Gene Expression and miR Profiles of Human Corneal Fibroblasts in Response to Dexamethasone

Abstract

Purpose.: Dexamethasone (DEX) is commonly used as a therapeutic agent for various ocular inflammatory diseases; however, its effect on resident naive cells is unknown. In this study, genome microarray and microRNA (miR) analyses were used to evaluate the global gene and miR expression of human corneal fibroblasts (HKFs) in response to treatment with DEX.

Methods.: Primary HKFs from three donors were treated with DEX for 16 hours. Treated and untreated cells were snap frozen for microarray and miR array analyses. Genes with a more than threefold change were classified into gene families using the DAVID web-based classification database, and six of these genes were validated using quantitative real-time PCR. Five miRs were also validated using miR-detection assays.

Results.: Of the 41,093 genes examined, 261 were upregulated and 123 were downregulated greater than threefold after DEX treatment. Real-time PCR confirmed upregulation of six genes, including oculocutaneous albinism II (OCA2), angiopoietin-like 7 (ANGPTL7), neuron navigator 2 (NAV2), neurofilament light chain polypeptide (NEFL), solute carrier family 16/member 12 (SLC16A12), and serum amyloid A1 (SAA1). Expression of several miR including miR-16, -21, and -29C were upregulated, whereas miR-100 was downregulated in fibroblasts by DEX.

Conclusions.: DEX can greatly change the global gene and miR profile of HKFs. DEX not only downregulates inflammatory genes, but can also induce expression of angiogenic and inflammatory genes. In addition, DEX may exert posttranscriptional gene regulation through miRs. These data support a complex role for DEX-induced changes in resident cells that may have implications in the clinical management of corneal inflammation with topical glucocorticoids.

Glucocorticoids (GCs) are important endocrine regulators of a wide range of physiological and pathologic responses. These include metabolism, salt and water balance, embryogenesis, behavior, neurobiology, inflammatory response, and programmed cell death together with cell proliferation and differentiation. ¹ At the tissue level, GCs are regulated by two enzymes, 11-beta hydroxysteroid dehydrogenase type 1 (11β-HSD1) and 11β-HSD2. The 11β-HSD1 reactivates cortisol (active glucocorticoid) from cortisone (inactivate glucocorticoid). ² This reaction requires nicotinamide adenine dinucleotide phosphate (NADPH) which is provided by the enzyme hexose-6-phosphate dehydrogenase (H6PDH) and the cellular response is mediated by the glucocorticoid receptor (GR). 11β-HSD2 inactivates cortisol to cortisone in secretory tissues where it protects the mineralocorticoid receptor from cortisol excess. Importantly, cortisol provides a “brake” during the initial inflammatory response as well as promoting resolution of inflammation. ^3,4  

Synthetic GCs such as DEX, prednisolone, and prednisone, are three of the most widely used drugs in the treatment of ocular inflammatory diseases. At a molecular level, GCs act through the glucocorticoid receptor (GR), a member of the nuclear hormone receptor superfamily of ligand-activated transcription factors. GRs can mediate transactivation of target genes by binding to their promoter region and can also interact with a multitude of transcription factors such as c-Jun and nuclear factor-kB (NF-κB). ⁵  

Stromal fibroblasts have multiple pathophysiological functions within the cornea, including maintenance of the extracellular matrix and corneal transparency; immunoprotection after injury or infection, and modulation of inflammation. The effect of GC treatment on human corneal fibroblasts is not well characterized; however, the control of gene expression is vital to any downstream response. In addition to direct gene expression, novel controlling molecules, microRNAs (miRs), have recently been described. miRs are processed from the primary transcripts by Drosha and DGCR8 to generate pre-miRs. These pre-miRs are then transported to the cytosol by exportin 5. Further processing by DICER leads to the formation of miR duplex, which is loaded onto the RISC complex. RISC is transported by importin 8 to its cognate mRNA, leading to repression of the target. ⁶  

In this study, we used a large-scale microarray to identify the human corneal fibroblast (HKF)–specific transcriptional effects of dexamethasone (DEX) and found that DEX can greatly change the global gene and miR profiles of HKFs. These data may have important implications for DEX treatment in some corneal disorders. 

Materials and Methods

Primary Human Corneal Fibroblast Culture

The collection of deceased donor peripheral corneal tissue (transplant waste) from penetrating and lamellar keratopathy surgical procedures for in vitro investigation were approved by the local research ethics committee, and all experiments were performed in accordance with the Declaration of Helsinki. According to U.K. National Guidelines for Organ Donation, only those donors who meet none of the absolute exclusion criteria, including active transmissible disease or infection, Creutzfeld-Jakob disease, intravenous drug abuse, or neurodegenerative disorders, and who have tissues that are suitable for transplantation can donate organs and tissues. Additional exclusions for eye donation include previous ocular surgery, inflammation, and tumors such as retinoblastoma. Only those tissues for which the donor or the donor's family had provided informed consent for research were used. Corneoscleral tissue was transported in tissue bank organ culture medium, and corneal tissue was dissected from the corneoscleral rims in a class II laminar flow culture hood. The corneal epithelial layer was removed from the stroma using Dispase enzyme digestion, and the endothelial layer was stripped under direct visualization with the aid of trypan blue. The remaining corneal stroma was chopped into small pieces and cultured in a 25-cm² flask with 2 mL RPMI with 50,000 U/L penicillin, 50,000 μg/L streptomycin, 1% (vol/vol) nonessential amino acid, 1% (vol/vol) sodium pyruvate, and 10% (vol/vol) fetal calf serum (Fibro medium; Invitrogen-Gibco, Grand Island, NY). Cells were grown at 37°C in a humid environment containing 5% CO₂, and the cell culture medium was changed every 2 to 3 days. Confluent cells were trypsinized, and only cells with up to four generations were used for subsequent experiments. Fibroblast phenotype was confirmed by positive immunostaining with anti-vimentin and anti-5B5 antibodies. 

Preparation of Cells for Microarray Analysis

To avoid confusion between estrogen–glucocorticoid interactions and androgen–glucocorticoid interactions, we restricted our studies to HKFs generated by three male donors. For each donor, two 25-cm² flasks of HKFs were grown until 90% confluence. One flask of cells was treated with 100 nM DEX in serum-containing medium (Fibro-medium; Invitrogen-Gibco) for 16 hours at 37°C with 5% CO₂, and the other flask of cells was used as a control in medium alone. After they were washed with cold PBS, the cells were trypsinized and centrifuged for 5 minutes at 1000 rpm. Cell pellets were snap-frozen by complete submersion in liquid nitrogen for 30 seconds. The frozen cell pellets were then shipped on dry ice to Miltenyi Biotec (Bergisch Gladbach, Germany) for two-color, whole-genome microarray and analyses (miRXxplore MicroRNA; Agilent, Palo Alto, CA). Control and DEX-treated samples from each donor were individually labeled with cyanine-3 (Cy3) and cyanine-5 (Cy5) and then hybridized on the same microarray. This experimental design enabled a direct comparison between the control and experimental samples, minimizing the variability between slides. 

Bioinformatic Analysis

Mean signal and local background intensities were obtained for each spot (ImaGene software; Biodiscovery Ltd., Auckland, New Zealand). Normalized and validated signal ratios are provided by an array analyzer (PIQOR; Miltenyi Biotec) using data from the software analysis. For each gene, a median value was calculated for the expression changes across the three donors. Genes exhibiting a greater than threefold change in expression were clustered with TIGR Multiple Experiment Viewer (TMEV) software for hierarchical clustering. ⁷ The DAVID web-based classification database (Laboratory of Immunopathogenesis and Bioinformatics, National Cancer Institute, Frederick, MD) was used to classify each cluster into gene families. ^8,9  

GC-Related Genes

To test the microarray analysis, the expression changes of known GC signaling–related genes were analyzed. These included (11-beta) hydroxysteroid dehydrogenase 1 (HSD11B1), (11-beta) hydroxysteroid dehydrogenase 2 (HSD11B2), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor alpha (NFKBIA), interferon gamma receptor 1 (IFNGR1), heat shock protein 90 kDa alpha class B member 1 (HSP90AB1), signal transducer and activator of transcription 5A (STAT5A), insulin-like growth factor 1 (IGF1), insulin-like growth factor 2 (IGF2), chemokine (C-X3-C motif) ligand 1 (CX3CL1), and Fas ligand (FASLG). 

Whole-Genome Microarray Validation

Genes that showed a greater than threefold change in expression from all three donors were examined using quantitative real-time PCR (qRT-PCR) for validation of the genome microarray data. The mRNAs of DEX-treated HKFs and vehicle cells from three new donors were extracted (RNeasy Mini Kit; Qiagen, Crawley, UK) according to the manufacturer's instructions. The cDNAs were reverse transcribed (TaqMan reverse transcription reagents; Applied Biosystems, Inc. [ABI], Warrington, UK). The primers and probes for the genes were purchased from ABI. Real-time PCR was performed (model 7700 genetic analyzer; Perkin-Elmer/ABI) in 20-μL reactions on 96-well plates in reaction buffer containing 10 μL universal PCR mastermix (TaqMan; ABI), 1 μL of mixed probe and primers, 1 μL of cDNA, and 8 μL of nuclease-free water. All reactions were single-plexed with primers specific for 18S rRNA, as an internal reference, and primers specific for the genes of interest. Data were obtained as Ct values (the cycle number at which logarithmic PCR plots cross a calculated threshold line) and used to determine dCt values (dCt = Ct of the target gene minus Ct of the internal reference, 18S). All target gene probes were labeled with the fluorescent label 6-carboxyfluorescein and 18S with VIC. To exclude potential bias caused by averaging data that had been transformed through the equation ^2−ΔΔCt, all statistics were performed at the Ct stage. Data are expressed as fold expression change in DEX-treated cells compared with control cells. 

miR Validation

Total mRNA was extracted from HKFs derived from three different male donors treated with either 100 nM DEX in serum-containing medium (Fibro-medium; Invitrogen-Gibco) or in the medium alone for 16 hours. Both miR labeling and detection kits were purchased from Marligen Biosciences, Inc. (Ijamsville, MD). The miRs analyzed including miR-16, -21, -27B, -29B, -29C, and -100 were labeled and detected with a plate reader (100 System; Luminex, Austin, TX), according to the manufacturer's instructions. In brief, the 3′ end of the RNA is poly(A) tailed followed by the ligation of biotinylated dendrimers (3DNA; Genisphere, Hatfield, PA) containing 15 biotins to provide signal amplification. The biotinylated samples were incubated with mixed fluorescently dyed beads. Each bead is coupled with a unique probe that recognizes a specific miR. Finally, the biotinylated miR hybridized to the beads were incubated with streptavidin-phycoerythrin (SAPE) and detected with the plate-reading system (100 System; Luminex). Water was used as a background control, and mixed synthetic miR-16, -21, and -100 (Marligen Biosciences, Inc.) were used as positive references. The median fluorescence intensity (MFI) of all examined samples was corrected with water control and the fold expression change in MFI was calculated, comparing DEX-treated cells with control cells. 

Potential miR Targets

The Enright Programme (Sanger Centre, Hinxton, UK) was used to classify miR target genes for each miR identified via the microarray. ¹⁰ The resultant genes were grouped according to function and enrichments scores which were derived using DAVID. 

Results

Global Transcriptional Changes of Primary Human Corneal Fibroblasts after DEX Treatment

The purity of the corneal fibroblast populations was confirmed by immunohistochemistry, which was positive for vimentin and 5B5, but negative for cytokeratin 3. By comparison, primary human corneal epithelial cells were positive for CK3 and vimentin, but negative for 5B5 (Supplementary Fig. S1). Of the 41,093 genes and transcripts examined in the HKFs, 23,186 were upregulated and 17,907 were downregulated after DEX treatment. Of these genes, 261 were upregulated and 123 were downregulated greater than threefold (Supplementary Table S1). 

Classification of Genes with Altered Signal Levels after DEX Treatment

Both threefold upregulated and downregulated genes were clustered into two main family gene groups, as shown in the hierarchal clustering diagrams in Figure 1. DAVID classification showed that those genes that were upregulated were mainly involved in biogenesis, development, cell adhesion, metabolism, inflammatory response, and angiogenesis (Fig. 1A), and those genes that were downregulated were involved in cellular functions including cellular processes, development, immune response, homeostasis, biological regulation, and metabolism (Fig. 1B; Supplementary Table S2). 

Figure 1.

 

Hierarchal diagrams of those genes exhibiting a threefold change in expression. Both threefold upregulated (A) and downregulated genes (B) were clustered into two main family gene groups by TEMV software. ⁷ DAVID classified the upregulated genes into cellular functions, including biogenesis, development, cell adhesion, metabolism, inflammatory response, and angiogenesis. The downregulated genes were mainly involved in cellular process, development, immune response, homeostasis, biological regulation, and metabolism.

The microarray data after DEX treatment showed that 11-beta hydroxysteroid dehydrogenase 1 (HSD11B1) was upregulated by 4.24-fold, and nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor alpha (NFKBIA) and interferon gamma receptor 1 (IFNGR1) were also upregulated more than twofold (i.e., 2.04 and 2.59 respectively). However, there was no effect on the expression of the GC-related genes 11-β-hydroxysteroid dehydrogenase 2 (HSD11B2), heat shock protein 90kDa alpha class B member 1 (HSP90AB1), signal transducer and activator of transcription 5A (STAT5A), insulin-like growth factor 1 (IGF1), insulin-like growth factor 2 (IGF2), chemokine (C-X3-C motif), ligand 1 (CX3CL1), and Fas ligand (FASLG). 

Validation of the Microarray Results via qRT-PCR

To validate the microarray data, the expression of six genes that were universally upregulated in all three donors—oculocutaneous albinism II (OCA2), angiopoietin-like 7 (ANGPTL7), neuron navigator 2 (NAV2), neurofilament light chain polypeptide (NEFL), solute carrier family 16/member 12 (SLC16A12), and serum amyloid A1 (SAA1)—were compared with new populations of HKFs in the presence or absence of 100 nM DEX by real-time quantitative (q)RT-PCR. The median expression changes for these genes were 214.82-, 18.89-, 2.15-, 18.66-, 15.64-, and 8.51-fold for OCA2, ANGPTL7, NAV2, SLC16A12, NEFL, and SAA1, respectively. Although these data validated the microarray findings, the absolute expression change of OCA2 after DEX treatment was much greater (qRT-PCR 214.82-fold vs. microarray 32.71-fold), and that of NAV2 was less (qRT-PCR 2.15-fold vs. microarray 18.85-fold) than that observed in the genome microarray (Fig. 2, Supplementary Table S3). 

Figure 2.

 

Validation of microarray data via qRT- PCR. Real-time qRT- PCR revealed that OCA2, ANGPTL7, NAV2, SLC16A12, NEFL, and SAA1 were all upregulated when human corneal fibroblasts were treated with DEX. This result is in agreement with the genome microarray, although the expression change of OCA2 was much greater and that of NAV2 was much lower than was observed in the microarray. Data are presented as the median (range, x–y ± SEM).

miR Analysis of Primary HKFs after Treatment with DEX

A total of 1376 miRs were examined by microarray screening, to assess the changes of miRs in HKFs after DEX treatment. An miR analysis of DEX-treated HKFs compared to untreated cells is shown in Figure 3, and the 12 miRs with the greatest changes, defined as >1.5-fold change in miR expression in at least one of the three donor cells after DEX treatment, are detailed in Figure 4. Of these, miR-100 expression was downregulated by >1.5-fold, and miR-29c, -886, and -1260 were upregulated in excess of 1.5-fold (2-, 1.75-, and 1.8-fold, respectively). These data support the ability of DEX to alter miR expression in HKFs. 

Figure 3.

 

miR expression after DEX treatment. Expression of 12 miRs was defined to be altered by DEX treatment by a minimum of 1.5-fold by miR microarray.

Figure 4.

 

miR expression after DEX treatment. The median changes for upregulated miR-29C, -886, and -1260 were greater than 1.5-fold and miR-100 was downregulated by more than 1.5-fold.

Validation of miR Changes by miR Detection Assay

To validate the data from the array, an miR detection assay was performed to compare the levels of miR-16, -21, -29B, -29C, and -100 between DEX-treated HKFs and control cells from three new donors. Although all examined miRs were upregulated in response to DEX treatment, only the levels of miR-16, -21, and -29c increased greater than 1.5-fold (Fig. 5). miR gene targets of interest were obtained from the Enright database (www.miRBase.org, Faculty of Life Sciences, University of Manchester, Manchester, UK). The resultant genes were separated into functional groups, and enrichment scores were obtained by DAVID. The results showed that the highest association for miR-16 was for genes involved in intracellular and organelle function; for miR-21, inflammation and responses to wounding or stress; and for miR-29C, extracellular matrix remodelling (Fig. 6). As expected, none of the genes upregulated by DEX in the global microarray were targets for the miRs found to be upregulated by treatment. These data therefore reflect that DEX treatment of HKFs potentially involves pathways important in both the pathogenesis of infectious keratitis and also the healing phase of the disease where therapeutic GCs are essential for clinical management. 

Figure 5.

 

Validation of miR data via miR detection assays. miR detection assay showed that DEX positively regulated all miRs examined, although only the changes in miR-16, -21, and -29C were greater than 1.5-fold.

Figure 6.

 

Enrichment scores for DEX-regulated miRs. Gene targets of miRs of interest were segregated into functional groups, and enrichment scores were obtained by DAVID, identifying metabolic processes (miR-100), intracellular organelle function (miR16), extracellular matrix remodeling (miR29b and -c), and inflammation and response to wounding (miR21).

Discussion

In this article, we have described the global gene expression changes induced by DEX treatment of HKFs, providing a comprehensive evaluation of the cellular processes in the cornea targeted by DEX treatment. In addition, we have also described changes in miR profile and offer a view on posttranscriptional regulation by DEX of genes in HKFs. To our knowledge, this is the first demonstration of the effect of DEX on HKFs by microarray and miR analyses. Consistent with others, we clearly showed that GCs are actively involved in the transcription, whether by upregulation or downregulation, of many genes. Of the 41,093 genes examined, 261 were upregulated and 123 were downregulated greater than threefold after DEX treatment. These genes were predominantly involved in cell development, cell adhesion, homeostasis, metabolism, and inflammation response, which reflects the diverse and potentially tissue-specific nature of GC action. Validation of new HKF samples confirmed increased expression of six consistently upregulated genes: OCA, ANGPTL7, NAV2, NEFL, SLC16, and SAA1. 

Of these, the ANGPTL family of proteins have high sequence and structural homology to the angiopoietins, and these proteins have been demonstrated to be involved in blood vessel formation including corneal angiogenesis. ¹¹ They also play an important role in lipid metabolism by inhibition of phospholipid lipase. ¹² A previous study indicated that corneoscleral explants are capable of secreting ANGPTL7 protein when stimulated by DEX and that stable overexpression of ANGPTL7 in a transfected immortalized trabecular meshwork cell line increased type 1 collagen secretion. Immunohistochemical analysis located ANGPTL7 immunoreactivity to the corneal stroma near the limbus, but with strong staining in the region of Bowman's layer, Schlemm's canal, and trabecular and scleral tissue. In addition, increased expression of ANGPTL7 was seen in analyses of aqueous humor taken from glaucoma patients versus healthy controls. ¹³ These data, taken together, suggest a link of ANGPTL7 expression with fibrosis and remodelling in TM cells, thereby increasing aqueous outflow facility. It is possible that similar pathways are involved in corneal cells, enabling corneal matrix remodelling after episodes of ulcerative keratitis. 

By contrast, the human acute-phase protein serum amyloid A (A-SAA), encoded by the SAA1 and SAA2 genes, is induced by proinflammatory cytokines during the acute-phase response to infection or injury. The regulation of human A-SAA by GCs is tissue and cell specific. In certain cell types, GCs greatly enhance the induction of SAA1 mRNA when prestimulated with proinflammatory cytokines, whereas in other cell types where SAA1 is constitutively expressed, GCs alone can upregulate A-SAA expression. ^14,15 Several functions of SAA1 have been reported. It can induce extracellular matrix degrading enzymes, which are important in repair processes after tissue damage, and can act as a chemoattractant for leukocytes. ^16,17 It is therefore possible that the synergistic effect of DEX and SAA1 provide a potent wound-healing function in the context of the human cornea, a process that may contribute to the preservation of corneal clarity by protecting corneal tissue from immune-mediated damage. 

In our study, OCA2 was upregulated to the highest level after DEX stimulation of HKFs. This gene was not classified into any of the cellular process or functional pathways. The protein encoded by OCA2 is believed to be an integral membrane protein involved in small molecule transport, specifically tyrosine: a precursor of melanin. ¹⁸ Mutations in this gene result in type 2 oculocutaneous albinism (OCA), a group of inherited disorders of melanin biosynthesis characterized by a generalized reduction in pigmentation of hair, skin, and eyes. ¹⁹ Its role in HKFs or human corneal pathophysiology, remains unknown. 

Although to date no microarray analyses evaluating the effects of DEX on HKFs have been conducted, similar experiments have been performed on human trabecular meshwork cells, highlighting a gene expression pattern analogous to that of the DEX-treated HKFs shown in our study. Several genes—OCA2, ANGPTL7, ANGPTL4, and SAA1—were strongly upregulated in both cell types, whereas γ-aminobutyric acid A receptor β1 (GABRB1) and period homolog3 (Per3), involved in ion channel activity and circadian rhythms, respectively, were downregulated in trabecular cells but not in our corneal fibroblasts. ²⁰  

It is important to note that variation of gene regulation was observed among the HKF cell lines generated from our three corneas of deceased donors and that this variation may be due to the complexities of GC signaling pathways and/or GR heterogeneity. GCs signal through genomic and nongenomic pathways. The well-known genomic actions of GCs are mediated by a direct interaction between a ligand-activated GR complex that translocates into the nucleus, dimerizing before binding to GC response elements found in the upstream regulatory regions of target genes, including AP-1 and NF-κB. ²¹ Nongenomic GC actions are less well defined, although their rapid actions control intracellular signaling cascades through other components of the multiprotein complex. ²²  

Another caveat regarding these data is the use of a single time point for analysis. Microarray analysis of keratocytes, fibroblasts, and myofibroblasts prepared from mouse corneas in serum-poor, serum-rich, and serum-TGFβ culture conditions, respectively, showed a wound-healing profile, indicated by (1) an upregulation of acute-phase responses, angiogenesis, and extracellular matrix remodelling in corneal fibroblasts; (2) an induction of macrophage-like gene expression (CD68, SAA3, CCL2, CCL7, CCL9, MMP3, and MMP12) in cultured keratocytes; and (3) late healing, together with profibrogenic TGFβ-responsive changes in the myofibroblasts. These data suggest that stromal cells have a dual role of first, corneal maintenance, and second, barrier protection via macrophage function. ²³ Because of the possibility of variation between the sexes and estrogen interaction with GCs, our study was restricted to HKFs derived from deceased male donors, and this too, may have had an influence on the data. 

Regardless of the donor variation, consistency in GC-specific target genes was seen. These included a twofold upregulation of NFKBIA and IFNGR1 and a 4.24-fold increase in HSD11B1 after treatment with DEX. The upregulation of 11B-HSD1 gene expression by DEX suggests that GCs have the ability to produce positive feedback on the induction of 11B-HSD1 in HKFs. We have also shown that the expression of proinflammatory genes, such as IL1a, IL1b, IL17RB, and IL11 in HKFs was attenuated by DEX, whereas inhibitory κBα (IκBα; NFKBIA) was upregulated twofold, the latter supporting evidence from a previous study identifying NFKBIA as one of the earliest DEX-induced genes in epidermal keratinocytes. ²⁴  

At the level of posttranscriptional control, gene regulation is typified by a large and growing category of 19 to 22 nucleotide noncoding RNAs, known as miRs, which have emerged and function as repressors in all known genomes. We have shown that 12 such miRs are regulated by DEX in HKFs. For example, miR-29 family members have been shown to target mRNA-encoding proteins involved in fibrosis and are downregulated in heart tissue close to the anatomic site of a myocardial infarction, leading to increased fibrosis. ²⁵ Similarly, transfection of miR-29b into human trabecular meshwork cells, results in the downregulation of genes encoding components of the extracellular matrix. On exposure to chronic oxidative stress, downregulation of miR-29b expression is seen in human trabecular meshwork, together with increased expression of extracellular matrix modeling genes. ²⁶ Conversely, our data show increases in miR-29b expression in DEX-treated HKFs, and we postulate that this is likely to inhibit matrix protein deposition and fibrosis, thereby maintaining corneal clarity and preserving sight. 

miR-21 mediates induction of a contractile phenotype in human vascular smooth muscle cells by TGFβ and BMP, by downregulating PDCD4, which acts as a negative regulator of smooth muscle contractile genes. TGFβ and BMP induce the rapid induction of mature miR-21. ²⁷  

In summary, GCs are the first-line treatment for a wide range of autoimmune and inflammatory ocular disorders, many by topical administration. Our results show the existence of a complex interaction of DEX with several pathways of HKF function. GCs may increase the transcription of genes coding for anti-inflammatory proteins such as interleukin (IL)-10, IL-1 receptor antagonist (IL-1ra), and neutral endopeptidase, but this is insufficient to explain the powerful immunosuppressive action of GCs, as most inflammatory genes regulated by GCs do not carry a classic binding site for regulation by the cognate GRα. ²⁸ The demonstration that DEX mediates expression of several miRs, identifies novel targets for further investigation into GC function, specifically, miR control of cell cycle apoptosis, immune defense, and tissue remodelling pathways, which are particularly relevant to our understanding GC regulation of corneal inflammation. 

Supplementary Materials

Figure sf01, DOC - Figure sf01, DOC 

Table st01, DOC - Table st01, DOC 

Table st02, DOC - Table st02, DOC 

Table st03, XLS - Table st03, XLS 

Footnotes

 Supported by Grant 2006-01 from Guide Dogs for the Blind, UK, and The Birmingham Eye Foundation.

Footnotes

 Disclosure: L. Liu, None; E.A. Walker, None; S. Kissane, None; I. Khan None; P.I. Murray, None; S. Rauz, None; G.R. Wallace, None