|Molecular Vision 2005;
Received 20 January 2005 | Accepted 21 September 2005 | Published 23 September 2005
Influence of cyclical mechanical strain on extracellular matrix gene expression in human lamina cribrosa cells in vitro
Ruaidhrí P. Kirwan,1
Cecilia H. Fenerty,2 John
Crean,1 Robert J. Wordinger,3
Abbot F. Clark,4
Colm J. O'Brien1,2
1Conway Institute of Biomolecular and Biomedical Research University College Dublin, Dublin, Ireland; 2Institute of Ophthalmology, Mater Misericordiae University Hospital, Dublin, Ireland; 3Department of Cell Biology and Genetics, University of North Texas Health Center at Fort Worth, Fort Worth, TX; 4Glaucoma Research, Alcon Research Ltd., Fort Worth, TX
Correspondence to: Dr. Ruaidhrí Kirwan, Institute of Ophthalmology, Mater Misericordiae University Hospital, No. 60 Eccles Street, Dublin 7, Ireland; email: firstname.lastname@example.org
Purpose: The mechanical effect of raised intraocular pressure is a recognised stimulus for optic neuropathy in primary open angle glaucoma (POAG). Characteristic extracellular matrix (ECM) remodelling accompanies axonal damage in the lamina cribrosa (LC) of the optic nerve head in POAG. Glial cells in the lamina cribrosa may play a role in this process but the precise cellular responses to mechanical forces in this region are unknown. The authors examined global gene expression profiles in lamina cribrosa cells exposed to cyclical mechanical stretch, with an emphasis on ECM genes.
Methods: Glial fibrillary acid protein negative primary LC cells were generated from the optic nerve head tissue of three normal human donors. Confluent cell passages (n=4) were exposed to 15% stretch at 1 Hz or static conditions for 24 h using the Flexercell system. Gene expression was assessed using Affymetrix U133A microarrays with pooled RNA. Expression levels were normalized using robust multi-chip average (RMA). Expression data was annotated using NIH DAVID software. ECM-related gene expression was validated in an independent experiment using quantitative real-time PCR and protein synthesis was measured using ELISA and immunohistochemistry.
Results: Compared with static controls, 805 genes were upregulated and 644 were downregulated by ±1.5 fold in stretched LC cells. Gene ontologies included ECM, cell proliferation, growth factor activity, and signal transduction. Differentially expressed ECM genes included elastin, collagens (IV, VI, VIII, IX), thrombospondin 1, perlecan, and lysl oxidase. Quantitative PCR demonstrated that the expression of TGF-β2, BMP-7, elastin, collagen VI, biglycan, versican, EMMPRIN, VEGF, and thrombomodulin were reproducible and consistent with the microarray data. VEGF and TGF-β2 protein levels were also significantly (p<0.05) increased in stretched cell media supernatants. Immunohistochemistry demonstrated increased EMMPRIN (an extracellular matrix metalloproteinase inducer) protein in human POAG optic nerve head tissue compared to nonglaucomatous controls.
Conclusions: These findings demonstrate that LC cells respond to mechanical stimuli in vitro by transcription of several components and modulators of the ECM. Some of the upregulated ECM genes identified are novel in the context of glaucomatous optic neuropathy (biglycan, versican, EMMPRIN, and BMP-7). The LC cell may represent both an important pro-fibrotic cell type in the optic nerve head and an attractive target for novel therapeutic intervention in POAG.
The lamina cribrosa is a distinct histological region of the optic nerve head (ONH). It is composed of approximately ten perforated (cribriform) connective tissue plates that permit the passage of retinal ganglion cell axons as they exit the ONH . It is a compliant tissue, which in normal human eyes sustains changes in intraocular pressure (IOP) without loss of structural or morphological integrity. This compliance reduces markedly with age due to increased collagen and reduced proteoglycan content in the lamina cribrosa extracellular matrix (ECM) [2,3]. Two important cell types that have been characterized in the lamina cribrosa include the ONH astrocyte and the lamina cribrosa (LC) cell. Both are members of the glial population of the ONH, however, the LC cell, unlike the astrocyte, does not express glial fibrillary acid protein (GFAP) . In terms of morphology, LC cells are broad, flat and polygonal, distinguishing them from the ONH astrocytes which are star shaped . ONH astrocytes are found throughout the ONH and separate the retinal ganglion cell axons from the cribriform plates. The LC cells, in contrast, are localized to the lamina cribrosa region and are situated within or between the cribriform plates .
Primary open angle glaucoma (POAG) is a progressive optic neuropathy characterized by raised IOP, retinal ganglion cell (RGC) axon loss, and excavation or cupping of the optic nerve head . It affects over 60 million people worldwide, representing one of the most common causes of irreversible blindness . Electron micrograph and immunohistochemistry studies demonstrate marked disruption to ECM architecture and composition in the glaucomatous lamina cribrosa. This includes backward displacement and distortion of the cribriform plates with increased amounts of collagen VI, elastin, and transforming growth factor-β2 [9-12]. Such ultrastructural changes may adversely affect the biomechanical properties of the lamina cribrosa, predisposing it to collapse as the IOP rises. An obvious consequence of lamina cribrosa collapse is the compression of RGC axons, which arrests their axoplasmic flow. This is clearly demonstrated in models of POAG in monkeys . It is likely that this is also the case in humans with POAG, offering a rational explanation for the interruption of optic nerve function and characteristic visual field disturbances. More recently, studies in monkeys have proposed that these mechanical insults to optic nerve function occur early in POAG, and support the hypothesis that the primary pathogenic event is mechanical failure of the lamina cribrosa .
It is not yet known which cell type is principally involved in ECM remodeling in the optic nerve head in POAG, however, the ONH astrocytes and LC cells have been implicated . One of the triggers for ECM remodeling may be the backward displacement of the lamina cribrosa (due to elevated IOP) which will deform and stretch its resident cells. Previous studies have shown that ONH astrocytes respond to hydrostatic pressure by upregulating elastin and heat shock protein-27 synthesis and that LC cells respond to hydrostatic pressure by upregulating collagen I mRNA synthesis or to cyclical strain by upregulation of TGF-β1 mRNA synthesis, TGF-β1 protein, and matrix metalloproteinase-2 activity [15-18]. These findings demonstrate a mechanosensory capability in both cell types which respond by release of components or modulators of the ECM. Unlike the LC cell, however, the ONH astrocyte has also been shown to release the neurotoxic mediator NO synthase-2 and may therefore directly induce axonal death in the optic nerve head .
While raised IOP is an established macro-stimulus for cupping of the ONH in POAG, the discrete transcriptional responses to mechanical forces in the lamina cribrosa in vivo are less clear. The emergence of microarray technology, which allows large scale profiling of gene expression in biological systems, has greatly enhanced our capacity to probe the molecular pathogenesis of POAG . In this present study, using microarrays, global changes in the LC cell transcriptome in response to 24 h of cyclical mechanical stretching were examined. Changes in selected ECM gene expression were then confirmed by quantitative PCR and corresponding protein changes by ELISA and immunohistochemistry. Some of the proteins of these genes have previously been characterized in the optic nerve head in POAG (e.g., collagen VI, elastin, and TGF-β2) whereas others are novel in this context (e.g., EMMPRIN, biglycan, VEGF, and BMP-7).
Primary cultures of human LC cells were generated from carefully dissected optic nerve head tissue from three normal donors and characterized as reported previously [5,21]. Cultures were confirmed to be glial fibrillary acid protein (GFAP) negative and maintained in 5% CO2-95% air at 37 °C. Four separate cultures of confluent cells were used for four mechanical stretching experiments (n=4). Each passage of cells was trypsinized and seeded equally (6x104 cells per well) in duplicate to flexible six well culture plates coated with laminin (Flexcell International Corp., Hillsborough, NC). Confluent monolayers formed every 7-10 days. Two cultures of cells were generated from donor 1 and one culture from donor 2 (total of 3 separate cultures), which were then used for the microarray analysis (Figure 1). A separate fourth culture of cells was generated from donor 3 and used for real-time PCR validation (Figure 1). All cultures used were between cell passages four and six. Cell media supernatants from all four experiments (n=4) were removed for analysis by enzyme linked immunosorbent assay (ELISA).
Application of cyclical mechanical stretch
Confluent LC cells were stretched using the Flexercell® Tension PlusTM FX-4000T system (Flexcell International Corp.) . This is a computer driven apparatus that creates a programmable bi-axial strain across laminin coated culture wells when positioned on a vacuum base station (Figure 2A). Control plates, also flexible and laminin coated, are not positioned on the base station but instead are placed adjacent to it in the same incubator. Each cycle of strain contains an upstroke to the predetermined percentage elongation and then falls to baseline (0% strain) with a distinct dicrotic notch (Figure 2B). Confluent cells were serum starved for 24 h prior to 1 s cycles (1 Hz) of 15% strain or static conditions for 24 h. All cells were maintained in 5% CO2-95% air at 37 °C for the duration of the experiment.
Isolation of RNA
Total RNA was isolated using a silica gel-based RNeasy spin column protocol (Qiagen, Valencia, CA). Total RNA was suspended in 30 μl of RNase-free H2O. RNA concentrations and purity were calculated from the absorbance at 260/280 nm and 5 μg of each sample was resolved on 1.5% agarose gels to assess their integrity.
To investigate the effect of 24 h of cyclical mechanical stretching on global gene expression in LC cells, Affymetrix HG-U133A microarrays containing 22,283 oligonucleotide target probes were used. Complimentary DNA (cDNA) was synthesized from total RNA using SuperScript Choice kit (Invitrogen, Paisley, UK). Biotin-labeled cRNA was prepared from template cDNA in an in vitro transcription (IVT) reaction using the Bioarray High-Yield RNA transcript labeling kit (Enzo Life Sciences, Farmingdale, NY). Labeled RNA from three separate stretch and control experiments (n=3) was pooled and then hybridized to an experimental or control microarray. Arrays were washed and fluorescently labeled before being scanned by a confocal laser scanner. Raw image files (CEL format) were obtained through Affymetrix Suite 5.0 software (MAS 5.0). Probe intensities from the two chips were then normalized by robust multi-chip analysis (RMA) using RMA Express. RMA Express is a stand alone graphic user interface package that analyzes directly from the Affymetrix microarray image file (CEL format) and comprises of a number of steps to background adjust, quantile normalize, log transform, and summarize the perfect match probe values . This form of normalization is called "complete normalization" as it uses the probe intensity data from all arrays in an experiment to form the normalizing relation. Expression data was filtered using Microsoft Excel. Genes displaying a signal log ratio of greater than 0.6 or less than -0.6 (±1.5 fold) were retained for further analysis and a subset of ECM/ECM-related genes were seleceted for independent validation by real time PCR.
Biologically relevant groups from the microarray data were generated using the Database for Annotation, Visualization and Integrated Discovery (DAVID; version 1.0) from the National Institutes of Health . This is a web-based software package that functionally annotates a list of uploaded gene identifiers and converts them into an ordered table of biological themes.
Quantitative real-time PCR
Total RNA (1 μg) was used to synthesize the first-strand cDNA using random hexamers and SuperScript II reverse transcriptase (Invitrogen, Paisley, UK). The cDNA was used for quantitative real-time PCR amplification with TaqManTM chemistry (Applied Biosystems, Foster City, CA). Fluorogenic probes and sequence specific primers for TGF-β2, BMP-7, elastin, collagen VI, biglycan, versican, EMMPRIN, thrombomodulin, VEGF, and the endogenous control 18S ribosomal RNA were designed and optimized in a preformulated assay (Assay-on-DemandTM, Applied Biosystems). The target gene names, Unigene identifiers and Applied Biosystems primer/probe assay identifiers are shown in Table 1. Duplicate cDNA template samples were amplified and analyzed in the Prism 7900HT sequence detection system (Applied Biosystems). Thermal cycler conditions were 10 min at 95 °C followed by 40 cycles of 30 s at 95 °C to denature the DNA and 30 s at 60 °C to anneal and extend the template. A standard curve of cycle thresholds using serial dilutions of cDNA samples were established and used to calculate the relative abundance of the target gene between stretch and control samples. Values were normalized to the relative amounts of 18S mRNA, which were obtained from a similar standard curve.
Enzyme linked immunosorbent assay (ELISA)
For quantification of total TGF-β2 or VEGF protein in stretched and static cell media supernatants (n=4), the quantitative sandwich immunoassay technique was used (both ELISA kits, R&D Systems, Minneapolis, MN). They were activated where appropriate with 1 N HCl according to the manufacturers instructions and standard concentrations of TGF-β2 or VEGF protein were prepared. Activated samples and standards were loaded to separate 96 well plates coated with monoclonal antibodies specific for human TGF-β2 or VEGF. After washing away any unbound substances, an enzyme-linked polyclonal antibody specific for TGF-β2 or VEGF was added to the wells. Following a second wash to remove any unbound antibody-enzyme reagent, a solution of hydrogen peroxide and tetramethylbenzidine was added to the wells after which a blue color developed in proportion to the amount of target protein bound in the initial step. The reaction was stopped using sulfuric acid and the optical density of each well was measured at 450 nm.
Double immunoflourescence histochemistry
Eyes from three human donors were obtained from regional eye banks and placed in 4% neutral buffered formalin within 4 h of death. Two of the donors had a documented history of glaucoma. The third donor had no history of ocular disease (normal control). The posterior segments were embedded in paraffin and 6 μm optic nerve head sections were cut and mounted on glass slides. After paraffin removal, the tissue was quenched for aldehydes by treatment in a 0.05 M glycine (Sigma, Dorset, UK) solution for 15 min. Nonspecific binding sites were then blocked for 15 min with phosphate buffered saline (PBS) solution containing 1% BSA and 1% serum from the species (rabbit or mouse) in which the secondary antibody was raised. The slides were washed in PBS prior to incubation with a mixture of anti-EMMPRIN antibody (diluted 1:25 in PBS; Abcam, Cambridge, MA) and anti-GFAP antibody (diluted 1:80 in PBS; Sigma, Dorset, UK) at 4 °C overnight. The sections were then washed 3 times in PBS. For detection of EMMPRIN, the sections were incubated with Alexa Fluor-488 labeled anti-mouse IgG for 1 h at room temperature (RT). For detection of GFAP, the sections were incubated with Alexa Fluor-594 labeled anti-rabbit IgG for 1 h at RT (both secondary antibodies were obtained from Molecular Probes, Leiden, The Netherlands). The tissue nuclei were then DAPI stained (300 nM in water) for 5-10 min. Images were captured with a Nikon Microphot FXA (Nikon, Inc., Melville, NY) equipped with a SenSys CCD camera (Photometrics, Tucson, AZ). Images were deconvoluted using Scanalytics IPLAB (Scanalytics, Fairfax, VA) and Vaytek Microtome (Vaytek, Fairfield, IA) software.
Data for the ELISA are summarized as the mean fold change±standard error (SE) calculated from four separate experiments (n=4). The paired student t-test was used to analyze the statistical significance (p<0.05) of differences between mean values.
Microarray and bioinformatic analysis
To investigate the effects of cyclical mechanical stretch on global gene expression in LC cells we used Affymetrix HG-U133A cRNA microarrays. A series of three cell culture experiments were performed generated from two separate human donors. Figure 3A shows a box plot of the control and stretch microarrays following normalization by Robust Multichip Analysis (RMA). The microarray data discussed in this manuscript is available at the Gene Expression Omnibus (GEO) database under the accession number GSE3166. Figure 3B shows the log/log pairwise scatter plot of the expression levels of the 22,283 genes from stretch and control arrays. Of the 22,283 genes scanned, 805 were upregulated by more than 0.6 signal log ratios (1.5 fold) and 644 were downregulated by more than -0.6 signal log ratios (1.5 fold) in stretch compared to control. In order to generate biologically relevant groups of genes from the microarray data, we used the NIH DAVID software. This separated the genes into three broad functional groups; biological process, molecular function, and cellular component (Figure 4). In Table 2 we list individual differentially expressed genes belonging to the categories of ECM, cell proliferation, growth factor activity, cell surface receptor linked signal transduction, and coagulation. These categories have been shown in detail, as they are the categories to which the nine validated genes belong. Of particular interest to our work was the ECM category, which showed that 30 ECM genes were differentially expressed in stretch compared to control (24 upregulated, 6 downregulated, Table 2). Components of the ECM within this category included elastin, biglycan, perlecan, collagen IV, collagen VI, collagen VIII, collagen IX, and collagen XIV. Modulators of the ECM within this category included ADAMTS2, ADAMTS9, lysl oxidase, MMP-23A, and thrombospondin 1.
Quantitative real time PCR
To validate the microarray analysis, we performed quantitative real-time PCR with TaqManTM chemistry using independent RNA (not used for the microarray analysis) from stretched and control LC cells generated from a third human donor. We analyzed 9 genes whose expression was upregulated (TGF-β2, BMP-7, elastin, collagen VI, biglycan, versican, and VEGF) or downregulated (thrombomodulin) in response to 24 h of cyclical mechanical stretch. While not all genes validated were assigned to the ECM category (TGF-β2, VEGF, BMP-7, EMMPRIN, and thrombomodulin) their functions do include maintenance or regulation of ECM metabolism. Figure 5A shows the fold expression changes observed using real-time PCR for the 9 validated genes. The results showed that the expression changes for all selected genes were reproducible and consistent with the direction of change predicted by microarray analysis. The magnitude of change determined by microarray and real-time PCR were somewhat different, a result which was not surprising considering the technical differences in the methods of analysis and normalization.
Enzyme linked immunosorbent assay (ELISA)
To investigate if the changes in mRNA reflected similar changes in protein synthesis, we used enzyme linked immunosorbent assay (ELISA). VEGF and TGF-β2 were chosen as representative genes whose expression was altered by cyclical mechanical stretch. Figure 5B shows that VEGF and TGF-β2 protein secretion were significantly increased in cell media supernatants of stretched LC cells compared to static controls. Mean control TGF-β2 concentration was 87±19 pg/ml (SE) and mean stretch TGF-β2 concentration was 115±30 pg/ml (SE; n=4; p=0.02). Mean control VEGF concentration was 241±66 pg/ml (SE) and mean VEGF stretch concentration was 407±148 pg/ml (SE; n=4; p=0.03). This correlated with the recorded direction of change in VEGF and TGF-β2 mRNA synthesis as measured by microarray and real-time PCR.
Double immunofluorescence histochemistry
EMMPRIN (green) and GFAP (red) immunoreactivity were both increased in the lamina cribrosa of glaucomatous tissue sections compared with the normal control sections. A representative micrograph is shown in Figure 6; glaucomatous lamina cribrosa tissue (Figure 6B), normal control lamina cribrosa tissue (Figure 6A). EMMPRIN protein was detected both distinct from, and co-localized (yellow) with GFAP in the glaucomatous sections. No staining was evident when the primary antibody was omitted (Figure 6C,D).
In this study, using microarray technology, we investigated global gene expression in LC cells in response to 24 h of cyclical mechanical stretch. Our results illustrate that the transcriptional profile of mechanically induced genes in LC cells contains several components and modulators of the ECM (e.g., elastin, collagen IV, VI, VIII, IX, XIV, fibulin 1, thrombospondin 1, VEGF, ADAMTS2, ADAMTS9, MMP-23A, and lysl oxidase). In order to validate the microarray results a number of genes were chosen for quantitative real time PCR analysis. This set of genes comprised eight upregulated (Collagen VI, elastin, TGF-β2, BMP-7, biglycan, versican, EMMPRIN, and VEGF) and one downregulated (thrombomodulin). They were selected because they are components or modulators of the ECM and as such may play a role in ECM remodeling in the optic nerve head in POAG. Three of the validated genes, collagen VI, elastin, and TGF-β2 have previously been characterized in the optic nerve head in POAG. The remaining six validated genes are, to our knowledge, novel in the context of ECM remodeling in POAG.
The parameters used in these experiments were 15% biaxial strain at 1 Hz maintained for 24 h. These are standard parameters that have been used extensively by other investigators using other cell types . The cyclic component (1 Hz) corresponded to ocular pulsations and the dicrotic waveform produced by the Flexercell apparatus created a physiological approximation to the central retinal artery pressure waveform (Figure 2B). A mathematical model of lamina cribrosa stresses in raised IOP as generated by Edwards et al.  demonstrated that the lamina cribrosa will experience strain of 12% at 40 mm Hg. While this relationship was not linear it is possible that the 15% strain used in our experiments could emulate the effect of pressures in excess of 40 mm Hg. Though these conditions are not an exact reproduction of the microenvironment to which LC cells are exposed in raised IOP in vivo, the Flexercell system does produce stimuli that are useful to investigators of POAG in vitro.
We chose a ±0.6 signal log ratio (±1.5 fold) cut off to filter the microarray data because the quantile normalization method of RMA which we used produces comparatively smaller signal log ratios than does the average scaling method of Affymetrix MAS 5.0 . Briefly, the differences between the analysis methods relates to background correction, Affymetrix MAS 5.0 uses a large background correction which adds more noise but gives larger fold change estimates. RMA gives smaller fold change estimates because smaller background corrections are made, but with minimal noise and more precision .
Collagen fibers constitute an important fraction of the human lamina cribrosa ECM. The center of each cribriform plate is composed of collagen I, III, V, and VI, surrounded by a layer of basement membrane type IV collagen to which the LC cells are attached [28,29]. Collagens behave in a synergistic manner with elastic fibers, and are thought to provide structural support to the lamina cribrosa and adjacent RGC axons . Another function of collagen VI is the organization of fibronectin protein in the ECM . Marked elevation in collagen VI protein has been demonstrated in the lamina cribrosa tissue of both monkeys with experimentally elevated IOP and humans with POAG [32,33]. Collagen type I and basement membrane type IV collagen are also increased in the lamina cribrosa in POAG . Our finding that LC cells upregulate collagen IV and VI expression in response to mechanical stretch is, therefore, important and indicates that this cell type may be a source for this protein in POAG. The other collagens which were upregulated by stretch were collagen VIII, IX, and XIV. These collagens may also play a role in ECM changes in the glaucomatous lamina cribrosa and provide new targets for immunohistochemistry studies of optic nerve head tissue.
Elastin, like collagen, is a major component of the lamina cribrosa ECM and is typically found in tissues that are exposed to mechanical stresses in vivo . Elastin is a classically upregulated protein in many fibrotic diseases, and in the glaucomatous lamina cribrosa, its fibers become curled and their architecture deranged, in a process referred to as elastotic degeneration . It is likely that the changes in optic nerve head elastin metabolism in POAG reflect a response to chronic elevation of IOP. Support for this comes from studies of experimental IOP elevation in monkeys where increased production of morphologically abnormal elastin has been shown in the lamina cribrosa of experimental animals, but not in normal IOP controls . In vitro studies using human optic nerve head astrocytes exposed to hydrostatic pressure have also shown that both elastin mRNA and protein expression are upregulated after 24 h in this environment . It may, therefore, be the case that elastotic degeneration adversely affects the lamina cribrosa's intrinsic compliance, predisposing it to collapse and compression as IOP rises. Two other genes that were downregulated in our system, MAGP-2 and fibulin-1, encode important proteins involved in elastin fiber assembly in the ECM [37,38] and also be may be important in the disturbed elastin production that occurs in the glaucomatous lamina cribrosa.
TGF-β2, a member of the TGF-β superfamily, is a multifunctional cytokine whose biological activities include induction of ECM synthesis, modulation of wound healing, and inhibition of cell growth . While the prototypical isoform, TGF-β1, has been detected in normal adult human lamina cribrosa tissue, TGF-β2 is the predominant type found in this tissue in POAG . In this setting, it may be responsible for the progressive accumulation and remodelling of the lamina cribrosa ECM as demonstrated by immunohistochemistry studies . We found both TGF-β2 mRNA and TGF-β2 protein expression to be increased in LC cells exposed to 24 h of cyclical mechanical stretch. We also found that TGF-β1 mRNA was upregulated in response to mechanical stretch, therefore, indicating that LC cells may be a significant source of TGF-β1 and TGF-β2 in the lamina cribrosa in POAG, possibly in response to the mechanical stimulus of raised IOP.
Bone morphogenetic proteins (BMPs) are also members of the TGF-β superfamily . They were originally detected as bone and cartilage inducing proteins but their other functions include regulation of developmental processes such as cell proliferation and apoptosis . Human LC cells in vitro express mRNA and protein for BMP-2, BMP-4, BMP-7, and the BMP receptor complex . Like their family members TGF-β1 and TGF-β2, cyclical mechanical stretch also upregulated the synthesis of BMP-1 and BMP-7 mRNA in our study. BMP-1 has been shown to modulate the formation of collagen I fibers and BMP-7 possesses anti-fibrotic properties [44,45]. Taken in this context, stretch induced BMP-1 and BMP-7 mRNA synthesis in LC cells may represent a response intended to modulate ECM accumulation in the optic nerve head.
Proteoglycans are another important component of the mammalian ECM and are present in both the human and monkey lamina cribrosa [46,47]. Elevated chondroitin sulfate proteoglycans are also evident in the lamina cribrosa of monkeys with laser-induced glaucoma, but not in normal IOP controls . This may represent a compensatory response to elevated IOP, as the expandable hydrodynamic volume which proteoglycans can generate is thought to aid buffering of intraocular/cerebrospinal fluid (translaminar) pressure gradients . We found that the expression of three proteoglycans, perlecan, biglycan, and versican were upregulated in LC cells in response to cyclical mechanical stretch. Perlecan is a heparan sulfate proteoglycan and a major component of basement membranes and functions in anchoring other matrix proteins to the basal lamina . Biglycan is a member of the family of small matrix proteogylcans which includes decorin and fibromodulin. It has several functions including extracellular organization of collagen VI fibers and support of neuronal growth . Versican is a large proteoglycan belonging to the hyaluronan-binding family that includes aggrecan and neurocan. Expression of the V3 isoform of versican was upregulated by stretch in our experiments. Verscian V3 can induce tropoelastin synthesis in arterial smooth muscle cells and correct impaired elastogenesis in fibroblasts [51,52]. These findings indicate that mechano-responsive perlecan, biglycan, and versican V3 expression may be relevant to basement membrane, collagen VI, and elastin metabolism in the ONH in POAG.
EMMPRIN (extracellular matrix metalloproteinase inducer) is a transmembrane glycoprotein and a member of the immunoglobulin superfamily. Its functions include induction of anchorage independent cell proliferation and stimulation of matrix metalloproteinase-2 (MMP-2) activity [53-55]. We found that EMMPRIN mRNA synthesis was upregulated in response to mechanical stretch in LC cells and that there was a marked increase in EMMPRIN protein staining in glaucomatous lamina cribrosa tissue compared to normal controls. EMMPRIN was identified distinct from, and co-localized with, GFAP in the glaucomatous sections. This staining pattern appears to indicate that both astrocytes and GFAP-negative cells (i.e., LC cells) may be responsible for upregulation of this MMP inducing protein in POAG. This is the first study examining EMMPRIN in the human optic nerve head, and we believe that its detection there is of relevance to the characteristic ECM changes in the glaucomatous lamina cribrosa. As mentioned above, EMMPRIN can induce anchorage-independent cell growth, which it achieves via a hyaluronan-dependent mechanism. This property of EMMPRIN may be important to reactive astrocytes in the glaucomatous optic nerve head, which in CNS injuries detach from their basement membranes and migrate to and proliferate at sites of axonal damage .
Thrombomodulin, another transmembrane glycoprotein, demonstrated downregulation in response to mechanical stretch in LC cells. This negative transcriptional response in the thrombomodulin gene has also been shown in human vascular smooth muscle cells exposed to mechanical strain in vitro . During experimental wound healing, heterozygous thrombomodulin deficient mice exhibit thicker and denser foci of collagen aggregates than wild type controls . As mentioned above, collagen VI deposition is increased in the lamina cribrosa in POAG. However, collagen I and IV are similarly elevated and form compacted bundles of fibers that are unlike those in normal tissue . That thrombomodulin expression was reduced in stretched LC cells is intriguing, and may contribute to an explanation for the disorganized collagen deposition that occurs in the lamina cribrosa in POAG.
VEGF is an established angiogenic, pro-proliferative growth factor whose transcription is driven by several stimuli such as mechanical strain and hypoxia [60,61]. VEGF induces a number of ECM remodelling enzymes including members of the MMP family . It also drives the expression of connective tissue growth factor (CTGF), a potent mediator of fibrosis, ECM synthesis, and angiogenesis . More recently, VEGF has been suggested as an important neurotrophic and neuroprotective agent in the central and peripheral nervous systems . For instance, it can stimulate axonal outgrowth in superior cervical and dorsal root ganglion cells and migration of astrocytes in vitro. We found that both VEGF mRNA synthesis and VEGF protein release were upregulated in LC cells following 24 h of cyclical mechanical stretch. It is possible that in POAG, stretch-induced VEGF may promote ECM remodeling, modulate astrocyte activity, and support RGC axon survival.
Our data shows consistency with previous in vitro studies conducted by Cui et al.  on human scleral fibroblasts and Yang et al.  on ONH astrocytes where microarray analysis identified similarly stretch responsive genes (e.g., VEGF and elastin). In contrast to the latter study, however, our results appear to show more transcription of ECM genes in LC cells than in ONH astrocytes, though it is acknowledged that the stretching methods were different in both studies. Other work conducted by Ahmed et al.  examined transcriptional patterns in retinal tissue from rats which had been exposed to chronically elevated IOP. Microarray analysis showed that expression of VEGF and several other ECM components (fibronectin and SPARC) and modulators (TIMP-1 and MMP-3) were upregulated in the experimental retinas compared with normal IOP controls. In common with our earlier work measuring TGF-β1 mRNA synthesis using real time PCR , this gene was also unregulated (Table 2, cell proliferation) in this present study. Though we had also found that MMP-2 protein activity was significantly increased in LC cells following 24 h of mechanical stretching, we did not observe similar changes at the mRNA level in the present study. Taken together, our data and those of these preceding studies strongly suggests that mechanical forces in the optic nerve head and retina are important in inducing ECM synthesis and turnover.
To our knowledge this study represents the first broad gene expression analysis of LC cells exposed to cyclical mechanical stretch. The data indicates that the LC cell's transcriptional response to mechanical stretch may be intended to bolster the lamina cribrosa's ECM, ultimately preventing its collapse during extended periods of elevated IOP. This represents an additional step towards understanding the role and mechanism of mechanical stretch stimuli in the optic nerve head in POAG. Future work will determine the precise pathological stimulus to which LC cells should be exposed, either varying the degree (e.g., <15%) or chronicity (e.g., >24 h) of strain. We also believe that our results provide new candidate ECM proteins to detect in glaucomatous optic nerve head tissue, proteins which may bear relevance to the pathogenesis of POAG (e.g., EMMPRIN). Finally, this study highlights a likely role for LC cells and mechanical stimuli in ONH ECM metabolism, particularly as LC cells express alpha smooth muscle actin and are, therefore, primed to respond to external stimuli in a pro-fibrotic manner . Our results also give sufficient rationale for investigating the anti-fibrotic and possible therapeutic potential of interrupting LC cell responses to mechanical stimuli in POAG.
Ruaidhrí P. Kirwan is supported by a grant from the Mater College for Post Graduate Education and Research 2003-2004. The authors wish to acknowledge the advice and technical assistance of William Howe, Debbie Lane (Glaucoma Research, Alcon Research Ltd., Fort Worth, TX), Anne-Marie Brun, and Dr. Lawrence Oakford (UNTHSC at Fort Worth, Fort Worth, TX).
1. Anderson DR. Ultrastructure of human and monkey lamina cribrosa and optic nerve head. Arch Ophthalmol 1969; 82:800-14.
2. Albon J, Purslow PP, Karwatowski WS, Easty DL. Age related compliance of the lamina cribrosa in human eyes. Br J Ophthalmol 2000; 84:318-23.
3. Albon J, Karwatowski WS, Easty DL, Sims TJ, Duance VC. Age related changes in the non-collagenous components of the extracellular matrix of the human lamina cribrosa. Br J Ophthalmol 2000; 84:311-7.
4. Hernandez MR, Igoe F, Neufeld AH. Cell culture of the human lamina cribrosa. Invest Ophthalmol Vis Sci 1988; 29:78-89.
5. Lambert W, Agarwal R, Howe W, Clark AF, Wordinger RJ. Neurotrophin and neurotrophin receptor expression by cells of the human lamina cribrosa. Invest Ophthalmol Vis Sci 2001; 42:2315-23.
6. Hernandez MR. The optic nerve head in glaucoma: role of astrocytes in tissue remodeling. Prog Retin Eye Res 2000; 19:297-321.
7. Quigley HA, Hohman RM, Addicks EM, Massof RW, Green WR. Morphologic changes in the lamina cribrosa correlated with neural loss in open-angle glaucoma. Am J Ophthalmol 1983; 95:673-91.
8. Quigley HA. Number of people with glaucoma worldwide. Br J Ophthalmol 1996; 80:389-93.
9. Coleman AL, Quigley HA, Vitale S, Dunkelberger G. Displacement of the optic nerve head by acute changes in intraocular pressure in monkey eyes. Ophthalmology 1991; 98:35-40.
10. Levy NS, Crapps EE. Displacement of optic nerve head in response to short-term intraocular pressure elevation in human eyes. Arch Ophthalmol 1984; 102:782-6.
11. Hernandez MR. Ultrastructural immunocytochemical analysis of elastin in the human lamina cribrosa. Changes in elastic fibers in primary open-angle glaucoma. Invest Ophthalmol Vis Sci 1992; 33:2891-903.
12. Pena JD, Taylor AW, Ricard CS, Vidal I, Hernandez MR. Transforming growth factor beta isoforms in human optic nerve heads. Br J Ophthalmol 1999; 83:209-18.
13. Quigley HA, Addicks EM. Chronic experimental glaucoma in primates. II. Effect of extended intraocular pressure elevation on optic nerve head and axonal transport. Invest Ophthalmol Vis Sci 1980; 19:137-52.
14. Bellezza AJ, Rintalan CJ, Thompson HW, Downs JC, Hart RT, Burgoyne CF. Deformation of the lamina cribrosa and anterior scleral canal wall in early experimental glaucoma. Invest Ophthalmol Vis Sci 2003; 44:623-37.
15. Hernandez MR, Pena JD, Selvidge JA, Salvador-Silva M, Yang P. Hydrostatic pressure stimulates synthesis of elastin in cultured optic nerve head astrocytes. Glia 2000; 32:122-36.
16. Salvador-Silva M, Ricard CS, Agapova OA, Yang P, Hernandez MR. Expression of small heat shock proteins and intermediate filaments in the human optic nerve head astrocytes exposed to elevated hydrostatic pressure in vitro. J Neurosci Res 2001; 66:59-73.
17. Yang JL, Neufeld AH, Zorn MB, Hernandez MR. Collagen type I mRNA levels in cultured human lamina cribrosa cells: effects of elevated hydrostatic pressure. Exp Eye Res 1993; 56:567-74.
18. Kirwan RP, Crean JK, Fenerty CH, Clark AF, O'Brien CJ. Effect of cyclical mechanical stretch and exogenous transforming growth factor-beta1 on matrix metalloproteinase-2 activity in lamina cribrosa cells from the human optic nerve head. J Glaucoma 2004; 13:327-34.
19. Liu B, Neufeld AH. Expression of nitric oxide synthase-2 (NOS-2) in reactive astrocytes of the human glaucomatous optic nerve head. Glia 2000; 30:178-86.
20. Hernandez MR, Agapova OA, Yang P, Salvador-Silva M, Ricard CS, Aoi S. Differential gene expression in astrocytes from human normal and glaucomatous optic nerve head analyzed by cDNA microarray. Glia 2002; 38:45-64.
21. Wordinger RJ, Lambert W, Agarwal R, Liu X, Clark AF. Cells of the human optic nerve head express glial cell line-derived neurotrophic factor (GDNF) and the GDNF receptor complex. Mol Vis 2003; 9:249-56 <http://www.molvis.org/molvis/v9/a36/>.
22. Banes AJ, Gilbert J, Taylor D, Monbureau O. A new vacuum-operated stress-providing instrument that applies static or variable duration cyclic tension or compression to cells in vitro. J Cell Sci 1985; 75:35-42.
23. Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 2003; 19:185-93.
24. Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 2003; 4:P3.
25. WuDunn D. The effect of mechanical strain on matrix metalloproteinase production by bovine trabecular meshwork cells. Curr Eye Res 2001; 22:394-7.
26. Edwards ME, Good TA. Use of a mathematical model to estimate stress and strain during elevated pressure induced lamina cribrosa deformation. Curr Eye Res 2001; 23:215-25.
27. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 2003; 31:e15.
28. Hernandez MR, Wang N, Hanley NM, Neufeld AH. Localization of collagen types I and IV mRNAs in human optic nerve head by in situ hybridization. Invest Ophthalmol Vis Sci 1991; 32:2169-77.
29. Ye H, Yang J, Hernandez MR. Localization of collagen type III mRNA in normal human optic nerve heads. Exp Eye Res 1994; 58:53-63.
30. Hernandez MR, Pena JD. The optic nerve head in glaucomatous optic neuropathy. Arch Ophthalmol 1997; 115:389-95.
31. Sabatelli P, Bonaldo P, Lattanzi G, Braghetta P, Bergamin N, Capanni C, Mattioli E, Columbaro M, Ognibene A, Pepe G, Bertini E, Merlini L, Maraldi NM, Squarzoni S. Collagen VI deficiency affects the organization of fibronectin in the extracellular matrix of cultured fibroblasts. Matrix Biol 2001; 20:475-86.
32. Fukuchi T. [Extracellular matrix of the optic nerve lamina cribrosa in monkey eyes with experimentally induced glaucoma]. Nippon Ganka Gakkai Zasshi 1991; 95:303-10.
33. Hernandez MR, Andrzejewska WM, Neufeld AH. Changes in the extracellular matrix of the human optic nerve head in primary open-angle glaucoma. Am J Ophthalmol 1990; 109:180-8.
34. Xu C, Zarins CK, Glagov S. Gene expression of tropoelastin is enhanced in the aorta proximal to the coarctation in rabbits. Exp Mol Pathol 2002; 72:115-23.
35. Pena JD, Netland PA, Vidal I, Dorr DA, Rasky A, Hernandez MR. Elastosis of the lamina cribrosa in glaucomatous optic neuropathy. Exp Eye Res 1998; 67:517-24.
36. Quigley HA, Brown A, Dorman-Pease ME. Alterations in elastin of the optic nerve head in human and experimental glaucoma. Br J Ophthalmol 1991; 75:552-7.
37. Penner AS, Rock MJ, Kielty CM, Shipley JM. Microfibril-associated glycoprotein-2 interacts with fibrillin-1 and fibrillin-2 suggesting a role for MAGP-2 in elastic fiber assembly. J Biol Chem 2002; 277:35044-9.
38. Roark EF, Keene DR, Haudenschild CC, Godyna S, Little CD, Argraves WS. The association of human fibulin-1 with elastic fibers: an immunohistological, ultrastructural, and RNA study. J Histochem Cytochem 1995; 43:401-11.
39. Yu L, Border WA, Huang Y, Noble NA. TGF-beta isoforms in renal fibrogenesis. Kidney Int 2003; 64:844-56.
40. Yan X, Tezel G, Wax MB, Edward DP. Matrix metalloproteinases and tumor necrosis factor alpha in glaucomatous optic nerve head. Arch Ophthalmol 2000; 118:666-73.
41. Yamashita H, Ten Dijke P, Heldin CH, Miyazono K. Bone morphogenetic protein receptors. Bone 1996; 19:569-74.
42. Hogan BL. Bone morphogenetic proteins: multifunctional regulators of vertebrate development. Genes Dev 1996; 10:1580-94.
43. Wordinger RJ, Agarwal R, Talati M, Fuller J, Lambert W, Clark AF. Expression of bone morphogenetic proteins (BMP), BMP receptors, and BMP associated proteins in human trabecular meshwork and optic nerve head cells and tissues. Mol Vis 2002; 8:241-50 <http://www.molvis.org/molvis/v8/a31/>.
44. Ge G, Seo NS, Liang X, Hopkins DR, Hook M, Greenspan DS. Bone morphogenetic protein-1/tolloid-related metalloproteinases process osteoglycin and enhance its ability to regulate collagen fibrillogenesis. J Biol Chem 2004; 279:41626-33.
45. Morrissey J, Hruska K, Guo G, Wang S, Chen Q, Klahr S. Bone morphogenetic protein-7 improves renal fibrosis and accelerates the return of renal function. J Am Soc Nephrol 2002; 13:S14-21.
46. Sawaguchi S, Iwata K, Kaiya T. [The distribution and characterization of sulfated-proteoglycans in the normal human lamina cribrosa]. Nippon Ganka Gakkai Zasshi 1991; 95:311-7.
47. Morrison JC, Rask P, Johnson EC, Deppmeier L. Chondroitin sulfate proteoglycan distribution in the primate optic nerve head. Invest Ophthalmol Vis Sci 1994; 35:838-45.
48. Fukuchi T, Sawaguchi S, Yue BY, Iwata K, Hara H, Kaiya T. Sulfated proteoglycans in the lamina cribrosa of normal monkey eyes and monkey eyes with laser-induced glaucoma. Exp Eye Res 1994; 58:231-43.
49. Irving-Rodgers HF, Harland ML, Rodgers RJ. A novel basal lamina matrix of the stratified epithelium of the ovarian follicle. Matrix Biol 2004; 23:207-17.
50. Wiberg C, Heinegard D, Wenglen C, Timpl R, Morgelin M. Biglycan organizes collagen VI into hexagonal-like networks resembling tissue structures. J Biol Chem 2002; 277:49120-6.
51. Merrilees MJ, Lemire JM, Fischer JW, Kinsella MG, Braun KR, Clowes AW, Wight TN. Retrovirally mediated overexpression of versican v3 by arterial smooth muscle cells induces tropoelastin synthesis and elastic fiber formation in vitro and in neointima after vascular injury. Circ Res 2002; 90:481-7.
52. Hinek A, Braun KR, Liu K, Wang Y, Wight TN. Retrovirally mediated overexpression of versican v3 reverses impaired elastogenesis and heightened proliferation exhibited by fibroblasts from Costello syndrome and Hurler disease patients. Am J Pathol 2004; 164:119-31.
53. Marieb EA, Zoltan-Jones A, Li R, Misra S, Ghatak S, Cao J, Zucker S, Toole BP. Emmprin promotes anchorage-independent growth in human mammary carcinoma cells by stimulating hyaluronan production. Cancer Res 2004; 64:1229-32.
54. Caudroy S, Polette M, Nawrocki-Raby B, Cao J, Toole BP, Zucker S, Birembaut P. EMMPRIN-mediated MMP regulation in tumor and endothelial cells. Clin Exp Metastasis 2002; 19:697-702.
55. Li R, Huang L, Guo H, Toole BP. Basigin (murine EMMPRIN) stimulates matrix metalloproteinase production by fibroblasts. J Cell Physiol 2001; 186:371-9.
56. Rhodes KE, Moon LD, Fawcett JW. Inhibiting cell proliferation during formation of the glial scar: effects on axon regeneration in the CNS. Neuroscience 2003; 120:41-56.
57. Feng Y, Yang JH, Huang H, Kennedy SP, Turi TG, Thompson JF, Libby P, Lee RT. Transcriptional profile of mechanically induced genes in human vascular smooth muscle cells. Circ Res 1999; 85:1118-23.
58. Peterson JJ, Rayburn HB, Lager DJ, Raife TJ, Kealey GP, Rosenberg RD, Lentz SR. Expression of thrombomodulin and consequences of thrombomodulin deficiency during healing of cutaneous wounds. Am J Pathol 1999; 155:1569-75.
59. Morrison JC, Dorman-Pease ME, Dunkelberger GR, Quigley HA. Optic nerve head extracellular matrix in primary optic atrophy and experimental glaucoma. Arch Ophthalmol 1990; 108:1020-4.
60. Leonard MO, Cottell DC, Godson C, Brady HR, Taylor CT. The role of HIF-1 alpha in transcriptional regulation of the proximal tubular epithelial cell response to hypoxia. J Biol Chem 2003; 278:40296-304.
61. Petersen W, Varoga D, Zantop T, Hassenpflug J, Mentlein R, Pufe T. Cyclic strain influences the expression of the vascular endothelial growth factor (VEGF) and the hypoxia inducible factor 1 alpha (HIF-1alpha) in tendon fibroblasts. J Orthop Res 2004; 22:847-53.
62. Unemori EN, Ferrara N, Bauer EA, Amento EP. Vascular endothelial growth factor induces interstitial collagenase expression in human endothelial cells. J Cell Physiol 1992; 153:557-62.
63. Suzuma K, Naruse K, Suzuma I, Takahara N, Ueki K, Aiello LP, King GL. Vascular endothelial growth factor induces expression of connective tissue growth factor via KDR, Flt1, and phosphatidylinositol 3-kinase-akt-dependent pathways in retinal vascular cells. J Biol Chem 2000; 275:40725-31.
64. Sondell M, Lundborg G, Kanje M. Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system. J Neurosci 1999; 19:5731-40.
65. Cui W, Bryant MR, Sweet PM, McDonnell PJ. Changes in gene expression in response to mechanical strain in human scleral fibroblasts. Exp Eye Res 2004; 78:275-84.
66. Yang P, Agapova O, Parker A, Shannon W, Pecen P, Duncan J, Salvador-Silva M, Hernandez MR. DNA microarray analysis of gene expression in human optic nerve head astrocytes in response to hydrostatic pressure. Physiol Genomics 2004; 17:157-69.
67. Ahmed F, Brown KM, Stephan DA, Morrison JC, Johnson EC, Tomarev SI. Microarray analysis of changes in mRNA levels in the rat retina after experimental elevation of intraocular pressure. Invest Ophthalmol Vis Sci 2004; 45:1247-58.