INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

PROSTAGLANDIN (PG) E₂ is a locally acting hormone with diverse biological functions (23, 25). This lipid mediator is produced from the enzymatic metabolism of arachidonic acid through the coordinated actions of cyclooxygenase (PGH synthase) and PGE₂ synthase. PGE₂ is normally produced in a number of tissues, including the gut, the prostate gland, and the kidney. In addition, its production is upregulated during inflammation; enhanced production of PGE₂ is thought to promote and shape a variety of inflammatory reactions (9). The diverse effects of PGE₂ are mediated by specific cell surface receptors that belong to the large family of G protein-coupled receptors (4, 8). Four classes of PGE₂ E-prostanoid (EP) receptors, designated EP₁ through EP₄, can be distinguished pharmacologically. EP receptors of each class have been cloned and sequenced, and each is the product of a distinct gene (4, 8). The tissue distribution and intracellular signaling pathways used by the subclasses of EP receptors differ substantially. These differences provide a mechanism to explain the wide array of effects that are induced by the single ligand, PGE₂.

Among the four classes of EP receptors, the EP₃ receptor is the most widely expressed (8). EP₃ receptors are found in the smooth muscle of the gastrointestinal tract, the uterus, and vascular tissues. In addition, high levels of the EP₃ receptor are expressed in gastric mucosa and in the kidney (5, 29). The EP₃ receptor is unique among prostanoid receptors in that multiple EP₃-receptor isoforms are generated through alternative splicing of the single EP₃-receptor gene (21, 30). These EP₃-receptor isoforms differ only in the COOH terminus. The differences in the COOH terminus produced by alternative splicing result in substantive alterations in intracellular signaling. At least two distinct EP₃-receptor isoforms are present in the kidney (31, 32).

Although some relatively specific agonists of the EP₃ receptor, such as misoprostol, have been identified, these compounds also interact in varying degrees with other EP receptors (8). There are no published EP₃-specific antagonists. Thus the physiological functions of the EP₃ receptor and its contribution to the actions of PGE₂ in vivo have been difficult to study. To examine the physiological functions of the EP₃ receptor, we produced EP₃ receptor-deficient (EP₃ /) mice by gene targeting. These mice are viable and do not exhibit abnormalities of their major organ systems. Because the EP₃ receptor is highly expressed in the renal medulla (5, 29) and because previous studies have suggested that PGE₂ may modulate the actions of vasopressin in the distal nephron (6, 12, 15, 16, 26, 27, 34), we studied urinary concentrating mechanisms in EP₃ / mice.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Disruption of the EP₃ receptor gene in embryonic cells. To clone the EP₃-receptor gene, a partial complementary DNA probe was prepared by polymerase chain reaction (PCR) after reverse transcription (RT) of total RNA isolated from mouse kidney. The EP₃-specific probe was amplified using the primers EP₃1F (5'-CCACATGAAGACTCGCGC-3') and EP₃2R (5'-GTTCAGCGAAGCCAGGCGAAC-3'). The primer sequences were derived from the published sequence of the mouse EP₃ cDNA (28). The resulting partial cDNA fragment corresponds to the region of the EP₃-receptor protein extending from the beginning of the fourth transmembrane domain to the middle of the seventh transmembrane domain.

View larger version (48K): [in this window] [in a new window]   Fig. 1.   Targeted disruption of EP3 locus. A: restriction map of endogenous EP3 gene locus, targeting vector, and targeted locus. Exon 1 of EP3 gene is indicated by shaded box, and phosphoglycerol kinase- (PGK) thymidine kinase (TK) and PGK-neomycin resistance cassettes (NEO) are indicated. Relevant restriction sites: E, EcoR V; Xb, Xba I; S, Sal I; Bg, Bgl I; Sm, Sma I; Xh, Xho I; N, Not I. Boxes below targeted locus indicate probes used to detect homologous recombination events by Southern blot analysis. B: Southern blot analyses of genomic DNA isolated from tail biopsies of progeny from EP3 +/- matings. Targeting event results in loss of a Bgl I site in 3' portion of exon 1 (left) and introduces a novel Bgl I site into locus in 5' end of NEO cassette (right) so that genomic fragment that hybridizes with probe 2 is reduced in size from 7.5 to 6.8 kb.

Generation of EP₃ receptor-deficient mice. Targeted ES cells were introduced into C57BL/6 mouse blastocysts using standard techniques (20). Male chimeras were mated with B6D2 (C57BL/6 × DBA/2 F1; Jackson Labs, Bar Harbor, ME) or 129/SvEv females to identify germ-line-competent chimeras that were capable of transferring the ES cell genome to their offspring. The targeted EP₃ allele was detected in offspring of these crosses by Southern blot analysis of genomic DNA isolated from tail biopsies as shown in Fig. 1B. Offspring carrying the mutant allele were intercrossed to obtain animals that were homozygous for the targeted mutation (EP₃ /).

RNA isolation and Northern analysis. Kidneys were harvested from neonatal EP₃ +/+ and EP₃ / mice, and RNA was extracted from the tissues using a commercially available reagent (RNAzol B; Tel-test). Twenty micrograms of each sample were fractionated by electrophoresis on a 1.2% agarose/formaldehyde gel and transferred to nitrocellulose membranes. The membranes were then hybridized at high stringency with a partial EP₃-receptor cDNA probe that was prepared by RT-PCR as described above and labeled with ³²P by random priming, using a commercially available kit (Amersham). After hybridization and washing, the filters were exposed to Kodak X-AR film for up to 1 wk at 80°C.

Urinary osmolality in mice lacking EP₃ receptors and the effects of the PG synthase inhibitor indomethacin. We first measured urine osmolalities in EP₃ +/+ (n = 7) and EP₃ / (n = 7) mice that were given free access to water. Urine osmolalities in these studies and all of the studies that follow were measured immediately after sample collection using a vapor-pressure osmometer (Wescor Instruments, UT). To assess the effects of the inhibition of PG synthesis on urinary osmolality in EP₃-deficient mice, we collected urine specimens from EP₃ +/+ (n = 7) and EP₃ / (n = 7) mice before and 12 h after the administration of indomethacin by gavage (10 mg/kg), and urine osmolalities were measured immediately.

Responses to a vasopressin analog. We measured the effects of the vasopressin analog desmopressin (DDAVP; Rhone-Poulenc Rorer, Collegeville, PA) on urine osmolality in EP₃ +/+ (n = 7) and EP₃ / (n = 7) mice. Before the experiments, animals were allowed free access to drinking water and 0.4% NaCl chow. After the collection of a baseline urine sample, mice were injected with 1.0 mg/kg DDAVP sc, and the water bottles were removed from their cages. Four hours after the DDAVP injections, urine samples were collected and urine osmolalities were immediately measured.

Effects of different levels of water intake on urinary volume and osmolality in EP₃ +/+ and EP₃ / mice. To examine the effect of the EP₃ mutation on urinary concentration in response to alterations in fluid intake, we housed EP₃ +/+ (n = 6) and EP₃ / (n = 6) mice in specially constructed metabolic cages. During an initial control period, urine volumes and osmolalities were measured while the animals had free access to water and 0.4% NaCl chow. After the control period, 10% dextrose was added to drinking water to increase water intake over a period of 48 h. Urine samples were free of glucose as determined by dipsticks (Glucochex). The 10% dextrose solution was then replaced with distilled water for a recovery period of 24 h. Water bottles were then removed during a 48-h period of water deprivation. Urinary flow rates were measured and expressed as milliliters per 24 hours per 20 g of body weight. Osmolality of urine was measured immediately after sample collection just before institution of the 10% dextrose or the onset of water deprivation and at 12- to 24-h intervals thereafter. During the entire experiment, all animals were weighed daily and were allowed free access to 0.4% NaCl chow.

Measurement of renal function. On the day of study, animals were anesthetized with 0.04 mg/g pentobarbital, and a polyethylene catheter (PE-90) was inserted into the trachea to facilitate spontaneous ventilation. The left carotid artery and left jugular vein were cannulated with polyethylene catheters (PE-10) for intravenous infusions, monitoring of mean arterial pressure (MAP) (Caldwell Systems model 8802; Rougemont, NC), and intermittent sampling of arterial blood. After surgery, normal saline (2.0% of body wt) was infused intravenously over 20 min to replace surgical losses. A priming dose of [carboxyl-¹⁴C]inulin and ³H-labeled p-aminohippuric acid ([glycyl-³H]PAH) was given, followed by infusion of pentobarbital, [carboxyl-¹⁴C]inulin, and [glycyl-³H]PAH in normal saline at a rate of 25 µl · min¹ · 100 g body wt¹. The bladder was cannulated via a suprapubic incision with a PE-50 catheter. After 30 min of equilibration, renal function was measured during two consecutive 30-min clearance periods. [Carboxyl-¹⁴C]inulin and [glycyl-³H]PAH in plasma and urine were measured in a liquid scintillation counter (Nuclear Chicago-TM Analytical, Elk Grove, IL), and clearances were calculated using standard formulas.

Data analysis. The values for each parameter within a group are expressed as means ± SE. For comparisons between EP₃ +/+ and EP₃ / groups, statistical significance was assessed using an unpaired t-test. A paired t-test was used for comparisons within groups.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Generation of EP₃-deficient mice. The gene encoding the EP₃ receptor was disrupted in the E14Tg2a ES cell line by homologous recombination with a targeting plasmid (shown in Fig. 1A). The targeting plasmid was designed to replace exon 1 of the EP₃ gene with a neomycin resistance cassette. Sequence analysis of our genomic clone suggested that the organization of the mouse EP₃-receptor gene is similar to that described for the human gene (17), wherein exon 1 contains the 5'-untranslated region of the EP₃ mRNA along with a large portion of the coding region extending from the initiation codon through the sixth transmembrane domain. The first intron begins in the sixth transmembrane domain, a position that is conserved between mouse and human genes (17). Because the amino acids encoded by the first exon are included in all of the known EP₃ isoforms, homologous integration of the targeting vector should prevent normal transcription of the EP₃ gene and should preclude expression of any functional EP₃-receptor isoform. Targeted ES cell lines were identified by Southern analysis and introduced into blastocysts to generate chimeric mice. Chimeras were bred to 129/SvEv or B6D2 mice, and offspring carrying the mutant allele were identified by Southern analysis, as shown in Fig. 1B.

EP₃ mRNA expression in mutant mice. To verify that the targeted mutation introduced into the EP₃ gene locus resulted in inactivation of the gene, we examined expression of the EP₃ mRNA by Northern blot. Total RNA was prepared from kidneys of EP₃ +/+ and EP₃ / animals, and Northern analysis was performed using an EP₃-receptor cDNA probe. As shown in Fig. 2, two bands that hybridized with the EP₃ probe can be detected in RNA prepared from the kidneys of EP₃ +/+ mice. The larger band is ~7 kb in size and the smaller band is 2.3 kb, consistent with previous reports that have documented multiple EP₃ transcripts in mouse kidney (28). Neither of these EP₃ mRNA transcripts could be detected in the kidneys of the EP₃ / animals.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Expression of EP3 mRNA in wild-type (EP3 +/+) and mutant (EP3 /) mice. A: RNA was isolated from kidneys of newborn EP3 +/+ and EP3 / littermates, and Northern analysis was performed using an EP3 cDNA probe. B: filter was stripped and rehybridized with an actin cDNA probe to confirm equivalency of RNA loading.

The effects of indomethacin on urine osmolality in EP₃ / mice. EP₃ receptors are highly expressed in the distal nephron, including medullary and cortical thick ascending limbs, as well as the cortical and medullary collecting duct (5, 29, 31, 33). Although previous studies have suggested that PGE₂ may modulate water reabsorption in these nephron segments (6, 12, 15, 16, 26, 27, 34), the specific EP receptors involved in these effects have not been identified. Thus we were interested in examining the effects of the PG-synthesis inhibitor indomethacin on urinary osmolality (U_Osm) in EP₃ / mice. When mice were given free access to water, there were no differences in U_Osm between the EP₃ +/+ and EP₃ / animals (2,323 ± 151 vs. 2,564 ± 161 mosmol/kgH₂O; P = 0.28).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of inhibiting PG synthesis on urinary osmolality. Administration of PG synthesis inhibitor indomethacin (10 mg/kg) caused significant increase in basal urine osmolality (UOsm) in EP3 +/+ mice () but did not significantly affect UOsm in EP3 / mice (). * P < 0.02 vs. control;  P < 0.05 vs. EP3 /.

DDAVP has similar effects on U_Osm in EP₃ +/+ and EP₃ / mice. To examine the role of EP₃ receptors in modulating responses to vasopressin, we administered the vasopressin analog DDAVP to EP₃ +/+ and EP₃ / mice. As depicted in Fig. 4, DDAVP rapidly increased urine osmolalities in EP₃ +/+ (+1,027 ± 175 mosmol/kgH₂O; P < 0.007 vs. pretreatment baseline) and in EP₃ / mice (+1,127 ± 93 mosmol/kgH₂O; P < 0.002 vs. pretreatment baseline). The effects of DDAVP on urinary osmolality were similar between the EP₃ +/+ and EP₃ / groups, and 4 h after administration of the vasopressin analog there was no significant difference in their urine osmolalities (3,482 ± 173 vs. 3,227 ± 91 mosmol/kgH₂O; P = 0.24).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Changes in urine osmolality after injection of vasopressin analog desmopressin (DDAVP). Urine osmolality was measured before and 4 h after subcutaneous injections of vasopressin analog DDAVP (1 µg/kg) in EP3 +/+ and EP3 / mice.

Renal concentrating and diluting mechanisms are intact in EP₃ / mice. To determine whether the absence of EP₃ receptors affected urinary concentration and dilution in response to physiological stimuli, we compared the effects of varied levels of water intake on urine volumes and osmolalities in EP₃ +/+ and EP₃ /. When 10% dextrose solution was added to their water, both groups increased their water intake significantly and to a similar extent (3.45 ± 1.42 vs. 11.62 ± 2.25 ml/day, P < 0.02, for EP₃ +/+ mice and 3.43 ± 1.11 vs. 11.65 ± 1.83 ml/day, P < 0.008, for EP₃ / mice). As depicted in Fig. 5A, this enhanced water intake caused both groups to excrete significantly larger volumes of dilute urine compared with the control period. The level of urine output and the minimal urine osmolalities achieved during this period were not different between the EP₃ +/+ and EP₃ / groups (Fig. 5B). Similarly, when they were deprived of water, both EP₃ +/+ and EP₃ / mice rapidly initiated an antidiuretic response. Both groups concentrated their urine to maximal levels within 24 h. As shown in Fig. 5, the responses of the EP₃ / mice to water deprivation were virtually identical to controls.

View larger version (15K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Fig. 5.   Urine volume and osmolality during water loading and deprivation. A: urine volumes in EP3 +/+ () and EP3 / () mice during control periods (when mice had free access to water), during water loading (when 10% dextrose was added to drinking water), and after 12, 24, and 48 h of water deprivation. B: urine osmolalities in EP3 +/+ () and EP3 / ( ) mice during experimental periods.

Renal function is normal in EP₃ / mice. Because PGE₂ can modulate renal hemodynamic function through its effects on renal vasculature (11), we measured glomerular filtration rate (GFR) and renal plasma flow (RPF) in anesthetized EP₃ / mice. We found that levels of GFR (10.54 ± 0.4 vs. 10.84 ± 1.14 ml · min¹ · kg¹; P = 0.82), measured as inulin clearance, and RPF (23.0 ± 2.54 vs. 24.98 ± 4.01 ml · min¹ · kg¹; P = 0.70), measured as PAH clearance, were similar in EP₃ +/+ and EP₃ / mice.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

PGE₂, a cyclooxygenase product of arachidonic acid, mediates a wide range of biological activities that are involved in the functions of virtually every organ system (23, 25). The diverse effects of PGE₂ are mediated by a family of EP receptors that can be divided pharmacologically into four subtypes (EP₁ through EP₄) (4, 8). Receptors representing these four subtypes have been cloned, sequenced, and found to be products of distinct genes. Each EP receptor is characterized by different patterns of coupling to intracellular signaling pathways, reflecting distinct physiological functions. However, the functional roles for the individual EP-receptor subtypes have been difficult to define in vivo because of the lack of potent and specific pharmacological agents that are capable of discriminating among the various EP receptors.

In these studies, we have used gene targeting to begin to study the functions of the most ubiquitous EP-receptor subtype, the EP₃ receptor. Proposed functions for EP₃ receptors include 1) promoting smooth muscle contraction in the gastrointestinal tract, uterus, and blood vessels; 2) inhibiting neurotransmitter release in autonomic nerves; 3) preventing lipolysis; 4) reducing acid secretion by gastric mucosa; and 5) promoting aggregation of platelets (4, 8). However, the specific contribution of EP₃ receptors to these functions in the intact organism has not been clearly established. In our studies, we found that mice lacking EP₃ receptors are viable, survive in expected numbers, and appear grossly normal and healthy. In addition, we found that pregnancies, gestation, and litter sizes are normal in EP₃ / females, suggesting that EP₃ receptors on the uterus and other reproductive tissues are not required for their normal function. The development of other major organ systems, including the gastrointestinal tract, cardiovascular system, and kidney, also appears to be normal in EP₃ / mice.

PGE₂ is the major cyclooxygenase metabolite produced by the kidney, and it is the predominant prostanoid product excreted in the urine (3, 11, 24). Among its effects in the kidney, PGE₂ inhibits the actions of the neuropeptide vasopressin (2, 6, 12, 16, 26, 27, 34). Vasopressin is the key regulator of water excretion by the kidney. This regulation is accomplished through the ability of vasopressin to stimulate water transport in segments of the distal nephron (1, 18). The increase in hydraulic permeability caused by vasopressin allows equilibration between luminal fluid and the regions of the inner renal medulla, where ambient osmolality is very high. This results in excretion of highly concentrated urine and conservation of body water. The effects of vasopressin on water transport are mediated through stimulation of cAMP production (1, 18). In cortical collecting tubules, PGE₂ inhibits cAMP production and reduces vasopressin-stimulated water reabsorption (6, 14, 22, 26, 27, 34). The EP₃ receptor is highly expressed in the renal medulla as well as the cortical collecting duct (5, 29, 33), and it couples to G_i proteins that inhibit adenylyl cyclase activity (4, 8). Accordingly, it has been suggested that the effects of PGE₂ to inhibit the actions of vasopressin are mediated by EP₃ receptors.

To examine the regulation of urinary concentration by the EP₃ receptor in vivo, we studied water homeostasis in EP₃ / mice. In EP₃ / mice provided water ad libitum, we found that the level of water intake and the volumes and osmolalities of urine were similar to controls. To determine the contribution of prostanoids to the regulation of urine osmolality, we measured urine osmolalities before and after administration of the cyclooxygenase inhibitor indomethacin. In EP₃ +/+ mice, acute inhibition of prostanoid production was associated with a significant rise in urine osmolality. A similar effect of acute cyclooxygenase inhibition has been described in humans (2, 19). In contrast, inhibiting PG synthesis had no effect on urine osmolality in mice lacking EP₃ receptors. This suggests that, under basal conditions, when mice have free access to water, PGE₂ modulates the osmolality of urine through the EP₃ receptor.

We considered that the difference in indomethacin response between EP₃ / and EP₃ +/+ mice might be related to the well-documented effect of PGE₂ to inhibit vasopressin-stimulated water reabsorption as mentioned above. To directly examine the role of EP₃ receptors in modifying the effects of vasopressin in the kidney, we compared the effects of a vasopressin analog on urine osmolality in EP₃ +/+ and EP₃ / mice. Although we anticipated that administration of DDAVP might increase urine osmolality in EP₃ / mice to a greater extent than controls, the change in urine osmolality caused by DDAVP was essentially identical between the two groups. This result suggests that binding of PGE₂ to the EP₃ receptor does not modulate changes in urine osmolality caused by acute increases in the circulating levels of vasopressin. These data, along with the indomethacin experiment, may indicate that the effects of PGE₂ on urinary concentration are not directly dependent on vasopressin.

Alterations in blood flow in the renal medulla can have profound effects on the generation of maximal osmolar gradients (7). Because PGE₂ acts as a vasodilator in the renal circulation, changes in medullary blood flow due to reduced PGE₂ synthesis might be expected to increase urine osmolality without a change in vasopressin activity. The contribution of such an effect by indomethacin has been demonstrated in rats (7). However, the role of EP₃ receptors in regulating medullary blood flow is not clear. Our finding that GFR and RPF are normal in EP₃ / mice argues against a primary role for EP₃ receptors in regulating hemodynamics in the whole kidney but does not rule out an effect in regulating regional blood flow. An alternative explanation of our indomethacin findings could be that the effects of PGE₂ on urine osmolality are detected at submaximal levels of vasopressin, but these are overcome at the very high levels that are present during dehydration or when pharmacological doses of vasopressin analog are administered.

After enhanced water intake or during water deprivation, urine volumes and osmolalities in EP₃ / mice were not different from controls. This suggests that the EP₃ receptor is not essential for the normal regulation of urinary concentrating mechanisms in response to various physiological stimuli. However, previous studies suggest that the effects of PGE₂ on water transport are complex. For example, PGE₂ increases basal water permeability in isolated cortical collecting ducts but inhibits water reabsorption in collecting tubules that are first exposed to vasopressin (12, 27). Thus the net effects of PGE₂ on water reabsorption may vary depending on experimental conditions. Moreover, the major effect of PGE₂ in these circumstances is to modulate the actions of vasopressin, rather than to function as a primary determinant of net water transport. It would not, therefore, be surprising that chronic compensatory responses might develop that could substitute for EP₃ receptors in this role. These responses might involve other hormonal influences on water transport or direct effects on the generation of medullary concentration gradients such as the alterations in medullary blood flow that were discussed previously. Further evidence for the existence and potential of compensatory mechanisms in this system is suggested by studies of humans treated with nonsteroidal anti-inflammatory drugs (NSAIDs), in these studies acute administration of cyclooxygenase inhibitors significantly alters urinary osmolality, but this effect disappears with time in subjects that are treated chronically with NSAIDs (2, 19).

Alternatively, EP receptors other than EP₃ may be involved in the inhibition of vasopressin's actions by PGE₂. These receptors may be capable of performing or compensating for this function when the EP₃ receptor is absent. In support of this hypothesis are a series of studies that suggest heterogeneity of PGE₂ binding and signaling in cortical collecting tubules. For example, PGE₂ can stimulate intracellular calcium release in isolated cortical collecting ducts (13, 22). The effects of PGE₂ to inhibit both vasopressin-stimulated water reabsorption and cAMP production in cortical collecting ducts can be partially reversed by an inhibitor of protein kinase C (14, 22), suggesting that this calcium signal might be linked to inhibition of vasopressin's actions in these segments. Moreover, at least two EP₃-receptor isoforms are present in vasopressin-sensitive nephron segments (31, 32). One of these isoforms (EP₃A) couples to G_i, and the other (EP₃B) is linked to intracellular calcium. The EP₁ agonist sulprostone inhibits cAMP generation in cortical collecting tubule cells (14, 27), and stimulation of the EP₁ receptor is commonly associated with increases in intracellular calcium (8). Thus the relative preservation of urinary concentrating functions in EP₃ / mice might be due, in part, to the compensatory effects of EP₁ receptors, which mediate functions in cortical collecting tubules that overlap with the EP₃ receptor. This hypothesis can be tested directly as EP₁ / mice become available.

In summary, our analysis of the phenotype of mice lacking EP₃ receptors for PGE₂ suggests that these receptors are not necessary for survival and normal development. Although stimulation of the EP₃ receptor by PGE₂ plays a role in modulating urine osmolality in mice with free access to water, the mechanism of this effect is not completely clear. Nonetheless, the EP₃ receptor is not required for normal regulation of urine volume and osmolality in response to various physiological stimuli. The preservation of these normal responses in the absence of EP₃ receptors may result from the compensatory effects of other systems, including the EP₁ receptor. Thus the EP₃ / mouse model should prove to be a valuable tool in defining these other mechanisms for regulating water metabolism.