The Effect of Nitric Oxide on Retinal Blood Flow during Hypoxia in Cats

Abstract

purpose. To investigate how nitric oxide (NO) contributes to the regulation of retinal circulation during rest and hypoxia in cats.

methods. N ^G-nitro-l-arginine-methyl ester (l-NAME; n = 7), an NO synthase inhibitor; N ^G-nitro-d-arginine methyl ester (d-NAME; n = 6), the inactive isomer; or phosphate-buffered saline (PBS; n = 7) was injected intravitreously. Hypoxia was induced in the cats by the administration of 10% oxygen. Vessel diameter and blood velocity were measured simultaneously in the anesthetized cats with a laser Doppler velocimetry system, and the retinal blood flow (RBF) and wall shear rate (WSR) were calculated as an index of shear stress on the retinal vessel wall.

results. After intravitreous injection of l-NAME, the vessel diameter (−8.1% ± 2.0%, P < 0.01), velocity (−17.0% ± 3.7%, P < 0.01), and RBF (−29.4% ± 4.6%, P < 0.01) significantly decreased compared with the preinjection level. In the PBS group, maximum increases above the prehypoxia diameter (13.7% ± 3.5%, P < 0.01), velocity (39.5% ± 8.6%, P < 0.01), RBF (73.9% ± 10.6%, P < 0.01), and WSR (28.3% ± 7.1%, P < 0.01) were observed during hypoxia. In the l-NAME group, those changes substantially decreased in response to hypoxia. d-name was inactive with regard to RBF during rest and hypoxia.

conclusions. These results suggest that NO may be continuously produced in the retina during rest and contributes to increased RBF during hypoxia. In addition, the increased WSR during hypoxia indicates that NO plays an important role in retinal hypoxic hyperemia through a flow-induced mechanism.

It is well known that a decreased blood oxygen level leads to vasodilation and increased blood flow, to maintain oxygen delivery to retinal tissues. When blood flow increases, blood vessels are dilated in the microvasculature by a flow-induced mechanism ^{1 2} that has been shown to be an adaptive response of the vessels to the change in blood flow that maintains a constant level of shear stress on the vascular wall. ³ The wall shear rate (WSR), used as an indicator of shear stress on the vessel wall, is calculated from the vessel diameter and blood velocity by assuming a parabolic velocity profile. ^{4 5 6} Although a conventional bidirectional laser Doppler velocimetry system has been used to measure the absolute blood velocity in retinal vessels in numerous studies, it is difficult to estimate the WSR with this technique, because velocity and diameter are not measured simultaneously in the same vessel. Recently, a new laser Doppler velocimetry system has been developed that enables the simultaneous measurement of vessel diameter and blood velocity. ^{7 8} This method may give us a new opportunity to evaluate the change in the shear stress on the retinal vessel wall by calculating the WSR. 

The endothelium is an important source of substances that constrict or relax the vascular smooth muscle. One important factor in vasodilation is an endothelium-derived relaxing factor, which has been identified as nitric oxide (NO). ⁹ Release of NO can be stimulated by physical factors, such as the shear stress that acts on the endothelial surface, ^{10 11} but the physiologic role of NO in the local regulation of blood flow remains to be determined. In retinal vessels, to the best of our knowledge, only two studies have examined the role of NO in retinal circulation during hypoxia, ^{12 13} but the involvement of NO in the retinal blood flow (RBF) in response to hypoxia is still controversial. 

The first goal of the present study was to test the hypothesis that a flow-induced mechanism is involved in the increase in RBF during hypoxia in vivo. The second goal was to test the hypothesis that NO in the retina plays a role in the regulation of RBF in response to hypoxia. For these purposes, we simultaneously measured the changes in vessel diameter and blood velocity in retinal vessels and calculated the WSR in response to hypoxia, by using bidirectional laser Doppler velocimetry and the intravitreous injection of NO synthase inhibitor. 

Materials and Methods

Animal Preparation

All experiments conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Twenty adult cats of either sex (weight range, 2.0–4.5 kg) were used in the study. With the animal in a closed box, anesthesia was induced with enflurane, oxygen, and nitrous oxide, followed by intraperitoneal injection of atropine (0.04 mg/kg). Tracheostomy was performed, and the animal was mechanically ventilated with 1.4% to 1.8% enflurane and room air. Catheters were placed in both femoral arteries and veins. Pancuronium bromide (0.1 mg/kg per hour; Sankyo Pharmaceutical Co., Tokyo, Japan) was infused continuously. With the animal prone, the head was fixed in a stereotaxic instrument. Arterial pH (pH), arterial partial carbon dioxide tension (PaCO₂), arterial partial oxygen tension (PaO₂), oxygen saturation (SaO₂), and hematocrit (Ht) were measured intermittently with a blood gas analyzer (Chiba Corning Co., Tokyo, Japan). The mean arterial blood pressure (MABP) and heart rate (HR) were monitored continuously. Rectal temperature was maintained at 37°C to 38°C with a heated blanket. The pupils were dilated with 0.5% tropicamide and 0.5% phenylephrine sulfate (Santen Pharmaceutical Co., Osaka, Japan). A 0-D contact lens was placed on the cornea, which was protected with a drop of sodium hyaluronate (Healon; Pharmacia & Upjohn Inc., Peapack, NJ). A butterfly needle (26-gauge) was inserted into the anterior chamber and connected to a pressure transducer to monitor the intraocular pressure (IOP). 

RBF Measurement

We measured RBF with a laser Doppler velocimetry system (Laser Blood Flowmeter model 100; Canon, Inc., Tokyo, Japan), which was based on bidirectional laser Doppler velocimetry, first described by Riva et al. ¹⁴ The system was customized for use in cats. The instrument, which is similar to a fundus camera, is designed to measure vessel diameter and blood velocity simultaneously in a retinal vessel and to calculate the RBF. ^{7 8} Diode (wavelength, 670 nm) and helium-neon lasers (wavelength, 543 nm) were used to measure the blood velocity and the vessel diameter, respectively. The diode laser was focused on the center of a vessel, and the helium-neon laser was positioned vertically to the vessel by direct visualization. 

The principles of the laser Doppler velocimetry system have been described in detail elsewhere. ^{7 8} Briefly, retinal blood velocity was measured using bidirectional laser Doppler velocimetry, which provides absolute measurements of the speed of the red blood cells (RBCs) that are flowing at discrete, selected sites in the retinal vessel, assuming Poiseuille flow. ^{15 16} The Doppler-shifted light that is scattered from the RBCs flowing in the retinal vessel is detected simultaneously in two directions separated by a fixed angle. The signals from the two photomultiplier tube detectors undergo computer-controlled spectrum analysis and sequential measurement of the maximum speed (V _max) of RBCs at the center of the vessel. In a laser Doppler velocimetry system, each pair of spectra was recorded, and the V _max was calculated automatically every 5 ms for 1 second during each measurement. The retinal blood velocity was defined as the average V _max during one cardiac cycle. Laser Doppler measurements were taken from the retinal arteries in front of the tapetum. Measurement sites were generally between the disc margin and the first bifurcation. The Doppler shift power spectra (DSPS), which have sharp and different cutoff frequencies in two scattering directions, were recorded from a retinal artery in front of the tapetum (Fig. 1) , suggesting that a laser Doppler velocimetry system can be used to determine the blood velocity in the retinal arteries in front of the tapetum. 

The vessel diameter of the retinal vessel is determined automatically by computer analysis of the signal produced by the image of the vessel. The vessel images were captured every 4 ms for 60 ms just before and after the measurement of blood velocity. The captured images were analyzed to obtain vessel diameter by using the half height of the transmittance profile to define the vessel edge, using microdensitometry. ¹⁷ The vessel diameter (D) was defined as the average of the diameters determined at each time point. 

RBF was calculated as RBF = S × V _mean, where S is the cross-sectional area of the retinal artery at the laser Doppler measurement site, assuming a circular cross section, and V _mean is the mean blood velocity calculated as V _mean = V _max/2. ¹⁶  

Because the ocular perfusion pressure (OPP) was calculated as \(\frac{2}{3}\) MABP − IOP, the retinal arterial vascular resistance (RVR) can be determined by RVR = OPP/RBF. ¹⁸  

WSR was used as an indicator of wall shear stress and was calculated from the vessel diameter and blood velocity data, assuming a parabolic flow profile. It was calculated as WSR = 8 × V _mean/D under the assumption of a parabolic velocity profile. WSR was represented as the unit S ⁻¹ (1/seconds). ⁴  

To evaluate the reproducibility of the laser Doppler system used in our animal experiments, we calculated the average coefficient of variation [100 (SD/mean)] during five baseline measurements before the intravitreous injection (n = 20). 

Effect of Intravitreous Injection of l-NAME during Rest

The intravitreous microinjection technique was performed by inserting a 30-gauge needle into the vitreous 7 mm posterior to the limbus. The injection was carefully performed with a 100-μL syringe (Hamilton, Reno, NV) with care taken not to damage the lens or retina. The head of a needle was positioned over the optic disc region. N ^G-nitro-l-arginine methyl ester (l-NAME) and N ^G-nitro-d-arginine methyl ester (d-NAME) were purchased from Sigma (St. Louis, MO). Given that the volume of the cat vitreous is approximately 2.5 mL, 50 μL l-NAME (100 mM; n = 7, l-NAME group) was injected into the vitreous to provide an extracellular concentration of 2 × 10⁻³ M in the vicinity of the retinal vessels. This concentration is adequate for the 50% inhibitory concentration (IC₅₀) of l-NAME. ¹⁹ PBS (n = 7, PBS group) was injected in the same manner as l-NAME. 

The RBF measurements started 15 minutes before the injection. The mean of five measurements at 1-minute intervals was defined as the baseline RBF before the injection. After the injection, measurements were performed every 10 minutes for 120 minutes. At each time point, three successive measurements were taken, and the average of the three measurements was used. 

Effect of Hypoxia

Hypoxia was induced 120 minutes after the injection of PBS or l-NAME by lowering the concentration of inspired oxygen from 21% to 10% for 10 minutes. The RBF measurements were performed every minute during and after hypoxia. Blood gas analysis was performed before the injection, before the initiation of hypoxia, at the end of hypoxia, and 10 minutes after the end of hypoxia. 

Confirmation of the Effect of the Intravitreous Injection of l-NAME

To establish that the effect of l-NAME was specific to the inhibition of retinal NO synthase (NOS), an additional group of animals (n = 6, d-name group) was intravitreously injected with the same concentration of the inactive isomer d-NAME (100 mM) and in the same manner as l-NAME. In addition, other cats (n = 2) were given l-arginine (300 mg/kg) intravenously 100 minutes after the intravitreous injection of l-NAME, and 10% hypoxia was induced for 10 minutes at 120 minutes after the injection. 

Statistical Analysis

All data are expressed as mean percentage ± SE. For statistical analysis, we used analysis of variance (ANOVA) for repeated measurements, followed by post hoc comparison with the Dunnett procedure. Group comparisons were performed using the unpaired Student’s t-test. P < 0.05 was considered statistically significant. 

Results

Reproducibility of Laser Doppler Velocimetry

The averaged values of the coefficient of variation during five baseline measurements before the intravitreous injection were as follows: vessel diameter, 2.4% ± 0.8%; blood velocity, 3.3% ± 0.9%; and RBF 5.9% ± 1.5% in retinal arteries (n = 20). 

Effect of Intravitreous Injection of l-NAME during Rest

There were no significant differences in any parameters obtained between groups before the injection (Tables 1 and 2) . The intravitreous injection of PBS or l-NAME did not alter the systemic parameters: pH, PaCO₂, PaO₂, SaO₂, Ht, MABP, or HR (Table 1) . Although the IOP decreased after the injection in each group, there were no differences in the change in IOP in the three groups. The OPP ( \(\frac{2}{3}\) MABP − IOP) did not significantly change in either group. 

After the injection, diameter, velocity, RBF, WSR, and RVR did not differ significantly from the preinjection level in the PBS group (Fig. 2) . At 120 minutes after injection, the intravitreous injection of l-NAME resulted in significant decreases in diameter (−8.1% ± 2.0%, P < 0.01), velocity (−17.0% ± 3.7%, P < 0.01), and RBF (−29.4% ± 4.6%, P < 0.01) and an increase in RVR (48.8% ± 12.7%, P < 0.01) compared with the preinjection levels by repeated-measures ANOVA, followed by the Dunnett procedure. Whereas the decrease in diameter did not occur until 100 minutes after the injection of l-NAME, the significant decreases in velocity and RBF occurred approximately 40 minutes after the injection. The WSR tended to decrease in the cats injected with l-NAME, but the change was not statistically significant. 

Effects of Hypoxia

In the present study, PaO₂ and SaO₂ were reduced to 20 to 30 mm Hg and 40% to 50% among the three groups in response to hypoxia (Table 1) . There were no differences in PaO₂ and SaO₂ among the groups during hypoxia. After the end of hypoxia, PaO₂ returned to baseline in the three groups. In the preliminary investigation in five other untreated cats, PaO₂ immediately decreased and reached the minimum level after approximately 5 minutes of hypoxia (data not shown). There was a slight variation in the change in MABP during hypoxia, but that was not statistically significant in either group. There were no significant changes in IOP or OPP in either group during hypoxia. The HR significantly increased in response to hypoxia in each group (Table 1) . 

After the initiation of hypoxia, the maximum percentage increases above the prehypoxia diameter (13.7% ± 3.5% at 9 minutes, P < 0.01), velocity (39.5% ± 8.6% at 7 minutes, P < 0.01), and RBF (73.9% ± 10.6% at 7 minutes, P < 0.01) were observed in the PBS group by repeated-measures ANOVA, followed by the Dunnett procedure (Fig. 3) . These maximum increases declined slightly but remained significantly elevated until the end of hypoxia. The diameter response occurred more slowly than did the velocity response and typically peaked after the velocity response stabilized in most cases. The WSR immediately increased and reached the maximum level (28.3% ± 7.1% at 4 minutes, P < 0.01) in accordance with the increase in velocity. After this increase, the WSR gradually declined and returned to the prehypoxia level at the end of hypoxia, in parallel with the increase in diameter. The RVR reached the minimum level (−41.2% ± 2.0%, P < 0.01) 8 minutes after the initiation of hypoxia. After the end of hypoxia, transient increases in diameter, velocity, RBF, and WSR were observed in the PBS group (Fig. 3) . All values gradually returned to the prehypoxia level within approximately 10 minutes after the end of hypoxia. 

In the l-NAME group, the changes observed in the PBS group were markedly suppressed in response to hypoxia (Fig. 3) . There were no significant changes in diameter, velocity, WSR, and RVR during hypoxia. Although there was a significant increase in RBF only at 5 minutes of hypoxia in the l-NAME group, the levels in the l-NAME group were significantly reduced compared with those in the PBS group at the same time point. Although velocity and WSR transiently increased after the end of hypoxia, the degree of these changes was smaller than in the PBS group. The transient increases in diameter and RBF observed in the PBS group were not observed after the end of hypoxia in the l-NAME group. 

In the d-name group, the time courses of the changes in retinal and systemic parameters observed after the injection and in response to hypoxia were similar to those in the PBS group. The intravitreous injection of d-NAME (100 mM) did not cause a significant change in RBF (Fig. 4) . After the initiation of hypoxia, the average maximum RBF was 9.0 ± 1.5 μL/min in the d-NAME group, whereas the maximum RBF was 5.4 ± 0.6 μL/min in the l-NAME group. The maximum RBF (9.9 ± 1.2 μL/min) observed in the PBS group was significantly reduced by the intravitreous injection of l-NAME (P < 0.01), but was not reduced by the intravitreous injection of d-NAME. 

Figure 5 shows the averaged time course of RBF obtained from the two cats injected intravenously with l-arginine (300 mg/kg) 100 minutes after the intravitreous injection of l-NAME. l-Arginine reversed the decreased RBF to the preinjection level during hypoxia. Hypoxia (10%) was induced in the cat at 120 minutes after the injection. The RBF then increased in response to hypoxia in the same time course and same magnitude as that observed in the PBS group. 

Discussion

The present study demonstrates that the intravitreous injection of l-NAME markedly reduced RBF during rest (Fig. 2) . This result suggests that NO is continuously produced in the retina and contributes substantially to the maintenance of RBF during rest. The previous in vivo studies yielded partially conflicting results—that is, that the administration of an NOS inhibitor showed either no effect on RBF ^{20 21} or a decrease in RBF ^{22 23 24} during rest. Our results are similar to the latter. The route of administration of NOS inhibitors may have caused this contradiction. In most studies in which the role of NO was examined, the NOS inhibitor was administered intravenously; however, intravenous administration of an NOS inhibitor increases systemic blood pressure ^{21 24 25} and modifies the tissue perfusion pressure and blood flow. Therefore, the systemic effects of intravenous administration of an NOS inhibitor seem to make it difficult to estimate the exact role of NO in controlling blood flow. ^{26 27} To our knowledge, only one study, by Donati et al., ¹³ demonstrated that the local administration of an NOS inhibitor in miniature pigs causes vasoconstriction of retinal arteries. Our results are in good agreement with their results. Our result that MABP did not significantly change after the intravitreous injection of l-NAME indicates that the intravitreous injection technique may minimize the systemic effects of the NOS inhibitor. 

The present result that the increase in RBF during hypoxia is greatly suppressed by the intravitreous injection of l-NAME (Fig. 3) strongly suggests that NO in the retina is involved in the increase in RBF during hypoxia. These results seem to be similar to those obtained previously in anesthetized miniature pigs by Donati et al. ¹³ In contrast, Gidday and Zhu ¹² reported that the retinal arteriolar diameter increased during hypoxia in newborn pigs pretreated with N ^G-monomethyl-l-arginine (l-NMMA, another nonselective NOS inhibitor). They concluded that NO does not contribute to the regulation of RBF during hypoxia in newborn pigs. Their findings do not agree with ours, possibly because of differences between their study and ours in species, methods of measurement, NOS inhibitors administered, and the ages of the animals. Another possible explanation is the time course from the injection of the NOS inhibitor to the induction of hypoxia. l-NMMA was administered 20 minutes before the initiation of hypoxia in their study, but l-NAME was administered 120 minutes before the initiation of hypoxia in our study. Our preliminary data indicate that the decrease in RBF induced by intravitreous injection of l-NAME reached the minimum level at 120 minutes and persisted for at least 3 hours under resting conditions (data not shown). In addition, it was reported that NOS activity in the brain was almost completely inhibited by 120 minutes and that it persisted for 6 hours after administration of l-NAME. ²⁸ Although we did not measure retinal NOS activity in the present study, the time course of the change in RBF induced by intravitreous injection of l-NAME during rest indicates that 120 minutes is sufficient for achieving almost complete NOS inhibition in the retina when l-NAME is administered intravitreously. 

The intravitreous injection of the inactive stereoisomer d-NAME at the same dose (100 mM) as that of l-NAME did not have a significant effect during rest and in response to hypoxia (Fig. 4) . In addition, we observed that the decreased RBF was immediately returned to the preinjection level by the intravenous administration of l-arginine (300 mg/kg) 100 minutes after the injection of l-NAME, and the RBF increased in response to hypoxia in the same manner observed in the PBS group (Fig. 5) . These results confirmed that the effect of l-NAME injected intravitreously is caused by inhibition of endogenous NOS in the retina and not by a nonspecific mechanism. 

The intravitreous injection of the drugs caused a slight increase in the IOP, but the IOP immediately returned to the preinjection level within a few minutes (data not shown). Our results that no change in RBF was induced by the intravitreous injection of PBS (Fig. 2) indicate that the intravitreous injection, per se, had little influence on the retinal circulation. In addition, we confirmed that there was no local effect of the insertion of the needle, because no inflammation was induced, and the RBF did not change for 3 hours after the insertion of the needle in the preliminary study (data not shown). These findings suggest that the intravitreous administration of the drugs and the needle insertion had little influence on the RBF observed in the present study. 

According to Kuo et al., ¹ three different vasoregulatory mechanisms—metabolic, myogenic, and flow-induced—coordinate the overall vascular response of the microvasculature from the downstream arteriole to the upstream arteries in the coronary circulation. In addition, these mechanisms have their predominant effects at different sites within the vascular tree. The metabolic, myogenic, and flow-induced mechanisms exert dominant effects on the downstream smallest-terminal arterioles (<20 μm in diameter), ²⁹ the medium-sized arterioles (20–30 μm), ³⁰ and the upstream large arterioles (∼100–150 μm), ²⁹ respectively. The diameters of the measured retinal arterioles in our study, which ranged from 60 to 100 μm, correspond approximately to that of the upstream large arterioles that have a dominant flow-induced mechanism. In retinal vessels, however, there have been no in vivo studies of whether a flow-induced mechanism is involved during functional hyperemia. A flow-induced mechanism is believed to be caused by the increased wall shear stress. ³ Shear stress, which is defined as the frictional force acting tangentially on the endothelial surface, has also been suggested to have a significant role, not only in the pathogenesis of atherosclerosis, but also in the physiologic adaptation of the vascular wall. ^{31 32 33} By simultaneous measurement of the vessel diameter and blood velocity, WSR was estimated as an indicator of shear stress on the vessel wall, assuming a parabolic velocity profile. ^{4 5 6} The present study provides the first documentation of the evaluation of the WSR of the retinal arterioles. During rest, the calculated WSR in cat retinal arterioles (Table 2) seems to agree with the data previously reported in cremaster muscle (1390 S ⁻¹) ⁴ and pial arterioles (2130 S ⁻¹). ⁶ Our results suggest that a laser Doppler velocimetry system is valuable for measuring not only the RBF but also the WSR in the retinal vascular wall by measuring the vessel diameter and the blood velocity of retinal arterioles. 

In addition, the time course of the WSR in the retinal arterioles in response to hypoxia was examined with a laser Doppler velocimetry system in the present study. Because shear stress is equal to shear rate multiplied by viscosity, and assuming that viscosity remained unchanged, the significant increase in the calculated WSR in the retinal arterioles (Fig. 3) strongly suggests that wall shear stress increases when RBF increases in response to hypoxia. Because the diameter significantly increased and the transiently increased WSR returned to the prehypoxia level in the late phase of hypoxia, it is likely that the delayed dilation of the retinal arterioles may be caused by a flow-induced mechanism to maintain a constant level of shear stress on the vascular wall. Therefore, our findings demonstrate that a flow-induced mechanism plays an important role in the increase in RBF in response to hypoxia. 

Significant increases in velocity, RBF, and WSR preceded the increase in diameter in response to hypoxia in the PBS group (Fig. 3) . In the early phase of hypoxia, the velocity and WSR increased significantly, whereas the diameter and OPP did not. Therefore, the initial increase in velocity may be caused by the preferential dilation of the downstream arterioles, resulting in the increases in RBF and WSR in the early phase of hypoxia. Because the downstream arterioles, especially the terminal arterioles, possess a dominant metabolic mechanism. ³⁴ we speculate that the initial dilation of the downstream arterioles may be the result of the metabolic mechanism. 

The present results showing that both changes in velocity and diameter observed in the PBS group were greatly suppressed in the l-NAME group indicate that these changes are NO-dependent responses. Because increased shear stress causes flow-induced vasodilation, ^{10 35} as a result of the release of NO by the endothelium, ³⁶ it is reasonable to conclude that increased wall shear stress in response to hypoxia causes the release of NO in the retina secondarily through a flow-induced mechanism. 

If NO is released secondarily by the stimuli of the increased shear stress, it must be noted what mechanism causes the initial increase in RBF, resulting in the increases in shear stress. Previous studies identified the vasoactive metabolites, i.e., adenosine, ³⁷ lactate, ³⁸ and a recently reported unknown factor from the surrounding retinal tissue, ³⁹ responsible for the increased RBF during hypoxia. Taken together, it is reasonable that these metabolic factors may trigger the initial increase of RBF through a metabolic mechanism, resulting in the release of NO from the endothelium, which is induced by the increased shear stress in the wall of the retinal arterioles through a flow-induced mechanism. 

There are two constitutively expressed isoforms of NOS: endothelial and neuronal (eNOS and nNOS, respectively). ⁴⁰ It has been reported that NO may be released not only from the endothelium ^{41 42} but also from NOS-containing neurons, dendrites, ⁴³ and astrocytes ⁴⁴ in response to hypoxia. In addition, a recent study of the cerebral circulation ⁴⁵ suggests that NO from nNOS may play an important role in hypoxic hyperemia using a topical administration of selective NOS inhibitor. Based on the difference of NOS, we speculate that NO from nNOS may be one of the factors that triggers the onset of vasodilation at the level of the downstream vessel (terminal arteriole and/or capillary), and NO from eNOS may amplify the increase in RBF through a flow-induced mechanism at the level of the measured retinal vessels (large arterioles). Because l-NAME is a nonselective NOS inhibitor, ⁴⁶ it was impossible to distinguish between the effect of nNOS and eNOS on RBF during hypoxia in the present study. Further studies using selective NOS inhibitors are needed to investigate the precise role of NOS. 

An unexpected transient increase in RBF, which was not observed in the retina previously, was observed in the PBS group after the end of hypoxia (Fig. 3) . At that point, transient increases in velocity, WSR, and MABP were observed in both groups, but increases in vessel diameter and RBF were observed only in the PBS group. These results indicate that the intravitreous injection of l-NAME reduced the flow-induced vasodilation in response to the increase in the WSR, which was probably induced by the transient increase in systemic blood pressure observed in both the l-NAME and PBS groups. Although we could not explain the exact mechanism of the transient increase in MABP after the end of hypoxia, our results indicate that NO may play a role in the transient increase in RBF after the end of hypoxia, probably through a flow-induced mechanism. 

In summary, the present study demonstrated two findings. First, NO was continuously produced in the retina during rest and contributed to the increase in RBF during hypoxia. Second, a flow-induced mechanism was involved with the hypoxic hyperemia in the retina. Based on these findings, we conclude that NO plays an important role in the increase in RBF in response to hypoxia probably through a flow-induced mechanism. 

 Supported by Grant-in-Aid for General Scientific Research C70125417.

 Submitted for publication December 31, 2001; revised April 19, 2002; accepted May 7, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

(A) Pair of DSPS obtained from two directions (left, Path 1; right, Path 2) from a cat retinal artery in the tapetum region. Arrows: the different cutoff frequencies, which were determined automatically by a laser Doppler velocimetry system. (B) Continuous recording of the absolute blood velocity in a cat retinal artery in the tapetum region during two cardiac cycles.

Table 1.

 

Changes in Systemic and Ocular Parameters

	PBS Group (n = 7)							l-NAME Group (n = 7)							d-NAME Group (n = 6)
		Preinjection	120 min	Hypoxia	After	Preinjection	120 min	Hypoxia	After	Preinjection	120 min	Hypoxia	After
pH	7.37 ± 0.01	7.36 ± 0.02	7.36 ± 0.02	7.36 ± 0.02	7.37 ± 0.01	7.36 ± 0.02	7.35 ± 0.02	7.35 ± 0.02	7.36 ± 0.01	7.35 ± 0.01	7.34 ± 0.02	7.34 ± 0.02
PaCO2 (mm Hg)	29.3 ± 1.0	30.3 ± 0.9	30.9 ± 1.0	31.1 ± 0.7	30.4 ± 0.5	31.4 ± 1.2	30.6 ± 1.9	31.5 ± 1.1	30.1 ± 0.4	29.4 ± 0.8	30.3 ± 1.0	29.9 ± 1.2
PaO2 (mm Hg)	110.7 ± 3.0	109.5 ± 2.8	28.6 ± 1.2, †	113.2 ± 2.7	114.3 ± 3.3	115.9 ± 2.5	29.7 ± 0.6, †	115.9 ± 4.7	110.1 ± 3.8	109.5 ± 1.9	27.4 ± 0.9, †	113.3 ± 2.7
SaO2 (%)	99.3 ± 0.3	99.4 ± 0.3	42.6 ± 3.0, †	99.3 ± 0.4	98.3 ± 0.9	98.5 ± 1.0	42.3 ± 1.6, †	99.1 ± 0.8	99.4 ± 0.9	99.2 ± 0.3	39.7 ± 2.7, †	99.2 ± 0.5
Ht (%)	29.3 ± 1.6	30.1 ± 1.8	30.1 ± 1.7	30.1 ± 1.3	30.0 ± 1.6	29.6 ± 1.3	30.0 ± 1.0	29.1 ± 1.5	29.2 ± 1.6	28.8 ± 2.6	29.5 ± 2.2	29.5 ± 2.2
MABP (mm Hg)	97.8 ± 3.4	95.3 ± 4.3	92.0 ± 5.3	104.3 ± 5.7	92.1 ± 4.8	88.8 ± 3.4	95.8 ± 4.2	90.9 ± 5.4	90.5 ± 3.8	87.4 ± 1.7	86.8 ± 2.0	91.1 ± 3.7
IOP (mm Hg)	13.4 ± 1.0	9.9 ± 1.4*	10.4 ± 1.4	10.4 ± 1.3	12.9 ± 1.2	8.9 ± 0.6*	9.9 ± 0.5	9.0 ± 0.5	13.0 ± 1.2	10.3 ± 1.0	12.1 ± 0.7	10.8 ± 1.1
OPP (mm Hg)	51.8 ± 2.1	53.7 ± 3.2	50.9 ± 3.6	59.1 ± 3.6	48.5 ± 3.6	50.4 ± 2.3	54.0 ± 2.8	51.6 ± 3.7	47.3 ± 3.3	47.9 ± 1.5	45.8 ± 2.0	50.1 ± 3.0
HR (beats/min)	144 ± 10	141 ± 11	162 ± 9	141 ± 10	152 ± 12	149 ± 12	167 ± 11	145 ± 11	141 ± 10	137 ± 11	166 ± 8	146 ± 9

 Data are expressed as the mean ± SE. Preinjection, before injection; 120 min, 120 minutes after injection; Hypoxia, 10 minutes of hypoxia; After, 10 minutes after the end of hypoxia.

*   P < 0.05 and

†   P < 0.05 vs. values at preinjection and values at 120 minutes after injection, respectively, by repeated-measures ANOVA followed by the Dunnett procedure.

Table 2.

 

Retinal Circulation at Rest

	PBS Group (n = 7)			l-NAME Group (n = 7)			d-NAME Group (n = 6)
		Preinjection	120 min	Preinjection	120 min	Preinjection	120 min
Diameter (μm)	84.4 ± 3.1	84.9 ± 2.8	82.2 ± 4.1	75.5 ± 4.0*	82.6 ± 3.5	83.9 ± 4.3
Velocity (mm/sec)	33.7 ± 2.5	34.2 ± 2.4	35.4 ± 1.9	29.2 ± 1.6*	31.7 ± 3.5	33.7 ± 4.0
RBF (μL/min)	4.1 ± 0.5	4.2 ± 0.5	4.6 ± 0.6	3.2 ± 0.3*	4.9 ± 0.9	5.0 ± 0.9
WSR (/sec)	1593 ± 102	1607 ± 98	1755 ± 143	1582 ± 132	1518 ± 114	1580 ± 118
RVR (mm Hg · min/μL)	11.1 ± 1.3	10.9 ± 1.1	9.0 ± 1.0	14.0 ± 2.2*	11.3 ± 2.1	11.0 ± 1.7

 Data are expressed as the mean ± SE. Preinjection, before the injection; 120 min, 120 minutes after the injection.

*   P < 0.05 vs. values at preinjection by unpaired Student’s t-test.

Figure 2.

 

Time course of the changes in retinal and systemic circulation after intravitreous injection of PBS (n = 7) and l-NAME (n = 7). Data are expressed as the mean percentage ± SE of the preinjection level. Arrow: time point of the intravitreous injection. *P < 0.05 compared with preinjection values by repeated-measures ANOVA followed by the Dunnett procedure.

Figure 3.

 

Time course of the changes in retinal and systemic circulation in response to hypoxia in PBS (n = 7) and l-NAME (n = 7) groups. Data are expressed as the mean percentage ± SE of the prehypoxia level. Solid bar: period of hypoxia. *P < 0.05 and †P < 0.05 compared with the prehypoxia values in the two groups PBS group and l-NAME group, respectively, by repeated-measures ANOVA, followed by the Dunnett procedure.

Figure 4.

 

Effect of the intravitreous injection of l-NAME and d-NAME on RBF during rest and in response to hypoxia. Symbols and error bars: mean ± SE. *P < 0.05 compared with the baseline by repeated-measures ANOVA followed by the Dunnett procedure. Baseline, before injection; 120 minutes, 120 minutes after injection (before hypoxia); hypoxia, peak response to hypoxia; and After, 10 minutes after the end of hypoxia. The peak response to hypoxia indicates the maximum change in RBF observed during hypoxia.

Figure 5.

 

The time course of RBF in cats given an intravenous injection of l-arginine 100 minutes after the intravitreous injection of l-NAME. All data are presented as a percentage of the preinjection level. Symbols and error bars: mean ± SE. Note the different time courses before and after the initiation of hypoxia.

The authors thank Kaoru Takakusaki, Tatsuya Habaguchi, Kazuya Saito, Junko Sugimoto, and Hiroyuki Kagokawa for their generous assistance.