Myocilin expression is regulated by retinoic acid in the trabecular meshwork-derived cellular environment

Experimental Eye Research 155 (2017) 91e98

Contents lists available at ScienceDirect

Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer

Research article

Myocilin expression is regulated by retinoic acid in the trabecular meshwork-derived cellular environment
cile Prat a, 1, Corinne Belville a, b, Aure
lie Comptour a, Geoffroy Marceau a, c, Ce
de
ric Chiambaretta a, c, d, Vincent Sapin a, c, *, Loïc Blanchon a Gael Clairefond a, Fre a EA7281 e Retinoids, Reproduction Developmental Diseases, School of Medicine, Clermont Universit
e, Universit
e d'Auvergne, F-63000 Clermont-Ferrand, France b GReD, CNRS UMR6293-Clermont Universit
e-INSERM U1103, Universit
e d'Auvergne, F-63000 Clermont-Ferrand, France c Biochemistry and Molecular Biology Department, CHU Clermont-Ferrand, F-63000 Clermont-Ferrand, France d CHU Clermont-Ferrand, Ophthalmology Department, F-63000 Clermont-Ferrand, France

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 September 2016 Received in revised form 17 January 2017 Accepted in revised form 25 January 2017 Available online 30 January 2017

Glaucoma is the leading cause of irreversible blindness and is usually classified as angle closure and open angle glaucoma (OAG). Primary open angle glaucoma represents the most frequent clinical presentation leading to ganglion cell death and optic nerve degeneration as a main consequence of an intraocular pressure’ (IOP) increase. The mechanisms of this IOP increase in such pathology remain unclear but one protein called Myocilin could be a part of the puzzle in the trabecular meshwork (TM). Previously described to be transcriptionally regulated by glucocorticoids, the comprehension of the trabecular regulation of Myocilin’ expression has only weakly progressed since 15 years. Due to the essential molecular and cellular implications of retinoids’ pathway in eye development and physiology, we investigate the potential role of the retinoic acid in such regulation and expression. This study demonstrates that the global retinoids signaling machinery is present in immortalized TM cells and that Myocilin (MYOC) expression is upregulated by retinoic acid alone or combined with a glucocorticoid cotreatment. This regulation by retinoic acid acts through the MYOC promoter which contains a critical cluster of four retinoic acid responsive elements (RAREs), with the RARE-DR2 presenting the strongest effect and binding the RARa/RXRa heterodimer. All together, these results open up new perspectives for the molecular understanding glaucoma pathophysiology and provide further actionable clues on Myocilin gene regulation. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Glaucoma Myocilin Trabecular meshwork Retinoids Retinoic acid Gene regulation

1. Introduction The Myocilin (MYOC) gene was independently identified from forward genetic approaches and in vitro expression studies. Indeed, in early 1990th, linkage study in a large family of juvenile-onset open-angle glaucoma (OAG) mapped a first locus designated GLC1A on the long arm of the chromosome 1 (Sheffield et al., 1993).


dicale, 4R3, Faculte
de * Corresponding author. Laboratoire de Biochimie Me
decine, 28 place Henri-Dunant, BP38, F-63001 Clermont-Ferrand Cedex, France. Me E-mail addresses: [email protected] (C. Prat), [email protected] (C. Belville), [email protected] (A. Comptour), geoffroy.marceau@ udamail.fr (G. Marceau), [email protected] (G. Clairefond), [email protected] (F. Chiambaretta), vincent.sapin@udamail. fr (V. Sapin), [email protected] (L. Blanchon). 1 Present address: Experimental Ophthalmology, HUG, Geneva, Switzerland. http://dx.doi.org/10.1016/j.exer.2017.01.006 0014-4835/© 2017 Elsevier Ltd. All rights reserved.

Then, the genes of this locus were screened and pathogenic variants in the Myocilin gene were found to segregate with juvenile and adult-onset OAG (Stone et al., 1997). Otherwise, because MYOC expression was first isolated by differential library screening of trabecular meshwork (TM) cell cultures after glucocorticoids exposure, MYOC was transiently referred as TIGR, standing for Trabecular meshwork-Induced Glucocorticoid Response (Nguyen et al., 1998; Polansky et al., 1997). The Myocilin gene, MYOC (MIM: *601652), is expressed in most of the human eye compartments. The resulting 504 amino acid secreted glycoprotein is detected in cornea, trabecular meshwork, aqueous humor, iris, ciliary body, choroid, sclera, retina and the axons of optic nerve ganglion cells (Karali et al., 2000). Since pioneer induction studies in TM cells (for example see, Polansky et al., 1997; Kirstein et al., 2000), the understanding of the tissue specific regulation of MYOC expression has progressed. Steroids treatment is associated

92

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

with increased levels of wild type Myocilin in trabecular meshwork cells, both in vitro and in vivo (Polansky et al., 1997; Clark et al., 2001; Ishibashi et al., 2002; Rozsa et al., 2006) whereas prostaglandin F2a decreases Myocilin expression (Lindsey et al., 2001). Interestingly, MYOC has been shown to be a delayed glucocorticoidresponsive gene in human trabecular meshwork cells (Shepard et al., 2001). Because of the cellular vitamin A final effectors (nuclear receptors belonging to the same superfamily than the glucocorticoids ones) are also upregulated by dexamethasone treatment (Ishibashi et al., 2002) and considering there absolute requirement for harmonious eye development as demonstrated by KO studies realized in mice (Kastner et al., 1994), they could be candidate actors of the delayed response. Retinoids signaling is mediated by the direct interactions of heterodimerized Retinoic Acid (RARs) and Retinoid X (RXR) nuclear receptors with specific consensus sequences (Retinoic Acid Responsive Elements; RAREs) in the promoter of a large set of genes. The vitamin A pathway displays pleiotropic functions such as cell differentiation, proliferation and apoptosis (Bastien and Rochette-Egly, 2004; Mark, 2007) that are essential for eye development and function (Cvekl and Wang, 2009; Samarawickrama et al., 2015). Together, these points lead us to focus on the interplay between the retinoids pathway and the Myocilin gene regulation in trabecular meshwork. Using a wellestablished TM derived cell line (TM5), we establish, that RAR/ RXR heterodimers trigger specific RARE on MYOC promoter and that retinoic acid and dexamethasone have a synergistic effect on MYOC induction. To the best of our knowledge, this is the first study demonstrating that retinoic acid signaling is able to regulate directly MYOC expression. These findings open up new perspectives for the understanding of POAG pathophysiology and the testing of alternative therapeutic strategies in this pathology. 2. Material and methods 2.1. Bioinformatics analysis MatInspector from Genomatix® program (http://www. genomatix.de) was used for in silico analysis to screen potential binding sites (RARE) for retinoic acid receptors. Briefly, the promoter sequence of the Myocilin gene was obtained from Ensembl (ENSG00000034971). The sequence comprised between nucleotides þ1 and 4500 was submitted to Matinspector by defining Homo sapiens transcription factor binding sites and V$RXRF matrix. Matrix similarities with consensus binding site for RAR/RXR heterodimers were kept for the rest of the study. 2.2. Chemicals Dimethyl sulfoxide (DMSO), all-trans retinoic acid (atRA) and dexamethasone were purchased from Sigma® (St Quentin Fallavier, France). 2.3. Cells culture TM5 immortalized cell line was kindly provided by Alcon® research ltd (Fort Worth, TX, USA). TM5 cells were subcultured at 37  C under 5% of CO2 in DMEM high glucose, with glutamax, with pyroxidine-HCL without sodium pyruvate supplemented with 2 mM of L-glutamine, 100 units of penicillin and 0.1 mg streptomycin (all products were purchased from Thermo-Fisher scientific (Saint-Aubin, France)). 2.4. RT-PCR and qPCR experiments Total RNA was extracted from TM primary cells or TM5 cells

using TRIZOL® after atRA (106M) or DMSO treatment for 24, 48 or 72 h (Thermo-Fisher scientific (Saint-Aubin, France)). cDNA was synthetized using a superscript III first strand synthesis system for RT-PCR (Thermo-Fisher scientific (Saint-Aubin, France)). PCR experiments were done using specific oligonucleotides previously checked for their ability to amplify (Table 1). Results were analyzed on a 2% agarose gel. Quantitative PCR experiments (oligonucleotides described in Table 1) were done as previously described (Prat et al., 2012) on a Light Cycler 480® (Roche, Meylan, France) according to the MIQE guide-lines (Bustin et al., 2009) with the use of geometric mean of two housekeeping genes (36B4/RPLP0 and hRPS17). Results are expressed as mean of three independent experiments that were run in duplicate. 2.5. Western blot analysis Protein extracts were obtained from TM primary cells or TM5 cells using RIPA buffer. Antibody against RARa, b, g (1/200, sc-551, sc-552, sc-550, Santa-Cruz®, Heidelberg, Germany) and RXRa, b, g (1/200, sc-553, sc-831, sc-555, Santa-Cruz®, Heidelberg, Germany) were used. For TM5 cells, this extraction was done after atRA (106M) or DMSO treatment for 24, 48 or 72 h. Briefly, cell pellets were rinsed three times in PBS and 400 mL of RIPA were added. The mix was then placed on ice for 30 min with regular vortex mixing. Supernatants were collected and frozen at 80  C. For westernblot, proteins were resolved on a SDS-PAGE gel electrophoresis (10%). Then, the transfer was done on nitrocellulose membrane (GE Healthcare®, Velizy, France) and saturated during 30 min with 5% skimmed milk in PBS-Tween (0, 1%). Antibody against Myocilin (1/ 200, sc20976, Santa-Cruz®, Heidelberg, Germany) or b-actin (1/ 1000, ab6276, Abcam®, Paris, France) were diluted in 5% skimmed milk in PBS-Tween (0, 1%) and incubated overnight at 4  C. The membrane, rinsed three times with distillated water, was incubated at RT with a HRP coupled secondary antibody anti-rabbit (1/ 5000, ab6885, Abcam®) or anti-mouse (1/5000, ab6808, Abcam®) for one hour. The revelation was done using ECL plus Western-Blot kit (GE 164 healthcare®). Quantity One software (BioRad®) was used for quantification. Results are expressed as mean of three independent experiments that were run in duplicate. 2.6. Retinoid transcriptional machinery tests in TM primary cells TM primary cells were transfected (using FugeneHD®, Roche, Meylan, France) with 1 mg DR5-tk-CAT (containing one copy of the retinoic acid-responsive element DR5 (direct repeat 5) upstream of the reporter gene CAT) and 0,15 mg of a CMV-b-Galactosidase vector serving as internal control to normalize variations in transfection efficiency. After a 24 h incubation, cells were rinsed and fresh medium containing the treatment was added (i.e.: retinol, atRA “retinol þ ethanol”, “retinol þ bisdiamine”, “retinol þ ketoconazole”) for 24 h. After washing the cells, CAT (Roche) and b-Galactosidase (Agilent technologies®, Les Ulis, France) assay were done on cell lysis. In all experiments, CAT activity was normalized by b-galactosidase activity according to both manufacturer protocols. Results are expressed as mean of three independent experiments that were run in duplicate. 2.7. Immunochemistry assays TM5 cells were cultured in Lab-tek™ culture chambers (MC2, Clermont-Ferrand, France). After treatment, they were fixed in paraformaldehyde 4% in PBS at room temperature (RT) for 10 min and washed three times in PBS. After incubation in bovine serum albumin 3% for 30 min, primary antibody against Myocilin (1/200,  4  C. After sc20976, Santa-Cruz®) was then applied overnight a

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

93

Table 1 List of primers used for this study: PCR and quantitative-PCR (upper panel), promoter cloning and mutagenesis (middle panel) and ChIP experiments (lower panel). Forward (5'-3')

Reve rse (5'-3')

For classial PCR RARa RARb RARg RXRa RXRb RXRg ADH1A ADH3 ADH4 DHRS4 DHRS9 RDH16 ALDH1A1 ALDH1A2 ALDH1A3 ALDH8A1 CYP26A1 CYP26B1 CYP26C1 CRBP1 CRBP2 CRABP1 CRABP2 MYOC RPLP0

AGTCCTCAGGCTACCACTAT ATGGATGTTCTGTCAGTGAG ACCAATAAGGAGCGACTCT GGATCCCACACTTCTCAG AGTACTGCCGCTATCAGAA CTACACAGATACCCCAGTGA TCTTGGTGGCTTTAAAAGTA ATGAAGTTCGCATTAAGATG AACCTGCCTGGTTAATTTAT TATCCTAGTCTCCAATGCTG AAACCTCAGAGAGACTTCGT CATACTGGACGTGAACTTGT TACTCACCGATTTGAAGATT TGACTTCCAGCAAGATAGAG ACCTGGAGGTCAAGTTCA TAGATTCTTACGACCCATCA AAGCTCTGGGACCTGTACT ACCTCTTTGAGGTCTACCAG GGAGAACCTCTTCTCACTG GATCTGACAGGCATAGATGA ATGAAAACTTTGAGGGCTAC ACCACTGAGATCAACTTCAA GAATTGCTCAAAGTGCTG TGGAGGAAGAGAAGAAGCGACTAA GACCTGGAAGTCCAACTACT

CCTCCTTCTTCTTCTTGTTT CATAGTGGTACCCTGATGAT ATCTCCTCTGAGCTGGTG GAGTCAGGGTTAAAGAGGAC GTTAGTCACAGGGTCATTTG GGGTAGTTCATGTTTCCAAT GTCTCAAAACATCAGAATGG TTTCAAGCAGTAGTTGCTCT ACTTTCAACTTCCCAGAACT CAAGGCTGTTTTACTGACAT TCTAGTGTCAGCCAGTCAGT CCAGGAAGCTCTTTAAGAAT TTGTCAACATCCTCCTTATC TGGATTATAGACAGGGAACA CTGTATCCATCGTCTCCAG TAAGGTTTTCCCTTGGTCT TGTAGATGAAGCCGTATTTC GCTCTCAATGAGGAGGTC CTTGCACTGTGAATGATTAG CTTCTTGAATACTTGCTTGC GAAATCCACATCATAGTTGC TCACTCTCGGACATAAATTC CTGCTCACAGACCATTTTAT TGTGTCATAAGCAAAGTTGA GTGATATCAAGCACTTCAGG

For qPCR MYOC RPLP0 hRPS17

ATCTCAGGAGTGGAGAGGGA AGGCTTTAGGTATCACCACT TGCGAGGAGATCGCCATTATC

CTGGCTGATGAGGTCATACTC GCAGAGTTTCCTCTGTGATA AAGGCTGAGACCTCAGGAAC

For Promoter Cloning in PGL3E MYOC -1147/þ1 MYOC -1148/-2422 MYOC-2423/-3697 MYOC -3698/-4298

AGTTTTGGTATATTTATTGGCTATTGCC TACTCAGCCCTGTGGTGGAC GAATGGTTTACTAAACCAACAGGG AATTTTATAAAGTCAGGCAT

GGGTGGCCTTGCTGGCTCAT GTCAGAAGTAACTTTAAGCCACTTG GTTGCCCAGAAGACATGAAAGA AAGTTATTCCTTATTAAACA

For Mutagenesis -3698/-4298 Mut1 -3698/-4298 Mut2 -3698/-4298 Mut3 -3698/-4298 Mut4

ACACACAGGCCCGATGTGTCGTGTTGCAGGAAGCATCTACAAC GTGTCTTACACCTACCTGTAGTAGTGGTTGACTCATGCACTGCAACC CATATTAGCCCGGCTGGTCTCTGGAGATCACTTGCAGGTGATCCA CAGGCATGAGTCACCGCGCCTAATTGCATAGACGTGTTTAATA

GACACATCGGGCCTGTGTGTGAGGGCTATGA TACAGGTAGGTGTAAGACACAGACCCTCACCCTC AGACCAGCCGGGCTAATATGGTGAAACCCCATCTC GGCGCGGTGACTCATGCCTGTAATCCCAGC

For ChIP experiment On GAPDH promoter On MYOC promoter

TACTAGCGGTTTTACGGGCG GCTGGGACTACAGGCGCACGC

TCGAACAGGAGGAGCAGAGAGCGA ACTCATGCCTGTAATCCCAGCACTTTAGGA

three washes in PBS, secondary antibody anti goat-Cy3 (1/5000) was incubated for one hour at RT. Cy3 amplification was applied for 5 min (TSA-Cy3 amplification kit™, 1/75, Perkin-Elmer®, Courtaboeuf, France). Slides were washed three times in PBS and incubated with DAPI (5min, dilution in PBS 1/5000). Finally, tissues sections were mounted in propyl gallate/PBS and examined under a Zeiss© Axiophot microscope. For negative controls, cells were incubated without primary antibody. 2.8. Reporter gene assays and expression plasmids For Myocilin promoter analysis, the different MYOC promoter constructs were obtained with oligonucleotide described in Table 1. First, cloned into PCRII-topo vector (Thermo-Fisher scientific SaintAubin, France), the promoter parts of Myocilin gene were then res, France) cloned in PGL3enhancer vector (Promega®, Charbonnie after digestion by XhoI and NotI restriction enzymes. The generation of 4 different mutated DR sites in pMyo (4298/-3698) were obtained using GeneTailor site direct mutagenesis system (ThermoFisher scientific (Saint-Aubin, France)). Expression plasmids for

human RARa and RXRa were cotransfected with pCH110 vector expressing b-galactosidase for normalization. For gene reporter experiments, reporter plasmid for Myocilin promoter (wild type or mutated; 1,45 mg), expression plasmid for RARa and RXRa (0,15 mg each) and for b-Galactosidase pCH110 (0,10 mg) were co-transfected using FugeneHD (Roche®, Meylan, France) in cells subcultured during 24 h with charcoal stripped FBV (PAA laboratories®, VelizyVillacoublay, France). After being incubated for 24 h more, cells were treated using equal volume of atRA (106M) or DMSO. Total cells were lysed and luciferase (Luciferase Reporter gene assay, Roche®, Meylan, France) or b-Galactosidase (High sensitivity bGal®, Agilent, Massy, France) activities were quantified. Each experiment was performed three times in duplicate. 2.9. Chromatin immunoprecipitation assays ChIP experiments were run as previously described (Borel et al., 2010). Oligonucleotides used for GAPDH promoter, Myocilin promoter site 3 or input control before immunoprecipitation, were described in Table 1. Antibodies used for immunoprecipitation were

94

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

respectively, Polymerase II (for positive control on GAPH promoter, sc899X, Santa Cruz®), RARa (on Myocilin promoter site 3, sc551X, Santa Cruz) and RXRa (on Myocilin promoter site 3, sc553X, Santa Cruz®) and Rabbit IgG (for negative control on both promoter, sc2027, Santa Cruz®). PCR products were resolved through a 2% agarose gel in TBE1X. 2.10. Statistical tests Results are expressed as means (±SEM). Comparisons were made using Mann-Whitney U test and statistical significance was set at p < 0.05, p < 0.01 or p < 0.001. 3. Results 3.1. The retinoids pathway actors are expressed in the trabecular meshwork immortalized TM5 cells and are able to induce the Myocilin gene expression The retinoids signaling is transduced from cytoplasmic retinol to

retinoid nuclear receptors through a multistep process based on cellular enzymes and binding proteins as represented in Fig. 1A. Qualitative RT-PCR experiments showed that a minimum of one actor for each metabolic activating step is expressed in the trabecular meshwork immortalized TM5 cells (Fig. 1B, upper panel). The presence of the final effectors of the vitamin A signaling pathway, the retinoid nuclear receptors subtypes (RARs and RXRs), were detected both at the mRNA and protein levels with the exception of RXRb and RXRg (Fig. 1B, middle and lower panel). Such pathway is also present and functional in TM primary cells (supplementary data 1). Then, the treatment of TM5 cells by atRA promoted a strong induction of MYOC expression (with a weak expression found at basal level) with highest induction level (8.7fold) at the 24th hour (Fig. 2A/left panel). Cycloheximidine’ treatment did not abolish this positive transcriptional induction (data not shown) concluding to a direct mechanism, in agreement with the mRNA’ early kinetic. This result was further supported by the regular increase (2e3-fold) of Myocilin protein levels between 24 and 72 h of treatment (Fig. 2A, left panel and right panel for a representative western-blot experiment). Taken together, these data indicated that the retinoids pathway is present and functional for retinoic acid in terms of regulation of the Myocilin gene with an induced MYOC expression. 3.2. Retinoic acid and dexamethasone have a synergic effect on Myocilin protein accumulation To further investigate the putative interplay between retinoids and glucocorticoids signaling in the Myocilin gene regulation, we cultured TM5 cells with dexamethasone and/or atRA during 48 and 72 h (Fig. 2B). On one hand, immunostaining experiments confirmed the weak presence of endogenous Myocilin as well as the neutral effect of DMSO. On the other hand, Myocilin expression was induced by glucocorticoids and active retinoids product, after 48 and 72 h of treatment. The most prominent result was the dramatic increase of cytoplasmic Myocilin’ presence under combined incubation. This later data strongly supported the idea of a synergic effect between steroids and retinoids signaling on Myocilin expression. 3.3. A 4-RAREs cluster is involved in the regulation of Myocilin by retinoic acid pathway

Fig. 1. (A) Overview of the metabolic steps of the vitamin A pathway. (B) Transcripts' expression of genes belonging to vitamin A pathway. First strand cDNA obtained with mRNA from TM5 cells were used to amplify by PCR all the members of the vitamin A pathway. Electrophoresis demonstrated that, at least, one member of each metabolic and nuclear step is present (upper part of the panel). Electrophoretic migration (middle part of the panel) and western-blot (lower part of the panel) confirmed the presence of the nuclear effectors of the vitamin A pathway in these cells: RAR and RXR subtypes (except b and g for RXR). Abbreviations: ADH: alcohol dehydrogenase; DHRS and RDH: retinol dehydrogenase; ALDH: retinal dehydrogenase; CYP26: cytochrome P26 class; CRABP: cellular retinoic acid binding protein; CRBP: cellular retinol binding protein; RAR: Retinoic Acid Receptor; RXR: Retinoic X Receptor.

Running the V$RXRF vertebrates list of matrix (matrix library version 9.0) on Genomatix® software, numerous putative RXR heterodimer binding sites were identified within a 4500bp promoter interval upstream the MYOC transcription start site (þ1). Among all of them, only four sites ehereafter referred as RARE1 to RARE4, from 50 to 3’- displayed the core characteristics of the RAR/ RXR heterodimer binding sites. These specific sequences were found to be several kilobase pairs distant from the transcription initiation site (Fig. 3A). However, the four RAREs clustered in a potential regulatory hot spot spanning over 600 base pairs (bp). Further in silico analysis demonstrated three distinct types of RARE: DR1 (RARE1 and RARE4), DR2 (RARE3) and DR5 (RARE4). RARE1 (DR1) and RARE3 (DR2) showed the highest scores of core similarity (1.00), while RARE3 had the best matrix similarity value (0.97 and data not shown). Using a series of expression constructs (Fig. 3A), reporter gene experiments were conducted. After transfection in TM5 cells and atRA incubation during 24, 48 and 72 h (Fig. 3B), the constructs falling in the 3697 and þ1 range failed to induce any significant luciferase signal compare to DMSO treatment. Conversely, the critical 600bp region (4298 and 3698) encompassing the four predicted RAREs was the only one to display a significant 3e6-fold induction from 24 to 72 h of treatment.

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

95

Fig. 2. (A) Induction of Myocylin transcripts and proteins by all-trans retinoic acid. In TM5 cells, Myocilin expression was induced at the mRNA and protein level after 24, 48 or 72 h of treatment by atRA (left panel). Results are normalized on the DMSO control condition. Results are expressed as mean of three independent experiments that were run in duplicate. Statistical differences between DMSO and 106M atRA treatment are symbolized by * and ** for p < 0.05 and p < 0.01 respectively. Representative western-blot assay used for the Myocilin protein quantification (right panel). (B) Synergistic induction of Myocylin by dexamethasone and all-trans retinoic acid. TM5 cells are immunostained for Myocilin (red) and counterstained with DAPI (blue) after treatment by dexamethasone and/or retinoid active derivative (atRA). Glucocorticoids and retinoids treatments increase independently the Myocilin expression levels after 48 and 72 h. A dramatic increase is further observed using combined treatment (“dexamethasone þ atRA” versus dexamethasone or atRA alone).

3.4. The DR2-RARE (RARE3) is the major retinoid-driven regulation site of MYOC expression Focusing on the above mentioned 600bp promoter region, we tested the activity of the four distinct RAREs. Each RARE was

selectively mutated and the residual luciferase activity was measured in TM5 cells (Fig. 3C). After 24 and 48 h of atRA incubation, all the constructs showed a decreased luciferase signal but only the RARE3 mutation (pGL3E 4298/-3698) displayed a significant 2-fold decrease. The signal decrease even reached 3-fold

96

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

Fig. 3. (A) Identification of RAREs in the promoter of Myocylin. The in silico analysis (MatInspector, Genomatix®, V$RXRF matrix) of the Myocilin promoter covered a 4500 base pairs (bp) region upstream the transcription start site (þ1). Four retinoic acid responsive elements (RARE) were identified in this interval and numbered from 50 to 3’. Precisely, the matrix demonstrated three distinct types of RARE: DR1 (RARE1 and RARE4), DR5 (RARE2) and DR2 (RARE3). The representation of the different Myocilin promoter constructs used for the transient transfection was mentioned just below. (B) Identification of atRA responsive portions of Myocylin’ promoter. TM5 cells were cotransfected with each promoter constructs and then treated by atRA (106M) or DMSO for 24 h, 48 h and 72 h. Results are normalized on the DMSO control condition and are expressed as mean of three independent experiments that were run in duplicate. The strongest induction was obtained using the pGL3E 4298/-3698 construct, which included the four predicted RAREs (*: p < 0,05; ***: p < 0,001). (C) Identification of efficient RAREs of Myocylin’ promoter. Site-directed mutagenesis of each predicted RARE sequences (RARE 1 to 4) of the pGL3E 4298/-3698 construct were used for transient transfection in TM5 cells. After 24 h, 48 h or 72 h of treatment by atRA (106M) or DMSO, results are normalized on the DMSO control condition and indicate that the DR2-RARE3 binding site, located between the 3787 and the 3811 position, is the main retinoic acid responsive element involved in the Myocilin gene induction (*: p < 0,05; **: p < 0,01). The exact sequence of the DR2 RARE3 was mentioned in accordance with the Genomatix® Bio-informatics study demonstrating the entire core matrix similarity (equal to 1 and calculated on the four most highest conserved position) and the strong global matrix similarity (equal to 0.973 compare to 1). ChIP experiments confirmed that RARa/RXRa heterodimer physically interacts with this DR2 RARE3 specific sequence (bottom part).

after 72 h of treatment (Fig. 3C). Finally, chromatin immunoprecipitation experiment in TM5 cells further demonstrated that the DR2-RARE3 is targeted by RARa/RXRa heterodimers (Fig. 3C).

Together, these results indicate that the DR2-RARE3 of the MYOC promoter is the main retinoic acid responsive element involved in the regulation of its transcription.

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

4. Discussion The regulation of the Myocilin transcription by different factors (environmental or endogenous) are part of the common POAG pathophysiology. Thus, a better understanding of MYOC regulation in trabecular meshwork has to be considered as an important step in order to explain the Myocilin induced glaucoma (Ricard and Tamm, 2005). Initially, some specific upstream sequence for protein binding were identified by Kirstein and colleagues (Kirstein et al., 2000). Then, some studies have described the (indirect) induction of this gene expression by glucocorticoids (Menaa et al., 2011; Faralli et al., 2015; Pattabiramann and Rao, 2015), or by the action of TGF-b2 (Wu et al., 2014). Thus, it has already been showed that dexamethasone enhances the expression of Myocilin gene (delayed induction, Shepard et al., 2001) and also the retinoic acid receptors in trabecular meshwork cells (Ishibashi et al., 2002). This last result permits to hypothesize that the active derivatives of vitamin A could be involved in MYOC regulation. This last fact could be move closer to the demonstration that the retinoids signaling pathway plays a critical role in the early eye development since mutations in the human retinol receptor gene (STRA6) and the retinoic acid receptor b (RARB) have been identified to cause a range of severe developmental defects (anophthalmia/microphthalmia) (Williamson and FitzPatrick, 2014). Furthermore, after birth, the role of retinoids is related to photon detection in the retinal photoreceptors (Palczewski, 2012) and to the ocular surface homeostasis (Samarawickrama et al., 2015; Nezzar et al., 2007). Here, we used a trabecular meshwork cell model (TM5) and showed that the enzymatic machinery, the binding proteins and the nuclear receptors of the retinoid pathway are expressed and functional in terms of gene’ regulation. Then, we demonstrated in such cell line that the retinoic acid is involved in the endogenous regulation of MYOC expression at the mRNA and protein level. But, the most striking result was the dramatic increase of the cytosolic Myocilin signal after co-treatment with dexamethasone and atRA. This last result suggests a synergic effect, in this case, between the glucocorticoid and retinoid pathways, as it was previously described in other cell lines (Lepar and Jump, 1992; Subramaniam et al., 2003). Several levels of interplay have been previously identified between these pathways. Dexamethasone enhances RXRa and RARb expression (Pascussi et al., 2000; Steineger et al., 1997; Yamaguchi et al., 1999), and retinoids and glucocorticoids are also able to induce overlapping set of target genes (Brossaud et al., 2013). This effect may be reached through the coexistence of RA and GR responsive elements in the promoter of these genes. In the Myocilin promoter, the region comprised between the transcription start site and the 2548 position has been shown to be required for glucocorticoids induction (Joe et al., 2011). On the other hand, our in silico screening identified a 600bp region ~1000bp upstream the previous “glucorticoids” one, which is critical for retinoids induction. Together, these data shows that retinoids and glucocorticoids act through distinct domains. One possibility is that independent but co-inductions occur on these domains by concomitant treatment using retinoids and glucocorticoid inducing a synergistic answer. Another one is that the physical distance between the two domains allows a functional
th et al., 2011). Alternately crosstalk as it was already described (To and its must be underlined for future studies on this promoter, RXR could be also an heterodimer partner of numerous other nuclear receptors such as PPAR (Peroxisome proliferator activator receptor), VDR (Vitamin D receptor), LXR (Liver-X-Receptor) and our in silico study identified 5 potential binding sites for VDR/RXR, one for LXR/ RXR and one for PPAR/RXR in the 2548 to þ1 region of the Myocilin promoter (data not shown). Hence, RXRa could be an additional bridge between the two pathways, by being induced by

97

glucocorticoids and engaged in the two distinct promoter domains. Due to the fact that retinoic acid signal acts through RAR/RXR receptors, we decided to use the specific RAR ligand (atRA) to ensure specific induction on canonic RAREs for RAR/RXR. Such retinoic acid responsive elements are featured by two direct repeated sequences (DR) called DR1, DR2 and DR5 with a respective spacing of 1, 2 and 5 base pairs. We show that the Myocilin promoter contains one DR5 but also two DR1 and one DR2. The four predicted DRs clustered in a specific 600bp interval approximatively 3500bp distant from the transcription start site. Transfection experiments confirmed that these DR enhance the Myocilin gene transcription. Of interest, further mutagenesis experiments demonstrated that the DR2-RARE site has the strongest inductive effects. DR5 are the most extensively studied and screened RAREs in gene promoters to predict a potential regulation by retinoic acid
e et al., 2011). The functional effect of the DR2-RAREs is (Laleve poorly documented in the literature (Donato and Noy, 2005; Jeong et al., 2006). Nevertheless in our case, a physical interaction of both RAR and RXR on this DR2-RARE was confirmed by ChIP experiments. 5. Conclusion Results obtained in the present study is the first report of the direct regulation (induction) of the Myocilin expression by the retinoids signaling pathway in immortalized TM cell line and bring some new clues concerning MYOC and its link with glaucoma as it was recently reviewed (Gemenetzi et al., 2012). Whenever, the Myocilin accumulation by trabecular cells and a direct link with an IOP increase still to be clarified, high protein levels in the aqueous humor of glaucoma affected dogs (Mackay et al., 2008a,b) and in patients suffering from glaucoma (Resch and Fautsch, 2009) underlined the importance of our finding. So and because the retinoid pathway is pharmacologically actionable (RAR agonists and antagonists) (Comptour et al., 2016), we anticipate from our study that Myocilin regulation by retinoic acid should become an important issue in the understanding and putative therapeutic strategies of the human POAG. Author contributions CP, CB, AC, GM and GC conducted experiments and acquired data. AC, MR, CB, LB and VS analyzed data. LB, VS and FC designed research studies. LB, VS and FC wrote the manuscript. Acknowledgements The authors would like to thank Alain Herbet for his critical reading of the manuscript. LB and AC thank THEA laboratory for granting the study. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.exer.2017.01.006. References Bastien, J., Rochette-Egly, C., 2004. Nuclear retinoid receptors and the transcription of retinoid-target genes. Gene 328, 1e16. Borel, V., Marceau, G., Gallot, D., Blanchon, L., Sapin, V., 2010. Retinoids regulate human amniotic tissue-type plasminogen activator gene by a two-step mechanism. J. Cell. Mol. Med. 14, 1793e1805. Brossaud, J., Roumes, H., Moisan, M.-P., Pallet, V., Redonnet, A., Corcuff, J.-B., 2013.

98

C. Prat et al. / Experimental Eye Research 155 (2017) 91e98

Retinoids and glucocorticoids target common genes in hippocampal HT22 cells. J. Neurochem. 125, 518e531. Bustin, S.A., Benes, V., Garson, J.A., Hellemans, J., Huggett, J., Kubista, M., Mueller, R., Nolan, T., Pfaffl, M.W., Shipley, G.L., et al., 2009. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55, 611e622. Clark, A.F., Steely, H.T., Dickerson, J.E., English-Wright, S., Stropki, K., McCartney, M.D., Jacobson, N., Shepard, A.R., Clark, J.I., Matsushima, H., et al., 2001. Glucocorticoid induction of the glaucoma gene MYOC in human and monkey trabecular meshwork cells and tissues. Invest. Ophthalmol. Vis. Sci. 42, 1769e1780. Comptour, A., Rouzaire, M., Belville, C., Bouvier, D., Gallot, D., Blanchon, L., Sapin, V., 2016 Oct. Nuclear retinoid receptors and pregnancy: placental transfer, functions, and pharmacological aspects. Cell. Mol. Life. Sci. 73 (20), 3823e3837. http://dx.doi.org/10.1007/s00018-016-2332-9. Epub 2016 Aug 9. Cvekl, A., Wang, W.-L., 2009. Retinoic acid signaling in mammalian eye development. Exp. Eye Res. 89, 280e291. Donato, L.J., Noy, N., 2005. Suppression of mammary carcinoma growth by retinoic acid: proapoptotic genes are targets for retinoic acid receptor and cellular retinoic acid-binding protein II signaling. Cancer Res. 65, 8193e8199. Faralli, J.A., Clark, R.W., Filla, M.S., Peters, D.M., 2015. NFATc1 activity regulates the expression of myocilin induced by dexamethasone. Exp. Eye Res. 130, 9e16. Gemenetzi, M., Yang, Y., Lotery, A.J., 2012. Current concepts on primary open-angle glaucoma genetics: a contribution to disease pathophysiology and future treatment. Eye 26, 355e369. Ishibashi, T., Takagi, Y., Mori, K., Naruse, S., Nishino, H., Yue, B.Y.J.T., Kinoshita, S., 2002. cDNA microarray analysis of gene expression changes induced by dexamethasone in cultured human trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 43, 3691e3697. Jeong, H., Kim, M.-S., Kim, S.-W., Kim, K.-S., Seol, W., 2006. Regulation of tyrosine hydroxylase gene expression by retinoic acid receptor. J. Neurochem. 98, 386e394. Joe, M.K., Sohn, S., Kim, T.E., Im, J., Choi, Y.R., Kee, C., 2011. Analysis of glucocorticoidinduced MYOC expression in human trabecular meshwork cells. Vis. Res. 51, 1033e1038. Karali, A., Russell, P., Stefani, F.H., Tamm, E.R., 2000. Localization of myocilin/ trabecular meshworkeinducible glucocorticoid response protein in the human eye. Invest. Ophthalmol. Vis. Sci. 41, 729e740. Kastner, P., Grondona, J.M., Mark, M., Ganamuller, A., LeMeur, M., Decimo, D., Vonesch, J.L., DollE, P., Chambon, P., 1994. Genetic analysis of RXR alpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987e1003. Kirstein, L., Cvekl, A., Chauhan, B.K., Tamm, E.R., 2000. Regulation of human myocilin/TIGR gene transcription in trabecular meshwork cells and astrocytes: role of upstream stimulatory factor. Genes Cells 5 (8), 661e676.
e, S., Anno, Y.N., Chatagnon, A., Samarut, E., Poch, O., Laudet, V., Benoit, G., Laleve Lecompte, O., Rochette-Egly, C., 2011. Genome-wide in silico identification of new conserved and functional retinoic acid receptor response elements (direct repeats separated by 5 bp). J. Biol. Chem. 286, 33322e33334. Lepar, G.J., Jump, D.B., 1992. Retinoic acid and dexamethasone interact to regulate S14 gene transcription in 3T3-F442A adipocytes. Mol. Cell. Endocrinol. 84, 65e72. Lindsey, J.D., Gaton, D.D., Sagara, T., Polansky, J.R., Kaufman, P.L., Weinreb, R.N., 2001. Reduced TIGR/myocilin protein in the monkey ciliary muscle after topical prostaglandin F(2alpha) treatment. Invest. Ophthalmol. Vis. Sci. 42, 1781e1786. €llberg, M.E., Gelatt, K.N., 2008a. Aqueous humor myocilin protein Mackay, E.O., Ka levels in normal, genetic carriers, and glaucoma Beagles. Vet. Ophthalmol. 11, 177e185. MacKay, E.O., Kallberg, M.E., Barrie, K.P., Miller, W., Sapienza, J.S., Denis, H., Ollivier, F.J., Plummer, C., Rinkoski, T., Scotty, N., et al., 2008b. Myocilin protein levels in the aqueous humor of the glaucomas in selected canine breeds. Vet. Ophthalmol. 11, 234e241. Mark, 2007. Function of retinoic acid receptors during embryonic development. Nucl. Recept. Signal 4. Menaa, F., Braghini, C.A., Vasconcellos, J.P.C.D., Menaa, B., Costa, V.P.,

Figueiredo, E.S.D., Melo, M.B.D., 2011. Keeping an eye on myocilin: a complex molecule associated with primary open-angle glaucoma susceptibility. Mol. Basel Switz. 16, 5402e5421. Nezzar, H., Chiambaretta, F., Marceau, G., Blanchon, L., Faye, B., Dechelotte, P., Rigal, D., Sapin, V., 2007. Molecular and metabolic retinoid pathways in the human ocular surface. Mol. Vis. 13, 1641e1650. Nguyen, T.D., Chen, P., Huang, W.D., Chen, H., Johnson, D., Polansky, J.R., 1998. Gene structure and properties of TIGR, an olfactomedin-related glycoprotein cloned from glucocorticoid-induced trabecular meshwork cells. J. Biol. Chem. 273, 6341e6350. Palczewski, K., 2012. Chemistry and biology of vision. J. Biol. Chem. 287, 1612e1619. Pascussi, J.M., Drocourt, L., Fabre, J.M., Maurel, P., Vilarem, M.J., 2000. Dexamethasone induces pregnane X receptor and retinoid X receptor-alpha expression in human hepatocytes: synergistic increase of CYP3A4 induction by pregnane X receptor activators. Mol. Pharmacol. 58, 361e372. Pattabiramann, P.P., Rao, P.V., 2015. Hic-5 regulates actin cytoskeletal reorganization and expression of fibrogenic markers and myocilin in trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 56 (9), 5656e5669. Polansky, J.R., Fauss, D.J., Chen, P., Chen, H., Lütjen-Drecoll, E., Johnson, D., Kurtz, R.M., Ma, Z.D., Bloom, E., Nguyen, T.D., 1997. Cellular pharmacology and molecular biology of the trabecular meshwork inducible glucocorticoid response gene product. Ophthalmol. J. Int. Ophtalmol. Int. J. Ophthalmol. Z. Für Augenheilkd 211, 126e139. Prat, C., Blanchon, L., Borel, V., Gallot, D., Herbet, A., Bouvier, D., Marceau, G., Sapin, V., 2012. Ontogeny of aquaporins in human fetal membranes. Biol. Reprod. 86, 48. Resch, Z.T., Fautsch, M.P., 2009. Glaucoma-associated myocilin: a better understanding but much more to learn. Exp. Eye Res. 88, 704e712. Ricard, C.S., Tamm, E.R., 2005. Focus on molecules: myocilin/TIGR. Exp. Eye Res. 81, 501e502. Rozsa, F.W., Reed, D.M., Scott, K.M., Pawar, H., Moroi, S.E., Kijek, T.G., Krafchak, C.M., Othman, M.I., Vollrath, D., Elner, V.M., et al., 2006. Gene expression profile of human trabecular meshwork cells in response to long-term dexamethasone exposure. Mol. Vis. 12, 125e141. Samarawickrama, C., Chew, S., Watson, S., 2015. Retinoic acid and the ocular surface. Surv. Ophthalmol. 60, 183e195. Sheffield, V.C., Stone, E.M., Alward, W.L., Drack, A.V., Johnson, A.T., Streb, L.M., Nichols, B.E., 1993. Genetic linkage of familial open angle glaucoma to chromosome 1q21-q31. Nat. Genet. 4, 47e50. Shepard, A.R., Jacobson, N., Fingert, J.H., Stone, E.M., Sheffield, V.C., Clark, A.F., 2001. Delayed secondary glucocorticoid responsiveness of MYOC in human trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 42, 3173e3181. Steineger, H.H., Arntsen, B.M., Spydevold, O., Sørensen, H.N., 1997. Retinoid X receptor (RXR alpha) gene expression is regulated by fatty acids and dexamethasone in hepatic cells. Biochimie 79, 107e110. Stone, E.M., Fingert, J.H., Alward, W.L., Nguyen, T.D., Polansky, J.R., Sunden, S.L., Nishimura, D., Clark, A.F., Nystuen, A., Nichols, B.E., et al., 1997. Identification of a gene that causes primary open angle glaucoma. Science 275, 668e670.
n, J., Rafter, I., Okret, S., 2003. Cross-talk between glucoSubramaniam, N., Campio corticoid and retinoic acid signals involving glucocorticoid receptor interaction with the homoeodomain protein Pbx1. Biochem. J. 370, 1087e1095.
th, K., Sarang, Z., Scholtz, B., Br

süs, L., To azda, P., Ghyselinck, N., Chambon, P., Fe Szondy, Z., 2011. Retinoids enhance glucocorticoid-induced apoptosis of T cells by facilitating glucocorticoid receptor-mediated transcription. Cell Death Differ. 18, 783e792. Williamson, K.A., FitzPatrick, D.R., 2014. The genetic architecture of microphthalmia, anophthalmia and coloboma. Eur. J. Med. Genet. 57, 369e380. Wu, Y., Chen, W., Guo, M., He, Q., Hu, Y., 2014. Effects of transforming growth factorb2 on myocilin expression and secretion in human primary cultured trabecular meshwork cells. Int. J. Clin. Exp. Pathol. 7, 4827e4836. Yamaguchi, S., Murata, Y., Nagaya, T., Hayashi, Y., Ohmori, S., Nimura, Y., Seo, H., 1999. Glucocorticoids increase retinoid-X receptor alpha (RXRalpha) expression and enhance thyroid hormone action in primary cultured rat hepatocytes. J. Mol. Endocrinol. 22, 81e90.