February 2004
Volume 45, Issue 2
Free
Glaucoma  |   February 2004
Selenium’s Effects on MMP-2 and TIMP-1 Secretion by Human Trabecular Meshwork Cells
Author Affiliations
  • Shannon M. Conley
    From the Departments of Ophthalmology and
  • Roberta L. Bruhn
    Epidemiology, University of Arizona, Tucson, Arizona.
  • Parham V. Morgan
    From the Departments of Ophthalmology and
  • W. Daniel Stamer
    From the Departments of Ophthalmology and
Investigative Ophthalmology & Visual Science February 2004, Vol.45, 473-479. doi:https://doi.org/10.1167/iovs.03-0767
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Shannon M. Conley, Roberta L. Bruhn, Parham V. Morgan, W. Daniel Stamer; Selenium’s Effects on MMP-2 and TIMP-1 Secretion by Human Trabecular Meshwork Cells. Invest. Ophthalmol. Vis. Sci. 2004;45(2):473-479. https://doi.org/10.1167/iovs.03-0767.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. Because of the observed increase in incidence of glaucoma among some individuals taking selenium as a dietary supplement, the present study was undertaken to investigate mechanisms of selenium-induced changes in homeostasis of human trabecular meshwork (HTM) cells. Specifically, the impact of selenium on matrix metalloproteinases (MMPs), their inhibitors (tissue inhibitors of metalloproteinases; TIMPs), and the second messengers that regulate MMP expression was investigated in an HTM cell culture model.

methods. HTM cell cultures were treated with an organic selenium compound (methyl seleninic acid), and changes in secretion and activity of MMPs and TIMPs were analyzed by Western blot and zymography. Changes in extracellular-signal–related kinases 1 and 2 (ERK1/2) and phospho-ERK1/2 levels were monitored by Western blot analysis of whole-cell lysates prepared from selenium-treated cells. Photographs of cultures over time were used to document selenium-induced changes in cell morphology.

results. Treatment of HTM cells with selenium for 24 hours at doses ranging from 1 to 10 μM caused a dose-dependent decrease in the secretion of MMP-2 and TIMP-1. Treatment for 6 hours revealed a significant decrease in MMP-2 and TIMP-1 at the highest dose. MMP-1, -3, and -9 and TIMP-2 were either not detected or their secretion was not consistently influenced by selenium treatment. Selenium treatment caused a significant decrease in ERK1/2 phosphorylation, but no change in overall ERK protein levels. Selenium treatment resulted in dose-dependent, reversible changes in HTM cell–matrix associations.

conclusions. Selenium-induced changes in MMP-2/TIMP-1 secretion may alter the balance of extracellular matrix turnover in the conventional outflow pathway and cause an increase in intraocular pressure that eventually leads to glaucoma.

Selenium is an essential element found in many foods and recently has been touted as a nutritional supplement with chemopreventive properties. Selenium targets cancer cells by stimulating apoptosis, 1 inhibiting angiogenesis, 1 2 3 and/or acting as an antioxidant by supporting enzyme systems such as the glutathione peroxidase system. In human umbilical vein endothelial cells (HUVECs), treatment with an organic selenium compound called methyl seleninic acid (MSeA) inhibits vascular endothelial growth factor and matrix metalloproteinase (MMP)-2, two proteins that contribute to the process of angiogenesis. 3 That selenium is an antiangiogenic agent is also supported by the observations that selenium treatment causes retraction of newly formed HUVEC capillary tubes in culture models 3 and that microvessel density in mammary tumors is significantly reduced after selenium treatment. 4 Selenium also affects tumor cell invasion by stimulating apoptosis or by altering the functionality or expression of adhesion molecules and cell–matrix-regulating enzymes as seen in HT1080 fibrosarcoma cells. 5  
The recent publication of a large epidemiologic study assessing the chemopreventive effects of selenium confirmed the clinical significance of in vitro evidence. The Nutritional Prevention of Cancer (NPC) clinical trial reported that selenium supplementation leads to significant reduction in overall cancer and specifically to a reduction in the incidence of prostate, lung, and colon cancers. 6 Unfortunately, the researchers also reported an increased incidence of glaucoma in some participants who received selenium supplements. 7 Initial review of adverse events in the NPC trial reported that selenium-supplemented study participants showed a significantly elevated hazard ratio (HR) for acquiring glaucoma (HR, 1.78; 95% confidence interval [CI] = 1.12–2.82). 7 The effect was accentuated when data from the portion of the trial when participants were allowed to choose whether to continue receiving supplements were analyzed. Participants who chose to maintain the selenium regimen had an elevated HR for glaucoma of 10.13 (95% CI = 1.32–77.62) compared with those who had never taken supplements. 7 To investigate this problem further, the NPC trial set up a Data Safety and Monitoring Board (DSMB) which reanalyzed the data and found that there was approximately a 40% increase in incidence of glaucoma among selenium-supplemented individuals, but that the estimates did not attain statistical significance. However, when the study participants were stratified by gender, the DSMB found that women had a statistically significant increase in risk of glaucoma, with an HR of 9.52 (95% CI = 1.20–75.31). 7  
Glaucoma is the second leading cause of irreversible blindness in the United States 8 9 and is generally characterized by death of retinal ganglion cells with subsequent loss of vision. Usually coincident with loss of retinal ganglion cells is increased intraocular pressure resulting from a decrease in aqueous humor outflow through the conventional outflow pathway. 10 Understanding mechanisms that control movement of aqueous fluid through the trabecular meshwork (TM) and out of the eye is critical to developing more effective therapy for people with glaucoma. 
The conventional outflow pathway is organized with a filter, consisting of trabecular lamellae covered with human TM (HTM) cells, in front of a resistor, consisting of juxtacanalicular HTM cells and the inner wall of Schlemm’s canal. Although no gross changes in HTM cells have been seen in glaucomatous eyes, 10 cellular changes that result in increased resistance to outflow are hypothesized to play a role in primary open-angle glaucoma, the most common form. 11 12 TM cells regulate the formation and turnover of extracellular matrix (ECM) in the conventional outflow pathway, and several research groups 13 14 15 16 hypothesize that defects in ECM turnover may lead to an accumulation of matrix materials over time and impede outflow. 
ECM turnover is tightly regulated by balancing degradation (by MMPs/TIMPs) with construction (by a variety of structural proteins). MMPs are a family of zinc-dependent enzymes secreted from many cell types, including those in vascular endothelia 3 and trabecular meshwork. 17 MMPs are responsible for digesting ECM and regulating ECM turnover. Changes in MMP levels can affect outflow in specific cases such as after laser trabeculoplasty 18 19 or on artificial manipulation of ECM turnover balance in the human anterior chamber perfusion system. 20 MMPs are secreted as zymogens and proteolytically cleaved to their active forms. In the TM several MMPs and tissue inhibitors of matrix metalloproteinase (TIMPs)—MMP-1, -2, -3, -9, and -14, TIMP-1, -2—are secreted and thought to help maintain homeostasis of the conventional outflow architecture. 13 14 17 21 22 23 Regulation of these proteins is particularly complex. For example, MMP-2 activation relies on a tight balance of activation through a TIMP-2–dependent ternary complex, and inhibition by TIMP-1 and -2 (under certain tissue conditions). 24  
The precise signaling mechanism by which selenium affects secretion of MMP is not known. There is evidence that the mitogen-activated protein kinase (MAPK) cascade, specifically the extracellular signal related kinases (ERK1/2), may be involved in MMP signaling in TM cells after stimulation with a variety of growth factors. 25 26 Platelet-derived growth factor and phorbol ester (TPA) both have been shown to activate ERK1/2 and stimulate MMP-2 secretion. 25 There is also evidence that selenium may interact with the ERK1/2 cascade, leading to as yet unspecified downstream events. 3 However, the specific molecular target through which selenium exerts its effects on the MAPK signaling pathway is not known, nor has the link between selenium, ERK signaling, and MMP secretion been defined. 
The observation that selenium interferes with MMP secretion in HUVECs indicates a mechanism by which selenium may decrease outflow through the trabecular meshwork. That is, selenium may adversely affect the balance of ECM degradation and formation in the TM and impede outflow facility. Thus, the purpose of the present study was to test the hypothesis that selenium adversely affects MMP and TIMP secretion by HTM cells. To test this hypothesis, cultured HTM cells were treated with MSeA and assayed for secretion of various MMPs and TIMPs and changes in the phosphorylation status of ERK1/2. 
Materials and Methods
Cell Culture
HTM cells were isolated with a blunt dissection technique followed by ECM digestion, and were cultured as described previously. 27 Six cell lines from human donors without a history of glaucoma were used for this study (TM26, TM37, TM60, TM61, TM83, and TM84). At least two different cell lines were tested for each experimental paradigm. Cells were grown in T75 culture flasks and seeded into six-well plates for experiments. HTM cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum, penicillin, streptomycin, and glutamine (100 U/mL, 0.1 mg/mL, 0.29 mg/mL respectively; Invitrogen, San Diego, CA) and grown in humidified air containing 5% CO2 at 37°C. HUVECs were purchased from Clonetics (San Diego, CA) and cultured according to the manufacturers’ recommendations. All HTM cells used in these experiments were confluent more than 1 week before testing. HUVECs were confluent (>1 week) or preconfluent (60%–80% confluent) as noted in the Results section. 
Selenium Treatment
Concentrated MSeA was the generous gift of Clement Ip of the Roswell Park Cancer Institute (Buffalo, NY), and was diluted to stock concentrations (1 mM) in sterile phosphate-buffered saline (PBS) and stored at −20°C (MSeA was never freeze thawed more than twice). All cells were washed three times with PBS and serum starved for 24 hours before selenium treatment. MSeA was added to fresh serum-free medium for final concentrations of 100 nM or 1, 2, 5, or 10 μM, as described. After cells were washed three times with PBS, the fresh, selenium-containing medium or serum-free medium (control) was added to the cells. Positive controls were treated with fresh serum-free medium containing 10 ng/mL TPA (Sigma-Aldrich, St. Louis, MO). After 6 or 24 hours, medium was collected, concentrated 30-fold using spin columns (Centricon; Millipore, Bedford, MA), and aliquoted. Cell lysates were prepared by the addition of 100 μL of 2× sample buffer (125 mM Tris-HCl [pH 6.8], 4% SDS, 20% glycerol, 0.01% bromophenol blue, 10% β-mercaptoethanol) to each well, followed by scraping. Sample buffer was added 1:1 to concentrated medium for Western blot analysis, and a 2:1 solution of sample and β-mercaptoethanol–free buffer was added to concentrated medium for zymography. Western blot samples and cell lysates were boiled for 10 minutes before storage at −20°C. 
Immunoblot Analysis
Whole-cell lysates or concentrated medium in 2× sample buffer were electrophoresed into 14% or 10% polyacrylamide gels containing 0.1% SDS. Fractionated proteins were transferred to nitrocellulose by a commercial system (Mini Transblot; Bio-Rad, Hercules, CA). Blots were blocked for 1 hour at approximately 25°C in Tris-buffered saline (137 mM NaCl, 25 mM Tris, and 2.7 mM KCl) containing 0.2% Tween-20 (TBST) and 5% (wt/vol) nonfat dry milk and then probed with anti-MMP-2 IgG (1:1000), anti-MMP-3 IgG (1:1000), anti-MMP-9 IgG (1:200), anti-MMP-1 IgG (1:200), anti-TIMP-1 IgG (1:1000), anti-TIMP-2 IgG (1:200), anti-ERK1/2 IgG (1:1000), or anti-phosphoERK1/2 (1:1000) overnight at 4°C (MMP/TIMP IgG from Oncogene Research Products, Boston, MA; ERK antibodies from Cell Signaling Technologies, Beverly, MA). The blots were washed (four times for 15 minutes each) in TBST and incubated for 1 hour at 25°C with horseradish peroxidase–conjugated secondary antibodies in 5% milk in TBST (goat anti-mouse, 1:5000; goat anti-rabbit, 1:5000). The blots were washed in TBST (four times for 15 minutes each), and enhanced chemiluminescence (Amersham Biosciences, Arlington Heights, IL) was used to visualize specific labeling. Blots were digitized with a gel-documentation system (EpiChemi II Darkroom; UVP, Upland, CA), and densitometry was performed on computer (Laboratory Works imaging software, ver. 4.0.0.8; Ultraviolet Products Inc., Upland, CA). 
Zymography
Concentrated medium in β-mercaptoethanol–free sample buffer was electrophoresed into 10% SDS-PAGE gels containing 0.1% gelatin. Gels were washed in 10 mM Tris-HCl (pH 7.4) with 2.5% Triton X-100 for 1 hour and incubated overnight at 37°C in incubation buffer (50 mM Tris-Cl [pH 7.5], 150 mM NaCl, 10 mM CaCl2, and 0.05% NaN3). Gels were stained with Coomassie blue (10% acetic acid, 25% isopropanol, 0.025% Coomassie blue dye) for 2 to 3 hours, and destained in methanol/acetic acid (10%/10%). Zymograms were imaged using the gel-documentation system (UVP), and densitometry was performed as for immunoblot analysis. 
Cell Morphology
Phase-contrast photography performed over time with an inverted microscope (IX70; Olympus, Tokyo, Japan) and image-processing software (Magna Fire software, ver. 2.1a; Optronics, Goleta, CA) enabled the documentation of morphologic changes after selenium treatment. Cells were imaged before serum starvation, after serum starvation, immediately after treatment, and 2 weeks after treatment. 
Results
Control Experiments with HUVECs
As a control, we treated confluent and preconfluent HUVEC cultures with MSeA for 24 hours at the following doses: 1, 2, 5, and 10 μM. Figure 1 summarizes our results. Figure 1A shows a representative Western blot and zymogram demonstrating that secretion of MMP-2 from preconfluent cells decreased after selenium treatment, as reported previously. In contrast, Figure 1B demonstrates that treatment with selenium did not cause a significant decrease in secretion of MMP-2 from confluent HUVECs. Figure 1C graphically depicts the magnitude of decrease in secretion of MMP-2 from preconfluent cells when compared with average densitometry results from confluent cells (n = 2). 
Long-Term, High-Dose Selenium Treatment
After confirming previously published results using HUVEC cultures, we tested confluent HTM cultures using the same doses as for the control. We chose to study only confluent HTM cell cultures (>1 week after confluence) because of the nonproliferative nature of TM cells in vivo and evidence that confluent and mature cell cultures in vitro most accurately model TM cells in vivo. 28 TPA, an indirect stimulator of MMP-2 activity, was used as a positive control. 21 Several MMPs secreted by HTM were examined, and Figure 2 shows representative results from experiments conducted in the presence of selenium for 24 hours. Figure 2A shows the effects of selenium on secretion of select proteins in concentrated conditioned medium. Several proteins were detected inconsistently: For example, MMP-9 was observed only in TPA-treated cells; MMP-3 was observed in approximately 50% of experiments, but never showed a reliable trend; TIMP-2 was detected in fewer than 20% of experiments; and MMP-1 was not detected at all (results not shown). Figure 2B is a representative Western blot (top) and zymogram (bottom) that shows decreased secretion of pro-MMP-2 from HTM cells after MSeA treatment, but no detectable active MMP-2. Mobility control for active MMP-2 in HT1080 cells is shown in lane 1. TIMP-1 secretion is shown in the representative blot in Figure 2C . TIMP-1 secretion also decreased as a function of selenium treatment. The changes in MMP-2 and TIMP-1 secretion were reliably documented in five separate experiments (using cell lines TM26, TM83) that are graphically summarized in Figure 2C . Statistical analysis indicates that secretion of MMP-2 and TIMP-1 was significantly decreased after treatment with MSeA in a dose-dependent fashion (two-tailed Student’s t-test, P < 0.05). Because selenium’s effects on TIMP-1 and MMP-2 were the most reliable results and because changes in MMP-2 secretion after selenium treatment had been documented, MMP-2 and TIMP-1 became the focus of future experiments. 
Long-Term, Low-Dose Selenium Treatment
After confirming that selenium significantly affects MMP-2 and TIMP-1 secretion at high doses, we wanted to determine whether similar effects would be apparent at lower, more physiologically relevant doses. To that end, cells were tested as described earlier at selenium concentrations of 100 nM and 1 μM. These concentrations closely mimic physiological levels and levels expected in selenium-supplemented individuals, respectively. Normal selenium concentrations in aqueous humor are approximately 10 ng/mL (∼100 nM). 29 Plasma selenium concentrations in selenium-supplemented individuals were typically 10 times higher (∼1 μM) than in nonsupplemented study participants. 7 However, in 15 experiments using three different HTM cell lines, we did not identify any reliable effects of selenium on secretion of MMP-2 or TIMP-1 (not shown), although both proteins were regularly detected. While the lower doses may be more physiologically relevant, our in vitro assay did not detect changes in protein secretion during the limited duration (24 hours) of the experiment. Because more than 48 hours in serum-free medium caused a dramatic decrease in the viability of our HTM cell cultures (data not shown), we were prevented from conducting experiments of longer duration. 
Short-Term, High-Dose Selenium Treatment
To see whether the observed decrease in secretion of MMP-2 after high doses of selenium occurs at time points earlier than 24 hours, cells were treated with 2, 5, or 10 μM MSeA, or with the control doses, and the samples were collected after 6 hours and analyzed. Figure 3A shows representative Western blots and a zymogram illustrating the results of the short-term experiments. Figure 3C shows that 6-hour treatment caused a significant decrease in MMP-2 secretion at the highest selenium dose (10 μM) but not at other doses. TIMP-1 secretion was similarly decreased at the highest doses (Fig. 3B) . As before, no active MMP-2 was detected. Figure 3D shows that at the early time point, zymography results were not consistent with Western blot results, showing no change in pro MMP-2. MMP-9, -3, and -1 and TIMP-2 were not detected in the short-term experiments (data not shown). 
MAPK Activation
Available evidence suggests that organic selenium compounds activate the ERK1/2 MAPK pathway. 3 To see whether this pathway is activated in HTM cells and to help determine intracellular mechanisms by which selenium may cause the observed decrease in MMP-2/TIMP-1 secretion, we conducted experiments in our cell culture model. Serum-starved cells (24 hours) were treated with MSeA (1, 2, 5, and 10) for an additional 24 hours. Whole-cell lysates were analyzed for both ERK1/2 and phospho-ERK1/2. Figure 4A shows representative Western blots demonstrating that total ERK protein remained unchanged after selenium treatment but that the quantity of phosphorylated ERK protein is dramatically decreased after treatment. Figure 4B displays graphic representations of average densitometry results from three experiments in two different cell lines. On average, phosphorylated ERK proteins decreased by approximately 60% after treatment with high doses of selenium (when compared with the control), and statistical analyses show that the decrease in phosphorylated ERK after selenium treatment is dose dependent and statistically significant (P < 0.05). 
Morphologic Changes
HTM cells were confluent for at least 1 week when used in the experiments. The cells were photographed after 24 hours of serum starvation, 24 hours after addition of selenium, and 2 weeks after treatment cessation. In all cases (n = 4), selenium treatment caused dose-dependent morphologic changes in the cell cultures. Figure 5 shows a representative set of micrographs demonstrating typical changes. Figure 5A shows a sample field from each of the dose ranges, with higher doses causing the most pronounced morphologic changes. Before application of selenium and after serum starvation, all wells looked similar to the negative control. Of interest, changes in the cultures were cell-line dependent. For example, one cell line exhibited severe changes (HTM 83), with treated cells lifting entirely off the plates and remaining tethered only by secondary cell associations (as seen in Fig. 5A , doses 5 and 10 μM). In another cell line, changes at the highest doses were less severe (HTM 61), with high-dose treated cells resembling the moderate-dose (1 and 2 μM) cells shown in Figure 5 . Morphologic changes were not associated with significant cell death, as evidenced by comparable levels of intracellular proteins (e.g., ERK1/2) in treated versus control cells (data not shown). Morphologic changes were reversible after removing the selenium. In all cases, cells reattached to the plates and began to spread out again and migrate. Figure 5B shows two representative fields from the 5-μM–treated well shown in Figure 5A , 2 weeks after selenium treatment ceased. In the image on the left, the mass of cells had re-adhered to the plate in a clump. The image on the right is more representative of the entire well after 2 weeks; as pictured, the cells were starting to return to a regular monolayer. 
Discussion
Results of the present study provide evidence that support a molecular mechanism underlying selenium-induced ocular hypertension. Treatment of HTM cultures with an organic selenium compound (MSeA) led to decreased secretion of MMP-2 and TIMP-1, two key players in ECM turnover in the conventional outflow pathway of humans. If malfunctions in ECM regulation lead to ocular hypertension and glaucoma, then selenium’s effects on ECM homeostasis may also lead to ocular hypertension. In addition, our experiments indicate that selenium affects cell–matrix associations and inhibits activation of the MAPK pathway, confirming a role for this signaling cascade in ECM regulation in the conventional outflow pathway. 
Several forms of selenium exist and are used for supplementation. In the Nutritional Prevention of Cancer trial, selenized yeast was used as the selenium supplement with placebo groups given a nonselenized yeast alternative. Selenized yeast contains selenized amino acids, such as selenomethionine and selenocysteine, which are metabolized by β-lyase to hydrogen selenide. 30 Hydrogen selenide is the essential selenium form in selenoproteins; however, it is also the form responsible for genotoxicity. Although most tested forms of selenium have some antineoplastic effects in animals, evidence suggests that methylselenol, a metabolite of hydrogen selenide and other selenocompounds such as selenomethylselenocysteine has very potent antineoplastic effects and does not cause the DNA strand breaks responsible for genotoxicity. 31 MSeA is a direct precursor of methylselenol that bypasses the need for β-lyase activation. This compound has been shown to have the typical antitumorigenic effects of selenium compounds, including inducing apoptosis and inhibiting cell accumulation. 31 32 Because the use of MSeA obviates the need for enzymatic activation, it is also a good choice for in vitro laboratory work. 
Differences in selenium’s effects on HUVECs versus HTM cells in our study reveal an interesting paradigm regarding the function of MMPs and TIMPs in their respective tissues. For example during angiogenesis, where HUVECs are used as a model, cells degrade matrix by secreting MMPs during their growth and migratory phases. For this reason, preconfluent cells are used in the HUVEC experimental setup. 3 In contrast, in TM models, HTM cells are typically studied only after confluence has been maintained for at least 1 week. 28 Careful studies by Alvarado et al. 28 have demonstrated that only after such a time did HTM cells in culture resemble their in vivo counterparts by a variety of measures. In our study, selenium decreased secretion and activity of MMP-2 in preconfluent HUVECs and had no effect on confluent cells. In contrast, selenium’s effects on MMP and TIMP secretion in HTM cells were in mature confluent monolayers. 
Although many ECM regulators are involved in TM homeostasis, we chose to focus on MMP-2 for two reasons: First, studies have shown a role for MMP-2 in selenium-mediated ECM regulation, 13 14 23 and, second, our results with other MMPs were inconsistent. MMP-2 activation is extremely complex and involves several steps. Because it is impossible to control all parts of the activation cascade, we focused only on the penultimate steps in the pathway: detecting the presence of pro-MMP-2 and active-MMP-2. Active MMP-2 was detected in medium only once (data not shown). The probable reason stems from the complexity in converting the secreted 72-kDa zymogen to the active 66-kDa form. This activation step requires the formation of a membrane-bound ternary complex involving membrane-bound MMPs (MT1-MMP) and TIMP-2. The lack of detection of TIMP-2 may in part explain why we did not see the 66-kDa active form of MMP-2. The presence of active MMP-2 may also be affected by the constitutive expression of the main MMP-2 inhibitor, TIMP-1; however, it is also intricately regulated. 
ERK signaling decreased in the presence of selenium, leading to the hypothesis that the MAPK pathway regulates secretion of MMP-2. This is consistent with previous studies showing that substances activating the ERK1/2 pathway in HTM cells can affect MMP-2 secretion. For example, platelet-derived growth factor and TPA both activate ERK1/2 and increase secretion of MMP-2. 25 The use of an MAPK pathway inhibitor eliminated the effects on MMP-2 secretion, providing further support for our hypothesis. 25 Similarly, TNFα has also been shown to increase phosphorylation of ERK1/2 in HTM cells affecting secretion of several MMPs, including MMP-1, -3, and -9. TNFα’s effects on MMPs were also eliminated in the presence of the MEK (MAPK pathway protein) inhibitor. 26 Of note, selenium’s effects on ERK1/2 in HUVECs appeared to occur after changes in MMP secretion were recorded. 3 Jiang et al., state that, in that case, inhibition of ERK signaling by MSeA was probably involved in downstream events, such as inhibition of cell proliferation, as opposed to directly regulating MMP secretion. It is therefore difficult to conclude at this time whether our results indicate that MSeA inhibits ERK1/2 activation that then inhibits MMP-2 secretion or whether MSeA inhibits MMP-2 secretion and the effects on ERK1/2 are precursors of a different set of events. As our data on ERK1/2 status came from only one time point, a temporal relationship is impossible to establish. Conversely, ERK phosphorylation is usually an early change, often peaking as soon as 30 minutes after treatment. It is therefore possible that ERK signaling occurs immediately after treatment and directly affects MMP secretion; however, more data are needed to establish this association. Future investigations may involve time-course experiments in HTM cells and the use of pharmacologic blockers for the ERK cascade. 
The morphologic changes in HTM cells after selenium treatment are novel. However, such changes are similar to those that occur after selenium treatment of newly formed HUVEC capillaries (in culture). In that case, selenium causes cell retraction and the formation of atypical spheroid bodies. 3 Although changes in HUVECs are visually distinct from the ones in HTM, it is possible that the mechanism is the same, perhaps resulting from targeted effects on cell–matrix complexes. Yoon et al. 5 reported that in HT1080 cells the inorganic selenium compound selenite causes a dose-dependent decrease in cell–matrix adhesions but does not affect cell–cell adhesions. These observations support our findings that treated HTM cells released from the plates but remained connected to one another. We hypothesize that these observed effects are a function of changes in integrin activation and that integrins may be a primary cellular target of selenium. There is evidence that integrins 33 interact with the MAPK cascade, making them plausible upstream signaling molecules for selenium’s effects on HTM cells. 
In conclusion, the evidence presented may explain mechanistically the observed increase in the risk of glaucoma associated with above-average selenium intake. However, further research is needed to make firm conclusions about causality. Future experiments may include (1) a closer examination of selenium’s effects on other MMPs that play a role in the complex regulation of ECM turnover; (2) the use of the anterior chamber perfusion system to assess the physiological relevance of selenium-mediated changes and their effect on IOP; (3) assessment of selenium’s effects on cell cycle arrest, apoptosis, and signaling mechanisms in TM; (4) examination of the effects of selenium on TM homeostasis when concurrently administered with 17β-estradiol, to understand the increased risk of selenium-induced glaucoma in women; and (5) investigation of the role of integrins in selenium signaling. Such experiments may enable us to elucidate causal relationships and identify doses of selenium that retain beneficial properties without inducing ocular toxicity. 
 
Figure 1.
 
MSeA’s effects on preconfluent and confluent HUVECs. HUVECs were serum starved for 24 hours and treated with various concentrations of MSeA or TPA (10 ng/mL) as a positive control. Medium was collected after 24 hours and analyzed. (A) Preconfluent HUVECs showed a dose-dependent decrease in secretion of MMP-2 and in active MMP-2. (B) MSeA had no effect on secretion of MMP-2 in confluent HUVECs. (C) Summary of the results presented in (A) and (B). n = 2.
Figure 1.
 
MSeA’s effects on preconfluent and confluent HUVECs. HUVECs were serum starved for 24 hours and treated with various concentrations of MSeA or TPA (10 ng/mL) as a positive control. Medium was collected after 24 hours and analyzed. (A) Preconfluent HUVECs showed a dose-dependent decrease in secretion of MMP-2 and in active MMP-2. (B) MSeA had no effect on secretion of MMP-2 in confluent HUVECs. (C) Summary of the results presented in (A) and (B). n = 2.
Figure 2.
 
Effects of selenium on MMP and TIMP secretion in HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. TPA (10 ng/mL) was used as the positive control; the negative control culture was serum starved but not selenium treated. HT1080 cells were treated with TPA and the medium used as an activity mobility control. Medium was collected and probed using zymography and Western blot analysis. (A) Many proteinases and inhibitors were examined but were detected inconsistently, including MMP-9, -3, and -1 (not shown) and TIMP-2. MMP-9 was detected only after induction by TPA. (B) Western blot (top) and zymogram (bottom) showing that MMP-2 exhibited a dose-dependent decrease in secretion. No active MMP-2 was detected (compared with lowest band in the HT1080 control lane). (C) Secretion of TIMP-1 also decreased inversely with MSeA treatment. (D) Summary of MMP-2 and TIMP-1 blot data (n = 5). *Significantly less than control (P < 0.05).
Figure 2.
 
Effects of selenium on MMP and TIMP secretion in HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. TPA (10 ng/mL) was used as the positive control; the negative control culture was serum starved but not selenium treated. HT1080 cells were treated with TPA and the medium used as an activity mobility control. Medium was collected and probed using zymography and Western blot analysis. (A) Many proteinases and inhibitors were examined but were detected inconsistently, including MMP-9, -3, and -1 (not shown) and TIMP-2. MMP-9 was detected only after induction by TPA. (B) Western blot (top) and zymogram (bottom) showing that MMP-2 exhibited a dose-dependent decrease in secretion. No active MMP-2 was detected (compared with lowest band in the HT1080 control lane). (C) Secretion of TIMP-1 also decreased inversely with MSeA treatment. (D) Summary of MMP-2 and TIMP-1 blot data (n = 5). *Significantly less than control (P < 0.05).
Figure 3.
 
Effects of high-dose, short-term selenium exposure on secretion of MMPs and TIMPs from HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. Medium was collected, concentrated, and analyzed after 6 hours of treatment. Control cultures were serum starved but not treated with selenium. (A) Representative Western blots (MMP-2 and TIMP-1) and zymogram (pro MMP-2) showing protein levels in HTM supernatants. (BD) Average protein levels. MMP-2: n = 4; TIMP-1: n = 3; Pro MMP-2: n = 4; Active MMP-2: not detected. *P < 0.05, when compared with control.
Figure 3.
 
Effects of high-dose, short-term selenium exposure on secretion of MMPs and TIMPs from HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. Medium was collected, concentrated, and analyzed after 6 hours of treatment. Control cultures were serum starved but not treated with selenium. (A) Representative Western blots (MMP-2 and TIMP-1) and zymogram (pro MMP-2) showing protein levels in HTM supernatants. (BD) Average protein levels. MMP-2: n = 4; TIMP-1: n = 3; Pro MMP-2: n = 4; Active MMP-2: not detected. *P < 0.05, when compared with control.
Figure 4.
 
Phosphorylation of ERK1/2 in HTM cells as a function of MSeA treatment. Confluent HTM cell cultures were serum starved for 24 hours and treated with MSeA for an additional 24 hours. Cell lysates were prepared and probed for total ERK protein and phosphorylated ERK. (A) Western blot analysis showing total ERK protein and phospho-ERK after MSeA treatment. The negative control was serum starved but not exposed to selenium. Cells were exposed to TPA (10 ng/mL) as a positive control for ERK phosphorylation. (B) Summary of ERK blot data. The quantity of total ERK protein did not change significantly as a function of MSeA treatment. By comparison, phosphorylated products decreased by as much as 60%. (*P < 0.05, §P < 0.01 versus negative control). Left: phospho-ERK1 and -2; right: ERK1 and -2. n = 3.
Figure 4.
 
Phosphorylation of ERK1/2 in HTM cells as a function of MSeA treatment. Confluent HTM cell cultures were serum starved for 24 hours and treated with MSeA for an additional 24 hours. Cell lysates were prepared and probed for total ERK protein and phosphorylated ERK. (A) Western blot analysis showing total ERK protein and phospho-ERK after MSeA treatment. The negative control was serum starved but not exposed to selenium. Cells were exposed to TPA (10 ng/mL) as a positive control for ERK phosphorylation. (B) Summary of ERK blot data. The quantity of total ERK protein did not change significantly as a function of MSeA treatment. By comparison, phosphorylated products decreased by as much as 60%. (*P < 0.05, §P < 0.01 versus negative control). Left: phospho-ERK1 and -2; right: ERK1 and -2. n = 3.
Figure 5.
 
Morphologic changes in MSeA-treated HTM 83 cells. HTM cells were serum starved and treated as described in Figure 3 . Cell associations changed in a dose-dependent manner. Starting at the 5-μM dose, the cells lifted off the plate and were tethered by secondary cell associations. (A) Micrographs of HTM cells immediately after cessation of 24-hour selenium treatment. (B) Two separate fields from the same well (5 μM selenium) 2 weeks after cessation of selenium treatment. Shown are results of one representative experiment of four in two different HTM cell lines. Scale bar, 50 μm.
Figure 5.
 
Morphologic changes in MSeA-treated HTM 83 cells. HTM cells were serum starved and treated as described in Figure 3 . Cell associations changed in a dose-dependent manner. Starting at the 5-μM dose, the cells lifted off the plate and were tethered by secondary cell associations. (A) Micrographs of HTM cells immediately after cessation of 24-hour selenium treatment. (B) Two separate fields from the same well (5 μM selenium) 2 weeks after cessation of selenium treatment. Shown are results of one representative experiment of four in two different HTM cell lines. Scale bar, 50 μm.
The authors thank Kathy Hardy, Emely Hoffman, Kristin Perkumas, Linda Meade-Tollin, and Clement Ip for contributions and assistance. 
Lu J. Apoptosis and angiogenesis in cancer prevention by selenium. Nutrition and Cancer Prevention. 2000;131–145. Kluwer Academic/Plenum Publishers New York.
Lu J, Jiang C. Antiangiogenic activity of selenium in cancer chemoprevention: metabolite-specific effects. Nutr Cancer. 2001;40:64–73. [CrossRef] [PubMed]
Jiang C, Ganther H, Lu J. Monomethyl selenium-specific inhibition of MMP-2 and VEGF expression: implications for angiogenic switch regulation. Mol Carcinog. 2000;29:236–250. [CrossRef] [PubMed]
Jiang C, Jiang W, Ip C, et al. Selenium-induced inhibition of angiogenesis in mammary cancer at chemopreventive levels of intake. Mol Carcinog. 1999;26:213–225. [CrossRef] [PubMed]
Yoon SO, Kim MM, Chung AS. Inhibitory effect of selenite on invasion of HT1080 tumor cells. J Biol Chem. 2001;276:20085–20092. [CrossRef] [PubMed]
Clark LC, Combs GF, Jr, Turnbull BW, et al. Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin: a randomized controlled trial: Nutritional Prevention of Cancer Study Group. JAMA. 1996;276:1957–1963. [CrossRef] [PubMed]
Lillico A, Jacobs B, Reid M, et al. Selenium Supplementation and Risk of Glaucoma in the NPC Trial Selenium and Cancer Projects Group. 2002; Arizona Cancer Center, University of Arizona Tucson, AZ.
Leske MC. The epidemiology of open-angle glaucoma: a review. Am J Epidemiol. 1983;118:166–191. [PubMed]
Quigley HA. Open-angle glaucoma. N Engl J Med. 1993;328:1097–1106. [CrossRef] [PubMed]
Epstein DL, Allingham R, Schuman J. Chandler and Grant’s Glaucoma. 1997;680. Williams & Wilkins Baltimore.
Hashimoto JM, Epstein DL. Influence of intraocular pressure on aqueous outflow facility in enucleated eyes of different mammals. Invest Ophthalmol Vis Sci. 1980;19:1483–1489. [PubMed]
Alvarado J, Murphy C, Juster R. Trabecular meshwork cellularity in primary open-angle glaucoma and nonglaucomatous normals. Ophthalmology. 1984;91:564–579. [CrossRef] [PubMed]
Yue BY. The extracellular matrix and its modulation in the trabecular meshwork. Surv Ophthalmol. 1996;40:379–390. [CrossRef] [PubMed]
Acott TS. Trabecular extracellular matrix regulation. 1992;125–127. Williams & Wilkins Baltimore.
Johnson M, Johnson DH, Kamm RD, et al. The filtration characteristics of the aqueous outflow system. Exp Eye Res. 1990;50:407–418. [CrossRef] [PubMed]
Ethier CR, Kamm RD, Palaszewski BA, et al. Calculations of flow resistance in the juxtacanalicular meshwork. Invest Ophthalmol Vis Sci. 1986;27:1741–1750. [PubMed]
Alexander JP, Bradley JM, Gabourel JD, et al. Expression of matrix metalloproteinases and inhibitor by human retinal pigment epithelium. Invest Ophthalmol Vis Sci. 1990;31:2520–2528. [PubMed]
Melamed S, Epstein DL. Alterations of aqueous humour outflow following argon laser trabeculoplasty in monkeys. Br J Ophthalmol. 1987;71:776–781. [CrossRef] [PubMed]
Van Buskirk EM, Pond V, Rosenquist RC, et al. Argon laser trabeculoplasty: studies of mechanism of action. Ophthalmology. 1984;91:1005–1010. [CrossRef] [PubMed]
Bradley JM, Vranka J, Colvis CM, et al. Effect of matrix metalloproteinases activity on outflow in perfused human organ culture. Invest Ophthalmol Vis Sci. 1998;39:2649–2658. [PubMed]
Alexander JP, Samples JR, Acott TS. Growth factor and cytokine modulation of trabecular meshwork matrix metalloproteinase and TIMP expression. Curr Eye Res. 1998;17:276–285. [CrossRef] [PubMed]
Alexander JP, Samples JR, Van Buskirk EM, et al. Expression of matrix metalloproteinases and inhibitor by human trabecular meshwork. Invest Ophthalmol Vis Sci. 1991;32:172–180. [PubMed]
Huang SH, Adamis AP, Wiederschain DG, et al. Matrix metalloproteinases and their inhibitors in aqueous humor. Exp Eye Res. 1996;62:481–490. [CrossRef] [PubMed]
Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases: evolution, structure and function. Biochim Biophys Acta. 2000;1477:267–283. [CrossRef] [PubMed]
Shearer T, Crosson CE. Activation of extracellular signal-regulated kinase in trabecular meshwork cells. Exp Eye Res. 2001;73:25–35. [CrossRef] [PubMed]
Alexander JP, Acott TS. Involvement of the Erk-MAP kinase pathway in TNFalpha regulation of trabecular matrix metalloproteinases and TIMPs. Invest Ophthalmol Vis Sci. 2003;44:164–169. [CrossRef] [PubMed]
Stamer WD, Seftor RE, Williams SK, et al. Isolation and culture of human trabecular meshwork cells by extracellular matrix digestion. Curr Eye Res. 1995;14:611–617. [CrossRef] [PubMed]
Alvarado JA, Wood I, Polansky JR. Human trabecular cells. II. Growth pattern and ultrastructural characteristics. Invest Ophthalmol Vis Sci. 1982;23:464–478. [PubMed]
McGahan MC, Grimes AM. Selenium concentration in ocular tissues and fluids. Ophthalmic Res. 1991;23:45–50. [CrossRef] [PubMed]
Ganther H. Selenium metabolism and mechanisms of cancer prevention. Nutrition and Cancer Prevention. 2000;119–130. Kluwer Academic/Plenum Publishers New York.
Lu J, Jiang C, Kaeck M, et al. Dissociation of the genotoxic and growth inhibitory effects of selenium. Biochem Pharmacol. 1995;50:213–219. [CrossRef] [PubMed]
Ip C, Thompson HJ, Zhu Z, et al. In vitro and in vivo studies of methylseleninic acid: evidence that a monomethylated selenium metabolite is critical for cancer chemoprevention. Cancer Res. 2000;60:2882–2886. [PubMed]
Alahari SK, Reddig PJ, Juliano RL. Biological aspects of signal transduction by cell adhesion receptors. Int Rev Cytol. 2002;220:145–184. [PubMed]
Figure 1.
 
MSeA’s effects on preconfluent and confluent HUVECs. HUVECs were serum starved for 24 hours and treated with various concentrations of MSeA or TPA (10 ng/mL) as a positive control. Medium was collected after 24 hours and analyzed. (A) Preconfluent HUVECs showed a dose-dependent decrease in secretion of MMP-2 and in active MMP-2. (B) MSeA had no effect on secretion of MMP-2 in confluent HUVECs. (C) Summary of the results presented in (A) and (B). n = 2.
Figure 1.
 
MSeA’s effects on preconfluent and confluent HUVECs. HUVECs were serum starved for 24 hours and treated with various concentrations of MSeA or TPA (10 ng/mL) as a positive control. Medium was collected after 24 hours and analyzed. (A) Preconfluent HUVECs showed a dose-dependent decrease in secretion of MMP-2 and in active MMP-2. (B) MSeA had no effect on secretion of MMP-2 in confluent HUVECs. (C) Summary of the results presented in (A) and (B). n = 2.
Figure 2.
 
Effects of selenium on MMP and TIMP secretion in HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. TPA (10 ng/mL) was used as the positive control; the negative control culture was serum starved but not selenium treated. HT1080 cells were treated with TPA and the medium used as an activity mobility control. Medium was collected and probed using zymography and Western blot analysis. (A) Many proteinases and inhibitors were examined but were detected inconsistently, including MMP-9, -3, and -1 (not shown) and TIMP-2. MMP-9 was detected only after induction by TPA. (B) Western blot (top) and zymogram (bottom) showing that MMP-2 exhibited a dose-dependent decrease in secretion. No active MMP-2 was detected (compared with lowest band in the HT1080 control lane). (C) Secretion of TIMP-1 also decreased inversely with MSeA treatment. (D) Summary of MMP-2 and TIMP-1 blot data (n = 5). *Significantly less than control (P < 0.05).
Figure 2.
 
Effects of selenium on MMP and TIMP secretion in HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. TPA (10 ng/mL) was used as the positive control; the negative control culture was serum starved but not selenium treated. HT1080 cells were treated with TPA and the medium used as an activity mobility control. Medium was collected and probed using zymography and Western blot analysis. (A) Many proteinases and inhibitors were examined but were detected inconsistently, including MMP-9, -3, and -1 (not shown) and TIMP-2. MMP-9 was detected only after induction by TPA. (B) Western blot (top) and zymogram (bottom) showing that MMP-2 exhibited a dose-dependent decrease in secretion. No active MMP-2 was detected (compared with lowest band in the HT1080 control lane). (C) Secretion of TIMP-1 also decreased inversely with MSeA treatment. (D) Summary of MMP-2 and TIMP-1 blot data (n = 5). *Significantly less than control (P < 0.05).
Figure 3.
 
Effects of high-dose, short-term selenium exposure on secretion of MMPs and TIMPs from HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. Medium was collected, concentrated, and analyzed after 6 hours of treatment. Control cultures were serum starved but not treated with selenium. (A) Representative Western blots (MMP-2 and TIMP-1) and zymogram (pro MMP-2) showing protein levels in HTM supernatants. (BD) Average protein levels. MMP-2: n = 4; TIMP-1: n = 3; Pro MMP-2: n = 4; Active MMP-2: not detected. *P < 0.05, when compared with control.
Figure 3.
 
Effects of high-dose, short-term selenium exposure on secretion of MMPs and TIMPs from HTM cells. Confluent (>1 week) HTM cells from three cell lines were serum starved for 24 hours and treated with MSeA as shown. Medium was collected, concentrated, and analyzed after 6 hours of treatment. Control cultures were serum starved but not treated with selenium. (A) Representative Western blots (MMP-2 and TIMP-1) and zymogram (pro MMP-2) showing protein levels in HTM supernatants. (BD) Average protein levels. MMP-2: n = 4; TIMP-1: n = 3; Pro MMP-2: n = 4; Active MMP-2: not detected. *P < 0.05, when compared with control.
Figure 4.
 
Phosphorylation of ERK1/2 in HTM cells as a function of MSeA treatment. Confluent HTM cell cultures were serum starved for 24 hours and treated with MSeA for an additional 24 hours. Cell lysates were prepared and probed for total ERK protein and phosphorylated ERK. (A) Western blot analysis showing total ERK protein and phospho-ERK after MSeA treatment. The negative control was serum starved but not exposed to selenium. Cells were exposed to TPA (10 ng/mL) as a positive control for ERK phosphorylation. (B) Summary of ERK blot data. The quantity of total ERK protein did not change significantly as a function of MSeA treatment. By comparison, phosphorylated products decreased by as much as 60%. (*P < 0.05, §P < 0.01 versus negative control). Left: phospho-ERK1 and -2; right: ERK1 and -2. n = 3.
Figure 4.
 
Phosphorylation of ERK1/2 in HTM cells as a function of MSeA treatment. Confluent HTM cell cultures were serum starved for 24 hours and treated with MSeA for an additional 24 hours. Cell lysates were prepared and probed for total ERK protein and phosphorylated ERK. (A) Western blot analysis showing total ERK protein and phospho-ERK after MSeA treatment. The negative control was serum starved but not exposed to selenium. Cells were exposed to TPA (10 ng/mL) as a positive control for ERK phosphorylation. (B) Summary of ERK blot data. The quantity of total ERK protein did not change significantly as a function of MSeA treatment. By comparison, phosphorylated products decreased by as much as 60%. (*P < 0.05, §P < 0.01 versus negative control). Left: phospho-ERK1 and -2; right: ERK1 and -2. n = 3.
Figure 5.
 
Morphologic changes in MSeA-treated HTM 83 cells. HTM cells were serum starved and treated as described in Figure 3 . Cell associations changed in a dose-dependent manner. Starting at the 5-μM dose, the cells lifted off the plate and were tethered by secondary cell associations. (A) Micrographs of HTM cells immediately after cessation of 24-hour selenium treatment. (B) Two separate fields from the same well (5 μM selenium) 2 weeks after cessation of selenium treatment. Shown are results of one representative experiment of four in two different HTM cell lines. Scale bar, 50 μm.
Figure 5.
 
Morphologic changes in MSeA-treated HTM 83 cells. HTM cells were serum starved and treated as described in Figure 3 . Cell associations changed in a dose-dependent manner. Starting at the 5-μM dose, the cells lifted off the plate and were tethered by secondary cell associations. (A) Micrographs of HTM cells immediately after cessation of 24-hour selenium treatment. (B) Two separate fields from the same well (5 μM selenium) 2 weeks after cessation of selenium treatment. Shown are results of one representative experiment of four in two different HTM cell lines. Scale bar, 50 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×