Involvement of Protein Kinase C in TNFα Regulation of Trabecular Matrix Metalloproteinases and TIMPs

Abstract

purpose. The cytokine TNFα is a strong modulator of trabecular meshwork (TM) matrix metalloproteinase (MMP) and tissue inhibitor (TIMP) expression. Studies were conducted to identify signal-transduction pathways involved.

methods. Porcine TM cells were treated with TNFα, and MMP and TIMP levels were evaluated by zymography and Western immunoblot. Inhibitors and activators of several signal-transduction pathways were used to select pathways that could be involved. Trabecular protein kinase C (PKC) isoforms were identified and localized by using Western immunoblots and confocal immunohistochemistry. Changes in subcellular distribution of PKC isoforms were evaluated. PKC isoform downregulation and additional inhibition profiles were used to refine the involvement pattern of different isoforms.

results. TNFα treatment increased MMP-1, -3, and -9 and TIMP-1 expression, whereas MMP-2 expression was not affected and TIMP-2 expression decreased. Agents that modulate protein kinase A (PKA) or inhibit phosphatidylinositol 3-kinase (PI3K) had minimal effects on trabecular MMP or TIMP induction by TNFα, whereas several agents that modulate PKC activity were effective. Trabecular cells expressed several PKC isoforms, which exhibited distinctive subcellular localization. TNFα treatment triggered some PKC isoform translocations. Exposure of trabecular cells to TNFα for 72 hours differentially downregulated several PKC isoforms. Treatment with a phorbol mitogen that stimulates most PKC isoforms produced strong increases in these MMPs. TNFα’s effects on MMP and TIMP expression were completely blocked by only one PKC inhibitor.

conclusions. The PKA and PI3K pathways appear not to be involved directly in transducing this TNFα signal, but at least one isoform of PKC seems to be required. Based on the inhibitor profiles and the downregulation and translocation studies, PKCμ appears to be critical in transducing this signal. Unraveling the remaining steps in this and in additional related TM signal-transduction pathways may provide targets for developing improved glaucoma treatments.

Ongoing trabecular meshwork (TM) extracellular matrix (ECM) turnover, initiated and controlled at least in part by the matrix metalloproteinase (MMP) family and their tissue inhibitors (TIMPs), is required to maintain normal aqueous humor outflow rates. ¹ Increased ECM turnover, initiated by elevated MMP levels in the juxtacanalicular region of the meshwork, may explain the efficacy of laser trabeculoplasty, a common treatment for glaucoma. ^{2 3} In at least a portion of cases of open-angle glaucoma, obstruction of aqueous outflow through the TM’s ECM is thought to be responsible for the elevated intraocular pressure and consequent glaucomatous optic nerve damage. ^{4 5 6} Approximately 2.47 million persons in the United States and 66.8 million worldwide are affected by this common blinding disease. ^{7 8} Understanding the normal regulation of the MMPs and TIMPs within the TM could thus have significant therapeutic implications. 

The MMPs and TIMPs are integrally involved in ECM turnover throughout the body. MMP activity is modulated by extracellular zymogen activation, by TIMP inhibition, and probably by changes in MMP protein interactions and turnover. ^{9 10 11} Intricate and complex transcriptional regulation of MMP and TIMP expression provides an additional level of ECM turnover regulation. The 5′-promoter regions of the various MMP and TIMP genes contain a variety of simple and complex enhancer elements, and their expression is modulated by numerous growth factors, cytokines, steroids, integrin ligation, and other extracellular information and conditions. ^{12 13 14 15 16 17 18} This regulation is mediated by signal-transduction pathways that have been identified in several specific cases and partially unraveled in a few others. Protein kinase C (PKC) involvement has been demonstrated in transcriptional regulation of the MMPs and TIMPs in several tissues, ^{19 20 21 22 23 24 25} although PKC’s involvement may not be a universal requirement, and different isoforms have been implicated in different tissues for different regulatory processes. 

A number of PKC isoforms have been identified that exhibit distinct regulation, subcellular distribution, and translocation patterns. They are differentially involved in diverse regulatory phenomena in various cell types. ^{26 27} These isoforms are grouped as conventional (α, βI/βII, and γ), novel (δ, ε, η, andθ ), and atypical (ζ and ι/λ). The μ isoform is somewhat unique in that it is membrane associated and does not fit completely into any of the previous categories. ²⁸ All isoforms require phosphatidylserine; the conventional, novel, and PKCμ isoforms are activated by diacylglycerols or their analogue, 12-tetradecanoylphorbol-13-acetate (TPA), and the conventional isoforms are activated by calcium. ²⁶ TPA activation of PKCζ is controversial and its direct activation of PKCι/λ is thought not to occur. Although there is some isoform variability, PKC activation generally involves phosphorylation, ligand binding, proteolytic removal of the autoinhibitory pseudosubstrate, and subcellular translocation, often directed to the membrane by diacylglycerol and phosphatidylserine. ²⁶ RACKs are proteins thought to target PKC isoforms to specific subcellular structures or substrates. ^{29 30 31}  

Trabecular cells respond to treatment with a variety of growth factors and cytokines by changing MMP and TIMP expression. ^{32 33} TNF, IL-1, and TPA are among the most effective agents we identified in producing these changes, after comparing a number of common extracellular signaling molecules. The effects of laser trabeculoplasty on trabecular MMP levels were recently shown to require mediation by IL-1 and/or TNFα. ³⁴ Thus, a study was undertaken to identify the signal-transduction pathways involved in the TNFα modulation of trabecular MMPs. 

Materials and Methods

Human eyes were obtained within 48 hours of death from the Portland Lion’s Eye Bank (Portland, OR); porcine eyes were from Carlton Packing (Carlton, OR); human recombinant TNFα and IL-1α and -1β were from R&D (Minneapolis, MN); TPA, 3-isobutyl-1-methylxanthine (IBMX), dibutyryl cAMP, H-89, KT-5720, forskolin, wortmannin, leupeptin, aprotinin, pepstatin, and fluorescein isothiocyanate (FITC)– and horseradish peroxidase–conjugated secondary antibodies were from Sigma (St. Louis, MO); GF 109203X (bisindolylmaleimide I or Gö 6850), Ro 31-8220, Gö 6976, and Gö 6983 were from CalBiochem (San Diego, CA); double-stranded DNA quantitation reagent (PicoGreen) was from Molecular Probes (Eugene, OR); protein kinase A, MMP, and TIMP antibodies were from Triple Point Biologics (Portland, OR); protein kinase C isoform and RACK-1 antibodies were from Transduction Laboratories (San Diego, CA); Dulbecco’s modified Eagle’s medium (DMEM), antibiotics, and antimycotics were from Gibco BRL (Grand Island, NY); fetal bovine serum was from HyClone (Logan, UT); chemiluminescence detection kits were from NEN Life Sciences (Boston, MA); and NIH-3T3 fibroblasts were from the American Type Culture Collection (Rockville, MD). 

Cell and Organ Culture, Treatments, and Extractions

Porcine and human TM cells and NIH-3T3 fibroblasts were cultured as previously described. ^{35 36} For one group of studies, stationary human anterior segment organ culture was used as previously described. ³⁷ The cultured trabecular cells were used as confluent monolayers at passage 3 and were maintained serum free for 48 hours before and during treatments. Except as specifically indicated, all the data shown are from porcine TM cells. The observations in Figures 1 4 and 5 were replicated in humans, showing no significant species differences. Five human cell lines and more than 20 different porcine cell lines each pooled from 20 to 40 eyes were studied. Double-stranded DNA analysis to estimate cell density in parallel flasks was conducted for some studies, as directed by the manufacturer. Because the differences between flasks were always less than ±10%, this procedure was not used in all studies. The lane-to-lane consistency of the protein-banding patterns on Western blot analysis (see description later), which were stained for 15 minutes (Ponceau S stain; Sigma), destained in 5% acetic acid, rinsed, and air-dried before probing, further verified uniform gel loading. MMP and TIMP analysis was conducted on culture medium collected 24, 48, or 72 hours after treatments and stored in aliquots frozen at −20°C until use. Analysis of PKC isoforms was conducted on extracts of cells at the times indicated. For these extractions, media were replaced with 0.5 ml of 4°C modified RIPA buffer ^{38 39} (2 mM EDTA, 2 mM EGTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate[ SDS], 100 mM NaF, 1 mM phenylmethylsulfonyl fluoride [PMSF], 20μ g/ml leupeptin, 20 μg/ml aprotinin, 20 μg/ml pepstatin, and 50 mM Tris, [pH 7.5]) per T-75 flask (BD Biosciences, Oxnard, CA); flasks were immediately placed on ice. Cells were scraped from the flasks, and the extract was sonicated, centrifuged, and frozen. For subcellular fractionation studies, cells were extracted at the indicated times by rinsing in 4°C phosphate-buffered saline, scraping the cells from the flasks in translocation buffer (2 mM EDTA, 2 mM EGTA, 50 mM NaF, 2 mM dithiothreitol [DTT], 1 mM sodium orthovanadate, 10 mM NaP₄O₇, 1 mM PMSF, 20μ g/ml leupeptin, 20 μg/ml aprotinin, 20 μg/ml pepstatin, and 30 mM Tris, [pH 7.4]) and sonicating on ice. The cytosolic fraction was the supernatant, after centrifugation at 100,000g for 30 minutes. The pellet was then resuspended in translocation buffer containing 0.1% Triton X-100, incubated at 4°C for 30 minutes, and centrifuged at 100,000g to separate the membrane and the insoluble particulate fractions. 

Zymograms and Western Immunoblots

Western immunoblots, transferred electrophoretically from standard SDS-PAGE gels to polyvinylidene fluoride (PVDF) membranes, were probed with the indicated primary antibodies, and for detection, the appropriate secondary antibodies were used with conjugated horseradish peroxidase and chemiluminescence, according to the manufacturer’s instructions. Gelatin was used as the substrate to detect gelatinase A and B (MMP-2 and -9, respectively), and β-casein was used as the substrate to detect stromelysin (MMP-3) in the zymograms (substrate, SDS-PAGE gels). ^{35 40} Confocal immunohistochemistry was conducted on trabecular cells that had been fixed for 20 minutes in 50% methanol and 50% acetone at −20°C, using an FITC-conjugated secondary antibody, with or without nuclear staining by propidium iodide, as previously described. ⁴¹  

All experiments presented were repeated at least three times and typical gels or micrographs were selected for presentation. 

Results

Trabecular Cell Response to TPA or TNFα Treatment

TPA or TNFα treatments produced dose-dependent increases in trabecular cell MMP-9 and -3 and TIMP-1 expression without affecting MMP-2 levels (Fig. 1) . TNFα but not TPA, decreased TIMP-2 expression. At higher concentrations of culture medium, modest TPA and significant TNFα induction of MMP-1 (interstitial collagenase) could be detected. These changes were also time dependent, becoming significant by 24 hours and reaching maximum changes by 72 hours at moderate doses (not shown). 

Involvement of PKC in TPA and TNFα Responses

Treatment of trabecular cells with agents that modulate PKA signal transduction, such as dibutyryl cAMP, forskolin, KT-5720, IBMX, or H-89 and attempts to block TPA’s or TNFα’s effects with these agents produced no significant effects, although trabecular cells express the typical PKA isoforms and subunits as detectable on Western immunoblots (not shown). A wide range of doses, based on common literature usage for each agent, was added for 24, 48, or 72 hours in these studies (Table 1) . Treatment with wortmannin was also ineffective in changing MMP or TIMP expression, in either the presence or absence of TPA or TNFα (data not shown). However, several relatively nonspecific PKC inhibitors, such as staurosporine, affected trabecular MMP and TIMP expression patterns as modulated by TNFα or TPA (data not shown). These observations, plus the strong responsiveness to TPA, prompted a more detailed analysis of trabecular PKC. 

Expression of PKC Isoforms by Trabecular Cells

Western blot analysis of porcine or human trabecular cells in cell or organ culture (Fig. 2A) showed that these cells expressed PKCα, -γ, -ι/λ, -θ, -ζ, and -μ at relatively high levels. PKCε was detectable, although at very low levels. Only human cells in culture produced detectable amounts of either βI or II. Low levels of the β isoform could be detected in highly concentrated extracts from human or porcine cells (data not shown). These cells also expressed high levels of the putative PKC-anchoring protein, RACK-1 (Fig. 2A) . All these PKC isoforms and RACK-1 were observed at the predicted M_r, based on observations in other tissues reported in the literature. Some small species, and possibly phosphorylation-dependent M_r differences, were apparent (data not shown). 

Subcellular Localization of Trabecular PKC Isoforms by Confocal Immunohistochemistry

To localize several of the PKC isoforms, confocal immunohistochemistry was used with untreated trabecular cells (Fig. 2B) . A significant portion of PKCα, -γ, -ι/λ, and -ζ immunostaining was associated with filamentous cytoplasmic strands, which was apparently the trabecular cytoskeleton. However, very distinctive patterns were seen with each isoform. Note that the different isoforms are shown in Figure 2 at different magnifications to accentuate the most distinctive aspect of their distribution, with the scale bar signifying 10 μm. PKCα, -ι/λ, and -ζ showed a punctate scattering throughout or at the surface of the cell. PKCα staining was most intense over and around the nucleus, whereas staining for PKCγ and -ι/λ were negative in the nucleus. From optical sections through the nucleus, it was apparent that PKCα was not actually predominantly within the nucleus (data not shown). Immunostaining for PKCι/λ, and to a lesser extent for PKCα and -ζ, was apparent at the cell periphery. PKCγ was clearly associated with filamentous strands, which are concentrated in some cells in a wide zone around the nucleus. Some of this immunostaining may also be associated with the Golgi–endoplasmic reticulum. Strong PKCμ immunostaining was associated with what appeared to be the Golgi apparatus in the confocal images. Punctate, probably cell-surface–associated, PKCμ immunostaining was apparent across the cells. Very distinctive PKCμ immunostaining also appeared within the nucleus, apparently surrounding and within the nucleoli. 

Treatment of trabecular cells with TPA or TNFα did not produce simple interpretable changes in these immunostaining patterns. Some differences were apparent, but they were modest and quantitative rather than qualitative or absolute (not shown). The TPA-triggered increase in membrane-associated PKCα, -γ, and -ζ observed in the translocation studies (described in the next section) was modestly apparent in the immunostaining localization also (data not shown). 

Differences in PKC Isoform Translocation Induced by TPA and TNFα

Because one step in the activation and action of some PKC isoforms involves translocation between subcellular compartments, we evaluated the distribution of several PKC isoforms among the membrane, cytosolic, or particulate fractions at various times after TPA or TNFα treatment. The membrane fraction at various times after treatment is shown in Figure 3 . Trabecular cells show an unusually high proportion of all the PKC isoforms in the particulate fraction and a very low proportion in the cytosolic fraction. More than 90% of trabecular PKCα, -γ, -ε, -ι/λ, and -ζ and more than 75% of trabecular PKCμ were found in the particulate fraction, and continuous or transient cytoplasmic levels were very low (data not shown). To eliminate the possibility that this was a methodologic rather that a cell-type phenomenon, we conducted parallel studies to compare trabecular cells to NIH-3T3 fibroblasts, with and without TPA and TNFα treatments (data not shown). In the fibroblasts, we found distributions similar to those normally reported in the literature. Fibroblast PKC is still predominantly particulate, but dramatically more is seen transiently in the cytosolic fraction after treatments (data not shown). 

TPA treatment produced a strong increase in membrane-bound (detergent extractable) PKCα, -γ, and -ζ over time, reaching a maxima at 30 minutes and declining modestly at 60 minutes (Fig. 3) . PKCε associated with the membrane increased by 5 minutes, reached a maximum at 30 minutes and declined slightly by 60 minutes after TPA treatment. TNFα, by contrast, caused a modest reduction in membrane-associated PKCα, -γ, -ε, and -ζ isoforms compared with control at early times with a return to baseline by 30 to 60 minutes. PKCι/λ levels in the membrane fraction were relatively high and appeared unchanged in response to both TPA and TNFα. The membrane fraction of PKCμ was high in controls and remained high after either TPA or TNFα treatment. After either treatment, although more pronounced after TPA, PKCμ became a doublet, with the appearance of a slightly slower-migrating band reaching a maximum at 30 minutes The cytosolic RACK-1 levels (data not shown) were several times as high as in the membrane fraction, and neither TPA nor TNFα changed this distribution dramatically. 

Downregulation of PKC Isoforms by Extended Treatments.

PKC downregulation often provides an indication of PKC involvement in a regulatory process. Although shorter treatment times had less dramatic effects on PKC isoform levels, by 72 hours of treatment several isoforms were strongly downregulated (Fig. 4) . PKCα, -γ, and -ε levels were almost undetectable, whereas PKCι/λ, -μ, or -ζ levels are only modestly affected by TPA treatment. TNFα had modest effects on PKCα, -γ, and -ι/λ levels; moderate effects on PKCε levels; and strong effects on PKCμ and -ζ levels. IL-1α, another important modulator of trabecular MMP and TIMP expression, had similar, but not identical, downregulating effects on these isoforms. 

Differential Effects of Synthetic PKC Inhibitors on TPA- and TNFα-Induced MMP and TIMP Expression.

Although a large number of PKC inhibitors have been developed and characterized, most exhibit only limited differential effects on the various PKC isoforms and most have limited specificity for PKCs over other kinase families or limited cell permeability. Our initial studies with staurosporine were suggestive, but the differential specificity of this inhibitor for PKCs over myosin light-chain kinase was only 2-fold and over PKA was only 10-fold, and interpretations are therefore difficult. In addition, added without TPA or TNFα, staurosporine had effects on MMP and TIMP expression; also, at low doses, it was synergistic with TPA (data not shown). When light-activated calphostin C was added, it killed trabecular cells before expression changes could be analyzed; thus, its usefulness in this study was limited. Several third-generation PKC inhibitors have been developed with increased PKC specificity compared with other protein kinases and that show significant PKC isoform differential effectiveness (Table 1) . ^{27 42 43 44 45 46 47 48}  

Thus, we evaluated the effects of these inhibitors on trabecular MMP and TIMP expression induced by TPA and TNFα (Fig. 5) . None of the inhibitors had appreciable affects on MMP or TIMP levels in the absence of the stimulatory agents. Bis I (GF109203X), Gö 6976, and Ro 31-8220 showed dose-dependent inhibition of all the changes in expression induced by TPA. However, differential effects were seen with these inhibitors’ ability to block TNFα’s effects on trabecular expression of these proteins (Fig. 5) . Bis I was unable to block any of TNFα’s effects. Gö 6976 was very potent, blocking the TNFα-induced increases in MMP-3 and -1 and TIMP-1 and the decrease in TIMP-2. It was only partially effective in blocking MMP-9 induction by TNFα. At the highest dose, approximately 10 times its 50% inhibitory concentration (IC₅₀), Ro 31-8220 slightly reduced the MMP-9 and TIMP-1 increases and markedly reduced MMP-3 and -1 immunostaining, without changing MMP-3 activity or the TIMP-2 level decrease caused by TNFα. To further evaluate the possible involvement of PKCζ in TNFα’s effect, we repeated similar studies with another PKC inhibitor, Gö 6983, which inhibited PKCα, -β, -γ, -δ, -ε, and -ζ at 6 to 10 nM and PKCμ at 20μ M. This inhibitor was effective against TPA’s effects, but had no effect on TNFα’s effects (data not shown). 

Discussion

TNFα’s induction of trabecular MMPs and TIMPs appears to require PKC and not PKA or phosphatidylinositol 3- or 4-kinases. ^{49 50 51} Based primarily on the inhibitor studies, TNFα appeared to cause trabecular MMP-3, -9, and -1 and TIMP-1 increases with an associated TIMP-2 decrease through a signal-transduction pathway(s) that included PKCμ as a required step. Although no selective PKCμ inhibitor is available, the combination of inhibitor specificities that we used have been studied in considerable detail. Based on the combination of these specificities, PKCμ appears to be the only PKC isoform that is required in this signal-transduction process. One caveat to this assignment is that the inhibition profile of PKCι/λ is incomplete (Table 1) ; thus, we cannot be absolutely certain that it is not involved. Because PKCι/λ is in the same family as PKCζ, it should share this isoform’s inhibition profile. However, this has not been demonstrated. 

The TM expressed a discrete but not particularly unique profile of PKC isoforms. We found PKCα, -γ, -ι/λ, -ε, -ζ, -μ, and -θ to be detectable in human and pig tissue and cell culture, each at the appropriate M_r, based on data in the literature. PKCβ was barely detectable, and we did not detect other isoforms. For reasons that are not apparent, we clearly saw several isoforms not detected in a previous study, in which only PKCα and -ε were found. ⁵² The apparent PKC anchoring protein, RACK-1, was present at high levels in the trabecular cytosolic, membrane, and particulate fractions. 

The trabecular subcellular distribution pattern of PKC isoforms suggests that each fulfills separate trabecular functions. In general, the trabecular cell localization of PKC is more distinctive by isoform than that reported for NIH 3T3 cells, ⁵³ which may reflect the highly differentiated state of trabecular cells. Changes in localization with these treatments did not provide clear indications of which isoforms were involved. The novel PKCμ distribution is intriguing. The distinctive apparent Golgi localization that we observed for PKCμ is compatible with a prior study, which localized this isoform specifically to the Golgi compartment. ⁵⁴ These investigators also suggested that it may be involved in glycosaminoglycan or glycoprotein posttranslational processing or at least in basal protein transport and secretion. The MMPs and TIMPs have been shown to exhibit strong vectorial secretion in the confluent endothelial cell. ⁵⁵ In addition, many of the MMPs and TIMPs have glycosylated and unglycosylated forms. However, this does not seem likely to be the primary site of the critical PKCμ involvement in this specific signal-transduction process, because the ratio of glycosylated and unglycosylated MMP forms is not affected by the PKC inhibitors. Thus, the portion of trabecular PKCμ that is critically involved in transducing this signal could be the apparent Golgi-associated fraction, but seems more likely to be the punctate fraction that is apparently dispersed on the cell surface. The apparent nucleolar PKCμ immunostaining was not observed in previous studies. ⁵⁴ This apparent nucleolar immunostaining was reproducible and specific for this antibody, but it could very well be an artifact. We have not attempted the difficult studies necessary to further clarify this point. 

When the subcellular distribution of PKCμ after TNFα or TPA treatment was evaluated, a significant bandshift was observed (Fig. 3) . Presumably, this transient upper band reflects a posttranslational modification, probably a phosphorylation. This could reflect an activation of this isoform and is probably of significance in the regulatory process. Multiple phosphorylation sites have been shown to be important in modulating this isoform’s activity. ⁵⁶  

The different translocation patterns of PKC isoforms observed in signal transduction by TPA and TNFα is intriguing. To the extent that isoform translocation provides information about isoform utilization, TPA could be using PKCα, -γ, -ε, -ζ and/or -μ to induce trabecular MMP and TIMP changes. Although TNFα produced a very small shift of several PKC isoforms away from the membrane fraction at early times, this may or may not reflect functional effects. If the shift was to a particulate position, this small change was not detectable on the very large background level in trabecular cells. 

The PKC isoform downregulation is also of interest. The implication, commonly accepted in the literature, is that a PKC isoform that is downregulated by extended treatment with an agent is probably actively involved in some aspect of signal transduction by this agent. These downregulation studies provide support for a PKCμ step in the process. However, it can be assumed that multiple PKC isoforms can be involved in several trabecular processes triggered by TPA or TNFα, whether or not they are required for this MMP-TIMP effect. It is interesting that the time required to achieve trabecular PKC isoform downregulation was longer than that observed in many other cells, suggesting a slower protein turnover rate. 

Because carefully regulated increases in trabecular ECM turnover by these MMPs would increase outflow facility, ¹ studies further unraveling the steps in this and in other pathways involved in this process may allow the development of improved therapies for glaucoma. 

 Supported by National Institutes of Health Grants EY03279, EY08247, and EY10572 and by grants from the Glaucoma Research Foundation, San Francisco, California; Research to Prevent Blindness, New York, New York; and Alcon Laboratories, Fort Worth, Texas.

 Submitted for publication February 20, 2001; revised July 13, 2001; accepted July 20, 2001.

 Commercial relationships policy: N.

 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.

 

Modulation of MMP and TIMP levels by the phorbol mitogen TPA and the cytokine TNFα. Porcine TM cells at confluence were maintained serum free for 48 hours before and during treatment with TPA or TNFα at the concentrations indicated. Media were collected after 72 hours of treatment and activity was evaluated by gelatin and β-casein substrate zymography or specific protein levels were evaluated by Western immunoblot analysis, as indicated. The zymograms were photographically reversed to enhance viewing contrast. S Free, serum free control.

Table 1.

 

Effects of Protein Kinase Inhibitors on TNFα Stimulation or of Protein Kinase Activators on Trabecular MMPs

Inhibitor or Activator	Kinase	IC50	Concentrations	Effect
GF109203X (I)	PKC α, β, γ	8–20 nM	10, 50, 100, and 200 nM	0
	PKC δ, ε	132–210 nM
	PKC ζ, ι/λ, μ	2–6 μM
Ro 31-8220 (I)	PKC α, β, γ, ε	5–27 nM	10, 50, 100, and 200 nM	0
Gö 6976 (I)	PKC α, β, μ	2–20 nM	10, 50, 100, and 200 nM	−
	PKC δ, ε, ζ	≫3 μM
Gö 6983 (I)	PKC α, β, γ	6–7 nM	10, 50, 100, and 200 nM	0
	PKC δ, ζ	1–60 nM
	PKC μ	20 μM
TPA/PMA (A)	PKC α, β, γδ, ε, μ	NA	1, 10, and 25 ng/ml	+
	PKC ζ, ι/λ	NA
Wortmannin (I)	PI 3-Kinase	5 nM	10, 25, and 100 nM	0
Dibutyryl-cAMP (A)	PKA	NA	1, 5, and 10 mM	0
IBMX (A)*	PKA	NA	1, 5, and 10 mM	0
KT-5720 (I)	PKA	56 nM	50, 200, and 500 nM	0
H-89 (I)	PKA	48 nM	10, 50, and 200 nM	0

 I, inhibitor; A, activator; −, inhibited TNFα-induced MMP increase; +, increased MMP without TNFα addition; 0, no effect on MMP levels with or without TNFα addition; NA, inhibition constants are not applicable for activators.

*  Indirect activation through inhibition of cAMP breakdown by phosphodiesterase.

Figure 2.

 

Expression and subcellular localization of PKC isoforms by trabecular cells. (A) Cellular extracts of porcine and human TM tissue and cultured human trabecular cells were analyzed by Western immunoblot analysis with isoform-specific PKC antibodies. Positive controls, when available, were provided by the supplier along with the antibodies. (B) Cultured porcine TM cells were fixed and immunostained with PKC isoform-specific antibodies, and isoform distributions were analyzed by fluorescence confocal microscopy. Scale bars, 10 μm.

Figure 3.

 

PKC isoforms in the membrane fraction after TPA and TNFα treatments. Particulate, cytosolic, and membrane fractions of trabecular cells were extracted 0, 5, 15, 30, and 60 minutes after treatment with 25 ng/ml of TPA or TNFα. The membrane fraction is shown at the times indicated. Fractions were evaluated by Western immunoblot analysis with isoform-specific antibodies as labeled. RACK-1 levels were also included in the analysis.

Figure 4.

 

Downregulation of PKC isoforms with extended TPA or TNFα treatment. Treatment of trabecular cells for 72 hours with nothing (S Free), TPA, TNFα, or IL-1α (all at 25 ng/ml) produced differential degrees of downregulation of the various PKC isoforms, as indicated.

Figure 5.

 

Effects of PKC inhibitors on TPA and TNFα induction of MMPs and TIMPs. Media from trabecular cells were evaluated by zymography and Western immunoblot analysis for MMP and TIMP levels 72 hours after treatment with the indicated concentrations of the PKC inhibitors, Bis (GF109203X), Go (Gö 6976), or Ro (Ro 31-8220), with and without simultaneous addition of 10 ng/ml TPA (A) or 25 ng/ml TNFα (B).

The authors thank the Microbiology and Molecular Immunology Core Facility for confocal microscopy.