Involvement of AP-1 in Interleukin-1α–Stimulated MMP-3 Expression in Human Trabecular Meshwork Cells

Abstract

purpose. Stromelysin-1 (MMP-3) degrades extracellular matrix and increases aqueous outflow. In the trabecular meshwork (TM), interleukin (IL)-1α is a potent inducer of MMP-3 expression. In different cells, IL-1α activates different signaling pathways, such as nuclear factor (NF)-κB–mediated protein expression, the phospholipase A₂ (PLA₂)–activated arachidonate cascade, and activator protein (AP)-1–associated transcription. In the present study, pharmacological tools were used to delineate the signaling mechanism involved in the effect of IL-1α on MMP-3 production in human TM cells compared with other ocular cells.

methods. Human TM and three other ocular cells (ciliary muscle, corneoscleral fibroblast, and lamina cribrosa) were cultured in 24-well plates in the presence or absence of IL-1α, with or without specific inhibitors of selected signaling pathways. Secreted proMMP-3 was quantified by ELISA, and MMP-3 activity was assayed by casein zymography.

results. IL-1α (5 ng/mL) increased proMMP-3 levels in human TM cells to 10- to 38-fold of control (P < 0.001). The effect of IL-1α was blocked by Gö6976, a protein kinase Cμ (PKCμ) inhibitor; PD98059, a mitogen-activated protein kinase kinase (MEK) inhibitor; SB202190, a p38 inhibitor; and SR11302, an AP-1 inhibitor; but not by inhibitors of casein kinase II, NFκB, PLA₂, phospholipase D (PLD), cyclooxygenases, lipoxygenase, or sphingomyelinase. SR11302 did not inhibit the effect of IL-1α on MMP-3 production in the other ocular cells tested.

conclusions. Based on the pharmacological effects of the inhibitors, the data indicate that activation of PKCμ, MEK, and p38 leading to the activation of AP-1 is critical to the IL-1α–stimulated upregulation of MMP-3 in human TM cells. Therefore, it is likely that compounds that activate the AP-1 pathway would upregulate the production of MMP-3 and improve aqueous outflow.

Matrix metalloproteinases (MMPs) are enzymes that catalyze the turnover of extracellular matrix. They are thought to regulate intraocular pressure (IOP) by improving the aqueous outflow through the trabecular meshwork (TM). A variety of MMPs such as MMP-1 (collagenase-1), MMP-2 (gelatinase A), MMP-3 (stromelysin-1), and MMP-9 (gelatinase B) have been shown to be produced by bovine, porcine, and human TM. ^{1 2 3 4 5} MMPs have also been detected in human aqueous humor. ⁶  

Evidence that MMPs may play a role in maintaining aqueous outflow was shown by Bradley et al. ⁷ who have reported that purified MMPs, such as MMP-3, increase the aqueous outflow rate in perfused human organ cultured eyes within 1 to 3 days of onset of treatment. Outflow-enhancing effects of the MMPs could be effectively blocked by synthetic MMP inhibitors, as well as by tissue inhibitors of metalloproteinases (TIMPs). The effect of MMP-3 is particularly intriguing, because stromelysins have the ability to degrade TM-associated extracellular matrix components such as proteoglycans, fibronectin, and laminin, in addition to gelatin and collagen. ⁸  

Furthermore, MMPs, especially MMP-3, have been implicated as mediators of the IOP-lowering effects of two popular glaucoma treatments: argon laser trabeculoplasty and topical application of prostanoids. For example, laser trabeculoplasty stimulates expression of MMP-3 in the juxtacanalicular region of the TM in human eye organ cultures. ³ Prostaglandin receptor agonists upregulate the expression of both MMP-3 and -1 in ciliary muscle cells. ^{9 10 11 12} This finding has been proposed as a means by which therapeutic prostanoids enhance uveoscleral outflow. 

MMP production or activation can be induced by cytokines and growth factors in a wide variety of tissues, ¹³ including those of the eye. ¹⁴ We and others have found that the cytokine interleukin (IL)-1α is a potent and highly efficacious activator of MMP-3 production in cultured human TM cells derived from both nonglaucomatous and glaucomatous donor eyes. ^{2 4 5} Based on this, it is therefore likely that IL-1α would increase aqueous outflow facility. Experimental results support this speculation. Perfusion of human organ culture eyes with IL-1α leads to a significant increase in aqueous outflow facility. ⁷ Intracameral injection of IL-1α in rats also enhances outflow rate in a dose-dependent manner. ¹⁵  

Despite its effect in aqueous outflow, potential local and systemic side effects of IL-1α effectively prohibit its usage as a therapeutic agent. Side effects reported for IL-1α include proinflammatory responses, induction of fever, destruction of cartilage matrix, bone resorption, thrombohemorrhagic lesions, hypercalcemia, exacerbation of autoimmune arthritis, central nervous system dysfunction, and muscle-wasting, among others. ^{16 17} Furthermore, topical ocular delivery of large molecules such as IL-1α is impractical because of corneal impenetrability. We hypothesize that, by understanding the signaling pathway(s) used by this cytokine in TM cells for the upregulation of MMP-3, it may be possible to develop other means to mimic the MMP-3 expression–inducing and outflow-enhancing effects of IL-1α while avoiding or minimizing its side effects. 

Biological actions of IL-1α involve the binding of the cytokine to its receptor, IL-1RI, which leads to the activation of multiple intracellular signaling mechanisms. ^{18 19 20 21 22 23 24} Based on the published information, the IL-1α–activated signaling mechanisms converge toward three major pathways: expression of the nuclear factor (NF)-κB regulation protein, phospholipase A₂ (PLA₂)–activated lipid transmitter production, and activator protein (AP)-1–associated transcription (Fig. 1) . Different cells may use different pathways, and it is not known which pathway is essential to the effect of IL-1α in TM. We used pharmacological tools to evaluate the contribution of the different mechanisms in the action of IL-1α in cultured human TM cells. We further investigated whether the same signaling mechanism is involved in other cultured human ocular cells, such as ciliary muscle cells, lamina cribrosa cells, and corneoscleral fibroblasts. 

Methods

Ten human TM cell lines, five each from donors with or without glaucoma, were used in these studies. They had been carefully isolated from explants of human donor eyes less than 24 hours after death and characterized as described. ^{25 26} Similarly, the human lamina cribrosa cell lines had been isolated and characterized previously. ^{25 27 28} Human corneoscleral fibroblasts, previously called corneal stroma cells, were characterized as described. ²⁷ Initial cultures of human ciliary muscle cells were a generous gift from Elke Lütjen-Drecoll ²⁹ and were subsequently adapted to the culture conditions in our laboratory. ³⁰  

Cell cultures were maintained at 5% CO₂ and 37°C in a medium consisting of Dulbecco’s modified Eagle’s medium with stabilized l-glutamine (Glutamax I; Invitrogen/Gibco, Grand Island, NY), supplemented with 10% fetal bovine serum (Hyclone, Logan, UT) and 50 μg/mL gentamicin (Invitrogen/Gibco). TM cells were routinely bead-passed (Cytodex 3 beads; Sigma-Aldrich, St. Louis, MO) and underwent brief (5–10 minutes) enzymatic dissociation with 0.05% trypsin-0.53 mM EDTA solution (Invitrogen/Gibco) solely for preparation of multiwell plates for assay. 

Twenty-four well plates of cell monolayers (estimated 90% confluence) were serum deprived for 24 hours and incubated with indicated compounds in serum-free medium for 24 hours in a final volume of 0.3 mL/well. Levels of secreted proMMP-3 were then evaluated in the cell supernatants, using a commercially available ELISA (The Binding Site, Birmingham, UK). ⁵ MMP-3 activity was assayed by casein zymography. ⁵ Cell morphology was routinely observed by light microscopy after treatment and viability of the cell monolayers was assessed by neutral red uptake. 

The selective AP-1 inhibitor retinoid SR11302 was provided by Mark Hellberg (Alcon, Fort Worth, TX). Diethyldithiocarbamic acid and indomethacin were obtained from Sigma-Aldrich. Arachidonyl trifluoromethyl ketone (AACOCF₃), bisindolylmaleimide I, curcumin, 5,6-dichloro-1-β-d-ribofuranosyl benzimidazole, fumonisin B1, Gö6976, Gö6983, NFκB-SN50, PD98059, SB202190, and SB203580 were purchased from Calbiochem (La Jolla, CA). 12-Epi-scalaradial was obtained from Biomol (Plymouth Meeting, PA). 

When data of two treatment groups were compared, they were evaluated by two-tailed Student’s t-test. One-way ANOVA followed by the Bonferroni test was used for multiple comparisons. Results were considered statistically significant if P < 0.05. 

Results

IL-1α is an efficacious and potent inducer of MMP-3 expression in the TM. ^{2 4 5} In cultured human TM cells, its effect was concentration-dependent, with an EC₅₀ of 0.42 ng/mL after a 24-hour incubation. Subsequent studies reported confirmed these findings (Table 1) . In TM cells derived from both nonglaucomatous and glaucomatous donor eyes, IL-1α at 5 ng/mL increased the proMMP-3 levels in the cell medium to 10 to 38 times control levels. In all cases, the stimulated levels were significantly different from the control levels (P < 0.001). 

Responses of the different human TM cell lines to IL-1α were very similar. Therefore, subsequent studies were performed with the TM35D cells, unless specified otherwise, because of its slightly faster proliferation rate. The concentrations of inhibitors used were approximately 10 times the reported IC₅₀ for the respective targeted proteins except for curcumin, 5,6-dichloro-1-β-d-ribofuranosyl benzimidazole, and indomethacin. These compounds were toxic to the cells at the higher concentrations, as indicated by reduction in neutral red uptake and change in cell morphology (data not shown). Therefore, they were used at a concentration equivalent to the respective reported IC₅₀, which was not cytotoxic. 

To determine whether the stimulatory effect of IL-1α on expression of MMP-3 depends on the activation of the NFκB pathway, cells were pretreated for 30 minutes with an NFκB inhibitor diethyldithiocarbamate (100 μM), ³¹ or NFκB-SN50 (20 μM), a cell-permeable peptide that inhibits translocation of active NFκB complexes into the nucleus, ³² before the addition of IL-1α. As seen in Table 2 , neither treatment decreased the MMP-3 expression induced by IL-1α; instead, NFκB-SN50 synergized with IL-1α in inducing MMP-3 expression, suggesting that the NFκB pathway may be involved in a feedback inhibition in the stimulatory effect of IL-1α but does not directly mediate its effect. Compounds affecting enzymes upstream of NFκB, such as the casein kinase II inhibitor 5,6-dichloro-1-β-d-ribofuranosyl benzimidazole (10 μM) or the protein kinase A activator forskolin (5 μM), also did not significantly affect the basal or IL-1α–stimulated production of MMP-3 (Table 2) . 

Compounds that interfere with PLA₂ activity were ineffective in blocking the IL-1α–induced expression of MMP-3, as well. For example, selective inhibitors such as AACOCF₃ (10 μM) and 12-epi-scalaradial (1 μM), as well as a nonselective PLA₂ inhibitor indomethacin (10 μM) did not significantly change the basal or stimulated proMMP-3 levels in the TM35D cells (Table 3) . Because indomethacin at the concentration used also suppressed cyclooxygenase activity, the absence of effect implies that activation of cyclooxygenase was also not essential to the action of IL-1α. In addition, curcumin, an inhibitor of both cyclooxygenase and lipoxygenase, at 500 μM, similarly did not prevent stimulation of MMP-3 expression by IL-1α (Table 3) . Hence, these data collectively indicate that the activation of PLA₂ and arachidonate cascade is not a necessity for the production of MMP-3 in human TM cells. 

By excluding two of the three major known signaling pathways of IL-1α, we expected to show that AP-1–associated transcription plays a crucial role in the stimulatory effect of IL-1α on MMP-3 expression in human TM cells. Indeed, pretreatment of the cells with the selective AP-1 inhibitor SR11302 ³³ (1 μM) significantly reduced the production of MMP-3 induced by IL-1α, without affecting the basal level (Table 4) . The inhibitory effect of SR11302 was concentration dependent with an IC₅₀ of approximately 10 nM (Fig. 2) . At 1 μM, a large portion of the effect of IL-1α was eliminated. SR11302 not only inhibited the IL-1α–induced increase in proMMP-3 levels, but also significantly suppressed the IL-1α–stimulated MMP-3 activity, as indicated by casein zymography (Fig. 3) . Inhibitors that affect certain upstream enzymes leading to the activation of AP-1 also downregulated the IL-1α–induced expression of MMP-3. For example, selective inhibitors of the p38 kinase (also known as p38^MAPK or HOG-1), SB202190, and SB203580, lowered the effect of IL-1α on proMMP-3 levels in a concentration-dependent manner (Table 4 , Fig. 4 ). Moreover, the PKC inhibitor Gö6976, which is selective for isozyme subtypes α, β, and μ, was effective in blocking the effect (Table 4 and Fig. 5 ). However, bisindolylmaleimide 1 (selective for PKCα, -β, -γ, -δ, and -ζ) and Gö6983 (selective for PKCα, -β, -γ, -δ, and -ε) were without effect (Table 4) , suggesting that the PKCμ isozyme is an important mediator of IL-1α activity in TM cells. 

Unfortunately, we were not able to test the effect of the nonselective PKC inhibitor staurosporine because of its toxicity to the cells (toxic at concentration as low as 10 nM; data not shown). Mitogen-activated protein kinase kinase (MEK; MAP/ERK kinase) inhibitor PD98059 also partially blocked the effect of IL-1α (Table 4) . In contrast, the sphingomyelinase inhibitor fumonisin B1 (0.3 μM; Table 4 ) and curcumin (500 μM; Table 3 ) which, in addition to being an inhibitor of cyclooxygenase and lipoxygenase, is also an inhibitor of c-Jun N-terminal kinase (JNK), did not affect proMMP-3 levels. These results suggest that, under the conditions in these studies, not all pathways leading to the activation of AP-1 were required for the induction of IL-1α. It appeared that the PKC/MEK and the p38 kinase pathways, but not the sphingomyelinase-ceramide or the JNK pathways, were indispensable. 

Taken together, the pharmacological profiles of these compounds demonstrate that AP-1–associated transcription is essential for the IL-1α–induced upregulation of MMP-3 (Fig. 6) . This was observed not only in the TM-35D cell line. Table 5 shows that SR11302 was also highly effective in blocking IL-1α–induced MMP-3 in other human TM cell lines, regardless of whether they were derived from glaucomatous or nonglaucomatous eyes. 

The AP-1–mediated, IL-1α–stimulated expression of MMP-3 appeared to be unique to the TM, because four human cell lines derived from three other ocular tissues did not share this property when tested under similar conditions. Table 6 shows that human lamina cribrosa cells (from glaucomatous and nonglaucomatous eyes), ciliary muscle cells, and corneoscleral fibroblasts exhibited a robust increase in proMMP-3 production after a 24-hour exposure to IL-1α (5 ng/mL). However, these increases were not inhibited by 1 μM SR11302. 

Discussion

In this study, confirming previous findings, IL-1α was an effective stimulator of MMP-3 production in cultured human TM cells. Use of selective, and in many cases redundant, inhibitors of crucial proteins or enzymes of each major signaling pathway related to the biological actions of IL-1α, showed that AP-1 activation is an important signaling pathway of the effect of IL-1α. The data also suggest that the NFκB-mediated transcription or the arachidonate cascade are probably not critical components, though we cannot completely exclude their involvement. 

The human MMP-3 gene is known to possess an NFκB-binding site in its promoter, ³⁴ but this transcription factor apparently is not involved in IL-1α induction of MMP-3 in the TM. Furthermore, an autoregulatory feedback loop has been shown to exist between NFκB and its endogenous inactivator IκBα in other cell types, ³⁵ where IL-1α stimulation of NFκB nuclear translocation resulted in a rapid (within 1 hour) synthesis of IκBα and the subsequent inactivation of NFκB itself. It is possible that a similar feedback loop exists in the TM and the potential initial effects of NFκB on MMP-3 expression in the TM may have been effectively negated during the 24-hour incubation period used for our studies. Alternatively, IL-1α may not stimulate the upstream enzymes that lead to NFκB activation in the human TM cell. 

Our data clearly support the importance of AP-1 in the stimulation of MMP-3 production by IL-1α in the TM cells. The AP-1 inhibitor SR11302 significantly suppressed the activity of IL-1α. This critical role of AP-1 in the effect of IL-1α on MMP-3 expression was also reported in other cells ^{36 37 38} and agrees well with the known presence of two AP-1 transcription factor–binding sites on the human MMP-3 gene. ³⁹ In the human TM cells, MMP-3 production was also significantly suppressed by inhibitors of three key enzymes in the signaling cascades leading to AP-1 formation: PKC, p38 kinase, and MEK. It is interesting to note that inhibition of any one of these enzymes was sufficient to reduce the effect of IL-1α significantly, suggesting that the concomitant activation of the PKC/MEK and p38 kinase pathways was necessary for the optimal activation of AP-1 and subsequent expression of MMP-3 (Fig. 6) . 

PKC is involved in all three of the major signaling pathways of IL-1α (Fig. 1) . It has been demonstrated that activation of PKC alone is sufficient to upregulate MMP-3 in the TM. ^{5 40} Furthermore, in the porcine TM cells, Alexander and Acott ⁴⁰ demonstrated that the PKCμ subtype, but not the other PKC isozymes, is critical for the increase in several MMPs induced by tumor necrosis factor (TNF)-α. Such a PKC isozyme–specific effect was also observed in the present study. In the human TM cells, the IL-1α–enhanced proMMP-3 level was significantly suppressed by Gö6976, an agent that inhibits the PKC subtypes α, and β, as well as μ (with high affinity; IC₅₀ = 20 nM). ⁴¹ In contrast, bisindolylmaleimide I, an inhibitor of PKCα, -β, -γ, -δ, and -ε ⁴¹ did not significantly affect TM cell response. Nor was the response significantly affected by the inhibitor Gö6983, which exhibits high affinity for the PKCα, -β, -γ, -δ, -ε, and -ζ subtypes. These data collectively suggest that the PKCμ isozyme is an important mediator of the effect of IL-1α on MMP-3 expression. It is possible that PKCμ serves as a final common pathway between the stimulatory effects of IL-1α and TNF-α on MMP-3 expression in the TM. 

SR11302 was effective in inhibiting the activity of IL-1α in all human TM cells tested, in cells derived from both nonglaucomatous and glaucomatous human donor eyes, though not necessarily to the same degree. This suggests that the essential involvement of AP-1–mediated transcription in IL-1α–induced MMP-3 is not an event peculiar to certain specific TM cell lines. Instead, it is probably a common phenomenon in the human TM. Moreover, this IL-1α dependence on AP-1 appears to be unique in the TM. SR11302 did not significantly affect the increased proMMP-3 levels induced by IL-1α in human cell lines derived from three other ocular tissues. Of the cell lines evaluated, only the response of human ciliary muscle cells appeared to be slightly reduced by SR11302; however, this reduction was not statistically significant. Nor was IL-1α–mediated MMP-3 production affected by SR11302 when tested in cultured human lamina cribrosa cells derived from either nonglaucomatous or glaucomatous eyes or in fibroblasts derived from the human corneal-scleral margin of a nonglaucomatous donor eye. Based on this limited survey, it seems that IL-1α can upregulate MMP-3 expression in many ocular cells, but different signaling pathways are recruited to mediate this effect. 

Identification of the AP-1 pathway as a critical mechanism in the effect of IL-1α on production of MMP-3 by human TM cells is a significant finding. IL-1α is a potent IOP-lowering cytokine. ^{7 15} Unfortunately, it cannot be used therapeutically because of its many other biological actions, especially those related to inflammation. Our findings may allow the identification of small-molecule AP-1 activators that will be useful in glaucoma therapy. These activators may be targeted at AP-1 itself or at mechanisms upstream of AP-1 activation, such as activators of p38 kinases, and the enzyme PKCμ. Compounds that selectively activate only the AP-1 pathway may offer the potential for minimization of the IL-1α side effects mediated by either the NFκB or arachidonate cascade pathways. Because the dependence of IL-1α on AP-1 appears to be selective in the TM but not in the other ocular cells tested, it is also probable that an AP-1 activator would increase MMP-3 expression only in the TM without affecting other ocular tissues. If this is true, AP-1 activators should not have side effects related to MMP-3 activation in other tissues. Nonetheless, the activation of AP-1 is known to affect the transcription of many other genes. There is a possibility that certain side effects are associated with the use of AP-1 activators. Regardless, this unique pharmacological mechanism is a very interesting and novel direction for the development of future compounds in the management of ocular hypertension. 

 Supported by Alcon Research, Ltd.

 Submitted for publication July 26, 2002; revised January 9, 2003; accepted February 20, 2003.

 Disclosure: D.L. Fleenor, Alcon Research, Ltd. (E, P); I.-H. Pang, Alcon Research, Ltd. (E, P); A.F. Clark, Alcon Research, Ltd. (E, P)

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

 

Figure 1.

 

Major signaling pathways involved in the various biological effects of IL-1α. This figure is a compilation of published findings from different cells and tissues. ^{18 19 20 21 22 23 24} AA, arachidonic acid; AP-1, activator protein 1; ATF2, activating transcription factor 2; CAPK, ceramide-activated protein kinase; Cer, ceramide; CK-II, casein kinase II; DAG, diacylglycerol; Elk-1, Eph-like kinase; ERK, extracellular signal-related kinase; IκB, inhibitor of nuclear factor κB; IKK, IκB-kinase; also known as conserved helix-loop-helix ubiquitous kinase, or “CHUK”; JNK, c-Jun N-terminal kinase; MAP, mitogen-activated protein; MEK, MAP/ERK kinase; MEKK, MEK kinase; MKK, MAP kinase kinase; NFκB, nuclear factor κB; NIK, NFκB-inducing kinase; p38, p38 kinase (also called p38^MAPK or HOG-1); PA, phosphatidic acid; PKA, protein kinase A; PKC, protein kinase C; PLA₂, phospholipase A₂; PLD, phospholipase D; Raf, a proto-oncogene; Rap1, GTP binding protein of the Ras superfamily; SM, sphingomyelinase; TAK1, transforming growth factor-β-activated kinase 1.

Table 1.

 

Effect of IL-1α on ProMMP-3 Production by Human TM Cells

Cell Line	Donor Age	Percent Vehicle Control
				Vehicle	IL-1α (5 ng/mL)
Nonglaucomatous eyes
TM-35D	6 mo	100 ± 1 (n = 198)	1878 ± 81 (n = 198)*
TM-16A	18 y	100 ± 4 (n = 24)	1610 ± 112 (n = 24)*
TM-75C	85 y	100 ± 5 (n = 12)	2578 ± 319 (n = 12)*
TM-79	58 y	100 ± 2 (n = 12)	983 ± 130 (n = 12)*
TM 332/344	47 y	100 ± 4 (n = 12)	1348 ± 110 (n = 12)*
Glaucomatous eyes
GTM-23D	67 y	100 ± 6 (n = 12)	1659 ± 137 (n = 12)*
GTM-76D	76 y	100 ± 4 (n = 12)	1205 ± 196 (n = 12)*
GTM-81C	88 y	100 ± 6 (n = 12)	3827 ± 324 (n = 12)*
GTM-83C	75 y	100 ± 5 (n = 12)	1582 ± 167 (n = 12)*
GTM-85B	85 y	100 ± 8 (n = 12)	1363 ± 145 (n = 12)*

Table 2.

 

Effects of Compounds Involved in the NFκB Pathway on MMP-3 Expression in HTM-35D Cells

Test Compound	Percent of Vehicle Control
		Control	Compound	IL-1α (5 ng/mL)	IL-1α+ Compound
Diethyldithiocarbamate (100 μM)	100 ± 12 (n = 9)	54 ± 16 (n = 9)	1368 ± 269* (n = 9)	1593 ± 391* (n = 9)
NFκB-SN50 (20 μM)	100 ± 7 (n = 9)	425 ± 134* (n = 9)	962 ± 49* (n = 9)	3867 ± 326* , † (n = 9)
5,6-dichloro-1-β-d-ribofuranosyl benzimidazole (10 μM)	100 ± 7 (n = 6)	136 ± 15 (n = 6)	1770 ± 257* (n = 6)	780 ± 369* (n = 6)
Forskolin (5 μM)	100 ± 4 (n = 12)	108 ± 5 (n = 12)	1242 ± 113* (n = 12)	Not tested

Table 3.

 

Effects of Inhibitors of the PLA₂ Pathway on MMP-3 Expression in HTM-35D Cells

Test Compound	Percent of Vehicle Control
		Control	Inhibitor	IL-1α (5 ng/mL)	IL-1α+ Inhibitor
AACOCF3 (10 μM)	100 ± 13 (n = 6)	85 ± 15 (n = 6)	1317 ± 268* (n = 6)	2578 ± 307* (n = 6)
12-Epi-scalaradial (1 μM)	100 ± 3 (n = 8)	171 ± 47 (n = 8)	1803 ± 398* (n = 8)	2037 ± 449* (n = 8)
Indomethacin (10 μM)	100 ± 7 (n = 6)	77 ± 7 (n = 6)	1359 ± 90* (n = 6)	1447 ± 294* (n = 6)
Curcumin (500 μM)	100 ± 10 (n = 7)	81 ± 10 (n = 7)	1659 ± 237* (n = 7)	1854 ± 81* (n = 7)

Table 4.

 

Effects of Inhibitors of the AP-1 Pathway on MMP-3 Expression in HTM-35D Cells

Test Compound	Percent of Vehicle Control
		Control	Inhibitor	IL-1α (5 ng/mL)	IL-1α+ Inhibitor
Fumonisin B1 (0.3 μM)	100 ± 10 (n = 7)	82 ± 13 (n = 7)	1659 ± 237* (n = 7)	1937 ± 101* (n = 7)
Bisindolylmaleimide 1 (0.1 μM)	100 ± 3 (n = 6)	102 ± 17 (n = 6)	2543 ± 606* (n = 6)	3433 ± 900* (n = 6)
Gö6983 (0.5 μM)	100 ± 8 (n = 10)	130 ± 26 (n = 10)	1281 ± 83* (n = 10)	2563 ± 258* , † (n = 10)
Gö6976 (0.2 μM)	100 ± 3 (n = 20)	122 ± 17 (n = 20)	2643 ± 313* (n = 20)	873 ± 136* , † (n = 20)
PD98059 (100 μM)	100 ± 5 (n = 8)	96 ± 13 (n = 8)	1237 ± 104* (n = 8)	807 ± 206* , † (n = 8)
SB202190 (0.1 μM)	100 ± 3 (n = 20)	107 ± 10 (n = 20)	2437 ± 326* (n = 20)	711 ± 94* , † (n = 20)
SB203580 (5 μM)	100 ± 7 (n = 6)	103 ± 15 (n = 6)	1469 ± 44* (n = 6)	293 ± 46* , † (n = 6)
SR11302 (1 μM)	100 ± 5 (n = 13)	89 ± 14 (n = 6)	1737 ± 279* (n = 13)	322 ± 62* , † (n = 13)

Figure 2.

 

Inhibition of IL-1α–stimulated MMP-3 expression by SR11302. Cells were pretreated with the indicated concentration of SR11302 for 30 minutes followed by a 24-hour incubation of IL-1α (5 ng/mL). Data are the mean ± SEM (n = 6).

Figure 3.

 

Effect of IL-1α with or without SR11302 on MMP-3 activity in TM35D cell supernatant. (A) A representative casein zymogram of concentrated supernatants of TM35D cells treated with vehicle, IL-1α (5 ng/mL), or IL-1α (5 ng/mL) with SR11302 (1 μM) for 24 hour. Numbers on the left side of gel image represent masses of molecular standards (in kilodaltons). (B) Summary of effect of IL-1α with or without SR11302 on casein hydrolytic activity of TM35D cell supernatant. Data are the mean ± SEM of results in five independent studies. *Significant difference (P < 0.05) between the groups by ANOVA with the Bonferroni test.

Figure 4.

 

Inhibition by SB202190 of IL-1α–stimulated MMP-3 expression. Human TM cells were pretreated with the indicated concentration of SB202190 for 30 minutes followed by a 24-h incubation of IL-1α (5 ng/mL). Data are the mean ± SEM (n = 6).

Figure 5.

 

Inhibition of IL-1α–stimulated MMP-3 expression by Gö6976. Human TM cells were pretreated with the indicated concentration of Gö6976 for 30 minutes followed by a 24-hour incubation of IL-1α (5 ng/mL). Data are the mean ± SEM (n = 6).

Figure 6.

 

Proposed critical signaling pathway for IL-1α–stimulated MMP-3 expression in cultured human TM cells. Abbreviations are defined in Figure 1 .

Table 5.

 

Effect of SR11302 on ProMMP-3 Level in Various HTM Cells

Cell Line	Percent of Vehicle Control
		Vehicle	IL-1α (5 ng/mL)	IL-1α (5 ng/mL)+ SR11302 (1 μM)
Nonglaucomatous eyes
TM-35D	100 ± 5 (n = 13)	1737 ± 279 (n = 13)*	322 ± 62 (n = 13)* , †
TM-16A	100 ± 8 (n = 12)	1873 ± 139 (n = 12)*	165 ± 19 (n = 12), †
TM-75C	100 ± 5 (n = 12)	2578 ± 319 (n = 12)*	1118 ± 248 (n = 12)* , †
TM-79	100 ± 2 (n = 12)	983 ± 130 (n = 12)*	259 ± 34 (n = 12)* , †
TM 332/344	100 ± 4 (n = 12)	1348 ± 110 (n = 12)*	231 ± 35 (n = 12)* , †
Glaucomatous eyes
GTM-23D	100 ± 6 (n = 12)	1659 ± 137 (n = 12)*	707 ± 93 (n = 12)* , †
GTM-76D	100 ± 4 (n = 12)	1205 ± 196 (n = 12)*	534 ± 67 (n = 12)* , †
GTM-81C	100 ± 6 (n = 12)	3827 ± 324 (n = 12)*	1332 ± 180 (n = 12)* , †
GTM-83C	100 ± 5 (n = 12)	1582 ± 167 (n = 12)*	617 ± 96 (n = 12)* , †
GTM-85B	100 ± 8 (n = 12)	1363 ± 145 (n = 12)*	209 ± 45 (n = 12), †

Table 6.

 

Effect of SR11302 on ProMMP-3 Levels in Other Human Ocular Cells

Cell Line	Donor Age	Percent of Vehicle Control
				Vehicle	IL-1α (5 ng/mL)	IL-1α (5 ng/mL)+ SR11302 (1 μM)
Nonglaucomatous eyes
Lamina cribrosa	82	100 ± 5 (n = 9)	2873 ± 434 (n = 9)*	2839 ± 381 (n = 9)*
Corneoscleral fibroblast	82	100 ± 2 (n = 9)	1859 ± 226 (n = 9)*	1792 ± 232 (n = 9)*
Ciliary muscle	67	100 ± 7 (n = 9)	2376 ± 430 (n = 9)*	1825 ± 382 (n = 9)*
Glaucomatous eyes
Lamina cribrosa	82	100 ± 8 (n = 9)	2210 ± 121 (n = 9)*	2459 ± 233 (n = 9)*

The authors thank Mari Engler and Sherry English-Wright for providing initial cultures of the human TM and lamina cribrosa cells; Karen David for providing the human corneoscleral fibroblasts; H. Thomas Steely for assistance in the analysis of zymographic images; Paula Billman and the Central Florida Lions Eye and Tissue Bank for the procurement of donor tissues, and Ted Acott for technical advice in zymography.