysis and visualization on 1 or 2% agarose gels stained with ethidium bromide.
In order to identify if primary cultures of human lens epithelial cells respond to GC treatment, primary cultures were created from explants of capsulorhexis specimens from surgery. Epithelial cells grew off the capsule and onto the plate (Figure 1). Growing explant cultures were then transfected with pGRE.Luc. Transfected cells were treated with 1 μM Dex or Veh for 24 h. Epithelial cell explants from four pairs of donor lenses were examined, in duplicate or triplicate, and, despite a large sample variation, each sample displayed greater than 4 fold increase in luciferase activity in Dex treated samples (Figure 2).
Glucocorticoid induced changes in gene expression have been reported to occur as early as 15-30 min after hormone administration . In order to identify an early time point to identify a primary response through GRE mediated genes, HLE B-3 cells were transfected with pGRE.Luc and treated with 1 μM Dex or Veh over a time course of 24 h. A significant increase in luciferase activity compared to vehicle was identified between 2-4 h and this was sustained over a 24 h period (Figure 3). Although GC mediated changes in gene expression may be occurring before 4 h, a 4 h treatment time period was chosen to examine by microarray. A later time of 16 h was also examined by microarray to identify genes that may be downstream of the 4 h response to lead to a better understanding of pathways involved in the GC response.
Microarray analysis was performed on total RNA extracts from tenth passage HLE B-3 cells treated with either 1 μM Dex or Veh for 4 or 16 h (in triplicate). The complete data profiles have been deposited in NCBIs Gene Expression Omnibus (GEO; accession number GSE3040).
The scaled data generated from Microarray Suite were imported into GeneSpring for fold change analysis, filtering, and cluster analysis. To identify highly reproducible changes, data were filtered based on select criteria. To be kept within the set, transcripts had to be modulated by at least 1.5 fold in at least two of three chips within a treatment group. All of the genes listed in Table 2 passed the filtering criteria in GeneSpring. However, GeneSpring and Microarray Suite use different algorithms for generating fold change. To determine the standard error and p values by ANOVA, the data filtered by GeneSpring was analyzed in Microarray Suite. Due to the difference in algorithms, several of the genes that met the 1.5 fold criteria in GeneSpring did not meet the same criteria when analyzed in Microarray Suite 5.0 (Table 2). However, genes that fell below the 1.5 fold cutoff or had a p value of greater than or equal to 0.05 were kept within the set in order to maximize the number of genes. Lens glucocorticoid responses have been difficult to elucidate. Since the genes met the criteria in GeneSpring, they were left within the set in order to avoid the possibility of excluding potential GC lens targets. The genes included by Genespring are useful as the purpose was to find a list of potentially interesting genes that will be confirmed by real-time PCR.
Hierarchical clustering with the Pearson Correlation was performed based on signal intensities generated from Microarray Suite. A condition cluster and dendrogram were generated based on the gene list created after filtering. The condition tree clustered based on the overall expression of each chip or treatment. Concurrently, a gene tree and dendrogram were also generated on the gene list. The gene tree clustered based on the expression of each gene across treatments. The treatments clustered together and demonstrated that the triplicates were reproducible (Figure 4).
A list of 136 genes passed the criteria for the 4 h data set and 86 genes for the 16 h data set (Table 2). Of the 136 genes from the 4 h data set, 93 transcripts were upregulated and 43 were downregulated by greater than 1.5 fold. Of these, 38 genes were upregulated and 1 was downregulated by greater than 3 fold. Of the 86 genes from the 16 h data set, 30 transcripts were upregulated and 56 were downregulated by greater than 1.5 fold. Of these, 9 transcripts were upregulated and 1 was downregulated by greater than 3 fold. Seven genes overlapped the two data sets (Table 3). A few of the genes fell below the 1.5 fold cut off due to the different algorithms used in the two programs, Microarray Suite and GeneSpring, used for analysis of the data. Although the fold changes were small (between 1 to 3 fold), many of them were statistically significant (p less than or equal to 0.05, Table 2).
To verify the results seen in the microarray, transcripts were examined in 4 or 16 h Dex or Veh treated HLE B-3 (in triplicate) and human lens explant cultures (created from donor lenses) treated for 4 h by RT-PCR. Human lenses treated for 16 h were not examined due to lack of material. The reaction was stopped during the log phase. Although the extent to which the fold change in gene expression differed between the microarray and real time PCR results, the general trends were consistent and correlated with the microarray data providing confidence in the microarray results, yet demonstrating the need for confirmation of microarray results (Table 4). In addition to SYBR green analysis, PCR products were visualized on 1 or 2% agarose gels stained with ethidium bromide (Figure 5) and these results were consistent with the RT-PCR data.
The 5' promoter region and full gene sequences of these target genes were analyzed by a signal scan program to determine if GREs were present. All of the genes examined contained putative GREs (data not shown) suggesting direct GC-GR signaling. Although no published gene sequence is available for DSIP, previously published reports indicate that DSIP is upregulated by GC treatment . DSIP has also been reported to be homologous to TSC22 . Our own alignment studies demonstrated that DSIP is 100% homologous to TSC22 and a signal scan of the TSC22 genomic sequence revealed putative GREs (data not shown).
To better understand what effect the change in gene expression may have on lens cells, microarray results were analyzed by functional clustering using EASE and NetAffx to identify biological and molecular pathways. EASE is an analysis tool that identifies enriched biological themes by prioritizing functional categories of genes that cluster together under a specific biological or molecular function based on the significance by determining gene sets which are statistically overrepresented. Overrepresentation is calculated based on the total number of genes assayed and annotated within each system. This is determined by the probability of seeing the number of "List Hits" (number of genes in the gene list that belong to the gene category) in the "List Total" (number of genes in the gene list) given the frequency of the "Population Hits" (number of genes in the total group of genes assayed that belong to the specific gene category) in the "Population Total" (number of genes in the total group of genes assayed that belong to any gene category) within the system. This is called the Fisher exact probability. EASE also determines an "EASE score" which is calculated by removing one gene in the category from the list and then calculating Fisher exact probability for that category . The EASE score is an adjustment of the Fisher exact that strongly penalizes the significance of categories supported by few genes and negligibly penalizes categories supported by many genes. The EASE score favors higher ranked categories (lists with larger numbers of genes) compared to the Fisher exact probability . The most significantly overrepresented categories that result from analysis of the gene lists are labeled "biological themes" or "molecular themes" (Figure 6, Figure 7, Table 5, and Table 6). Those categories with an EASE score of less than or equal to 0.05 are underlined. Some of the genes are unclassified, which means no annotations are available for that gene identifier under the specified parameters although an annotation for that gene may exist. The size of the pie slices and the number of genes in any one functional category do not reflect biological importance. They only reflect the results of EASE analysis and give an understanding of the biological and molecular functions in which Dex induced changes in gene expression may be involved.
Although multiple biological processes were represented, only coagulation and metabolism had an EASE score of less than or equal to 0.05 and were significantly modulated in lens epithelial cells treated with Dex for 4 h. All the categories listed in the pie chart contained genes that were both upregulated and downregulated, except for the categories of coagulation and death, in which all the genes were upregulated. Although the categories contained genes that were both upregulated and downregulated, the majority were upregulated.
Response to stimulus, coagulation, and metabolism were significantly modulated in lens epithelial cells treated with Dex for 16 h. All the categories listed in the pie chart contained genes that were both upregulated and downregulated, except for the categories of morphogenesis and death, in which all the genes were downregulated. The categories contained relatively equal numbers of genes that were both upregulated and downregulated except for metabolism in which the majority of genes expressed were downregulated.
The genes modulated with 4 h Dex treatment clustered into several molecular categories, however, none of the categories identified received an EASE score that was considered significant. Many of the categories represented contained genes that were both upregulated and downregulated, although some only contained genes that were upregulated. The molecular function category identified as significant was receptor binding in 16 h Dex treated lens cells. Once again, the categories contained both genes that were upregulated and downregulated, however, some categories only contained genes that were downregulated. Additional analysis of microarray results was performed with SAM, GeneSpring, and NetAffx (data not shown). Clustering with different programs resulted in expression of similar categories, and parallel to EASE results, the categories were very few and broad.
EASE was also used to identify pathways through the use of KEGG pathways terms associated with the differentially expressed genes. EASE analysis of 4 and 16 h modulated genes through use of KEGG pathways revealed that the phosphatidyl inositol signaling system, complement and coagulation cascades, cell cycle, integrin mediated cell adhesion, purine metabolism, and pyrimidine metabolism all contained 3-6 modulated genes that functionally clustered into these pathways. Some of the genes are unclassified in the KEGG pathways, which means no annotations are available for that gene identifier under the specified parameters designated although an annotation for that gene may exist. The size of the pie slices and the number of genes in any one category do not reflect the biological importance, only the relative number of genes with modulated expression within the pathways category. Some genes may be represented in more than one pathway (Figure 8). The transcripts from the microarray that fall into some of the different pathways are listed in Table 7.
Of the pathways represented, the phosphatidyl inositol signaling system, complement and coagulation cascades, purine metabolism, and circadian rhythm pathways contained transcripts that were upregulated from the 4 or 16 h data sets. Integrin mediated cell adhesion, pyrimidine metabolism, cell cycle, and complement and coagulation cascades contained transcripts that were downregulated from the 4 or 16 h data sets.
In the current study, we have demonstrated that primary cultures of human lens epithelial cells express a functionally active GR. Furthermore, using microarray technology, we examined the global gene expression of HLE B-3 cells treated with GC to identify GC mediated gene targets and were able to verify the results in immortalized and primary cultures of lens epithelial cells.
Previously, we identified the expression of classical GR mRNA and protein in freshly isolated human lens epithelial cells . In this study, we demonstrate that the GR in primary cultures of human lens epithelial cells is able to induce transcription of a GRE mediated gene. This is similar to the luciferase expression we previously reported in the HLE B-3 immortalized human lens epithelial cell cultures . This is the first report of GR activity in primary cultures of human lens epithelial cells. These data demonstrate that primary cultures of hLECs contain a transcriptionally active GR able to induce changes in gene transcription and provide further proof that the lens GR is functionally active.
Glucocorticoids are known to elicit divergent biological outcomes and a lens glucocorticoid response has been difficult to characterize [18,26,28]. Because of the association with cataract, it is important to understand the biological significance of the GC effect in lens epithelial cells. Although GCs have been proposed to affect lens epithelial cells through a variety of mechanisms, including oxidative stress [11-15], nonspecific GC binding [16-18], and membrane steroid receptors , the identification of a transcriptionally active intracellular GR suggests that GCs may be modulating the transcription of target genes. To gain a better understanding of glucocorticoid induced gene changes in lens epithelial cells, microarray technology serves as a useful method to identify changes in global gene expression that can lead to identification of biological functions or signaling pathways [44,45]. A previous published report has demonstrated global changes in gene expression in lens epithelial cells after 24 and 48 h of treatment with GC, however it has not identified early glucocorticoid responses or signaling pathways involved in a GC response . Understanding of a GC response requires an understanding of short term and long term effects of treatment.
GC induced changes in gene expression due to GR binding to a GRE have been reported to occur as early as 15-30 min after hormone administration but can also occur as late as 4-24 h in other cell types [41,46-48]. Although glucocorticoid effects on lens epithelial cells, such as changes in protein expression , the formation of a cataract [1-6], and changes in gene expression  have only been observed after long term treatment, a transcriptionally active receptor suggests that there may be changes in gene expression after a short treatment time. In our experiments, a significant increase in luciferase activity of Dex compared to Veh treated cells was identified as early as 2 h and this was sustained over a 24 h period. These findings demonstrate an early GC response in lens epithelial cells suggesting that GCs can play a role in gene expression as soon as 2 h. Although GC mediated changes in gene expression may be occurring as early as 2 h, at 4 h the increase in luciferase activity was more robust and was chosen for examination on microarray. This time point provides information about GC targets, however, a second treatment time of 16 h was also examined by microarray in order to better understand signaling pathways involved in a GC response.
Three biological replicates were created and analyzed for each of the treatments (Veh or Dex) at each of the time points (4 h or 16 h). The analysis process can be broken down into two parts; identification and verification of differentially expressed genes and the determination of the biological significance of groups of genes through analysis of the biological and molecular themes with respect to Gene Ontology and KEGG pathway terms associated with these genes. Although the overall expression of each of the chips was similar and the triplicates were reproducible, differences were observed. It is normal to observe differences in replicate microarrays . To identify specific gene expression changes that are significantly modulated and are reproducible, in the first step of analysis the triplicates were stringently filtered. Although three biological replicates were created, genes were kept within the set if they were modulated by 1.5 fold in at least two of three chips in order to identify a greater number of genes. Based on the difficulty of demonstrating a direct effect of glucocorticoids on the lens, we predicted that glucocorticoid effects on gene expression would be few and small. The first step of analysis resulted in identifying changes in 136 transcripts after 4 h of treatment and 86 transcripts after 16 h of treatment. Only 7 transcripts overlapped the two groups, suggesting continual signaling and transcription of the 7 genes. Of the 7, the two most notable were the DSIP and the PER 1 genes which were upregulated by at least 5 fold in both data sets. It is interesting to note that of the 7 transcripts expressed at 4 and 16 h of Dex treatment in our results, it has been reported that by 48 h of Dex treatment, DSIP, PAI-1, and DKFZP586A0522 all continue to have >2 fold upregulated expression while Chemokine ligand and pleckstrin homology like domain are downregulated by >2 fold compared to Veh in HLE B-3 cells . Period 1, which continued to exhibit nearly 5 fold expression after 16 h of Dex treatment, was not identified on the 48 h microarray . Coagulation factor II receptor was immediately downregulated from 4 fold expression at 4 h to nearly 1 fold expression at 16 h. This demonstrates changes in signaling as the GC response continues over a long time period indicating the need to understand the effect of short term and long term treatment times.
Studies looking at changes in gene expression by microarray at early time points (2 to 8 h) after glucocorticoid treatment observed a markedly greater number of upregulated transcripts than downregulated transcripts [49,50], similar to our results. In our own studies, we observed 93 transcripts upregulated and 43 transcripts downregulated in lens epithelial cells treated with glucocorticoid for 4 h. It is interesting to note that after 16 h of treatment, 30 transcripts were upregulated while 56 were downregulated. It has been reported that after 24 h of GC treatment 57 transcripts were upregulated and 50 were downregulated and by 48 h 92 were upregulated and 42 were downregulated . By 24 h, it appears that the number of up regulated and down regulated transcripts is nearly equal but by 48 h, the number of up regulated genes is markedly increased. Glucocorticoids appear to be modulating genes differently at early and later times.
Glucocorticoid actions at the genomic level have been reported to involve transactivation, due to the binding of the activated GR to a GRE, and transrepression, due to the binding of an activated GR to a negative GRE . It is possible that the large number of upregulated transcripts at early time points could be due to positive regulation, or transactivation, of genes, and at a later time point of 16 h the large number of downregulated transcripts could be due to transrepression of genes. Glucocorticoid responses can be divided into two types, primary and secondary responses. In a primary response, the activated GR binds directly to a GRE to activate gene transcription and no new protein synthesis is needed. A secondary response involves the elapse of time during which the products of the primary response act as positive and negative transcription factors or co-factors . It is possible that the large number of downregulated transcripts at 16 h could be due to a secondary glucocorticoid response with the expression of co-repressors that were upregulated at 4 h. DSIP is expressed at both the 4 and 16 h time points and has been reported to act as a transcriptional repressor inhibiting the transcription of PPAR-γ2 in adipocytes . There are also examples of promoters containing GREs responding both positively and negatively to GC, depending on the conditions . This may be due to differences in GR cofactor expression over time or due to different cell types and may determine whether GREs are positively or negatively regulated by GR.
Several of the transcripts were analyzed by RT-PCR and revealed trends of expression similar to that seen on the microarray and thus confirmed the accuracy of the microarray results in the immortalized HLE B-3 human lens epithelial cell line and in a primary culture of human lens epithelial cells created from donor lens explants. However, although the data are statistically significant and correlate with the microarray data, the differences in fold changes seen between the microarray and RT-PCR demonstrate the need to verify microarray results after both the first and second steps of microarray analysis. Transcripts, for which genomic sequences were available, TSC22, PER1, HSP-70, PRKAR1A, F2R, PAI-1, GADD45, SGK, PHLDA, IER3, NGF, MCP-1, SCNN1A, DUSP-1, CCND1, and CCK were examined and found to contain putative GREs in the promoter regions suggesting that the modulation of these genes could be due to direct binding of the ligand bound GR to the GRE of these promoters in hLECs.
Primary cultures of LECs were created from donor lenses and capsulorhexis specimens obtained from cataract surgery. The capsulorhexis specimen was from a type of cataract not associated with prolonged glucocorticoid use, however, we do not know if LECs obtained from cataractous specimens would respond to glucocorticoid treatment differently than noncataractous specimens. Human specimens, whether they be from donor lenses or from surgical specimens, are difficult to control for because each sample will be different depending on the patient's genetics, environmental factors, and medical history. Ideally, a microarray could be repeated with freshly isolated human lens epithelial cells. However, previous reports of microarray studies with human lenses reported the use of between 12-108 samples in order to obtain sufficient material to hybridize to just one microarray chip [33,54]. Due to a limited amount of sample, we chose to verify the results through RT-PCR of RNA obtained from primary cultures of single lenses. To minimize the contribution of lens pathology, the response was measured from vehicle or dexamethasone treatment of the same specimen. We have demonstrated that primary cultures of human lens epithelial cells, from capsulorhexis specimens and donor lenses, respond to glucocorticoid treatment similarly to immortalized human lens epithelial cell cultures. Human specimens, whether they are from donor or cataractous lenses, are limited so immortalized cell cultures can be used to examine hypotheses, which can then be confirmed in primary cultures. We demonstrated that the results obtained from the HLE B-3 cell line could be confirmed by real time PCR in primary cultures of human lens epithelial cells. This is the first demonstration of a glucocorticoid induced change in gene expression in primary cultures of human lens epithelial cells.
Microarray technology through the use of simultaneous monitoring of the expression levels of thousands of genes identified gene expression changes that are small but distinct and significant, suggesting that glucocorticoids have a direct effect on lens cell function. Although a total of 215 transcripts were identified, this information alone does not provide much information about the biological response of human lens epithelial cells to GC treatment. The real understanding and discovery come from the data analysis.
EASE data analysis revealed that the genes modulated are involved in a wide array of biological and molecular processes such as coagulation, response to stimulus, and metabolism. As these categories were further subdivided, the specificity and the number of categories increased but fewer and fewer genes clustered together under a single function. It is clear that GCs modulate genes in hLECs that are involved in a wide array of processes. GCs appear to be modulating a number of genes, clearly demonstrating a glucocorticoid response, but they do not appear to be modulating a specific biological or molecular function collectively. This may account for the difficulty researchers have had in identifying a functional response to glucocorticoid treatment in lens epithelial cells and demonstrates the difficulty in understanding glucocorticoid action in LECs.
Results from the reported 24 and 48 h microarray studies also demonstrate that GC induced changes in gene expression cluster into broad categories. However, it was not reported if the number of modulated genes that clustered into any of these categories compared to the number of genes represented in the category were considered significant . It is interesting to note that there is some overlap with our results in the identification of certain functional groups including transcription factor activity, cell cycle, and apoptosis.
Our analysis with EASE was confirmed through analysis of data with NetAffx, GeneSpring, and SAM. It is important to note that neither EASE nor any of the other programs used, determines the expression on the chip, nor distinguishes between upregulated and downregulated transcripts or positive or negative regulators of the selected functional category. Therefore it is important to correlate the results of the array with the functional groups and the literature database in order to understand the effect the modulation of gene expression has on a particular cell function, especially if large groups of genes cluster together under a single function.
Although the transcripts do not cluster together under a specific functional group, it does not mean they do not coordinately function together. Modulated genes may not be affecting a specific cell function, but they may be involved in signaling pathways. EASE was also used to identify pathways through the use of KEGG pathways terms associated with the differentially expressed genes. EASE analysis of 4 and 16 h modulated genes through use of KEGG pathways revealed that the phosphatidyl inositol signaling system, complement and coagulation cascades, and cell cycle contained several genes that functionally clustered into these pathways.
EASE and KEGG terms are not absolute. Although a gene may be listed as belonging to a particular pathway, careful literature searches reveal that the same gene is involved in another pathway as well. The MAP Kinase and Phosphatidyl Inositol pathways illustrate this. In this data set, EASE only identified one transcript, MEKK3 (Affy ID 218311_at) as playing a role in the MAP Kinase Pathway (data not shown). EASE identified 6 transcripts that play a role in the phosphatidyl inositol pathway and one of them is the Dual Specificity Phosphatase-1 (Affy ID 201041_s_at). Careful literature searches reveal that Dual Specificity Phosphatase-1 is also known as Map Kinase Phosphatase-1 . Alignment of cDNA transcripts reveal that the two are 100% homologous (data not shown). Map Kinase Phosphatase-1 is an important regulator of the MAP Kinase Pathway . Further examination of the microarray results and literature reveals several other modulators of the MAP Kinase pathway that were not identified by EASE, including Plasminogen Activator Urokinase which stimulates MAP Kinases , Serum and Glucocorticoid Regulated Kinase which inhibits an upstream activator of MAP Kinase , and chemokine monocyte chemoattractant protein 1/Monocyte chemoattractant protein-1 which stimulates MAPK activation . It appears that glucocorticoids are modulating signaling pathways in lens cells.
The phosphatidyl inositol pathway is involved in cellular responses such as cell survival, cell proliferation, cell growth, and transformation [60,61]. The MAP Kinase pathway is involved in cell proliferation, cell differentiation, cell motility, cell survival, and apoptosis . Both of these pathways are important in cell stress responses and appear to crosstalk and play roles in similar functions [63,64]. Glucocorticoid treatment of lens epithelial cells does not appear to be modulating one specific biological function, but instead appears to be playing a small role in a variety of functions. It is possible that glucocorticoids affect many cellular functions through the modulation of components and regulators of the phosphatidyl inositol and MAP Kinase pathways. Regulators of the MAP Kinase and phosphatidyl inositol pathways were reported to be modulated in the 24 and 48 h microarrays . Specifically, Map Kinase Phosphatase-1, which exhibited greater than 2 fold expression on our 16 h microarray, continued to demonstrate 2 fold expression at 48 h. Like glucocorticoids, both of these pathways are involved in a multitude of cellular and molecular functions, which may be a reason that researchers have been unable to identify a specific short term glucocorticoid response. It is possible that prolonged modulation of these pathways could lead to abnormal lens epithelial cell proliferation, differentiation, motility, survival, or apoptosis, all of which have been implicated in the formation of a steroid induced cataract. HES1 (Affy ID 203395_s_at) is down regulated by nearly 2 fold in the 4 h microarray. HES1 is a transcriptional repressor that inhibits proliferation in PC12 and MCF-7 cells [65,66]. However at 16 h, cyclin D1, which is involved in proliferation in the G1 to S transition of the cell cycle, is downregulated. TNFRSF6 (AFFY ID 204780_s_at), a member of the TNF-receptor superfamily, has a >3 fold expression in the 4 h microarray and activation of the receptor is known to be involved in a wide range of responses, including cell death, cell proliferation, inflammation, and differentiation . CCL2 (Affy 216598_s_at), which is downregulated in both the 4 and 16 h microarray by >2 fold has been shown to be involved in the migration of endothelial cells . However, IL8, which is increased by 3 fold in the 16 h microarray, has been suggested to be involved in migration . SGK (Affy ID 201739_at) was upregulated with GC treatment of LECs and appears to be involved in a glucocorticoid receptor mediated protection from apoptosis in a mammary cell line . Furthermore many of these genes have connections with the MAP Kinase or phosphatidyl inositol pathways [71-76]. These seemingly diverse changes in gene expression may be due to glucocorticoid modulation of signaling pathways.
A decrease in protein expression of E cadherin and N cadherin was reported after prolonged treatment with Dex without changes in mRNA expression . We did not see any changes in gene expression of the cadherins on the 4 or 16 h micorarrays. It is interesting to note that both these signaling pathways have been reported to be involved upstream and downstream of cadherin expression [77-79]. We are currently investigating the effect of glucocorticoid in hLECs on the MAP Kinase and phosphatidyl inositol pathways.
Our results presented here demonstrate conclusively that human lens epithelial cells express a transcriptionally active GR, demonstrate early GC responses, and have identified endogenous changes in gene expression that play a role in very important cell functions including cell signaling, cell communication, cell metabolism, cell proliferation, and cell death. Since glucocorticoids are used in the clinical setting for treatment of acute and chronic disorders and a negative side affect of prolonged GC use is the formation of a cataract, it is important to understand a lens GC response. The studies reveal that glucocorticoids do not appear to modulate a specific cellular function but instead modulate important biological pathways that may be involved in the lens glucocorticoid response. This will lead to a better understanding of glucocorticoid signaling in lens epithelial cells, the effect of prolonged glucocorticoid treatment and the formation of a steroid induced cataract.
The authors would like to thank Ilene Seguino for help with obtaining donor lenses and Dr. Bin Tian for help with microarray and SAM analysis. This work was supported in part by NIH-NEI grant EY02299 (to BJW) and an unrestricted grant from Research to Prevent Blindness, Inc., NY (to the Department of Ophthalmology).
1. Black RL, Oglesby RB, Von Sallmann L, Bunim JJ. Posterior subcapsular cataracts induced by corticosteroids in patients with rheumatoid arthritis. JAMA 1960; 174:166-71.
2. Havre DC. Cataracts in children on long-term corticosteroid therapy. Arch Ophthalmol 1965; 73:818-21.
3. Williamson J, Paterson RW, McGavin DD, Jasani MK, Boyle JA, Doig WM. Posterior subcapsular cataracts and glaucoma associated with long-term oral corticosteroid therapy. In patients with rheumatoid arthritis and related conditions. Br J Ophthalmol 1969; 53:361-72.
4. Urban RC Jr, Cotlier E. Corticosteroid-induced cataracts. Surv Ophthalmol 1986; 31:102-10.
5. Kaye LD, Kalenak JW, Price RL, Cunningham R. Ocular implications of long-term prednisone therapy in children. J Pediatr Ophthalmol Strabismus 1993; 30:142-4.
6. Greiner JV, Chylack LT Jr. Posterior subcapsular cataracts: histopathologic study of steroid-associated cataracts. Arch Ophthalmol 1979; 97:135-44.
7. Geley S, Fiegl M, Hartmann BL, Kofler R. Genes mediating glucocorticoid effects and mechanisms of their regulation. Rev Physiol Biochem Pharmacol 1996; 128:1-97.
8. Reichardt HM, Tronche F, Berger S, Kellendonk C, Schutz G. New insights into glucocorticoid and mineralocorticoid signaling: lessons from gene targeting. Adv Pharmacol 2000; 47:1-21.
9. Aranda A, Pascual A. Nuclear hormone receptors and gene expression. Physiol Rev 2001; 81:1269-304.
10. Webster JC, Cidlowski JA. Mechanisms of Glucocorticoid-receptor-mediated Repression of Gene Expression. Trends Endocrinol Metab 1999; 10:396-402.
11. Watanabe H, Kosano H, Nishigori H. Steroid-induced short term diabetes in chick embryo: reversible effects of insulin on metabolic changes and cataract formation. Invest Ophthalmol Vis Sci 2000; 41:1846-52.
12. Nishigori H, Lee JW, Yamauchi Y, Iwatsuru M. The alteration of lipid peroxide in glucocorticoid-induced cataract of developing chick embryos and the effect of ascorbic acid. Curr Eye Res 1986; 5:37-40.
13. Anderson EI, Wright DD, Spector A. The state of sulfhydryl groups in normal and cataractous human lens proteins. II. Cortical and nuclear regions. Exp Eye Res 1979; 29:233-43.
14. Lou MF, Dickerson JE Jr, Garadi R, York BM Jr. Glutathione depletion in the lens of galactosemic and diabetic rats. Exp Eye Res 1988; 46:517-30.
15. Reddy VN. Glutathione and its function in the lens--an overview. Exp Eye Res 1990; 50:771-8.
16. Bucala R, Fishman J, Cerami A. Formation of covalent adducts between cortisol and 16 alpha-hydroxyestrone and protein: possible role in the pathogenesis of cortisol toxicity and systemic lupus erythematosus. Proc Natl Acad Sci U S A 1982; 79:3320-4.
17. Manabe S, Bucala R, Cerami A. Nonenzymatic addition of glucocorticoids to lens proteins in steroid-induced cataracts. J Clin Invest 1984; 74:1803-10.
18. Jobling AI, Stevens A, Augusteyn RC. Binding of dexamethasone by alpha-crystallin. Invest Ophthalmol Vis Sci 2001; 42:1829-32.
19. Zhu XL, Sexton PS, Cenedella RJ. Characterization of membrane steroid binding protein mRNA and protein in lens epithelial cells. Exp Eye Res 2001; 73:213-9.
20. Southren AL, Gordon GG, Yeh HS, Dunn MW, Weinstein BI. Receptors for glucocorticoids in the lens epithelium of the calf. Science 1978; 200:1177-8.
21. Tchernitchiv A, Wenk EJ, Hernandez MR, Weinstein BI, Dunn MW, Gordon GG, Southren AL. Glucocorticoid localization by radioautography in the rabbit eye following systemic administration of 3H-dexamethasone. Invest Ophthalmol Vis Sci 1980; 19:1231-6.
22. Stokes J, Noble J, Brett L, Phillips C, Seckl JR, O'Brien C, Andrew R. Distribution of glucocorticoid and mineralocorticoid receptors and 11beta-hydroxysteroid dehydrogenases in human and rat ocular tissues. Invest Ophthalmol Vis Sci 2000; 41:1629-38.
23. Suzuki T, Sasano H, Kaneko C, Ogawa S, Darnel AD, Krozowski ZS. Immunohistochemical distribution of 11beta-hydroxysteroid dehydrogenase in human eye. Mol Cell Endocrinol 2001; 173:121-5.
24. Thompson EB. The structure of the human glucocorticoid receptor and its gene. J Steroid Biochem 1987; 27:105-8.
25. Rousseau GG, Baxter JD. Glucocorticoid receptors. Monogr Endocrinol 1979; 12:49-77.
26. Jobling AI, Augusteyn RC. Is there a glucocorticoid receptor in the bovine lens? Exp Eye Res 2001; 72:687-94.
27. Gupta V, Wagner BJ. Expression of the functional glucocorticoid receptor in mouse and human lens epithelial cells. Invest Ophthalmol Vis Sci 2003; 44:2041-6.
28. Wang-Su ST, McCormack AL, Yang S, Hosler MR, Mixon A, Riviere MA, Wilmarth PA, Andley UP, Garland D, Li H, David LL, Wagner BJ. Proteome analysis of lens epithelia, fibers, and the HLE B-3 cell line. Invest Ophthalmol Vis Sci 2003; 44:4829-36.
29. Fleming TP, Song Z, Andley UP. Expression of growth control and differentiation genes in human lens epithelial cells with extended life span. Invest Ophthalmol Vis Sci 1998; 39:1387-98.
30. van Venrooij WJ, Groeneveld AA, Bloemendal H, Benedetti EL. Cultured calf lens epithelium. II. The effect of dexamathasone. Exp Eye Res 1974; 18:527-36.
31. Lyu J, Kim JA, Chung SK, Kim KS, Joo CK. Alteration of cadherin in dexamethasone-induced cataract organ-cultured rat lens. Invest Ophthalmol Vis Sci 2003; 44:2034-40.
32. Reichardt HM, Tuckermann JP, Gottlicher M, Vujic M, Weih F, Angel P, Herrlich P, Schutz G. Repression of inflammatory responses in the absence of DNA binding by the glucocorticoid receptor. EMBO J 2001; 20:7168-73.
33. Hawse JR, DeAmicis-Tress C, Cowell TL, Kantorow M. Identification of global gene expression differences between human lens epithelial and cortical fiber cells reveals specific genes and their associated pathways important for specialized lens cell functions. Mol Vis 2005; 11:274-83 <http://www.molvis.org/molvis/v11/a32/>.
34. James ER, Fresco VM, Robertson LL. Glucocorticoid-induced changes in the global gene expression of lens epithelial cells. J Ocul Pharmacol Ther 2005; 21:11-27.
35. Hosack DA, Dennis G Jr, Sherman BT, Lane HC, Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Biol 2003; 4:R70.
36. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000; 25:25-9.
37. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001; 98:5116-21. Erratum in: Proc Natl Acad Sci U S A 2001; 98:10515.
38. Grewal A, Conway A. Tools for Analyzing Microarray Expression Data. JALA 2000; 5:62-64.
39. Engelbrecht Y, de Wet H, Horsch K, Langeveldt CR, Hough FS, Hulley PA. Glucocorticoids induce rapid up-regulation of mitogen-activated protein kinase phosphatase-1 and dephosphorylation of extracellular signal-regulated kinase and impair proliferation in human and mouse osteoblast cell lines. Endocrinology 2003; 144:412-22.
40. Brenier-Pinchart MP, Vigan I, Jouvin-Marche E, Marche PN, Pelet E, Gross U, Ambroise-Thomas P, Pelloux H. Monocyte chemotactic protein-1 secretion and expression after Toxoplasma gondii infection in vitro depend on the stage of the parasite. FEMS Microbiol Lett 2002; 214:45-9.
41. Shepard AR, Jacobson N, Fingert JH, Stone EM, Sheffield VC, Clark AF. Delayed secondary glucocorticoid responsiveness of MYOC in human trabecular meshwork cells. Invest Ophthalmol Vis Sci 2001; 42:3173-81.
42. Westrin A, Ekman R, Traskman-Bendz L. High delta sleep-inducing peptide-like immunoreactivity in plasma in suicidal patients with major depressive disorder. Biol Psychiatry 1998; 43:734-9.
43. Vogel P, Magert HJ, Cieslak A, Adermann K, Forssmann WG. hDIP--a potential transcriptional regulator related to murine TSC-22 and Drosophila shortsighted (shs)--is expressed in a large number of human tissues. Biochim Biophys Acta 1996; 1309:200-4.
44. Wu W, Chaudhuri S, Brickley DR, Pang D, Karrison T, Conzen SD. Microarray analysis reveals glucocorticoid-regulated survival genes that are associated with inhibition of apoptosis in breast epithelial cells. Cancer Res 2004; 64:1757-64.
45. Leclerc N, Luppen CA, Ho VV, Nagpal S, Hacia JG, Smith E, Frenkel B. Gene expression profiling of glucocorticoid-inhibited osteoblasts. J Mol Endocrinol 2004; 33:175-93.
46. Thompson EB, Webb MS, Miller AL, Fofanov Y, Johnson BH. Identification of genes leading to glucocorticoid-induced leukemic cell death. Lipids 2004; 39:821-5.
47. Wang JC, Derynck MK, Nonaka DF, Khodabakhsh DB, Haqq C, Yamamoto KR. Chromatin immunoprecipitation (ChIP) scanning identifies primary glucocorticoid receptor target genes. Proc Natl Acad Sci U S A 2004; 101:15603-8.
48. Duanmu Z, Kocarek TA, Runge-Morris M. Transcriptional regulation of rat hepatic aryl sulfotransferase (SULT1A1) gene expression by glucocorticoids. Drug Metab Dispos 2001; 29:1130-5.
49. Agbemafle BM, Oesterreicher TJ, Shaw CA, Henning SJ. Immediate early genes of glucocorticoid action on the developing intestine. Am J Physiol Gastrointest Liver Physiol 2005; 288:G897-906.
50. Planey SL, Abrams MT, Robertson NM, Litwack G. Role of apical caspases and glucocorticoid-regulated genes in glucocorticoid-induced apoptosis of pre-B leukemic cells. Cancer Res 2003; 63:172-8.
51. Schaaf MJ, Cidlowski JA. Molecular mechanisms of glucocorticoid action and resistance. J Steroid Biochem Mol Biol 2002; 83:37-48.
52. Shi X, Shi W, Li Q, Song B, Wan M, Bai S, Cao X. A glucocorticoid-induced leucine-zipper protein, GILZ, inhibits adipogenesis of mesenchymal cells. EMBO Rep 2003; 4:374-80.
53. Kino T, Nordeen SK, Chrousos GP. Conditional modulation of glucocorticoid receptor activities by CREB-binding protein (CBP) and p300. J Steroid Biochem Mol Biol 1999; 70:15-25.
54. Hawse JR, Hejtmancik JF, Horwitz J, Kantorow M. Identification and functional clustering of global gene expression differences between age-related cataract and clear human lenses and aged human lenses. Exp Eye Res 2004; 79:935-40.
55. Givant-Horwitz V, Davidson B, Goderstad JM, Nesland JM, Trope CG, Reich R. The PAC-1 dual specificity phosphatase predicts poor outcome in serous ovarian carcinoma. Gynecol Oncol 2004; 93:517-23.
56. Wu JJ, Bennett AM. Essential role for mitogen-activated protein (MAP) kinase phosphatase-1 in stress-responsive MAP kinase and cell survival signaling. J Biol Chem 2005; 280:16461-6.
57. Nicholl SM, Roztocil E, Davies MG. Urokinase-induced smooth muscle cell responses require distinct signaling pathways: a role for the epidermal growth factor receptor. J Vasc Surg 2005; 41:672-81.
58. Zhang BH, Tang ED, Zhu T, Greenberg ME, Vojtek AB, Guan KL. Serum- and glucocorticoid-inducible kinase SGK phosphorylates and negatively regulates B-Raf. J Biol Chem 2001; 276:31620-6.
59. Werle M, Schmal U, Hanna K, Kreuzer J. MCP-1 induces activation of MAP-kinases ERK, JNK and p38 MAPK in human endothelial cells. Cardiovasc Res 2002; 56:284-92.
60. Osaki M, Oshimura M, Ito H. PI3K-Akt pathway: its functions and alterations in human cancer. Apoptosis 2004; 9:667-76.
61. Krasil'nikov MA, Shatskaya VA, Stavrovskaya AA, Erohina M, Gershtein ES, Adler VV. The role of phosphatidylinositol 3-kinase in the regulation of cell response to steroid hormones. Biochim Biophys Acta 1999; 1450:434-43.
62. Roux PP, Blenis J. ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions. Microbiol Mol Biol Rev 2004; 68:320-44.
63. Chan A. Ras-MAPK pathways. Sci STKE 2005; 2005:tr5.
64. Clark AR, Lasa M. Crosstalk between glucocorticoids and mitogen-activated protein kinase signalling pathways. Curr Opin Pharmacol 2003; 3:404-11.
65. Muller P, Kietz S, Gustafsson JA, Strom A. The anti-estrogenic effect of all-trans-retinoic acid on the breast cancer cell line MCF-7 is dependent on HES-1 expression. J Biol Chem 2002; 277:28376-9.
66. Castella P, Sawai S, Nakao K, Wagner JA, Caudy M. HES-1 repression of differentiation and proliferation in PC12 cells: role for the helix 3-helix 4 domain in transcription repression. Mol Cell Biol 2000; 20:6170-83.
67. Beyaert R, Van Loo G, Heyninck K, Vandenabeele P. Signaling to gene activation and cell death by tumor necrosis factor receptors and Fas. Int Rev Cytol 2002; 214:225-72.
68. Weber KS, Nelson PJ, Grone HJ, Weber C. Expression of CCR2 by endothelial cells: implications for MCP-1 mediated wound injury repair and In vivo inflammatory activation of endothelium. Arterioscler Thromb Vasc Biol 1999; 19:2085-93.
69. Morland CM, Morland BJ, Darbyshire PJ, Stockley RA. Migration of CD18-deficient neutrophils in vitro: evidence for a CD18-independent pathway induced by IL-8. Biochim Biophys Acta 2000; 1500:70-6.
70. Mikosz CA, Brickley DR, Sharkey MS, Moran TW, Conzen SD. Glucocorticoid receptor-mediated protection from apoptosis is associated with induction of the serine/threonine survival kinase gene, sgk-1. J Biol Chem 2001; 276:16649-54.
71. Sriuranpong V, Borges MW, Ravi RK, Arnold DR, Nelkin BD, Baylin SB, Ball DW. Notch signaling induces cell cycle arrest in small cell lung cancer cells. Cancer Res 2001; 61:3200-5.
72. Choi J, Park SY, Joo CK. Hepatocyte growth factor induces proliferation of lens epithelial cells through activation of ERK1/2 and JNK/SAPK. Invest Ophthalmol Vis Sci 2004; 45:2696-704.
73. Kotone-Miyahara Y, Yamashita K, Lee KK, Yonehara S, Uchiyama T, Sasada M, Takahashi A. Short-term delay of Fas-stimulated apoptosis by GM-CSF as a result of temporary suppression of FADD recruitment in neutrophils: evidence implicating phosphatidylinositol 3-kinase and MEK1-ERK1/2 pathways downstream of classical protein kinase C. J Leukoc Biol 2004; 76:1047-56.
74. Fritz EA, Jacobs JJ, Glant TT, Roebuck KA. Chemokine IL-8 induction by particulate wear debris in osteoblasts is mediated by NF-kappaB. J Orthop Res 2005; 23:1249-57.
75. Bian ZM, Elner SG, Yoshida A, Elner VM. Differential involvement of phosphoinositide 3-kinase/Akt in human RPE MCP-1 and IL-8 expression. Invest Ophthalmol Vis Sci 2004; 45:1887-96.
76. Kobayashi T, Deak M, Morrice N, Cohen P. Characterization of the structure and regulation of two novel isoforms of serum- and glucocorticoid-induced protein kinase. Biochem J 1999; 344:189-97.
77. Shao J, Evers BM, Sheng H. Roles of phosphatidylinositol 3'-kinase and mammalian target of rapamycin/p70 ribosomal protein S6 kinase in K-Ras-mediated transformation of intestinal epithelial cells. Cancer Res 2004; 64:229-35.
78. Han YS, Bang OS, Jin EJ, Park JH, Sonn JK, Kang SS. High dose of glucose promotes chondrogenesis via PKCalpha and MAPK signaling pathways in chick mesenchymal cells. Cell Tissue Res 2004; 318:571-8.
79. Juliano RL. Signal transduction by cell adhesion receptors and the cytoskeleton: functions of integrins, cadherins, selectins, and immunoglobulin-superfamily members. Annu Rev Pharmacol Toxicol 2002; 42:283-323.