Genes Upregulated in the Human Trabecular Meshwork in Response to Elevated Intraocular Pressure

Abstract

purpose. To identify genes upregulated in perfused, intact human trabecular meshwork (TM) in response to elevated intraocular pressure (IOP).

methods. Two pairs of anterior segments of normal human eyes from postmortem donors were placed in culture and perfused 24 hours at constant flow (3μ l/min). After reaching baseline, the flow of one eye from each pair was raised to obtain an incremental pressure (ΔP) of 50 mm Hg for 6 hours. The anterior segments were then quickly frozen in liquid nitrogen, and their TMs were dissected for RNA extraction. SMART cDNA libraries were generated from control and high-pressure human TM RNAs and hybridized to sets of identical high-density cDNA gene arrays. These arrays contained 18,376 human expressed sequence tags (ESTs), corresponding to both characterized and unknown genes. Differentially expressed genes were identified by different-intensity hybridization signals and confirmed by semi-quantitative polymerase chain reaction.

results. Eleven genes were found to be consistently upregulated in the human TM by elevated IOP: interleukin-6, preprotachykinin-1, secretogranin-II, cathepsin-L, stromelysin-1, thymosin-β₄, α-tubulin,α B-crystallin, glyceraldehyde-3-phosphate dehydrogenase, metallothionein and Cu/Zn superoxide dismutase. The products of these genes are involved in vascular permeability, secretion, extracellular matrix remodeling, cytoskeleton reorganization, and reactive oxygen species scavenging.

conclusions. Elevated IOP induced specific upregulation of 11 physiologically relevant genes. On the basis of their known activities, the products of each of these genes might predict homeostatic mechanisms similar to those involved in the regulation of blood vessel permeability. We hypothesize that similar mechanisms might be involved in regulating flow through Schlemm’s Canal endothelium.

Glaucoma, a disease characterized by the degeneration of retinal ganglion cells, is commonly associated with an elevated intraocular pressure (IOP). ^{1 2} This increase in IOP results from an increase in resistance to aqueous humor outflow. ³ In normal physiological conditions, aqueous humor is secreted by the ciliary body, flows through the pupil into the anterior chamber, and leaves the eye, mostly through the conventional outflow pathway, the trabecular meshwork (TM) and Schlemm’s Canal (SC). Earlier experimental evidence suggests that the normal outflow resistance, responsible for maintaining IOP, resides deep in the conventional outflow pathway at the level of the juxtacanalicular tissue and/or in the endothelium of the inner wall of the SC. ³ The cause for the increase in resistance leading to an elevated IOP in the glaucomatous eye remains unknown but it is believed to reside at the same locus. ³ Some have hypothesized that there are abnormalities of the turnover of the extracellular matrix (ECM) in glaucoma. ^{4 5}  

Although most aqueous humor probably flows through the TM by a mechanism not dependent on cellular energy, ^{6 7} cells of the outflow pathway could modulate the rate of this flow through a variety of mechanisms. These would include (1) phagocytosis of debris that could cause mechanical blockage, ^{8 9 10} (2) secretion of proteins and enzymes that could alter the composition of the ECM, ⁵ (3) changes in cytoskeletal organization that could affect cell shape and thus influence the dimensions or direction of the outflow pathway for aqueous humor, ^{11 12} and/or (4) release of factors that could influence the permeability of the inner walls of the SC. 

To maintain a physiological IOP in the living eye, it would be desirable for the TM–SC system to have in place feedback mechanisms capable of altering outflow resistance. Such mechanisms could include both sensing mechanisms for differential pressure or fluid distortion, and response mechanisms that would be able to send signals to neighboring cells. However, although there is morphologic evidence for the possible presence of mechanoreceptors at nerve endings in adjacent tissues, ^{13 14} there is very little information at the molecular level of possible homeostatic responses to changes in IOP. 

One way to begin understanding the possible molecular mechanisms that could regulate IOP would be to study those genes that are upregulated in the outflow pathway when the tissue is subjected to a high-IOP insult. We applied an exponential polymerase chain reaction (PCR)-gene array hybridization strategy, which is among the current techniques available to study such differential gene expression. To our knowledge no such study has been performed on intact human TM tissue. One of the reasons for this lack of data has been the technical difficulty associated with the small size of the human TM. To date, this problem has precluded the construction of good human TM cDNA libraries and the characterization of relevant genes. 

We have recently developed in our laboratory a strategy that allows us to perform the analysis of differential gene expression in very small tissue samples. ¹⁵ In this study, we have applied these new techniques to examine the pattern of genes upregulated by a high-IOP (HIOP) stimulus to the human TM, lasting for 6 hours in the perfused human anterior segment organ culture. ^{16 17 18}  

Our results show that changes in IOP do induce changes in gene expression in the human TM. Some of these TM genes, activated by an increase in IOP, code for proteins involved in regulation of inflammation, secretion, ECM digestion, cytoskeleton organization, and responses to oxidative and other types of stresses. Genes involved in similar functions have been shown to be very important in the regulation of permeability of vascular endothelium. 

Methods

Perfused Human Anterior Segment Organ Culture

Two pairs of normal human eyes were obtained from The National Disease Research Interchange (NDRI), a nonprofit organization engaged in the procurement and distribution of human tissues for biomedical research in the United States. These were obtained with the signed consent of the patients, and procedures followed the Tenets of the Declaration of Helsinki. The age of the donors was 69 years (individual 1) and 92 years (individual 2). The eyes were dissected within 30 to 40 hours of death, and neither subject had been diagnosed with glaucoma. Organ cultures were prepared as described in earlier techniques. ^{16 17 18} Briefly, eyes were bisected at the equator and the lens, iris, and vitreous were removed. The anterior segment was then clamped to a modified Petri dish and perfused at 3μ l/min constant flow using a microinfusion pump. Culture medium was Dulbecco’s modified Eagle’s medium containing 100 U/ml penicillin, 0.1 mg/ml streptomycin, 170 μg/ml gentamicin, and 250 μg/ml amphotericin B (DMEM+ ). Anterior segments were maintained at 37°C and 5% CO₂. IOPs were monitored continuously with a pressure transducer connected to the dish’s second cannula and recorded with an automated computerized system. After reaching baseline values (usually 24 hours), the flow of one eye was raised to obtain aΔ P of 50 mm Hg for a period of 6 hours. The flow of the contralateral eye, which served as the control, was maintained at 3 μl/min. At the end of the experiment, anterior segments were frozen in liquid nitrogen within 2 minutes of turning off the perfusion pumps and stored at− 70°C for RNA extraction. 

RNA Extraction and Construction of Exponential cDNA Probes

RNA extraction and construction of cDNA libraries were performed basically as detailed in Gonzalez et al. ¹⁵ Briefly, TM tissues were obtained under a dissecting microscope from the frozen anterior segments before the complete thawing of the specimen. The isolated tissue was placed into a 1.5-ml microcentrifuge tube containing 350 μl of guanidine thiocyanate buffer, homogenized with a disposable pestle, and loaded onto a QIAshredder column (QIAGEN, Chatsworth, CA). Extraction continued using the RNeasy kit (QIAGEN), and RNA molecules selectively bound to the silica gel base were eluted with 30 μl RNAase-free water. 

One fourth of the RNA sample was lyophilized and used as template for the construction of the exponential cDNA libraries as described. ¹⁵ Briefly, the total RNA was reverse-transcribed directly using the SMART PCR cDNA synthesis kit (Clontech, Palo Alto, CA) according to the manufacturer’s protocol. Annealing was conducted using a modified oligo(dT) at 70°C for 2 minutes in the presence of the SMART II oligonucleotide in total volume of 5 μl. The reaction was followed by the addition of Superscript II (200 units), RNase H⁻ reverse transcriptase (RT) (Gibco-BRL, Gaithersburg, MD) and incubated at 42°C for 1 hour. The reaction was stopped by adding 40 μl of Tris-EDTA buffer and heating at 72°C for 7 minutes (final volume, 50 μl). Representative double-stranded cDNAs were then generated by exponential PCR amplification. The optimal number of cycles for each sample was determined by analyzing the PCR products of a series PCR amplification using different numbers of cycles. 

Two microliters from the 50-μl single-stranded cDNA stocks were amplified in 100-μl reactions using the SMART PCR primer and the predetermined exponential number of cycles. Amplified SMART cDNAs from the control and high-pressure TMs were purified using the QIAquick PCR Purification kit (QIAGEN) and labeled by random priming (Ready To Go DNA labeling kit; Pharmacia Biotech, Piscataway, NJ). 

Differential Hybridization to Human cDNA Arrays

Approximately 1 to 3 × 10⁸ cpm from control and HIOP TM cDNAs were hybridized to identical membranes containing high-density human cDNA arrays. The membrane arrays used here were the human gene discovery array (GDA) version 1.2 (Genome Systems, St. Louis, MO), which contain 18,376 nonredundant human cDNA clones from the I.M.A.G.E. collection. ¹⁹ Hybridizations were performed at 42°C in a 10-ml solution of 50% deionized formamide, 6× sodium saline citrate (SSC), 5× Denhardt’s, and 1% sodium dodecyl sulfate (SDS) as previously described. ¹⁵ Ten milligrams of sheared salmon sperm DNA (Research Genetics, Huntsville, AL) and 50 μg of Cot1 DNA (Gibco-BRL) were used as carriers in all hybridization steps. Membrane filters were washed four times in 500 ml of 2× SSC, 1% SDS at 65°C for 20 minutes and exposed to x-ray film (Eastman Kodak, Rochester, NY) at 4°C. The time of exposure of the membranes was adjusted to obtain the same general intensity in both of them. ¹⁵  

Identification of Differentially Expressed Genes

Within one small square of the GDA filter grid, Genome Systems spotted each individual clone twice. Positive hybridization signals corresponding to true positive clones are those where the intensity of the two spots corresponding to the same gene are equivalent. Comparison between the same clone in the two membranes was done taking into account the overall positive signal of the membrane. Identification of each specific clone was conducted by matching the spot location in the membrane filter grid with the Genome Systems data bank (http://www.genomesystems.com/GDA/geneID.html). 

Confirmation of the Differentially Expressed Genes

Differential expression of the genes selected by hybridization to the arrays was verified by semi-quantitative PCR of the control and HIOP cDNAs. Target genes were compared from the two conditions using primers specifically designed to anneal with sequences located in different exons. 

PCR reactions were performed in a Perkin Elmer, GeneAmp PCR System 2400 with advantage cDNA polymerase mix (Clontech, Palo Alto, CA). Amplifications were conducted using a program with a precycle of 94°C for 15 seconds, cycles of 94°C for 15 seconds, 60°C for 30 seconds, 72°C for 30 seconds, and an extension of 72°C for 7 minutes. Normalization of the SMART cDNA samples from normal and HIOP TMs was carried out with two genes that did not show differences of hybridization in the cDNA arrays: hypoxanthine phosphororibosyl transferase (HPRT) and major histocompatibility complex class I (MHC-I). ¹⁵ The selection of internal standards that do not show changes in expression in the arrays is particularly critical, because the use of some other genes frequently chosen as controls could be misleading. Thus, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a commonly used internal standard, has been shown to be upregulated in a variety of experimental conditions. ²⁰ Amplification of each of the differentially expressed genes was performed at exponential phase (number of cycles determined for each tested gene) using the normalized template values. PCR products were analyzed in 1.5% Super AcrylAgarose (DNA Technologies Inc., Gaithersburg, MD) gels containing 0.25 μg/ml of ethidium bromide. Images of the gels were captured with a Progress-Research-3012 digital camera (Konitron Elektronik, Eching, Germany) and quantified using the Image Analysis Software (NIH). Identification of the analyzed PCR products was confirmed by DNA sequencing. Each amplified fragment was cloned into the pCR2.1-TOPO cloning vector (Invitrogene, Carlsbad, CA) and sequenced with an M13 universal reverse primer. 

Table 1 summarize the sequences and sources of all primers used in this study as well as the expected size of the amplified products of the induced genes. 

Results

Generation of High-IOP–Treated TM Tissue

To examine the differential responsiveness of the TM to fluid mechanical forces, we subjected human postmortem anterior segments to a higher flow for 6 hours. This elevated flow resulted in an additional increase in IOP of 48.8 mm Hg (baseline, 7.9 mm Hg) for the first pair and 50.7 mm Hg (baseline, 25.4 mm Hg) for the second one. The contralateral eyes were maintained at the physiological flow rate of 3μ l/min, and their pressures at the beginning and end of the experiment were 1.3/1.3 mm Hg for pair 1 and 13.0/11.3 mm Hg for pair 2. 

Exponential cDNAs Libraries

Analytical amplifications of the first-strand cDNA from normal and HIOP TM tissues were conducted as described in the Methods section. The exponential phase was determined in the gel by the disappearance of distinct bands and increased high-molecular-weight smear in the lanes corresponding to the next set of cycles. Figure 1 shows the results of the step-wise amplification of both cDNA samples from individual 2. The number of PCR cycles selected (exponential phase) were 20 and 17, respectively, for the control and HIOP from the first individual and 21 for both eyes in the second (shown in Fig. 1 ). After this number of cycles, the cDNA molecules corresponding to the most abundant mRNA populations have ended their exponential growth. Preparative exponential libraries were then obtained from both individuals at the same number cycles as that of the analytical sample. 

Differentially Expressed Genes

Approximately 80 ESTs of the gene arrays showed clear and consistent differences in the intensity of hybridization between the control and the HIOP cDNAs. The majority of the spots represented upregulated genes. Despite the incidental low pressure of one of them, the contralateral eyes from individuals 1 and 2 showed a very similar pattern of hybridization. Analysis of the spots using the Genome Systems sequence data bank identified 11 upregulated genes and classified the remaining ones as cDNAs corresponding to unknown genes or genes of unknown function. None of the downregulated ESTs corresponded to genes of known function. The information of the uncharacterized ESTs has been stored in our laboratory for future analysis as soon as the characterization of those genes becomes available. For this study, we concentrated our efforts in identifying and confirming the differential expression of the 11 upregulated genes with a well-characterized function. The proteins encoded by these genes can be grouped into molecules with five different roles. The first two genes, interleukin-6 (IL-6) and preprotachykinin-A (PPT-A) code for signaling molecules involved in the regulation of the inflammatory response. One, secretogranin-II, is a marker of the regulated secretory pathway. The next two, cathepsin-L and stromelysin-1, are enzymes of the extracellular proteinase family. Two of the proteins, thymosin-β₄ and α-tubulin, are involved in the organization of the cell’s cytoskeleton. The last four upregulated genes encode αB-crystallin, GAPDH, metallothionein, and Cu/Zn-dependent superoxide dismutase (Cu/Zn SOD), all proteins known to be induced by different types of stresses. Some of the spots that showed different hybridization corresponded to different ESTs from the same gene, reaffirming the true induction of those genes. In the membranes, there were 9 different EST spots corresponding to metallothioneins, 2 for the α-tubulin, and 3 for the αB-crystallin. 

The results of the hybridization of the SMART cDNAs from the normal and HIOP TMs to the gene arrays are shown in Figure 2 . The membranes shown correspond to samples obtained from the anterior segment cultures from donor 1, a 69-year-old, white man. Reproducible patterns were obtained with the cDNA from donor 2, a 92-year-old white woman, subjected to the same pressure insult (not shown). Figure 2A shows the fields containing HPRT and MHCH-I, two nondifferentially expressed housekeeping genes used as internal controls. Figure 2B shows the partial six-square fields of gene membrane arrays containing the hybridization spots corresponding to the 11 induced genes. With the exception of the α-tubulin and αB-crystallin genes, the small squares containing the spots of the upregulated or equally expressed genes are located in the central square of each depicted field. Positions of the double spots corresponding to each gene are shown in Figure 2C . 

Confirmation of Differentially Expressed Genes

The 11 cDNAs that appeared to be preferentially expressed in the GDA blots were further identified by sequencing their specific products from semi-quantitative PCR reactions. In every case, the nucleotide sequence corresponded to the clone identified on the Genome Systems data bank. 

The differential expression between both conditions was verified by semi-quantitative PCR on normalization of the amount of template present in the HIOP and normal pressure SMART cDNA samples. Normalization was accomplished by using primers from housekeeping HPRT and MHC-I genes (same intensity in the hybridization membranes) and for the MHC-I as follows: (a) cDNAs from control and HIOP TMs were diluted 10 times and amplified (1 μl) at different number of cycles with the MHC-I primers; (b) resulting PCR products were analyzed by gel electrophoresis and the cycle number at which amplifications were exponential for both conditions was determined (Fig. 3A ); (c) 1 μl from a 1/10 dilution of the HIOP and 1 μl from serial dilutions (1/10 to 1/60) of the normal pressure cDNAs were amplified at the determined exponential cycle number; and (d) resulting PCR products were analyzed by gel electrophoresis (Fig. 3B) , and band intensities were compared by densitometry. Final normalization values of the templates were 1:6 (control:HIOP) for individual 1 and 5:1 (control:HIOP) for individual 2. For corroboration, template normalization was repeated using HPRT-specific primers in a similar manner (not shown). Results were identical, as were the ones obtained for MHC-I. 

The PCR quantification of the 11 genes was then analyzed on reactions stopped at the exponential phase (determined for each tested gene) and using the normalized template values. The primers chosen to analyze the PPT-A mRNA amplify together the three known alternatively spliced products of this gene, α, β, and γ. ²¹ The transcriptα codes for the substance P (SP) precursor, whereas the β and γ forms code for the precursors of substance K (neurokinin A). ²² All three products were found to be present in the TM. Of the three, the PPT-A β transcript was the more abundant in both individuals (approximately 80% of the total PPT-PCR product), whereas PPT-Aα was approximately 20%, and PPT-Aγ, though present, was hardly visible in the ethidium bromide–stained gels. Given the high similarities of the 9 metallothioneins identified in the membranes, we also chose, for this gene, primers that would amplify undistinctively all the isoforms. Because there is not clear evidence on what the differences in function between them may be, we made no attempt to further characterize the different isoforms present in the TM. 

Figure 4 shows the final results of the semi-quantitative PCR in the normal and HIOP cDNAs of the two individuals analyzed. The relative level of induction of each gene was different in both individuals examined, but importantly, all 11 genes analyzed were upregulated with pressure in both of them. 

Table 2 includes a summary of the 11 differentially expressed genes, the location of their respective ESTs in the gene array, the number of cycles used for semi-quantitative PCR and the relative level of induction for both individuals analyzed. 

Discussion

One of the drawbacks in using primary cultured cells for the study of differential expression derives from the fact that once in culture, the cells may change considerably the expression of some of their genes. Recently, we have shown that certain TM genes such asα B-crystallin and fibronectin are down- and upregulated, respectively, in cultured cells, whereas the expression of others, such as elastase-IIIA and osteopontin, completely disappear. ¹⁵ In addition, all mechanical forces acting on the outflow pathway tissue ^{23 24} are not present in traditional cell culture techniques. Recent studies have applied mechanical stretch to human TM cells. ²⁵ However, in contrast to cultured cells, the organ culture model used in these studies ^{18 26} preserves the architecture of the outflow tissue and keeps intact potential sites for differential pressure effects. The different layers of the TM and SC that may be altered by mechanical stress are maintained. 

In this study, we have constructed cDNAs libraries from the TMs of paired normal and HIOP perfused human anterior segments. We have confirmed our results by repeating the library construction and screening in a second, unrelated individual. Similar mechanical stresses were likely applied to the experimental eyes of each of the two pairs studied (ΔDPs = 48.8 mm Hg for the first and 50.7 mm Hg for the second), and the described 11 genes were induced in both cases. We also observed that the level of induction of these genes was different for each of the individuals studied. That result is not unexpected because in studies with human tissues, individual variations due to, for example, age, race, and sex, are unavoidable. The comparison of cDNA libraries from normal and HIOP in paired eyes from the same individual does allow us, however, to identify and distinguish relevant inductions. 

The first two of the genes upregulated with elevated pressure in the TM, IL-6 and SP, are well-known markers of inflammation. ^{27 28} One of the important actions of these two molecules is their ability to increase vascular permeability in vitro ^{29 30 31} and in vivo. ^{32 33 34} This effect may be produced by mechanisms such as the induction of intercellular gaps, ^{31 34} increased number of pinocytic vesicles and vacuoles, ³⁵ and release of ECM proteinases. ^{36 37} IL-6 can be induced by ROS, ³⁸ cyclic adenosine monophosphate (cAMP), ³⁹ forskolin, ⁴⁰ transforming growth factor beta ⁴¹ and bradykinin, ⁴² and it is noteworthy that most of these agents also influence aqueous humor outflow facility. ^{42 43 44 45 46 47} SP can be induced by noxious stimuli and in turn promote the release of prostaglandin E₂, ⁴⁸ inflammatory cytokines (including IL-6), ⁴⁹ and superoxide. ⁵⁰ Initially considered to be exclusively of neuronal origin, ⁵¹ SP has been subsequently identified in a variety of nonneuron cell types. ^{52 53 54} In the eye, SP has been described inside the nerves of ocular tissues, including those of the TM. ^{55 56 57} Because our organ culture tissue preparation would exclude neuronal bodies, our results suggest a potential endogenous production of this peptide by TM cells. Interestingly, human umbilical vein endothelial cells release SP in response to increased flow. ⁵⁸  

Given the functions of IL-6 and SP in the vascular system, it is very attractive to speculate that these two molecules might play a similar role in relation to SC. The known embryological vascular origin of SC ^{59 60 61 62} might support such a hypothesis. We postulate that, in response to elevated IOP, the cells of the TM might release factors to which the SC could respond in a way similar to other vascular cells. Permeability would be momentarily increased and physiological IOP would be restored. This SC response might have been so programmed during morphogenesis. 

Another induced gene, secretogranin-II, codes for a protein involved in the regulation of the secretory pathway. ⁶³ Secretogranin-II was initially described in the anterior pituitary ⁶⁴ and later was shown to be present throughout the endocrine and nervous systems. ⁶⁵ The specific functions of secretogranin and its derived neuropeptide secretoneurin ⁶⁶ are not clear. A potential role in the regulation of aqueous humor outflow has been proposed, based on secretogranin expression in the ciliary epithelium and the presence of secretoneurin in the aqueous humor. ⁶⁷ Our results on the increase in secretogranin-II mRNA in the TM cells in response to IOP could simply represent an activation of the secretory activity of these cells. This enhancement could influence the release of other substances, such as ECM proteinases or signaling molecules such as SP and IL-6. 

Remodeling of ECM by the secretion of matrix metalloproteinases (MMPs) (including cathepsin-L and stromelysin-1) has been known to play a major role in regulating vascular permeability in other systems. For instance, invading tumor cells upregulate expression of MMPs to degrade ECM synthesized by vascular endothelial cells, ⁶⁸ and connective tissue cells secrete stromelysin to regulate protein permeability. ⁶⁹ In the eye, MMPs and their inhibitors are present in aqueous humor, ⁷⁰ and perfusion of human anterior segments with MMPs produces an increase in outflow facility. ⁷¹ There is also a reported excess accumulation of ECM components in glaucomatous eyes ⁴ as well as a reported increase in stromelysin activity after laser trabeculoplasty. ⁷² Our finding that stromelysin and cathepsin are upregulated by HIOP potentially supports the concept that these MMPs might be involved in aqueous humor outflow resistance regulation. 

The induction of thymosin-β₄ and α-tubulin reinforces the potential importance of the cytoskeleton on the regulation of IOP. Thymosin is an actin-binding protein that triggers destabilization of stress fibers, ⁷³ and α-tubulin is a component of microtubules. Changes in actin and tubulin have been associated with changes in cell contractility. ^{74 75} Our results might suggest that in response to HIOP, TM cells might decrease contractility and tension (relax). Several studies have shown that cytoskeletal-related drugs that induce either relaxation or contraction are associated with an increase in outflow facility. ^{11 76 77} Perhaps there are many potential regulatory mechanisms relating to temporal changes in cellular tension. An alternative explanation to a potential influence in IOP by the induction of these genes, would be that an increase in microtubule formation might also facilitate cell secretion ⁷⁸ and thereby contribute to remodeling of the ECM or release of other extracellular enzymes. In this regard, it may be important that TIGR/MYOC, a protein genetically linked to glaucoma, ⁷⁹ has been localized to cell vesicles ⁸⁰ and therefore may be involved in secretion. 

The last set of HIOP upregulated genes, αB-crystallin, GAPDH, metallothioneins, and Cu/Zn SOD, have all been reported to be induced by different types of stress in several systems. αB-Crystallin, localized in the juxtacanalicular region, is induced after heat shock, oxidative stress, ^{81 82} and mechanical stretching of primary TM cells. ⁸³ GAPDH is commonly used as mRNA internal standard with the assumption that its expression levels remain relatively constant in different experimental conditions. However, recent data demonstrated upregulation of GAPDH in different tissues in response to a variety of stress conditions. In particular, it has been shown to be induced by oxidative stress in endothelial cells. ⁸⁴ Metallothioneins (MTs), cysteine-rich proteins with metal-binding activity, are inducible by metals, glucocorticoids, inflammatory stress signals, and oxidative stress. ^{85 86 87 88} Superoxide dismutase (Cu/Zn SOD) expression also responds to oxidative stress. ⁸⁹ The activation of these genes in the TM ^{90 91} could be an indication of the importance of regulating the TM redox state. It is also noteworthy that upregulation of redox-sensitive genes has been reported in human vascular endothelial cells exposed to oscillatory and shear stress. ⁹²  

Different types of mechanical stress activate both prooxidant and antioxidant compensatory genes. Thus, cyclic strain ^{93 94} and shear flow stresses ^{92 95} induce the production of ROS that may subsequently act as signaling molecules and influence cell adhesion. ⁹⁶ The cyclic strain is particularly interesting, because this type of stress is potentially similar to the one applied to the TM in situ in our studies. It is conceivable that HIOP-induced TM cells would respond in a similar manner and generate ROS. ROS (superoxide free radicals and H₂O₂) also mediate the production of MMPs during reorganization of cell shape ⁹⁷ and cell retraction with formation of intercellular gaps. ⁹⁸ To this effect, perfusion of low concentrations of H₂O₂ in presence of a catalase inhibitor has been reported to produce an increase in outflow facility. ⁹⁹  

In summary, we have found that a given increase in IOP (ΔDP ≅ 50 mm Hg, for 6 hours) in a perfused human anterior segment model, can induce selective upregulation of 11 physiologically relevant TM genes (IL-6, PPT-A, secretogranin, cathepsin, stromelysin, thymosin, α -tubulin,α B-crystallin, GAPDH, MTs, and Cu/Zn SOD). Based on their known activities, the products of each of these genes would be predictive to have a potential effect in aqueous humor outflow resistance. These genes involved in vascular permeability, secretion, degradation of ECM, reorganization of the cytoskeleton, and ROS scavenging all represent potential candidates for the regulation of IOP at different levels. Their coordinate induction provides a possible mechanism to maintain homeostasis of the outflow pathway system. Given the vascular nature of SC, we hypothesize that some of these mechanisms may have evolved from those that regulate the permeability of the vascular endothelium. We believe that these and other yet to be characterized genes that exhibit differential regulation in the outflow pathway may provide new insights into the potential regulation of resistance to aqueous humor outflow in the human eye. 

 Supported by National Eye Institute Grants EY11906 and EY01894, the American Health Assistance Foundation, and Research to Prevent Blindness. TB is a Jules and Doris Stein Research to Prevent Blindness Professor Awardee.

 Submitted for publication June 4, 1999; revised August 26, 1999; accepted September 15, 1999.

 Commercial relationships policy: N.

 

Table 1.

 

Oligonucleotides Used for Semi-Quantitative PCR

Gene Product	Sequence (5′-3′)	Product Size (bp)
HPRT*	5′-GGCGTCGTGATTAGTGATGATGAACC-3′	467
	5′-CTTGCGACCTTGACCATCTTTGGA-3′
MHC-I*	5′-GCAAGGATTACATCGCCCTGAACGAG-3′	566
	5′-CATCATAGCGGTGACCACAGCTCCAA-3′
IL-6	5′-CTGGTCTTTTGGAGTTTGAGGTATACC-3′	294
	5′-CCATGCTACATTTGCCGAAGA-3′
PPTA-1	5′-GCAGCGACCAGATCAAGGAGGAA-3′	208, 217, 262
	5′-GCATTGCACTCCTTTCATAAGCCA-3′
Secretogranin-II	5′-ATGGCATATGAAAACCTGAACGAC-3′	774
	5′-GTCCAGTACTCCACAAAACGCTACA
Cathepsin-L	5′-CGTAAAGATGGCCAAAGACCG-3′	191
	5′-CACTCGGATCTTCAATGATTCGA-3′
Stromelysin-1	5′-CATCAGAGGAAATGAGGTACGAGC-3′	478
	5′-AAGTTCCCTTGAGTGTGACTCGAG-3′
Thymosin-β4	5′-CATGTCTGACAAACCCGATATGG-3′	170
	5′-GGAATGTACAGTGCATATTGGCG-3′
α-Tubulin*	5′-CACCCGTCTTCAGGGCTTCTTGGTTT-3′	527
	5′-CATTTCACCATCTGGTTGGCTGGCTC-3′
αB-Crystallin	5′-TCACCTAGCCACCATGGACATCGCCA-3′	540
	5′-CAAAAGCTTATTACTATTTCTTGGGGGCTG-3′
GAPDH*	5′-ACCACAGTCCATGCCATGAC-3′	452
	5′-TCCACCACCCTGTTGCTGTA-3′
Metallothioneins	5′-GCAAAGAGTGCAAATGCACCTC-3′	124
	5′-TCAGGCACAGCAGCTGCACT-3′
Cu/Zn SOD	5′-AGGGCATCATCAATTTCGAGC-3′	429
	5′-CAAGGGAATGTTTATTGGGCG-3′

*  Purchased from Clontech.

Figure 1.

 

Selection of the number of cycles for the exponential PCR amplification of the cDNA from control (C) and high-intraocular pressure (HIOP) trabecular meshwork tissues (individual 2). Fifteen-microliter aliquots from the amplification reactions after 17, 20, and 23 cycles were analyzed in a 1.5% Super AcrylAgarose gel (DNA Technologies, Inc.). Overcycled reactions can be identified by the disappearance of the distinct bands corresponding to abundant cDNAs and the increased smear molecular weight. Preparative amplifications were subsequently performed at 21 cycles for both eyes of this individual. Line M: 200 ng of 1-kb DNA ladder size markers (Gibco BRL).

Figure 2.

 

Differential hybridization of the GDA filter gene arrays with the SMART cDNA probes from the control and high-intraocular pressure (HIOP) trabecular meshwork tissues. Partial areas of different fields are shown, with small squares containing the spots corresponding to the induced genes. For easy identification, with the exception of αB-crystallin and α-tubulin, the identified genes are located in the central square of each depicted field. (A) Two noninduced genes, HPRT and MHC-I, used as internal standards and for normalization of the SMART cDNA samples. Their GDA coordinates are as follows: HPRT: field 6, location o17, pattern 8; and MHC-I: field 4, location g6, pattern 4. (B) Eleven genes upregulated in the HP sample. The GDA coordinates for each of the genes are as follows: IL-6: field 1, location m14, pattern 7; PPT-A: field 3, location o11, pattern 6; secretogranin-II: field 2, location d11, pattern 2; cathepsin-L: field 1, location a24, pattern 7; stromelysin-1: field 4, location h2, pattern 7; thymosin-β₄: field 1, location n4, pattern 6;α B-crystallin: field 1, location m5, pattern 6; α-tubulin: field 1, location h10, pattern 7; GADPH: field 1, location h11, pattern 8; metallothionein: field 5, location k2, pattern 4; Cu/Zn SOD: field 6, location i10, pattern 6. (C) Each small square of the GDA filters includes eight different genes. Each gene is spotted twice according to the pattern number shown in this panel.

Figure 3.

 

Normalization of the control and high-intraocular pressure (HIOP) SMART cDNAs (individual 2, MHC-I gene). (A) Serial PCR amplifications (21–39 cycles) of 1-μl aliquots from 10× dilutions of control and HIOP SMART cDNAs using primers from control gene MHC-I. PCR cycle 29 resulted to be on exponential phase in the cDNAs from both tissues. (B) One microliter from a serial dilution (1/10 to 1/60) of the control SMART cDNA and 1 μl from a 1/10 dilution of the HIOP SMART cDNA were amplified for 29 cycles. Their resulting PCR products are analyzed by 1.5% Super AcrylAgarose gel electrophoresis. Densitometry measurements of the HIOP band intensity at 1/10 were equal to that of control at 1/50 (normalization value control:HIOP, 5:1). Same normalization values were obtained using the primers from the second control gene HPRT (not shown).

Figure 4.

 

Confirmation of the differential expression in both individuals by semi-quantitative PCR. Amplification of the 11 genes selected by hybridization to gene arrays was performed using their specific primers shown in Table 1 and their predetermined exponential number of cycles shown in Table 2 . Two-lane numbers show the amplification from the control (left) and HIOP (right) SMART cDNAs for each of the genes analyzed. 1, MHC-I (control); 2, HPRT (control); 3, IL-6; 4, PPT-A-1; 5, secretogranin-II; 6, cathepsin-L; 7, stromelysin-1; 8, thymosin-β₄; 9, α-tubulin; 10,α B-crystallin; 11, GADPH; 12, metallothionein; 13, Cu/Zn SOD. Line M: 200 ng of 1-kb DNA ladder size markers (Gibco BRL). All 11 genes were induced in both individuals.

Table 2.

 

Genes Upregulated in Human TM in Response to High Intraocular Pressure

Genes	Location in the GDA Filters	Cycles Used for Semi-quantitative PCR for Individual			Approximate Level of Induction for Individual
				1	2	1	2
Internal controls
HPRT	6.o.17.8	26	29	1×	1×
MHC-I	4.g.6.6	27	29	1×	1×
Induced genes
IL-6	1.m.14.7	24	22	6×	2×
PTTA-1	3.o.11.6	19	20	2×	3×
Secretogranin-II	2.d.11.2	24	22	4×	2×
Cathepsin-L	1.a.24.7	24	24	4×	5×
Stromelysin-1	4.h.2.7	18	20	6×	2×
Thymosin-β4	1.n.4.6	20	21	5×	10×
α-Tubulin	6.j.23.7 1.h.0.7	21	22	3×	5×
αB-Crystallin	5.c.22.3 2.b.15.7 1.m.5.6	18	22	5×	4×
GAPDH	1.h.11.8	20	24	2×	3×
Metallothioneins	5.k.2.4 1.i.17.5 6.i.19.4 5.p.1.4 1.e.18.7 5.k.2.4 4.d.14.1 4.d.11.5 3.p.17.5	18	22	2×	6×
Cu/Zn SOD	6.i.10.6	23	20	3×	3×

The authors thank Laura Leigh Rowlette for her excellent technical assistance, especially with the perfused organ culture system.