Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Thiazide (TZ) diuretics exert an antidiuretic action in central diabetes insipidus (CDI) (Crawford and Kennedy, 1959) as well as in nephrogenic diabetes insipidus (NDI) (Forrest et al., 1974). The exact mechanism behind this "paradoxical" antidiuretic response remains elusive. The major hypothesis proposed by several investigators and cited in authoritative textbooks (Ives, 2001) suggests that TZ-induced contraction of the extracellular fluid volume leads to a baroreceptor-mediated reduction in the flow of tubular fluid output delivered to the diluting segment (distal delivery) and hence production of reduced amounts of a more concentrated urine. This is supported by studies showing that TZ treatment reduces distal delivery in normal and CDI rats as measured by micropuncture (Weinman and Eknoyan, 1975; Shirley et al., 1982; Walter and Shirley, 1986) or indirectly by lithium clearance (Thomsen and Schou, 1973; Petersen et al., 1974; Lunau et al., 1994). Furthermore, in anesthetized CDI rats sodium replacement prevented the antidiuresis as well as the reduction in distal delivery induced by hydrochlorothiazide (HCTZ) (Shirley et al., 1978; Walter et al., 1979).

However, a number of observations in Brattleboro rats with CDI are not entirely consistent with the above-mentioned hypothesis: 1) sodium depletion induced by sodium restriction alone is not accompanied by antidiuresis, and the antidiuresis induced by HCTZ administration ceases when the drug is withdrawn despite sustained sodium depletion (Walter and Shirley, 1983); 2) the antidiuresis induced by chronic bendroflumethiazide (BFTZ) administration is not related to changes in lithium clearance (Grønbeck et al., 1998); and 3) in our laboratory BFTZ-induced antidiuresis in conscious diabetes insipidus rats could not be prevented when sodium losses were replaced by use of a computerized servo-control system (Spannow et al., 1997). For these reasons we concluded that at least in rats with CDI, sodium depletion cannot be the sole mechanism eliciting TZ-induced antidiuresis, i.e., that thiazides must exert some additional effects on distal nephron water reabsorption (Grønbeck et al., 1998). In line with this, it was recently reported that HCTZ added to the perfusion fluid in vitro stimulated water permeability in inner medullary collecting ducts isolated from Brattleboro rats (Cesar and Magaldi, 1999). However, since BFTZ treatment did not alter renal expression or collecting duct intracellular distribution of AQP2 water channels, it seems unlikely that thiazides should exert an antidiuretic hormone-like action on the collecting duct cells (Grønbeck et al., 1998).

Whereas most studies on the mechanism of TZ-induced antidiuresis have been carried out in Brattleboro rats with CDI, the major clinical application of thiazides as antidiuretics is in patients with NDI, which can either be congenital (Bichet et al., 1997) or induced by drugs such as lithium (Jackson and Dousa, 1990). In rats with lithium-induced NDI, BFTZ treatment (10 mg/kg/day) caused a two-thirds reduction in urine flow associated with generation of slightly hypertonic urine (Christensen, 1976). A similar antidiuretic response was observed when chlorothiazide was administered to patients with lithium-induced polyuria (Forrest et al., 1974). Nevertheless, so far no study has addressed the mechanism of the antidiuretic response to thiazides in NDI.

The objective of the present study was to investigate the role of sodium depletion for TZ-induced antidiuresis in NDI. We used two different animal models: rats with NDI induced by chronic lithium administration, and rats with NDI induced by acute i.v. infusion of the vasopressin V₂ receptor antagonist OPC-31260. For comparison, we also examined the response of TZ in rats undergoing water diuresis. Clearance experiments were conducted in conscious rats instrumented with permanent catheters, and the renal hemodynamic and excretory responses to i.v. infusion of BFTZ (0.25 mg/h) were assessed with and without replacement of urinary sodium losses.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Animals and Physical Environment

Female Wistar rats [Crl:(WI)BR; 210-230 g] were purchased from Charles River, Sulzfeld, Germany. The animals were housed in a temperature- (22-24°C) and moisture (40-70%)-controlled room with a 12-h light/dark cycle (light on from 6:00 AM to 6:00 PM). All animals were given free access to tap water and a pelleted rat diet containing approximately 140 mmol/kg of sodium, 275 mmol/kg of potassium, and 23% protein (catalog no. 1310; Altromin International, Lage, Germany).

Animal Preparation

One week before the renal clearance experiments all rats were anesthetized with 1% halothane in a gas mixture of N₂O/O₂ (2 + 1) and chronic medical grade Tygon catheters were implanted into the abdominal aorta and into the inferior caval vein via a femoral artery and vein. A silicone-coated stainless steel chronic suprapubic bladder catheter was implanted into the urinary bladder and sealed with a screw after flushing the bladder with 0.6 mg/ml ampicillin (Anhypen; Nycomed Pharma, Oslo, Norway). Catheters were produced, fixed, and sealed as previously described (Petersen et al., 1991). All surgical procedures were performed during aseptic conditions. To relieve postoperative pain, rats were treated with 0.2 mg/kg of b.wt. i.p. buprenorfin, twice daily for 2 days (Anorfin; GEA A/S, Copenhagen, Denmark). After instrumentation, the animals were housed individually.

Rats with NDI Induced by Lithium Treatment

The rats were fed lithium-containing rat pellets containing 40 mmol/kg of lithium for the first 2 weeks, thereafter 60 mmol/kg. Lithium was added as lithium citrate. The rats were given tap water ad libitum and free access to solid sodium chloride to prevent lithium toxicity (Thomsen et al., 1974). The treatment period was 4 to 5 weeks to induce maximal polyuria and impairment of renal concentrating ability (Christensen et al., 1982). Daily water intake was measured as an index of polyuria.

Rats with NDI Induced by Acute V₂ Receptor Blockade

In this model, polyuria was induced by i.v. infusion of the V₂ receptor antagonist OPC-31260 at a rate of 0.2 mg/h from the beginning of the clearance experiment (Jonassen et al., 1998).

Rats with Water Diuresis

The rats were infused with 150 mM glucose at a rate of 8 ml/h from the beginning of the clearance experiment.

Renal Clearance Technique

The renal hemodynamic and tubular response to BFTZ was examined by clearance techniques in conscious rats. Prior to the clearance experiments all rats were adapted to the restraining cage used for these experiments by training them for two periods of 2 h each on separate days. Clearance of [¹⁴C]tetraethylammonium bromide was used as a marker for the effective renal plasma flow (Rasmussen et al., 1990), clearance of [³H]inulin as a marker for glomerular filtration rate (Shalmi et al., 1991) and clearance of lithium as a marker for distal delivery (Thomsen and Shirley, 1997). Whereas the former substances were administered in tracer amounts by constant i.v. infusion, lithium was added in small amounts (12 mmol/kg) to the diet of control rats for 2 to 3 days prior to experiments to avoid acute effects on renal function (Shalmi and Thomsen, 1989).

Clearance experiments were started at 8:00 AM and clearance markers were infused at a constant rate (1.5 ml/h) throughout experiments. A 90-min equilibration period was followed by 2 × 30-min control periods and 6 × 30-min experimental periods. BFTZ was infused i.v. with a constant rate of 0.25 mg/h (0.5 ml/h). All infusions were made through the venous catheter. The arterial catheter was used for blood samples (300 µl each being immediately replaced with heparinized donor blood) and for monitoring of mean arterial blood pressure during experiments.

Study Design and Clearance Protocols

For both models of NDI, three separate subgroups were investigated.

Time-Control Subgroup. Lithium-pretreated rats were given constant infusion of 150 mM glucose (8 ml/h) during equilibration and control periods and switched to servo-controlled replacement of urinary volume losses during the rest of the experiment. In the group receiving acute administration of V₂ antagonist, servo-controlled volume replacement with 150 mM glucose was on from the beginning of the experiment and maintained throughout. The water-loaded group received constant infusion of 150 mM glucose (8 ml/h) during equilibration and control periods and were switched servo-controlled replacement of urinary volume losses during the rest of the experiment.

BFTZ-Treated Subgroup. These rats were treated like the time-control rats but at the end of the second control period, BFTZ infusion was started.

BFTZ-Treated Subgroup with Sodium Replacement. This subgroup was treated like the BFTZ rats, but in addition urinary sodium losses were replaced (Spannow et al., 1997) by a servo system during BFTZ infusion.

Analytical Methods

Electrolytes in samples of plasma and urine were determined by atomic absorption spectrometry, using a PerkinElmer series 2380 or a PerkinElmer Analyst 300 atomic absorption spectrometer. [³H]Inulin and [¹⁴C]tetraethylammonium were determined by double label scintillation counting in a Packard 2250 CA liquid scintillation counter. Urinary osmolality was determined by freezing point depression, using an Advanced model 3CII freezing depression osmometer.

Drugs and Chemicals

Bendroflumethiazide was a gift from Leo Pharmaceuticals, Copenhagen, Denmark. It was dissolved in 150 mM glucose (0.5 mg/ml) by addition of 0.1% ethanolamine. OPC-31260 was a gift from Otsuka Pharmaceutical Co., Tokushima, Japan. It was dissolved in 150 mM glucose (0.2 mg/ml). [³H]Inulin was purchased from Amersham (Hillerod, Denmark) (catalog no. TRA324). [¹⁴C]Tetraethylammonium bromide was purchased from Dupharma A/S (Kastrup, Denmark) (catalog no. NEC298). Other chemicals were of analytical grade from standard suppliers.

Calculations

Renal clearances (C) and fractional excretions (FE) of the clearance markers were calculated by the standard formula C = U · V/P; FE = C/GFR, where V is urine flow rate, U is concentration in urine, and P is plasma concentration. Lithium clearance was used as an index of distal delivery (Thomsen and Shirley, 1997). Thus, V/C_Li signifies the fraction of distal delivery being excreted as urine (fractional distal water excretion).

Data Presentation and Statistics

All values are presented as means ± S.E.M. Overall statistical comparisons were performed by one-way analysis of variance (between groups), one-way analysis of variance for repeated measures (within group), or two-way analysis of variance for repeated measures for two-way classified data (group and time). Subsequently, the effect of the individual treatments was evaluated by comparison of the last control period with the average of the two last experimental periods using Student's paired t test. Differences were considered statistically significant at the 0.05 level.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

In the sodium-replaced groups the cumulated sodium balances during the 3 h with BFTZ infusion were 23 ± 14 µmol in the lithium rats and +5 ± 26 µmol in the rats with V₂ receptor blockade, indicating that sodium losses were completely compensated for by the servo-control system.

Rats with NDI Induced by Chronic Lithium Administration (Fig. 1). During the days before the clearance experiment, rats treated with lithium for 4 to 5 weeks showed polyuria corresponding to a daily water consumption of 70 to 100% of their body weight. The plasma lithium concentration was 0.54 ± 0.06 mmol/l. During experiments, the urine flow rate was maintained by infusion of 150 mM glucose solution (8 ml/h, corresponding to 85% of the body weight per 24 h) in equilibration and control periods, and thereafter at a rate controlled by the volume servo. As seen from Fig. 1, the urine flow rate remained stable throughout in the time control group when the volume servo was activated.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Studies in rats with NDI induced by chronic lithium administration. Urinary sodium excretion (UNaV), V, urine osmolality (Uosm), CLi, urine flow as percentage of distal delivery (V/CLi), ERPF, and GFR during two control periods (0-60 min) and during infusion of vehicle or BFTZ with or without servo control of the sodium balance (60-240 min). Control period 2 versus two last experimental periods (paired Student's t test): *p < 0.05, **p < 0.01, ***p < 0.001.

Rats with NDI Induced by Acute V₂ Receptor Blockade (Fig. 2). In this model, polyuria was induced by i.v. infusion of the V₂ receptor antagonist OPC-31260 along with servo-controlled replacement of urinary volume losses with 150 mM glucose solution. As reported previously (Jonassen et al., 1998), this procedure induced a urine flow rate of about 100 µl/min/100 g of b.wt., which in the time control group remained stable throughout.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Studies in rats with NDI induced by acute V2 receptor blockade. The renal variables and treatments are the same as described in legend to Fig. 1.

Rats with Water Diuresis (Fig. 3). Water diuresis was induced by infusion of 150 mM glucose solution (8 ml/h for 150 min) and maintained by the volume servo. In the time control group, the urine flow rate was stabilized at about 80 µl/min/100 g of b.wt. with a urine osmolality of about 100 mosmol/kg. BFTZ induced natriuresis without significant changes in urine flow rate, leading to an increase in urine osmolality. EPRF, GFR, and C_Li remained unaltered whereas V/C_Li decreased. Because BFTZ did not cause antidiuresis in water diuretic rats, the group with sodium replacement was omitted.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Studies in rats with diabetes insipidus due to infusion of fluid (water diuresis). The renal variables and treatments are the same as described in legend to Fig. 1 with the only exception that a BFTZ + Na group was omitted because BFTZ did not cause antidiuresis.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

In this article we investigated the role of sodium depletion for the acute antidiuresis induced by BFTZ in two different rat models of NDI, using computer-controlled independent replacements of urinary losses of sodium and water. The major new observation is that in NDI, whether induced by lithium or acute V₂ receptor blockade, BFTZ-induced antidiuresis can be completely prevented by sodium replacement. This contrasts our recent findings in rats with CDI (Spannow et al., 1997) using an identical protocol.

Mechanism of Antidiuresis Induced by BFTZ in Rats with NDI. In this study we chose two different models of drug-induced NDI: long-term lithium administration known to inhibit vasopressin V₂ receptor-stimulated cyclic AMP formation in the collecting duct principal cells (Christensen et al., 1985), and acute blockade of V₂ receptors with the selective vasopressin antagonist OPC-31260. In both models, polyuria is due to inability of vasopressin to stimulate insertion of AQP2 water channels in the luminal collecting duct membrane (Marples et al., 1995; Christensen et al., 1998).

Response to BFTZ in Rats with Water Diuresis. Rats with water diuresis induced by sustained water loading develop polyuria due to suppression of AVP release. It is well known that water diuretic individuals will not respond to thiazide administration with antidiuresis, and this was confirmed in the present study. BFTZ induced natriuresis and negative sodium balance both in water diuretic rats and in NDI rats, but distal delivery was only decreased in the latter. One major difference between diabetes insipidus and water diuresis is the state of body hydration: whereas diabetes insipidus is usually associated with negative water balance, as signaled by plasma hyperosmolality (Valtin, 1967), individuals with water diuresis are per definition overhydrated. Thus, the explanation for the absence of an antidiuretic response to thiazides in water diuretic rats could simply be that the overhydration prevents TZ-induced contraction of the intravascular volume, and thereby the baroreceptor-mediated reduction in distal delivery. It seems physiologically reasonable that the compensatory fall in distal delivery in response to thiazide-induced sodium depletion is more readily activated in states with volume deficiency than in conditions with volume excess.