Free
Lens  |   June 2012
MMP2 Activity is Critical for TGFβ2-Induced Matrix Contraction—Implications for Fibrosis
Author Affiliations & Notes
  • Julie A. Eldred
    School of Biological Sciences, University of East Anglia, Norwich, Norfolk, UK;
  • Lisa M. Hodgkinson
    School of Biological Sciences, University of East Anglia, Norwich, Norfolk, UK;
  • Lucy J. Dawes
    School of Biological Sciences, University of East Anglia, Norwich, Norfolk, UK;
  • John R. Reddan
    and Oakland University, Rochester, Michigan.
  • Dylan R. Edwards
    School of Biological Sciences, University of East Anglia, Norwich, Norfolk, UK;
  • I. Michael Wormstone
    School of Biological Sciences, University of East Anglia, Norwich, Norfolk, UK;
  • Corresponding author: I. Michael Wormstone, School of Biological Sciences, University of East Anglia, Norwich NR4 7TJ, UK; i.m.wormstone@uea.ac.uk
Investigative Ophthalmology & Visual Science June 2012, Vol.53, 4085-4098. doi:10.1167/iovs.12-9457
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Julie A. Eldred, Lisa M. Hodgkinson, Lucy J. Dawes, John R. Reddan, Dylan R. Edwards, I. Michael Wormstone; MMP2 Activity is Critical for TGFβ2-Induced Matrix Contraction—Implications for Fibrosis. Invest. Ophthalmol. Vis. Sci. 2012;53(7):4085-4098. doi: 10.1167/iovs.12-9457.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose.: The fibrotic lens disorder posterior capsule opacification (PCO) develops in millions of patients following cataract surgery. PCO characteristics are extensive extracellular matrix (ECM) production and contraction of the posterior lens capsule, resulting in light-scattering ECM modification (wrinkling). The pro-fibrotic cytokine transforming growth factor beta (TGFβ) is central to PCO development. This study aimed to elucidate the role of the ECM modulators matrix metalloproteinases (MMPs) in TGFβ-mediated PCO formation.

Methods.: The human lens epithelial cell-line FHL-124 and human capsular bag models were employed. Gene expression of MMP family members was determined by oligonucleotide microarray and quantitative real-time RT-PCR. MMP2 and MT1-MMP protein levels were analyzed by ELISA and Western blotting, respectively. Matrix contraction was determined using an FHL-124 patch contraction assay; at end-point, cells were stained with Coomassie brilliant blue and area was determined using image analysis software. Cell coverage and wrinkle formation on the posterior capsule were also assessed using human capsular bag models.

Results.: Active TGFβ2 (10 ng/mL) increased gene and protein levels of MMP2 and MT1-MMP and induced matrix contraction in FHL-124 cells. Specific siRNA inhibition of MT1-MMP did not suppress TGFβ2-induced matrix contraction. Active TGFβ2-mediated contraction was prevented by broad-spectrum MMP inhibitor GM6001 (25 μM), MMP2 siRNA, and MMP2 neutralizing antibody (4 μg/mL). TGFβ2-induced wrinkle formation was attenuated in human capsular bags treated with MMP2 neutralizing antibody (20 μg/mL).

Conclusions.: MMP2 plays a critical role in TGFβ2-mediated matrix contraction, which appears to be independent of MT1-MMP. MMP2 inhibition provides a novel strategy for the treatment of PCO and potentially other fibrotic disorders.

Introduction
Fibrotic diseases are characterized by scarring due to excessive production, deposition, and contraction of extracellular matrix (ECM). 1 Posterior capsule opacification (PCO) is a fibrotic condition of the lens that affects millions of people following cataract surgery, resulting in a secondary loss of vision. 
During cataract extraction surgery, the majority of anterior lens epithelial cells that reside on the internal surface of the anterior lens capsule are removed; however, any remaining cells begin to grow across denuded regions of the anterior capsule and the previously cell-free posterior capsule. 2 In addition, many of these cells undergo the process of epithelial to mesenchymal transformation becoming myofibroblastic. 3 These actively growing myofibroblasts secrete excessive ECM and develop a contractile phenotype that can result in matrix-filled wrinkles on the posterior capsule that cause light scatter and thus secondary visual loss. 4  
In 2002 the World Health Organization (WHO) estimated that globally 37 million people were blind, with 48% of this burden attributable to cataract. 5 The WHO Vision 2020 proposal estimated that by the year 2020 greater than 30 million people will require cataract surgery. 6 However, 2 years post cataract extraction surgery approximately 15% of patients will develop PCO, rising to approximately 31.9% by 4 years postsurgery 7 ; in younger patients (<40 years) the 4 year postsurgery incidence is approximately 70%. 8 PCO is therefore a significant socioeconomic problem and presents a considerable burden to global healthcare budgets. 
A critical factor associated with PCO development is the cytokine transforming growth factor beta (TGFβ) often referred to as the key mediator or master switch of fibrosis. 9  
Activation of TGFβ is integral to PCO development and results from the wound-healing response initiated by the trauma of cataract extraction surgery. 3,10,11 TGFβ typically resides in normal healthy aqueous humor in a latent form 12 but can be activated by inflammatory response factors such as lipopolysaccharides, thrombospondins, and plasmin proteases such as matrix metalloproteinases (MMPs). 1315 Furthermore, total TGFβ levels in postoperative aqueous humor of rabbits have been shown to increase 2-fold compared with preoperative aqueous humor. 16 In an established model of PCO using human donor lenses, 17 addition of active TGFβ2 resulted in posterior capsular wrinkling that was representative of the PCO observed in donor lenses that had previously undergone surgery and were implanted with an intraocular lens (IOL). 18 Additionally, TGFβ has also been implicated in the progression of anterior subcapsular cataract (ASC), 19 a disorder with many phenotypic similarities to PCO. 
Interestingly, as well as being a potent inducer of matrix contraction, TGFβ is also known to regulate expression of members of the MMP family in human lens tissue. 18,20 MMPs play an essential role in both normal biological processes such as embryogenesis and wound healing and pathological disorders such as arthritis, cancer, nephritis, and importantly fibrosis. 21 The substrates of MMPs include lens ECM components such as collagen type IV, 22 and it is therefore possible that the MMPs might contribute to the matrix deforming effects observed in response to elevated levels of active TGFβ. In 2002 Wormstone et al. 18 determined that media from human lenses that underwent surgical trauma in sham cataract operations, contained transiently elevated levels of MMP2 and 9 in the subsequent 4 days of culture. 18 Additionally, pro-form MMP2 and 9 levels were enhanced and sustained by treatment with 10 ng/mL active TGFβ2. 18 Treatment of rat lenses with TGFβ to induce ASC has also been shown to increase the levels of pro-form MMP2 and 9 and active MMP2 in bathing media. 23 This study also demonstrated that the broad-spectrum MMP inhibitor GM6001 and a specific MMP2/9 inhibitor reduced the formation of TGFβ-induced ASC. 23 MT1-MMP (MMP14) expression is also reported in the lens 24 and is generally proposed to play a role in the activation of MMP2 through proteolytic cleavage to remove the pro-domain. 25 MT1-MMP, pro-form MMP2, and tissue inhibitor of matrix metalloproteases 2 (TIMP2) form a trimolecular complex, 26 which results in removal of the inhibitory pro-domain from MMP2 by an adjacent free MT1-MMP molecule. 
The current body of work has identified that MMP inhibition prevents active TGFβ2-induced matrix contraction by lens epithelial cells. Specifically MMP2 was found to be essential in this process. Furthermore, we have demonstrated that the activation of pro-MMP2 to active MMP2 occurs in a MT1-MMP independent fashion. Specific deletion or neutralization of MMP2 activity following cataract surgery may prove beneficial in suppressing TGFβ-induced matrix contraction. This could reduce light scatter incidence in cataract patients and maintain a lens epithelial phenotype. Therefore, inhibition of MMP2 serves as a valuable therapeutic target against PCO. 
Materials and Methods
FHL-124 Cells
The fetal human lens epithelial cell-line FHL-124 was used throughout this study because the cells have been previously shown to share 99.5% homology with native human lens epithelium. 27 FHL-124 cells were kindly provided by Prof J.R. Reddan (Oakland University, Rochester, MI). 
All reagents were from Sigma-Aldrich (Poole, Dorset, UK) unless otherwise stated. 
Patch Contraction Assay
The assay previously described by Dawes et al. 28 was employed to assess matrix contraction. FHL-124 cells were seeded at four distinct sites on a 35-mm tissue culture dish (Corning Incorporated Life Sciences, Corning, NY) at a density of 5000 cells/25 μL Eagle's minimum essential medium (EMEM) supplemented with 5% fetal calf serum (FCS, Gibco, Paisley, Scotland, UK) and 0.05 mg/mL gentamicin and maintained in a 35°C, 5% CO2 incubator until confluent patches spanning 5 mm developed. The medium was then replaced with nonsupplemented (serum-free [SF]) EMEM and the cells were cultured for a further 24 hours at 35°C in a 5% CO2 incubator. Subsequently medium was removed from four patch culture dishes and cells fixed for 30 minutes with 4% formaldehyde at room temperature followed by washing in PBS, this represented a “t = 0” reference control. All remaining cell cultures were exposed to experimental conditions for up to 3 days. Experiments were concluded once cell-free regions (holes) appeared within the central region of the TGFβ2 treated patches. Patch assay dishes were subsequently fixed for 30 minutes with 4% formaldehyde at room temperature. The patch assays (including t = 0 controls) were washed with PBS and stained with Coomassie brilliant blue (a total protein dye; Merck, Darmstadt, Germany) for 30 minutes, which enabled patches to be visualized and measured. The cells were washed several times in PBS to remove excess Coomassie blue protein dye, and images of patches were captured using a charged coupled device (CCD) camera and grabber software (Synoptics, Cambridge, UK) and analysis was performed using PC Image (Foster Findlay, Newcastle-upon-Tyne, UK). 
Oligonucleotide Microarray Analysis
A retrospective analysis was performed on oligonucleotide microarray data sets established by Dawes et al. 29 These data compared the level of gene expression in FHL-124 cells in the presence and absence of TGFβ. With respect to the current study we specifically examined these data for expression patterns of the following MMP and TIMP family members: MMP1 (accession number M13509); MMP2 (M55593); MMP3 (X05232); MMP4 (U10324); MMP7 (Z11887); MMP8 (J05556); MMP9 (J05070); MMP10 (X07820); MMP11 (X57766); MMP12 (L23808); MMP13 (X75308); MMP14 (Z48481); MMP15 (Z48482); MMP16 (D83646); MMP17 (NM_016155); MMP18 (U38320); MMP20 (Y12779); MMP23A (AF056200); MMP23B (AB010962); MMP24 (AB021227); TIMP1 (X03124); TIMP2 (AL110197); TIMP3 (AL023282); and TIMP4 (NM_003256). 
TaqMan Analysis
FHL-124 cells were seeded onto 35-mm tissue culture dishes (Corning Incorporated Life Sciences) at a density of 35,000 cells/1.5 mL in EMEM supplemented with 5% FCS and 0.05 mg/mL gentamicin antibiotic and maintained in a 35°C, 5% CO2 incubator until dishes were >70% confluent with cells. The medium was then replaced with nonsupplemented SF EMEM and the cells were cultured for a further 24 hours at 35°C in a 5% CO2 incubator. Following serum starvation, cells were treated for a further 24 hours with nonsupplemented SF EMEM (control) or TGFβ2 (10 ng/mL). Total RNA was subsequently prepared using an RNeasy minikit (Qiagen, West Sussex, UK) and quantitative real-time PCR was used to analyze mRNA expression of MMP family members as described by Hodgkinson et al. 24 ; for MMP and TIMP primer/probe sets used see Nuttall et al. 30  
ELISA Analysis of Total MMP2 Protein Levels
FHL-124 cells were seeded and maintained as described for oligonucleotide microarray analysis. Following serum starvation cells were placed into experimental conditions for a further 48 hours. Culture media was collected from the dishes at end point (48 hours) and stored at −20°C until ELISA analysis with a commercially available ELISA kit (R&D Systems, Abingdon, UK). MMP2 concentration in culture media was determined by comparison to a standard series using a spectrophotometer multiwell plate reader set at a wavelength of 450 nm (Victor; EC&G Wallac, Cambridge, UK). 
siRNA Transfection
Silencer Validated MMP2 siRNA, Silencer Predesigned MMP14 (MT1-MMP) siRNA, and Silencer siRNA Negative control #2 (Universal scrambled siRNA) sense 5′-3′, antisense 5′-3′, were employed in this study (Ambion, Huntingdon, UK). FHL-124 cells were seeded onto 35-mm tissue culture dishes at either 35,000 cells/1.5 mL for protein extraction or as four patches of 5000 cells/25 μL, for patch assay analysis. Cells were maintained at 35°C in a 5% CO2 incubator in EMEM supplemented with 5% FCS and 0.05 mg/mL gentamicin antibiotic for 3 days and then serum starved for 1 day. Transfections were performed with 100 nM siRNAs according to the manufacturer's instructions. One microliter of either MMP2, MT1-MMP (MMP14), or negative control siRNA (final concentration of 100 nM) was added to 184 μL reduced-serum medium (Optimem; Invitrogen, Paisley, UK). A separate solution containing 5 μL oligofectamine (Invitrogen) and 10 μL of Optimem reduced-serum medium was also prepared for each siRNA used. The separate solutions were incubated at room temperature for 5 minutes then each separate siRNA containing solution was combined with an oligofectamine-containing solution and mixed by gentle agitation to create a transfection mix for each distinct siRNA. The combined solutions were then further incubated at room temperature for 15 to 20 minutes. During this incubation period serum-containing medium was aspirated from cell preparations and the cells washed with 2 mL of Optimem reduced-serum medium. This solution was aspirated and replaced with 800 μL of fresh reduced-serum medium. Following the incubation period, 200 μL of each siRNA transfection mix was added to appropriate culture dishes. The cells were incubated at 35°C in a 5% CO2 atmosphere for 4 hours to initiate transfection. After 4 hours a further 500 μL of Optimem reduced-serum medium was added to each dish. Following a further 20-hour incubation at 35°C in a 5% CO2 incubator, cell preparations were placed into experimental conditions such that 10 ng/mL TGFβ2 was added to the appropriate dishes. At appropriate time-points cell lysis for RNA (24 hour) or protein (48 hour) was carried out, whilst patch assays were terminated after 3 days in experimental conditions. 
Western Blotting
Protein Extraction
Following the transfection experimental period, culture media were collected in 1.5-mL Eppendorf tubes, snap frozen in liquid nitrogen, and stored at −80°C. FHL-124 cells were then immediately washed briefly with ice-cold PBS. Ice-cold PBS was aspirated and replaced with Daubs lysis buffer containing 50 mM HEPES pH 7.5, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 10% glycerol, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 10 mM sodium fluoride, 1 mM phenylmethylsulfonylfluoride, and 10 μg/mL aprotinin. 31 Following a 10-minute incubation on ice, cells were detached from the dishes using a cell-scraper and the culture dishes were further incubated on ice for another 10 minutes. The resultant lysates were carefully collected and transferred to 1.5-mL Eppendorf tubes then centrifuged for 10 minutes at 16,060 g. The soluble fractions were collected from each tube and transferred to fresh 0.5-mL Eppendorf tubes and stored at −20°C, whilst the insoluble fractions were discarded. 
To ensure equal amounts of protein for each sample could be analyzed a protein assay of each lysate was performed. A protein standard serial dilution set ranging from 0 to 1000 μg/mL was created from a 2 mg/mL BSA in lysis buffer stock solution. Samples of each standard and unknown sample (10 μL) were pipetted into separate wells of a transparent 96-well microtiter plate, along with 40 μL of double distilled water. All standards were tested in triplicate, whilst all unknown samples were tested in duplicate. Two hundred microliters of a working reagent (BCA protein assay kit; Thermoscientific, Cramlington, Northumberland, UK) was added to each well (this was made at a concentration of 1:50); the plate was gently shaken on a microtiter plate shaker for 1 minute prior to a 1-hour incubation at 35°C. After this time the plate was left at room temperature to cool for 5 minutes and then analyzed in a spectrophotometer plate reader at a setting of 565 nm. The protein content of each unknown sample could then be calculated from the standard dilution series. Equal amounts of protein from each sample were loaded onto 10% SDS-PAGE gels for electrophoresis, and then transferred onto PVDF membrane (NEN Life Science Products, Boston, MA) with a Trans-Blot semi-dry Transfer Cell (Bio-Rad, Hemel Hempstead, Hertfordshire, UK). Membranes were washed with PBS containing Tween and milk protein to block nonspecific sites. Specific proteins were probed using anti-MMP-14 and anti-β-actin (Cell Signalling Technology, Hitchin, Hertfordshire, UK) and further detected using the ECL+ blotting analysis system (GE Healthcare, Little Chalfont, Buckinghamshire, UK) and Hyperfilm ECL photographic paper (GE Healthcare). Resultant protein bands were scanned and measured using Kodak 1D 3.5 imaging software (Eastman Kodak Company, Rochester, NY). 
Total RNA Extraction and cDNA Generation
FHL-124 cells were seeded onto 35-mm tissue culture dishes at a density of 35,000 cells in 1.5 mL 5% FCS-supplemented EMEM containing 0.05 mg/mL gentamicin. Cells were left to reach 70% confluency of the dish and then media were aspirated and replaced with SF EMEM containing 0.05 mg/mL gentamicin for a further 24 hours. After this period of serum starvation cells were left in either control (SF EMEM) conditions or treated with 10 ng/mL TGFβ2 for 24 hours. Total RNA was subsequently extracted from FHL-124 cells using an RNeasy kit (RNeasy microkits; Qiagen) in accordance with the manufacturer's instructions. In the initial step, RLT buffer (containing beta mercaptoethanol) was added directly to PBS-washed FHL-124 monolayers and the cells were detached from the dishes using a cell scraper. The resultant cell lysates were transferred to Eppendorf tubes and the lysates passed at least seven times through a 20-gauge needle. The remainder of the protocol was as described by the manufacturer and included a DNAse step. RNA was quantified with a spectrophotometer (ND-1000; NanoDrop, Wilmington, DE) the ratio of absorbance at 260 nm and 280 nm was measured, this ranged from 1.8 to 2.2 (mean 2.0), which is indicative of a pure (uncontaminated) RNA sample. Where possible, total RNA was immediately used for cDNA generation or was briefly stored at −80°C. Generation of cDNA was performed with reverse transcriptase (Superscript II; Invitrogen) and random primers (Promega, Southampton, UK), according to standard protocols. 
Real-Time PCR
Quantitative real-time PCR was used to analyze mRNA expression for all MMP genes (for a comprehensive description of MMPs and TIMPs primer/probe sets used, see Nuttall et al.30). Assuming 100% efficiency in the reverse transcription reaction, either 1 or 5 ng cDNA was used in real-time PCR performed using a real-time PCR machine (ABI7700; Applied Biosystems). Reagent-based assays (TaqMan Universal PCR Master Mix, No AmpErase UNG; Applied Biosystems) and PCR reagents were used according to the manufacturers' instructions. The amount of amplification associated with priming from genomic DNA contamination was evaluated with control reverse transcription reactions containing all reagents without reverse transcriptase. Conditions for the PCR amplification were 2 minutes at 50°C, 10 minutes at 95°C, and then 40 cycles, each consisting of 15 seconds at 95°C and 1 minute at 60°C. The cycle number at which amplification entered the exponential phase (cycle threshold) was determined, and this number was used as an indicator for the amount of target RNA in each tissue analyzed. To determine the relative RNA levels in the samples, standard curves for each primer/probe set were prepared by taking cDNA from one sample and making 2-fold serial dilutions covering the range equivalent to 20 to 0.625 ng RNA (for 18S analysis, the range was 1 to 0.03125 ng). Differences in the total amount of RNA present in each sample were normalized to endogenous 18S ribosomal RNA gene expression, as previously described. 
MMP2 Biotrak Assay
FHL-124 cells were routinely seeded on to 35-mm tissue culture dishes (Corning Incorporated Life Sciences) at a density of 35,000/1.5 mL in 5% FCS-supplemented EMEM with 0.05 mg/mL gentamicin. Cells were maintained at 35°C in a 5% CO2 incubator for 72 hours, after which time they were washed briefly with Optimem media and transfected with either MT1-MMP or scramble siRNA (see siRNA transfection protocol). Cells were maintained for a further 24 hours at 35°C in a 5% CO2 incubator before addition of 10 ng/mL TGFβ2 to the appropriate dishes. Dishes were incubated for an additional 48 hours at 35°C in a 5% CO2 incubator, after which time media was collected from the dishes and stored at −20°C until analysis using a MMP2 Biotrak assay kit (GE Healthcare). All reagents listed were supplied with the Biotrak kit unless otherwise stated. Biotrak assay buffer, wash buffer, and the 96-well plate were slowly warmed to 20°C to 27°C prior to use. The assay buffer was reconstituted in 100 mL distilled water and 1 mL of assay buffer added to supplied Biotrak standard to create a 24 ng/mL stock and stored on ice. Additionally, the wash buffer was diluted with 500 mL distilled water and also kept on ice. A standard dilution series was created by adding 500 μL stock standard (24 ng/mL) to 500 μL assay buffer to form a 12 ng/mL standard, this procedure was repeated until a series containing 3, 1.5, 0.75, 0.38, and 0.19 ng/mL standards was formed (standards were vortexed in between each dilution step). The 12 and 6 ng/mL standards were then discarded, and the dilution series stored on ice until required. All standards and samples were run in duplicate on the plate. A 100-μL volume of assay buffer was loaded into two wells on the plate to create a 0 ng/mL standard (blank). Subsequently 100 μL of each standard (0.1–2 ng/mL) was loaded into appropriate wells using polypropylene tips. A 100-μL volume of each media sample was then loaded in duplicate onto the plate and the plate was covered with the supplied lid and incubated overnight at 2°C to 8°C. Following overnight incubation, 1 mL of dimethylsulfoxide was added to the supplied p-aminophenylmercuric acetate (APMA) and vigorously vortexed to form a 1 M stock. A 5-μL dose of 1 M APMA stock was added to 10 mL of room-temperature assay buffer to create a 0.5 mM APMA stock. The supplied substrate solution was diluted with 5.1 mL of assay buffer, gently mixed, and stored on ice. The detection enzyme was allowed to thaw and kept on ice until required. All solutions were aspirated from the plate and four washes with wash buffer were performed ensuring all wells were completely filled and emptied during each washing step. To wells containing standards and samples where pro and active MMP2 was to be measured a 50-μL volume of 0.5 mM APMA was added. All other wells were loaded with 50 μL of assay buffer. Immediately before use, 100 μL of detection enzyme was added to the reconstituted substrate and mixed gently to create the detection reagent. All wells were then loaded with 50 μL of detection reagent and shaken on a microtiter titramax 100 plate shaker (Heidolph Instruments, Schwabach, Germany) for 20 seconds. The plate was then promptly read at 405 nm on a FLUOstar omega spectrophotometer plate reader (BMG Labtech, Aylesbury, UK) to obtain a t = 0 starting value for each well. The plate was subsequently covered with the supplied lid and incubated at 37°C for 6 hours then shaken again for 20 seconds on a microtiter titramax 100 plate shaker and once more read at 405 nm on an omega spectrophotometer plate reader to obtain a final value for each well. The t = 0 results were subtracted from the final plate reading (6 hours) and concentrations of active MMP2 determined from the standard curve results. 
Capsular Bag Analysis
The human capsular bag model is an advantage to PCO research because it provides the same cellular organization as in vivo and allows cells to grow on their natural matrix. 
Sham cataract operations were performed to create capsular bags from human donor lenses that were obtained with national research ethics committee approval and used in accordance with the tenets of the Declaration of Helsinki. A small rhexis was made in the center of the anterior lens capsule through which the fiber-cell mass of the lens was removed by hydro-expression. This left the entire “cell-free” posterior capsule and a ring of anterior capsule with its associated anterior lens epithelial cells. The capsular bag was then dissected from the globe and pinned onto 35-mm petri dishes (Corning Incorporated Life Sciences) using entomological pins as previously described. 17 All capsular bags contained similar starting populations of anterior lens epithelial cells. Capsular bags were subsequently maintained in nonsupplemented SF EMEM and 10 ng/mL TGFβ2 ± 20 μg/mL MMP2 neutralizing antibody for up to 21 days in a 35°C, 5% CO2 incubator. Experimental conditions were replenished every 3 to 4 days and ongoing observations were performed using a Nikon Eclipse TE200 phase-contrast microscope (Nikon Instruments Europe, Surrey, UK). End-point analysis of cell coverage on the previously cell-free posterior capsule was assessed with ImageJ analysis software (http://rsbweb.nih.gov/ij/). Using the area of the anterior capsulorhexis formed during the sham cataract operation as a control such that complete cell coverage of the posterior capsule as observed thorough the created anterior capsular rhexis was scored as 100% cell coverage, the effect of experimental conditions on the progression of anterior lens epithelial cells onto the posterior capsule could be evaluated. Wrinkling of the posterior capsule was quantified by transformation of phase contrast micrographs into binary images by defined thresholding techniques with Adobe Photoshop (Adobe Systems Europe Ltd., Berkshire, UK). Analysis of the black area of the binary images was then determined using ImageJ analysis software. 10  
Immunocytochemistry
Capsular bags were cultured in EMEM containing TGFβ2 (10 ng/mL) ± 20 μg/mL MMP2 neutralizing antibody until complete cell coverage on the posterior capsule had been achieved in the TGFβ2-treated capsular bags. At the end point (14–21 days), culture medium was aspirated from capsular bag culture dishes and the tissue was washed three times in quick succession with PBS. Subsequently capsular bags were fixed in 1.5 mL of 4% formaldehyde in PBS solution for 30 minutes. The capsular bags were then bisected, and each section was transferred to a new 35-mm petri dish and washed for a further three times with a solution of PBS containing 0.02% BSA and 0.05% Igepal (Sigma, Poole, Dorset, UK). The cell membranes were permeabilized in PBS containing 0.5% Triton-X100 for 30 minutes. A further three washes in PBS containing 0.02% BSA and 0.05% Igepal were completed for a time period of 15 minutes per wash on a plate shaker. Nonspecific cellular sites were blocked with normal goat serum (1:50) in 1% BSA in PBS, and incubated for 1 hour at 37°C. Capsular bags were then washed for 15 minutes with shaking in PBS containing 0.02% BSA and 0.05% Igepal, and this was repeated twice more. Following the final wash, 75 μL of the actin filament stain Texas Red X-Phalloidin (Molecular Probes, Eugene, OR) 1:100 in PBS containing 1% BSA was added to the cover-slips for 10 minutes in the dark at room temperature. The capsular bags were washed a further three times in PBS containing 0.02% BSA and 0.05% Igepal with shaking in the dark for 10-minute durations. Capsular bags sections were mounted on microscope slides (Chance Propper Ltd., Smethwick, West Midlands, UK) using Hydromount (National Diagnostics, Atlanta, GA) mounting solution and left to dry for 30 minutes at room temperature in the dark and were subsequently viewed under a Zeiss CCD Upright epifluorescent microscope and analyzed using Axiovision 4.1 imaging software (Carl Zeiss Ltd., Cambridge, UK). 
Statistical Analysis
Statistical differences were determined using Student's t-test or ANOVA (with Tukey). A P value of ≤ 0.05 was considered significant. 
Results
Broad-Spectrum MMP Inhibition by GM6001 Prevents TGFβ2-Induced Matrix Contraction
FHL-124 patch assays exposed to active TGFβ2 (10 ng/mL) for 3 days demonstrated the appearance of multiple contractile holes throughout the entire patch area (Fig. 1A); this was associated with a reduced stained area, such that patch area was decreased by 43.46% ± 5.52% in comparison to nontreated controls (Fig. 1B). This decrease in patch area was not attributable to losses in cell population by apoptosis or cell detachment from the culture dish (data not shown), but by matrix configuration and distribution changes in association with cells as previously determined by Dawes et al. 28 using periodic acid Schiff (PAS) or aniline blue collagen staining. The addition of the broad-spectrum MMP inhibitor GM6001 (25 μM) resulted in no changes in patch area when compared with untreated control (Figs. 1A, 1B). Furthermore, FHL-124 patches co-treated with GM6001 (25 μM) and TGFβ2 (10 ng/mL) for 3 days exhibited no contractile hole formation (Fig. 1A). Moreover, patch areas were comparable to untreated control dishes (Fig. 1B). Importantly, the patch area of dishes treated with GM6001 (25 μM) in the presence of TGFβ2 (10 ng/mL) was significantly greater than dishes treated with TGFβ2 (10 ng/mL) alone (Figs. 1A, 1B), indicating that GM6001 abolished the contractile ability of TGFβ2 and demonstrates a requirement for MMPs for this process. 
Figure 1. 
 
(A, B) Broad-spectrum MMP inhibitor GM6001 (25 μM) prevented active TGFβ2 (10 ng/mL)-induced matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 7). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2- and TGFβ2 + GM6001–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 1. 
 
(A, B) Broad-spectrum MMP inhibitor GM6001 (25 μM) prevented active TGFβ2 (10 ng/mL)-induced matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 7). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2- and TGFβ2 + GM6001–treated groups (P ≤ 0.05, ANOVA with Tukey test).
TGFβ2 Upregulates MMP2 and MT1-MMP Expression in FHL-124 Cells
It was important to determine which MMPs were expressed by FHL-124 cells and upregulated by active TGFβ2. Oligonucleotide microarray and quantitative real-time RT-PCR (TaqMan) analysis for the MMP and TIMP families showed that TGFβ2 (10 ng/mL) significantly increased mRNA expression of both MMP2 and MT1-MMP (MMP14) compared with controls (Figs. 2A, 2B). Additionally, microarray analysis also revealed that TGFβ2 (10 ng/mL) treatment led to increases in MT2-MMP (MMP15) and TIMP3 (Fig. 2A). While the TaqMan analysis examination also demonstrated a TGFβ2-mediated increase in MT6-MMP (MMP25) (Fig. 2B). 
Figure 2. 
 
(A, B) Oligonucleotide microarray and TaqMan analysis showing the effects of 24-hour active TGFβ2 (10 ng/mL) stimulation on MMP gene expression levels (n = 4). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, two-tailed t-test).
Figure 2. 
 
(A, B) Oligonucleotide microarray and TaqMan analysis showing the effects of 24-hour active TGFβ2 (10 ng/mL) stimulation on MMP gene expression levels (n = 4). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, two-tailed t-test).
ELISA analysis of culture media taken at endpoint from FHL-124 cells treated with active TGFβ2 (10 ng/mL) for 48 hours revealed that TGFβ2 significantly increased total MMP2 protein levels to 468.50% ± 101.34% when compared with untreated control media (Fig. 3A). Similarly, Western blot analysis demonstrated that TGFβ2 (10 ng/mL) stimulation caused MT1-MMP protein level to significantly increase to 195.32% ± 40.17% in comparison with control lysates (Figs. 3B, 3C). 
Figure 3. 
 
(AC) Active TGFβ2 (10 ng/mL) caused significant increases in MMP2 and MT1-MMP protein levels. Quantification by ELISA and Western blotting, respectively. Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, one-tailed t-test).
Figure 3. 
 
(AC) Active TGFβ2 (10 ng/mL) caused significant increases in MMP2 and MT1-MMP protein levels. Quantification by ELISA and Western blotting, respectively. Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, one-tailed t-test).
Validation of siRNA Knockdown of MMP2 and MT1-MMP
Evaluation of the TaqMan screen and microarray analysis presented MMP2 and MT1-MMP as targets for specific inhibition. FHL-124 cells were transfected with either MMP2- or MT1-MMP–specific siRNA for 24 hours, whilst siRNA for a random gene sequence with no specific target was used as an experimental control (Scramble siRNA) for both MMP analyses. 
Real-time PCR for MMP2 and MT1-MMP was performed to verify inhibition at the mRNA level. Treatment of FHL-124 cells with MMP2-targeted siRNA resulted in a 68.44% ± 1.63% decrease in MMP2 mRNA expression when compared to scramble siRNA controls (Fig. 4A). Specific knockdown of MT1-MMP using siRNA led to an 83.16% ± 1.76% decrease in MT1-MMP mRNA expression (Fig. 4B). Thus, specific inhibition of MMPs can be achieved using siRNA knockdown techniques. 
Figure 4. 
 
(AE) Validation of siRNA-targeted inhibition of MMP2 and MT1-MMP gene expression using QRT-PCR (24 hour) and protein determined by ELISA (MMP2) and Western blotting (MT1-MMP) (48 hour). Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between Scramble siRNA–treated and MMP2/MT1-MMP siRNA–treated groups (P ≤ 0.05, two-tailed t-test).
Figure 4. 
 
(AE) Validation of siRNA-targeted inhibition of MMP2 and MT1-MMP gene expression using QRT-PCR (24 hour) and protein determined by ELISA (MMP2) and Western blotting (MT1-MMP) (48 hour). Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between Scramble siRNA–treated and MMP2/MT1-MMP siRNA–treated groups (P ≤ 0.05, two-tailed t-test).
It was important to confirm that targeted siRNA knockdown also significantly reduced the protein levels of MMP2 and MT1-MMP. Therefore, FHL-124 cells were transfected with Scramble, MMP2, or MT1-MMP siRNA for 48 hours. Subsequently culture media were collected to assess MMP2 protein levels by ELISA. Protein extraction was performed to determine MT1-MMP protein by immunoblotting techniques. ELISA analysis of FHL-124 cells treated with MMP2-targeted siRNA resulted in a 71.73% ± 3.68% decrease in MMP2 protein levels when compared with Scramble-treated control cells (Fig. 4C). Immunoblot investigation of MT1-MMP siRNA transfected FHL-124 cells demonstrated an 89.23% ± 6.67% inhibition in MT1-MMP protein (Figs. 4D, 4E). 
Specific MMP2 Inhibition Suppresses TGFβ2-Induced Matrix Contraction
In order to assess the effects of MMP2 on TGFβ2-induced matrix contraction, FHL-124 cells were transfected with MMP2-targeted siRNA for 24 hours prior to active TGFβ2 (10 ng/mL) addition for a further 72 hours. Scramble siRNA–transfected patches treated with TGFβ2 (10 ng/mL) alone had a significant 21.91% ± 3.78% reduction in patch area compared with Scramble siRNA–transfected untreated controls (Fig. 5B). In cells treated with MMP2 siRNA, MMP2 knockdown had no effect on matrix contraction in the absence of TGFβ2 when compared to Scramble siRNA controls (Figs. 5A, 5B). Furthermore, the results clearly show that MMP2 knockdown cells exposed to active TGFβ2 (10 ng/mL) did not develop matrix holes and did not significantly differ from control cells (Figs. 5A, 5B). 
Figure 5. 
 
(AD) MMP2 inhibition using siRNA and neutralizing antibody (4 μg/mL) prevents active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 8 MMP2 siRNA and n = 3 MMP2 neutralizing antibody). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2 and TGFβ2 + MMP2 inhibition groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 5. 
 
(AD) MMP2 inhibition using siRNA and neutralizing antibody (4 μg/mL) prevents active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 8 MMP2 siRNA and n = 3 MMP2 neutralizing antibody). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2 and TGFβ2 + MMP2 inhibition groups (P ≤ 0.05, ANOVA with Tukey test).
In additional experiments, MMP2 activation was prevented with a specific MMP2 neutralizing antibody that inhibits MMP2 pro-domain cleavage. FHL-124 patches were pretreated for 15 minutes with 4 μg/mL MMP2 antibody prior to active TGFβ2 (10 ng/mL) stimulation for 72 hours. No significant effect on contraction was observed with addition of MMP2 antibody (4 μg/mL) alone compared to untreated control cells (Fig. 5C). Stimulation with 10 ng/mL TGFβ2 led to a significant 19.02% ± 2.79% decrease in patch area when compared to untreated control (Figs. 5C, 5D). Cells pretreated with MMP2 antibody (4 μg/mL) and subsequently exposed to 10 ng/mL TGFβ2 did not display significant matrix contraction, such that they did not significantly differ from untreated controls (Figs. 5C, 5D) but did show a significant inhibition of the TGFβ2-induced matrix contraction (Figs. 5C, 5D). 
Specific Suppression of MT1-MMP Does Not Prevent TGFβ2-Induced Matrix Contraction
Since MT1-MMP is considered to be the principal activator of MMP2, loss of MT1-MMP would be expected to prohibit active MMP2 generation and thus once more prevent TGFβ2-mediated matrix contraction. MT1-MMP mRNA and protein expression were suppressed using targeted siRNA for 24 hours prior to active TGFβ2 (10 ng/mL) addition for a further 72 hours; cells transfected with Scramble siRNA were once more used as an experimental control. 
FHL-124 cells transfected with Scramble siRNA showed no matrix contraction in the end-point analysis (Figs. 6A, 6B), in contrast Scramble siRNA–transfected cells exposed to TGFβ2 for 72 hours exhibited a marked reduction of 21.59% ± 6.54% in patch area (Fig. 6B) when compared to Scramble siRNA controls. However, in contrast to the results seen with knockdown of MMP2, when MT1-MMP siRNA–treated cells were exposed to TGFβ2 for 72 hours, matrix contraction was still evident with a significant 38.69 % ± 12.27% decrease in patch area when compared to Scramble siRNA controls (Fig. 6B). Thus these data indicate that suppression of MT1-MMP did not impede TGFβ2-controlled matrix contraction. 
Figure 6. 
 
(A, B) MT1-MMP inhibition using siRNA did not prevent active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 5). *Significant difference between control and TGFβ2+MT1-MMP siRNA–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 6. 
 
(A, B) MT1-MMP inhibition using siRNA did not prevent active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 5). *Significant difference between control and TGFβ2+MT1-MMP siRNA–treated groups (P ≤ 0.05, ANOVA with Tukey test).
TGFβ2-Induced Pro-MMP2 Activation Is Independent of MT1-MMP
The persistence of the contractile effects of TGFβ2 when MT1-MMP levels were significantly reduced suggests that MMP2 activation was independent of MT1-MMP. To test this notion, FHL-124 cells were transfected with either MT1-MMP targeted or Scramble siRNA for 24 hours prior to addition of 10 ng/mL active TGFβ2 for a further 48 hours. Experimental bathing media were collected at the end of the experimental period and analyzed for active MMP2 on a Biotrak microtiter plate (GE Healthcare). TGFβ2 (10 ng/mL) significantly increased active MMP2 levels to 220.9% ± 30.8% in culture media from cells treated with Scramble siRNA when compared with cells treated with Scramble siRNA alone (Fig. 7). Addition of 10 ng/mL active TGFβ2 to MT1-MMP siRNA–treated cells also resulted in a significant elevation of active MMP2 to 203.1% ± 58.1% relative to MT1-MMP knockdown controls (Fig. 7). 
Figure 7. 
 
MT1-MMP inhibition by targeted siRNA does not prevent MMP2 activation by active TGFβ2 (10 ng/mL) assessed by Biotrak analysis. Data expressed as mean ± SEM (n = 3). *Significant difference between control and TGFβ2 treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 7. 
 
MT1-MMP inhibition by targeted siRNA does not prevent MMP2 activation by active TGFβ2 (10 ng/mL) assessed by Biotrak analysis. Data expressed as mean ± SEM (n = 3). *Significant difference between control and TGFβ2 treated groups (P ≤ 0.05, ANOVA with Tukey test).
MMP2 Inhibition Prevents TGFβ2-Induced Matrix Contraction in Human Lens Capsular Bags
Sham cataract operations were carried out on human donor eyes, which generated pairs of capsular bags from a single donor. Capsular bags were exposed to 10 ng/mL active TGFβ2 ± 20 μg/mL MMP2 neutralizing antibody (14–21 days). End-point examination of the posterior capsule of capsular bags stimulated with TGFβ2 (10 ng/mL) revealed a significant coverage of cells across the previously cell-free posterior capsule (Figs. 8A, 8C, 8E). These cells appeared elongated and demonstrated a myofibroblastic phenotype; moreover, extensive wrinkle formation was observed on the posterior capsule (Figs. 8A, 8C). Immunocytochemistry revealed F-actin was assembled into stress fibers (Fig. 8E). In contrast the posterior capsule of capsular bags co-treated with TGFβ2 (10 ng/mL) and 20 μg/mL MMP2 neutralizing antibody demonstrated significantly less cell coverage with only 21.76% ± 13.66% of the observed posterior capsule area containing cells, compared with 92.92% ± 7.08% in the capsular bags treated with TGFβ2 alone (Figs. 8B, 8D, 8F, 9A). Significantly, capsular bags co-treated with TGFβ2 (10 ng/mL) and 20 μg/mL MMP2 neutralizing antibody demonstrated a significant reduction in capsular wrinkling compared with TGFβ2-treated capsular bags, such that matrix wrinkling was inhibited by 88.95% ± 3.87% (Fig. 9B). In preparations treated with MMP2 neutralizing antibody, cells retained an epithelial morphology (Figs. 8B, 8D) with “cobblestone” cortical distribution of F-actin and an absence of stress filaments (Fig. 8F). 
Figure 8. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) (B, D, F) prevents active TGFβ2 (10 ng/mL)-induced (A, C, E) formation of wrinkles and actin stress fibers in human capsular bags (n = 3). Images captured at the end point.
Figure 8. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) (B, D, F) prevents active TGFβ2 (10 ng/mL)-induced (A, C, E) formation of wrinkles and actin stress fibers in human capsular bags (n = 3). Images captured at the end point.
Figure 9. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) reduces active TGFβ2-mediated (10 ng/mL) cell coverage (A) and wrinkling (B) of the posterior capsule of human capsular bags assessed by ImageJ analysis of phase contrast images captured at the end point (n = 3). *Significant difference between TGFβ2-treated and TGFβ2 + anti-MMP2–treated groups (P ≤ 0.05, two-tailed t-test).
Figure 9. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) reduces active TGFβ2-mediated (10 ng/mL) cell coverage (A) and wrinkling (B) of the posterior capsule of human capsular bags assessed by ImageJ analysis of phase contrast images captured at the end point (n = 3). *Significant difference between TGFβ2-treated and TGFβ2 + anti-MMP2–treated groups (P ≤ 0.05, two-tailed t-test).
Discussion
The work presented here indicates that the pro-fibrotic cytokine TGFβ2 stimulates a significant increase in production of the matrix modulators MMP2 and MT1-MMP (MMP14) in FHL-124 cells. Furthermore, we established that inhibition of MMP2 activity alone was sufficient to prevent TGFβ2-induced matrix contraction and consequent wrinkling of the human posterior lens capsule, suggesting that MMP2 is critical in TGFβ-mediated fibrotic events. 
TGFβ is regarded as the central mediator of fibrosis 1 and is implicated in the progression of multiple fibrotic disorders such as renal fibrosis, 32 idiopathic pulmonary fibrosis, 33 and systemic sclerosis. 34 In the human lens capsular bag model, transdifferentiation and matrix production/contraction were significantly enhanced by active TGFβ, resulting in amplification of the light scatter due to wrinkle formation that is the hallmark of PCO. 18 TGFβ is able to generate excessive amounts of multiple matrix components including collagens, fibronectin, and integrins. 29,35 Furthermore, addition of active TGFβ2 to both human capsular bag models and FHL-124 cells has been shown to result in autocrine production of TGFβ210,29 and can also synthesize many of its own activators including thrombospondin-1 (TSP-1), furin, chymase, αVβ6 integrin, and, important to our research, MMPs. 36  
Active MMP2 is capable of cleaving ECM components such as collagen type IV, exposing sites of attachment that can promote cell migration. 37 This is significant because collagen type IV is a main structural component of the lens capsule that may be modified during PCO development. 38,39 Furthermore, the ECM is acknowledged as an effective store for large quantities of inactive TGFβ, 40 and MMP2 can proteolytically release TGFβ from its inactive ECM-bound complex. 14 Active MMP2 cleavage of the latent TGFβ binding protein that tethers TGFβ to the ECM allows inactive TGFβ to be activated by removal of the latency-associated protein (LAP) by proteins such as TSP-1 and furin, 15,41 which are associated with inflammatory responses. In addition, proteolytic cleavage by active MMP2 can release the active TGFβ from LAP itself. 36  
MMP2 is secreted as an inactive pro-form, and activation of pro-MMP2 is proposed to occur via the membrane tethered MMP, MT1-MMP. 4244 Interestingly, in this study attenuation of MT1-MMP did not prevent TGFβ2-stimulated matrix contraction or active MMP2 production, signifying that TGFβ2-mediated pro-form MMP2 activation occurs independently of MT1-MMP. 
Moreover, real-time PCR and oligonucleotide analysis also revealed no significant increase in TIMP2 following TGFβ stimulation. TIMP2 is required along with MT1-MMP to activate pro-form MMP2. 45 Our data indicate that despite TGFβ2 addition to FHL-124 cells increasing MMP2 and MT1-MMP expression it did not augment TIMP2 gene expression leading us to hypothesize that an MT1-MMP/TIMP2 dependent activation of pro-form MMP2 is not required in the lens. Interestingly, our results did show an increase in the natural inhibitor of MT1-MMP (i.e., TIMP3), indicating the turnover of MT1-MMP is still under strict control. 
One remaining question therefore is how active MMP2 is generated in lens epithelial cells. Oligonucleotide analysis revealed that following active TGFβ2 exposure, increases in MT2-MMP were observed. Additionally, TaqMan studies indicated increased production of MT6-MMP following active TGFβ2 stimulation. MT2-MMP and MT6-MMP like MT1-MMP are membrane type MMPs. The role of MT6-MMP remains to be determined, but MT2-MMP is expressed in human anterior lens epithelium along with MT1-MMP, MT3-MMP, MT4-MMP, and MT5-MMP. 24 A study by Morrison et al. 46 using TIMP2 knockout mice fibroblast cells revealed that heightened levels of MT2-MMP were sufficient to activate pro-form MMP2 despite the lack of TIMP2. This was dependent on MMP2 binding to the cell surface via the hemopexin domain most likely by association with integrin αvβ6. This analysis surmised that MT1-MMP required a certain threshold of TIMP2 to be capable of activating the pro-form MMP2, and in conditions of very low TIMP2 expression MT1-MMP was unable to function effectively. Moreover, Morrison et al. 46 concluded that MT2-MMP has no need for TIMP2 because it was able to efficiently activate pro-form MMP2 in the TIMP2 knockout fibroblast cells. 46 Furthermore, MT2-MMP activation of pro-form MMP2 occurred much more rapidly than by MT1-MMP. This could be due to the presence of an RGD recognition sequence in the MT2-MMP hemopexin domain allowing MT2-MMP to adhere to integrins and form a receptor complex with similarly bound MMP2. Additionally, in MT1-MMP−/− human skin fibroblasts, Ruangpanit et al. 47 showed that although a reduction in MMP2 activation was observed in MT1-MMP null cells compared with wild-type fibroblasts, a significant residual amount of non–MT1-MMP-controlled activation of MMP2 still occurred. 
Our human lens capsular bag model illustrated that addition of active TGFβ2 following sham cataract operations resulted in anterior lens epithelial cell migration, matrix contraction, and wrinkle formation. These events are hallmarks of PCO. 3,11 However, the addition of an MMP2 inhibitor not only prevented wrinkle formation but cells found on the posterior capsule in capsular bags treated with TGFβ2 and MMP2 neutralizing antibody exhibited a morphology characteristic of anterior cuboidal lens epithelial cells. Furthermore, progression of anterior lens epithelial cells exposed to TGFβ2 and MMP2 antibody on to the posterior capsule was attenuated compared to capsular bags treated with TGFβ2 alone. These findings are comparable to a study by Wong et al. 48 in which donor human capsular bag cultures were treated with 10% serum supplemented media in the presence and absence of the broad-spectrum MMP inhibitor GM6001 . Following GM6001 (100 μM) treatment for 10 days a 5-fold difference in cell migration was observed compared to controls, along with minimal capsular wrinkling. By day 15 only 5% capsular contraction was exhibited in GM6001-treated cultures, whilst a 60% capsular contraction was detected in cultures maintained in control conditions. 48 Additionally, Seomun et al. 49 determined that human lens epithelial cells (HLE B3s) treated with (2R)-2-[(4-biphenylylsulfonyl)amino]-3-phenylpropionic acid (a MMP2/9 inhibitor) in the presence of TGFβ1 prevented the elongated and scattered cell morphology associated with TGFβ1 treatment. Furthermore, strong immunoreactivity for αSMA and MMP2 was correlated with TGFβ1-induced anterior subcapsular plaque formation. 49 Our findings could be due to several factors that are influenced by MMP2. The ECM is an excellent store of bound growth factors including TGFβ, 35,36 along with other growth factors such as fibroblast growth factor (FGF) 50 and insulin growth factor (IGF). 51 This accrual of growth factors can provide a considerable reserve of bioavailable cytokines under appropriate conditions, which can exacerbate TGFβ effects in the capsular bag. For example, in a whole rat-lens study of anterior subcapsular plaque formation, FGF was found to significantly enhance the effects of TGFβ-induced ASC; however, FGF alone was incapable of instigating plaque formation. 52 FGF is, however, an inducer of lens cell proliferation. 53,54 In our capsular bags treated with TGFβ2 and MMP2 inhibitor, the lack of active MMP2 production could also be inhibiting FGF release from the collagenous capsule, thus preventing subsequent stimulation of anterior lens cell proliferation and migration. 
Interestingly, there were morphological differences between cells residing on the capsular bags in the presence or absence of MMP2 neutralizing antibody. Phase-contrast and immunocytochemical analysis of the TGFβ2-treated capsular bags clearly showed that advancing anterior lens epithelial cells had an elongated myofibroblastic morphology; furthermore, distinct cell-free spaces could be identified between the extended cells. This morphology was not observed in the capsular bags exposed to TGFβ2 and MMP2 inhibitor, in which the cells retained tight cell to cell connections and a cuboidal epithelial cell structure. An investigation into ASC formation in excised rat lenses by Dwivedi et al. 23 showed that the bathing media of lenses treated with TGFβ2 had significant amounts of E-cadherin fragments caused by E-cadherin shedding. The production of these fragments was significantly attenuated by the application of MMP inhibitors {GM6001 and MMP2/9 inhibitor (2R)-[(4biphenylylsufonyl)amino]-N-hydroxyl-3-phenylpropionamide)}. Importantly, E-cadherins are involved in cell to cell adhesion and thus the study by Dwivedi et al strengthens our premise that MMP2 could be involved in loss of cell to cell adhesion in TGFβ2-treated capsular bags. 
The current study has determined a critical role for MMP2 in TGFβ-controlled matrix contraction and PCO development. We therefore propose that MMP2 is an excellent target for new approaches to PCO prevention. An important consideration is how one can deliver an effective agent. With respect to PCO there are several options. An inhibitor of MMP2 could take the form of a pharmacological antagonist, siRNA, or a neutralizing antibody. In each case the agent would have to be delivered at the time of surgery and localized to the lens cell population. Drug delivery systems are in place that can seal the opening in the lens capsular bag created by surgery and allow an agent to be delivered to lens cells for a predetermined period and then be removed. 56 This approach reduces the risk of affecting other ocular tissues and has been used clinically. Another approach is to modify an artificial IOL that is inserted in to the capsular bag during surgery to restore visual power. This drug delivery method has been used in the past for a number of molecules. 57,58 Interestingly, the effect of the MMP2 inhibitors GM6001, MMP2/9 inhibitor I, and MMP2/9 inhibitor II on ocular cell migration and viability were tested, along with long-term release rates from silicon-coated disks (a material commonly used to make IOLs). 59 Silicon disks coated with high concentrations of MMP2/9 inhibitor II reduced migration of FHL-124 cells by > 42%. Importantly after an initial release of up to 25% of inhibitor from silicon-coated disks there was a steady and long-lasting release for a further 144 days. 59 Such strategies therefore could deliver an inhibitor of MMP2 to lens cells for a sufficient period of time to reduce PCO development. 
Acknowledgments
The authors thank Diane Alden for technical assistance and Pamela Keeley, Debbie Busby, Sam Major, and Mary Tottman at the East Anglian Eye Bank for their invaluable contribution to our studies. 
References
Leask A Abraham DJ . TGF-beta signaling and the fibrotic response. FASEB J . 2004;18:816–827. [CrossRef] [PubMed]
Wormstone IM . Posterior capsule opacification: a cell biological perspective. Exp Eye Res . 2002;74:337–347. [CrossRef] [PubMed]
Eldred JA Dawes LJ Wormstone IM . The lens as a model for fibrotic disease. Philos Trans R Soc Lond B Biol Sci . 2011;366:1301–1319. [CrossRef] [PubMed]
van Bree MCJ van der Meulen IJE Franssen L Imaging of forward light-scatter by opacified posterior capsules isolated from pseudophakic donor eyes. Invest Ophthalmol Vis Sci . 2011;52:5587–5597. [CrossRef] [PubMed]
Resnikoff S Pascolini D Etya'ale D Global data on visual impairment in the year 2002. Bull WHO . 2004;82:844–851. [PubMed]
Taylor HR . Cataract: how much surgery do we have to do? Br J Ophthalmol . 2000;84:1–2. [CrossRef] [PubMed]
Laurell CG Zetterstrom C . Effects of dexamethasone, diclofenac, or placebo on the inflammatory response after cataract surgery. Br J Ophthalmol . 2002;86:1380–1384. [CrossRef] [PubMed]
Moisseiev J Bartov E Schochat A Blumenthal M . Long-term study of the prevalence of capsular opacification following extracapsular cataract-extraction. J Cataract Refract Surg . 1989;15:531–533. [CrossRef] [PubMed]
Sime PJ O'Reilly KMA . Fibrosis of the lung and other tissues: new concepts in pathogenesis and treatment. Clin Immunol . 2001;99:308–319. [CrossRef] [PubMed]
Wormstone IM Anderson IK Eldred JA Dawes LJ Duncan G . Short-term exposure to transforming growth factor beta induces long-term fibrotic responses. Exp Eye Res . 2006;83:1238–1245. [CrossRef] [PubMed]
Wormstone IM Wang LX Liu CSC . Posterior capsule opacification. Exp Eye Res . 2009;88:257–269. [CrossRef] [PubMed]
Schlotzer-Schrehardt U Zenkel M Kuchle M Sakai LY Naumann GO . Role of transforming growth factor-beta1 and its latent form binding protein in pseudoexfoliation syndrome. Exp Eye Res . 2001;73:765–780. [CrossRef] [PubMed]
Ohta K Yamagami S Taylor AW Streilein JW . IL-6 antagonizes TGF-beta and abolishes immune privilege in eyes with endotoxin-induced uveitis. Invest Ophthalmol Vis Sci . 2000;41:2591–2599. [PubMed]
Yu Q Stamenkovic I . Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF-beta and promotes tumor invasion and angiogenesis. Genes Dev . 2000;14:163–176. [PubMed]
Frazier WA . Thrombospondins. Curr Opin Cell Biol . 1991;3:792–799. [CrossRef] [PubMed]
Wallentin N Wickstrom K Lundberg C . Effect of cataract surgery on aqueous TGF-beta and lens epithelial cell proliferation. Invest Ophthalmol Vis Sci . 1998;39:1410–1418. [PubMed]
Liu CSC Wormstone IM Duncan G Marcantonio JM Webb SF Davies PD . A study of human lens cell growth in vitro—a model for posterior capsule opacification. Invest Ophthalmol Vis Sci . 1996;37:906–914. [PubMed]
Wormstone IM Tamiya S Anderson I Duncan G . TGF-beta 2-induced matrix modification and cell transdifferentiation in the human lens capsular bag. Invest Ophthalmol Vis Sci . 2002;43:2301–2308. [PubMed]
Hales AM Schulz MW Chamberlain CG McAvoy JW . TGF-beta-1 induces lens cells to accumulate alpha-smooth muscle actin, a marker for subcapsular cataracts. Curr Eye Res . 1994;13:885–890. [CrossRef] [PubMed]
Wride MA Geatrell J Guggenheim JA . Proteases in eye development and disease. Birth Defects Res C Embryo Today . 2006;78:90–105. [CrossRef] [PubMed]
Visse R Nagase H . Matrix metalloproteinases and tissue inhibitors of metalloproteinases - Structure, function, and biochemistry. Circ Res . 2003;92:827–839. [CrossRef] [PubMed]
Mandal M Mandal A Das S Chakraborti T Chakraborti S . Clinical implications of matrix metalloproteinases. Mol Cell Biochem . 2003;252:305–329. [CrossRef] [PubMed]
Dwivedi DJ Pino G Banh A Matrix metalloproteinase inhibitors suppress transforming growth factor-beta-induced subcapsular cataract formation. Am J Pathol . 2006;168:69–79. [CrossRef] [PubMed]
Hodgkinson LM Duncan G Wang LX Pennington CJ Edwards DR Wormstone IM. MMP . and TIMP expression in quiescent, dividing, and differentiating human lens cells. Invest Ophthalmol Vis Sci . 2007;48:4192–4199. [CrossRef] [PubMed]
Seiki M . Membrane-type matrix metalloproteinases. APMIS . 1999;107:137–143. [CrossRef] [PubMed]
Kandasamy AD Chow AK Ali MA Schulz R . Matrix metalloproteinase-2 and myocardial oxidative stress injury: beyond the matrix. Cardiovasc Res . 2010;85:413–423. [CrossRef] [PubMed]
Wormstone IM Tamiya S Eldred JA Characterisation of TGF-beta 2 signalling and function in a human lens cell line. Exp Eye Res . 2004;78:705–714. [CrossRef] [PubMed]
Dawes LJ Eldred JA Anderson IK TGF beta-induced contraction is not promoted by fibronectin-fibronectin receptor interaction, or alpha SMA expression. Invest Ophthalmol Vis Sci . 2008;49:650–661. [CrossRef] [PubMed]
Dawes LJ Elliott RM Reddan JR Wormstone YM Wormstone IM . Oligonucleotide microarray analysis of human lens epithelial cells: TGF beta regulated gene expression. Mol Vis . 2007;13:1181–1197. [PubMed]
Nuttall RK Pennington CJ Taplin J Elevated membrane-type matrix metalloproteinases in gliomas revealed by profiling proteases and inhibitors in human cancer cells. Mol Cancer Res . 2003;1:333–345. [PubMed]
Daub H Wallasch C Lankenau A Herrlich A Ullrich A . Signal characteristics of G protein-transactivated EGF receptor. EMBO J . 1997;16:7032–7044. [CrossRef] [PubMed]
Bottinger EP Bitzer M . TGF-beta signaling in renal disease. J Am Soc Nephrol . 2002;13:2600–2610. [CrossRef] [PubMed]
Khalil N O'Connor R . Idiopathic pulmonary fibrosis: current understanding of the pathogenesis and the status of treatment. Can Med Assoc J . 2004;171:153–160. [CrossRef]
Krieg T Takehara K . Skin disease: a cardinal feature of systemic sclerosis. Rheumatology . 2009;48:14–18. [CrossRef]
Roberts AB McCune BK Sporn MB . TGF-beta-regulation of extracellular-matrix. Kidney Int . 1992;41:557–559. [CrossRef] [PubMed]
Jenkins G . The role of proteases in transforming growth factor-beta activation. Int J Biochem Cell Biol . 2008;40:1068–1078. [CrossRef] [PubMed]
Cheng S Pollock AS Mahimkar R Olson JL Lovett DH . Matrix metalloproteinase 2 and basement membrane integrity: a unifying mechanism for progressive renal injury. FASEB J . 2006;20:1898–1900. [CrossRef] [PubMed]
Dehm P Kefalides NA . The collagenous component of lens basement membrane. The isolation and characterization of an alpha chain size collagenous peptide and its relationship to newly synthesized lens components. J Biol Chem . 1978;253:6680–6686. [PubMed]
Wederell ED de Iongh RU . Extracellular matrix and integrin signaling in lens development and cataract. Semin Cell Dev Biol . 2006;17:759–776. [CrossRef] [PubMed]
Annes JP Munger JS Rifkin DB . Making sense of latent TGF beta activation. J Cell Sci . 2003;116:217–224. [CrossRef] [PubMed]
Munger JS Harpel JG Gleizes PE Mazzieri R Nunes I Rifkin DB . Latent transforming growth factor-beta: structural features and mechanisms of activation. Kidney Int . 1997;51:1376–1382. [CrossRef] [PubMed]
Page-McCaw A Ewald AJ Werb Z . Matrix metalloproteinases and the regulation of tissue remodelling. Nat Rev Mol Cell Biol . 2007;8:221–233. [CrossRef] [PubMed]
Boyd PJ Doyle J Gee E Pallan S Haas TL . MAPK signaling regulates endothelial cell assembly into networks and expression of MT1-MMP and MMP-2. Am J Physiol Cell Physiol . 2005;288:C659–C668. [CrossRef] [PubMed]
Hernandez-Barrantes S Toth M Bernardo MM Binding of active (57 kDa) membrane type 1-matrix metalloproteinase (MT1-MMP) to tissue inhibitor of metalloproteinase (TIMP)-2 regulates MT1-MMP processing and pro-MMP-2 activation. J Biol Chem . 2000;275:12080–12089. [CrossRef] [PubMed]
Wang ZP Juttermann R Soloway PD . TIMP-2 is required for efficient activation of proMMP-2 in vivo. J Biol Chem . 2000;275:26411–26415. [CrossRef] [PubMed]
Morrison CJ Butler GS Bigg HF Roberts CR Soloway PD Overall CM . Cellular activation of MMP-2 (gelatinase A) by MT2-MMP occurs via a TIMP-2-independent pathway. J Biol Chem . 2001;276:47402–47410. [CrossRef] [PubMed]
Ruangpanit N Price JT Holmbeck K MT1-MMP-dependent and -independent regulation of gelatinase a activation in long-term, ascorbate-treated fibroblast cultures: regulation by fibrillar collagen. Exp Cell Res . 2002;272:109–118. [CrossRef] [PubMed]
Wong TT Daniels JT Crowston JG Khaw PT . MMP inhibition prevents human lens epithelial cell migration and contraction of the lens capsule. Br J Ophthalmol . 2004;88:868–872. [CrossRef] [PubMed]
Seomun Y Kim J Lee EH Joo CK . Overexpression of matrix metalloproteinase-2 mediates phenotypic transformation of lens epithelial cells. Biochem J . 2001;358:41–48. [CrossRef] [PubMed]
Tholozan FMD Gribbon C Li Z IFGF-2 release from the lens capsule by MMP-2 maintains lens epithelial cell viability. Mol Biol Cell . 2007;18:4222–4231. [CrossRef] [PubMed]
Fowlkes JL Enghild JJ Suzuki K Nagase H . Matrix metalloproteinases degrade insulin-like growth factor-binding protein-3 in dermal fibroblast-cultures. J Biol Chem . 1994;269:25742–25746. [PubMed]
Cerra A Mansfield KJ Chamberlain CG . Exacerbation of TGF-beta-induced cataract by FGF-2 in cultured rat lenses. Mol Vis . 2003;9:689–700. [PubMed]
Ishida I Saika S Okada Y Ohnishi Y . Growth factor deposition in anterior subcapsular cataract. J Cataract Refract Surg . 2005;31:1219–1225. [CrossRef] [PubMed]
Wormstone IM Del Rio-Tsonis K McMahon G FGF: an autocrine regulator of human lens cell growth independent of added stimuli. Invest Ophthalmol Vis Sci . 2001;42:1305–1311. [PubMed]
Maloof AJ Pandey SK Neilson G Milverton EJ . Selective death of lens epithelial cells using demineralized water and Triton X-100 with PerfectCapsule sealed capsule irrigation: a histological study in rabbit eyes. Arch Ophthalmol . 2005;123:1378–1384. [CrossRef] [PubMed]
Duncan G Wang L Neilson GJ Wormstone IM . Lens cell survival after exposure to stress in the closed capsular bag. Invest Ophthalmol Vis Sci . 2007;48:2701–2707. [CrossRef] [PubMed]
Behar-Cohen FF David T D'Hermies F In vivo inhibition of lens regrowth by fibroblast growth factor 2-saporin. Invest Ophthalmol Vis Sci . 1995;36:2434–2448. [PubMed]
Duncan G Wormstone IM Liu CS Marcantonio JM Davies PD . Thapsigargin-coated intraocular lenses inhibit human lens cell growth. Nat Med . 1997;3:1026–1028. [CrossRef] [PubMed]
Morarescu D West-Mays JA Sheardown HD . Effect of delivery of MMP inhibitors from PDMS as a model IOL material on PCO markers. Biomaterials . 2010;31:2399–2407. [CrossRef] [PubMed]
Footnotes
 Supported by The Dunhill Medical Trust, The Humane Research Trust, and the BBSRC.
Footnotes
 Disclosure: J.A. Eldred, P; L.M. Hodgkinson, P; L.J. Dawes, P; J.R. Reddan, None; D.R. Edwards, P; I.M. Wormstone, P
Figure 1. 
 
(A, B) Broad-spectrum MMP inhibitor GM6001 (25 μM) prevented active TGFβ2 (10 ng/mL)-induced matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 7). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2- and TGFβ2 + GM6001–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 1. 
 
(A, B) Broad-spectrum MMP inhibitor GM6001 (25 μM) prevented active TGFβ2 (10 ng/mL)-induced matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 7). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2- and TGFβ2 + GM6001–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 2. 
 
(A, B) Oligonucleotide microarray and TaqMan analysis showing the effects of 24-hour active TGFβ2 (10 ng/mL) stimulation on MMP gene expression levels (n = 4). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, two-tailed t-test).
Figure 2. 
 
(A, B) Oligonucleotide microarray and TaqMan analysis showing the effects of 24-hour active TGFβ2 (10 ng/mL) stimulation on MMP gene expression levels (n = 4). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, two-tailed t-test).
Figure 3. 
 
(AC) Active TGFβ2 (10 ng/mL) caused significant increases in MMP2 and MT1-MMP protein levels. Quantification by ELISA and Western blotting, respectively. Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, one-tailed t-test).
Figure 3. 
 
(AC) Active TGFβ2 (10 ng/mL) caused significant increases in MMP2 and MT1-MMP protein levels. Quantification by ELISA and Western blotting, respectively. Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between control and TGFβ2-treated groups (P ≤ 0.05, one-tailed t-test).
Figure 4. 
 
(AE) Validation of siRNA-targeted inhibition of MMP2 and MT1-MMP gene expression using QRT-PCR (24 hour) and protein determined by ELISA (MMP2) and Western blotting (MT1-MMP) (48 hour). Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between Scramble siRNA–treated and MMP2/MT1-MMP siRNA–treated groups (P ≤ 0.05, two-tailed t-test).
Figure 4. 
 
(AE) Validation of siRNA-targeted inhibition of MMP2 and MT1-MMP gene expression using QRT-PCR (24 hour) and protein determined by ELISA (MMP2) and Western blotting (MT1-MMP) (48 hour). Data expressed as mean ± SEM (n = 4 MMP2 and n = 3 MT1-MMP). *Significant difference between Scramble siRNA–treated and MMP2/MT1-MMP siRNA–treated groups (P ≤ 0.05, two-tailed t-test).
Figure 5. 
 
(AD) MMP2 inhibition using siRNA and neutralizing antibody (4 μg/mL) prevents active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 8 MMP2 siRNA and n = 3 MMP2 neutralizing antibody). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2 and TGFβ2 + MMP2 inhibition groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 5. 
 
(AD) MMP2 inhibition using siRNA and neutralizing antibody (4 μg/mL) prevents active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 8 MMP2 siRNA and n = 3 MMP2 neutralizing antibody). *Significant difference between control and TGFβ2-treated groups; #Significant difference between TGFβ2 and TGFβ2 + MMP2 inhibition groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 6. 
 
(A, B) MT1-MMP inhibition using siRNA did not prevent active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 5). *Significant difference between control and TGFβ2+MT1-MMP siRNA–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 6. 
 
(A, B) MT1-MMP inhibition using siRNA did not prevent active TGFβ2-induced (10 ng/mL) matrix contraction assessed by patch assay. Data expressed as mean ± SEM (n = 5). *Significant difference between control and TGFβ2+MT1-MMP siRNA–treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 7. 
 
MT1-MMP inhibition by targeted siRNA does not prevent MMP2 activation by active TGFβ2 (10 ng/mL) assessed by Biotrak analysis. Data expressed as mean ± SEM (n = 3). *Significant difference between control and TGFβ2 treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 7. 
 
MT1-MMP inhibition by targeted siRNA does not prevent MMP2 activation by active TGFβ2 (10 ng/mL) assessed by Biotrak analysis. Data expressed as mean ± SEM (n = 3). *Significant difference between control and TGFβ2 treated groups (P ≤ 0.05, ANOVA with Tukey test).
Figure 8. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) (B, D, F) prevents active TGFβ2 (10 ng/mL)-induced (A, C, E) formation of wrinkles and actin stress fibers in human capsular bags (n = 3). Images captured at the end point.
Figure 8. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) (B, D, F) prevents active TGFβ2 (10 ng/mL)-induced (A, C, E) formation of wrinkles and actin stress fibers in human capsular bags (n = 3). Images captured at the end point.
Figure 9. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) reduces active TGFβ2-mediated (10 ng/mL) cell coverage (A) and wrinkling (B) of the posterior capsule of human capsular bags assessed by ImageJ analysis of phase contrast images captured at the end point (n = 3). *Significant difference between TGFβ2-treated and TGFβ2 + anti-MMP2–treated groups (P ≤ 0.05, two-tailed t-test).
Figure 9. 
 
MMP2 inhibition using neutralizing antibody (20 μg/mL) reduces active TGFβ2-mediated (10 ng/mL) cell coverage (A) and wrinkling (B) of the posterior capsule of human capsular bags assessed by ImageJ analysis of phase contrast images captured at the end point (n = 3). *Significant difference between TGFβ2-treated and TGFβ2 + anti-MMP2–treated groups (P ≤ 0.05, two-tailed t-test).
×
×

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.

×