PGF2α is one of the major prostanoids produced by the kidney. The cellular effects of PGF2α are mediated by a G protein-coupled transmembrane receptor designated the FP receptor. Both in situ hybridization and β-galactosidase knocked into the endogenous FP locus were used to determine the cellular distribution of the mouse FP receptor. Specific labeling was detected in the kidney, ovary, and uterus. Abundant FP expression in ovarian follicles and uterus is consistent with previous reports of failed parturition in FP−/− mice. In the kidney, coexpression of the mFP mRNA with the thiazide-sensitive cotransporter defined its expression in the distal convoluted tubule (DCT). FP receptor was also present in aquaporin-2-positive cortical collecting ducts (CCD). No FP mRNA was detected in glomeruli, proximal tubules, or thick ascending limbs. Intrarenal expression of the FP receptor in the DCT and CCD suggests an important role for the FP receptor regulating water and solute transport in these segments of the nephron.
prostanoids,including PGE2, PGD2, prostacyclin (PGI2), TxA2, and PGF2α, modulate a diverse spectrum of physiological processes, including reproduction, inflammation, microvascular resistance, and epithelial ion transport rates (7, 33). Despite originating from a common precursor, PGH2, the effects of these derivative prostanoids may either oppose each other, as in the case of the prothrombotic action of TxA2 vs. the antithrombotic effects of PGI2 (16), or exert functionally complementary effects such as the smooth muscle constrictor effects of TxA2 and PGF2α (48). These cellular and physiological effects are mediated by the selective interaction of each prostanoid with unique G protein-coupled receptors (GPCRs) (8, 33, 46). Genetic disruption of GPCR prostanoid receptors has not only firmly established roles for these receptors as critical mediators of prostanoid action, but it also revealed significant new biology related to the roles of prostaglandins (33, 46). In the case of the FP receptor for PGF2α, these studies revealed that the FP receptor is highly expressed in the ovary and its function is essential for normal parturition (47).
The FP receptor is also highly expressed in the kidney (2,44). Furthermore, PGF2α is a major product of cyclooxygenase-mediated arachidonate metabolism in the kidney (14), and renal synthesis of PGF2α is regulated by sodium depletion, potassium depletion, and adrenal steroids (35, 40). Infusion of exogenous PGF2α modulates renal salt excretion and urine flow (42). Despite this evidence supporting a role for the FP receptor in the kidney, the intrarenal sites of expression or mechanism of these PGF2α-activated GPCRs in the kidney remain poorly characterized. The purpose of the present studies was to map the intrarenal distribution of the FP receptor in the kidney.
Generation of RNA fragments.
RNA probes were generated by RT-PCR to amplify a 399-bp fragment spanning the 5′-UTR to Arg100 in the coding region of the mouse FP receptor cDNA from kidney RNA. The sense primer was 5′-AACCACTCAGTGGCTCAGGA-3′, and the antisense primer was 5′-GCGGATCCAGTCTTTATC3′. The identity of the amplified product was directly confirmed by sequencing and alignment with the mouse FP receptor (BLAST, NCBI) and ligated into the transcription vector pCR2.1 (Invitrogen). The two distinct clones were isolated, which allowed transcription of either the sense or antisense cDNAs using the T7 promoter. The plasmids were linearized and RNAs transcribed from the flanking T7 promoter in the presence of [α-35S]UTP. RNA (5 × 105 cpm/μl) was used for in situ hybridization.
C57BL/6J mice weighing between 20 and 30 g were anesthetized using intraperitoneal ketamine and xylazine (200 mg and 15 mg/kg, respectively). After surgical anesthesia was achieved, mice were killed by cervical dislocation and kidney, stomach, liver, ovary, and uterus were harvested.
For in situ hybridization studies, tissues were fixed in 4% paraformaldehyde. Tissues were imbedded in paraffin and 7-μm sections were cut. Before hybridization, sections were deparaffinized, refixed in paraformaldehyde, treated with proteinase K (20 μg/ml), washed with PBS, refixed in 4% paraformaldehyde, and treated with triethanolamine plus acetic anhydride (0.25% vol/vol). Finally, sections were dehydrated in 100% ethanol.
Anti-sense RNA was hybridized to the sections at 50–55°C for ∼18 h as described previously (11). After hybridization, sections were washed at 50°C in 5× SSC + 10 mM β-mercaptoethanol for 30 min. This was followed by a wash in 50% formamide, 2× SSC, and 100 mM β-mercaptoethanol for 60 min. After additional washes in 10 mM TRIS, 5 mM EDTA, 500 mM NaCl (TEN), sections were treated with RNase (10 μg/ml), at 37°C for 30 min, followed by another wash in TEN (37°C). Sections were then washed twice in 2× SSC and then twice in 0.1× SSC (50°C). Slides were dehydrated with graded ethanols containing 300 mM ammonium acetate.
For detection of the hybridized probe, slides were dipped in photo emulsion (Ilford K5, Knutsford, UK) diluted 1:1 with 2% glycerol/water and exposed for 7 days at 4°C. After development in Kodak D19, slides were counterstained with hematoxylin and eosin. Photomicrographs were taken using a Zeiss Axioskop using both bright- and darkfield optics.
Preparation of tissue for β-galactosidase staining and immunohistochemistry.
Multiple organs including liver, spleen, stomach, duodenum, lung, and kidneys of the double transgenic mice were harvested at death. After fixation with 4% paraformaldehyde plus 0.25% glutaraldehyde in PBS for 2 h at 4°C, tissue sections were cut with a vibratome into 200-μm slices. To detect β-galactosidase (β-gal) activity, these slices were bathed in permeabilization solution (2 mM MgCl2, 0.01% sodium deoxycholate, 0.02% NP-40 in PBS) for 30 min × 2 and then stained with 1 mg/ml 5-bromo-4-chloro-3-indolyl-d-galactopyranoside (X-gal; Sigma, St. Louis, MO) in staining solution (2 mM MgCl2, 5 mM potassium ferricyanide, potassium ferrocyanide, 20 mM Tris, pH 7.4 in PBS) at room temperature in the dark for 48 h (5,36). Tissues were washed, dehydrated through graded ethanol series, and embedded in paraffin, using standard procedures. Serial 5-μm sections were cut and examined by light microscopy.
To define the nephron segments that expressed FP receptor mRNA, in situ hybridization was followed by immunostaining of the tissue sections with a rabbit anti-collecting duct antibody or a goat anti-human Tamm-Horsfall antibody, which specifically recognizes medullary and cortical thick ascending limb (mTAL and cTAL) as well as the early portion of the distal tubule. To define the β-gal-positive nephron segments, sections were co-stained using a goat anti-human Tamm-Horsfall antibody (1:2,500, Organon-Technika) that specifically recognizes mTAL and cTAL as well as the early portion of the distal tubule (28, 49). A commercially available anti-aquaproin-2 (AQP2) antibody was used to specifically identify collecting duct principal cells (AQP21-A, Anti-Rat AQP2 IgG no. 2, Alpha Diagnostic International, San Antonio, TX) (26). To define distal convoluted tubule segments (DCT), an anti-thiazide-sensitive NaCl cotransporter (TSC) antibody was used [generously provided by Dr. M. Knepper (30)]. Staining was localized using a biotinylated anti-IgG secondary antibody applied to β-gal-stained sections. Biotin was identified using streptavidin coupled to horseradish peroxidase and was visualized with diaminiobenzidine (Vector Vectastain ABC kit). Sections were viewed and imaged with a Zeiss Axioskop and Spot-Cam digital camera (diagnostic instruments).
Intrarenal distribution of the FP receptor.
Autoradiograms of the kidney, with an anti-sense FP receptor riboprobe (Fig. 1), showed intense labeling of subpopulations of epithelial tubules in the renal cortex. No specific labeling was obtained with a sense mRNA probe (data not shown). There was light and diffuse labeling of the outer medulla. There was no detectable labeling of the papilla. A similar pattern of FP mRNA expression was obtained by mapping β-gal activity in heterozygous FP +/− mice (Fig. 1 B).
Segmental expression of the mFP receptor was most abundant in tubules that colabeled with antibodies to the TSC- and AQP2 antibody (collecting duct specific)-positive tubules. There was no evidence for hybridization of the FP antisense fragment to either proximal straight tubules or thick ascending limb (Fig.2). No labeling of papillary or inner medullary structures was observed.
Abundant β-gal expression was detected in stromal surrounding the ureteral smooth muscle (Fig. 3). Epididymus possesses endogenous β-gal activity in control animals, complicating the interpretation of this tissues. However, this endogenous activity was not present in any other organs from wild-type animals examined including the distal vas deferens and luminal cells of the testis where low levels of β-gal activity were detected. In the female genital tract, β-gal expression was detected in ovary corpora luteal cells and the smooth muscle cells lining the fallopian tubule and uterus. Patches of intense β-gal labeling in tissues obtained from FP +/− mice were associated with dermal hair follicles. Liver failed to show any β-gal staining in hepatocytes, however, labeling of vascular tissue was detected.
The kidney is a site of robust prostaglandin synthesis and expresses abundant prostanoid receptors (8, 10, 46). Renal expression of the FP, EP1, EP3, and TP receptor mRNAs is particularly high (13, 25, 46). Furthermore, many of the signaling pathways activated by this subset of receptors are similar (12, 33), allowing for the possibility that these receptors subserve functionally redundant roles. In this regard, it is of note that striking similarities between the renal effects of PGE2 and PGF2α exist. Similar to PGE2, intrarenal infusion of PGF2α is associated with natriuresis and diuresis, without altering glomerular filtration rate or renal hemodynamics (50). Furthermore, basolateral addition of either PGF2α or PGE2can antagonize ADH-stimulated water absorption in microperfused collecting ducts (43). Nonetheless, because PGF2α potently activates both prostaglandin FP and EP3 receptors (1, 31), it is difficult to attribute these renal affects specifically to activation of the FP receptor. Furthermore, we are unaware of any published studies examining the renal effects of FP receptor-selective agonists. For these reasons, it is important that the present studies now demonstrate segmental expression of FP receptor mRNA along the mouse nephron.
The present studies used both in situ hybridization and a β-gal reporter knocked into the endogenous FP locus (47) to map the distribution of the FP receptor. FP receptor expression determined using these two different techniques was mutually supportive. In the kidney, the most intense labeling was detected over a subpopulation of cells in the cortex. The mouse FP receptor mRNA was most abundant in distal nephron segments colabeling with antibodies to the TSC1 and the vasopressin-stimulated water channel AQP2. In mice, TSC1 is expressed only in the DCT, where it mediates NaCl absorption (15). FP receptor activation could inhibit salt absorption in this nephron segment, thereby contributing to the natriuretic effects of PGF2α. The DCT is also a major site of calcium absorption (32, 38), so it is also conceivable that PGF2α plays a role in modulating Ca2+absorption by the kidney. Similarly, the detection of the FP receptor in AQP2-immunoreactive cells demonstrates its expression in the collecting duct (20, 21), representing another site where its activation could contribute to PGF2α-induced natriuresis and diuresis.
Although the presence of low levels of FP mRNA in the thick ascending limb cannot be excluded, it seems clear that the expression of FP transcripts in the thick ascending limb is markedly less than in either the DCT or cortical collecting duct (CCD). Interestingly, there appears to be a gradient for the intensity of FP gene expression along the distal tubule, with greater levels of expression in the DCT/connecting tubule, > CCD >>MCD. This is in contrast to EP3 mRNA, which is more abundant in medullary CD than CCD and expressed in mTAL as well (9, 11, 45). Finally, the EP1 receptor is most abundant in the papillary collecting duct (25,45). This axial heterogeneity of the prostanoid receptors is consistent with a major role for PGF2α action in the renal cortex as opposed to the medulla, where PGE2 action may predominate.
The cellular effects of the FP receptor in distal renal epithelia remain uncharacterized. In fibroblasts, smooth muscle cells, or cells transfected with the FP receptor, PGF2α activates a signaling pathway coupled to increased cell calcium and phosphatidylinositol hydrolysis (3, 23, 24). A similar signaling pathway is activated by the EP1 receptor in the collecting duct, and this signaling pathway contributes the capacity of PGE2 to inhibit vasopressin-stimulated water flow and sodium absorption (19, 25, 27). Activation of a Ca2+-coupled signaling pathway by the FP receptor in the collecting duct could therefore contribute to natriuresis and diuresis caused by PGF2α infusion. Other studies in transfected cells show that the FP receptor can activate a β-catenin-coupled signaling pathway (18), however, the significance of this pathway in differentiated renal epithelial is uncharacterized. Alternatively, of the known prostanoid receptors, the FP receptor protein sequence is most closely related to the EP3 receptor (37) that preferentially couples to Gi and inhibits vasopressin-stimulated cAMP generation and water flow via this pertussis toxin-sensitive mechanism (25, 27, 41). Additional studies will be required to determine which, if any of these pathways, is activated by the FP receptor in these nephron segments.
As previously reported, β-gal expression was abundant in ovarian corpus luteum where FP activation appears to play a critical role in parturition, initiating the perinatal decline in progesterone secretion (47). The expression of β-gal in corpora luteal cells provides additional validation for concordance of β-gal expression with FP mRNA expression since abundant expression FP mRNA has been demonstrated in corpora lutea of mice by both techniques (44,47). FP receptor is also expressed in uterine smooth muscle (6), consistent with the present studies demonstrating β-gal in this tissue. Robust β-gal activity was also detected along the male genital tract, particularly in the epithelia lining the lumen of the epididymis and vas deferens. Because the epididymis possesses endogenous β-gal activity (17, 22) detected in the wild type (not shown), the significance of staining in this segment of the male genital tract remains uncertain. In contrast, we did not detect β-gal activity in wild-type testis, so FP recetor could be expressed in this tissue and its activation contributes to previous reports that in vivo administration of PGF2α to mice causes atrophy of epididymal epithelium (39).
β-Gal expression was not apparent in hepatocytes, despite reported effects of PGF2α on hepatic glucose output (34), consistent with the possibility of an unrelated receptor or pharmacological target for PGF2α in mediating these effects. Interestingly, intense expression of β-gal activity was observed in hepatic vasculature, where it could mediate the capacity of PGF2α to induce nitric oxide-dependent vasodilatation (4). Finally, the present studies also identified a restricted pattern of FP receptor in skin, particularly in the dermal papillae. This site of expression may be important in mediating the stimulatory effect of latanaprost, an FP-selective agonist, on hair growth (29).
In summary, the present studies demonstrate high levels of expression of mRNA for the FP receptor in kidney distal tubules, including the DCT and CCD. The intrarenal distribution of FP receptor mRNA corresponds with the known effects of PGF2α on salt and water transport in the kidney. The FP receptor is expressed along both male and female genitourinary tracts.
This work was funded in part by an American Heart Association fellowship award (to O. Saito), an American Diabetes Association award (to Y. F. Guan), and VA Merit Award and National Institutes of Health Grant DK-37097 (to M. D. Breyer).
Present address of M. Kömhoff: Dept. of Pediatrics, Philips-Univ. Marburg, Deutschhausstrasse 12, D35053 Marburg, Germany.
Address for reprint requests and other correspondence: M. D. Breyer, Division of Nephrology and Dept. of Medicine, Vanderbilt Univ., F427-ACRE Bldg., Dept. of Veterans Affairs Medical Center, Nashville, TN 37212 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published March 11, 2003;10.1152/ajprenal.00441.2002
- Copyright © 2003 the American Physiological Society