Gremlin Utilizes Canonical and Non-Canonical TGFΒ Signaling

Experimental Eye Research 113 (2013) 117e127

Contents lists available at SciVerse ScienceDirect

Experimental Eye Research

journal homepage: www.elsevier.com/locate/yexer

Gremlin utilizes canonical and non-canonical TGFb signaling to induce lysyl oxidase (LOX) genes in human trabecular meshwork cellsq

Anirudh Sethi 1, Robert J. Wordinger, Abbot F. Clark*

Department of Cell Biology & Anatomy, North Texas Eye Research Institute, University of North Texas Health Science Center, 3500 Camp Bowie Blvd., Fort Worth, TX 76107, USA article info abstract

Article history: The TGFb/BMP signaling pathways are involved in glaucomatous damage to the trabecular meshwork Received 15 February 2013 (TM) leading to elevated intraocular pressure (IOP), which is a major risk factor for the development and Accepted in revised form 15 May 2013 progression of glaucoma. The BMP antagonist gremlin is elevated in glaucomatous TM cells and tissues Available online 5 June 2013 and can directly elevate IOP. Gremlin utilizes the TGFb2/SMAD pathway to induce TM extracellular matrix (ECM) proteins. The purpose of this study is to determine whether expression of the ECM cross- Keywords: linking lysyl oxidase (LOX) genes is regulated by gremlin in cultured human TM cells. Human TM cells primary open angle glaucoma were treated with recombinant gremlin, and expression of the LOX genes was examined by quantitative trabecular meshwork extracellular matrix RT-PCR and western immunoblotting. TM cells were pretreated with TGFBR inhibitors (LY364947 or gremlin SB431542), an inhibitor of the SMAD signaling pathway (SIS3), or with JNK (SP600125) and p38 MAPK lysyl oxidase (SB203580) inhibitors to identify the signaling pathway(s) involved in gremlin induction of LOX protein lysyl oxidase like expression. All ﬁve LOX genes (LOX and LOXL1e4) were induced by gremlin. Gremlin induction of LOX transforming growth factor beta genes and protein expression was blocked by TGFBR inhibitors as well as by inhibitors of the SMAD3, JNK and p38 MAPK signaling pathways. We conclude that gremlin employs both canonical TGFb/SMAD and the non-canonical JNK and p38 MAPK signaling pathways to induce LOX genes and proteins in cultured human TM cells. Increased LOX levels may be at least partially responsible for gremlin-mediated IOP elevation and increased aqueous humor outﬂow resistance leading to glaucoma. Ó 2013 The Authors. Published by Elsevier Ltd. All rights reserved.

1. Introduction extracellular matrix (ECM) material within the TM. Therefore, glaucoma can be viewed as a fibrotic disorder of the TM. Glaucoma is a leading cause of irreversible visual impairment Numerous studies have shown that TGFb2 levels are elevated in and blindness in the world, with primary open-angle glaucoma the AH (Inatani et al., 2001; Tripathi et al., 1994) and TM of POAG (POAG) being the major form of glaucoma (Quigley and Broman, patients (Tovar-Vidales et al., 2011). TGFb2 is a profibrotic growth 2006). Elevated intraocular pressure (IOP) is a major risk factor factor linked to fibrotic diseases of the lung, kidney, skin, and liver. for the development and progression of glaucoma (AGIS, 2000; TGFb2 has a number of effects on the TM, most notably increasing Weinreb and Khaw, 2004). Ocular hypertension is attributed to AH outflow resistance and elevating IOP in perfusion cultured hu- increased aqueous humor (AH) outflow resistance in the trabecular man and porcine eyes (Bachmann et al., 2006; Fleenor et al., 2006; meshwork (TM), a tissue in the anterior segment of the eye. Gottanka et al., 2004) as well as in rodent eyes (Shepard et al., Elevated IOP has been associated with increased deposition of 2010). TGFb2 also modulates TM ECM metabolism. This growth factor increases the expression of a variety of ECM proteins, including fibronectin (FN), collagen (COL), elastin (ELN), and pro- teoglycans as well as plasminogen activator inhibitor-1 (PAI1) and q This is an open-access article distributed under the terms of the Creative tissue inhibitor of metalloproteinase-1 (TIMP1) (Fuchshofer and Commons Attribution-NonCommercial-No Derivative Works License, which per- Tamm, 2012). PAI-1 and TIMP-1 suppress proteolytic degradation mits non-commercial use, distribution, and reproduction in any medium, provided of the ECM (Fuchshofer and Tamm, 2012). In addition, TGFb2 in- the original author and source are credited. creases expression of the ECM cross-linking enzymes * Corresponding author. Tel.: þ1 817 735 2094. transglutaminase-2 (TGM2) (Tovar-Vidales et al., 2008; Welge- E-mail address: [email protected] (A.F. Clark). e 1 Current address: Department of Biochemistry, University of Texas South- Lussen et al., 2000), lysyl oxidase (LOX), and LOX like 1 4 western Medical Center, Dallas, TX 75390, USA. (LOXL1e4) (Sethi et al., 2011b). Thus, parallels can be easily drawn

0014-4835/$ e see front matter Ó 2013 The Authors. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.exer.2013.05.011 118 A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127

SMAD as well as c-Jun N-terminal Kinase (JNK) signaling pathways Abbreviations to induce LOX genes and proteins in cultured human TM cells (Sethi et al., 2011b). Both SMAD and mitogen activated protein kinase POAG Primary open-angle glaucoma (MAPK) signaling pathways, including JNK signaling, have been IOP Intraocular pressure associated with fibrosis (Ma et al., 2009). However, very little is AH Aqueous Humor known about the role of gremlin and the signaling mechanism(s) TM Trabecular Meshwork used to induce LOX and LOXL. ECM Extracellular matrix The purpose of the present study was to determine: (1) whether TGFb Transforming growth factor beta gremlin induces LOX gene expression in the TM cells, and (2) which FN Fibronectin signaling pathway(s) regulate gremlin-induced LOX expression in COL Collagen the TM. ELN Elastin PAI1 Plasminogen activator inhibitor-1 2. Materials and methods TIMP1 Tissue inhibitor of metalloproteinase-1 TGM2 Transglutaminase 2 2.1. TM cell culture LOX Lysyl Oxidase LOXL Lysyl Oxidase like Human TM cells were isolated from carefully dissected human BMP Bone morphogenetic proteins TM tissue explants derived from donor eyes and characterized as BMPR Bone morphogenetic proteins receptor previously described (Fleenor et al., 2006; Wordinger et al., 2002, MAPK Mitogen activated protein kinase 2007). All donor tissues were obtained from regional eye banks JNK c-Jun N-terminal Kinase and managed according to the guidelines in the Declaration of Helsinki for research involving human tissue. Isolated TM cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen-Gibco, Grand Island, NY) containing L-glutamine between the profibrotic effects of TGFb in the TM and the TGFb- (0.292 mg/ml; Gibco BRL Life Technologies), penicillin (100 units/ mediated fibrosis that occurs in other cells and tissues. ml)/streptomycin (0.1 mg/ml); (Gibco BRL Life Technologies), and We have previously reported that TM cells express several 10% fetal bovine serum (Gibco BRL Life Technologies). members of the bone morphogenetic (BMP) family, including BMP ligands (BMP2, BMP4, BMP5 and BMP7), BMP receptors (BMPR1a, 2.2. TM cell treatments BMPR1b, BMPR2), and BMP antagonists such as gremlin (Wordinger et al., 2002, 2007). BMPs are members of the TGFb TM cells were grown to 100% confluency and then kept in superfamily of proteins that control multiple functions in a variety serum-free medium for 24 h prior to gremlin or inhibitor treatment of cell types. Interestingly, BMP4 and BMP7 block the TGFb2- to avoid the confounding effects of serum proteins. TM cells were induction of a variety of ECM proteins in TM cells, including incubated with fresh medium containing specific signaling in- fibronectin, collagens IV & VI, TSP-1, and PAI1 (Fuchshofer et al., hibitors for 1e6 h, prior to the addition of varying concentrations of 2007; Wordinger et al., 2007). recombinant gremlin protein (R&D System, Minneapolis, MN). BMP antagonists, such as gremlin, tightly regulate BMP cellular Small molecule inhibitors LY364947 (5 mM, Tocris biosciences, activity. Gremlin directly binds to specific BMP ligands and blocks Ellisville, MO) and SB431542 (5 mM, SigmaeAldrich, St. Louis, MO) BMP binding to their receptors (Wordinger et al., 2008). We have were used to examine the effects of inhibition of TGFb Receptor-1/2. reported greater levels of gremlin in glaucomatous TM cells and SMAD-3 phosphorylation inhibitor SIS3 (10 mM, SigmaeAldrich, St. tissues (Wordinger et al., 2007). Elevated gremlin levels also occur Louis, MO), JNK inhibitor SP600125 (10 mM, SigmaeAldrich, St. in fibrotic disease of the kidney (Wordinger et al., 2008). We pre- Louis, MO), and p38 MAPK inhibitor SB203580 (5 mM, Tocris Bio- viously reported that gremlin antagonized BMP4 inhibition of sciences, Ellisville, MO) were used to examine effects of inhibition TGFb2-induced ECM proteins like FN and PAI1 in TM cells and also on canonical SMAD, JNK, and p38 signaling pathways, respectively elevated IOP in perfusion cultured human anterior segments (Sethi et al., 2011b). (Wordinger et al., 2007). Recently, we also demonstrated that gremlin alone can induce fibrosis-like activities in TM cells. Gremlin 2.3. RNA isolation induced expression of FN, COL1, ELN and PAI1 genes and proteins in cultured TM cells utilizing the TGFb2/SMAD signaling pathway Total cellular RNA was extracted from cultured TM cells using (Sethi et al., 2011a). TRI Reagent RT extraction (MRC Inc., Cincinnati, OH), and the Su- The LOX family contains 5 genes, LOX and LOXL1e4, encoding perScript VILO cDNA Synthesis kit (Invitrogen) was used for first enzymes that covalently cross-link elastin and collagens via gen- strand cDNA synthesis. PCR primers for the various LOX genes eration of aldehydes on lysine residues. This cross-linking reaction were designed using Primer3 software (http://frodo.wi.mit.edu/ provides additional mechanical strength to the ECM and also makes primer3/). The primer pairs are listed in Table 1. the ECM more resistant to degradation. LOX and LOXL are associated with several abnormalities related to an imbalance in ECM 2.4. Quantitative real time PCR synthesis and/or degradation such as fibrotic disorders of connective tissues of the heart (atrial fibrosis and myocardial fibrosis), Real-time PCR was performed as described previously (Sethi vasculature (atherosclerosis, vascular aneurysms and arterial et al., 2011b). Briefly, 2.5 ml of cDNA was used in a reaction con- fibrosis), lungs (pulmonary fibrosis), skin (fibrosis, hypertrophic sisting of 1.5 units per reaction of antibody-bound Taq enzyme scarring, keloids, and scleroderma), kidney (diabetic nephropathy, (Jump Start; SigmaeAldrich, St. Louis, MO), 10 PCR buffer, 1.5 mM nephritis), liver (liver stiffness preceding liver fibrosis), mouth MgCl2, 200 nM dNTP mix, 100 nM PCR primers (Table 1), 2.5 ml (inflamed oral tissue, gingival atrophy), and colon (intestinal green nucleic acid dye (EvaGreen; Biotium, Hayward, CA), as well as fibrotic disease) (Sethi et al., 2012). The TM expresses enzymatically 30 nM passive reference dye (Rox; USB, Cleveland, OH) per 50-ml active LOX and LOXL proteins, and TGFb2 utilizes both canonical reaction. PCR was performed on a real-time thermal cycler (model A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127 119

Table 1 were blocked with 5% Fat-free Dry Milk in TBST for 1 h and then List of the primers used for PCR studies. incubated overnight with primary antibodies (Table 2). The mem- Gene Primer (50 / 30) branes were washed with TBST and processed with corresponding

LOX Left CGACCCTTACAACCCCTACA horseradish peroxidase-conjugated secondary antibodies (Table 2). Right AAGTAGCCAGTGCCGTATCC The proteins were then visualized in a Fluor ChemTM 8900 imager LOXL1 Left AGAGCCTCTCTGTCCACCAG (Alpha Innotech) using ECL detection reagent SuperSignal West Right GTACACCTGCCCGTTGTTCT Femto Maximum Sensitivity Substrate (Pierce Biotechnology). To LOXL2 Left CCTGGGGAGAGGACATACAA ensure equal protein loading, the same blot was subsequently Right CTCGCAGGTGACATTCTTCA LOXL3 Left CAACGCGGCCTTCTACAG developed for GAPDH expression. Right GGTGTCATTGGCACGATAGA LOXL4 Left CGACAGCCACTACTACAGGAAA 2.6. Statistical analysis Right CTGGTGGATCCAGAAGGAGTT GAPDH Left CCACCCATGGCAAATTCCATGGCA Right TCTAGACGGCAGGTCAGGTCCACC For comparing results between two groups, the two-tailed student’s t test was performed. One-way ANOVA was employed for comparison of results between more than two groups.

Mx3000p; Stratagene, La Jolla, CA), with cycling parameters of 3. Results initial denaturation at 95 C; 40 cycles of 95 C for 30 s, 60 C for 30 s, and 72 C for 60 s, and a denaturation cycle for the creation of a 3.1. Gremlin induces LOX mRNA and protein expression in TM cells dissociation curves. Reactions for each sample were run in dupli- cate, cycle thresholds (Ct) were normalized to GAPDH expression as We previously reported that gremlin induces the ECM proteins a housekeeping gene, and comparative quantitation was performed FN, COL1, PAI1 and ELN in TM cells (Sethi et al., 2011a). Therefore, using MxPro ver. 4.0 software (Stratagene). The delta delta Ct we first determined the effect of gremlin on LOX and LOXL method was used for quantification of the data. Only individual PCR expression. Treatment with gremlin (1 mg/ml) for 24 h significantly samples with single-peak dissociation curves were selected for data induced LOX and LOXL1e4 mRNA expression (n ¼ 3, p < 0.05) analysis. (Fig. 1A). Gremlin also induced protein expression of cell associated and secreted LOX, LOXL1, LOXL2, and LOXL4 in cell lysates (Fig. 1B 2.5. Protein extraction and western blot analysis and D) and conditioned medium (Fig. 1C and E). LOXL3 was not assayed due to lack of a commercially consistently reliable Secreted proteins: LOX proteins secreted by TM cells were antibody. determined by western immunoblot analysis. Conditioned medium was collected from human TM cells after 24-h treatment with 3.2. Gremlin induces LOX genes and proteins in a concentration- gremlin in serum-free medium containing 0.5 mg/ml BSA. Proteins and time-dependent fashion were separated on a 10% denaturing polyacrylamide gel and transferred by electrophoresis to a PVDF membrane. Membranes TM cell strains (n ¼ 3) were treated with increasing concen- were blocked with 5% Fat-free Dry Milk in tris-buffered saline trations of gremlin (0e2 mg/ml) for 24 h. The mRNA and protein tween buffer (TBST) for 1 h and then incubated overnight with expression of LOX and LOXL1e4 were determined using qRT-PCR primary antibodies (Table 2). The membranes were washed with and western immunoblotting, respectively. Gremlin induced the TBST and processed with corresponding horseradish peroxidase- expression of all 5 LOX genes (Fig. 2A), as well as cell-associated conjugated secondary antibodies (Table 2). The proteins were (Fig. 2B and D) and secreted LOX proteins (Fig. 2C and E) in a then visualized in a Fluor ChemTM 8900 imager (Alpha Innotech, concentration-dependent manner. TM cells were also treated with San Leandro, CA) using ECL detection reagent SuperSignal West gremlin for 6, 12 and 24 h to determine the time dependence of LOX Femto Maximum Sensitivity Substrate (Pierce Biotechnology and LOXL mRNA induction. Gremlin significantly (p < 0.01) induced Rockford, IL). LOX and LOXL mRNA expression, although the time course of in- Cell-associated proteins: Total cellular protein was extracted duction varied for each gene (Fig. 3A). By six hours, gremlin from TM cells using mammalian protein extraction buffer (MPER, significantly induced all LOX genes except LOXL3. Similarly, TM cell Pierce Biotechnology), containing protease inhibitor (Pierce strains (n ¼ 3) were treated with gremlin (1 mg/ml) for 3, 6, 12, 24 Biotechnology) and phosphatase inhibitor (Pierce Biotechnology) and 48 to evaluate the effects of exogenous gremlin on LOX protein cocktails. Protein concentration was determined using the Bio-Rad expression. Gremlin induced both cell-associated (Fig. 3B and D) Dc protein assay system (Bio-Rad Laboratories, Richmond, CA). The and secreted (Fig. 3C and E) LOX proteins as early as 6 h and cellular proteins were separated on denaturing polyacrylamide gels maintained this induction for up to 48 h. We were unable to get and then transferred to PVDF membranes by electrophoresis. Blots consistent western immunoblot results for LOXL3. The gremlin induction of LOXL2 and LOXL4 proteins peaked as soon as 3 h and was maintained at 48 h. Therefore, gremlin induction of LOX and Table 2 LOXL mRNA and proteins was both time and concentration List of antibodies used for western immunoblotting studies. dependent.

Antibody Ab. Dilution Source 3.3. TGFb signaling in gremlin induction of LOX proteins Rabbit anti-LOX 1:10,000 Novus Biologicals (Cat. # NB100-2530) Rabbit anti-LOXL1 1:500 Abnova (Cat. # H00004016-D01P) Mouse anti-LOXL2 1:2000 R&D Systems (Cat. # MAB2639) We previously used various small molecule inhibitors to explore Rabbit anti-LOXL3 1:1000 SigmaeAldrich (Cat # SAB2107508) the involvement of TGFb signaling pathway(s) in gremlin-mediated Mouse anti-LOXL4 1:250 Abcam (Cat. # ab25904) ECM induction. We discovered that gremlin utilizes TGFb receptors Rabbit anti-GAPDH 1:1000 Cell Signaling (Cat. # 2118) to induce ECM proteins (Sethi et al., 2011a). We employed a similar Donkey anti-mouse IgG 1:10,000 Santa Cruz Biotechnology (Cat. # sc-2314) approach to study the role of TGFb receptors/signaling pathways in Goat anti-rabbit IgG 1:10,000 Santa Cruz Biotechnology (Cat. # sc-2004) regulating gremlin induction of LOX and LOXL proteins. SB431542 is Fig. 1. Gremlin induces LOX mRNA and proteins in TM cells. (A) Induction of LOX, LOXL1, LOXL2, LOXL3 and LOXL4 mRNA in six TM cell strains treated with gremlin (1 mg/ml) for 24 h vs. untreated control samples. Values represent the fold induction of LOX genes normalized to GAPDH. Three replicates of each sample were employed. Gremlin signiﬁcantly induced all 5 LOX genes in all the six cell lines. (BeC) Western immunoblots of cell-associated (B) and secreted (C) LOX, LOXL1, LOXL2 and LOXL4 proteins in two TM cell strains treated with gremlin (1 mg/ml) for 24 h. Cell-associated LOX proteins were normalized to GAPDH (B), while equal volumes of conditioned medium were loaded for secreted LOXs (C). Gremlin increased the expression of both cell associated and secreted LOXs. Similar results were observed in four additional TM cell strains. The western blot images are representative of three independent experiments. (DeE) Densitometry analysis of western immunoblots for cell associated (D) and secreted (E) LOXs. Two-tailed student’s t test was used to compare control and gremlin-treated samples **0.0001 < p < 0.01 and ***p < 0.0001.

Fig. 2. Concentration-dependent gremlin-induction of LOX mRNA and proteins. Concentration dependent induction of LOX mRNA (A), cell-associated protein (B), and secreted proteins (C) by 0e2 mg/ml gremlin in cultured TM cell strains (n ¼ 3). qRT-PCR values (A) represent fold LOX induction normalized to GAPDH and compared to controls. Three replicates of each sample were employed. Each line represents the individual LOX gene. Concentration dependent induction of cell associated (B) and secreted LOX proteins (C) by 0e2 mg/ml gremlin in cultured TM cell strains. Similar results were observed in two additional TM cell strains. The western blot images are representative of two independent experiments in each TM cell strain. (DeE) Densitometry analysis of western immunoblots for cell associated (D) and secreted (E) LOXs are plotted as lines. One-Way ANOVA was used for statistical analyses for comparing the samples. *0.01 < p < 0.05, **0.0001 < p < 0.01 and ***p < 0.0001. A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127 121

Fig. 3. Time-dependent gremlin-induction of LOX proteins. Time course induction (0e48 h) of LOX mRNA (A), cell-associated proteins (B), and secreted LOX proteins (C) after treatment of cultured TM cell strains (n ¼ 3) with gremlin (1 mg/ml). qRT-PCR values (A) represent fold LOX induction normalized to GAPDH compared to controls. Three replicates of each sample were employed for qRT-PCR. Each line represents the individual LOX gene. Time dependent induction of cell associated (B) and secreted LOX proteins (C) by gremlin in cultured TM cell strains. Similar results were observed in two additional TM cell strains. The western blot images are representative of two independent experiments in each TM cell strain. GAPDH served as loading control (B), while equal volumes of conditioned medium were used for secreted proteins (C). (DeE) Densitometry analysis of western immunoblots for cell associated (D) and secreted (E) LOXs are plotted as continuous lines. One-Way ANOVA was used for statistical analyses for comparing the samples. *0.01 < p < 0.05, **0.0001 < p < 0.01 and ***p < 0.0001. a selective TGFBR1 and TGFBR2 receptor inhibitor, while LY364947 We previously demonstrated that TGFb2 utilizes the JNK is relatively more selective for the TGFBR2 receptor than the signaling pathway to regulate LOX and LOXL induction in TM cells TGFBR1 receptor. We pretreated TM cell strains (n ¼ 3) for 1 h with (Sethi et al., 2011b). Therefore, we wanted to test whether gremlin or without 5 mM SB431542 or LY364947 followed by treatment with also utilizes MAPK signaling pathways to regulate LOX protein recombinant gremlin (1 mg/ml) for 24 h. Gremlin elevated cell- expression. We used SP600125, a selective small molecule inhibitor associated and secreted LOX and LOXL protein expression of JNK, to determine the potential role of the JNK signaling pathway compared to untreated or inhibitor only-treated samples. Pre- in gremlin-induction of LOX proteins. Three TM cell strains were treatment with either of the two inhibitors, LY364947 (Fig. 4)or pretreated with SP600125 (10 mM) for one hour prior to treating SB431542 (Fig. 5) blocked gremlin-mediated LOX induction in all with recombinant gremlin for an additional 24 h. Untreated cells TM cell strains tested (p < 0.001). Treatment with the inhibitors and SP600125-alone treated cells served as negative controls. alone did not have any effect on the expression of LOX or LOXL Gremlin-induction of cell-associated (Fig. 7A and C) and secreted proteins. The inhibitors alone or in combination with gremlin also (Fig. 7B and D) LOX and LOXL1 protein expression was inhibited by did not affect the health of TM cells as measured by LDH cytotox- SP600125 pretreatment. However, SP600125 only mildly blocked icity assay (Data not shown). gremlin induction of cell-associated LOXL2 and LOXL4 proteins, but We also previously reported that gremlin induced phosphory- did not affect their secreted forms. lation of SMAD2 and SMAD3 proteins within 15 min in TM cells and TGFb2 also phosphorylates and activates p38 MAPK in cultured maintained this phosphorylation up to 4 h. SMAD2 and 3 together TM cells as early as 15 min and maintains this phosphorylation state or individually form a complex with co-SMAD4 to regulate tran- for 4 h (Sethi et al., 2011b). To determine whether the p38 pathway scription of target genes (Shi and Massague, 2003). To determine regulates LOX protein expression, we employed SB203580, a se- whether SMAD3 transcriptionally regulates LOX proteins, we lective small molecule inhibitor of p38. Three TM cell strains were employed SIS3, a selective small molecule inhibitor of SMAD3. pretreated with SB203580 (10 mM) for one hour prior to treating Three TM cell strains were pretreated with SIS3 (10 mM) six hours with recombinant gremlin for an additional 24 h. Untreated cells prior to treating with recombinant gremlin for an additional 24 h to and SB203580-alone treated cells served as negative controls. study the expression of LOX and LOXL proteins. Untreated cells and Gremlin-induction of cell-associated LOXL1 and LOXL4 proteins SIS3 alone treated cells served as negative controls. Gremlin- was completely inhibited by SB203580 pretreatment (Fig. 8A and induction of cell-associated (Fig. 6A and C) and secreted (Fig. 6B C). SB203580 mildly decreased gremlin induction of LOX and only and D) LOX protein expression was inhibited by SIS3 pretreatment. marginally blocked gremlin-induced LOXL2. SB203580 also blocked Therefore, we concluded that gremlin induction of LOXs is medi- gremlin induction of secreted LOXL1, LOXL2 and LOXL4 proteins and ated by SMAD3 signaling. had little effect on blocking LOX secretion (Fig. 8B and D). 122 A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127

Fig. 4. TGFb receptor II inhibitor LY364947 blocks gremlin-induction of LOX proteins. Western immunoblots determining the effect of TGFBR inhibitor LY364947 on gremlin induction of cell-associated (A) and secreted (B) LOX proteins. Western immunoblots of TM cells pretreated with 5 mM of LY364947 followed by treatment with 1 mg/ml of gremlin for 24 h. Untreated and inhibitor only treated cells served as negative controls. GAPDH was used as a loading control for cell associated proteins. Blots shown are representative data obtained in one TM strain. The experiment was performed independently two times in three different TM cell strains. (CeD) Densitometry analysis of western immunoblots for cell associated (C) and secreted (D) LOXs are plotted as bar graphs. Two-tailed student’s t test was used to compare gremlin- and gremlin þ LY364549-treated samples. ***p < 0.0001. A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127 123

Fig. 5. TGFb receptor I/II inhibitor SB431542 blocks gremlin-induction of LOX proteins. Western immunoblots determining the effect of TGFBR inhibitor SB431542 on gremlin induction of cell-associated (A) and secreted (B) LOX proteins. Western immunoblots of TM cells pretreated with 5 mM of SB431542 followed by treatment with 1 mg/ml of gremlin for 24 h. Untreated and inhibitor only treated cells served as negative controls. GAPDH was used as loading control for cell associated proteins. Blots shown are representative data obtained in one TM strain. The experiment was performed independently two times in three different TM cell strains. (CeD) Densitometry analysis of western immunoblots for cell associated (C) and secreted (D) LOXs are plotted as bar graphs. Two-tailed student’s t test was used to compare gremlin- and gremlin þ SB431542-treated samples. **0.0001 < p < 0.01, ***p < 0.0001.

Collectively, these data illustrate a complex regulation of LOX Gremlin’s role in development and in several fibrotic diseases is proteins in TM cells, which may have significant implications in quite well known. It is not uncommon to find developmental genes fibrosis and glaucoma pathogenesis. re-expressed in pathological conditions, including cancer and inflammation. Prior to this work, our understanding of gremlin’s 4. Discussion involvement in TM fibrosis is currently limited to TGFb signaling- mediated gremlin induction of ECM proteins (Sethi et al., 2011a). LOX and LOXL oxidize peptidyl lysine and hydroxylysine resi- However, the role of gremlin in regulating ECM crosslinking has not dues in collagen and lysine residues in elastin to produce peptidyl been studied previously. alpha-aminoadipic-delta-semialdehydes. These aldehyde residues Our current results demonstrate that exogenous gremlin in- can spontaneously combine with vicinal peptidyl aldehydes or with duces mRNA and protein expression of LOX and LOXL genes in epsilon-amino groups of peptidyl lysine to form the covalent cross- multiple TM cell strains. We also demonstrated that gremlin uti- links that stabilize and decrease the solubility of fibrillar collagen lizes the TGFb signaling pathway to regulate LOX and LOXL proteins and elastin fibers in the ECM (Kagan and Li, 2003). Also, LOX and and that both the canonical SMAD as well as non-SMAD JNK1/2 and LOXL are involved in several connective tissue fibrotic diseases p38 MAPK signaling pathways are involved in this gremlin induc- (Sethi et al., 2012). It is likely that these enzymes also play a crucial tion. The results have been illustrated in Fig. 9, which suggests that role in other fibrotic diseases such as glaucoma. gremlin functions by inhibiting the endogenous BMP antagonism of Elastin and collagen proteins are elevated in the glaucomatous TGFb signaling in cultured TM cells. This leads to LOX and LOXL TM (Acott and Kelley, 2008), and a recent report used atomic force induction. These data highlight the significance of ECM cross- microscopy to demonstrate that glaucomatous TM tissues were linking by BMP antagonists like gremlin in the TM cells and high- significantly stiffer than the TM from non-glaucoma eyes (Last light that perturbations of the TGFbeBMP signaling pathways et al., 2011). Greater cross-linking of ECM proteins within the regulation of the ECM environment may lead to TM fibrosis and TM could lead to greater tissue stiffness. Overexpression of pro- glaucoma. Since TGFb and BMP pathways are endogenously active fibrotic molecules like TGFb2 has been shown to increase ECM in TM cells (Wordinger et al., 2002, 2007), the data raise important crosslinking (Sethi et al., 2011b; Tovar-Vidales et al., 2011; Welge- questions about the existence and regulation of a plethora of Lussen et al., 2000). downstream players including other intracellular and extracellular 124 A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127

Fig. 6. SMAD3 inhibitor SIS3 blocks gremlin-induction of LOX proteins. Representative western immunoblots of SMAD3 inhibitor SIS3 effects on gremlin induction of cell-associated (A) and secreted (B) LOX proteins. Western immunoblots of TM cells pretreated with 10 mM of SIS3 followed by treatment with 1 mg/ml of gremlin for 24 h. Untreated and inhibitor only treated cells served as negative controls for cell-associated proteins. GAPDH was used as loading control for cell associated proteins. Blots shown are representative data obtained in one TM strain. The experiment was performed independently two times in three different TM cell strains. (CeD) Densitometry analysis of western immunoblots for cell associated (C) and secreted (D) LOXs are plotted as bar graphs. Two-tailed student’s t test was used to compare gremlin- and gremlin þ SIS3-treated samples. **0.0001 < p < 0.01, ***p < 0.0001. A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127 125

Fig. 7. JNK inhibitor SP600125 blocks gremlin-induction of several LOX proteins. Representative western immunoblots on the effect of JNK inhibitor SP600125 on gremlin induction of cell-associated (A) and secreted (B) LOX proteins. Western immunoblots of TM cells pretreated with 10 mM of SP600125 followed by treatment with 1 mg/ml of gremlin for 24 h. Untreated and inhibitor only treated cells served as negative controls. GAPDH was used as loading control for cell associated LOXs. Blots shown are representative data obtained in one TM strain. The experiment was performed independently two times in three different TM cell strains. (CeD) Densitometry analysis of western immunoblots for cell associated (C) and secreted (D) LOXs are plotted as bar graphs. Two-tailed student’s t test was used to compare gremlin- and gremlin þ SP600125-treated samples. **0.0001 < p < 0.01, ***p < 0.0001. 126 A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127

Fig. 8. p38 MAPK inhibitor SB203580 blocks gremlin-induction of several LOX proteins. Representative western immunoblots of p38 MAPK inhibitor SB203580 effects on gremlin induction of cell-associated (A) and secreted (B) LOX proteins. Western immunoblots of TM cells pretreated with 10 mM of SB203580 followed by treatment with 1 mg/ml of gremlin for 24 h. Untreated and inhibitor only treated cells served as negative controls. GAPDH was used as loading control for cell associated LOXs. Blots shown are representative data obtained in one TM strain. The experiment was performed independently two times in three different TM cell strains. (CeD) Densitometry analysis of western immunoblots for cell associated (C) and secreted (D) LOXs are plotted as bar graphs. Two-tailed student’s t test was used to compare gremlin- and gremlin þ SB203580-treated samples. **0.0001 < p < 0.01, ***p < 0.0001.

Fig. 9. Proposed mechanism of gremlin regulation of LOXs in TM. Gremlin blocks BMP signaling pathway, which is unable to inhibit the endogenous TGFb signaling in TM cells. The endogenously active TGFb pathway can activate Smad, P38 and JNK pathways to induce different LOX genes and proteins in TM cells. A. Sethi et al. / Experimental Eye Research 113 (2013) 117e127 127 proteins that are involved in maintaining TM homeostasis, distur- de la Cruz-Merino, L., Henao-Carrasco, F., Garcia-Manrique, T., Fernandez- bance of which may lead to diseases like glaucoma. Salguero, P.M., Codes-Manuel de Villena, M., 2009. Role of transforming growth factor beta in cancer microenvironment. Clin. Transl. Oncol. 11, 715e720. We have previously reported that both TGFb2 and gremlin Fleenor, D.L., Shepard, A.R., Hellberg, P.E., Jacobson, N., Pang, I.H., Clark, A.F., 2006. phosphorylate and activate SMAD2/3 signaling in TM cells. TGFbeta2-induced changes in human trabecular meshwork: implications for e Knocking down the SMAD signaling pathway blocked TGFb2 in- intraocular pressure. Invest. Ophthalmol. Vis. Sci. 47, 226 234. Fuchshofer, R., Tamm, E.R., 2012. The role of TGF-beta in the pathogenesis of pri- duction of LOX and LOXL (Sethi et al., 2011b). By blocking SMAD mary open-angle glaucoma. Cell Tissue Res. 347, 279e290. signaling with SIS3, we also observed that gremlin induction of LOX Fuchshofer, R., Yu, A.H., Welge-Lussen, U., Tamm, E.R., 2007. Bone morphogenetic proteins is inhibited. These data not only imply that gremlin em- protein-7 is an antagonist of transforming growth factor-beta2 in human trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 48, 715e726. ploys the canonical SMAD pathway to regulate LOXs, but also em- Gottanka, J., Chan, D., Eichhorn, M., Lutjen-Drecoll, E., Ethier, C.R., 2004. Effects of phasizes the profibrotic effects of SMAD signaling in the TM. We TGF-beta2 in perfused human eyes. Invest. Ophthalmol. Vis. Sci. 45, 153e158. have previously observed that TM cells maintain basal phosphor- Inatani, M., Tanihara, H., Katsuta, H., Honjo, M., Kido, N., Honda, Y., 2001. Trans- forming growth factor-beta 2 levels in aqueous humor of glaucomatous eyes. ylation levels of both JNK and p38 MAPK (Sethi et al., 2011a, 2011b). Graefes Arch. Clin. Exp. Ophthalmol. 239, 109e113. Crosstalk and interaction between SMAD and MAPK pathways Javelaud, D., Mauviel, A., 2005. Crosstalk mechanisms between the mitogen- has been observed in several cell types and in a variety of normal activated protein kinase pathways and Smad signaling downstream of TGF- e and pathological conditions (de la Cruz-Merino et al., 2009; Jav- beta: implications for carcinogenesis. Oncogene 24, 5742 5750. Kagan, H.M., Li, W., 2003. Lysyl oxidase: properties, specificity, and biological roles elaud and Mauviel, 2005). Our data indicate that the basal level of inside and outside of the cell. J. Cell Biochem. 88, 660e672. MAPK kinase activity may be important in regulating LOX and LOXL Last, J.A., Pan, T., Ding, Y., Reilly, C.M., Keller, K., Acott, T.S., Fautsch, M.P., Murphy, C.J., in TM cells. Whether the basal MAPK kinase activity regulates LOX Russell, P., 2011. Elastic modulus determination of normal and glaucomatous human trabecular meshwork. Invest. Ophthalmol. Vis. Sci. 52, 2147e2152. enzymatic activity is a question that needs to be addressed. Ma, F.Y., Sachchithananthan, M., Flanc, R.S., Nikolic-Paterson, D.J., 2009. Mitogen Several additional questions are raised by our current results. activated protein kinases in renal fibrosis. Front. Biosci. (Schol. Ed.) 1, 171e187. First, it was surprising to find that all 5 LOX family genes are Quigley, H.A., Broman, A.T., 2006. The number of people with glaucoma worldwide in 2010 and 2020. Br. J. Ophthalmol. 90, 262e267. induced by gremlin. The LOX and LOXL enzymes may have different Sethi, A., Jain, A., Zode, G.S., Wordinger, R.J., Clark, A.F., 2011a. Role of TGFbeta/Smad specific roles in the TM including differences in substrate specificity signaling in gremlin induction of human trabecular meshwork extracellular and/or specific localization patterns. The potential relationship matrix proteins. Invest. Ophthalmol. Vis. Sci. 52, 5251e5259. fl Sethi, A., Mao, W., Wordinger, R.J., Clark, A.F., 2011b. Transforming growth factor- between the LOX proteins in regulating AH out ow in gremlin- beta induces extracellular matrix protein cross-linking lysyl oxidase (LOX) induced ocular hypertension and POAG is not known. It is also genes in human trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 52, not clear which LOX protein is important for normal TM homeo- 5240e5250. Sethi, A., Wordinger, R.J., Clark, A.F., 2012. Focus on molecules: lysyl oxidase. Exp. stasis and if any of the LOX proteins are directly involved in path- Eye Res. 104, 97e98. ogenesis of glaucoma. Future in vivo studies are needed to address Shepard, A.R., Millar, J.C., Pang, I.H., Jacobson, N., Wang, W.H., Clark, A.F., 2010. this question. The role of MAPK signaling in TM fibrosis and in Adenoviral gene transfer of active human transforming growth factor-{beta}2 fl regulating TM LOX enzymatic activity also needs further study. Our elevates intraocular pressure and reduces out ow facility in rodent eyes. Invest. Ophthalmol. Vis. Sci. 51, 2067e2076. current results provide a foundation to address these issues in Shi, Y., Massague, J., 2003. Mechanisms of TGF-beta signaling from cell membrane future studies. to the nucleus. Cell 113, 685e700. Tovar-Vidales, T., Clark, A.F., Wordinger, R.J., 2011. Transforming growth factor-beta2 utilizes the canonical Smad-signaling pathway to regulate tissue trans- Acknowledgments glutaminase expression in human trabecular meshwork cells. Exp. Eye Res. 93, 442e451. The authors would like to acknowledge grant support from the Tovar-Vidales, T., Roque, R., Clark, A.F., Wordinger, R.J., 2008. Tissue transglutaminase expression and activity in normal and glaucomatous human National Institute of Health-National Eye Institute (EY-017374). The trabecular meshwork cells and tissues. Invest. Ophthalmol. Vis. Sci. 49, 622e authors would also acknowledge Ankur Jain and Tara Tovar-Vidales 628. at the North Texas Eye Research Institute, UNT Health Science Tripathi, R.C., Li, J., Chan, W.F., Tripathi, B.J., 1994. Aqueous humor in glaucomatous eyes contains an increased level of TGF-beta 2. Exp. Eye Res. 59, 723e727. Center for his help in this project. We would also like to thank Lions Weinreb, R.N., Khaw, P.T., 2004. Primary open-angle glaucoma. Lancet 363, 1711e Eye Institute for Transplant and Research (Tampa, FL) for providing 1720. donor eyes used for preparing primary TM cell cultures. Welge-Lussen, U., May, C.A., Lutjen-Drecoll, E., 2000. Induction of tissue transglutaminase in the trabecular meshwork by TGF-beta1 and TGF-beta2. Invest. Ophthalmol. Vis. Sci. 41, 2229e2238. References Wordinger, R.J., Agarwal, R., Talati, M., Fuller, J., Lambert, W., Clark, A.F., 2002. Expression of bone morphogenetic proteins (BMP), BMP receptors, and BMP Acott, T.S., Kelley, M.J., 2008. Extracellular matrix in the trabecular meshwork. Exp. associated proteins in human trabecular meshwork and optic nerve head cells Eye Res. 86, 543e561. and tissues. Mol. Vis. 8, 241e250. AGIS, 2000. The Advanced Glaucoma Intervention Study (AGIS): 7. The relationship Wordinger, R.J., Fleenor, D.L., Hellberg, P.E., Pang, I.H., Tovar, T.O., Zode, G.S., between control of intraocular pressure and visual field deterioration. The AGIS Fuller, J.A., Clark, A.F., 2007. Effects of TGF-beta2, BMP-4, and gremlin in the Investigators. Am. J. Ophthalmol. 130, 429e440. trabecular meshwork: implications for glaucoma. Invest. Ophthalmol. Vis. Sci. Bachmann, B., Birke, M., Kook, D., Eichhorn, M., Lutjen-Drecoll, E., 2006. Ultra- 48, 1191e1200. structural and biochemical evaluation of the porcine anterior chamber perfu- Wordinger, R.J., Zode, G., Clark, A.F., 2008. Focus on molecules: gremlin. Exp. Eye sion model. Invest. Ophthalmol. Vis. Sci. 47, 2011e2020. Res. 87, 78e79.

Gremlin Utilizes Canonical and Non-Canonical TGF&Beta; Signaling

Gremlin Utilizes Canonical and Non-Canonical TGFΒ Signaling