Immunolocalization of the Na-K-Cl Cotransporter in Bovine Ciliary Epithelium

Abstract

purpose. Recent evidence suggests that Na-K-Cl cotransport plays a major role in blood-to-aqueous anion transport across the ciliary body epithelium. The present study was undertaken to determine the location of the Na-K-Cl cotransporter in fixed sections of bovine eye.

methods. Sections of paraformaldehyde-fixed adult and calf bovine eyes were treated with a monoclonal antibody to mammalian Na-K-Cl cotransporter and a fluorescent secondary antibody and examined under a fluorescent microscope. Na-K-Cl cotransporter protein was detected on immunoblots of dissected tissue and purified nonpigmented ciliary epithelial (NPE) and pigmented ciliary epithelial (PE) cells.

results. Cotransporter immunofluorescence was most intense along the basolateral surfaces of the PE cells. Anterior pars plicata possessed the greatest PE immunofluorescence, and this diminished posteriorly toward the pars plana. Quantitation of immunofluorescence images indicated 7- to 10-fold more cotransporter protein in pars plicata PE than in pars plana PE. Diffuse cytoplasmic fluorescence was seen in the NPE cells, which was also brightest in anterior pars plicata. Immunoblots of separated PE and NPE cells from anterior pars plicata showed that PE contain four times more 170-kDa cotransporter protein than NPE. This confirmed fluorescence quantitation estimates. Cotransporter was barely detectable in cells isolated from pars plana in either cell layer. Immunoblots of the Na,K-ATPase catalytic (alpha) subunit in separated NPE and PE cells showed that NPE cells possessed approximately eight times more alpha subunit protein than PE. Immunofluorescence indicated a similar distribution of alpha subunits and indicated a basolateral membrane location for the subunit on both cell types. Na-K-Cl cotransporter fluorescence patterns showed more variability in adult animals than in calves, which may be related to aging and/or disease. Distinctive patterns of cotransporter fluorescence were also seen in the cornea, iris, and retina.

conclusions. Localization of the Na-K-Cl cotransporter to the plasma membrane on the blood side of the ciliary epithelium tight junctions supports a role for the Na-K-Cl cotransporter in ciliary epithelium as a chloride entry mechanism involved in blood-to-aqueous chloride transport. The concentration of Na,K-ATPase catalytic subunits on NPE basolateral membranes could provide net Na⁺ efflux into the aqueous humor.

Aqueous humor is formed by fluid filtration across the fenestrated endothelium of the ocular capillaries and transepithelial ion transport across the ciliary epithelium. ¹ Chloride is transported from blood to aqueous humor by the ciliary bilayer and likely contributes to fluid formation by the tissue. ^{2 3} Chloride transport by the ciliary epithelium is largely electroneutral, which implies that cation flux accompanies chloride transport, ^{2 3} but a net transepithelial cation flux across ciliary epithelium has not yet been demonstrated. ^{2 4} The transporters responsible for chloride transport by the ciliary bilayer include the Na-K-Cl cotransporter in both bovine and rabbit ciliary epithelium. In rabbit, the Na-K-Cl cotransporter contributes to about half of blood-to-aqueous chloride transport. ³ In cow, Na-K-Cl cotransport may be responsible for >80% of blood-to-aqueous chloride transport. ² Anion exchangers have also been proposed to play a role in chloride entry into the rabbit bilayer, ⁵ whereas chloride efflux into the aqueous humor appears to involve chloride channels on the aqueous side of the ciliary bilayer in both species. ^{2 3 6}  

The ciliary epithelium is composed of a pigmented epithelial layer (PE) overlain on the aqueous side by a nonpigmented epithelial (NPE) layer. The PE and NPE layers are aligned with their apical surfaces facing each other, with the PE basolateral surfaces facing the blood and NPE basolateral surfaces facing the aqueous humor. ⁷ Tight junctions link cells of the NPE layer but not the PE layer. ⁸ NPE and PE cell layers communicate via gap junctions and appear to function as a syncytium because small molecules rapidly diffuse from one layer to the other. ^{9 10} The ciliary epithelium extends from just below the iris, where it is a highly convoluted structure termed the pars plicata, toward the retina, where the ciliary epithelium flattens and becomes the pars plana. ⁷ Whether the pars plicata and pars plana play different roles in aqueous humor formation is not known, although anatomic considerations have suggested that the pars plicata is the primary site of fluid formation by the tissue. ⁷  

Na-K-Cl cotransporters are a family of glycosylated integral membrane proteins of core protein MW 110 to 130 kDa, which use the standing sodium gradient maintained by the Na,K-ATPase to transport one sodium ion, one potassium ion, and two chloride ions in an electroneutral fashion. ¹¹ In fluid transporting tissues such as kidney, intestine, and retinal pigment epithelium (RPE), Na-K-Cl cotransport contributes to fluid transport by providing chloride entry into the epithelial layer. ¹¹  

The syncitial nature of the ciliary epithelium predicts that entry mechanisms for ions involved in blood-to-aqueous ion transport will be located on the blood side of the tight junctions, whereas ion efflux mechanisms will be present on the lumenal side of the bilayer, which is the NPE basolateral membrane. Studies of separated PE and NPE layers have found that Na-K-Cl cotransporter activity ¹⁰ and immunoreactivity ^{3 12} localize predominantly to the PE cells. However, NPE cells also possess cotransporter immunoreactivity, ¹² and cotransporter activity has been described in cultured NPE cells. ^{12 13 14} To further characterize the physical location of the Na-K-Cl cotransporter in ciliary epithelium, we used a monoclonal antibody to mammalian Na-K-Cl cotransporter ¹⁵ to carry out immunofluorescence localization studies of the Na-K-Cl cotransporter in bovine eye. We report here that the Na-K-Cl cotransporter localizes primarily to the PE basolateral membrane in bovine ciliary epithelium. Cotransporter is also detectable in bovine cornea, iris, and retina. Some of these results have been previously reported in abstract form. ¹⁶  

Materials and Methods

Immunofluorescence Microscopy

Tissue was prepared for microscopy as follows. Enucleated eyes from adult cows and 1- to 4-day-old calves were obtained within 2 hours of death from a local abbattoir and kept on ice until use. Eyes were sliced at the cornea and fixed in paraformaldehyde-lysine-periodate (Sigma Chemical Co., St. Louis, MO) overnight at 4°C. Unless otherwise indicated, cornea and lens were removed from fixed eyes, and the eyes were opened with a longitudinal cut at the edge of the eye, then washed with phosphate-buffered saline (PBS), pH 7.4 (Gibco BRL, Grand Island, NY) five times for 10 minutes, and incubated in a 30% (wt/vol) sucrose solution overnight at 4°C. The tissue was then mounted in OCT compound (Allegiance Health Care, Hayward, CA) and snap-frozen in dry ice–cooled 2-methylbutane (Sigma-Aldrich, Milwaukee, WI). Eight-micron-thick cryosections were cut on a Leica 600 microtome and transferred to silane-coated microscope slides (Polysciences, Inc., Warrington, PA). At least 20 sections were cut from each animal. 

Cryosections were rinsed in PBS five times for 10 minutes, followed by a 5-minute incubation with a 1% (wt/vol) SDS/8% (vol/vol) 2-mercaptoethanol in PBS solution. ¹⁵ After three rinses for 5 minutes each in PBS, sections were incubated for 1 hour at room temperature in blocking solution (10% fetal calf serum [Summit Biotechnology, Ft. Collins, CO]/2.5% goat serum [Vector Laboratories, Burlingame, CA] in PBS). Sections were incubated with monoclonal antibodies to either mammalian Na-K-Cl cotransporter (T4c ¹⁵ ) or the alpha subunits of Na,K-ATPase (alpha5 ¹⁷ ; Developmental Hybridoma Bank, Iowa City, IA) at 0.8 μg/ml in blocking solution for 1 hour at room temperature. Sections were then rinsed five times for 10 minutes each in PBS and incubated with either Cy3- or FITC-conjugated goat anti-mouse secondary antibodies at 1:2000 and 1:100 (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), respectively, for 1 hour at room temperature. Coverslips were mounted in 75% glycerol/25% PBS with 0.1% diphenylamine (Sigma Chemical Co.) added to retard bleaching. Immunolabeled sections were viewed and photographed using a Nikon eclipse TE200 fluorescent microscope equipped with Nikon H-III camera. Cryosections in which 0.8 μg/ml of preimmune mouse immunoglobulins (Jackson ImmunoResearch Laboratories) substituted for primary antibody were used as controls. Immunolocalization of antigen was evaluated in at least five sections from each animal. After fluorescence photography, coverslips were removed and sections were stained with Harris hemotoxylin and eosin (Sigma Chemical Co.) for photography. 

For image quantitation, the areas quantified were from photographs representative of six animals examined and were taken at an exposure time such that pixels/area measurements remained within the linear region of a pixels/area versus exposure time curve (not shown). Photographic images were scanned into an IBM desktop computer, and areas of cotransporter fluorescence were converted to pixels/area using Sigmagel software (Jandel Scientific, San Rafael, CA). At least 100 pixels/area measurements were taken from each cell type. Pixils/area measurements were taken at the basolateral surface in the PE cells and throughout the cytoplasm in the NPE cells. 

Separation of PE and NPE Cells

Enucleated eyes were used within 4 hours after death. Ciliary processes were dissected and trypsinized, and a mixed population of PE and NPE was obtained as described by Edelman et al. ¹⁰ PE and NPE separation was then carried out as previously described. ¹² The clear cell layer obtained from Percoll gradients contained 98% ± 3% NPE cells (identified by size ¹² ), whereas the dark layer contained 90% to 95% PE cells (identified by pigmentation and size). Cell viability, as estimated by trypan blue exclusion, was routinely >95%. Representative yields were 1 × 10⁶ NPE per 20 eyes and 3 × 10⁶ PE per 20 eyes. 

Immunoblotting

Eyes were opened along the equator, with the anterior segment placed face down and the vitreous humor removed. The ring of ciliary epithelium was dissected either as pars plicata or as anterior pars plicata, posterior pars plicata, anterior pars plana, and posterior pars plana. Tissue or separated NPE or PE cells were placed in 0.5 ml ice-cold deionized H₂O containing protease inhibitors (Complete Protease Inhibitor Cocktail Tablets; Boehringer–Mannheim, Indianapolis, IN) and homogenized using the Tissue Tearor (Fisher Scientific, Pittsburgh, PA) at a setting of 2 for approximately 15 seconds. The homogenates were centrifuged at 5000g for 15 minutes, and the resulting supernatants were centrifuged in a fixed angle rotor in a Sorvall superspeed centrifuge at 37,000g for 60 minutes at 4°C to pellet cell membranes. Cell membrane pellets were resuspended in 1% SDS with protease inhibitor cocktail and 10 μl saved for protein assay by the method of Peterson. ¹⁸  

Immunoblotting ¹⁹ was carried out as previously described. ¹² After electrotransfer, blots were coated with blocking buffer (5% Carnation nonfat dry milk in PBS/0.5% cold water fish gelatin/0.1% Tween 20) and exposed overnight at 4°C to Na-K-Cl cotransporter antibody (1:10,000) or Na,K-ATPase antibody (1:50) in wash buffer (0.5% fish gelatin/0.1% Tween 20). After the blot was washed five times in wash buffer, it was incubated for 1 hour at room temperature with 1:2500 horseradish peroxidase–labeled goat anti-mouse IgG (Amersham International, Arlington Heights, IL), followed by four washes in wash buffer and a final wash in ddH₂O. Immunolabeled proteins were detected using an ECL kit (Amersham International). 

Images on x-ray films were scanned into an IBM desktop computer, and protein bands of interest were quantified within the linear range of a concentration/signal graph using Sigmagel software. 

Results

Figure 1 shows a section of calf pars plicata treated with a monoclonal antibody against the Na-K-Cl cotransporter. ¹⁵ Intense fluorescence along the basolateral edge of all cells in the PE layer was seen (Fig. 1B) . Weaker immunofluorescence in the NPE layer was diffuse and predominantly localized to the cytoplasm. Little immunofluorescence in either cell layer was detected when mouse IgG was used as primary antibody (Fig. 1C) . To determine whether the dense pigmentation in the PE layer obscured cytoplasmic cotransporter immunofluorescence, as has been reported for other antigens, ²⁰ cryosections of pars plicata containing PE cells lacking heavy pigmentation were examined (Fig. 1D) . Cotransporter immunofluorescence was clearly restricted to the basolateral surface, with minimal fluorescence visible in the cytoplasm. 

Immunofluorescence in pars plicata was compared with that in pars plana (Fig. 2) . A schematic representation of ciliary epithelium (Fig. 2A) indicates four regions that were examined: anterior and posterior pars plicata and anterior and posterior pars plana. Anterior pars plicata (Fig. 2B) showed intense immunofluorescence on the basolateral border of PE cells, with fainter and more diffuse fluorescence within the NPE layer. Cotransporter fluorescence in posterior pars plicata (Fig. 2C) was similar to that in anterior pars plicata, although somewhat less intense in both cell layers. Anterior pars plana (Fig. 2D) possessed little cotransporter fluorescence in either layer. Posterior pars plana contained little cotransporter fluorescence in the PE layer, but near the retina NPE cytoplasmic immunofluorescence increased noticeably, whereas PE immunofluorescence was quite weak (Fig. 2E) . Fluorescence quantitation using Sigmagel software indicated that PE fluorescence in anterior pars plicata was approximately three times as great as that in NPE on a per area basis (Fig. 3A ). In posterior pars plicata (excluding the lumenal edge), PE immunofluorescence was five to six times that in the NPE, although fluorescence in both cell layers was less than that in the corresponding layers of anterior pars plicata. In anterior pars plana, PE immunofluorescence fell to <10% of that in anterior pars plicata PE. In posterior pars plana, immunofluorescence in PE remained low but NPE fluorescence was four to five times greater than that in PE. 

Cotransporter immunofluorescence in pars plicata also varied along a lateral axis, from sclera to lumen (Fig. 4) . NPE cotransporter fluorescence was faint throughout pars plicata but became markedly brighter at the lumenal (inner) edge of the tissue (Fig. 4A) . As elsewhere, lumenal NPE immunofluorescence was primarily cytoplasmic and punctate, with the brightest signal being perinuclear and diminishing in the direction of the basolateral surface (Fig. 4B) . In contrast, PE plasma membrane fluorescence was intense throughout pars plicata but diminished close to the scleral (outer) edge (Fig. 4C) . Fluorescence quantitation indicated that immunofluorescence on the PE basolateral membrane at the scleral edge was ∼10% of that seen in both middle and lumenal locations (Fig. 3B) , whereas NPE fluorescence was approximately eightfold greater along the lumenal edge than in middle pars plicata. NPE fluorescence increased from 15% to 20% of PE immunofluorescence in middle pars plicata to ∼130% of PE fluorescence at the lumenal edge. 

The 170-kDa cotransporter protein in the four regions of ciliary epithelium was also quantitated by immunoblot (Fig. 5) . Cotransporter protein was most concentrated in anterior and posterior pars plicata, with anterior pars plicata containing ∼82% of that in posterior pars plicata. Anterior pars plana contained ∼56%, and posterior pars plana ∼17% of the cotransporter protein in posterior pars plicata. These data provide additional evidence that pars plicata contains significantly more cotransporter protein than pars plana. 

Immunoblots of cotransporter protein from purified PE and NPE cells indicated that in pars plicata PE contained approximately four times the 170-kDa protein found in pars plicata NPE (Fig. 6 , left), confirming previous immunofluorescence data. In pars plana, the cotransporter protein could only be faintly detected in either cell layer (Fig. 6 , right). 

We compared Na-K-Cl cotransporter immunofluorescence patterns in ciliary epithelium from adult and from young animals. Immunofluorescence in pars plicata from six calves varied only slightly among animals, in pattern or intensity (not shown). However, in adult animals cotransporter fluorescence varied considerably both in pattern and in intensity (Fig. 7) . Tissue from 4 of 10 adult animals exhibited fluorescence patterns similar to those found in the calf (Fig. 7A) . In the remaining six, PE immunofluorescence was either cytoplasmic and punctate (Fig. 7B) or basolateral but greatly diminished (Fig. 7C) . In the NPE layer (Figs. 7B 7C) fluorescence was either of equivalent intensity or slightly greater than that in PE. 

Na,K-ATPase in Bovine NPE and PE

Na,K-ATPase is an ion transporter that is required for aqueous humor formation. ²¹ Na,K-ATPase catalytic (alpha) subunits have been identified on the basolateral membranes of both PE and NPE layers by Ghosh et al. ^{22 23} NPE cells were found to contain more alpha 1 and alpha 2 subunits than PE cells in pars plicata. ²³ Alpha 3 subunits were not quantitated. We determined the amounts of total alpha subunits in PE and NPE using a monoclonal antibody that recognized all alpha subunit subtypes. ¹⁷ Figure 8 shows that more alpha subunit protein was detected in NPE than in PE cells on immunoblots. Image quantitation indicated that 7.6 ± 0.6 times more catalytic subunit was present in NPE than in PE (n = 3; P < 0.0001). Immunofluorescence studies supported this assessment (Fig. 9) . Intense alpha subunit immunofluorescence was seen along the NPE basolateral border, whereas PE basolateral immunofluorescence was significantly fainter (Fig. 9B) . NPE Na,K-ATPase basolateral fluorescence appeared quite different from Na-K-Cl cotransporter immunofluorescence in NPE (cf. Fig. 1 ), underscoring the non–plasma membrane pattern of cotransporter distribution in this cell layer. In PE, Na-K-Cl cotransporter was concentrated on the basolateral membrane, similar to the fainter alpha subunit fluorescence in this cell type (Fig. 9A) . 

Na-K-Cl Cotransporter Immunofluorescence in Cornea, Iris, and Retina

Na-K-Cl cotransporter immunofluorescence was detected in bovine cornea, iris, and retina. In cornea, cotransporter immunofluorescence was visible in both the corneal epithelium (Fig. 10A ) and endothelium (Fig. 10D) , but not in the stroma. Immunofluorescence was strongest in the outer layers of corneal epithelium (Fig. 10A , arrows), diminishing toward the internal epithelial layers. Substitution of mouse IgG for the cotransporter antibody resulted in minimal immunofluorescence in both corneal cell layers (Figs. 10B 10E) . 

Na-K-Cl cotransport has been detected electrophysiologically in bovine RPE. ²⁴ Na-K-Cl cotransporter immunofluorescence was clearly visible in the RPE (Fig. 11A ). It did not appear to localize to one side of the polarized RPE cell. Immunofluorescence was also visible in the outer plexiform layer (OPL). A similar retinal field treated with mouse IgG showed no immunofluorescence in these regions when photographed at the same exposure time (Fig. 11C) . Immunofluorescence was visible in the outer nuclear layer and photoreceptors, but this was also present in mouse IgG controls. 

Iris immunofluorescence could be seen in both the apical and basolateral sides of iris pigmented epithelium (Fig. 12A ). Little fluorescence was seen with mouse IgG controls (Fig. 12B) . The heavy pigmentation of this cell layer prevented examination of the cell cytoplasm for cotransporter immunofluorescence. 

Discussion

The role of Na-K-Cl cotransport by the ciliary epithelium has been controversial for over a decade. Early electrophysiological studies found no evidence consistent with a Na-K-Cl cotransporter involved in blood-to-aqueous anion transport, ^{25 26 27 28 29 30} although inhibition of short circuit current by loop diuretics added to the aqueous (but not blood) side of the epithelium was reported. ^{27 28} A recent study of ciliary epithelial cell chloride found no role for the Na-K-Cl cotransporter as an entry mechanism for blood chloride under physiological conditions. ⁵  

In contrast, several recent reports have suggested that the Na-K-Cl cotransporter does contribute significantly to blood-to-aqueous chloride transport and fluid flow across the ciliary bilayer and that the cotransporter is located primarily on PE cells. ^{2 3 10} Thus, isolated bovine PE but not NPE cells displayed a bumetanide-inhibitable regulatory volume increase. ¹⁰ Bumetanide inhibited short circuit current across a rabbit ciliary epithelial bilayer, suggesting a role for the cotransporter in maintaining that current. ³ Bumetanide was more potent when added to the PE side than to the NPE side of the tissue, indicating a PE side location for the Na-K-Cl cotransporter. Bumetanide blocked most of blood-to-aqueous chloride transport across bovine ciliary epithelium ² and ∼50% of blood-to-aqueous chloride transport across rabbit ciliary epithelium. ³ And finally, more cotransporter protein was immunologically detected in PE cells than in NPE cells isolated from either bovine or rabbit. ^{3 12}  

The present immunolocalization studies were undertaken to define the subcellular location of the cotransporter within the ciliary epithelium. The central finding of this study is that the Na-K-Cl cotransporter is concentrated on the basolateral membrane of the PE layer. It is present minimally if at all on the NPE basolateral membrane. This distribution is consistent with a function of the Na-K-Cl cotransporter as a chloride entry mechanism involved in blood-to-aqueous chloride transport. An alternative model ⁵ has postulated involvement of an NPE Na-K-Cl cotransporter in chloride release into the aqueous. Because this would require cotransporter on NPE basolateral membranes, the absence of detectable cotransporter on NPE basolateral membranes (cf. Figs. 1 and 3 ) argues against this model. The model ⁵ further proposes a role for a PE Na-K-Cl cotransporter in chloride release from PE into the blood. Our immunologic data are not inconsistent with this proposal, but recent studies of ³⁶Cl transport across rabbit ciliary epithelium ³ found no role for the cotransporter in aqueous-to-blood chloride transport. A clear role in blood-to-aqueous chloride transport was demonstrated, however. 

Both immunofluorescence and immunoblot analysis revealed an anterior-to-posterior concentration gradient of cotransporter protein within ciliary epithelium. Pars plicata PE contained approximately 10 times more cotransporter protein than pars plana PE, and there was substantially more cotransporter protein in pars plicata overall than in pars plana. Immunofluorescence studies indicated that the cotransporter was most concentrated in anterior pars plicata, whereas immunoblots of dissected tissue indicated that similar amounts of 170-kDa cotransporter protein were present in anterior and posterior pars plicata. The reason for this disparity is not known. Recent immunolocalization studies of Na,K-ATPase subunits in ciliary epithelium also found a greater concentration of alpha subunits in pars plicata than in pars plana. ²³ Given the likely involvement of these two ion transporters in fluid formation, these data support the view ^{31 32} that the pars plicata is the primary site of aqueous humor formation within the ciliary epithelium. 

NPE cytoplasmic immunofluorescence was also most intense in anterior pars plicata and least intense in pars plana. However, in posterior pars plana immediately anterior to the retina, NPE fluorescence sharply increased. We were unable to confirm this finding by immunoblotting because only a faint cotransporter signal in either purified cell type from pars plana could be detected. At present the significance of posterior pars plana NPE immunofluorescence is unclear. 

In addition to varying along an anterior-to-posterior gradient, Na-K-Cl cotransporter immunofluorescence in pars plicata varied along a lateral plane. We were not able to quantitate by immunoblot the amount of cotransporter protein from the lumenal and scleral edges of pars plicata, so we cannot confirm that changes in immunofluorescence reflected changes in membrane-bound 170-kDa protein. Two additional antibodies to the Na-K-Cl cotransporter also detected the changes in PE immunofluorescence observed with the T4 monoclonal, but much less cytoplasmic immunofluorescence in lumenal NPE was observed (data not shown). The significance of NPE cytoplasmic immunofluorescence remains unclear. 

Na,K-ATPase is required for aqueous humor formation in man since the specific Na,K-ATPase inhibitor ouabain reduces aqueous flow. ²¹ Na,K-ATPase could contribute to aqueous formation in two ways: first, by maintaining an inward sodium gradient that provides the electrochemical conditions favorable for transepithelial ion flow via other ion transport mechanisms; and second, by contributing directly to transepithelial sodium transport via active extrusion of sodium ions from the NPE into the aqueous humor. The first function, which the Na,K-ATPase performs in most fluid-transporting epithelia, does not require a specific plasma membrane location for the ATPase. ³³ The second function however requires that there be greater Na,K-ATPase activity on the NPE basolateral membrane than on the blood side of the tight junctions. Na,K-ATPase is composed of an alpha catalytic subunit and a beta subunit, ³⁴ and both subunits have been localized to the basolateral membranes of PE and NPE cells. ^{23 35} These studies have suggested that NPE contain more Na,K-ATPase than PE, and Riley and Hirata ³⁶ found that rabbit NPE cells contained twice the Na,K-ATPase activity as in PE cells. We found that the difference between NPE and PE was surprisingly large: approximately eightfold more alpha subunit in NPE than in PE in pars plicata. These results extend the findings of Ghosh et al. ²³ and Riley and Hirata ³⁶ and support the possibility that NPE Na,K-ATPase might be capable of generating a net blood-to-aqueous sodium flux and thereby contribute to fluid flow across the tissue. Attempts to detect a net sodium flux have so far been unsuccessful. ^{2 4}  

The rate of aqueous humor formation declines in humans with increasing age. ³⁷ The underlying reasons for this decline are unknown. PE basolateral immunofluorescence was reduced or altered in 60% of adult animals, compared with strong PE basolateral immunostaining seen in 100% (6/6) of young animals. This raises the possibility that a declining Na-K-Cl cotransporter concentration or altered cotransporter distribution could contribute to lowered aqueous humor formation. Adult cows are reported to have very low rates of aqueous formation. ³⁸ The adult animals used in this study were dairy cows that had stopped producing milk and often had mastitis, whereas the young animals are considered a much healthier population. Thus, age and/or disease may contribute to the variability of Na-K-Cl cotransporter fluorescence in the adult population. This may be relevant to age-related diseases of the eye, such as glaucoma and cataract. 

Na-K-Cl cotransport activity has been reported in rabbit and bovine corneal epithelium ^{39 40} but not rabbit corneal endothelium. ⁴¹ We found cotransporter immunofluorescence in both bovine corneal tissues. This disparity may be due to species differences. No immunofluorescence was detectable in corneal stroma. No physiological studies have been previously carried out on Na-K-Cl cotransport in bovine iris, where we detected cotransporter staining as well. 

In the RPE, the Na-K-Cl cotransporter serves as a chloride entry mechanism, contributing to absorption of fluid from the subretinal space, ²⁴ so its immunologic detection was expected. The detection of immunofluorescence in the outer plexiform layer was unexpected, however. This layer is composed primarily of synapses between photoreceptor cells and dendrites of horizontal cells and bipolar cells as well as synapses between interplexiform cells and horizontal cells. ⁴² Anion exchanger immunostaining has also been reported in this layer. ⁴³  

In summary, the Na-K-Cl cotransporter can be immunologically detected in several fluid-transporting tissues in the eye, as well as in nontransporting tissues. In ciliary epithelium, the cotransporter is primarily localized to the basolateral surface of the pigmented cell layer. This is consistent with its proposed function as a chloride entry mechanism in aqueous-directed fluid flow underlying aqueous humor formation. ^{2 3 10} Concentration of cotransporter protein within the pars plicata suggests that the pars plicata is the primary site of aqueous humor formation. In light of evidence that the Na-K-Cl cotransporter is hormonally regulated both in intact ciliary epithelium ³ and in cultured ciliary epithelial cells, ^{12 13 14 44 45} further study of this transporter may lead to new strategies for pharmacological modulation of aqueous humor formation. 

 Supported by National Institutes of Health Grant EY10332, American Heart Association Grant 94015270 (CL), That Man May See, Inc., and Research to Prevent Blindness.

 Submitted for publication June 21, 2000; revised September 27, 2000; accepted October 6, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Na-K-Cl cotransporter immunofluorescence in bovine pars plicata. (A) Light micrograph of calf pars plicata stained with hemotoxylin and eosin. (B) Pars plicata stained with a monoclonal antibody to mammalian Na-K-Cl cotransporter. Strong immunofluorescence along the basolateral membrane of the PE (arrowheads), and diffuse, cytoplasmic fluorescence in the NPE is seen. (C) Section treated with mouse IgG as primary antibody and photographed under identical conditions as in (B). (D) Cotransporter immunofluorescence in PE containing little pigmentation. PE immunofluorescence is limited to the plasma membrane (arrowheads). Bar, (A through C) 25 μm; (D) 1.2 μm. St, stroma; Aq, aqueous humor.

Figure 2.

 

Regional differences in Na-K-Cl cotransporter immunofluorescence. (A) Diagram of a longitudinal section of the anterior segment of bovine eye showing the four areas of ciliary epithelium examined: (B) anterior pars plicata; (C) posterior pars plicata; (D) anterior pars plana; (E) posterior pars plana. (B) Anterior pars plicata. Intense fluorescence along the basolateral surface of the PE layer, and diffuse cytoplasmic fluorescence in the NPE layer is seen. (C) Posterior pars plicata. A similar pattern to that in (B), although somewhat decreased immunofluorescence in both cell layers is seen. (D) Anterior pars plana. Little immunofluorescence in the PE layer is seen. (E) Posterior pars plana. Fluorescence is mainly in the NPE, with little fluorescence in the PE. Bar, 25 μm.

Figure 3.

 

Quantification of regional Na-K-Cl cotransporter fluorescence. (A) The micrographs in Figure 2 and two other sets were scanned, and brightness was quantitated as described in Methods. Values are means ± SE. Black, PE; white, NPE. (B) Immunofluorescence at the scleral edge and middle and lumenal edge of pars plicata. The micrograph in Figure 4 and two others were scanned, and brightness was quantitated as in Methods. NPE different from PE in middle, P < 0.0001. PE at the scleral edge different from PE in the middle, P < 0.0001. PE different from NPE at the lumenal edge, P < 0.05. Color code as in (A).

Figure 4.

 

Cotransporter immunofluorescence in pars plicata from sclera to lumen. (A) Immunofluorescence increases in NPE at the lumenal edge (arrows). (B) Immunofluorescence of NPE at the lumenal edge. Fluorescence in NPE appears punctate and perinuclear. (C) PE immunofluorescence at the scleral edge. Note weak basolateral PE signal. Bar, (A) 25 μm; (B) 1μ m; (C) 75 μm.

Figure 5.

 

(A) Immunoblot of cell membranes from four regions of ciliary epithelium. Tissue was dissected, and cell membranes were prepared and immunoblotted as described in Methods. Twenty-five micrograms of protein of each sample was electrophoresed. (A) Anterior pars plicata; (B) posterior pars plicata; (C) anterior pars plana; (D) posterior pars plana. Arrow: 170-kDa cotransporter protein. (B) Quantitation of 170-kDa cotransporter band in each region of ciliary epithelium. n = 3; P < 0.001, (C) different from (A) or (B); P < 0.0001, (D) different from (C).

Figure 6.

 

Immunoblots of Na-K-Cl cotransporter in separated PE and NPE cells. (A) PE and NPE from anterior pars plicata. Twenty-five micrograms of each membrane preparation was electrophoresed. (B) PE and NPE from pars plana. One hundred twenty micrograms of each membrane preparation was electrophoresed. Arrow: 170-kDa protein.

Figure 7.

 

Na-K-Cl cotransporter staining patterns in ciliary epithelium from adult animals. (A) Pattern similar to calf (4 of 10 animals). (B) Pattern found in 3 of 10 animals. Immunofluorescence in PE was punctate and cytoplasmic, with approximately the same intensity as in NPE. (C) Pattern found in 3 of 10 animals. Less PE immunofluorescence relative to NPE was seen. NPE fluorescence in (B) and (C) was more punctate than in (A). St, stroma. Bar, 25 μm.

Figure 8.

 

Immunoblot of Na,K-ATPase alpha subunits in separated NPE and PE cells. NPE and PE cells were purified and immunoblotted as in Methods. Arrows: alpha1 (96 kDa) and alpha2 and alpha3 (105 kDa) proteins. ³⁴

Figure 9.

 

Double-labeled immunofluorescence patterns of Na,K-ATPase alpha subunit and Na-K-Cl cotransporter in anterior pars plicata. An FITC-linked secondary antibody was used to detect the Na-K-Cl cotransporter, and a cy3-linked secondary antibody was used to detect Na,K-ATPase. (A) Na-K-Cl cotransporter immunofluorescence. NKCC, Na-K-Cl cotransporter; (B) Na,K-ATPase alpha subunit fluorescence; (C) a field stained with hematoxylin and eosin. Bar, 25μ m.

Figure 10.

 

Na-K-Cl cotransporter immunofluorescence in bovine cornea. (A through C) Epithelium and stroma; (D through F) endothelium and stroma. (A and D) Na-K-Cl cotransporter immunofluorescence (arrows); (B and E) mouse IgG controls; (C and F) similar fields stained with hematoxylin and eosin. Arrows: areas of greatest Na-K-Cl cotransporter immunofluorescence. Bar, 25μ m.

Figure 11.

 

Na-K-Cl cotransporter immunofluorescence in retina. (A) Na-K-Cl immunofluorescence was detected in both RPE, and the outer plexiform layer. Some fluorescence was also seen at the border of the inner plexiform layer and inner nuclear layer. (B) Similar field stained with hematoxylin and eosin. (C) Mouse IgG control. No immunofluorescence in RPE and little in OPL or IPL is seen. GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; PR, photoreceptors; RPE, retinal pigment epithelium. Bar, 25 μm.

Figure 12.

 

Na-K-Cl cotransport immunofluorescence in iris. (A) Immunofluorescence. White arrow is lumenal border of pigmented epithelial (PE) cells. (B) A similar field treated with mouse IgG. (C) A similar field stained with hematoxylin and eosin. St, stroma. Bar, 25 μm.

The authors thank Jennifer LaVail and Jon Polansky for valuable discussions.