1

NF- B mediated nitric oxide production and activation of caspase-3 cause retinal ganglion cell death in the hypoxic neonatal retina

Gurugirijha Rathnasamy1, Viswanathan Sivakumar1, Parakalan Rangarajan1, Wallace

S Foulds, 2, 3 , Eng Ang Ling1, Charanjit Kaur 1,2*

1. Department of Anatomy, Yong Loo Lin School of Medicine, Blk MD10, 4 Medical

Drive, National University of Singapore, Singapore 117597

2. Singapore Eye Research Institute c/o Singapore National Eye Centre, 11 Third

Hospital Avenue, Singapore 168751

3. Emeritus Professor, University of Glasgow, Glasgow, Scotland G12 8QQ

*Corresponding author

E-mail: antkaurc@nus.edu.sg

Fax: 65-67787643

Phone: 65-65163209

Running title: nNOS mediates apoptosis of RGCs in hypoxic retina

IOVS Papers in Press. Published on August 19, 2014 as Manuscript iovs.13-13718

Copyright 2014 by The Association for Research in Vision and Ophthalmology, Inc.

2

Abstract 1

Purpose: Hypoxic insult to the developing retina results in apoptosis of retinal ganglion cells 2

(RGCs) through production of inflammatory mediators, nitric oxide (NO) and free radicals. 3

The present study was aimed at elucidating the pathway through which hypoxia results in 4

overproduction of NO in the immature retina, and its role in causing apoptosis of RGCs. 5

6

Methods: One-day-old Wistar rats were exposed to hypoxia and their retinas were studied at 7

3h to 14 days after exposure. The protein expression of nuclear factor- B (NF- B) and 8

neuronal nitric oxide synthase (nNOS) in the retina and primary cultures of RGCs was 9

analysed using western blotting and double-immunofluorescence; whereas, the concentration 10

of NO was determined calorimetrically. In cultured RGCs, hypoxia-induced apoptosis was 11

evaluated by caspase-3 immunolabelling. 12

13

Results: Following hypoxic exposure, NF- B mediated expression of nNOS, which was 14

localized to the RGCs, and subsequent NO production was significantly increased in the 15

developing retina. In primary cultures of RGCs subjected to hypoxia, the up-regulation of 16

nNOS and NO was significantly suppressed when treated with 7-nitroindazole (7-NINA), a 17

nNOS inhibitor or BAY, a NF- B inhibitor. Hypoxia-induced apoptosis of RGCs, which was 18

evident with caspase-3 labelling, was also suppressed when these cells were treated with 7-19

NINA or BAY. 20

21

Conclusion: Our results suggest that in RGCs, hypoxic induction of nNOS is mediated by 22

NF- B and the resulting increased release of NO by RGCs, results in their apoptosis through 23

caspase-3 activation. It is speculated that targeting nNOS could be a potential neuroprotective 24

strategy against hypoxia-induced RGCs death in the developing retina. 25

3

26

Key words: hypoxia; nitric oxide; neuronal nitric oxide synthase; 7-nitroindazole; NF- B; 27

retinal ganglion cells; caspase-3 28

29

Introduction 30

Hypoxic insult to the immature retina results in death of retinal ganglion cells (RGCs) 1, 2 and 31

leads to visual impairments in the neonate. Growing evidence suggests that several factors 32

such as apnoea, placental insufficiency, pulmonary dysfunction and respiratory distress and 33

cyanotic heart disease, all of which can result in hypoxia, are important etiological factors in 34

the development of retinal damage in the immature eye 3-6 35

We have shown recently that hypoxic damage in the developing retina, through 36

enhanced production of free radicals, nitric oxide (NO) and inflammatory mediators results in 37

the death of RGCs 7, 8. Nitric oxide (NO), synthesized from L-arginine by nitric oxide 38

synthase (NOS), is known to mediate a wide range of physiological processes such as 39

vasodilation, neurotransmission and host cell defense 9-11. In response to various stimuli, all 40

three isoforms of NOS (endothelial NOS (eNOS), neuronal NOS (nNOS) and inducible NOS 41

(iNOS)), have been demonstrated in the developing retina 1, 12-16. Among these isoforms, 42

nNOS has been reported to be expressed in RGCs 13, 14, 17. We have earlier reported an 43

increased expression of nNOS in the RGCs of retinas of hypoxic neonatal rats 1 and it has 44

been proposed that nNOS contributes significantly to the death of RGCs 17. 45

Although NO regulates many physiological and cellular processes, a high concentration of 46

NO has been reported to be potentially cytotoxic 18-20 and has been postulated as a key factor 47

mediating various forms of retinopathy 21. Koeberle and Ball 22 have previously demonstrated 48

the death of RGCs in adult retina following the intraocular administration of an NO donor. 49

NO has been demonstrated to have toxic effects on RGCs in vitro under hypoxic conditions 50

4

23, and suppression of NOS under such conditions has been reported to protect RGCs. 51

Although a toxic role for NO has been proposed, its role in mediating apoptosis of RGCs, in 52

the hypoxic developing retina, has not been elucidated. 53

The present study was aimed at demonstrating the pathway involved in hypoxia-54

mediated up-regulation of nNOS in RGCs and the subsequent production of NO, which may 55

result in apoptosis of RGCs. Previous studies have suggested that nuclear factor kappa B 56

(NF-kB) might be involved in the up-regulation of nNOS 24-26 and hypoxia has been reported 57

to enhance the nuclear translocation of NF-kB in neural tissues 27, 28. The present study 58

evaluated the role of NF- B, in inducing nNOS, by treating hypoxic RGCs with NF- B 59

specific inhibitor, BAY. In addition, the nNOS specific inhibitor, 7-nitroindazole (7-NINA), 60

was used to evaluate the role of NO in causing the death of RGCs. 61

Materials and Methods 62

Animals 63

Forty-seven 1-day-old Wistar rats were exposed to hypoxia by placing them for two hours 64

in a multi-gas chamber (Model: MCO 18M; Sanyo Biomedical Electrical Co, Ltd, Tokyo, 65

Japan), filled with a gas mixture of 5% oxygen and 95% nitrogen. The rats were then allowed 66

to recover under normoxic conditions for 3h, 24h, 3d, 7d or 14d before sacrifice. Another 67

group of 41 rats kept outside the chamber was used as age matched controls. A total of 40 (6-68

8 days old) rats were used for the preparation of primary cultures of RGCs. The study was 69

approved by the Institutional Animal Care and Use Committee of National University of 70

Singapore and was conducted in accordance with the ARVO Statement for the Use of 71

Animals in Ophthalmic and Vision Research. 72

73

74

5

Primary cultures of retinal ganglion cells 75

Preparation of retinal suspensions 76

Retinas were dissociated enzymatically to make a suspension of single cells, essentially 77

as described previously29, 30. Briefly, the retinas derived from 6 to 8 day-old Wistar rats were 78

incubated at 37°C for 30 min in Eagle’s balanced salt solution containing papain (15U/ml), 79

collagenase (70U/ml), bovine serum albumin (BSA; 0.2mg/ml; Sigma-Aldrich, MO, USA) 80

and DL-cysteine (0.2mg/ml). To yield a suspension of single cells, the tissue was then 81

triturated sequentially through a narrow-bore Pasteur pipette in a solution containing 2ng/ml 82

ovomucoid, 0.004% DNase I, and 1mg/ml BSA. After centrifugation at 800 rpm for 5 min, 83

the cells were rewashed in a solution containing ovomucoid and BSA (10mg/ml of each). The 84

cells were then re-suspended in 0.1% BSA in phosphate buffered saline (PBS). 85

Preparation of panning tubes 86

Culture flasks (25 cm2) were incubated with OX-42 antibody (1:50; Harlan Sera-Lab Ltd, 87

Edinburgh, UK) diluted in 2.5 ml of PBS at 4°C overnight. Corning polypropylene tubes 88

(50ml) were incubated with Thy-1 antibody diluted in 3 ml of PBS (1:200; Santacruz 89

Biotechnology Inc., Santacruz, USA). The flasks and tubes were washed twice with 3 ml of 90

PBS. To prevent non-specific binding of cells to the panning flasks and tubes, 3-4 ml of 0.1% 91

BSA was placed on the coated area. 92

Panning procedure 93

The retinal cell suspension was incubated in the OX-42-coated flasks at room temperature 94

for 30 min. The suspension was gently shaken every 10 min to ensure access of all cells to 95

the surface of the coating area. Non-adherent cells were removed and placed in the Thy-1 96

coated tubes. The cells were incubated for 30 min and the tubes were then washed gently five 97

times with 3 ml of PBS. Finally adherent cells on Thy-1 coated tubes were washed with 98

culture medium (described below) and after centrifugation at 800 rpm for 5 min, the 99

6

supernatant was carefully discarded and the cells were seeded on to 12-mm glass cover slips 100

that had been coated with poly-L-lysine (50μg/ml; Sigma-Aldrich). The purity of the RGCs 101

in cultures was determined by staining with the antibody Thy-1, a specific RGC marker 31. 102

The percentage of RGCs in the cultures was about 90%. 103

Culture of purified retinal ganglion cells 104

Purified RGCs were plated at a low density of approximately 500-2000 cells/well and 105

were cultured in 400μl of a serum-free Neurobasal medium containing with glutamine (1mM; 106

Sigma-Aldrich), gentamicin (10μl/ml; Invitrogen Life Technologies, CA, USA), B27 107

supplement (1:50; Invitrogen Life Technologies), brain-derived neurotrophic factor (50ng/ml; 108

Sigma-Aldrich), ciliary neurotrophic factor (50ng/ml; Sigma-Aldrich), and forskolin (10μM; 109

Sigma-Aldrich). The cultures were maintained at 37°C in a humidified atmosphere 110

containing 5% CO2 and 95% air. 111

Treatment of retinal ganglion cells 112

To study the effects of hypoxia on the expression of nNOS and NF- B in RGCs, the cells 113

were exposed to hypoxia in a chamber (Model MCO 18M; Multi-gas incubator, Sanyo 114

Company Pte Ltd, Japan) for 4h at 37°C in a 3% oxygen, 5% CO2 and 92% nitrogen mixture. 115

In all the experiments, RGCs in matching controls were incubated in an incubator at 37°C 116

with 95% air and 5% CO2. In addition, serum-free medium containing 10μM of 7-117

nitroindazole (7-NINA; 32 Tocrisis Bioscience, MO, USA) or BAY (NF- B inhibitor; 33 118

Sigma-Aldrich, MO, USA) was added to each well for 3h immediately after hypoxic 119

exposure. In various groups, the concentration of NO was measured, and cell death was 120

investigated by caspase-3 labeling. 121

Western blotting 122

Retinas (2 retinas from each rat) were removed from rats exposed to hypoxia (n=5 at each 123

time point) and their corresponding controls (n=5 at each time point). Protein was extracted 124

7

from the retinas using tissue protein extraction reagent and from cultured RGCs using 125

mammalian protein extraction reagent (Thermo Scientific, MA, USA) containing protease 126

inhibitors. All procedures were carried out at 4°C. Homogenates were centrifuged at 15000 g 127

for 10 min and the supernatant was collected. Cytoplasmic and nuclear extracts of cultured 128

RGCs were isolated using a NE-PER© (Thermo Scientific, MA, USA) kit following the 129

manufacturer's instructions. Protein concentrations were determined by the Bradford method 130

34 using BSA (Sigma-Aldrich) as a standard. Samples of supernatants containing 20μg 131

protein were heated to 95°C for 5 min and were separated by SDS-PAGE in 10% SDS gels, 132

in a Mini-Protean 3 apparatus (Bio-Rad, CA, USA). Protein bands were electro-blotted onto 133

0.45μm polyvinylindene difluoride membranes (Bio-Rad) and then blocked with 5% non-fat 134

milk for one hour at room temperature. The membranes were washed and subsequently 135

incubated with anti-nNOS (1:1000; BD Biosciences, USA) antibodie diluted in blocking 136

solution (5% non-fat milk), overnight at 4°C. The membranes were then washed and were 137

incubated with secondary antibody; anti-rabbit IgG, 1:5000 (Thermo Scientific, MA, USA) 138

conjugated with horseradish peroxidase. Specific binding was revealed by an enhanced-139

chemiluminescence kit (Thermo Scientific, MA, USA) following the manufacturer’s 140

instructions. For loading control, after intensive washing, the membranes were incubated with 141

monoclonal mouse anti- -actin (1:5000; Sigma-Aldrich). X-ray films (Thermo Scientific, 142

MA, USA) were scanned with a computer-assisted G-710 densitometer (Bio-Rad) to quantify 143

band optical density using Quantity One software (Bio-Rad). 144

Nitric oxide colorimetric assay 145

The total amount of NO in the retinal tissue supernatant from the control and hypoxic rats 146

(n=5 at each time point) was determined by the Griess reaction using a colorimetric assay kit 147

(US Biological, Swampscott, MA, USA) according to the manufacturer’s instructions. The 148

8

optical density was measured at 520nm with a precision microplate reader (Molecular 149

Devices Corporation, CA, USA) that detects nitrite (NO2-), a stable reaction product of NO. 150

The concentration of NO in the culture medium from RGCs subjected to hypoxia and 151

those subjected to hypoxia and of 7-NINA/BAY was measured using the above colorimetric 152

assay kit. 153

NF- B assay 154

The level of NF- B in the cytosolic/nuclear fractions of cultured RGCs from the control, 155

hypoxia, hypoxia + BAY and Hypoxia + 7-NINA groups was determined with NF- B p65 156

(pSer536) phosphotracer ELISA kit (Abcam) according to the manufacturer's instructions. 157

The relative fluorescence of the samples was measured at 530nm excitation/590nm emission 158

using a SpectraMaxM5 microplate reader (Molecular Devices Corporation, CA, USA). 159

Double immunofluorescence 160

Rats at 3d after hypoxic exposure and their corresponding controls (n=3 in each group) 161

were used for double immunofluorescence studies. Following deep anesthesia with 6% 162

pentobarbital, the rats were sacrificed by perfusion with 2% paraformaldehyde in 0.1M 163

phosphate buffer, pH 7.4. Frozen coronal sections of the retina with a thickness of 40μm 164

were cut with a cryostat (Model 3050; Leica Instruments GmbH, NUBLOCH, Germany) and 165

rinsed in PBS. Endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide in 166

methanol for 30 min and the sections were subsequently washed with PBS. Sections were 167

then incubated at room temperature with a cocktail mix of two primary antibodies: nNOS 168

(1:500; BD Biosciences, USA)/NF- B (1:100; Santacruz Biotechnology Inc, USA) and NeuN 169

(1:200; Millipore, MA, USA) the latter being a specific marker for RGCs 35. Subsequent 170

antibody detection was carried out with a cocktail mix of two secondary antibodies: Cy3-171

conjugated goat anti-rabbit IgG and FITC-conjugated sheep anti-mouse IgG (1:100; Sigma-172

Aldrich). After three washes with PBS, the sections were mounted with a fluorescent 173

9

mounting medium (DakoCytomation, Glostrup, Denmark). Co-localization of nNOS/NF- B 174

with NeuN was observed under a confocal microscope (Olympus, FV 1000 Olympus Optical 175

Co. Ltd, Tokyo, Japan). The isotypic control confirmed the specificity of all primary 176

antibodies used (data not shown). 177

Purified RGCs were fixed in 4% paraformaldehyde in 0.1M phosphate buffer, pH 7.4 for 178

20 min, and blocked with 3% normal goat serum and 1% BSA for 30 min. The cells were 179

then incubated overnight at 4°C with a mixture of two primary antibodies against NF- B / 180

nNOS and Thy1.1 separately. Thy1.1 is a specific marker for RGCs. Subsequent antibody 181

detection was carried out with the mixture of secondary antibodies; Cy3-conjugated goat anti-182

rabbit IgG and FITC-conjugated sheep anti-mouse IgG (1:100; Sigma-Aldrich) and processed 183

as described above. 184

Caspase-3 labelling in Retinal Flatmounts and in cultured RGCs 185

Retinal flatmounts were prepared from retinas collected from rats at 3d after hypoxia 186

along with their age matched controls and hypoxic rats treated with BAY (20mg/kg36)/ 7-187

NINA (10mg/kg37) (n=3 in each group), following the instruction as described previously 38. 188

For the detection of apoptosis, the flatmounts were washed with PBS and the endogenous 189

peroxide activity was blocked with 0.3% hydrogen peroxide for 30 minutes. The flatmounts 190

were then incubated overnight at room temperature with a cocktail of two primary antibodies: 191

anti-caspase-3 (1:200; Cell Signaling Technology, Inc., MA, USA) and anti-Thy1.1 192

antibodies. Subsequently the flatmounts were washed in PBS and incubated with a cocktail 193

mix of secondary antibodies and were mounted with a fluorescent mounting medium (Dako 194

Cytomation) following the steps detailed above. For apoptosis detection in primary cultures, 195

RGCs were incubated at 4°C overnight with a cocktail mix of anti-caspase-3 and anti-Thy1.1 196

antibodies and processed as described above. The number of caspase-3 positive RGCs was 197

obtained by counting cells in six randomly selected microscopic fields obtained from each 198

10

slide at X40 magnification. The percentage of caspase-3 positive RGCs against the total 199

number of RGCs was calculated and averaged. 200

Statistical analysis 201

The data are presented as mean ± SD. One-way ANOVA followed by post-hoc analysis 202

using Dunnett’s test (GraphPad Software, San Diego, CA, USA) was used to determine the 203

statistical significance of differences between normal vs hypoxic and between hypoxic vs 204

hypoxic+7-NINA/BAY groups. A value of P<0.05 (*) was considered statistically 205

significant. 206

Results 207

nNOS protein expression by western blotting 208

Protein expression of nNOS showed a significant difference between the control and 209

hypoxic groups. An immunoreactive band for nNOS was detected at 155 kDa (Fig.1A) and 210

was significantly increased at 24h and 3d after the hypoxic exposure (Fig. 1B) but decreased 211

below control levels at 7 and 14d. 212

Nitric Oxide assay 213

The concentration of NO in the retina was significantly increased at 24h and 3d after hypoxic 214

exposure when compared with controls (Fig. 1C). However, the changes in NO levels 215

observed at 3h, 7d and 14d were not significant. 216

Cellular localization of nNOS and NF- B 217

Expression of nNOS was localised in NeuN labelled cells in the ganglion cell layer (GCL) 218

of the retina that were identified as RGCs. A weak immunoexpression of nNOS was observed 219

in the GCL (Fig. 2A-C) of control rat retinas. The expression of nNOS in the GCL (Fig, 2D-220

F) was enhanced at 3d following hypoxic exposure when compared to the controls. At 3d 221

after hypoxic exposure, parallel to nNOS the expression of NF- B was also increased in 222

NeuN labelled RGCs in the hypoxic retina when compared to controls (Fig 2G-L) and there 223

11

was increased nuclear translocation of NF- B into the nucleus of RGCs in the hypoxic retina 224

(Fig 2 K, L). 225

Hypoxia-induced nuclear translocation of NF- B 226

ELISA analysis indicated a significant increase in both nuclear and cytoplasmic NF- B in 227

RGCs exposed to hypoxia when compared to controls (Fig. 3A). However, this increase was 228

significantly suppressed when hypoxic RGCs were treated with NF- B inhibitor BAY. In 229

hypoxic RGCs treated with 7-NINA there was no significant change in NF- B levels when 230

compared to hypoxic RGCs. 231

Double immunofluorescence showed the cytoplasmic localization of NF- B in the control 232

group of cells (Fig. 3B-D) and the nuclear translocation in cells subjected to hypoxia (Fig. 233

3E-G). 234

NF- B mediated the expression of nNOS and NO production in hypoxic RGCs 235

Western blot analysis showed that the protein expression of nNOS was increased in 236

primary cultures of RGCs subjected to 4h of hypoxia when compared to that of controls (Fig. 237

4A, B) and was suppressed by both 7-NINA and BAY. NO levels in RGCs culture media as 238

determined by the colorimetric assay were significantly increased after 4h of hypoxic 239

exposure when compared to control cell culture medium (Fig. 4C). NO levels were reduced 240

in the RGCs culture medium from hypoxic+7-NINA and hypoxic+BAY groups. 241

Double immunofluorescence showed that the nNOS protein expression was enhanced in 242

RGCs subjected to 4h of hypoxia (Fig. 5 D-F) compared to that of control cells (Fig. 5A-C). 243

Hypoxia-induced nNOS expression was diminished by 7-NINA (Fig. 5G-I) and by BAY 244

(Fig. 5J-L), confirming the results obtained from western blot and colorimetric analyses. 245

Caspase-3 activation in retinal flatmounts and in cultured RGCs 246

In retinal flatmounts (Fig. 6A) obtained from control rats, only a few Thy-1 247

immunoreactive RGCs were positive for caspase-3 (Fig. 6B-D). There was a significant 248

12

increase in caspase-3 positive RGCs in retinal flatmounts obtained from hypoxic rats (Fig. 249

7E-G). However, there was reduced caspase-3 labelling in retinal flatmounts from hypoxic 250

rats treated with either BAY (Fig. 6H-J) or 7-NINA (Fig. 6K-M). The percentage of caspase-251

3 positive RGCs was significantly increased in hypoxic group when compared to the control 252

group. This was however; significantly reversed when hypoxic RGCs were treated with 7-253

NINA or BAY (Fig. 6N). Similar results were obtained in RGC cultures in control, hypoxia, 254

hypoxia+BAY and hypoxia+7-NINA groups (Fig. 7A-M). 255

Discussion 256

In this study, we have shown that in neonatal retina, apoptosis of RGC following a hypoxic 257

exposure is associated with hypoxia-mediated nuclear translocation of NF- B and increased 258

expression of nNOS. It appears that the nuclear translocation of NF- B played a role in the 259

increased expression of nNOS in hypoxic RGCs. This notion lends its support from the fact 260

that nNOS expression in cultured RGC exposed to hypoxia was significantly reduced by a 261

NF- B specific inhibitor, BAY. Our study further indicates that an increased production of 262

NO through the enhanced nNOS expression in hypoxic RGCs causes the death of the RGCs 263

by activated caspase-3 mediated apoptosis. 264

In the developing retina, NOS and NO are required for the timely maturation of the inner 265

plexiform layer 12 and for early retinal differentiation 15 but an excessive induction of NOS 266

isoforms has been implicated in damage to the retina 17, 39. In response to hypoxic insult, in 267

the developing retina, although excessive production of inflammatory mediators 7 and 268

destructive effects of free radicals 8 are implicated in death of RGCs, it appears that increased 269

nNOS expression in RGCs in the hypoxic retina and the subsequent production of NO may 270

also result in their apoptosis via caspase-3 activation. Hypoxia-mediated nNOS expression is 271

well documented 1, 39, 40 and NO produced from nNOS has been demonstrated as being 272

13

highly detrimental to RGCs in adult retina 17, 22, yet the mechanisms involved remain to be 273

fully characterized. 274

It has been speculated that hypoxia-mediated activation of transcription factor, NF- B 27, 28, 275

may play a critical role in the regulation and activation of genes involved in inflammation, 276

oxidative stress and apoptosis 27, 41-43. Under physiological conditions, NF-κB is localized in 277

the cytoplasm and its activation is inhibited by IκB 44. Hypoxic exposure causes degradation 278

of IκB and results in activation and translocation of NF-κB 28, 45 to the nucleus, where it 279

regulates the expression of target genes. In the present study, following hypoxic exposure, the 280

expression of NF- B was increased in the RGCs of developing retina. This was further 281

supported by the finding from cultured RGCs, wherein the concentration of NF- B was up-282

regulated in both the cytoplasmic and nuclear fractions of cultured hypoxic RGCs. However, 283

this increase was abolished when hypoxic cultured RGCs were treated with BAY. Hypoxia-284

induced NF- B activation has been reported previously in human retinal progenitor cells 46 285

and in RGC-5, a RGC cell line 47, 48 and the activation of NF- B in RGCs has been implicated 286

in the apoptosis of these cells 49 50 48. In light of the above and from our results, it appears that 287

the nuclear translocation of NF- B in RGCs, in response to hypoxia, could lead to the 288

transcription of genes that might result in the death of RGCs. 289

Additional support for the role of NF- B in the hypoxia induced expression of nNOS 290

comes from a previous study, which reported the presence of NF- B binding site in the 291

promoter of the nNOS gene 25. The suppression of nNOS expression and NO production in 292

hypoxic RGCs treated with BAY or 7-NINA also supports the view that hypoxia-mediated 293

nuclear translocation of NF- B is essential for the induction of nNOS and the subsequent 294

production of NO in hypoxic RGCs. 295

14

A number of reports claim that an excess production of NO through nNOS expression could 296

mediate RGC death. These include the ability of NO to induce apoptosis in cultured retinal 297

neurons when treated with advanced glycation end products 51/ S-nitroso-N-acetyl-298

penicillamine (SNAP), a NO donor 23 and the increased survival of cultured RGCs against 299

NO mediated neurotoxicity, by the addition of NOS inhibitors such as L-NAME to the 300

culture medium 52. Previously, NO was shown to induce the pro-apoptotic cascade, in 301

hypoxic neural tissues, by phosphorylating Bcl-2 53. Once phosphorylated, Bcl-2 loses its 302

anti-apoptotic potential and its ability to heterodimerize with the pro-apoptotic protein Bax, 303

resulting in Bax-mediated activation of caspases and initiation of apoptosis 54-56. NO 304

mediated injury to the RGCs is believed to occur via a caspase dependent pathway. The 305

addition of caspase inhibitor, Z-VAD-FMK, to SNAP treated hypoxic RGC-5 cells resulted 306

in partial protection 23. In the present study, following hypoxia, parallel to the increased NO 307

production there was increased expression of caspase-3 in RGCs in the developing retina. 308

Our in vitro study also depicted the same; wherein there was increased caspase-3 labelling in 309

hypoxic cultured RGCs. This increase in caspase-3 positive RGCs however, was reduced 310

when treated with 7-NINA or BAY, both in vivo and in vitro. The results support the view 311

that excess NO produced by nNOS in hypoxic RGCs leads to their apoptosis through 312

activation of caspase cascade. 313

Conclusion 314

Taken together, our results indicate that in hypoxic immature retina, activation of NF- B 315

in the RGCs results in the increased expression of nNOS, which subsequently leads to 316

increased production of NO. This enhanced production of NO in turn causes the death of 317

RGCs through caspase-3 activation. Inhibitors of nNOS and NF-κB, such as 7-NINA and 318

BAY significantly reduced hypoxia-induced nNOS expression and NO production and 319

15

decreased the death of RGCs following hypoxia, suggesting that they could be potential 320

therapeutic agents against hypoxia associated damage in the developing retina. 321

Acknowledgements 322

This study was supported by a research grant R-181-000-120-213 from National Medical 323

Research Council (NMRC) of Singapore and R-181-000-148-750 from National University 324

Health system (NUHS), Singapore. The authors do not have any conflict of interest. 325

326

References 327

1. Kaur C, Sivakumar V, Foulds WS, Luu CD, Ling E-A. Cellular and vascular changes 328 in the retina of neonatal rats after an acute exposure to hypoxia. Investigative ophthalmology 329 & visual science 2009;50:5364-5374. 330

2. Sivakumar V, Foulds WS, Luu CD, Ling EA, Kaur C. Hypoxia-induced retinal 331 ganglion cell damage through activation of AMPA receptors and the neuroprotective effects 332 of DNQX. Experimental Eye Research 2013;109:83-97. 333

3. Johns KJ, Johns JA, Feman SS, Dodd DA. Retinopathy of prematurity in infants with 334 cyanotic congenital heart disease. American journal of diseases of children (1960) 335 1991;145:200-203. 336

4. Shah VA, Yeo CL, Ling YLF, Ho LY. Incidence, risk factors of retinopathy of 337 prematurity among very low birth weight infants in Singapore. Annals of the Academy of 338 Medicine, Singapore 2005;34:169-178. 339

5. Akkoyun I, Oto S, Yilmaz G, et al. Risk factors in the development of mild and severe 340 retinopathy of prematurity. J AAPOS 2006;10:449-453. 341

6. Lad EM, Nguyen TC, Morton JM, Moshfeghi DM. Retinopathy of prematurity in the 342 United States. Br J Ophthalmol 2008;92:320-325. 343

7. Sivakumar V, Foulds WS, Luu CD, Ling E-A, Kaur C. Retinal ganglion cell death is 344 induced by microglia derived pro-inflammatory cytokines in the hypoxic neonatal retina. The 345 Journal of pathology 2011;224:245-260. 346

8. Kaur C, Sivakumar V, Robinson R, Foulds WS, Luu CD, Ling E-A. Neuroprotective 347 effect of melatonin against hypoxia-induced retinal ganglion cell death in neonatal rats. J 348 Pineal Res 2013;54:190-206. 349

9. Christopherson KS, Bredt DS. Nitric oxide in excitable tissues: physiological roles 350 and disease. The Journal of clinical investigation 1997;100:2424-2429. 351

16

10. MacMicking J, Xie QW, Nathan C. Nitric oxide and macrophage function. Annual 352 review of immunology 1997;15:323-350. 353

11. Nathan C. Inducible nitric oxide synthase: what difference does it make? The Journal 354 of clinical investigation 1997;100:2417-2423. 355

12. Oh S-J, Kim K-Y, Lee E-J, et al. Inhibition of nitric oxide synthase induces increased 356 production of growth-associated protein 43 in the developing retina of the postnatal rat. Brain 357 research Developmental brain research 2002;136:179-183. 358

13. Kim KY, Ju WK, Oh SJ, Chun MH. The immunocytochemical localization of 359 neuronal nitric oxide synthase in the developing rat retina. Experimental brain research 360 Experimentelle Hirnforschung Expérimentation cérébrale 2000;133:419-424. 361

14. Tsumamoto Y, Yamashita K, Takumida M, et al. In situ localization of nitric oxide 362 synthase and direct evidence of NO production in rat retinal ganglion cells. Brain research 363 2002;933:118-129. 364

15. Patel JI, Gentleman SM, Jen LS, Garey LJ. Nitric oxide synthase in developing 365 retinas and after optic tract section. Brain research 1997;761:156-160. 366

16. Griffith RM, Li H, Zhang N, et al. Next-generation sequencing analysis of gene 367 regulation in the rat model of retinopathy of prematurity. Documenta ophthalmologica 368 Advances in ophthalmology 2013;127:13-31. 369

17. Katsuki H, Yamamoto R, Nakata D, Kume T, Akaike A. Neuronal nitric oxide 370 synthase is crucial for ganglion cell death in rat retinal explant cultures. Journal of 371 pharmacological sciences 2004;94:77-80. 372

18. Iadecola C. Bright and dark sides of nitric oxide in ischemic brain injury. Trends in 373 neurosciences 1997;20:132-139. 374

19. Hardy P, Peri KG, Lahaie I, Varma DR, Chemtob S. Increased nitric oxide synthesis 375 and action preclude choroidal vasoconstriction to hyperoxia in newborn pigs. Circ Res 376 1996;79:504-511. 377

20. Hardy P, Dumont I, Bhattacharya M, et al. Oxidants, nitric oxide and prostanoids in 378 the developing ocular vasculature: a basis for ischemic retinopathy. Cardiovascular research 379 2000;47:489-509. 380

21. Osborne NN, Casson RJ, Wood JPM, Chidlow G, Graham M, Melena J. Retinal 381 ischemia: mechanisms of damage and potential therapeutic strategies. Progress in retinal and 382 eye research 2004;23:91-147. 383

22. Koeberle PD, Ball AK. Nitric oxide synthase inhibition delays axonal degeneration 384 and promotes the survival of axotomized retinal ganglion cells. Experimental neurology 385 1999;158:366-381. 386

23. Sato T, Oku H, Tsuruma K, et al. Effect of hypoxia on susceptibility of RGC-5 cells 387 to nitric oxide. Investigative ophthalmology & visual science 2010;51:2575-2586. 388

17

24. Harvey BH, Bothma T, Nel A, Wegener G, Stein DJ. Involvement of the NMDA 389 receptor, NO-cyclic GMP and nuclear factor K-beta in an animal model of repeated trauma. 390 Human psychopharmacology 2005;20:367-373. 391

25. Li Y, Li G, Li C, Zhao Y. Identification of nuclear factor-kappaB responsive element 392 within the neuronal nitric oxide synthase exon 1f-specific promoter. Acta biochimica et 393 biophysica Sinica 2007;39:247-254. 394

26. Li Y, Li C, Sun L, et al. Role of p300 in regulating neuronal nitric oxide synthase 395 gene expression through nuclear factor-kappaB-mediated way in neuronal cells. 396 Neuroscience 2013;248:681-689. 397

27. Oliver KM, Garvey JF, Ng CT, et al. Hypoxia activates NF-kappaB-dependent gene 398 expression through the canonical signaling pathway. Antioxidants & redox signaling 399 2009;11:2057-2064. 400

28. Culver C, Sundqvist A, Mudie S, Melvin A, Xirodimas D, Rocha S. Mechanism of 401 hypoxia-induced NF-kappaB. Molecular and cellular biology 2010;30:4901-4921. 402

29. Otori Y, Kusaka S, Kawasaki A, Morimura H, Miki A, Tano Y. Protective effect of 403 nilvadipine against glutamate neurotoxicity in purified retinal ganglion cells. Brain research 404 2003;961:213-219. 405

30. Otori Y, Wei JY, Barnstable CJ. Neurotoxic effects of low doses of glutamate on 406 purified rat retinal ganglion cells. Investigative ophthalmology & visual science 1998;39:972-407 981. 408

31. Barnstable CJ, Dräger UC. Thy-1 antigen: a ganglion cell specific marker in rodent 409 retina. Neuroscience 1984;11:847-855. 410

32. Huang H-M, Shen C-C, Ou H-C, et al. Neuroprotective MK801 is associated with 411 nitric oxide synthase during hypoxia/reoxygenation in rat cortical cell cultures. Journal of 412 cellular biochemistry 2002;84:367-376. 413

33. Deng YY, Lu J, Ling EA, Kaur C. Monocyte chemoattractant protein-1 (MCP-1) 414 produced via NF-kappaB signaling pathway mediates migration of amoeboid microglia in the 415 periventricular white matter in hypoxic neonatal rats. Glia 2009;57:604-621. 416

34. Bradford MM. A rapid and sensitive method for the quantitation of microgram 417 quantities of protein utilizing the principle of protein-dye binding. Analytical biochemistry 418 1976;72:248-254. 419

35. Buckingham BP, Inman DM, Lambert W, et al. Progressive ganglion cell 420 degeneration precedes neuronal loss in a mouse model of glaucoma. The Journal of 421 neuroscience : the official journal of the Society for Neuroscience 2008;28:2735-2744. 422

36. Murugan M, Sivakumar V, Lu J, Ling EA, Kaur C. Expression of N-methyl D-423 aspartate receptor subunits in amoeboid microglia mediates production of nitric oxide via NF-424

B signaling pathway and oligodendrocyte cell death in hypoxic postnatal rats. Glia 425 2011;59:521-539. 426

18

37. Tjong YW, Ip SP, Lao L, et al. Role of neuronal nitric oxide synthase in colonic 427 distension-induced hyperalgesia in distal colon of neonatal maternal separated male rats. 428 Neurogastroenterology and motility : the official journal of the European Gastrointestinal 429 Motility Society 2011;23:666-e278. 430

38. Tual-Chalot S, Allinson KR, Fruttiger M, Arthur HM. Whole mount 431 immunofluorescent staining of the neonatal mouse retina to investigate angiogenesis in vivo. 432 Journal of visualized experiments : JoVE 2013;e50546. 433

39. Sennlaub F, Courtois Y, Goureau O. Inducible nitric oxide synthase mediates retinal 434 apoptosis in ischemic proliferative retinopathy. The Journal of neuroscience : the official 435 journal of the Society for Neuroscience 2002;22:3987-3993. 436

40. Rey-Funes M, Ibarra ME, Dorfman VB, et al. Hypothermia prevents nitric oxide 437 system changes in retina induced by severe perinatal asphyxia. Journal of neuroscience 438 research 2011;89:729-743. 439

41. Yoshida A, Yoshida S, Khalil AK, Ishibashi T, Inomata H. Role of NF-kappaB-440 mediated interleukin-8 expression in intraocular neovascularization. Investigative 441 ophthalmology & visual science 1998;39:1097-1106. 442

42. Li Q, Verma IM. NF-kappaB regulation in the immune system. Nature reviews 443 Immunology 2002;2:725-734. 444

43. Kratsovnik E, Bromberg Y, Sperling O, Zoref-Shani E. Oxidative stress activates 445 transcription factor NF-kB-mediated protective signaling in primary rat neuronal cultures. 446 Journal of molecular neuroscience : MN 2005;26:27-32. 447

44. Beg AA, Baldwin AS. The I kappa B proteins: multifunctional regulators of Rel/NF-448 kappa B transcription factors. Genes & development 1993;7:2064-2070. 449

45. Koong AC, Chen EY, Giaccia AJ. Hypoxia causes the activation of nuclear factor 450 kappa B through the phosphorylation of I kappa B alpha on tyrosine residues. Cancer 451 research 1994;54:1425-1430. 452

46. Kumar R, Harris-Hooker S, Kumar R, Sanford G. Co-culture of Retinal and 453 Endothelial Cells Results in the Modulation of Genes Critical to Retinal Neovascularization. 454 Vascular cell 2011;3:27. 455

47. Hong S, Lee JE, Kim CY, Seong GJ. Agmatine protects retinal ganglion cells from 456 hypoxia-induced apoptosis in transformed rat retinal ganglion cell line. Bmc Neurosci 457 2007;8:81. 458

48. Tulsawani R, Kelly LS, Fatma N, et al. Neuroprotective effect of peroxiredoxin 6 459 against hypoxia-induced retinal ganglion cell damage. Bmc Neurosci 2010;11:125. 460

49. Choi JS, Sungjoo KY, Joo CK. NF-kappa B activation following optic nerve 461 transection. Korean journal of ophthalmology : KJO 1998;12:19-24. 462

50. Castagné V, Lefèvre K, Clarke PG. Dual role of the NF-kappaB transcription factor in 463 the death of immature neurons. Neuroscience 2001;108:517-526. 464

19

51. Kobayashi T, Oku H, Komori A, et al. Advanced glycation end products induce death 465 of retinal neurons via activation of nitric oxide synthase. Experimental eye research 466 2005;81:647-654. 467

52. Nichol KA, Schulz MW, Bennett MR. Nitric oxide-mediated death of cultured 468 neonatal retinal ganglion cells: neuroprotective properties of glutamate and chondroitin 469 sulfate proteoglycan. Brain research 1995;697:1-16. 470

53. Mishra OP, Zubrow AB, Ashraf QM. Nitric oxide-mediated activation of extracellular 471 signal-regulated kinase (ERK) and c-jun N-terminal kinase (JNK) during hypoxia in cerebral 472 cortical nuclei of newborn piglets. Neuroscience 2004;123:179-186. 473

54. St Clair EG, Anderson SJ, Oltvai ZN. Bcl-2 counters apoptosis by Bax 474 heterodimerization-dependent and -independent mechanisms in the T-cell lineage. The 475 Journal of biological chemistry 1997;272:29347-29355. 476

55. Haldar S, Chintapalli J, Croce CM. Taxol induces bcl-2 phosphorylation and death of 477 prostate cancer cells. Cancer research 1996;56:1253-1255. 478

56. Hu ZB, Minden MD, McCulloch EA. Phosphorylation of BCL-2 after exposure of 479 human leukemic cells to retinoic acid. Blood 1998;92:1768-1775. 480

481

Figure Legends 482

Figure. 1 Western blot analysis showing the protein expression of neuronal nitric oxide 483

synthase (nNOS) in the retina of postnatal rats at 3, 24h, 3, 7 and 14d after hypoxic exposure 484

and their corresponding controls. The upper panel A shows the immunoreactive bands of 485

nNOS (155kDa) and -actin (43kDa). (B) Bar graph showing significant changes in the 486

optical density following hypoxic exposure. Each bar represents the mean ± SD. The 487

experiment was repeated three times and a representative blot is shown here. Significant 488

differences in protein level between hypoxic and control groups are expressed as *P<0.01. 489

(C) Nitric oxide (NO) content in the retina of postnatal rats at 3, 24h, 3, 7 and 14d after 490

hypoxic exposure and their corresponding controls. Significant differences in NO level 491

between hypoxic and control groups are expressed as *P<0.01. 492

Figure. 2 Confocal images showing the distribution of NeuN (A, D; G, J: green), neuronal 493

nitric oxide synthase (nNOS; B, E; red) and nuclear factor- B (NF- B: H, K) in the retinal 494

20

ganglion cells (RGCs; arrows) in the ganglion cell layer (GCL) in the retina at 3d after 495

hypoxic exposure. Co-localized labeling of NeuN with nNOS/NF- B in RGCs is detected in 496

C, F, I, L. Note the increased expression of nNOS/NF- B in RGCs in hypoxic rats (D-F; J-L) 497

when compared to control rats (A-C; G-I) and the increased nuclear translocation of NF- B 498

in the RGCs in hypoxic retina (J-L). Scale bar, A-F = 20 m. 499

Figure. 3 Panel A represents an Enzyme-linked immunosorbent assay analysis showing 500

cytoplasmic and nuclear NF-κB levels in primary cultured retinal ganglion cells (RGCs) in 501

control, hypoxia, hypoxia + 10 M BAY (Hyp+BAY) and hypoxia + 10 M 7-NINA (Hyp+7-502

NINA) groups. Data represents mean ± SD of the fluorescent intensity of p-P65 subunit of 503

NF-κB in cultured RGCs. Significant differences in NF-κB levels in comparison to control 504

group is indicated by *P<0.05, **P<0.01 and with respect to hypoxia as #P<0.05, ##P<0.01. 505

B-G - Confocal images showing the localisation of NF-κB (C, F; red) in Thy1 (green) 506

labelled control (B-D) and hypoxic (E-G) cultured retinal ganglion cells (arrows). Nuclear 507

translocation of NF-κB is evident following hypoxia (F, G). Scale bar, B-G = 20 m. 508

Figure. 4 A. Western blot analysis showing the protein expression of neuronal nitric oxide 509

synthase (nNOS) in retinal ganglion cells (RGCs) of control, hypoxia, hypoxia+10 M 7-510

NINA (Hyp+7-NINA) and hypoxia+10 M BAY (Hyp+BAY) groups. The upper panel shows 511

the immunoreactive bands of nNOS and its corresponding -actin band. Bar graph in panel B 512

shows the significant difference in optical density between control and treatment groups and 513

are indicated by *P<0.01 and #P<0.01. 514

Figure. 5 Confocal images showing the distribution of Thy-1 (A, D, G, J) and neuronal nitric 515

oxide synthase (nNOS; B, E, H, K) in primary cultures of retinal ganglion cells (RGCs) in 516

control, hypoxia, hypoxia +10 M 7-NINA (Hyp+7-NINA) and hypoxia + 10 M BAY 517

(Hyp+BAY) groups. The co-localized expression of nNOS with Thy-1 immunoreactive cells 518

(arrows) can be seen in C, F, I and L. Following hypoxia nNOS expression is upregulated (E, 519

21

F) which is prevented in hypoxic+7-NINA (H, I) and hypoxic+BAY (K, L) groups. Scale bar 520

(A-F): 20 m. 521

Figure. 6 Panel A shows the confocal image of retinal flatmount prepared from a 4d old 522

control rat. B-M- Confocal images showing the apoptosis of Thy-1 (B, E, H, K) positive 523

RGCs (arrows), as marked by caspase-3 labelling (C, F, I, L), on retinal flatmounts prepared 524

from control, hypoxia, hypoxia+20mg/kg BAY (Hyp+BAY) and hypoxia+10mg/kg 7-NINA 525

(Hyp+7-NINA) groups of rats at 3d following hypoxic exposure. The co-localized expression 526

of caspase-3 positive cells and Thy-1 can be seen in D, G, J, M. Bar graph in N represents the 527

significant differences in the percentage of caspase-3 positive RGCs in various groups. 528

Significant differences with respect to control is indicated by *P<0.05, **P<0.01 and with 529

respect to hypoxia as #P<0.05, ##P<0.01. Scale bar, A= 500 m; B-M = 50 m. 530

Figure. 7 Confocal images showing apoptotic cells labeled with Thy-1 (A, D, G, J) and 531

caspase-3 (B, E, H, K) in primary cultured retinal ganglion cells (RGCs) in control, hypoxia, 532

hypoxic+10 M 7-NINA (Hyp+7-NINA) and hypoxic+10 M BAY (Hyp+BAY) groups. The 533

co-localized expression of caspase-3 positive cells and Thy-1 can be seen in C, F, I and L. 534

Panel M, bar graph represents the significant differences in the mean percentage of caspase-3 535

positive RGCs. When hypoxic RGCs were treated with 7-NINA and BAY, the incidence of 536

caspase-3 positive cells is significantly decreased as indicated by *P<0.01; #P<0.01. Scale 537

bar, A-L = 20 m. 538

539

Abbreviations 540

541

BSA : bovine serum albumin 542

Caspase-3 : cas-3 543

GCL : ganglion cell layer 544

22

HIF-1 : hypoxia inducible factor-1 545

NF- B : nuclear factor-kappaB 546

NO : nitric oxide 547

nNOS : neuronal nitric oxide synthase 548

iNOS : inducible nitric oxide synthase 549

eNOS : endothelial nitric oxide synthase 550

7-NINA : 7-nitroindazole 551

PBS : phosphate buffered saline 552

RGCs : retinal ganglion cells 553