Protective Effects of Dietary Docosahexaenoic Acid against Kainate-Induced Retinal Degeneration in Rats

Abstract

purpose. To investigate the role played by docosahexaenoic acid (DHA) in the retina, and more specifically, its ability to protect the retina from kainic acid (KA)-induced retinal damage.

methods. Three-week-old female Wistar rats were used. DHA (1000 mg/kg per day) was fed to the rats for 7, 14, and 28 days, and the concentrations of DHA and arachidonic acid (AA) in the retina and serum were measured. In another group of rats, the right eyes were injected intravitreally with 3.12 nanomoles KA after DHA supplementation for 14 days. Electroretinograms (ERGs) elicited by different stimulus intensities were recorded before and on days 1, 7, and 14 after the KA injection. The amplitudes and implicit times of the a- and b-waves were compared. The number of cells in the ganglion cell layer (GCL) and inner nuclear layer (INL) were compared by histopathologic examination.

results. The concentration of DHA in the serum and retina increased after DHA supplementation. The concentration of AA in serum decreased with DHA supplementation, but the concentration of AA in retina did not show any significant change. The b-waves of the ERGs recorded after KA injection were significantly attenuated in both groups of rats. However, the attenuation was significantly less in the DHA-supplemented rats than in gum arabic–supplemented control rats. The numbers of cells in the INL and GCL were significantly higher in DHA-supplemented rats.

conclusions. These results indicate that DHA supplementation can partially counteract KA neurotoxicity in the rat retina. DHA may play a role in modulating neuronal excitability by reducing KA-induced responses in the retina.

Glutamate is a major excitatory neurotransmitter in the vertebrate central nervous system. In the retina, l-glutamate is highly concentrated in the photoreceptors, the bipolar and ganglion cell layer (GCL). ^{1 2 3} It was shown as early as 1957 that an intraperitoneal injection of glutamate induces retinal lesions in newborn mice. ⁴ Intravitreal injection of kainic acid (KA), a structural analogue of l-glutamate, also induces rapid and selective lesions in the inner retina of rats with sparing of the photoreceptor cells. ^{5 6}  

Glutamate and KA act on postsynaptic cells by binding to receptors. The glutamate receptors have been divided into ionotropic and metabotropic receptor subtypes. ⁷ The ionotropic glutamate receptors have been named according to the preferred agonist: N-methyl-d-aspartate (NMDA),α -amino-3-hydroxyl-5-methyl-isoxazol-4-propionic acid (AMPA), and KA subtypes. 

Docosahexaenoic acid (DHA) is found in high concentrations in mammalian retinas ⁸ and, although there is evidence that it is active in various aspects of retinal physiology, its exact role in retinal physiology remains unclear. DHA is required for optimal retinal function in animals. ^{9 10 11 12 13} It has been demonstrated to be important for the maturation of retinal photoreceptors and for preventing apoptosis of the photoreceptors ¹⁴ in the developing retina. ¹⁵ Dietary studies on omega-3 fatty acids in rats, ^{9 10 11 16} monkeys, ^{12 17} and human infants ^{13 14 18} have demonstrated that DHA deficiency results in delayed retinal development, visual impairment, electroretinographic (ERG) abnormalities and disruption of rod outer segment membrane renewal. In addition, deficiency in DHA is associated with polydipsia and behavioral and cognitive disturbances. ¹⁹ Clinically, the level of DHA in red blood cells ²⁰ and sperm ²¹ is significantly lower in patients with retinitis pigmentosa. 

Many of the studies on excitatory amino acid toxicity have reported that high endogenous levels of glutamate are associated with the degeneration of photoreceptor cells, ^{22 23} ischemic damage, ^{24 25 26} and glaucoma. ²⁷ Excessive release of excitatory amino acids, particularly glutamate, results in a marked increase in the calcium concentration in postsynaptic cells, which has a potentially fatal effect. The toxicity can be due to an indirect effect through the release of nitric oxide and arachidonic acid (AA). ²⁸  

Relevant to this study, dietary supplementation of omega-3 fatty acids has been demonstrated to decrease ischemic and excitotoxic brain damage in rats in vivo. ²⁹ In addition, NMDA-induced responses are potentiated by AA ³⁰ and DHA, ³¹ and non-NMDA responses are reduced by both AA and DHA in rats. This suggests that DHA alters the activation of the non-NMDA (kainate-induced) receptors. 

In the present study, we investigated whether the reduction of excitotoxic brain damage by supplementary omega-3 fatty acids ²⁹ can also be demonstrated in the retina. We used the ERGs to monitor retinal function, and increased the retinal and serum levels of DHA by dietary supplementation of DHA. The results showed that the b-wave depression caused by KA-induced degeneration was significantly less in the DHA-supplemented rats. 

Materials and Methods

Animals

Three-week-old female Wistar rats weighing 30 to 50 g were used. The rats were housed with a 12:12-hour light–dark schedule, and the mean light level in the room was 436 lux. They were allowed free access to food (Rodent Laboratory Diet; Oriental Yeast, Tokyo, Japan) and water. 

All the procedures in this investigation conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the procedures were approved by the Animal Care Committee of Chiba University. 

Measurement of Fatty Acid Concentration in Serum and Retina

Eighteen rats were used in this experiment. The DHA (1000 mg/kg per day) was dissolved in gum arabic solution and was given by gastric intubation for 28 days (between 3 and 7 weeks of age) in four rats, for 14 days (between 5 and 7 weeks of age) in five rats, and for 7 days (between 6 and 7 weeks of age) in four rats. Five other rats served as control animals and were treated with gum arabic solution alone in the same volume for 28 days between 3 and 7 weeks of age. 

All rats were deeply anesthetized with an intraperitoneal injection of pentobarbital and killed before the blood and retinas were collected. Lipids were extracted according to Folch et al. ³² The concentrations of fatty acid were measured according to Saito et al. ³³ by gas–liquid chromatography (model 6890; Hewlett–Packard, Palo Alto, CA). Major fatty acid profiles were measured in control and DHA-supplemented rats for 14 days. The concentrations of DHA and AA in the serum and retina were measured in control and DHA-supplemented rats for 7, 14, and 28 days. Statistical analysis was performed with the two-tailed t-test. 

KA Study

Twenty rats were used for the KA study, 11 of which were supplemented with DHA. The DHA (1000 mg/kg per day) was administered by gastric intubation for 14 successive days between 3 and 5 weeks of age. Intravitreal injection of KA was performed at 5 weeks of age. The same amount of DHA was administered daily after the intravitreal injection of KA for the duration of the experiment between 5 and 7 weeks of age. The remaining nine rats served as control animals and received gum arabic solution alone in the same volume on the same schedule. 

KA (Sigma, St. Louis, MO) was dissolved in sterile normal saline solution, and 3.12 nanomoles KA in 5 μl solution was injected intravitreally into the right eye with a microsyringe (Hamilton, Reno, NV). ³³ After anesthesia, paracentesis was performed with a 27-gauge needle, and a 30-gauge needle was used for the intravitreal injections. These procedures were performed under an operating microscope. 

ERG recordings were performed before, and on days 1, 7, and 14 after the KA injection. The procedures for the ERG recordings have been described in detail earlier. ³⁴ Briefly, a MacLab system (Scope 3.5; AD Instruments, Castle Hill, Australia) was used, and eight responses were averaged with an analysis time of 600 msec. The interstimulus interval was set at 5 seconds, and the interval between each recording was set at 1 minute. 

The anesthetized rat was placed in an electrically shielded cage with its head fixed in place with surgical tape and dark adapted for 30 minutes. The rectal temperature was maintained at 38°C by a heating pad. ERGs were recorded by a fixed stimulus set in which the intensity was altered in 1.0-log step with neutral density (ND) filters. The value of the ND filters varied from −7.0 log units to 0 (full stimulus intensity). 

The statistical significance of the difference between the gum arabic– and the DHA-supplemented rats after the KA injection was determined by repeated measures analysis of variance. 

Histopathologic Examination

Twelve rats were used for histopathologic examinations. After the ERG recordings on day 14 after the KA injection, the animals were perfused through the heart with 10% buffered formalin while under deep anesthesia. The enucleated eyes were kept in the same solution for 30 minutes, and the anterior segments were removed. The eyecups were embedded in paraffin and 6-μm sections were cut for hematoxylin and eosin staining. The retinal histoarchitecture was evaluated by light microscopy. Measurements were made on photographs of the histologic sections obtained from the KA-injected eyes and from left eyes as control samples without KA. Sections perpendicular to the retinal surface were measured at two adjacent locations along the vertical meridian within 1 to 2 mm of the optic disc. The number of the cells was counted for a fixed area (approximately 125 μm length) on each section. Retinas of six KA-injected right eyes and six control eyes were measured. Statistical analysis was performed with the two-tailed t-test. 

Results

Fatty Acid Profiles

The ratios of major fatty acid to total fatty acid in the serum of control and DHA-supplemented rats with for 14 days are shown in Table 1 . The ratio of stearic acid and AA decreased significantly, and the ratio of eicosapentaenoic acid, docosapentaenoic acid, and DHA increased significantly. 

DHA in Serum and Retina

In the serum, the mean concentration of DHA increased with the duration of DHA supplementation (Fig. 1A ). The mean ± SEM of the concentration of DHA in control rats was 37.8 ± 3.3 μg/ml, and it increased to 82.0 ± 19.7 μg/ml with supplementation of DHA for 14 days (P = 0.09). With DHA supplementation for 28 days, the DHA concentration increased to 97.3 ± 10.2 μg/ml, which was significantly higher than the control level (P = 0.005). 

The mean ± SEM of the concentration of DHA in the retina also increased for the duration of DHA supplementation (Fig. 1B) . The mean DHA concentration was 1.06 ± 0.087 μg/mg in the retina of the control rats, and it increased significantly to 1.36 ± 0.062μ g/mg with 14 days’ supplementation of DHA (P = 0.025). 

AA in Serum and Retina

The mean ± SEM of the concentrations of AA in serum and retina are shown in Figure 2 . The mean ± SEM of the concentration of AA was 270.6 ± 25.8μ g/ml in the serum of the control rats, and it decreased to 197.5 ± 16.1 μg/ml (P = 0.047) with 7 days’ supplementation of DHA. The level decreased further with longer periods of supplementation of DHA. The mean AA concentration in the retina also decreased with DHA supplementation, but the difference was not statistically significant, even after 4 weeks of DHA supplementation. 

Effect of KA on ERG with and without DHA Supplementation

The ERGs recorded in a DHA-supplemented and a control rat before and on day 7 after the KA injection are shown in Figure 3 . The number on the left of the ERGs represents the value of the ND filter that was used to reduce the full-intensity stimulus. 

For all intensities, the amplitudes of the b-waves were reduced in both groups on day 7. However, the b-wave amplitudes were larger in the DHA-supplemented rat at all stimulus intensities. The amplitude of a-wave did not show any difference. 

From ERGs such as these, the b-wave amplitudes were measured for all the rats in the two groups. The means ± SEM of the b-wave amplitudes are plotted in Figure 4 . Before KA injection, the amplitude of the b-wave did not show any significant difference between the DHA-supplemented and control rats (P = 0.86; Fig. 4A ). On day 1 after KA injection, the amplitude of b-waves in both groups decreased. However, the mean b-wave amplitudes of the DHA-treated rats were larger than those of the control rats, but the difference was not statistically significant (P = 0.10; Fig. 4B ). On day 7, the mean b-wave amplitudes of the control rats were significantly smaller than those from the DHA-supplemented rats at all stimulus intensities (P = 0.008; Fig. 4C ). On day 14, the mean amplitudes of the b-waves of the DHA-supplemented rats were significantly larger (P = 0.005; Fig. 4D ). 

For the a-waves, there was a slight decrease in the amplitudes after KA, but the changes were not significant at any day after injection (Table 2) . 

The a- and b-wave peak latencies with stimulus intensity of −2 log NDF are shown in Tables 3 and 4 . The peak latency of both waves increased with time, but the increase was not significant at any day after injection. 

Effect of KA on Retinal Histology in Rats with and without DHA Supplementation

The histologic examinations at 14 days after KA injection showed that without injection of KA there were no differences in the histology between the DHA-supplemented (Fig. 5B ) and the gum arabic–supplemented rats (Fig. 5A) . After injection of KA, the thickness of the outer nuclear layer (ONL) was not changed compared with eyes without KA. However, the inner nuclear cell layer (INL) became thinner, and the number of cells in the INL was reduced. The number of nuclei in the GCL was also reduced (Fig. 5C) ; however, the reductions of cells in these two layers were less in the rats with DHA supplementation (Fig. 5D) . 

The mean ± SEM of the number of cells in the GCL without KA was 30.2 ± 3.0 with DHA supplementation and 29.8 ± 2.4 with gum arabic supplementation. After KA injection, the mean numbers of cells in the GCL was significantly reduced to 17.2 ± 1.3 in the DHA-supplemented rats (P = 0.003) and to 8.0 ± 0.4 in the control rats (P < 0.001). The difference between the two groups with KA injection was statistically significant (P < 0.001). 

The mean ± SEM number of cells in the INL was 161.7 ± 6.2 in the DHA-supplemented rats and 155.2 ± 7.4 in the gum arabic–supplemented control rats. After KA injection, the mean number decreased to 132 ± 7.25 in the DHA-supplemented rats (P = 0.011) and 81.33 ± 0.98 in the control rats (P < 0.001). The difference between these two groups was statistically significant (P < 0.001). 

Discussion

In an earlier study, we demonstrated that an intravitreal injection of 3.12 nanomoles KA depressed the amplitudes of the b-wave. ³⁴ These results were in good agreement with the observations of Goto et al. ⁶ and Vaegan et al. ^{35 36} In the present study, we showed that this depression of the b-wave amplitude by KA was significantly less, and the numbers of cells in the GCL and INL were significantly higher in the DHA-supplemented rats. We injected the KA 14 days after the beginning of DHA supplementation when the DHA concentration in the retina was significantly higher than in the control animals. Thus, we found that the retinal alterations induced by KA injection were reduced by DHA supplementation electrophysiologically and histologically. 

The depression of the b-waves of the ERGs can be accounted for by the loss of cells in the INL, as noted in the histologic sections and as reported earlier in rats ⁶ and other species. ^{37 38 39 40 41} This site of cell loss in the INL and GCL is consistent with the anatomic distribution of KA receptors. ^{42 43} The mRNA of the different KA receptor subunits in the rat retina are distributed throughout the INL and GCL, for example, on the horizontal cells, bipolar cells, amacrine cells, and ganglion cells. ⁴²  

Although kainate neurotoxicity has been extensively studied in cat ^{35 36} and chicken ^{37 38 39 40} retinas, the exact mechanism for the cell death has not been established. It has been suggested that KA binds to specific receptors on the membrane of susceptible retinal neurons which leads to a prolonged opening of the membrane channels. This results in a massive entry of sodium ions and the loss of potassium ions. ⁴¹ This alteration of the intracellular ionic milieu and/or cell swelling leads to cell death. In this regard, it is important to note that KA did not lead to a significant depression of the a-waves. 

How do we account for the protective effect of DHA against KA toxicity? The action of DHA and AA on glutamate receptors has been reported by several groups. Thus, Miller et al. ³⁰ reported that AA potentiated the current through NMDA receptor channels and slightly reduced the current through non-NMDA receptor channels in isolated cerebellar granule cells. Nishikawa et al. ³¹ reported that AA potentiated the NMDA-induced responses in a concentration-dependent manner in pyramidal neurons of rat cerebral cortex. In our study, the concentration of AA in the retina was reduced by DHA supplementation, although it was not statistically significant. The results of Miller et al. ³⁰ show, however, that the effect of KA injection must be stronger in DHA-supplemented rats. Nishikawa et al. ³¹ also reported that DHA potentiated the NMDA-induced responses but reduced the non-NMDA (kainate-induced) responses by approximately 30%. In the present study, the concentration of DHA was increased significantly by the oral supplementation, and we suggest that the higher concentrations of DHA attenuated the action of KA on non-NMDA receptors. 

In summary, DHA dietary supplementation reduced KA damage in the retina. These physiologic findings were supported by histologic observations. Taken together, the findings in this study suggest that DHA may play an important role in modulating neuronal excitability by reducing KA-induced responses in the retina. Thus, it may be possible to use DHA clinically to protect or rescue retina from damage induced by excessive release of glutamate as in glaucoma and degenerative or ischemic diseases of the retina. ^{22 23 24 25 26 27}  

 Supported by Grant 10357015 from the Japanese Ministry of Education.

 Submitted for publication May 22, 2000; revised September 5, 2000; accepted September 20, 2000.

 Commercial relationships policy: N.

 

Table 1.

 

Fatty Acid Profiles

	Control (%)	DHA Supplementation (%)*
Palmitic acid (16:0)	12.1 ± 0.53	12.9 ± 1.63
Palmitoleic acid (16:1)	0.62 ± 0.06	0.93 ± 0.28
Stearic acid (18:0)	17.1 ± 1.04	15.1 ± 0.87†
Oleic acid (18:1, n-9)	5.1 ± 0.75	6.5 ± 1.76
Vaccenic acid (18:1, n-7)	1.5 ± 0.12	1.6 ± 0.22
Linoleic acid (18:2)	9.6 ± 0.70	10.1 ± 3.09
Arachidonic acid (20:4, n-6)	33.7 ± 1.92	22.5 ± 3.26, ‡
Eicosapentaenoic acid (20:5, n-3)	0.82 ± 0.20	2.6 ± 0.51, ‡
Docosapentaenoic acid (22:5, n-3)	0.26 ± 0.06	0.48 ± 0.14†
Docosahexaenoic acid (22:6, n-3)	6.9 ± 0.75	11.7 ± 1.65, ‡

 Data are mean ± SD.

 Readings after 14 days’ supplementation.

†   P < 0.05.

‡   P < 0.005.

Figure 1.

 

Concentrations of DHA in serum (A) and in retina (B) in DHA-supplemented and control rats. In DHA-supplemented rats, the concentration of DHA in the serum increased significantly 28 days after the beginning of DHA supplementation (A). The concentration of DHA in the retina was significantly higher 14 days after the beginning of DHA supplementation (B). *P < 0.05, **P < 0.01.

Figure 2.

 

Concentrations of AA in serum (A) and in retina (B). The concentration of AA in serum of DHA-supplemented rats was significantly lower 7 days after the beginning of DHA supplementation (A). The concentration of AA in retina did not show any significant changes during this period (B).* P < 0.05, **P < 0.01.

Figure 3.

 

ERGs recorded before and 1 week after intravitreal injection of 3.12 nanomoles KA in a DHA-supplemented and a control rat. Number (left) represents the value of the ND filters in log units used to reduce the full-intensity stimulus. The maximum illuminance (ND = 0) was 1.5 × 10⁵ lux.

Figure 4.

 

The relationship between stimulus intensity and the mean ± SEM of b-wave amplitude of DHA-supplemented and control rats before intravitreal KA injection (A) and on days 1 (B), 7 (C), and 14 (D) after KA injection.

Table 2.

 

Amplitude of a-Wave with Stimulus Intensity of −2 Log NDF

	Control (μV)	DHA Supplementation (μV)
Before	555.5 ± 81.3	568.1 ± 99.5
Day 1	505.8 ± 100.9	383.3 ± 100.5
Day 7	465.1 ± 106.4	455.7 ± 107.1
Day 14	412.2 ± 89.5	430.0 ± 124.8

 Data are mean ± SEM.

Table 3.

 

Peak Latency of a-Wave with Stimulus Intensity of −2 Log NDF

	Control (msec)	DHA Supplementation (msec)
Before	35.1 ± 1.9	35.1 ± 1.6
Day 1	41.3 ± 2.8	40.5 ± 2.5
Day 7	41.5 ± 2.2	46.3 ± 5.3
Day 14	47.3 ± 5.8	50.8 ± 7.0

 Data are mean ± SEM.

Table 4.

 

Peak Latency of b-Wave with Stimulus Intensity of −2 Log NDF

	Control (msec)	DHA Supplementation (msec)
Before	84.6 ± 4.4	87.1 ± 5.9
Day 1	84.3 ± 3.1	87.5 ± 3.7
Day 7	91.3 ± 3.2	98.6 ± 4.9
Day 14	107.0 ± 14.9	118.8 ± 13.5

 Data are mean ± SEM.

Figure 5.

 

Light micrographs showing the histologic appearance of the rats fed gum arabic solution (A, C) or DHA (B, D) for 28 days (14 days before and 14 days after the KA injection). (A, B) Retinas from the left eyes without KA injection; (C, D) retinas from KA-injected right eyes. Bar, 20 μm.

The authors thank Duco I. Hamasaki for critical review of the manuscript.