Signaling Pathways Used in Trabecular Matrix Metalloproteinase Response to Mechanical Stretch

Abstract

purpose. Trabecular meshwork (TM) matrix metalloproteinase (MMP), and tissue inhibitor (TIMP) changes in response to mechanical stretching appear to be central to intraocular pressure (IOP) homeostasis. Studies were conducted to define the signal transduction pathway responsible for the increases in MMP-2 and -14 that occur in response to mechanical stretching of TM cells.

methods. Porcine TM cells were subjected to mechanical stretching, and changes in MMP-2 and -14 levels were determined by gelatin zymography and Western immunoblot analysis. Effects of signal transduction pathway inhibitors on MMP levels were analyzed. Phosphospecific antibodies were used to identify phosphorylation changes in select pathway intermediates. In silico secondary structure analysis was conducted on the 5′ untranslated regions (UTRs) of MMP-2 and -14 mRNAs.

results. The increases in MMP-2 and -14 that occur 24 hours after sustained mechanical stretching of TM cells were blocked by rapamycin. Wortmannin blocked the MMP-2 but not the MMP-14 increase. Protein kinase B (PKB) phosphorylation on S473 and T308 was increased significantly by stretching. Rapamycin-sensitive phosphorylation of T389 in p70/p85 S6 kinase was also increased. The phosphorylations of the translation initiation factor eIF-4E on S209 and of its inhibitory binding protein 4E-BP1 on T70 were both increased by stretch. The calculated free energies of secondary structures of the 5′ UTRs of the mRNAs for MMP-2 and -14 were negative and relatively large. MMP-2 also had pyrimidine tracts in the extreme 5′ region of its UTR.

conclusions. The increases in TM MMP-2 and -14 protein levels in response to mechanical stretching appear to be transduced at least in part by mTOR, the mammalian target of rapamycin (mTOR). The wortmannin sensitivity implicates phosphoinositide 3-kinase as a modulator of the MMP-2 but not the MMP-14 increase. Integrin-linked kinase (ILK), phosphoinositide-dependent kinase (PDK-1), and PKB are implicated in the MMP-2 increase. Translational initiation involving eIF-4E and its inhibitory binding protein 4E-BP1 appear to be involved in both the MMP-2 and -14 increases with stretching and are normally regulated by mTOR. The high degree of secondary structure in the 5′ UTRs of these transcripts is typically an indicator of genes specifically sensitive to regulation through this pathway. P70/p85 S6 kinase is probably involved downstream from mTOR and PKB in regulating translation of MMP-2, which has pyrimidine tracts in its 5′ UTR. Manipulation of these transduction pathways may provide new approaches to therapeutic IOP regulation.

The mechanism that provides normal IOP homeostasis is poorly understood. A relatively effective homeostatic mechanism must exist, because only approximately 3% to 5% of people exhibit pathologic elevations in IOP with subsequent optic nerve damage, even at advanced ages. ^{1 2} We have hypothesized that trabecular meshwork (TM) cells can adjust the outflow resistance over the time scale of hours to days by modulating TM extracellular matrix (ECM) turnover and subsequent biosynthetic replacement. ^{3 4 5 6} Manipulation of the activity of a family of TM ECM turnover enzymes, the matrix metalloproteinases (MMPs), reversibly modulates outflow facility. ⁷ Inhibition of the endogenous ECM turnover by these MMPs increases the outflow resistance, thus ongoing ECM turnover must be necessary for normal IOP maintenance. ⁷ In addition, laser trabeculoplasty, a common treatment for glaucoma, appears to owe its efficacy to producing relatively sustained TM MMP elevations, particularly within the juxtacanalicular region of the meshwork. ^{8 9 10}  

Much of the aqueous humor outflow resistance appears to reside within the deepest portion of the TM, ¹¹ and thus the resistance is functionally distributed across Schlemm’s canal. Any changes in the outflow resistance sufficient to change IOP thus causes changes in the degree of stretching and distortion of the juxtacanalicular TM. A likely sensing mechanism for TM cells, allowing them to determine whether to increase or decrease the outflow resistance to maintain appropriate IOP homeostasis, would be the sensing of mechanical stretching. This IOP-sensing could be mediated and transduced through integrin–ECM interactions. ^{12 13} We and others have found that TM cells can sense mechanical stretching and produce several distinctive responses. ^{14 15 16 17 18 19 20 21 22} In support of our overall hypothesis of IOP homeostasis, TM cells respond to pressure elevations or to mechanical stretching by increasing MMP-2 and -14 activity or levels, while dramatically reducing levels of their primary inhibitor, tissue inhibitor of matrix metalloproteinase (TIMP)-2. ¹⁴ Increased TM MMP activity results in reduced outflow resistance ⁷ and thus should restore the IOP to lower normal levels. Thus, the components of a self-contained TM IOP homeostasis mechanism are present and functional. Others have found somewhat similar evidence for IOP homeostasis. ²³  

Detailed studies of the signal transduction pathways involved in MMP responses to mechanical stretching in other cell types are limited, and even less information is available defining the TM responses. Some other TM responses and signal transduction triggered by mechanical stretching have been reported. ^{15 19} Studies of TM MMP and TIMP regulation and signal transduction have been focused primarily on growth factor and cytokine responses, which appear to use a different subset of this family of proteinases. ^{8 9 10 24 25 26 27} These MMP and TIMP changes appeared to be regulated primarily at the transcriptional level. However, in separate studies (Zhu et al., manuscript in preparation), we found that the MMP-2, MMP-14, and TIMP-2 protein changes, which occur in response to mechanical stretching of TM cells, are not associated with changes in their mRNA levels. Thus, studies were undertaken to identify signal transduction pathways that are important in transducing these TM MMP changes in response to mechanical stretching. Because the respective mRNA levels do not change, special attention was paid to potential translational regulatory mechanisms. 

Materials and Methods

Porcine eyes were obtained from Carlton Packing (Carlton, OR) within 2 to 5 hours of death; leupeptin, aprotinin, pepstatin, proteinase inhibitor cocktail, gelatin, phenylmethylsulfonyl fluoride (PMSF), and horseradish peroxidase-conjugated secondary antibodies were from Sigma-Aldrich (St. Louis, MO); GF 109203X (bisindolylmaleimide I or Gö 6850), wortmannin, and rapamycin were from CalBiochem (San Diego, CA); a dsDNA quantitation reagent (PicoGreen) was from Molecular Probes (Eugene, OR); MMP-14 antibodies were from Triple Point Biologics (Portland, OR) or Oncogene (La Jolla, CA); phospho T308 and S473 protein kinase B (PKB) antibodies; eIF-4E and phospho S209 eIF-4E antibodies; phospho T389 p70/p85 S6 kinase antibody; and 4E-BP1 and phospho T37, S65, and T70 4E-BP1 antibodies were from Cell Signaling Technologies (Beverly, MA); Dulbecco’s modified Eagle’s medium (DMEM), antibiotics, and antimycotics were from Invitrogen-Gibco (Grand Island, NY); fetal bovine serum was from HyClone (Logan, UT); Super Signal chemiluminescence detection kits were from Pierce (Rockford, IL); and cell culture membrane inserts (Falcon PET track-etched 3-μm pore membranes in six-well format) were from BD Biosciences (Franklin Lakes, NJ). 

TM Cell Culture, Mechanical Stretching, Treatments, and Extractions

Porcine TM cells were cultured as previously described. ^{5 28 29} By passage 3 to 5, cells were plated at a density of approximately 90% confluence onto cell culture insert membranes in six-well culture plates. ¹⁴ After 3 to 5 days, serum-free medium was added to the cells for 24 hours, before and during the stretching experiments. To apply mechanical stretch and distortion, a glass bead was placed beneath the insert in the center of the membrane, and weight was applied to the lid of the plate to force the lip of the insert down onto the upper lip of plate. ¹⁴ To avoid breaking the membranes, the weight was applied slowly, over several minutes. This produced a defined upward deflection of the center of the membrane, which increased the surface area by an estimated 6%. The membranes are 25 mm in diameter and the displacement 4.5 mm. The distorted membrane approximates the shape of a cone when stretched, in that the bead contacts only a small central portion of the membrane. 

More than 20 different porcine cell lines, each pooled from 20 to 40 eyes, were studied. For the various studies shown, control refers to the samples where the membranes were not stretched. However, in all the inhibitor studies, parallel samples were evaluated in which vehicle alone or inhibitor was added without mechanical stretching. These were not different from the nonstretched controls, and thus these data are not shown. DNA analysis (PicoGreen; Molecular Probes) to estimate cell density on parallel membranes was conducted for some studies as directed by the manufacturers. Because the differences were always less than 10%, the analysis was not conducted for all studies. At the indicated times after initiation of stretching, media were collected and stored in aliquots frozen at −20°C until use. Cellular proteins were extracted from membranes that had been immediately rinsed with ice cold phosphate-buffered saline, with a modified radioimmunoprecipitation assay (RIPA) buffer ^{24 30 31} (2 mM EDTA, 2 mM EGTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 50 mM NaF, 2 mM dithiothreitol [DTT], 1 mM sodium orthovanadate, 10 mM NaP₄O₇, 1 mM PMSF, 20 μg/mL leupeptin, 20 μg/mL aprotinin, 20 μg/mL pepstatin, and 50 mM Tris [pH 7.5]) on ice. In some experiments, MMP-14 was extracted on ice with 1% Triton X-114, proteinase inhibitor cocktail, 10 mM Tris (pH 7.4), and 150 mM NaCl, and centrifuged at 13,000g at 4°C. Extract was then overlaid on a similar solution containing 6% sucrose, warmed to 37°C, and centrifuged for 3 minutes at 500g. ³² The supernatant was then subjected to a second Triton extraction. The two lower Triton layers were then pooled and analyzed for MMP-14, as described in the next section. 

Zymograms and Western Immunoblots

Western immunoblots, transferred electrophoretically from standard SDS-PAGE gels to nitrocellulose or PVDF membranes, were probed with the indicated primary antibodies, and detection was performed with the appropriate secondary antibodies with conjugated horseradish peroxidase and chemiluminescence according to the manufacturer’s instructions. To verify uniform cell numbers, extraction efficiency or sampling, gel loading, and transfer efficiency, blots were stained for total protein (Ponceau stain; Sigma-Aldrich) after blotting but before adding primary antibody or blocker solutions. Any blots that did not exhibit uniformity of protein banding were not used. Gelatin was used as the substrate to detect MMP-2 (gelatinase A) in the zymograms (substrate SDS-PAGE gels). ^{5 14 24 33} Gels, stained blots, or autoradiographs were scanned and relative band density analyzed ³⁴ using densitometry computer programs (BioImage, Ann Arbor, MI, or UVP, Upland, CA). Student’s t-test or Mann-Whitney rank sum analysis was used to determine significance when comparing treatment results. All experiments presented were repeated at least three times, and representative gels were selected for presentation. 

Secondary Structure Analysis

The 5′ UTRs of human MMP-2 (AJ298926; all accession numbers are GenBank; http://www.ncbi.nlm.nih.gov/GenBank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD), MMP-14 (AF158733), MNK-1 (XM_003720), ornithine decarboxylase 1 (XM_002679), α1-actin (AF182035) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; NM_002046) were subjected to in silico secondary structural analysis using mFold (ver. 3.1, created by Michael Zuker, Rensselaer Polytechnic Institute, Troy, NY, and available at www.bioinfo.math.rpi.edu). ^{35 36} Temperature was set at 37°C and salt was 1 M NaCl with no divalent ions. 

Results

Effects of Stretching and Inhibitors on TM MMP-2 and -14 Levels

TM cells respond to sustained mechanical stretching by increasing medium MMP-2 levels and extractable MMP-14 levels within 24 to 48 hours after stretching. ¹⁴ The effects of several signal transduction pathway inhibitors on these changes were evaluated. When the phosphoinositol-3-kinase (PI3 kinase) inhibitor wortmannin or the mammalian target of rapamycin (mTOR) inhibitor rapamycin was added 1 hour before and during the stretching, the MMP-2 increases were blocked (Figs. 1A 1B) . However, the protein kinase C inhibitor bisindolylmaleimide I had no effect on this increase even at these relatively high doses (Fig. 1C) . Time courses and dose–response curves (not shown) were used to define the optimum TM responses of these inhibitors and when added without stretching, they did not change MMP-2 levels (data not shown). The t-test significance levels shown are for the comparison of the stretched sample to the nonstretched control (Figs. 1A 1B) and for the stretched sample compared to the stretched sample with the respective inhibitor (Fig. 1) . 

In similar experiments, evaluating the effects of wortmannin and rapamycin on TM MMP-14 levels after mechanical stretching, only rapamycin was effective in blocking the action (Fig. 2) . Mechanical stretching caused a 30% to 70% increase in MMP-14 levels, rapamycin completely blocked the increases and wortmannin did not significantly reduce the increases. Some slight effects of dimethyl sulfoxide (DMSO), even when diluted 1:10,000, were observed on MMP-14 levels. Several doses and times were also evaluated in these and other experiments (data not shown). 

Effects of Mechanical Stretching on PKB Phosphorylation

Because PKB is often centrally involved in signal transduction downstream from PI3 kinase, we evaluated the phosphorylation state of this enzyme. Integrin-linked kinase (ILK) and phosphoinositide-dependent kinase (PDK-1) normally phosphorylate PKB on S473 and T308, respectively; thus phospho-specific antibodies for these sites in PKB were used. Western immunoblots showed that the phosphorylation state of both sites was elevated significantly after 24 hours of sustained mechanical stretching of TM cells (Fig. 3) . 

Effects of Mechanical Stretching on p70 S6 Kinase Phosphorylation

Several rapamycin-sensitive phosphorylated proteins with important roles in protein translational regulation have been identified. One of these is p70 S6 kinase. The phosphorylation of T389 in p70 S6 kinase was dramatically elevated at 24 hours, and rapamycin was very effective in blocking this increase (Fig. 4) . Two isoforms of this kinase, the 70- and the 85-kDa forms as labeled in the figure, were similarly affected. The bands were scanned separately and together, but are presented together because the changes are essentially parallel. 

Effects of Mechanical Stretching on Translational Initiation Factor Phosphorylation

Another group of common rapamycin-sensitive downstream targets are the translational initiation factors, ^{37 38 39 40} and because we had detected MMP-2 and -14 protein level changes but no mRNA level changes (Zhu et al., manuscript in review), we evaluated phosphorylation of eIF-4E and its inhibitory binding protein 4E-BP1. The phosphorylation level of eIF-4E at S209 was dramatically elevated at 24 hours after initiation of mechanical stretching (Fig. 5) . The doublet at approximately 27 kDa was scanned together, because the changes were parallel. 

The eIF-4E inhibitory binding protein 4E-BP1 is also often phosphorylated in the process of upregulating translational initiation. Several phosphorylation sites have been identified on 4E-BP1 and several different protein kinases are reported to phosphorylate it. Mechanical stretching increased the phosphorylation level of T70 in this protein in the TM at several time points (Fig. 6) . There was a rapid transient initial T70 phosphorylation increase and then a more sustained increase lasting at least 24 hours. Phosphorylation of 4E-BP1 on S65 and T37 was detectable at all these times, but not significantly increased with stretch (data not shown). The protein normally migrates at approximately 17 kDa, but the multiple phosphorylations shift its apparent molecular weight to approximately 21 kDa. We observe a doublet with or without stretching at approximately 20 kDa, which is the approximate size of the hyperphosphorylated forms. ⁴¹  

Predicted MMP-2 and -14 mRNA 5′ UTR Secondary Structure

The highly regulated translational initiation complex involving eIF-4E also includes an mRNA helicase (eIF-4A) and selectively affects translation of transcripts with high degrees of secondary structure in their 5′ UTRs. ^{37 38 42 43 44} Thus, we conducted in silico analysis of the MMP-2 and -14 transcripts to determine the degree of structure in their 5′ UTR. For comparison, two genes with known low amounts of secondary structure, GAPDH and α1-actin, and two genes with known high degrees of structure, MNK-1 and ODC, were compared with MMP-2 and -14, using the third-generation structure prediction program, mFold. ^{35 36} The most stable predicted structure and the calculated free energy for that structure for each of these genes are shown (Fig. 7) . GAPDH and α1-actin have small negative free energies (ΔG), −13.87 and −30.8 kcal/mole respectively, whereas MNK-1 and ODC have very large negative free energies, −104.7 and −151.07 kcal/mole, respectively. MMP-2 and -14 also exhibit extensive secondary structure (Fig. 7) and very large negative free energies: −131.86 and −111.93 kcal/mole, respectively. In addition, GAPDH and α1-actin have only, respectively, three and one different secondary structure variations that are sufficiently stable to reach the default stability cutoff. However, MNK-1, ODC, and MMP-2 and -14 have, respectively, 6, 10, 19, and 12 different folding structures that reach this stability cutoff level. 

Another characteristic of the 5′ UTRs of selectively translated transcripts is the presence of tracts of pyrimidines (5′ TOP) at the extreme 5′ end of the transcript. ^{37 44 45} MMP-2 has two pyrimidine tracts 5 bases long with one purine between them and one 7 bases long slightly downstream from these (Fig. 8) . MMP-14 has only one pyrimidine tract, 8 bases long, and 84 bases into the transcript (Fig. 8) . MMP-2 also has one short open reading frame (ORF) with good context for translation initiation in the middle of the 5′ UTR (Fig. 8) , which is also important in selective translational regulation. ^{37 44 46}  

Hypothetical Pathway for TM IOP Homeostasis Regulation

Although only portions of the total pathway(s) have been demonstrated at this point, based on our data and observations from the literature on other tissues and cell types, a hypothetical transduction pathway is proposed (Fig. 9) . In our working hypothesis, TM cells sense mechanical stretching and distortion produced by the effects of changes in IOP on the outflow resistance, ¹⁴ which is thought to reside in the juxtacanalicular region of the TM. ¹¹ Stretching produces tension and relative movement of juxtacanalicular ECM macromolecules, which are directly or indirectly attached to TM cells through specific integrins. ^{12 13 47 48 49} The first transduction step probably involves adaptor proteins and kinases associated with the cytoplasmic tails of the integrins, where they interact with the cytoskeleton and signaling complexes at focal complexes or focal contacts. ^{48 49 50 51} This triggers activation of PI 3 kinase, which produces phosphatidylinositol-3,4,5-trisphosphate (PI3). ^{52 53 54} PI 3 kinase is specifically inhibited by wortmannin. PI3 facilitates activation of PKB by phosphorylation on S473 by ILK and on T308 by PDK1. ^{54 55 56 57} PI3 can also directly and/or indirectly activate mTOR, but mTOR can also be activated by several other wortmannin-insensitive mechanisms. ^{37 38 41 58 59} The p70 S6 kinase is phosphorylated on T389 by mTOR and on T412 by PKB and other kinases. ^{45 57 58 60} Activated p70 S6 kinase increases protein translation of selective genes, particularly those with tracts of pyrimidines (TOPs) in their extreme 5′ UTR, by mechanisms that are only partially understood. ^{38 44 60 61} The transcriptional initiation factor eIF-4E is thought to be phosphorylated primarily by MNK-1, which docks on the large adaptor/scaffold protein eIF-4G. The eIF-4E inhibitory binding protein 4E-BP1 is phosphorylated by mTOR and several other kinases. ⁴¹ Hyperphosphorylated 4E-BP1 releases eIF-4E, which binds directly to the 5′ cap of the mRNA to be translated and recruits the mRNA to a heterotrimeric complex with the large scaffold protein eIF-4G and the RNA helicase eIF-4A. This helicase is particularly important in regulating translation of genes with high levels of secondary structure in their 5′ UTRs. ^{37 38 43 44 61} Translational initiation and elongation follows with inclusion of several other factors, many of which are phosphorylated by mTOR and other kinases. 

Discussion

Both MMP-2 and -14 protein levels are elevated at 24 to 48 hours after mechanical stretch. ¹⁴ Rapamycin, which is a very specific inhibitor of mTOR, blocks increases of both of these proteins. Thus, our data indicate that mTOR is central to both of these protein level increases. These increases also appear to be due to translational regulation, because mRNA levels of neither protein show an increase with stretching (Zhu et al., manuscript in preparation). The 5′ UTRs of both of these mRNAs appear to have considerable secondary structure with ΔG = −131.86 and −111.93 kcal/mol, respectively, for MMP-2 and -14. Thus, these genes are very strong candidates for selective regulation by eIF-4E activity modulation and selective translational initiation. ^{37 43 44 61} The increased phosphorylation eIF-4E on S209 and of 4E-BP1 on T70 with stretching are apparently sufficient to cause release and activation of eIF-4E allowing mRNA cap binding and initiation of translation of MMP-2 and -14. T37 and S65 of 4E-BP1 appear to be phosphorylated in TM cells without stretching (data not shown). Based on the migration of 4E-BP1 at approximately 20 kDa, T46 is also likely to be phosphorylated before stretching. A hierarchical phosphorylation mechanism compatible with our observations is reported for the activation of this protein. ⁴¹  

Because wortmannin blocks increases in MMP-2 but not in MMP-14 levels, an additional transduction pathway appears to be involved in the MMP-2 increases. Wortmannin specifically inhibits PI3 kinase, and PI3 is critical for the activation of PKB by ILK and PDK1. ^{52 55 57} Although PKB can regulate mTOR, it is also essential in the phosphorylation and activation of p70 S6 kinase. Phosphorylation of p70 S6 kinase (and p85) in response to stretch on T389 is rapamycin sensitive in the TM. However, p70 S6 kinase phosphorylation on T412 appears to be due to activation of PKB. ⁵⁸ Because MMP-2 has 5′ TOP pyrimidine tracts at the extreme 5′ end of its UTR, translation of its mRNA is very likely to be selectively regulated by p70 S6 kinase. ^{44 45 60 61} The pyrimidine tract found in the 5′ UTR of MMP-14 is not near the transcription initiation site and is thus less likely to be important in its translational regulation. Although the rapamycin sensitivity of the MMP-2 increase could be due strictly to the p70 S6 kinase, which is probably wortmannin and rapamycin sensitive, the high degree of secondary structure in MMP-2s 5′ UTR argues that eIF-4E is also involved in its translational regulation. 

Extrapolation of these results to the human eye requires some caution, because this mechanical stretch model approximates only the forces that apply to a TM cell or ECM molecule in vivo. The increase in surface area of the membranes when stretched seems reasonable, compared to micrographs of eye bank eyes with no venous backpressure exposed to near or slightly above physiologic perfusion pressures ⁶² (e.g., 5 mm Hg in Figs. 5 7 8 9 , and 15 mm in Fig. 11 of Ref. ⁶² ). In these micrographs, the deflections of the juxtacanalicular region into Schlemm’s canal appear to be roughly similar in relative magnitude to those obtained when the bead is used to stretch the membrane in our model system. ⁶² However, the extent to which our model system reflects the in vivo distortions and movements of ECM macromolecules relative to TM cells as induced by changes in IOP or outflow resistance has not been established. 

These studies provide possible new sites for drug-based manipulation of the trabecular resistance to aqueous outflow and may allow correction of glaucomatous outflow obstructions. They also extend our understanding of TM biology, outflow regulation and the hypothetical homeostatic IOP regulation mechanism. The involvement of translational regulation in TM IOP homeostasis was surprising and is in sharp contrast to the transcriptional regulation that mediates growth factor–, cytokine-, and laser trabeculoplasty–induced changes in TM MMPs. ^{8 9 10 24 25} The conclusions of this study thus support our working hypothesis that changes in IOP cause mechanical stretching, which triggers TM MMP increases through translational mechanisms and increases ECM turnover, providing homeostatic restoration of normal outflow facility and subsequently adjusting the outflow resistance to correct the IOP. 

 Supported by National Institutes of Health Grants EY03279, EY08247, and EY10572 and by grants from the Glaucoma Research Foundation (San Francisco, CA), Research to Prevent Blindness (New York, NY), Alcon Labs (Fort Worth, TX), Kettering Family Foundation (Denver, CO), Culpeper Biomedical Pilot Initiative (New York, NY), Northwest Health Foundation (Portland, OR), American Federation for Aging Research/Pfizer (New York, NY), and Collins Medical Trust (Portland, OR).

 Submitted for publication March 3, 2003; revised July 29, 2003; accepted August 18, 2003.

 Disclosure: J.M.B. Bradley, Alcon Labs (F); M.J. Kelley, Alcon Labs (F); A. Rose, Alcon Labs (F); T.S. Acott, Alcon Labs (F)

 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.

 

Effects of inhibitors on TM MMP-2 level changes with mechanical stretching. Porcine TM cells on membrane inserts were not stretched (Con) or were subjected to mechanical stretching (Str) for 24 hours in the presence of the indicated signal-transduction pathway inhibitors. Media were analyzed for MMP-2 activity by gelatin zymography, and activity is presented as relative band density from densitometric scans of the gels. (A) Wortmannin (Wort) was added at 10 ng/mL 1 hour before stretching was initiated. Mean values from analysis of scans of the bands with standard deviations for the indicated n are shown. Significance levels are for unpaired t-tests comparing Con versus Str and Str versus Str+Wort, as indicated. Results from two separate experiments, one with triplicate samples and one with duplicate samples, are shown. (B) Rapamycin (Rap) was added at 5 ng/mL 1 hour before stretch was initiated. Significance is for unpaired t-tests comparing Con versus Str and Str versus Str+Rap, as indicated. Results from three separate experiments with duplicate samples are shown. (C) Bisindolylmaleimide I (Bis) was added at 100 and 200 nM, as indicated, 1 hour before stretch was initiated. Significance values are shown from unpaired t-tests comparing Str with Str+Bis at either concentration. Results from two separate experiments with duplicate samples, are shown. After the statistical analysis was conducted, the relative band densities of these three sets of studies were normalized to give a control value of 1.0.

Figure 2.

 

Effects of inhibitors on TM MMP-14 level changes with mechanical stretching. Porcine TM cells were not stretched (Con) or were stretched (Str) for 24 hours. Wortmannin (Wort) at 10 ng/mL or rapamycin (Rap) at 5 ng/mL was added as indicated 1 hour before stretching was initiated. MMP-14 was extracted from the cells and analyzed by Western immunoblot. The results of densitometric scans of blots are shown, with the mean and the SD, sample number (n), and the significance, as determined by paired t-tests. Two separate experiments with two or three replicate samples are shown. The major band of activated MMP-14, which migrates at approximately 58 kDa, was scanned.

Figure 3.

 

Effects of mechanical stretching on TM PKB phosphorylation. Porcine TM cells were not stretched (Con) or were stretched (Str) for 24 hours, and the phosphorylation of PKB on residues S473 (A) or T308 (B) determined by probing Western immunoblots of cell extracts with the respective phospho-specific antibody. Means of relative band densities with standard deviations are plotted, and typical immunoblots are shown. The band migrates at approximately 60 kDa. Significance values are from unpaired t-tests with the n as indicated. In each case, results from two separate experiments with triplicate (A) or duplicate (B) samples are shown.

Figure 4.

 

Effects of mechanical stretching and rapamycin on phosphorylation of p70 and p85 S6 kinase. The degree of phosphorylation of p70 S6 kinase on T389 after 24 hours of mechanical stretching was determined by Western immunoblot with a phosphospecific antibody. As indicated, 5 ng/mL of rapamycin (Rap) was added 1 hour before and during stretching (Str). The phosphorylation state was determined for nonstretched control (Con), stretched (Str) samples, and stretched plus rapamycin (Str+Rap)–treated samples. The two bands, p70 and p85, migrated at approximately 70 and 85 kDa, respectively. Both bands were included in the mean and SD shown for one of two completely separate experiments, each with triplicate samples.

Figure 5.

 

Effects of mechanical stretching on phosphorylation of the eukaryotic translational initiation factor eIF-4E. The degree of phosphorylation of eIF-4E on S209 after 24 hours of mechanical stretching is presented, as is a typical immunoblot. The mean and SD is shown for one of two separate experiments with triplicate samples. Both bands of the doublet migrating at approximately 26 kDa were scanned together.

Figure 6.

 

Effects of mechanical stretching on phosphorylation of the translational initiation factor binding protein 4E-BP1. The degree of phosphorylation of 4E-BP1 on T70 after 5 or 15 minutes or 1, 6, or 24 hours of mechanical stretching is presented, as indicated, expressed as the mean ± SD. The significance from paired t-tests comparing nonstretched control (Con) with stretched (Str) samples at the same time is shown above each time point. Three separate experiments with duplicate or triplicate samples are shown. Scans are for a pair of bands migrating at approximately 20 kDa.

Figure 7.

 

Predicted secondary structure and calculated folding stability of 5′ UTRs of MMP-2 and -14 with several comparators. The 5′ UTRs of human glyceraldehyde-3-phosphate dehydrogenase (GAPDH), α1-actin, MAP kinase-interacting kinase (MNK-1), ornithine decarboxylase (ODC), and MMP-2 and -14 were folded using the mFold program. The most stable structure predicted is shown for each, as indicated. The calculated free energy for each structure is given shown in kilocalories per mole.

Figure 8.

 

Sequences of the 5′ UTRs for MMP-2 and -14. The 5′ UTRs for both mRNAs showing the translation start sites, ATG, upstream start for a short ORF, ATG, tracts of four or more pyrimidines (C or T) underscored and the translation of the short upstream ORF from MMP-2 with bases providing key context points.

Figure 9.

 

Hypothetical signal transduction pathways involved in TM IOP homeostasis.

The authors thank John W. Samples and Natalie Hernandez for technical assistance.