Expression and Localization of FP and EP Prostanoid Receptor Subtypes in Human Ocular Tissues

Abstract

purpose. To determine the expression and precise cellular and subcellular localization of the EP prostanoid receptor subtypes EP₁ through EP₄ and the FP receptor in normal human ocular tissues on the protein and mRNA levels.

methods. Expression of EP and FP receptor proteins was examined by immunohistochemistry on the light microscopic level, using subtype-specific antibodies on frozen and paraffin-embedded tissue sections of 10 normal human donor eyes. The subcellular distribution of the receptor proteins was studied by electron microscopic immunogold labeling. mRNA expression in various ocular tissues was analyzed by reverse transcription-polymerase chain reaction, using subtype-specific primers.

results. The highest expression of the EP₁ receptor protein was found in the epithelia of the cornea, conjunctiva, lens, and the ciliary body; trabecular cells; iris vessels; and retinal ganglion cells. EP₂ receptor labeling was most prominent in the corneal epithelium and choriocapillaries. EP₃ and EP₄ receptor labeling was primarily observed in the corneal endothelium and keratocytes, trabecular cells, ciliary epithelium, and conjunctival and iridal stroma cells, and EP₃ was found, in addition, in retinal Müller cells. The highest expression of FP receptor protein was found in the corneal epithelium, ciliary epithelium, the circular portion of ciliary muscle, and iris stromal and smooth muscle cells. Immunoelectron microscopy showed a subcellular distribution of all prostanoid receptors along plasma membranes and the nuclear envelope. EP and FP receptor mRNA expression largely paralleled the proteins’ expression patterns.

conclusions. The findings demonstrate a wide distribution but differential expression of FP and EP prostanoid receptor subtypes in human ocular tissues. EP₁ through EP₄ receptor subtype expression in human outflow pathways could be significant for future pharmacologic management strategies for the glaucomas.

Prostaglandins (PGs) are local mediators that regulate a great variety of cellular functions by acting on specific G-protein-coupled cell surface receptors. In the eye, they have been implicated in the regulation of blood flow, in the regulation of intraocular pressure, and in mediating inflammatory processes, although they also possess anti-inflammatory and immunomodulatory activities. ^{1 2 3} Topically administered PGs and their analogues have been shown to lower intraocular pressure significantly. ^{2 4 5}  

In a previous study, we have demonstrated PGE₂ to be a major PG in the aqueous humor of patients undergoing cataract surgery. ⁶ The diverse actions of PGE₂ are mediated by specific E-prostanoid (EP) receptors, which can be subdivided into at least four subtypes: EP₁ through EP₄. ⁷ The four EP receptor subtypes differ in structure, binding profiles, and coupling to signal-transduction pathways. 

EP receptors have been shown to be present in ocular cells and tissues of different animal species, mostly in the ciliary muscle or ciliary epithelial cells. ^{8 9 10 11 12 13 14 15} Only a few studies have examined EP receptor localization in human ocular cell lines and isolated ocular cell types and tissues, such as lens epithelial cells ¹⁶ or ciliary epithelial and ciliary muscle cells, ^{17 18 19 20 21 22} by using molecular biological, radioligand-binding, or pharmacologic approaches. However, the specific distribution and precise localization of all four EP receptor subtypes in the human eye are not well characterized and have not yet been completely documented. 

Because knowledge of the specific EP receptor distribution would provide a better understanding of the ocular effects of PGs and their analogues, we investigated the expression and precise cellular and subcellular localization in human ocular tissues of the four known EP receptors. For comparison, the PGF_2α receptor FP was also studied. We used reverse transcription-polymerase chain reaction (RT-PCR) to identify specific EP receptor subtype mRNAs and immunohistochemistry with EP subtype-specific antibodies to localize the protein distribution on the light and electron microscopic level. This study is the first to demonstrate the presence and differential expression of all four EP prostanoid receptor subtypes and the FP receptor in all human ocular tissues and provides evidence for the existence of intracellular PG receptors, suggesting a role in gene transcription. 

Methods

Tissues

Immunohistochemistry was performed on 10 normal human donor eyes (age range, 12–86 years) obtained at autopsy and fixed less than 5 hours after death. The eyes had no history or morphologic evidence of any known ocular disease. 

For RT-PCR, fresh cornea, trabecular meshwork, iris, ciliary body including ciliary muscle, choroid, and retina from three normal-appearing donor eyes (ages 64, 71, and 81 years; two males, one female; 4 to 6 hours after death) and one eye surgically enucleated because of malignant melanoma of the posterior choroid (age 60 years; female) were dissected from surrounding tissues and homogenized. The protocol of the study adhered to the tenets of the Declaration of Helsinki for experiments involving human tissue. 

Antibodies

Affinity-purified rabbit polyclonal antibodies raised against synthetic peptides specific for each EP receptor subtype were used in this study. Generation and characterization of the antibodies were performed as previously described. ²³ In addition, an affinity-purified rabbit polyclonal antibody raised against a synthetic peptide sequence from the human FP receptor (FP2: VYASDKEWIRFDQSNV) was used. Specificity of the antibody was determined by Western blot analysis and by preadsorption experiments. 

Immunohistochemistry

The polyclonal antibodies (diluted 1:20–1:500) were used for immunolocalization of EP₁, EP₂, EP₃, EP₄, and FP receptor proteins on frozen sections, paraffin-embedded sections, and resin-embedded (LR White; Electron Microscopy Sciences, Fort Washington, PA) ultrathin sections of human ocular tissues, by means of immunofluorescence, immunoenzyme labeling, and postembedding immunogold labeling. 

For indirect immunofluorescence, ocular tissues were embedded in optimal cutting temperature (OCT) compound and frozen in isopentane-cooled liquid nitrogen. Cryostat-cut sections (6 μm) were fixed in cold acetone, blocked with 10% normal goat serum, and incubated in primary antibody diluted in phosphate-buffered saline (PBS) overnight at 4°C. Antibody binding was detected by phalloidin-conjugated goat anti-rabbit secondary antibodies (Alexa 488; Molecular Probes, Eugene, OR). 

For immunostaining of 4-μm paraffin-embedded sections, tissues were fixed in 4% paraformaldehyde in PBS at 4°C overnight and embedded in paraffin. Immunolabeling experiments were performed by the peroxidase-labeled streptavidin-biotin method (LSAB Plus-kit; DAKO, Glostrup, Denmark) according to the manufacturer’s instructions. 3-Amino 9-ethyl carbazol (AEC) was used as a chromogenic substrate and Mayer hemalun as a counterstain. 

For postembedding immunogold labeling, tissue specimens were fixed in 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 to 5 hours at 4°C. Specimens were dehydrated serially to 70% ethanol at −20°C and embedded in resin (LR White; Electron Microscopy Sciences). Ultrathin sections were successively incubated in Tris-buffered saline (TBS), 0.05 M glycine in TBS, 0.5% ovalbumin and 0.5% fish gelatin in TBS, primary antibody diluted in TBS-ovalbumin overnight at 4°C, and finally 10 nm gold-conjugated secondary antibody (BioCell, Cardiff, Wales, UK) diluted 1:30 in TBS-ovalbumin for 1 hour. After they were rinsed, the sections were stained with uranyl acetate and examined with an electron microscope. 

In negative control samples, the primary antibody was replaced by PBS or equimolar concentrations of nonimmune rabbit IgG or an irrelevant primary antibody. 

Reverse Transcription-Polymerase Chain Reaction

Extraction of total RNA was performed with a kit (RNeasy Mini-Kit; Qiagen, Hilden, Germany) according to the manufacturer’s instructions. After DNase I digestion, 1 μg total RNA was reverse transcribed in 20-μL reaction volumes, using 500 ng oligo dT primer and 200 U reverse transcriptase (SuperScript II; Life Technologies, Gaithersburg, MD). 

Normalization of cDNAs from the different ocular tissues was performed in 20-μL PCR reaction volumes using primers for the housekeeping gene GAPDH and 2.5 μL of dilutions (1:5–1:50) of the first-strand products. Dilutions resulting in the same band intensity were used for analytic amplifications. 

Specific intron-spanning primers were designed to amplify unique regions in the human EP receptor subtypes and the FP receptor (Table 1) . Amplification of EP and FP cDNAs was performed at the exponential phase in 25-μL reaction volumes, using the normalized template data and a program with a denaturation step of 95°C for 1 minute, and 40 cycles of 95°C for 15 seconds, 55°C (or 58°C and 61°C, respectively) for 30 seconds, and 72°C for 90 seconds. PCR products (10 μL) were analyzed in 1.2% agarose gels containing 250 ng/mL ethidium bromide. Images were obtained with an image-capture system (Cs-1; Cybertech, Berlin, Germany), and differences of band intensities were compared. Identity of the PCR fragments was subsequently confirmed by sequence analysis (Prism 310 gene sequence analyzer; PE-Applied Biosystems, Foster City, CA). 

Results

Expression of FP and EP Receptor Proteins in Human Ocular Tissues

Light Microscopy.

Light microscopic immunolabeling results are shown in Figures 1 2 3 4 5 6 and are summarized in Table 2 . Immunofluorescence and imunoenzyme techniques on cryosections and paraffin-embedded sections of 10 human eyes yielded largely identical results. 

All four EP and the FP receptor proteins were generally detected in various cell types of the cornea and conjunctiva, particularly in the limbal region (Figs. 1 2 3 4 5 6) , sclera and episclera, trabecular meshwork (Figs. 2 6) , iris (Fig. 3) , ciliary body (Figs. 4 6) , lens (Fig. 6) , choroid, and retina (Figs. 5 6) . Positive immunolabeling often appeared not to be distributed uniformly but to be confined to individual cells within connective tissues and smooth muscle tissues, as well as to part of blood vessels within certain tissues. Both endothelial cells and smooth muscle cells of blood vessels in the conjunctiva, sclera, episclera, iris, ciliary body, and choroid were labeled. In addition, intrascleral nerve fiber bundles were generally labeled with all antibodies. A prominent perinuclear labeling pattern was evident for all PG receptor proteins, notably visible in the corneal epithelium (Figs. 1A 1E) and stromal cells (Figs. 1H 3D) by immunofluorescence microscopy, whereas nuclear labeling was observed only with the antibody to EP₂. Antibody binding was abolished completely when nonimmune serum or PBS was used instead of the primary antibody (Fig. 6H) . 

EP₁ Receptor.

The antibody to the EP₁ receptor subtype reacted strongly with all ocular tissues and predominantly labeled the epithelia of the cornea (Figs. 1A 6A) , conjunctiva (Fig. 6B) , lens (Fig. 6C) , and ciliary body (Figs. 4A 6E 6F) , trabecular endothelial cells (Figs. 2A 6D) , vascular endothelial cells of the iris (Figs. 3A 3B) and retinal ganglion cells and photoreceptor cells (Figs. 5A 5B 6G ; Table 2 ). In the trabecular meshwork, the trabecular cells, particularly in the outer portions of the meshwork, and the cells lining the Schlemm canal, collector channels, and scleral aqueous veins were prominently labeled (Figs. 2A 6D) . Positive EP₁ immunoreactivity in the iris was localized to vascular endothelial cells, and, to a lesser extent, to pigment epithelial cells, dilator and sphincter muscle cells, and individual stromal cells, presumably melanocytes (Fig. 3A) . In the ciliary body, nonpigmented and pigmented epithelial layers showed an equally prominent immunoreaction, particularly at the ciliary crests (Figs. 4A 6E) , whereas the ciliary muscle showed moderate labeling intensity (Fig. 6F) . In the neuroretina, all layers were weakly positive, with the most prominent staining in the ganglion cell layer, the inner nuclear layer, and the photoreceptor inner segments, whereas the photoreceptor outer segments and Müller cells were completely negative (Figs. 5A 5B 6G) . Positive immunoreactivity was also detected in the endothelial cells of the retinal vasculature. 

EP₂ Receptor.

EP₂ receptor labeling was only moderately present in ocular tissues. The most prominent staining for the EP₂ receptor protein was observed in the corneal epithelium (Fig. 1B) ; the conjunctival epithelium at the limbus, notably in its basal layers (Fig. 1F) ; and the vascular endothelium of the choriocapillaries (Fig. 5C ; Table 2 ). Whereas trabecular cells were only weakly labeled, the outer wall of the Schlemm canal and scleral cells in the immediate periphery of the outer wall, collector channels, and scleral aqueous veins showed more intense immunoreactions (Fig. 2B) . Specific EP₂ labeling was also observed on the ciliary epithelium, which showed a dotlike labeling pattern along the basal aspects of the pigmented layer in the pars plicata and of both pigmented and nonpigmented layers in the pars plana of the ciliary body (not shown). The ciliary muscle stained only weakly. In the neuroretina, staining appeared weak and diffuse throughout all layers, but was most abundant in the ganglion cell and nerve fiber layers as well as retinal capillaries (Fig. 5C) . In addition to individual stromal cells, immunopositive fiberlike structures resembling nerve fibers or cell processes were generally observed in the sclera, the trabecular meshwork, and the iris stroma. 

EP₃ Receptor.

EP₃ receptor subtype labeling was primarily observed in the corneal endothelium and keratocytes (Fig. 1C) ; in endothelial cells of the trabecular meshwork (Figs. 2C 2D) ; in the ciliary epithelium (Fig. 4B) ; in stromal cells of the conjunctiva, particularly in the limbal area (Figs. 1G 1H) ; in the iris (Figs. 3C 3D) ; and in Müller cells of the retina (Fig. 5D) . The trabecular endothelial cells showed a marked immunopositive reaction, whereas endothelial cells lining the Schlemm canal and collector channels were moderately positive (Figs. 2C 2D) . Scleral cells in the periphery of collector channels and aqueous veins were also labeled. The iris showed a moderate reaction of dilator and sphincter muscles, but a prominent perinuclear labeling pattern of stromal cells (Figs. 3C 3D) . Part of the vessels in the iris stroma, particularly in the iris root, were also positive. In the ciliary body, most of the immunoreaction was located in the basal aspects of the pigmented epithelial cells in the pars plicata (Fig. 4B) and in the nonpigmented epithelial cells in the pars plana of the ciliary body. Individual muscle cells in the ciliary body were distinctly positive. Strong and specific labeling was exclusively present on Müller cells and their processes within the retinal layers, whereas all neuronal structures appeared to be largely negative and retinal vessels only weakly positive (Fig. 5D) . In addition to cellular staining, immunopositive fiberlike structures resembling nerve fibers or cell processes were observed in the sclera; in the conjunctival stroma penetrating the basal layer of the conjunctival epithelium (Fig. 1G) ; in the stroma of iris (Fig. 3C) , ciliary body, and choroid; and in the ciliary muscle (Fig. 4E) . 

EP₄ Receptor.

The strongest expression of the EP₄ receptor protein was observed in the corneal endothelium and keratocytes (Fig. 1D) , in endothelial cells of trabecular meshwork (Fig. 2E) , in the ciliary epithelium (Fig. 4C 4D) , and in cells of the iris stroma (Fig. 3E) and conjunctival stroma in the limbal region. In the trabecular meshwork, prominent immunoreactivity was present on the trabecular cells and endothelial cells lining the Schlemm canal, collector channels, and aqueous veins (Fig. 2E) . In addition to smooth muscle cells and stromal cells of the iris (Fig. 3E) , part of the iridal vessels were markedly positive, particularly in the iris root area. In the ciliary body, the basal aspects of the pigmented epithelium were found to be strongly labeled in the pars plicata (Fig. 4C) , whereas the nonpigmented epithelium was markedly positive in the pars plana region (Fig. 4D) . The ciliary muscle showed only moderate immunopositivity, but individual muscle cells and blood vessels within and anterior to the ciliary muscle were heavily labeled. Blood vessels in the ciliary processes were negative, however. In the retina, prominent labeling was limited to the nerve fiber layer, whereas all other retinal layers were only weakly labeled (Fig. 5E) . Müller cells and retinal vessels were essentially negative. Strongly immunofluorescent globular structures resembling corpora amylacea were found occasionally in the inner plexiform layer. Fiberlike immunopositive structures resembling nerve fibers were again observed in the stroma of iris (Fig. 3E) , ciliary body, and choroid; in the ciliary muscle; and in the sclera surrounding blood vessels. 

FP Receptor.

The highest expression of FP receptor protein was found in the corneal epithelium, revealing a clear perinuclear labeling pattern (Fig. 1E) ; in the ciliary epithelium; in the circular portion of the ciliary muscle (Fig. 4F) ; and in the stromal cells and smooth muscle cells of the iris, particularly the sphincter muscle (Fig. 3F) . The most striking feature of FP receptor-like immunoreactivity was the focal labeling of individual cells (e.g., in the conjunctival stroma at the limbal region, the corneal stroma, the sclera, the iris stroma, the ciliary muscle, and the trabecular meshwork). 

Trabecular meshwork cells, particularly in the outer portions of the meshwork and endothelial cells of the Schlemm canal, collector channels, and aqueous veins showed moderate immunopositivity (Fig. 2F) . Positive immunoreactivity was also localized to smooth muscle cells and individual cells in the iris stroma, presumably melanocytes (Fig. 3F) , particularly concentrated along the anterior border layer. Specific labeling was also detected in both pigmented and nonpigmented layers of the ciliary epithelium, particularly in the pars plicata. In addition, specific immunoreactivity was detected in some stromal cells anterior to the ciliary muscle, in individual ciliary muscle cells (Fig. 4F) , and in vascular endothelial cells lining blood vessels within and anterior to the ciliary muscle, whereas capillaries in the ciliary processes were negative. The choroidal stroma had groups of large, heavily stained cells resembling ganglion cells adjacent to large vessels in the deeper stroma (Fig. 5F) . In the retina, all neuroretinal layers and retinal vessels were weakly positive, whereas Müller cells appeared to be unstained. Brightly fluorescent globular structures resembling corpora amylacea were present in the ganglion cell layer and inner plexiform layer. 

Electron Microscopy.

Immunogold labeling confirmed positive cellular staining for EP and FP receptor proteins in all ocular cell types (e.g., in corneal epithelial cells; Fig. 7A ), trabecular endothelial cells and the endothelia of the Schlemm canal (Fig. 7B) , iris stromal melanocytes (Fig. 7C) , vascular endothelial cells (Fig. 7D) , and ciliary nonpigmented (Figs. 7E 7F) and pigmented epithelial cells (Fig. 7G) . 

The subcellular labeling pattern was rather similar for all EP receptor subtypes and the FP receptor protein: The gold marker clearly revealed a prominent clusterlike localization of receptor proteins along apical (Figs. 7A 7B 7D) or basolateral (Figs. 7E 7F 7G) plasma membranes. In the cytoplasm, the most prominent localization of the gold marker was observed in the nuclear envelope (Figs. 7A 7B 7C 7D) and occasionally on outer mitochondrial membranes, on rough endoplasmic reticulum, and in cytoplasmic vesicles. The nuclei proper were labeled only with antibodies to EP₂, but were largely negative for the other EP receptor subtypes. 

Expression of EP and FP Receptor mRNA in Human Ocular Tissues

Reverse transcription followed by PCR was performed, using total RNA extracted from different ocular tissues and primer sets specific for the individual FP and EP receptor subtypes. The PCR reactions with the RNA isolated from the three donor eyes and the eye surgically enucleated for melanoma yielded largely identical results. Fig. 8 shows ethidium bromide-stained agarose gels with the PCR products obtained with total RNA isolated from two representative human eyes. The identity of the PCR products of predicted size was confirmed by sequence analysis. Negative control experiments without templates did not yield any products (not shown). 

In all four eyes analyzed, EP₁ receptor mRNA expression was strongest in the ciliary body and trabecular meshwork, followed by the iris, choroid, and retina and was weakest in the cornea. The EP₂ mRNA was expressed at much lower levels than the other mRNA species. EP₂ receptor expression was strongest in the trabecular meshwork and iris, followed by the choroid and the ciliary body, and weakest in the retina and the cornea. EP₃ receptor expression was found to be similarly strong in the cornea, ciliary body, retina, and choroid and was slightly weaker in trabecular meshwork and iris. EP₄ receptor expression was equally strong in ciliary body and choroid, followed by iris, trabecular meshwork, and cornea, and was weakest in the retina. The highest FP receptor expression was found in the ciliary body, followed by the choroid, iris, and trabecular meshwork, with the weakest found in the cornea and retina. 

When the tissue-specific mRNA expression of the different receptor subtypes for each eye were compared, the cornea showed uniformly the highest expression of EP₃, followed by EP₄ and FP, and the lowest expression of EP₁ and EP₂. This largely paralleled the protein expression, which was most prominent for EP₃ and EP₄, and weaker for EP₁, EP₂, and FP. mRNA expression in the trabecular meshwork differed somewhat between the eyes: Whereas the highest expression was consistently found for the EP₁ and EP₃ subtypes, expression of EP₂, EP₄, and FP showed interindividual differences with low expression of EP₂ and EP₄ in two of the eyes, and low expression of EP₄ and FP in the other two. The mRNA expression in the iris showed the highest levels for EP₃, EP₄, and FP and lower expression levels for EP₁ and EP₂, again reflecting the general expression pattern of the proteins. The ciliary body showed consistently in all four eyes the highest mRNA expression for the FP receptor, followed by an equal level of expression of EP₁, EP₃, and EP₄, and the lowest expression for EP₂. This expression pattern was well reflected by the general expression pattern of the proteins. mRNA expression in the retina of all four eyes was highest for the EP₃ receptor, followed by EP₁ and FP, and was lowest for EP₂ and EP₄, again matching the expression patterns of the proteins. The choroid showed a consistent expression pattern for mRNA in all eyes analyzed, with an equally high expression of mRNA for EP₃, EP₄, and FP, and a lower expression for EP₂ and EP₁, again matching the expression pattern of the proteins. 

Discussion

In the present study, we analyzed for the first time the expression and regional distribution of the FP and all four EP receptor subtypes in human ocular tissues, using subtype-specific primers and antibodies. This combined approach confirmed the presence and differential expression of both messages and proteins of all EP receptor subtypes and FP receptors in virtually all ocular tissues, suggesting specific binding sites for PGE₂ and PGF_2α in cornea, conjunctiva, sclera, trabecular meshwork, lens, iris, ciliary body, retina, and choroid of the human eye. Whereas RT-PCR identified receptor mRNA expression in the various tissues, immunolabeling with subtype-specific antibodies permitted a more precise localization of tissue distribution and subcellular localization of receptor proteins. The immunohistochemical findings were largely in agreement with the molecular biological findings, and marked differences between mRNA and protein distribution, as reported in a previous study, ²⁴ were not noted. 

Distribution of PG Receptors in Ocular Tissues

Prior studies in animals have identified prostanoid EP and FP receptors in ocular tissues, primarily the iris-ciliary body complex, using radioligand binding and functional assays, ^{8 9 10 11 12 13 14 15 24 25} but comparison of these results with the present findings in human tissues is difficult. For instance, all four EP receptor subtypes were shown to be expressed in bovine ciliary epithelium ¹⁴ : The EP₁ and EP₄ receptors were found primarily in nonpigmented epithelial cells, whereas the EP₂ receptor was localized to the pigmented epithelial cells, and the EP₃ receptor subtype was localized to both the pigmented and the nonpigmented cells. This distribution differs from the human situation, in which all four EP receptor subtypes were obviously expressed in both epithelial layers, predominantly in the pigmented layer in the pars plicata and in the nonpigmented layer in the pars plana of the ciliary body. However, the findings of FP receptor distribution in monkey eyes largely corresponds to the present findings in human ocular tissues: FP receptor protein and message were found to be most prominent in the corneal and conjunctival epithelium and in the ciliary muscle and ciliary processes and to be moderately prominent in the retina, iris, and connective tissues of monkey eyes. ²⁴  

Molecular biological and functional studies on prostanoid receptor distribution in human ocular tissues in situ and in vitro have shown that trabecular meshwork cells express FP receptors, ²⁶ lens epithelium expresses EP₄ and FP receptors, ¹⁶ and ciliary epithelial and ciliary muscle cells express EP₂, EP₄, and FP receptors. ^{17 18 19 20 21 27} These isolated data are largely confirmed by the findings of the present study. However, it has to be taken into account that the literature is somewhat confusing regarding the EP₂ receptor’s nomenclature, because before 1995, when the human EP₂ receptor was cloned, the cloned EP₄ receptor was misclassified as the EP₂ receptor. ²⁸ Ligand binding and quantitative autoradiographic visualization of FP receptors yielded the highest receptor concentration in the longitudinal ciliary muscle, the iris sphincter muscle, and the retina of human eyes, ²⁹ which is partly in agreement with the present findings; however, in this study, high concentrations of FP receptors were mainly detected in the circular portion of the ciliary muscle and also in the corneal and ciliary epithelium. In situ hybridization revealed the presence of high levels of EP₁ receptor mRNA in blood vessels of the iris, ciliary body, and choroid, in the iris sphincter, ciliary muscle, and retina (photoreceptors, inner and outer nuclear layers, ganglion cells). FP receptor mRNA was predominantly present in the circular portion of the ciliary muscle and in the ciliary stroma, iris vasculature, and iris sphincter, but was absent in the retina and choroidal vasculature. ²² Apart from the absence of signals for FP receptor mRNA in retina and choroid, these data are again in good agreement with the findings of receptor distribution in the present study. 

The present study also provides evidence for the existence of intracellular PG receptors EP₁ through EP₄ and FP in the nuclear envelope of various ocular cell types, suggesting an intracellular action of PGs and a direct influence of PGs on gene transcription. These findings are supported by recent data in the literature showing nuclear association of functional EP₁, EP₃, and EP₄ receptors in porcine cerebral tissue (i.e., in vascular endothelial cells and neurons), in fibroblast cell lines, and in human embryonic kidney cells. ^{30 31} Stimulation of these nuclear receptors has been shown to modulate nuclear calcium and gene transcription. Consistent with the recently demonstrated localization of cyclooxygenase-2 in the perinuclear region of ocular cells, ⁶ it is conceivable that locally produced intracellular PGs can activate nuclear EP receptors and modulate gene transcription in various ocular cell types. 

Functional Significance of PG Receptors in Ocular Tissues

Prostanoid FP and EP receptor subtypes are coupled to different signal transduction systems, and stimulation of these receptors triggers different cellular signals: FP and EP₁ stimulate intracellular Ca²⁺ mobilization, EP₃ inhibits adenylate cyclase, and EP₂ and EP₄ activate adenylate cyclase. ³ Thus, FP, EP₁, and EP₃ receptors are generally considered constrictor receptors, whereas EP₂ and EP₄ are considered receptors with relaxant properties. In view of these specific signal-transduction mechanisms, knowledge of the specific PG receptor distribution is needed to facilitate interpretation of the action of substances. 

PGE₂, the major PG in human aqueous humor, ⁶ has a broad range of biological actions, including the contraction or relaxation of vascular and nonvascular smooth muscles, stimulation or suppression of neurotransmitter secretion, regulation of intraocular pressure, and regulation of cell growth. ⁷ PGE₁ has been shown to stimulate growth and melanogenesis of iris melanocytes, by acting on the EP₂ receptor. ³² Cell-type-specific signal-transduction pathways may account for the diverse biological effects induced by PGE. The presence and differential distribution of various EP receptor subtypes in human ocular tissues, as demonstrated in this study, offer a potential explanation for the multiple effects of PGE₂. 

Topically administered PGs reduce intraocular pressure in animals and in normal and glaucomatous human eyes. ^{1 2 4 5} It is generally accepted that PGF_2α analogues such as latanoprost reduce intraocular pressure mainly by increasing the uveoscleral outflow, ^{33 34} perhaps through the mediation of FP receptors on ciliary muscle cells. ^{18 20 22 27 35 36} A high expression of FP receptor mRNA in the ciliary body and a prominent localization of FP receptor proteins in the circular portion of the human ciliary muscle, but only moderate expression and localization in the trabecular meshwork, were confirmed by the present study and are consistent with this proposed mechanism of action. The increase in uveoscleral outflow is thought to occur by activation of intracellular signal-transduction pathways, including cAMP formation, and induction of c-Fos and c-Jun expression. These signals are believed to cause increased synthesis of matrix metalloproteinases and, consistently, increased degradation of extracellular matrix components in the intercellular spaces. ³⁶ However, it is also conceivable that induction of the nuclear transcription factors c-Fos and c-Jun occurs by direct intracellular action of PGs on nuclear prostanoid receptors, influencing gene transcription and induction of metalloproteinase production. 

Topically applied PGE₂ and its analogues also markedly reduce intraocular pressure in humans and animal species, which may be preceded by a transient pressure elevation, ^{2 37 38} but the underlying mechanism of the pressure-lowering effects of PGE₂ is, in contrast to that of PGF_2α, not fully understood. Studies in the perfused human anterior segment have shown that PGE₁ causes a dose-dependent increase in trabecular outflow, ³⁹ suggesting different mechanisms of action for PGF_2α and PGE₁. Intraocular pressure reduction in animals was shown to be aided by selective agonists for most prostanoid EP receptors, implying the involvement of EP₁, EP₂, and EP₃ receptor subtypes, ^{40 41} and the presence of EP₂ receptors in the bovine trabecular meshwork and ciliary muscle has been suggested by a functional study. ⁴² Specific prostanoid receptors present on human trabecular meshwork cells may mediate some of the pressure-lowering effect of PGE. Consistent with this, a high expression of EP₁, EP₂, EP₃, and EP₄ receptor mRNA in the human trabecular meshwork and expression of all EP receptor proteins, most prominently EP₁, EP₃, and EP₄, in trabecular cells and endothelial cells lining the Schlemm canal and collector channels was demonstrated in this study. Whereas the expression of FP receptors on human and bovine trabecular cells has already been reported, ^{26 42} the additional presence of all EP receptor subtypes on human trabecular cells may be important for future pharmacologic management strategies for the glaucomas. Based on the present findings, EP₁, EP₃, and EP₄ receptor agonists, in particular, may substantially affect aqueous outflow and may be promising ocular hypotensive agents. 

However, the ocular hypotensive action of PGE₂ and its analogues may be due not only to an increased trabecular outflow facility, because all EP receptor subtypes are also expressed in the human ciliary muscle (although to a lesser extent than the FP receptor) and may thus also contribute to regulation of uveoscleral aqueous outflow. Stimulation of EP₂ and EP₄ receptors is thought to mediate relaxation of smooth muscle cells, whereas stimulation of EP₁ and EP₃ receptors is believed to cause smooth muscle contraction. ^{3 7} Thus, the coexistence of all four EP receptor subtypes in human ciliary muscle and trabecular cells, which also possess smooth muscle-like contractile properties, ⁴³ suggests that regulation of aqueous outflow by PGE₂ is based on complex antagonistic mechanisms involving multiple receptor subtypes. However, mechanisms of action other than contraction of smooth muscle cells are also possible. 

There is also evidence that that a significant proportion of the ocular hypotensive action of PGF_2α and its analogues is attributed to an additional PGE₂ agonistic activity, mainly mediated by EP₁ and EP₃ receptors, ^{3 41} both of which were shown to be prominently expressed in the human trabecular meshwork and ciliary muscle. 

Although the intraocular-pressure-reducing effect of PGs and their analogues has been well documented, the microvascular effects are less well known. ^{3 44} Wide expression of EP and FP receptors in vascular endothelial and smooth muscle cells in virtually all human ocular tissues examined suggests that EP and FP receptors also mediate vascular reactions of PGF_2α and PGE₂. Although the EP₁ receptor has been detected immunohistochemically in blood vessels of the human conjunctiva, sclera, iris, ciliary body, choroid, and retina in the present study, functional studies in monkey eyes have suggested that neither the EP₁ nor the FP receptor is involved in the regulation of ocular blood flow. ³ However, a selective EP₂ receptor agonist causes a marked drop in vascular resistance of the choroid, indicating that EP₂ receptors mediate vasodilation in the primate eye. This assumption is supported by the prominent immunolabeling for the EP₂ receptor subtype in choriocapillaries and retinal vessels in human eyes, as shown in the current study. Based on the present findings, the receptors thought to be involved in mediating the effects of PGE₂ on the retinal and choroidal microvasculature of human eyes are predominantly of the EP₁ and EP₂ subtype that may elicit vasoconstriction and vasodilation, respectively. This observation is in agreement with findings in pig eyes: Although porcine retinal vessels have been described to contain EP₁, EP₂, EP₃, and FP receptors, the receptors mainly involved in regulating retinal and choroidal blood flow are the EP₁ and EP₂ receptor subtypes. ⁴⁵  

Although both EP₂ and EP₃ receptors have been detected on porcine neurites, ⁴⁶ only the EP₂ subtype was found in plexiform and nerve fiber layers of the human retina, whereas the EP₃ subtype appeared to be confined to Müller cells. However, the additional presence of EP₃ receptors on neuroretinal structures cannot be completely excluded, because a clear differentiation between Müller cell processes and nerve fibers was not feasible on the light microscopic level. The significance of the expression of EP receptors in the cornea, conjunctiva, sclera, and lens, is not yet known. 

In conclusion, all four known EP receptor and FP receptor proteins and mRNAs were found to be widely distributed but differentially expressed in human ocular tissues, suggesting an array of effects of PGE₂ and PGF_2α in the eye. Although the functional significance and precise roles of specific EP receptor subtypes in the human eye remain to be determined, the findings in the present study provide detailed baseline data for the assessment of the pathophysiological roles of these prostanoid receptors and could be an important tool to facilitate the use of EP subtype-specific agonists or antagonists in human ocular diseases. 

 Supported by Grant SFB-539 from Deutsche Forschungsgemeinschaft, Bonn, Germany.

 Submitted for publication August 10, 2001; revised January 10, 2002; accepted January 25, 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.

 

Table 1.

 

Primers Used in RT-PCR

Target	Sequence (5′ to 3′ Direction)	Annealing Temperature (°C)	Expected PCR Product Size (bp)	Accession No.
EP1
Sense	CGCTATGAGCTGCAGTACCC	55	508	NM_000955
Antisense	CCAGGATCTGGTTCCAGGAG
EP2
Sense	CAACCTCATCCGCATGCAC	55	419	NM_000956
Antisense	CTCAAAGGTCAGCCTG
EP3
Sense	CGCCTCAACCACTCCTACACA	61	837	NM_000957
Antisense	GCAGACCGACAGCACGCACAT
EP4
Sense	TGGTATGTGGGCTGGCTG	55	434	NM_000958
Antisense	GAGGACGGTGGCGAGAAT
FP
Sense	CACAAGCTGCCAGACGGAAAAC	55	510	NM_000959
Antisense	TTGTAGAAACACCAGGTCCTCG
GAPDH
Sense	ACCACAGTCCATGCCATCAC	58	490	M33197
Antisense	TCCACCACCCTGTTGCTGTA

Figure 1.

 

Immunofluorescence labeling of FP and EP prostanoid receptor subtypes in normal human cornea (A–E) and conjunctiva (F–H). Note the distinct perinuclear labeling pattern in (A), (E), and (H). EP₁ (A) and EP₂ (B) receptor immunoreaction was mainly localized to the corneal epithelium. Positive immunoreaction for EP₃ (C) and EP₄ (D) was predominantly present in stromal keratocytes. (E) FP receptor localization in the corneal epithelium. (F) Immunofluorescence for the EP₂ receptor was present in the basal layers of the conjunctival epithelium at the limbus. EP₃ immunoreactivity in cells of the conjunctival stroma; note immunopositive fiberlike structures penetrating the basal layers of the conjunctival epithelium in (G) and the distinct perinuclear labeling pattern of stromal cells in (H). Magnification, (A, B, E–H) ×180; (C, D) ×90.

Figure 2.

 

Immunofluorescence labeling of FP and EP prostanoid receptor subtypes in normal human trabecular meshwork tissue. (A) Positive staining for EP₁ was seen on trabecular cells of the outer portion of the meshwork and cells lining the Schlemm canal. (B) EP₂ receptor immunoreaction was mainly localized to the outer wall and periphery of the Schlemm canal. Strong positive immunoreaction for EP₃ was present on trabecular cells of the entire meshwork (C); (D) shows labeling of trabecular endothelial cells in detail. (E) Prominent EP₄ immunoreactivity was also observed on trabecular cells throughout the meshwork. (F) Moderate immunofluorescence for the FP receptor was detected on trabecular endothelial cells in the outer portions of the meshwork and along the Schlemm canal and collector channels. AC, anterior chamber; CC, collector channel; IR, iris tissue apposed to the meshwork; SC, Schlemm canal. Magnification: (A, C, E, F) ×90 (B, D) ×180.

Figure 3.

 

Immunofluorescence labeling of FP and EP prostanoid receptor subtypes in normal human iris tissue. Expression of EP₁ receptor protein in the pigment epithelium, dilator muscle, individual stromal cells, and stromal vessels (arrow) of the iris (A); labeling of vascular endothelial cells is shown in detail in (B). EP₃ receptor immunoreaction was particularly observed in cells and fiberlike structures in the iris stroma and also in the dilator muscle (C); note the distinct perinuclear labeling pattern of stromal cells in (D). (E) Strong positive immunoreaction for the EP₄ receptor subtype was also seen in the dilator muscle and in cells and fiberlike structures in the iris stroma. (F) Labeling for the FP receptor was mainly localized to individual cells in the iris stroma. DI, dilator muscle; PE, pigment epithelium. Magnification: (A–D, F) ×180; (E) ×90.

Figure 4.

 

Immunofluorescent labeling of FP and EP prostanoid receptor subtypes in normal human ciliary body tissue. (A) Positive staining for EP₁ was present in both nonpigmented and pigmented epithelial layers of ciliary processes. (B) Staining for EP₃ was seen along the basal aspects of the pigmented epithelial cells of a ciliary process. EP₄ receptor immunoreactivity was also observed along the pigmented epithelial layer in the pars plicata (C), but mainly on the nonpigmented cells in the pars plana of the ciliary body (D). (E) In the ciliary muscle, EP₃ receptor immunoreaction was mainly present on fiberlike structures between the muscle cells. (F) FP receptor immunoreactivity in the ciliary muscle was localized to individual cells (arrows) in the circular portion of the muscle. PC, posterior chamber; ST, stroma of ciliary body. Magnification, ×180.

Figure 5.

 

Immunofluorescence labeling of FP and EP prostanoid receptor subtypes in normal human retina. Predominant staining for the EP₁ receptor was seen in the ganglion cell layer and inner nuclear layer (A); (B) labeling of ganglion cells and cells in the inner nuclear layer shown in detail. (C) Expression of EP₂ receptor protein was most prominent in the choriocapillaris (CC) and less prominent in the retinal ganglion cell and nerve fiber layer. (D) EP₃ receptor immunoreaction was exclusively found in retinal Müller cells. (E) Positive immunoreaction for the EP₄ receptor was mainly observed in the retinal ganglion cell and nerve fiber layers. (F) In the choroidal stroma, FP receptor labeling was seen in the walls of larger vessels (CV) and on groups of large cells in the perivascular stroma (ST). GC, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; PR photoreceptor inner segments. Magnification, ×180.

Figure 6.

 

Immunolocalization of the EP₁ receptor subtype in paraffin-embedded sections of normal human ocular tissues by the peroxidase-labeled streptavidin-biotin method. Positive labeling of the corneal (A) and conjunctival epithelium (B), notably in the basal layers. (C) Positive immunoreaction in the lens epithelium. (D) In the trabecular meshwork, the EP₁ receptor was expressed in the outer portions and along the Schlemm canal, collector channels, and intrascleral aqueous veins. Positive immunoreaction for EP₁ in both layers of the ciliary epithelium in the pars plicata (E) and pars plana (F) of the ciliary body and in the ciliary muscle (F). (G) In the retina, EP₁ receptor labeling was mainly seen in ganglion cells and, to a lesser extent, in the inner nuclear layer and photoreceptor layer. (H) Negative control of ciliary body tissue, using nonimmune serum instead of the primary antibody, showed no reaction. AC, anterior chamber; CM, ciliary muscle; GC, ganglion cell layer; INL, inner nuclear layer; LC, lens capsule; ONL, outer nuclear layer; PC, posterior chamber; PR, photoreceptor inner segments. Magnification: (A–C, G) ×180; (D–F) ×90; (H) ×45.

Table 2.

 

Immunolocalization of EP and FP Prostanoid Receptors in Normal Human Ocular Tissues (n = 10)

	Cornea					Conjunctiva					Sclera			Trabecular Meshwork					Lens Epithelium
		Epithelium	Keratocytes	Endothelium	Epithelium	Stromal Cells	Vessels	Stromal Cells	Vessels, †	Trabecular Cells	Schlemm Canal Cells	Collector Channels
EP1	++	±	±	++	±	+	±, ‡	+	++	++	++	++
EP2	++	±	±	+*	±*	+*	±	+	+	±	±	+
EP3	±	++	++	±	++*	+	±	+	++	+	+	+
EP4	±	++	++	±	++*	+	±, ‡	+	++	+	+	+
FP	++	±	±	+*	±*	+*	±	+	+	+	+	+

	Iris							Ciliary Body									Choroid					Retina
		Epithelium	Muscles	Stromal Cells	Vessels	Epithelium	Muscle	Stromal Cells	Vessels, §	Capillaries, ∥	Stromal Cells	Vessels	Capillaries	RPE	PR, #	ONL	INL	PL	GC	NFL	MC	Vessels
EP1	+	+	+	++	++	+	+	+	+	±	+	+	−	++	+	++	±	++	+	−	+
EP2	+	+	±	+	+	±	±	+	+	±	+	++	−	±	±	±	±	+	+	−	+
EP3	−	+	++	±	++	+	+	+	±	+	+	−	−	−	−	−	−	−	−	++	±
EP4	−	+	++	+	++	+	+	+	−	+	+	−	−	±	±	±	+	±	++	−	−
FP	±	++	+	+	++	++, ¶	+	+	−	+	+	−	−	+	+	+	±	+	+	−	±

 ++ strong reaction; + moderate reaction; ± weak reaction; − negative reaction; RPE, retinal pigment epithelium; PR, photoreceptors; ONL, outer nuclear layer; INL, inner nuclear layer; PL, plexiform layers; GC, ganglion cell layer; NFL, nerve fiber layer; MC, Müller cells.

*  Mainly in the limbal region.

†  Intra- and episcleral vessels.

‡  In the periphery of blood vessels.

§  Within and anterior of the ciliary muscle, and in the inner connective tissue layer.

∥  In the ciliary processes.

¶  Circular portion of ciliary muscle.

^#  Photoreceptor inner segments.

Figure 7.

 

Immunogold localization of EP receptor proteins in various human ocular cell types visualized by transmission electron microscopy. (A) EP₁ receptor immunoreactivity on the apical and lateral plasma membranes (arrows) and nuclear envelope (arrowheads) of a corneal epithelial cell. Inset: association of the gold marker indicating EP₁ within the nuclear envelope in detail (arrowheads). (B) EP₃ receptor immunolocalization was predominantly present along the apical plasma membrane (arrows) and nuclear envelope (arrowheads) of an endothelial cell lining the Schlemm canal (SC). (C) Immunogold labeling for EP₄ receptor along the nuclear envelope (arrowheads) of an iris stroma melanocyte. (D) EP₄ receptor immunoreactivity on the apical plasma membrane (arrows) and the nuclear envelope (arrowheads) of a vascular endothelial cell in the iris stroma. (E) Immunolocalization of EP₁ receptor along the basolateral membranes of the nonpigmented ciliary epithelium in the pars plicata. (F) Detail of EP₁ receptor immunolocalization on the basolateral membranes (arrow) of the nonpigmented ciliary epithelium. (G) EP₂ receptor immunoreactivity along the basal aspects of the pigmented ciliary epithelium in the pars plicata. Gold labeling was observed in the basal cytoplasm and the basal plasma membranes (arrows). Nu, nucleus; Pl, plaque material; Pe, pericyte; Bm, basement membrane. Scale bars: (A, B) 0.3 μm; (A, inset; F) 0.2 μm; (C–E) 0.5 μm; (G) 0.4 μm.

Figure 8.

 

RT-PCR analysis of prostaglandin FP and EP receptor mRNAs expressed by various human ocular tissues from two representative eyes of the four eyes examined. Total RNA was extracted from the tissues, and RT-PCR was performed using specific primers for the human FP and EP receptor subtypes. Lane 1: cornea RNA; lane 2: trabecular meshwork RNA; lane 3: iris RNA; lane 4: ciliary body RNA; lane 5: retina RNA; lane 6: choroid RNA; lane M: molecular size marker.

The authors thank Carmen Hofmann-Rummelt and Elke Meyer for excellent technical assistance.