Preventive Effects of Omega-3 and Omega-6 Fatty Acids on Peroxide Mediated Oxidative Stress Responses in Primary Human Trabecular Meshwork Cells

Pathologic processes in glaucoma include increased apoptosis, accumulation of extracellular material in the trabecular meshwork and optic nerve, condensations of the cytoskeleton and precocious cellular senescence. Oxidative stress was shown to generate these alterations in primary ocular cells. Fatty acids omega-3 and -6 are alleged to constitute a prophylaxis against these deleterious effects. Here, we tested actual preventive effects omega-3 and -6 against peroxide induced stress responses in primary human trabecular meshwork cells. Changes of mitochondrial activity, proliferation, heat shock proteins, extracellular matrix components, and inflammatory markers were evaluated. Alterations of the cytoskeleton were evaluated by phalloidin labeling. Here we report a repressive effect of omega-6 on metabolic activity and proliferation, which was not detected for omega-3. Both agents were able to prevent the anti-proliferative effect of H2O2, but only omega-3 prevented metabolic repression. Expression of heat shock protein 27 was unaltered by both fatty acids, whereas heat shock protein 90 was significantly induced by both. Omega-6 increased fibronectin and connective tissue growth factor synthesis, as well as the amount of secreted fibronectin. Omega-3, instead, induced plasminogen activator inhibitor 1 synthesis. H2O2 further increased fibronectin production in omega-6 supplemented cells, which was not the case in omega-3 treated cells. H2O2 stimulation of plasminogen activator inhibitor 1 and connective tissue growth factor was repressed by both fatty acids. Both fatty acids appeared to abolish H2O2 mediated stimulation of nuclear factor κB and IL-6, but not IL-1α and IL-8. H2O2 induced formation of cross-linked actin networks and stress fibers, which was reduced by preemptive application of omega-3. Omega-6, in contrast, had no protective effect on that, and even seemed to promote condensation. Based on the observed side effects of omega-6, omega-3 appears to be the more beneficial fatty acid in respect of prophylactic intake for prevention of a glaucomatous disease.


Introduction
The trabecular meshwork (TM) accounts for about 70-90% of total aqueous humor (AH) outflow from the anterior chamber in the adult human eye. Moreover, the TM constitutes an outflow resistance and thereby determines the intraocular pressure (IOP). Increase of outflow resistance and IOP, eventually, is a major risk factor in primary open angle glaucoma (POAG) [1]. POAG patients exhibit specific morphological peculiarities as accumulations of extracellular matrix material (ECM) and condensation or clustering of the cytoskeleton. Frequently, signs of subclinical inflammation are reported. Moreover, deregulations on the cellular level affecting signaling pathways regulating apoptosis, cellular senescence and cell cycle control were observed in ocular cells of POAG patients [1,2,3,4].
In vitro, such glaucoma-characteristic alterations were detected due to artificially induced oxidative stress in cultured human TM cells (hTM) [5,6,7,8]. This lead to the hypothesis that oxidative stress might either be responsible or at least involved in the onset and progression of the TM changes observed in POAG.
Supporting that, evidence for a higher frequency of impaired mitochondrial function has been reported in POAG patients [8,9,10,11]. Tanwar et al. (2010) identified conspicuous variations in the mitochondrial DNA of patients with primary congenital glaucoma affecting the respiratory chain and leading to increased production of reactive oxygen species (ROS) [12]. Izzotti et al. (2003) found a deletion mutation of the glutathione S-transferase Mu 1 (GSTM1) gene in POAG patients which correlated with an increase of 8-hydroxydeoxyguanosine (8-OH-dG), a marker of DNA damage and precocious senescence, in TM cells [13]. Sacca et al. (2005) described a distinct correlation between oxidative DNA damage in the TM, visual field reduction and IOP increase [14]. Previously, Abu-Amero et al. (2006) had already observed a significant reduction of the mitochondrial respiratory activity in patients with POAG [9]. Mitochondrial dysfunctions and a reduced mitochondrial respiratory activity favor accumulation of ROS. Studies to evaluate the total reactive anti-oxidative potential of the aqueous humor state a significantly decreased anti-oxidative capacity in patients with POAG [15,16,17,18]. All these findings suggest a constantly heightened oxidative stress level in patients with POAG [1]. Consequently, the role of oxidative stress in the pathogenesis of POAG has become focus of experimental studies [14,18,19,20] and became a potential new target for therapeutic approaches.
In this context, preemptive application of dietary supplements with alleged preventive capacities from oxidative stress was proposed in terms of a prophylaxis or even therapy of ocular degenerative diseases [21]. Amongst these supplements, essential fatty acids omega (v)-6 and v-3 are of special interest due to their reported anti-inflammatory, antithrombotic, hypolipidemic, and vasodilatory capacities [22,23]. There is already consistent evidence that v-3 fatty acids are protective agents against ischemia-, light-, oxygen-, inflammation-, and age-associated pathologies of the vascular and neural retina [24,25]. But there are also reports that indicate the requirement of a fine-tuned balance of fatty acid intake, as a misbalanced v-3/ v-6 ratio or excessive amounts of v-6 fatty acids are suspected to promote cardiovascular, inflammatory and autoimmune diseases or cancer [26,27,28].
In the presented study, the effects of v-6 and v-3 fatty acids and their preventing capacities against oxidative stress induced glaucoma-associated expression changes were analyzed in cultivated hTM, the most vulnerable ocular tissue to oxidative stress [7,11,29].

Primary hTM cell culture and treatment
Explant cultures of hTM were obtained from the eye bank of the Ludwig-Maximilians-University, Munich, Germany. Methods of securing human tissue were humane, included proper consent and approval, complied with the Declaration of Helsinki and were approved by the ethic committee of the Department of Medicine of the Friedrich-Alexander-University Erlangen-Nuremberg. The consent statement was written (EK_No. 4346-CH). Monolayer cultures were established from eyes obtained 4 to 8 hours post mortem of five human donors (40-50 years) without any history of eye diseases. Cells were propagated in complete F10 (Ham's F10 medium supplemented with 10% fetal bovine serum [FBS], 10 U/ ml penicillin, 10 mg/ml streptomycin, and 0.25 mg/ml Fungizone Mix; all from PAN TM Biotech GmbH, Aidenbach, Germany) under standard cell culture conditions in 6-well (RNA/ protein extraction) or 24 well (CCK-8/BrdU/IF-labeling) cell culture plates (Techno Plastic Products AG, Trasadingen, Switzerland).
To test the effects of v-3 or v-6 fatty acids, confluent hTM of passages 3 to 5 were pre-incubated in low F10 (Ham's F10/1% FBS) for 24 hours. Then the medium was substituted by low F-10 supplemented to nontoxic 50 mM v-3 [30] or 16 mM v-6 [31] fatty acids (both from Sigma-Aldrich, Taufkirchen, Germany). After 24 hours, medium was replaced by fresh v-3 or v-6 containing medium for an additional 24 hours incubation. After 48 hours in total, oxidative stress was induced by exposure to nontoxic 300 mM hydrogen peroxide [6] (H 2 O 2 , Sigma-Aldrich) for 1 hour. Afterwards the cells were washed with PBS and further cultured with the distinct media for 1 hour. In control cultures, the medium was changed at the same time points, but no H 2 O 2 was added.

RNA isolation and real-time PCR
Total RNA was isolated with a RNA isolation kit (RNeasy Fibrous Tissue Mini Kit; Qiagen N.V., Hilden, Germany) according to the manufactures' instructions. Structural integrity of the RNA samples was confirmed by electrophoresis in 1% Tris-acetate-EDTA (TAE)-agarose gels [32]. Yield and purity were determined photometrically. 200 ng of mRNA were transcribed to cDNA by reverse transcription using a reverse transcription-PCR kit (Access RT-PCR Introductory System; Promega Corporation, Madison, USA). Real-time PCR quantification was performed in 40 cycles in a thermocycler (Light-Cycler System; Roche Diagnostics, Penzberg, Germany). The selected primers for FN, PAI-1, CTGF, Hsp27, Hsp90, IL-1a, IL-6, IL-8, NFkB and glycerinaldehyd-3-phosphat-dehydrogenase (GAPDH) were purchased from Metabion (Metabion International AG, Martinsried, Germany); primer sequences are summarized in Table 1. Corresponding probes were selected with the ProbeFinder v2.04 software (Roche). The standard curve was obtained from probes of three different untreated hTM cultures. As internal control GAPDH was processed simultaneously in each assay and levels of FN-1, PAI-1, CTGF, Hsp27, Hsp90, IL-1a, IL-6, IL-8 and NFkB mRNAs were determined as relative ratios (RR) by division by GAPDH. Ratios in non-supplemented cells without H 2 O 2 treatment were set to one and expression levels of treated cells are given as fold of that. All experiments were performed in triplicates with TM cultures from three different donors. Values represent mean averages 6 SD (n = 9).

Protein isolation and western blot
Media of hTM were collected separately and concentrated six fold by centrifugation (Vivaspin20; Sartorius Stedim GmbH, Goettingen, Germany). Cells were directly lysed in RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% DOC, 0.1% SDS, 50 mM Tris [pH 8.0], 4 mM DTT, 0.5 mM NaVanadate, 2 mM NaF, 2 mM phenylmethylsulfonyl fluoride) containing protease inhibitors (complete protease inhibitor cocktail; Roche). Protein contents of concentrated media and cell lysates were determined by the bicinchoninic acid (BCA) protein assay (Pierce, Rockford, USA). Samples were supplemented with one fourth SDS-loading buffer (Roti-load-1; Roth, Karlsruhe, Germany) and aliquots containing equal proteins were separated by SDS-polyacrylamide gel electrophoresis (PAGE). Proteins were transferred onto a nitrocellulose membrane (Protran Ba-183; Whatman, Dassel, Germany) by semi-dry or tank blotting. Further procedure was done as previously described [33] and primary antibodies ( Table 2) were added over night at 4uC. Corresponding secondary alkaline phosphatase (AP) -conjugated antibodies ( Table 2) were incubated for 30 minutes at room temperature. After substrate incubation (CDP-star; Roche) the signals were visualized by exposure to light sensitive films (Hyperfilm ECL; GE Healthcare, Munich, Germany), which were digitized and densitometrically quantified with the Multi Gauge V3.1 software (Fujifilm, Duesseldorf, Germany). All experiments were performed in triplicates with hTM cultures from three different donors. Values represent mean averages 6 SD (n = 9).

Fibronectin ELISA
Medium contents for FN were analyzed by ELISA according to the manufacturer's instructions (QuantiMatrix Human Fibronectin ELISA KIT; Millipore, Schwalbach, Germany). Aliquots of 50 ml of six-fold concentrated cell media were set in. Absorbance at 450 nm was measured with a spectrophotometer (MultiscanH Spectrum; Thermo). Measurements were done as triplicates of media from hTM cultures of 3 different donors in 3 independent experiments. Values represent mean averages 6 SD (n = 9).

Interleukin ELISAs
Amounts of secreted Interleukins (IL) -6 and -8 were analyzed by ELISA according to the manufacturer's instructions (Quanti-Glo Chemiluminescent ELISA; R&D Systems, Minneapolis, USA). Aliquots of 100 ml and 50 ml of six-fold concentrated cell media were used for IL-6 and IL-8 determination, respectively. Light emission was measured with a luminometer (Fluoroskan Ascent FL; Thermo). Measurements were done as triplicates of hTM cell media from 3 different donors in 3 independent experiments. Values represent mean averages 6 SD (n = 9).

Nuclear Factor (NF) kB Assay
Nuclear content of NFkB was tested in nuclear extracts with a NoShift TM NFkB Transcription Factor Assay (Novagen/Merck, Darmstadt, Germany) according to the manufacturer's instructions. For nuclear extracts, cells were collected from plates with Trypsine/EDTA (PAN TM ), washed three times in Hank's buffered salt solution (HBBS; PAN TM ) and lysed in three times packed cell volumes of low-salt hypotonic cell lysis buffer (20 mM HEPES pH 7.5, 10 mM KCl, 5 mM MgCl2, 0.5 mM EDTA, 0.1% TritonX-100, 10% glycerol, protease inhibitor cocktail [Roche]) for 10 min on ice. Nuclei were pelleted by centrifugation for 10 sec at 4uC and cytosolic fractions (supernatants) were discarded. Nuclei were washed once in low-salt hypotonic cell lysis buffer, and extracted using high-salt hypotonic cell lysis buffer (low-salt supplemented to 500 mM NaCl) for 10 min on ice. Debris was sedimented by centrifugation for 30 min at 4uC and nuclear extracts were transferred to fresh vials. After BCA protein determination extracts were stored at 280uC until use. For assays, equal masses of proteins were set in. Quantification was done by measurement of the absorbance at 450 nm with a spectrophotometer (MultiscanH Spectrum; Thermo). Measurements were done as triplicates of nuclear extracts of hTM cell cultures from 3 different donors in 3 independent experiments. Values represent mean averages 6 SD (n = 9).

Proliferation analysis
Controls had a mean BrdU incorporation rate of 168634% after 48 hours normalized to t = 0 hours. Cells preincubated with v-3 showed a similar increase of incorporation rate as the controls up to 184617% at 48 hours (p = 0.2447). In contrast, v-6 preincubated cells' incorporation rate remained rather constant over 48 hours reaching a final value of 96616%, which was significantly lower than in the controls and the v-3 supplemented cells (***p#0.001; Fig. 2A).
Realtime PCR analysis of NFkB expression did not reveal any significant influences of both fatty acids on this transcription factor (Fig. 6A). Data suggested that H 2 O 2 slightly stimulated NFkB, but statistical significance was not reached (p = 0.1087; Fig. 6A). Notably, such an indication of increase was not detected in fatty acid supplemented hTM (Fig. 6A). On the protein level, a similar regulation was observed. Fatty acids alone did not affect the nuclear NFkB level, whereas H 2 O 2 , though insignificantly, stimulated nuclear contents (Fig. 6B). Again, such a slight increase was not observed in v-3 or v-6 supplemented hTM (Fig. 6B).

Changes of the F-actin cytoskeleton
Phalloidin labeling of F-actin revealed no explicit differences between controls and fatty acid supplemented cells with respect to formation of cross-linked actin networks (CLANs; Fig. 7A). It appeared that v-6 treated cells tended to form increased numbers of stress fibers than the controls as well as v-3 treated cells (Fig. 7A). H 2 O 2 exposure promoted accumulation of stress fibers and CLAN formation (Fig. 7B) in the controls. In v-6 preincubated cells, CLAN formation and stress fibre accumulation appeared even more pronounced (Fig. 7B) than in the stimulated controls and in v-3 pre-treated cells. v-3 pre-treated cells showed a similar frequency of CLANs and stress fibres as the unstimulated controls (Fig. 7B).

Discussion
The current picture of glaucoma pathogenesis suggests that combinatory events act synergistically on the basis of an individual predisposition leading to the onset of the disease [34]. One factor in the focus of investigative glaucoma research is oxidative stress [35], as one of the main problems in glaucoma is an imbalance of ROS formation and decomposition [1] resulting in ROS accumulation [15,19,20] as well as general reduction of antioxidant capacities. TM cells were shown to be especially sensitive to oxidative stress [8,19] and typical responses affect ECM synthesis [6], and basic cellular processes, like cell cycle control, proliferation, apoptosis and cellular metabolism [36,37]. In this respect, prophylactic intake of dietary supplements with alleged antioxidant capacities is recommended for protection from deleterious effects of oxidative stress and prevention of glaucoma thereby.
The aim of this study was to test v-3 and v-6 fatty acids for their abilities to antagonize H 2 O 2 induced glaucomatous effects on cellular activity, proliferation, stress response and ECM synthesis using an in-vitro model based on primary human TM cells.
We found that v-6 inhibited the normal increase of metabolic activity and proliferation during cultivation that was observed in controls and v-3 supplemented cells. This suggests an antiproliferative, cytostatic capacity of v-6, which would agree to reports about rather deleterious effects of excessive amounts of v-6, including promotion of cardiovascular disease, inflammation, autoimmune diseases or cancer [26,27,28], [38,39]. Though, extended studies evaluating cell cycle and cell death are necessary to make concrete assertions. In respect of their capacities to prevent TM cells from H 2 O 2 mediated reductions of metabolism and proliferation, we also found differences for v-6 and v-3 fatty acids. BrdU incorporation was efficiently stabilized by both, but only v-3 rendered hTM unsusceptible against H 2 O 2 mediated reduction of mitochondrial activity.
Taken together, our data indicate inhibitory side effects of v-6 on metabolism and proliferation and a limited effectiveness in prevention from oxidative stress. In conclusion, v-3 appeared to be more beneficial for cellular protection.
Another cellular response to elevated oxidative stress levels is an augmented synthesis of Hsps, which constitute active components of cellular protection and rescue mechanisms [6,40,41,42]. Here we analyzed two representatives of this protein family, Hsp27 and Hsp90, which were found to be strongly expressed in glaucomatous tissues, particularly in the TM. Hsp27 is considered an early marker of cellular stress responses, and is alleged to be especially protective against oxidative damage. Moreover, it was shown that Hsp27 conveys an anti-apoptotic effect by modulation of the nuclear factor (NF) kB-pathway. Hsp90 is also activated by oxidative stress and acts as an essential chaperone maintaining protein stability [42], including transcription factors that regulate anti-apoptotic signaling-pathways.
Our data suggest that Hsp27 is not a direct target of v-6 and v-3 fatty acids, as the observed regulations of the mRNA, although statistically significant, did not manifest on the protein level. Hsp90 in contrast, was strongly activated on the RNA level, and also significantly increased on protein level by both supplements. Based on the known functions of Hsp90, this could be interpreted as a kind of cellular alert condition, with a constantly activated mild stress response rendering the cells prepared for potential exogenous threats. In the context of glaucoma prevention, this would be favorable and argue for beneficial effects of v-6 and v-3 fatty acids in respect of stress control. Partial support is also added by our data on NFkB and Interleukin synthesis. One feature of oxidative stress is the onset of inflammatory processes that promote disease progression in many neurodegenerative conditions. Key players in this processes are NFkB, regulating expression/secretion of pro-inflammatory cytokines [43]. Here we found that fatty acids appeared to abolish the H 2 O 2 mediated stimulation of nuclear NFkB, which therefore could be interpreted as an anti-inflammatory effect. Also the stimulation of IL-6, a key pro-inflammatory interleukin and circumscribed oxidative stress marker [44], was repressed by v-3 and v-6, for the former even to a significant extent. However, our data on IL-1a and IL-8 did not indicate any effect of fatty acids. Moreover, the observed changes in protein levels were rather subtle, so we cautiously propose a beneficial effect, denoting that this hypothesis requests extended experimental approaches in future studies.
Another typical hallmark of the glaucomatous TM is the accumulation of extracellular material in consequence of an increased synthesis and concurrent repression of proteolytic degradation. It has been already shown that oxidative stress induces synthesis of various ECM components [5,35] and accordingly we detected an increase of FN, PAI-1 and CTGF mRNA and protein upon H 2 O 2 exposition. The most obvious effects were observed for FN, which was strongly increased in the medium of v-6 supplemented and even more increased when the cells were additionally exposed to H 2 O 2 , indicating a potential synergistic effect. CTGF synthesis was also stimulated by v-6, although significantly less than FN. Notably, there were no indications for a synergistic effect of v-6 and H 2 O 2 . v-3 fatty acid did not have significant effects on FN and CTGF expression, but antagonized the H 2 O 2 mediated stimulation of both proteins. An obvious effect, however, was detected on PAI-1 expression, which was activated by v-3. Notably, exposition to H 2 O 2 resulted in lower PAI-1 expression than in the corresponding controls, indicating a similar antagonizing effect as observed with CTGF and FN. Summarized, it appeared that v-6 fatty acid alone seemed to stimulate ECM synthesis and v-3 fatty acid seemed to prevent ECM degradation via activation of PAI-1, both effects that would favor ECM accumulation in the context of glaucoma disease. Paradoxically, both fatty acids seem to have the ability to antagonize the H 2 O 2 mediated stimulations, thus indicating an overall protective effect.
Morphologically, we observed an increased formation of CLANs and intracellular stress fibers after H 2 O 2 stimulation, a frequent finding in glaucomatous TM cells [5,45,46,47,48,49,50]. Here, only preemptive application of v-3 had a preventive effect on formation of these stress indicators.
To sum up, v-6 was efficient in preventing H 2 O 2 mediated antiproliferative effects, but displayed a repressive effect on mitochondrial activity and proliferation. For v-3, we observed no negative side effects but an effective potential to prevent H 2 O 2 mediated anti-proliferative/-metabolic effects. Both agents induced Hsp90, which can be interpreted in terms of a cellular precaution to forthcoming insults. Considering matrix synthesis, both fatty acids were pro-fibrotic, but still could antagonize H 2 O 2 stimulation. Lastly, v-3 was effective in prevention from CLAN and stress fiber formation.
Based on this, we conclude that v-3 to seems be the more beneficial fatty acid, whereas v-6 appears more critical and not unreservedly recommendable. This agrees with reported cytotoxic side effects of high-dose v-6 [26,27,28,38,39,51]. Future studies including other vulnerable ocular cell types will have to prove if preemptive dietary with v-3 helps to prevent deleterious effects of oxidative stress in glaucoma and other age-associated degenerative diseases, and will have to further challenge the eligibility of v-6 as a protective nutritional supplement.