Thirty-five adult male albino Sprague-Dawley rats weighing 200 to 250 g were used for this study. Animals were anesthetized with 4% chloral hydrate (1 mL/100 g body weight). The pupils were dilated with 1% tropicamide eye drops. The intraocular pressure (IOP) was raised to 90 to 120 mmHg by cannulation of the anterior chamber with a 30-gauge needle connected to a hydrostatic pressure device, and the elevated IOP was maintained for 60 minutes. The applied pressure of 90 to 120 mmHg is in the range, or slightly above, systemic systolic pressure in the rat. Following removal of the cannula, recirculation began immediately and the IOP decreased to normal values within 5 minutes. Animals were sacrificed by an overdose of chloral hydrate at 12 hours, 24 hours, 3 days, 7 days, and 14 days after reperfusion. The anterior segments of the eyeballs were removed. For western blot analysis, retinal tissues were quickly dissected on an ice-cold plate, frozen on dry ice, and stored at -70℃. For immunocytochemistry and quantitative analysis, the eyecups were fixed by immersion in fixative (4% paraformaldehyde/0.2% picric acid in 0.1 M phosphate buffer [PB, pH 7.4]), for 2 to 3 hours. Following fixation, the retina was carefully dissected and transferred to 30% sucrose in PB, for 24 hours at 4℃. They were then frozen in liquid nitrogen, thawed, and rinsed in 0.01 M phosphate buffered saline (PBS, pH 7.4). The central region near the optic disc was excised, dehydrated in a series of graded alcohol, and embedded in wax.

Western blot analysis was performed on the retinal extracts, which were homogenized in 10 volumes of 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Triton ×-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 0.02% sodium azide, 1 mM PMSF, and 5 µg/mL leupeptin. Protein concentration for each sample was assayed by the Lowry method, in duplicate, and the results were averaged. Duplicate sets of protein standards containing 0, 1, 3, 5, 10, 20, 40, and 60 mg of bovine serum albumin were assayed by the same method. The results were averaged and graphed to produce a linear equation that was used to estimate the protein contents in the retinal extracts. The optical density of each sample was measured at 660 nm, using a spectrometer (Spectronic 20; Bausch & Lomb, Rochester, NY, USA). Tissue samples corresponding to 25 g of total protein were heated at 100℃ for 10 minutes with an equivalent volume of sample buffer (containing 4% SDS and 10% mercaptoethanol) and loaded onto 10% polyacrylamide gels. The proteins were electrotransferred to a nitrocellulose membrane in Tris-glycine-methanol buffer. The membrane was blocked for 1 hour at room temperature in a blocking solution containing 5% nonfat dry milk, 0.05% Tween-20, and PBS. The membrane was then incubated for 15 hour at 4℃ with a goat polyclonal antibody directed against NKCC2 (1 : 1,000; Chemicon, Temencula, CA, USA) in the blocking solution. The membrane was rinsed with 0.05% Tween-20 in PBS for three washes of 10 minutes and incubated for 1 hour at room temperature in a 1 : 200 dilution of biotinylated donkey anti-rabbit IgG (Vector Laboratories, Burlingame, CA, USA). The blot was washed three times for 10 minutes each and then processed for analysis using an enhanced chemiluminescence detection kit (Amersham, Arlington Heights, IL, USA).

For the immunocytochemistry of the NKCC2, 50-µm-thick vibratome sections, prepared as described above, were used. Endogenous peroxidase activity was blocked by placing the sections in 1.4% H₂O₂ in methanol for 10 minutes. The sections were then incubated in 10% normal serum (Vector Laboratories) in PBS for 1 hour at room temperature in order to block nonspecific binding sites. Sections were then incubated in a solution of antibodies to the NKCC2 (1 : 1,000) for 1 day at 4℃. After washing in PBS for 45 minutes (3 × 15 minutes), the retinal tissues were incubated for 12 hours in biotinylated donkey anti-rabbit IgG (1 : 50, Vector Laboratories) with 0.5% Triton X-100 at 4℃, rinsed in PBS, and subsequently incubated in Avidin-biotin-peroxidase complex (Vector Laboratories) in PBS for 1 day at 4℃. Retinas were rinsed in two changes of PBS and three changes of 0.05M Tris-HCl buffer (TB, pH 7.4) for 5 minutes at room temperature, and were then incubated in 0.05% 3,3-diaminobenzidine tetrahydrochloride in TB for 10 minutes. Hydrogen peroxide was added to the incubation medium to a final concentration of 0.01% H₂O₂, and the container was gently shaken as the reaction proceeded. After 1.2 minutes of reaction, as determined by the degree of staining, the reaction was stopped with several washes of TB and PB. The stained tissues were mounted onto gelatin-coated slides and coverslipped with glycerol.

To identify whether cells demonstrating NKCC2 immunoreactivity were ganglion cells, double labeling techniques using antibodies against the NKCC2 and Thy1.1, a specific marker for ganglion cells, were performed. Sections were incubated overnight in monoclonal Thy1.1 antibody (1 : 500; Chemicon, Temencula, CA, USA) with 0.5% Triton X-100 in 0.1M PB at 4℃. Sections were rinsed for 30 minutes with 0.01 M PBS, and the retinal tissues were incubated for 2 hours in biotinylated donkey anti-mouse IgG (1 : 50, Vector Laboratories), rinsed in PBS, and subsequently incubated for 2 hours in biotinylated donkey anti-mouse FITC (1 : 50, Vector Laboratories). After rinsing for 30 minutes with 0.01 M PB and for 30 minutes with 0.01 M PBS and with subsequent incubation for 1 hour in normal donkey serum, the retina was incubated overnight in anti-NKCC2 at 4℃. Sections were then rinsed for 30 minutes with 0.01 M PBS and incubated for 2 hour in biotinylated donkey anti-rabbit IgG (1 : 50, Vector Laboratories), rinsed in PBS, and incubated for 2 hour in biotinylated Cy-3 (1 : 500, Vector Laboratories).

To identify whether cells demonstrating NKCC2 immunoreactivity were horizontal cells, double labeling techniques using antibodies against the NKCC2 and calbindin, a specific marker for horizontal cells, were performed. After incubation overnight in monoclonal calbindin antibody (1 : 3,000, Chemicon) with 0.5% Triton X-100 in 0.1 M PB at 4℃, sections were incubated for 2 hours in biotinylated donkey anti-mouse FITC (1 : 50, Vector Laboratories), rinsed for 30 minutes with 0.01 M PB and 30 minutes with 0.01 M PBS, and then incubated for 2 hours in biotinylated donkey anti-mouse IgG (1 : 50, Vector Laboratories) again. We incubated retinal tissues overnight in biotinylated donkey anti-mouse FITC (1 : 50, Vector Laboratories) and rinsed for 30 minutes with 0.01 M PB and 30 minutes with 0.01 M PBS. In order to inhibit nonspecific reactions, sections were incubated in normal donkey serum for 1 hour. After overnight incubation in anti-NKCC2 at 4℃, sections were rinsed for 30 minutes with 0.01 M PBS. The retinal tissues were incubated for 2 hours in biotinylated donkey anti-rabbit IgG (1 : 50, Vector Laboratories) and then for 2 hours in biotinylated Cy-3 (1 : 500, Vector Laboratories).

Sections were analyzed using a Bio-Rad Radiance Plus confocal scanning microscope (Bio-Rad, Hemel Hempstead, UK), installed on a Nikon Eclipse E600 fluorescence microscope (Nikon, Tokyo, Japan). FITC and Cy3 signals were always detected separately. The FITC labeling was excited using the 488 nm line of the Argon ion laser and detected after passing an HQ513/30 (Bio-Rad) emission filter. For detection of the Cy3 signal, the 543 nm line of the green HeNe laser was used in combination with the 605 / 32 (Bio-Rad) emission filter. Images were imported into Adobe Photoshop ver. 7.0 (Adobe Systems, San Jose, CA, USA) and printed on photo paper (Seiko Epson, Tokyo, Japan). For presentation, all manipulations (brightness and contrast only) were carried out equally for all images.

For whole-mount immunostaining, the same immunocytochemical procedures described above were used, but with a longer incubation time. For semi-quantitative analysis of post-ischemic changes in the NKCC2 immunoreactive ganglion cells, the density (per unit area, 0.25 mm × 0.2 mm) of the labeled ganglion cells was counted in the mid temporal region of the whole-mount retinas. The number of labeled cells was divided by the unit area to obtain mean densities of labeled cells per mm². Data are given as mean ± standard deviation. Statistical significance was assessed using one-way ANOVA.

To evaluate the postlesional changes in the NKCC2 levels quantitatively, we performed an immunoblot analysis on membranes prepared from control and ischemic rat retina. As shown in Fig. 1, antibodies to the NKCC2 demonstrated a single band with a molecular mass of 100 to 150 kDa, the intensity of which increased up to 3 days following ischemia/reperfusion and gradually declined thereafter, in accordance with the immunocytochemical observations. Densitometric analysis revealed that the NKCC2 level was upregulated to approximately 115% of controls by 1 day and to approximately 415% by 3 days. The NKCC2 levels decreased to 275% and 313% at 7 and 14 days post ischemia/reperfusion, respectively.

The thickness of the inner retina gradually decreased with increasing time after ischemia, in agreement with previous reports [14-16]. In normal rat retina immunostained for the NKCC2, labeling was mainly observed in the ganglion cell layer and outer plexiform layer (Fig. 2). In the ganglion cell layer, axonal staining was stronger than cell body staining. In the outer plexiform layer, labeled profiles were horizontally oriented. After ischemia/reperfusion, the NKCC2 immunoreactivity increased compared with controls.

To identify the nature of the labeled profiles in the ganglion cell and outer plexiform layers, double labeling experiments using antisera against the NKCC2 were performed with anti-Thy1.1 for the ganglion cell layer and anti-calbindin for the outer plexiform layer. In the merged picture (Fig. 3C and 3F), the NKCC2 labeled profiles in the ganglion cell layer also demonstrated Thy1.1 immunoreactivity. Two weeks after injury, the ischemic retina (Fig. 3D-3F) demonstrated a decreased ganglion cell population compared with the normal retina (Fig. 3A-3C). In the other merged picture (Fig. 4C and 4F), the NKCC2 labeled profiles in the inner retina also demonstrated calbindin immunoreactivity. Two weeks after injury, the ischemic retina (Fig. 4D-4F) demonstrated increased horizontal cell process immunoreactivity compared with the normal retina (Fig. 4A-4C).

In the present study, all labeled profiles that lay in the ganglion cell and outer plexiform layers of the control retina and retinae of experimental groups were immunoreactive to antibodies against Thy1.1 and calbindin, respectively. These findings clearly indicate that the NKCC2 is expressed in retinal ganglion and horizontal cells.

The density of the NKCC2-labeled ganglion cells was calculated in unit areas (0.25 mm × 0.2 mm) in the mid-temporal region of the retina (Figs. 5 and 6). In the normal retina, the mean density of labeled ganglion cells was 1255 ± 109 cells per unit area. These values correspond to the previous report of Lafuente et al. [14], Lafuente et al. [15], and Lafuente et al. [16]. The density of labeled cells was altered at 3 days and 2 weeks after reperfusion. The density of the labeled cells was 391 ± 49 (31.2% of normal density) and 185 ± 37 cells/unit area (14.7%) by 3 days and 2 weeks, respectively. These results are in agreement with a prior report by Lafuente et al. [14], Lafuente et al. [15], and Lafuente et al. [16], which demonstrated that retinal ganglion cell loss after retinal ischemia is an ongoing process and that most ganglion cells die after ischemia/reperfusion.