In adult animals and humans, activation of κ-opioid receptors results in a diuresis. The aim of the present study was to investigate whether κ-opioids are also diuretic early in life and whether this is altered during postnatal maturation. Therefore, the renal effects of the κ-opioid-receptor agonist U-50488H were measured in two separate age groups of conscious lambs at two stages of postnatal maturation (∼1 wk and ∼6 wk) under physiological conditions. To evaluate whether the renal responses to U-50488H resulted from receptor-dependent effects, responses to U-50488H were also tested in the presence of the specific κ-opioid-receptor antagonist 5′-guanidinonaltrindole (GNTI). Urinary flow rate, free water clearance, and electrolyte excretions and clearances were measured for 30 min before and for 90 min after intravenous injection of U-50488H or vehicle. An increase in urinary flow rate accompanied by an increase in free water clearance occurred in response to administration of U-50488H but not vehicle. There were no effects of U-50488H on electrolyte excretions or clearances at either 1 or 6 wk of postnatal life. Although there were no effects of GNTI on any of the measured or calculated variables, the aforementioned diuretic response to U-50488H was abolished by pretreatment with GNTI in both age groups. We conclude that κ-opioid receptors are diuretic early in life and that this response does not appear to be altered as postnatal maturation proceeds. Therefore, these data provide evidence that activation of κ-opioid receptors early in life may lead to alterations in fluid balance.
- renal function
endogenous opioid peptides are involved in a variety of physiological processes, including cardiovascular and endocrine regulation, and extend to feeding, pain, substance abuse, traumatic brain injury, and hemorrhagic shock. These effects are elicited through four families of peptides (the dynorphins, β-endorphins, enkephalins, and nociceptins/orphanin FQ), which act through specific opioid-receptor subtypes designated as kappa (κ), mu (μ), and delta (δ) and the N/OFQ peptide receptor, respectively.
It is generally well recognized that activation of κ-opioid receptors leads to a diuresis in rats (13, 23–26), mice (36), dogs (42), and humans (21, 31). The diuretic effect of κ-opioid-receptor activation appears to result from a combination of 1) its direct inhibitory effects on the tubular action of arginine vasopressin (AVP) in promoting water reabsorption and 2) suppression of AVP release from the neurohypophysis (2–4, 9, 25–26, 46). Despite the diuretic response to activation of κ-opioid receptors by either receptor agonists or the naturally occurring ligand, dynorphin A, there appears to be no consistent accompanying increase in electrolyte excretion (1–2, 48). For this reason, κ-opioid peptides are considered aquaretic.
Little is known, however, regarding the renal responses to activation of κ-opioid receptors during the perinatal period or whether they are in fact aquaretic during this time. Jackson and Kitchen (14) reported an increase in urine output (by manual bladder emptying) in response to intraperitoneal administration of U-50488H to 10-day-old rat pups, although no other measurements were made and no additional age groups were studied. There have been no other studies into the possible role of κ-opioid receptors in modulation of kidney function during the perinatal period.
The present experiments were carried out to measure renal responses to activation of κ-opioid receptors during postnatal maturation in conscious lambs. To this end, the specific agonist U-50488H was administered in two separate age groups of lambs at two stages of postnatal maturation under physiological conditions in order to 1) define renal responses to κ-opioid-receptor activation and 2) determine whether these effects are altered during postnatal maturation. Experiments were also carried out in the presence of the selective κ-opioid-receptor antagonist 5′-guanidinonaltrindole (GNTI) to evaluate renal effects of GNTI and the response to activation of κ-opioid receptors by U-50488H.
Experiments were carried out in two separate groups of conscious, chronically instrumented sheep aged ∼1 and ∼6 wk at the time of study. Animals were obtained from a local source (Woolfitt Acres, Olds, Alberta, Canada) and housed with their mothers in individual pens in the vivarium, except during surgery and experiments. All surgical and experimental procedures were carried out in accordance with the Guide to the Care and Use of Experimental Animals provided by the Canadian Council on Animal Care and with the approval of the Animal Care Committee of the University of Calgary.
Using aseptic techniques and previously detailed methods (43, 45), we performed surgery on newborn lambs 2–5 days after birth and on older lambs a minimum of 4 days before the start of experiments. Briefly, anesthesia was induced with a mask and halothane (3–4%) in oxygen, the trachea was intubated, and anesthesia was maintained by ventilating the lungs with halothane (0.5–1%) in a mixture of nitrous oxide and oxygen (3:2). Under sterile conditions, polyethylene catheters were inserted into left and right femoral veins and arteries (1.19 mm ID, 1.70 mm OD; Intramedic), advanced to the inferior vena cava and abdominal aorta for later intravenous infusions and arterial sampling, and tunneled subcutaneously to exit the lamb on right and left flanks. Through an abdominal midline incision, the bladder was then exposed, and a catheter was inserted directly across the bladder wall by use of a specially adapted feeding tube (Medi-Craft) for collection of urine and measurement of urinary flow rate during experiments.
After incisions were closed, all catheters were secured in a body jacket (Lomir, Montreal, Canada) for safe storage between experiments. Antibiotics (5.0 mg/kg enrofloxacin; Baytril) were administered intramuscularly at surgery and at 12-h intervals thereafter for 48 h. Lambs were allowed to recover from surgery in a critical care unit for small animals (Shor-line, Schroer Manufacturing), with adjustable oxygen supply. All lambs were able to stand within 60 min of completion of surgery, after which they were returned to the vivarium.
Experiments were not started until a minimum of 3 days had elapsed after the day of surgery. During this time, animals were trained to rest comfortably in a supportive sling in the laboratory environment for 1–2 h daily to allow them to become accustomed to their surroundings before the start of experiments.
Three different protocols were carried out in the two separate age groups of lambs, as detailed below. On the day of an experiment, the lamb was removed from the vivarium and placed in a supportive sling in the laboratory environment for at least 60 min. During this time, the bladder was allowed to drain, and an intravenous infusion of 5% dextrose in 0.9% sodium chloride was started at a rate of 4.17 ml·kg−1·h−1 and continued for the duration of the experiment to assist in maintenance of fluid balance. After each experiment, the lamb was returned to the vivarium where it was housed with the ewe until the next study day. After all experiments were performed according to the protocols described below, lambs were euthanized with pentobarbital sodium. Both kidneys were removed and immediately weighed.
Lambs aged 9 days (SD 2) [7.6 kg body wt (SD 1.4); n = 11] and 41 days (SD 5) [12.6 kg body wt (SD 2.3); n = 13] were studied in the following dose-response experiments (with n = 3–6 animals for each dose). This protocol was designed to establish the relationship between U-50488H dose and cumulative urinary flow in the two age groups of animals and to determine the maximal effective dose (ED100), defined as the minimum dose that produces the maximum response to a drug, for both age groups, to be applied in protocols 2 and 3.
Each experiment consisted of consecutive urinary collection periods for 30 min before (control) and 90 min after intravenous administration of the specific κ-opioid-receptor agonist U-50488H. In each animal, at intervals of 24–48 h and in random order, one of the following doses of U-50488H was tested: 0, 0.01, 0.1, 0.5, 1.0, and 5.0 mg/kg until a maximum of three doses had been tested in each lamb. At the end of each 30-min collection period, urine volume was recorded for later calculation of cumulative urinary flow rate.
Lambs aged 7 days (SD 2) [8.2 kg body wt (SD 1.4); n = 7] and 40 days (SD 1) [12.3 kg body wt (SD 2.5); n = 7] were studied. The purpose of this protocol was to measure renal responses to ED100 of U-50488H as well as vehicle at two postnatal ages. Two experiments were carried out in each animal in random order at intervals of 24–48 h as follows: U-50488H (ED100 selected from protocol 1, experiment 1) and vehicle (and equivalent volume of 0.9% saline, experiment 2). Each experiment consisted of consecutive 10-min urinary collection periods for 30 min before (control, collection 1–3) and 90 min after intravenous injection of U-50488H or vehicle (collection 4–12). At the end of each 10-min urinary collection period, urine volume was recorded and samples were stored at −70°C for later determination of urinary electrolytes (Na+, K+, Cl−) and urinary osmolality (UOsm). At the midpoint of the subsequent urinary collection periods (collection 2, 5, 7, 9, 11), arterial blood (2.5 ml) was removed and placed immediately into 10 U of heparin in chilled polypropylene tubes; samples were centrifuged, and supernatant was removed and stored at −70°C for later determination of plasma electrolytes (Na+, K+, Cl−) and plasma osmolality (POsm).
Lambs aged 16 days (SD 3) [8.7 kg body wt (SD 1.1); n = 8] and 45 days (SD 5) [15.2 kg body wt (SD 1.6); n = 10] were studied. The purpose of this third protocol was to determine the role of κ-opioid receptors in eliciting the renal responses to U-50488H measured in protocol 2 by evaluating its response before and after GNTI. The duration of the inhibitory effect of GNTI was also evaluated as follows. Responses to ED100 of U-50488H were measured 24 h before (control) and 1, 24, 48, 72, and 96 h after intravenous injection of GNTI or vehicle. A maximum of three time points after control was tested in each animal, with the five time points assigned randomly, so that for each time point there were three to five experiments carried out in different lambs in the two age groups. (The ED100 for GNTI was determined in preliminary experiments over the range of doses of 0 to 5.0 mg/kg as the minimum dose required for maximum inhibition of the aquaretic effect of U-50488H and was 0.25 mg/kg in 1-wk-old lambs and 0.5 mg/kg in 6-wk old lambs.)
Details of Drug Selection and Preparation
U-50488H is (+/−)-trans-3,4-dichloro-N-methyl-N-[2-(1-pyrrolidinyl)cyclohexyl]-benzene acetamide methane sulfonate salt (Sigma-Aldrich, St. Louis, MO). U-50488H was selected because it is a highly selective κ-opioid-receptor agonist and relatively inactive at the μ-receptor (22, 47). U-50488H was dissolved in 0.9% sodium chloride to make a stock solution of 20 mg/ml. Different doses of U-50488H were prepared by serial dilution and stored in aliquots at −20°C until later use. Each dose was administered over 5 s in a volume of 0.35 ml/kg followed by intravenous infusion of ∼5 ml of 0.9% sodium chloride.
GNTI is 5′-guanidinyl-17-(cyclopropylmethyl)-6,7-dehydro-4,5α-epoxy-3,14-dihydroxy-6,7–2′,3′-indolomorphinian dihydrochloride (Tocris, Ellisville, MO). GNTI is a potent κ-opioid-receptor antagonist displaying 208- and 799-fold selectivity over μ- and δ-receptors, respectively; a greater antagonist potency than the prototypical κ-opioid-receptor antagonist norbinaltorphimine; and with a considerably shorter half-life (15–16). GNTI was dissolved in sterile water to make a solution of 10 mg/ml.
Urine and plasma samples were thawed to room temperature, and urinary and plasma electrolyte concentrations (Na+, K+, Cl−) and osmolalities were measured in each sample in duplicate with a flame photometer (IL-943, Lexington), titrator (Labconco Digital Chloridometer, Kansas City, MO), and microsmometer (model 3MO; Advanced Instruments, Needham Heights, MA).
Clearances of electrolytes (Cx) were calculated as follows: Cx = (Ux·V̇)/Px, where U and P refer to urinary concentrations and V̇ refers to urinary flow rate. Free water clearance (CH2O) was calculated as the difference between urinary flow rate and osmolar clearance. The ratio of urinary Na+ concentration to urinary K+ concentration was also calculated. The transtubular K+ gradient was calculated as UK/PK/(UOsm/POsm), where UOsm/POsm is the ratio of UOsm to POsm and UK and PK are urinary and plasma K+. To normalize renal function between the two age groups of lambs, parameters of renal function were corrected for total kidney weight. For protocol 2, data were averaged over 30-min intervals to control, 30, 60, and 90 min.
For protocol 1, dose-dependent effects of U-50488H on urinary flow rate were evaluated by one-way ANOVA for each age group (49), and the ED100 was determined as the dose at which no further change from control occurred. For protocol 2, multivariate ANOVA for repeated measures over time was applied to the measured variables, with factors being age (1 wk and 6 wk) and treatment (U-50488H and vehicle). Where the F value was significant, a Student-Newman-Keul's multiple-comparison procedure was applied to determine where the significant difference(s) occurred (49). For protocol 3, one-way ANOVA was applied to determine the effects of GNTI on the renal response to U-50488H in each age group. Where the F value was significant, a Holm's Sidak multiple-comparison procedure was applied to determine where the significant difference(s) occurred (27). Significance was accepted at the 95% confidence interval. All values are presented as means (and SD).
Figure 1 illustrates the U-50488H dose-to-cumulative urinary flow rate relationship over 90 min in both age groups. There was an overall effect of dose on cumulative urinary flow rate in both age groups, with the point at which no further increase in cumulative urinary flow rate occurred being defined as the ED100, which was 0.5 mg/kg at 1 wk and 1.0 mg/kg at 6 wk (see Fig. 1); these doses were applied in protocols 2 and 3.
There were no differences in any of the measured or calculated variables between vehicle- and U-50488H-treated animals measured during control in either age group of lambs (Table 1).
Figure 2 illustrates the increase in urinary flow rate that occurred at 60–90 min at 1 wk and 30 and 90 min at 6 wk. Although there was an overall effect of U-50488H treatment (F = 5.23; P = 0.031) and time over 90 min (F = 4.35; P = 0.007), there were no interactions between treatment and time. Figure 3 illustrates the increase in CH2O that occurred at 60–90 min at 1 wk and at 90 min at 6 wk. For CH2O, there was an effect of treatment (F = 4.95; P = 0.036) and time (F = 8.36; P < 0.0001), as well as an interaction between treatment and time (F = 5.39; P = 0.002). For both variables, there were no effects of age. There was, however, an interaction between age, treatment, and time (F = 2.81; P = 0.045) for UOsm and for UOsm/POsm (F = 2.82; P = 0.045) (Table 2). Therefore, for these two variables (UOsm and UOsm/POsm), the effect of U-50488H was different at 1 wk compared with that at 6 wk.
There were no effects of U-50488H or vehicle on the excretion rates of Na+, K+, and Cl− or on the clearances of Na+, K+, and Cl− and no effects of U-50488H or vehicle on ratio of urinary Na+ concentration to urinary K+ concentration or on transtubular K+ gradient.
There were no effects of GNTI on any of the measured baseline values in either age group. The aforementioned aquaretic response to U-50488H was, however, abolished by pretreatment with GNTI at 1 h; this inhibitory effect was sustained for 48 h (Fig. 4) in both age groups. In contrast, the response to U-50488H was not altered by administration of vehicle at any of the time points tested (data not shown).
The present study was carried out in conscious lambs to investigate the renal effects of κ-opioid-receptor activation during postnatal maturation. Administration of the specific κ-opioid-receptor agonist U-50488H to lambs aged ∼1 and ∼6 wk elicits a diuresis that is abolished by the selective κ-opioid-receptor antagonist GNTI. Because the diuresis was unaccompanied by any alterations in electrolyte excretion and there was an accompanying increase in CH2O, the effects of κ-opioid-receptor activation were in fact aquaretic. These data provide evidence that, in the newborn period, activation of κ-opioid receptors may lead to alterations in fluid balance.
It has been known for more than 20 years that, at least in adults, κ-opioid compounds are diuretic (17). Slizgi and Ludens (41) first reported a diuretic response to subcutaneous administration of the partial κ-opioid-receptor agonist and the benzomorphan analog ethylketocyclazocine (EKC) to conscious rats. Leander (23) described a diuresis following subcutaneous administration of bremazocine, EKC, and ketazocine, which was inhibited by the selective κ-opioid-receptor antagonists WIN-44,441 and MR-2266BS. Huidobro-Toro and Parada (13) also measured a diuretic response to intraperitoneal administration of κ-opioid-receptor agonists to hydrated conscious rats, which was abolished by WIN-44,441 and MR-2266BS but not by the nonselective antagonist naloxone (13). In pentobarbital sodium-anesthetized dogs, Slizgi and Ludens (41) reported a diuretic response to intravenous administration of the selective κ-opioid-receptor agonist U-50488H in the absence of any changes in glomerular filtration rate (GFR), determined as the clearance of inulin. Dykstra et al. (7) also reported a dose-dependent diuretic response to subcutaneous administration of U-50488H in conscious rhesus monkeys. Again, this was abolished by MR-2266BS. More recently, Kramer et al. (21) showed in human subjects that administration of the κ-opioid-receptor agonist asimadoline elicits a diuresis in the absence of changes in GFR, measured as the clearance of endogenous creatinine. Together, these studies provide definitive evidence that, in all adult species studied to date including human, activation of κ-opioid receptors elicits a diuresis.
The aforementioned diuretic effects of activation of the κ-opioid receptor appear to 1) result from a combination of peripheral and centrally mediated effects and 2) involve inhibition of the antidiuretic hormone AVP, leading to an increased CH2O (17). In conscious Sprague-Dawley rats, Slizgi and Ludens (39) reported diuretic effects of various κ-opioid peptides and correlated their diuretic activity to EKC binding sites within the rat kidney. They also measured a diuresis in response to subcutaneous administration of EKC to conscious Sprague-Dawley but not Brattleboro rats (40), who lack endogenous AVP, and reported an inhibitory effect of EKC on the antidiuretic response to exogenous AVP in Brattleboro rats, suggesting a renal involvement of κ-opioid-induced diuresis. Yamada et al. (48) measured renal responses to intraperitoneal injection of U-62066E, a compound structurally related to U-50488H, in conscious Brattleboro and Long-Evans rats in vivo. In addition, U-62066E was applied in vitro to inner medullary collecting duct cells isolated and perfused from the kidney of Sprague-Dawley rats. The diuresis elicited by U-62066E in Long-Evans rats was abolished by pretreatment with MR-2266BS and did not occur in Brattleboro rats. Moreover, U-62066E did not influence either basal or AVP-stimulated osmotic water permeability in inner medullary collecting duct cells, suggesting that its primary site of action is on AVP release from the pituitary. Further evidence to support a predominantly central nervous system effect of κ-opioid peptides was provided by Ellis and Adam (8) who showed that, in the isolated perfused rat kidney, ketocyclazocine had no direct renal effects.
Similarly, administration of U-50488H to conscious newborn lambs in the present study produced a diuresis that resulted from an increase in CH2O, therefore defined as an aquaresis. Although all animals responded to U-50488H with an aquaresis, the time at which this response occurred and the magnitude of the aquaretic response between different animals were variable. Although we have no definitive explanation, one possibility is that the altered responses correlate to differences in circulating AVP levels during control, before administration of the κ-opioid-receptor agonist. Because this aquaretic response to U-50488H was abolished 1 h after pretreatment with GNTI, with effects persisting for 48 h (Fig. 4), we can conclude that it results from the direct action of κ-opioid receptors, possibly on AVP release. Plasma levels of AVP were not measured in the present study; however, the fact that the effects of U-50488 on both UOsm and UOsm/POsm were different at 1 wk compared with at 6 wk provides evidence to support the postulate that the AVP response to U-50488H might also be age dependent. In addition, we previously observed that administration of U-50488H was in fact associated with age-dependent alterations in systemic and renal hemodynamics (34), which could impact the filtration fraction. In the present experiments, however, glomerular ultrafiltration, renal plasma flow, and filtration fraction were not measured.
κ-Opiate receptors exist on hypothalamic and neurohypophysial tissue, and dynorphin is colocalized with AVP within neurosecretory vesicles. Rossi et al. (37) showed in hypothalamo-neurohypophysial explants of Long-Evans rats that AVP gene expression and osmotically induced AVP secretion were inhibited by the κ-opiate-receptor agonist RU-51599 at the level of the hypothalamus. In neurosecretosomes removed from Wistar rat pituitaries, Zhao et al. (50) showed that U-50488H and dynorphin A1–13 inhibited AVP secretion evoked by direct depolarization of the nerve terminals, thus providing evidence that κ-opiate receptors are located on the nerve terminals of magnocellular neurons. In hypothalamic neurons in culture derived from 70-day-gestation ovine fetuses, AVP was secreted basally and in response to K+-induced depolarizations (5). Dynorphin also inhibited basal and K+-stimulated AVP release, demonstrating that the negative feedback control of AVP by activation of κ-opioid receptors is intact before birth (5) in the developing ovine.
A role for κ-opioid peptides in influencing tubular electrolyte reabsorption in the adult has also been implicated, although the literature in this regard is inconsistent. Huidobro-Toro and Parada (13) demonstrated a dose-dependent antinatriuresis after intraperitoneal administration of U-50488H to hydrated conscious rats but no consistent antikaliuretic effect. In the isolated perfused rat kidney, however, administration of the κ-opioid agonist ketocyclazocine had no effect on Na+ or K+ excretion rates (8). Kapusta and Obih (18) showed that U-50488H administered intracerebroventricularly elicited a decrease in Na+ excretion, which occurred concomitantly with an increase in renal sympathetic nerve activity; this antinatriuresis was abolished in the absence of renal sympathetic nerves. Ashton et al. (1) also measured an antinatriuresis and antikaliuresis effect and an increase in GFR after subcutaneous injection of U-50488H in inactin-anesthetized rats. In contrast, an increase in Na+ excretion occurred after intravenous administration of U-50488H to pentobarbital sodium-anesthetized dogs (42), whereas administration of the stable dynorphin analog E-2078 to conscious rats elicited no change in Na+ excretion yet K+ excretion decreased (10). In the present experiments, both age groups of conscious lambs appeared to have a low, albeit consistent, plasma K+ concentration. In previous studies carried out in conscious lambs, our group reported plasma K+ concentrations of 3.6 (SD 0.3) (30), 3.8 (SD 0.4) (38), and 3.8 (SD 0.3) (44). We do not place any significance on the slightly lower levels measured in the present study [3.3 (SD 0.3) at 1 wk and 3.3 (SD 0.6) at 6 wk] other than a variability between different populations of animals. Both Na+ and Cl− levels and POsm values were within the normal range.
Although, initially, opiate-like activity was identified in kidney tissues (12, 29), it is now apparent that there are considerable species differences in the intrarenal localization of specific opioid receptors (35, 39). Radioligand binding studies performed with kidney homogenates and autoradiographic studies revealed no opioid binding sites in the rat kidney, along with a definitive lack of renal μ- or κ-opioid receptors in the guinea pig kidney (6). A homogeneous δ-opioid-receptor population was, however, identified in the guinea pig kidney, primarily in the region of the corticomedullary junction (6). In the opossum kidney cell line, a model of proximal tubular cells, a high-affinity κ1-opioid receptor was identified (11). This is now known as the KOR1 receptor and is believed to be the receptor through which the endogenous ligand dynorphin A and the structurally similar nociceptin elicit their physiological effects (28). In the opossum kidney cell line, κ-opioids do not appear to modulate Na+-PO43− cotransport but do inhibit cellular proliferation in a dose-dependent manner (11). In addition, gene expression of endorphin precursors proenkephalin A and B is detectable in the kidney of rat pups but not in adult rats (19, 20) and in piglets but not in adult pigs (32–33), providing evidence that opioid receptor distribution may be age dependent. The distribution of κ-opioid receptors in the kidney during the perinatal period, including that of the ovine kidney, is not yet known.
In conclusion, the present study provides new information that, under physiological conditions in conscious newborn animals, activation of κ-opioid receptors results in an aquaresis, as previously observed in adult animals and humans. Additional studies are required to further elucidate the role of AVP in initiating the observed aquaretic response to κ-opioid-receptor activation during postnatal maturation.
This work was supported by an Operating Grant provided by the Canadian Institutes for Health Research. During the tenure of these experiments, Francine G. Smith was a Heritage Medical Senior Scholar supported by the Alberta Heritage Foundation for Medical Research.
The authors gratefully acknowledge the excellent assistance provided by Lucy Yu.
A portion of this work was presented in poster format at the Experimental Biology 2002 annual meeting (Qi W, Smith FG, FASEB J 15: A446, 2002).
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.
- Copyright © 2007 the American Physiological Society