Roles of Constitutive Nitric Oxide Synthase in ... - Semantic Scholar

Roles of Constitutive Nitric Oxide Synthase in Postischemic Rat Retina Masanori Hangai,1 Kazuaki Miyamoto,x Kano Hiroi,x Akitaka Tujikawa,x Yuichiro Ogura,1 Yoshihito Honda,x and Nagahisa Yoshimura2 Nitric oxide is a reactive species that could be protective or destructive to the retina depending on the stage of the evolving ischemic process. This study was conducted to obtain a better understanding of the roles of constitutive nitric oxide synthase (cNOS) during reperfusion after ischemia in rat retina.

PURPOSE.

METHODS. Retinal ischemia was induced

for 60 minutes in Sprague-Dawley rats by ligating the optic nerve. Gene expression for endothelial and neuronal nitric oxide synthases (eNOS and nNOS) was studied by reverse transcription-polymerase chain reaction (RT-PCR). To inhibit cNOS, A^'-nitroL-arginine (L-NNA) was injected intraperitoneally four times (every 6 hours) beginning 2 hours after reperfusion, for a total dose of 80 mg/kg. Retinal damage was assessed by the rate of a- and b-wave recovery on electroretinograms and by the thickness of the retinal layers. Retinal circulation and vessel diameter were evaluated by the dye-dilution technique. RESULTS. After ischemia ended, eNOS mRNA initially decreased until 6 hours, then increased to a peak at 12 hours, and decreased progressively beyond 24 hours until the final measurement at 96 hours of reperfusion. nNOS mRNA decreased to nearly undetectable levels during the same measurement periods. L-NNA treatment enhanced reduction of a- and b-wave amplitudes and increased thinning of the inner retina in postischemic eyes. Retinal mean circulation time was markedly prolonged in L-NNA-treated postischemic eyes. Arterial mean transit times were 2.1-fold and 4.5-fold longer in L-NNA-treated postischemic eyes than in L-NNA-treated nonischemic eyes and in D-NNA-treated postischemic eyes, respectively. CONCLUSIONS. This

study shows that postischemic inhibition of NOS worsens retinal damage after ischemia-reperfusion and alters postischemic retinal circulation. Nitric oxide may play an important role in protecting the retina from ischemic injury, possibly by preventing postischemic hypoperfusion. (Invest Ophthalmol Vis Set. 1999;4O:45O-458)

N

itric oxide (NO) is a free radical that plays a variety of roles in the pathophysiology of many diseases. '" 4 It is synthesized by three isoforms of NO synthase (NOS): neuronal (nNOS), endothelial (eNOS), and inducible (iNOS) isoforms; the former two share enzymatic characteristic and are termed constitutive isoforms (cNOS).5'7 All isoforms have been shown to be present or inducible in neural retina.8'12 NO could be generated by a variety of retinal cells, such as retinal neurons, 81013 Miiller cells,1014 astrocytes,12 retinal pigment epithelium,9 endothelial cells,11 pericytes," optic nerves,15 and peripheral ocular nerve fibers.8 There is increasing evidence that NO plays an important role in physiologic maintenance and pathologic processes in the retina.16 Implication of

From the 'Department of Ophthalmology and Visual Science, Kyoto University Graduate School of Medicine; and the department of Ophthalmology, Shinshu University School of Medicine, Matsumoto, Japan. Supported in part by Grant-in-Aid 10470567 for Scientific Research (NY) from the Ministry of Education, Science, Sports and Culture, Tokyo, Japan. Submitted for publication December 8, 1997; revised April 27 and August 20, 1998; accepted September 30, 1998. Proprietary interest category: N. Reprint requests: Nagahisa Yoshimura, Department of Ophthalmology, Shinshu University School of Medicine, Matsumoto 390-8621, Japan.

450

NO in retinal ischemic injury seems to vary depending on the experimental conditions, because previous attempts to reveal the significance of NO in retinal ischemia-reperfusion injury have yielded contradictory results.17"20 Because ischemiareperfusion consists of at least three stages—the ischemic stage, early-stage reperfusion (hyperemic stage), and late-stage reperfusion (posthyperemic stage)—the significance and sources of NO may vary with the stages. Ischemia and reperfusion are thought to mediate retinal damage after retinal ischemic episodes, such as central retinal arterial occlusion and branch retinal arterial occlusion. In general, reperfusion induces a variety of phenomena, including hypoperfusion; synthesis of inflammatory mediators, oxygenderived free radicals and lipid mediators; and activation of blood-derived cells.21"27 To date, inflammatory cytokines and iNOS have been shown to be upregulated at the transcriptional level in postischemic retina during late-stage reperfusion.2829 Neutrophil infiltration into retina has been observed at this stage.28'30 Furthermore, the blood-retinal barrier is known to break down, with a peak 1 day after reperfusion.31 Considering the well-defined nature of NO as a vasoregulatory mediator, it is possible that NOS, especially cNOS, plays a major role in some of the processes associated with retinal reperfusion. In the present study, we focused on the role of cNOS during late-stage reperfusion and examined effects of NOS inhibition Investigative Ophthalmology & Visual Science, February 1999, Vol. 40, No. 2 Copyright © Association for Research in Vision and Ophthalmology

IOVS, February 1999, Vol. 40, No. 2 by A^'-nitro-L-arginine (L-NNA) on the outcome of retinal damage and on retinal blood flow. MATERIALS AND METHODS

Animals All studies were in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Sixtyseven male adult Sprague-Dawley rats weighing approximately 250 g were used. Rats were anesthetized by intramuscular injection of 60 mg/kg ketamine hydrochloride and 10 mg/kg xylazine. The pupils were dilated with topical 1% atropine sulfate and 0.5% phenylephrine. Rats were killed by an overdose of pentobarbital. Sixteen rats were used for the electroretinographic (ERG) experiments and 10 for blood flow analysis. Rats used for the measurement of retinal thickness were the same as those used for the ERG experiments. Thirty-three rats were used for polymerase chain reaction (PCR) and eight for light microscopic studies to observe histologic retinal changes in the L-NNA-treated postischemic eyes such as retinal edema and neutrophil infiltration. All animals were fed standard laboratory chow ad libitum and allowed free access to water in an air-conditioned room with a 12-hour-light-12-hourdark cycle.

Induction of Ischemia and Reperfusion Retinal ischemia and reperfusion were induced as previously described.28 In brief, after the optic nerve of the right eye was exposed, a 6-0 nylon suture was passed behind the optic nerve and tightened until blood flow ceased in all the retinal vessels. Complete nonperfusion was confirmed through a surgical microscope after tightening and before releasing the ligation. Sixty minutes later the suture was removed, and reperfusion of the vessels was observed through the surgical microscope. In preliminary experiments, complete retinal ischemia during the ligature was confirmed by obtaining a flat wave in ERG in all three rats examined.

Reverse Transcription-Polymerase Chain Reaction Reverse transcription-polymerase chain reaction (RT-PCR) was performed as previously described.28 Briefly, first-strand cDNA was synthesized from 5 ng total RNA pretreated by RNase-free DNase (Promega, Madison, WT), and then amplified by a stepped cycle (95°C, 30 seconds; 55°C, 30 seconds; 72°C, 60 seconds) for 25 or 35 cycles (/3-actin, 25 cycles; eNOS and nNOS, 35 cycles). The primers used in this experiment were: 5'-CACCCTCAGGTTCTGTGTGTT-3' (sense) and 5'-GTAGCCTGGAACATCTTCCGT-3' (antisense) for eNOS (GenBank accession no. U02534), 5'-AATCCAGGTGGACAGAGACC-3' (sense) and 5'-TCCTTGAGCTGGTAGGTGCT-3' (antisense) for nNOS,32 and 5'-AGCTGAGAGGGAAATCGTGC-3' (sense) and 5'-ACCAGACAGCACTGTGTTGG-3' (antisense) for /3-actin. The PCR products obtained were extracted from the agarose gel after electrophoresis and were subcloned by using T-vector. Nucleotide sequencing of the subcloned DNA was performed as previously described.28

Drug Administration and Electroretinogram L-NNA and D-NNA were sonicated in normal saline. Rats received four intraperitoneal administrations of L-NNA (20 mg/kg

Constitutive NOS in Postischemic Retina

451

every 6 hours for a total dose of 80 mg/kg)33 beginning 2 hours after cessation of ischemia. In our preliminary experiments monitoring the conversion of [3H]arginine to [3H]citrulline in rat retina, intraperitoneal injection of L-NNA 20 mg/kg inhibited NOS activity to less than 50% at least for 6 hours (15% of NOS at 1 hour, 26% at 3 hours, 37% at 6 hours, and 79% at 9 hours; unpublished data). Other rats received similar administrations of D-NNA as the control. Electroretinograms were recorded as previously described.12 In brief, flash ERGs were recorded from each eye using a photostimulator lamp placed in front of the eye with maximum light intensity (3500 lux on the surface of die cornea). After dark adaptation for at least 60 minutes, a carbon electrode (NEC San-ei, Tokyo, Japan) was placed on the cornea, and stainless steel needle electrodes (NEC San-ei) were placed under the skin of the nose and the tail, which served as a reference and ground, respectively. The responses of four trials were amplified and averaged.

Light Microscopic Examination and Quantification of Retinal Thickness Immediately after enucleation, the eyes were cut open and fixed in 1.5% formaldehyde and 1% glutaraldehyde in phosphate buffer. Sagittal 5-ju.m-thick paraffin sections through the optic nerve were obtained and stained with hematoxylin and eosin. Retinal ischemic damage was evaluated by measuring the thickness of the retina. Parameters used were mean thickness from the outer limiting membrane to inner limiting membrane (OLM-ILM), that of inner retina from the ILM to the interface of the outer plexiform layer and the outer nuclear layer (ONL); and that of the outer nuclear layer.3435 In each of the superior and inferior hemispheres, five sets of color photographs were taken with a Coolscan camera (Nikon, Tokyo, Japan) and transferred to the computer. The thickness of each layer was determined by measuring six regions of each of the five photographs on the computer screen. Finally, the retinal thickness values were obtained from the total of 60 measurements per each eye. An average retinal thickness for each eye was obtained by averaging the 60 thickness values.

Scanning Laser Ophthalmoscope-Based Fluorescein Angiography Retinal blood flow was examined 24 hours after reperfusion, as previously described.36 The animals were anesthetized with a 1:1 mixture of 10 mg/kg ketamine hydrochloride and 4 mg/kg xylazine, and the pupils were dilated with 0.5% tropicamide and 2.5% phenylephrine hydrochloride. In each rat, a 24-gauge catheter was inserted into the tail vein. After a 10-/xl bolus of 10% sodium fluorescein (Fluorescite; Alcon, Fort Worth, TX) was injected rapidly (in approximately 0.1 seconds) into the tail vein catheter, fluorescein fundus angiograms with the optic disc centered in the field of view were observed with a scanning laser ophthalmoscope in the 40° field and were recorded on a high-speed videotape at the video rate of 30 frames/sec. At least three fluorescein angiographic recordings were repeated in the same animal. The femoral arterial pressure was measured by direct cannulation. The mean arterial pressure was approximately 100 mm Hg to 110 mm Hg in each rat. In preliminary experiments there was no difference in the mean arterial blood

452

Hangai et al.

IOVS, February 1999, Vol. 40, No. 2

pressure measured before the injection compared with that measured after the injection of fluorescein.

Video Image Analysis The recorded images were digitized through an analog-digital converter board (Video Vision; Radius, San Jose, CA) and entered into a computer. The digitized images were composed of 640 X 480 pixels with a gray scale of 256 levels. To determine the gray level as an indication of the fluorescence intensity of each retinal artery, the measurement window, a circle with a diameter slightly larger than the diameter of the vessel,37 was located at the edge of the optic disc on the vessel. The average gray levels indicating the average fluorescence intensities within the measurement windows were measured frame by frame. When the data obtained were plotted with the time on the abscissa and the intensity on the ordinate, they generated a typical dye-dilution curve. The recirculation phase was almost negligible because of the small close of injected fluorescein dye. Resultant dye- dilution curves were regressed to a log-normal distribution function. Regressed parameters of the log-normal function were used to determine the arterial mean transit time (AMTT). The AMTT was denned as the time between the appearance of the dye in the measurement window and the transit time of half the amount of the dye. The appearance time of the dye was determined as the time that fluorescence intensity within the measurement window was greater than the background intensity level. The AMTT values were obtained generally from six arteries. An average AMTT for each angiogram was obtained by averaging the six AMTT values. The AMTT for each rat was calculated as the average of the AMTT values obtained from at least three angiograms. Vessel diameters were measured on a computer monitor in units of pixels and then compared. Measurements were obtained at a 1-disc-diameter distance from the center of the optic disc in monochromatic images recorded before fluorescein angiography. Each vessel diameter was calculated as the distance between the half-height points determined separately on each side of the density profile of the vessel image. The averages of the individual arterial and venous diameters were used as the arterial and venous diameters for each rat. The observers were masked to which treatment was administered to the rat until after the analysis had been completed. Statistical Analysis The values of ERG a- and b-wave amplitudes, measurements of the retinal layers, AMTTs, and vessel diameters were compared between groups by one-way analysis of variance followed by Fisher's protected least significant difference test or Scheffe's post hoc test and Student's f-test, as appropriate. The comparison of ERG a- and b-wave amplitude between preischemic and postischemic levels were made by the paired f-test. All values are expressed as mean ± SD. RESULTS Gene Expression of eNOS and nNOS in Postischemic Retina RT-PCR analysis using specific primers for eNOS and nNOS yielded bands of the expected sizes (eNOS, 357 bp; nNOS, 498 bp; Figs. 1A, IB). By nucleotide sequencing of the PCR prod-

Time after Reperfusion (Hours)

I

I

N 0 1/4 1/2 1 3 6 12 24 48 96

eNOS nNOS (1-actin

i

FIGURE 1. RT-PCR analysis of gene expression for eNOS and nNOS. Specific amplification of eNOS (A) and nNOS (B) mRNA. First-strand cDNAs used in this experiment were obtained from normal retina. For PCR amplification, 35 cycles were used for eNOS and nNOS. Each DNA was confirmed to be derived from the target cDNA sequences by nucleotide sequencing. 0X174 Hae\W fragments were used as size markers. (C) Representative electrophoresis pattern of PCR products, The gel was stained with ethidium bromide. Lane N, retina from nonsurgical rats. Time 0 indicates the retina that was isolated at the end of ischemia. |3-Actin was used as the control (25 cycles). At least three experiments were performed at each time point. ucts, w e confirmed t h e amplicons to b e derived from t h e target

cDNAs (graphic data not shown). To monitor levels of gene expression for eNOS and nNOS, we performed a time course study of the gene expression (Fig. 1Q. Expression of the j3-actin gene showed steady levels when observed in gels stained by ethidium bromide. nNOS gene expression was detected in normal sensory retina, but there was much less expression as early as 15 minutes after ischemia. Expression continued to decrease and reached the lowest level 12 hours after ischemia. This low level of expression continued for 2 days. The eNOS gene was moderately expressed in normal retina. eNOS gene expression appeared to decrease from the end of ischemia and to show lower levels than preischemic levels until 3 hours after reperfusion, but then increased to greater than preischemic levels from 6 to 24 hours, with peak expression at 12 hours. eNOS gene expression appeared to be equal to or lower than preischemic levels by 48 to 96 hours. Effects of L-NNA on the Rate of Electroretinogram a- and b-Wave Amplitude The amplitude of a- and b-waves was significantly decreased by the L-NNA treatment in nonischemic eyes (a-wave: P = 0.031



A •

D-NNA + Y77\ D-NNA + v%m L-NNA + ^m Ischemia " ^ Fellow Fellow

L-NNA + Ischemia

400 -

453

20% and 22% decreases in a-wave amplitudes of r>NNA-treated postischemic eyes on days 1 and 3 after reperfusion, respectively (Fig. 2A). There were approximately 49% and 62% decreases in a-wave amplitudes of L-NNA-treated postischemic eyes on days 1 and 3, respectively (Fig. 2A). The data indicate that L-NNA treatment decreased a-wave amplitudes by 2,4-fold (P = 0.043) and 2.8-fold (P = 0.011) compared with the effect of D-NNA treatment on days 1 and 3, respectively. There were approximately 59% and 48% decreases in b-wave amplitudes of D-NNA-treated postischemic eyes compared with amplitudes in their fellow eyes on days 1 and 3 after reperfusion, respectively (Fig. 2B). There were approximately 83% and 77% decreases in b-wave amplitudes of L-NNA-treated postischemic eyes compared with amplitudes in their fellow eyes on days 1 and 3, respectively (Fig. 2B). These indicate that L-NNA treatment decreased b-wave amplitudes by 1.4-fold (P = 0.0031) and 1.6-fold (P = 0.0020) compared with amplitudes after D-NNA treatment on days 1 and 3, respectively.

Time after reperfusion

Effects of L-NNA on Retinal Histology and Thickness

B I

To see whether there were histologic changes in the retina and vessels of L-NNA-treated postischemic eyes, a light micro-

1 D-NNA +F77I D-NNA + Fellow k - " 1 Ischemia

1200 -

*»«" Jfes pre

1 day

3 days

Time after reperfusion

FIGURE 2, Recording of ERG a-wave (A) and b-wave (B) amplitudes in eyes 1 and 3 days after ischemia. Amplitudes of a- and b-waves were significantly decreased by the L-NNA treatment in nonischemic eyes. The percentages shown are the rate of decrease in a- or b-wave amplitudes of postischemic retinas compared with their fellow eyes. There were 2.4- and 2.8-fold decreases in a-wave amplitudes on days I and 3, respectively, in L-NNA-treated groups compared with amplitudes in D-NNA-treated groups. There were 1.4- and 1.6-fold decreases in b-wave amplitudes on days 1 and 3, respectively, in i.-NNA-treated groups compared with those in i>NNA-treated groups. Results represent mean ± SD; n = 8. (A) *P = 0.031; **P = 0.0046; ***P = 0.043; ••••/» = 0.011. (B) V = 0.005; **P = 0.0099; ***P = 0.0031; ****P = 0.0020. Differences were determined by analysis of variance followed by Fisher's protected least significant difference between groups and by paired /-test between preischemic and postischemic levels. Fellow indicates nonischemic control retinas from fellow eyes.

on day 1,P = 0.0046 on day 3; b-wave; P = 0.005 on day l,P = 0.0099 on day 3; Fig. 2), although not by the D-NNA treatment. We therefore compared the amplitude of ischemic eyes with that of their fellow eyes. L-NNA treatment enhanced the decrease in postischemic amplitude of the a- and b-waves when compared with D-NNA treatment. There were approximately

GCL ¥

iNLrfy;< FIGURE 3. Representative photographs showing aggregates of blood cells and prolonged retinal edema. Shown are sections from nits killed at 24 hours after reperfusion. (A) D-NNA-treated postischemic retina. (B) L-NNA-treated postischemic retina. Apparent retinal edema in the nerve fiber layer, ganglion cell layer (GCL), and the inner plexiform layer was observed in (B), but not in (A). Note that aggregates of a large number of red blood cells and a few neutrophils (arrows) can be seen. Asterisks indicate retinal large vessels. INL; inner nuclear layer. Scale bar, 20 ftm.

454

Hangai et al.


GCL

RPEA FIGURE 4. Light micrographs showing retinal thickness on day 7 after ischemia. Representative photographs taken in the region 0.5 mm to 1 mm from the optic disc. (A) D-NNA-treated nontschemic eye, (B) i.-NNA-treated nonischemic eye, (C) D-NNA-treated postischemic eye, and (D) i.-NNA-treated postischemic eye. Five-micrometer sagittal sections through the optic nerve head were prepared and stained with hematoxylin and eosin. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; RPE, retinal pigment epithelium. Scale bar, 20 p.m.

scopic study was performed. Strong edema was observed in L-NNA-treated postischemic retinas, with a peak at 24 hours, but not in r>NNA-treated postischemic retinas (Tig. 3)- Edema was prominent in the inner plexiform layer, the ganglion cell layer, and the nerve fiber layer throughout the retina. Retinal vessels were filled with aggregates of a large number of red blood cells and a few neutrophils in L-NNA-treated postischemic retinas, but the aggregates were not seen in D-NNAtreated postischemic retinas. To further determine whether L-NNA treatment exacerbates ischemia-reperfusion injury, mean thickness of the retinal layers was quantified in eyes 7 days after reperfusion (n = 7 or 8; Figs. 4, 5). There were no significant differences in the mean thickness of all three layers between D-NNA- and L-NNAtreated nonischemic eyes. But, there were approximately 16% and 33% decreases in overall retinal thickness (OLJVHLM) in the i>NNA- and L-NNA-treated postischemic eyes, respectively, compared with retinal thickness in their fellow eyes. This means that L-NNA treatment decreased overall retinal thickness twofold compared with D-NNA treatment. There were approximately 23% and 50% decreases in the inner retinal layers in the D-NNA- and i.-NNA-treated postischemic eyes, respectively, compared with thickness in their fellow eyes. L-NNA treatment decreased the inner retinal layers by twofold compared with i>NNA treatment. No significant changes in thickness of the ONL were detected between the groups, although there was disorganization in the pattern of photoreceptor nuclei and the inner and outer segments in some of areas without uniformity in ischemic groups.

Fundus Observation Fundus observation was performed through the surgical microscope at 6, 12, 18, and 24 hours after reperfusion. Venous blood flow in L-NNA-treated postischemic eyes was obviously abnormal, with slow flow of clusters of red cells through the large veins followed by empty gaps within the tubular space.

Effects of L-NNA on the Retinal Blood Flow In an attempt to explain the abnormal blood flow, we performed scanning laser ophthalmoscope- based fluorescein an-

giography (« - 4 or 5). The AMTTs were quantitated in all groups, but the venous mean transit times (VMTTs) and, as a result, the mean circulation times (MCTs) could not be determined in L-NNA-treated postischemic eyes, because venous flow was too slow,38 and because there were multiple defects in fluorescence on the venous images (n = 5; Fig. 6E). The multiple defects on the venous images were not observed in any other eyes in this study, and, to our knowledge, have not been reported. As shown in Figure 6, however, it was obvious

D-NNA + r m D-NNA + V7m L--NNA + ^m L-NNA +

OLM-ILM

Inner Retina

ONL

FIGURE 5. Thickness of the retinal layers in eyes 7 days after ischemia. There were no significant differences in mean thickness of all three layers between D-NNA- and L-NNA-treated nonischemic eyes. The percentages shown are the rate of decrease in the thickness of postischemic retinas compared with their fellow eyes. There was a twofold decrease in overall retinal thickness (OLM-ILM) in L-NNA-treated groups compared with decrease in thickness in o-NNA-treated groups. There was a 2.2-fold decrease in the inner retinal layers in L-NNAtreated groups compared with the decrease in D-NNA-treated groups. There were no significant changes in thickness of the ONL between any groups. Results represent mean ± SD (n = 7 or 8). *P = 0.009; **P = 0.0124. Differences were determined by analysis of variance followed by Scheffe's test. Fellow indicates nonischemic control retinas from fellow eyes.,



455

that the mean circulation time in i.-NNA-treated postischemic eyes was extremely prolonged. As shown in Figure 7A, the AMTT in the L-NNA-treated nonischemic group was 2.4-fold longer than that in the r>NNAtreated nonischemic group. The AMTT in the i.-NNA-treated postischemic group was 2.1-fold longer than that in the L-NNAtreated nonischeniic group and was 4.5-fold longer than that in the D-NNA-treated postischemic group. There were no statistically significant differences in the AMTTs between D-NNAtreated postischemic and o-NNA-treated nonischemic groups.

B

D

Arteries

FIGURE 6. Representative sequential fluorescein angiographic flame obtained with a scanning laser ophthalmoscope. The numbers in the lower right indicate the times (in seconds) after the appearance of dye in the central retinal artery. At least three fluorescein angiographic recordings were repeated in the same animal. (A) Normal retina, (B) i>NNA-treated postischemic retina, (C) L-NNA-treated nonischemic retina, and (D) L-NNA-treated postischemic retina. Retinal circulation in i.-NNA-treated postischemic eyes was markedly prolonged compared with that in the other groups. Note the long-lasting arterial phase and marked delay in appearance of the venous phase'in L-NNA-treatcd postischemic retina. (E) Venous phase in L-NNA-treated postischemic retina. The venous flow was slow, and there were multiple defects of fluorescence, probably aggregated blood cells, in the venous phase images. The arrow indicates the same filling defect; it took 6 seconds to flow from the position in the upper left to the one shown in the lower right.

Veins

FIGURE 7. Effects of L-NNA on arterial mean transit times (AMTTs) and vessel diameters. (A) The AMTT in the L-NNA-treated nonischeniic group was significantly longer than that in the [>NNA-treated nonischemic group. The AMTT in the L-NNA-treated postischemic group was significantly prolonged compared with that in the i.-NNA-treated nonischemic group and the D-NNA-treated postischemic group. There were no significant differences in the AMTTs between i>NNA-treated postischemic and D-NNA-treated nonischemic groups. *P = 0.0002 by unpaired £-test; **P = 0.018 versus L-NNA-treated nonischemic eyes and 0.0012 versus D-NNA-treated postischemic eyes by analysis of variance (ANOVA) followed by Scheffe's test. (B) The arterial diameters in the L-NNA-treated postischemic group were significantly constricted compared with those of the i>NNA-treated nonischeniic group. The venous diameters in the i>NNA-treated postischemic group were significantly dilated compared with those of the D-NNAtreated nonischemic group (.P < 0.0001 by ANOVA followed by Scheffe's test). This dilation was significantly inhibited by L-NNA treatment. Fellow indicates nonischemic control retinas from fellow eyes. *P = 0.0095 by unpaired /-test; "P < 0.0001 by ANOVA followed by Scheffe's test; ***P = 0.0001 by unpaired t-test. All results represent mean ± SD; n — 4 or 5-

456

Hangai et al.

Effects of L-NNA on Retinal Vessel Diameters The venous diameter in the D-NNA-treated postischemic group increased 1.6-fold compared with dilation in the D-NNAtreated nonischemic group (Fig. 7B). There was a significant 33% reduction in retinal venous diameter in L-NNA-treated postischemic eyes compared with i>NNA-treated postischemic ones. There were no significant differences between L-NNA-treated postischemic eyes and i>NNA-treated nonischemic eyes, indicating that venous dilation after ischemia was reduced by L-NNA.

DISCUSSION The ERG experiments showed that the postischemic NOS inhibition by L-NNA increased the reduction of a- and b-wave amplitudes (Fig. 2). However, the L-NNA treatment reduced the amplitudes especially of the b-wave not only in postischemic eyes but also in nonischemic eyes. This reduction of b-wave amplitude after NOS inhibition in nonischemic eyes is consistent with the findings of Ostwald et al.20'39 The decrease in ERG wave amplitudes in nonischemic eyes may result from inhibition of physiologic retinal functions such as visual transduction.l6 NOS inhibition seems to have such direct effects on retinal functions, and it therefore may not be appropriate to monitor retinal damages using ERG alone when NOS inhibitors are used,39 although it is still a useful parameter for evaluating retinal damage that can reflect retinal phototransductive activities. To verify further that NOS inhibition by L-NNA increases retinal ischemia-reperfusion injury, we measured thickness of various retinal layers (Figs. 4, 5). The L-NNA treatment itself did not change thickness of any layer examined in this study. The ischemic insult reduced the thickness of the inner retina, and the L-NNA treatment significantly enhanced this reduction. Together with the results from ERG experiments, these data demonstrate that postischemic NOS inhibition by the L-NNA treatment aggravated retinal ischemia-reperfusion injury, particularly in the inner retina. In our previous study, retinal ischemia-reperfusion injury was partially ameliorated by another NOS inhibitor, L-NIO, administered on the same time schedule.12 It has been reported that L-NNA showed a marked preference for inhibiting cNOS (nNOS and eNOS) and that TV-iminoethyl-L-ornithine (L-NIO) far more effectively inhibits ,iNOS compared with L-NNA. 40 " 42 The varying results with types of NOS inhibitors suggest that the roles of NO in retinal ischemia-reperfusion injury vary with the particular NOS isoforms. In parallel with the decrease in ERG a- and b-wave amplitudes and retinal thickness, retinal reperfusion was markedly inhibited in L-NNA-treated postischemic eyes (Figs. 6, 7). Conflicting results have been reported about the effects of NOS inhibition on the resting retinal blood flow.20'43"45 Ostwald et al.39 observed that NOS inhibition decreased retinal blood flow when combined with adenosine receptor blockade. In the present study, L-NNA treatment seemed to increase retinal circulation times in nonischemic eyes (Figs. 6A, 6C). However, inhibition of retinal circulation after L-NNA treatment in postischemic eyes was by far more severe than in nonischemic eyes (Figs. 6C, 6D). Postischemic eyes treated with L-NNA showed apparently abnormal retinal blood flow, which to our knowledge has not been reported previously.

IOVS, February 1999, Vol. 40, No. 2 The mechanisms by which the L-NNA treatment caused the severe inhibition of postischemic circulation are uncertain but seem to result from local vascular responses. The prolongation of AMTT suggests increased vascular resistance. The prolongation of AMTT induced by the L-NNA treatment was more in postischemic eyes than in nonischemic fellow eyes, although ischemia-reperfusion alone caused no significant delay in the AMTT (Fig. 7). Although increased systemic blood pressure may be involved in the prolonged AMTTs in the L-NNA-treated rats, this marked difference in AMTT between L-NNA-treated postischemic and fellow eyes implies that some local vascular events that increase retinal vascular resistance and can be inhibited by NO occurred after L-NNA treatment in postischemic retina. Retinal vascular resistance increases if vasoconstriction occurs in retinal vessels. In the present study, L-NNA treatment inhibited postischemic venous dilation and may have caused an increase in retinal vascular resistance. It has been suggested that NO plays a role in the regulation of retinal microcirculation possibly through modulating the tone of pericytes.'1 ' 46 ' 47 It may be, therefore, that vasoconstriction at the microvascular level occurs in the presence of L-NNA treatment and leads to increased local vascular resistance. Rheologic responses, such as increased capillary plugging of neutrophils and decreased erythrocyte deformability, may be involved in the postischemic vascular effects that increase retinal vascular resistance. NO inhibits aggregation of platelet and neutrophils and adhesion of neutrophils on endothelium,48 and it has been suggested that NO also preserves erythrocyte deformability.49 Aggregates of a large number of erythrocytes and a few neutrophils observed in L-NNA-treated eyes may show one of the rheologic abnormalities that could be induced by NOS inhibition in postischemic eyes (Fig. 3). Results in a number of studies, including those in which knockout mice were deficient in nNOS or eNOS, suggest that nNOS is toxic and eNOS is beneficial in brain ischemia,50'51 although plenty of conflicting results have been also reported. Because L-NNA was used only during late-stage reperfusion, the beneficial role of eNOS that this study suggests is confined to late-stage reperfusion. Vorwerk et al.52 could find no beneficial effects in retinal ischemic injury between eNOS-deficient and control mice. This discrepancy may suggest either that eNOS is toxic in the ischemic stage or that nNOS in the peripheral nerve fibers in and around the choroid8 and in nerve fibers leading to the ophthalmic artery53 is important in the regulation of postischemic retinal circulation. Alternatively, if there are eNOS homologues expressed in retina, genetic deletion of the known eNOS gene could fail to inhibit the unknown retinal homologues, which might compensate the deficient roles of the deleted known eNOS gene. The results of this study raise a possibility that NO donors may be one of the promising agents that can improve retinal circulation after fibrinolytic therapy in central and branch retinal artery occlusion. Nitric oxide donors must be used carefully, because high concentrations of NO could be toxic to neural retina.54 However, inhibition of NOS has received much attention as one of the promising neuroprotective therapies.51 Our findings suggest that NOS inhibition must be effected without inhibiting the beneficial effects of NO on retinal reperfusion.


References 1. Schmidt HH, Walter U. NO at work. Cell. 1994;78:919-925. 2. Stewart AG, Phan LH, Grigoriadis G. Physiological and pathophysiological roles of nitric oxide. Microsurgery. 1994;15:693-702. 3. Billiar TR. Nitric oxide. Novel biology with clinical relevance. Ann Surg. 1995;221:339-349. 4. Kuo PC, Schroeder RA. The emerging multifaceted roles of nitric oxide. Ann Surg. 1995;221:220-235. 5. Dawson TM, Snyder SH. Gases as biological messengers: nitric oxide and carbon monoxide in the brain. J Neurosci. 1994;14: 5147-5159. 6. Marietta MA. Nitric oxide synthase: aspects concerning structure and catalysis. Cell. 1994;78:927-930. 7. Nathan C, Xie QW. Nitric oxide synthases: roles, tolls, and controls. Cell. 1994;78:915-918. 8. Yamamoto R, Bredt DS, Snyder SH, Stone RA. The localization of nitric oxide synthase in the rat eye and related cranial ganglia. Neuroscience. 1993:54:189-200. 9. Goureau O, Lepoivre M, Becquet F, Courtois Y. Differential regulation of inducible nitric oxide synthase by fibroblast growth factors and transforming growth factor b' in bovine retinal pigmented epithelial cells: inverse correlation with cellular proliferation. Proc NatlAcadSci USA. 1993:90:4276-4280. 10. Liepe BA, Stone C, Koistinaho J, Copenhagen DR. Nitric oxide synthase in Miiller cells and neurons of salamander and fish retina. J Neurosci. 1994;l4:764l-7654. 11. Chakravarthy U, Stitt AW, McNallyJ, Bailie JR, Hoey EM, Duprex P. Nitric oxide synthase activity and expression in retinal capillary endothelial cells and pericytes. Curr Eye Res. 1995:14:285-294. 12. Hangai M, Yoshimura N, Hiroi K, Mandai M, Honda Y. Inducible nitric oxide synthase in retinal ischemia-reperfusion injury. Exp Eye Res. 1996;63:501-50913. Venturini CM, Knowles RG, Palmer RM, Moncada S. Synthesis of nitric oxide in the bovine retina. Biochem Biopbys Res Commun. 1991:180:920-925. 14. Goureau O, Hicks D, Courtois Y, De Kozak Y. Induction and regulation of nitric oxide synthase in retinal Miiller glial cells. J Neurochem. 1994;63:310-317. 15. Yamamoto R, Bredt DS, Dawson TM, Snyder SH, Stone RA. Enhanced expression of nitric oxide synthase by rat retina following pterygopalatine parasympathetic denervation. Brain Res. 1993; 631:83-88. 16. Goldstein IM, Ostwald P, Roth S. Nitric oxide: a review of its role in retinal function and disease. Vision Res. 1996;36:2979-2994. 17. Maynard Kl, Chen D, Arango PM, Ogilvy CS. Nitric oxide produced during ischemia improves functional recovery in the rabbit retina. Neuroreport. 1996;8:81 - 85. 18. Veriac S, Tissie G, Bonne C. Oxygen free radicals adversely affect the regulation of vascular tone by nitric oxide in the rabbit retina under high intraocular pressure. Exp Eye Res. 1993:56:85-88. 19. Geyer O, Almog J, Lupu-Meiri M, Lazar M, Oron Y. Nitric oxide synthase inhibitors protect rat retina against ischemic injury. FEBS Lett. 1995:374:399-402. 20. Ostwald P, Goldstein IM, Pachnanda A, Roth S. Effect of nitric oxide synthase inhibition on blood flow after retinal ischemia in cats. Invest Ophtbalmol Vis Sci. 1995:36:2396-2403. 21. Hossmann KA. Ischemia-mediated neuronal injury. Resuscitation. 1993:26:225-235. 22. Conger JD, Weil JV. Abnormal vascular function following ischemia-reperfusion injury. J Invest Med. 1995:43:431-442. 23- Mullane KM, Westlin W, Kraemer R. Activated neutrophils release mediators that may contribute to myocardial injury and dysfunction associated with ischemia and reperfusion. Ann NYAcad Sci. 1988:524:103-121. 24. Zimmerman BJ, Granger DN. Mechanisms of reperfusion injury. Am J Med Sci. 1994;307:284-292. 25. Hansen PR. Role of neutrophils in myocardial ischemia and reperfusion. Circulation. 1995:91:1872-1885. 26. Williams FM. Neutrophils and myocardial reperfusion injury. Pharmacol Tber. 1996;72:1-12.


457

27. Kuroda S, Siesjo BK. Reperfusion damage following focal ischemia: pathophysiology and therapeutic windows. Clin Neurosci. 1997; 4:199-212. 28. Hangai M, Yoshimura N, Yoshida M, Yabuuchi K, Honda Y. Interleukin-1 gene expression in transient retinal ischemia in the rat. Invest Ophthalmol Vis Sci. 1995:36:571-578. 29- Hangai M, Yoshimura N, Honda Y. Increased cytokine gene expression in rat retina following transient ischemia. Ophthalmic Res. 1996;28:248-254. 30. Szabo ME, Droy-Lefaix MT, Doly M, Carre C, Braquet P. Ischemia and reperfusion-induced histologic changes in the rat retina. Demonstration of a free radical-mediated mechanism. Invest Ophthalmol Vis Sci. 1991:32:1471-1478. 31. Wilson CA, Berkowitz BA, Funatsu H, et al. Blood-retinal barrier breakdown following experimental retinal ischemia and reperfusion. Exp Eye Res. 1995;6l:547-557. 32. Bredt DS, Hwang PM, Glatt CE, Lowenstein C, Reed RR, Snyder SH. Cloned and expressed nitric oxide synthase structurally resembles cytochrome P450 reductase. Nature. 1991:351:714-718. 33. Traystman RJ, Moore LE, Helfaer MA, et al. Nitro-L-arginine analogues. Dose- and time-related nitric oxide synthase inhibition in brain. Stroke. 1995:26:864-869. 34. Unoki K, LaVail MM. Protection of the rat retina from ischemic injury by brain-derived neurotrophic factor, ciliary neurotrophic factor, and basic fibroblast growth factor. Invest Ophthalmol Vis Sci. 1994;35:907-915. 35. Unoki K, Ohba N, Arimura H, Muramatsu H, Muramatsu T. Rescue of photoreceptors from the damaging effects of constant light by midkine, a retinoic acid-responsive gene product. Invest Ophthalmol Vis Sci. 1994:35:4063-4068. 36. Miyamoto K, Ogura Y, Nishiwaki H, et al. Evaluation of retinal microcirculatory alterations in the Goto-Kakizaki nit. A spontaneous model of non-insulin-dependent diabetes. Invest Ophthalmol Vis Sci. 1996:37:898-905. 37. Riva CE, Feke GT, Ben-Sira 1. Fluorescein dye- dilution technique and retinal circulation. AmJ Physiol. 1978;234:H315-H322. 38. Blair NP, Feke GT, Morales-Stoppello J, et al. Prolongation of the retinal mean circulation time in diabetes. Arch Ophthalmol. 1982; 100:764-768. 39- Ostwald P, Park SS, Toledano AY, Roth S. Adenosine receptor blockade and nitric oxide synthase inhibition in the retina: impact upon post-ischemic hyperemia and the electroretinogram. Vision Res. 1997:37:3453-3461. 40. Ishii K, Chang B, Kerwin J Jr, Huang ZJ, Murad F. Nw-nitro-Larginine: A potent inhibitor of endothelium-derived relaxing factor formation. Eur J Pharmacol. 1990;176:219-223. 41. Lambert LE, Whitten JP, Baron BM, Cheng HC, Doherty NS, McDonald IA. Nitric oxide synthesis in the CNS endothelium and macrophages differs in its sensitivity to inhibition by arginine analogues. Life Sci. 1991;48:69-75. 42. McCall TB, Feelisch M, Palmer RM, Moncada S. Identification of N-iminoethyl-L-ornithine as an irreversible inhibitor of nitric oxide synthase in phagocytic cells. Br J Pharmacol. 1991 ;102:234 -238. 43. Deussen A, Sonntag M, Vogel R. L-arginine-derived nitric oxide: a major determinant of uveal blood flow. Exp Eye Res. 1993;57:129 134. 44. Seligsohn EE, Bill A. Effects of NG-nitro-L-arginine methyl ester on the cardiovascular system of the anaesthetized rabbit and on the cardiovascular response to thyrotropin-releasing hormone. Br J Pharmacol. 1993:109:1219-1225. 45. Gidday JM, Zhu Y. Nitric oxide does not mediate autoregulation of retinal blood flow in newborn pig. Am J Physiol. 1995;269: H1065-H1072. 46. Haefliger IO, Zschauer A, Anderson DR. Relaxation of retinal pericyte contractile tone through the nitric oxide-cyclic guanosine monophosphate pathway. Invest Ophthalmol Vis Sci. 1994;35: 991-997. 47. Roufail E, Stringer M, Rees S. Nitric oxide synthase immunoreactivity and NADPH diaphorase staining are co-localised in neurons closely associated with the vasculature in rat and human retina. Brain Res. 1995:684:36-46.

458

Hangai et al.

48. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA. 1991; 88:4651-4655. 49. Korbut R, Gryglewski RJ. Nitric oxide from polymorphonuclear leukocytes modulates red blood cell deformability in vitro. Eur J Pharmacol. 1993;234:17-22. 50. Samdani AF, Dawson TM, Dawson VL. Nitric oxide synthase in models of focal ischemia. Stroke. 1997;28:1283-1288. 51. ladecola C. Bright and dark sides of nitric oxide in ischemic brain injury. Trends Neurosci. 1997;20:132-139.

IOVS, February 1999, Vol. 40, No. 2 52. Vorwerk CK, Hyman BT, Miller JW, et al. The role of neuronal and endothelial nitric oxide synthase in retinal excitotoxicity. Invest Ophthalmol Vis Sci. 1997;38:2038-2044. 53. Elsas T, Uddman R, Sundler F. Vasoactive intestinal peptide- and nitric oxide synthase-containing nerve fibers in the rat ophthalmic artery have different origins. Graefes Arch Clin Exp Ophthalmol. 1996;234:125-130. 54. Kashii S, Mandai M, Kikuchi M, et al. Dual actions of nitric oxide in N-methyl-D-aspartate receptor-mediated neurotoxicity in cultured retinal neurons. Brain Res. 1996;711:93-101.