INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

RENAL SODIUM EXCRETION IS generally considered to be regulated mainly via changes in body fluid volume. However, a number of animal studies indicate that stimulation of osmoreceptors may induce natriuresis by a mechanism distinct from the volume-dependent regulation. Experimental results from rats (17), sheep (7, 23), and dogs (12) indicate a larger natriuresis when a sodium load is administered as a hypertonic solution than when the same amount of sodium is given as an isotonic solution. Additionally, increased sodium excretion in response to water restriction has been shown in several animal species (16, 22-25, 35, 38). During dehydration the natriuresis is mediated by an increase in plasma tonicity that is able to override the antinatriuretic signal arising from the concomitant volume depletion. The natriuresis after infusion of hypertonic saline and dehydration could arise from stimulation of specific osmoreceptors involved in sodium homeostasis.

Results of studies in cats (34), sheep (5), and dogs (10, 39) indicate that infusion of hypertonic saline into the common carotid arteries elicits a natriuresis. It has therefore been argued that central osmoreceptors located within the area supplied by these arteries are responsible for the osmoregulatory control of sodium balance. Natriuresis induced by stimulation of these osmoreceptors is probably mediated by a humoral substance, since renal denervation does not abolish the renal response to intracarotid administration of hypertonic saline (11).

Although the evidence for a contribution of osmoregulatory control of sodium excretion from animal studies is substantial, osmoregulatory control of sodium balance is poorly elucidated in humans. The purpose of the present study was to test whether osmostimulation can induce natriuresis in humans. Identical amounts of sodium were infused as a hypertonic solution (stimulation of osmoreceptors plus volume receptors/baroreceptors) or as an isotonic solution (stimulation of volume receptors/baroreceptors only). Experiments were conducted after 4 days of low or high salt intake to further investigate the role of the sodium status of the subjects.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Experiments were performed in six healthy men. Subjects were 23-29 yr old and weighed 64.7-91.2 kg. All gave informed consent, and the study was approved by the Ethics Committee of Copenhagen (j.nr. KF 01-040/95).

Investigations were performed after 4 days of a relatively low dietary sodium intake of <75 mmol NaCl/day (LoNa⁺) or a high sodium intake of 300 mmol NaCl/day (HiNa⁺). On the night before the experiment the subject slept at the laboratory. At 0700 he was awakened and consumed a light, standardized, low-salt breakfast (3 slices of toast, 15 g of marmalade, and 250 ml of low-fat milk).

At 0800 the subject was placed in the supine position and instrumented with catheters. After local anesthesia with 2% lidocaine, a 70-cm catheter (Cavafix Certo, B. Braun Melsungen) was introduced percutaneously through a cubital vein until the tip of the catheter was positioned in the superior vena cava. This catheter was used for recording of central venous pressure (CVP) and blood sampling. The position of the catheter was confirmed by typical pressure waveforms and by characteristic responses to respiratory maneuvers. A short catheter (Venflon, 18 gauge) was inserted into a superficial vein of the opposite arm for infusion of saline. After the insertion of the catheters the subject emptied his bladder and was weighed. Subsequently, he was placed in an armchair with a seat height of 47 cm and a back angle of ~65°. The subject kept his feet on the floor and was allowed to stand only for micturition. The subject was given an oral water load of 10 ml/kg body wt of tap water preheated to 37°C. The subject drank water in amounts equal to those voided plus 1 ml/min as replacement for the insensible water loss to maintain the degree of hydration throughout the experiment. Room temperature was kept at ~24°C. The measurements were started 90 min after hydration.

Each experiment lasted 5 h and was divided into two control periods of 30 min, four infusion periods of 30 min, and four postinfusion periods of 30 min. Two infusion experiments plus a time control were performed on separate days after LoNa⁺ and HiNa⁺. At least 2 wk of recovery were allowed between each experiment. In the infusion experiments, identical amounts of sodium (3.08 mmol/kg body wt) were infused over 120 min as 0.9% saline (Isot) or as 5% saline (Hypr). Thus, during Isot and Hypr, 20 and 3.5 ml/kg body wt, respectively, were infused. Infusions were administered by an automatic infusion pump (LifeCare, model 4, Abbott). Time control series without saline infusion were otherwise identical to infusion series.

Urine was collected by voluntary micturition during the last minute of each period. After each urine collection the urinary concentration of sodium was measured. The excreted amount of sodium was replaced by intravenous injection of isotonic saline in the beginning of the subsequent period to maintain the sodium load. The injected volume was subtracted from the volume of water administered orally so that water balance also remained unchanged.

Blood samples were obtained from the central venous catheter, and the withdrawn blood was replaced by twice the amount of saline. In the middle of each urine sampling period, 10 ml of blood were sampled in heparinized tubes (15 IU/ml blood) for measurements of sodium, osmolality, Hb, hematocrit, and oncotic pressure. During periods 2, 4, 6, and 10, additional blood samples for hormone analyses (20 ml) were collected in prechilled polyethylene tubes containing aprotinin (300 kallikrein-inactivating units/ml) and EDTA (3 µmol/ml). The blood samples were centrifuged at +4°C, and plasma was stored at 18°C until later analysis for concentrations of ANG II, atrial natriuretic peptide (ANP), arginine vasopressin (AVP), and oxytocin.

Hematocrit was determined by centrifugation (Microfuge, Christ), and Hb concentrations were measured spectrophotometrically. Hematocrit and Hb concentrations were used to calculate relative changes in plasma volume by the method described by Harrison (15). Oncotic pressure was determined by a colloid osmometer (model 4400, Wescor). Urine and plasma sodium concentrations were measured by flame photometry (model 343, Instrumentation Laboratories, Lexington, MA) and osmolality by freezing-point depression (model 3DII, Advanced Instruments, Needham Heights, MA). Concentrations of creatinine were measured by conventional spectrophotometry using the method of Bonsnes and Taussky (6).

Arterial systolic and diastolic pressures were recorded semiautomatically four times during each experimental period (Propaq 102, Protocol Systems), and the means of these values were used to calculate mean arterial pressure (MAP) as follows: MAP = diastolic pressure +  × pulse pressure.

CVP was measured through the central venous catheter connected by a 200-cm fluid-filled tube to a pressure transducer (model P23XL, Spectramed) placed at the level of the fourth intercostal space and the midclavicular line. Placement of the transducer was controlled immediately before each recording. The transducer was connected to an amplifier (strain indicator CA660, Peekel) and a strip chart recorder (model ES 1000, Gould). CVP was measured in the middle of each period over 1 min. Similarly, heart rate was determined from electrocardiogram recordings and presented as the mean value of recordings over 1 min.

Hormone concentrations in plasma and urine samples were quantified by RIA after extraction. Thawed samples were acidified with 4% acetic acid, and peptides were extracted by use of C₁₈ Sep-Pak cartridges (Waters, Milford, MA), as previously described (9). After elution, the samples were dried and stored at 18°C in tubes topped with N₂ until analysis for hormone immunoreactivity. AVP was quantified as previously described (9) with the use of an antibody (Ab 3096) recently produced in our laboratory (final dilution 1:800,000). Detection limit was 0.22 pg/ml sample. Intra- and interassay coefficients of variation were 9.3 and 7.9%, respectively. Recovery of unlabeled AVP added to plasma was 71%. Immunoreactivity of ANP was determined according to the assay procedure described earlier (30) by use of an antibody (RAS-8798) purchased from Peninsula Laboratories. Detection limit was 1.5 pg/ml sample. Intra-assay coefficient of variation was 6%, and recovery was 85%. Immunoreactivity of ANG II was determined by use of an antibody (Ab 5-030682) produced in our laboratory (final dilution 1:100,000). The assay procedure was described by Kappelgaard et al. (20). Detection limit was 1.4 pg/ml sample. Intra-assay coefficient of variation and recovery were 5 and 83%, respectively. Concentrations of oxytocin were determined as described by Engstrøm and Vilhardt (13). Sensitivity of the assay was 0.2 pg/ml sample. Intra- and interassay coefficients of variation were 7.8 and 7.7%, respectively. The procedure for measuring endothelin-1 concentrations in extracted urine samples has recently been described (10). The detection limit was 0.16 pg/ml urine, and extraction recovery was 91%. Intra- and interassay coefficients of variation were 5.2 and 6.5%, respectively. Urodilatin was measured by an assay procedure recently developed in our laboratory. Briefly, an antibody (S-1969) kindly donated by Biomedica (Vienna, Austria) was used in a final dilution of 1:625,000. Cross-reactivities with human ANP-(99126), human ANP-(428), rat ANP, endothelin-1, AVP, and ANG II were <0.001%. After Sep-Pak extraction, samples were incubated with antibody for 24 h and then incubated with tracer (¹²⁵I-urodilatin) for another 48 h. Urodilatin immunoreactivity was measured on the supernatant after sedimentation of free antigen by charcoal-plasma suspension and centrifugation. Detection limit was 0.32 pg/ml urine. Intra-assay coefficient of variation and extraction recovery of unlabeled urodilatin added to urine samples were 10 and 85%, respectively. The results have not been corrected for incomplete recovery.

Statistics. Values are means ± SE. Data were subjected to one-way ANOVA for repeated measures (37). In case of significantly large F values, all possible differences were evaluated by a sequential variant of the Studentized range method (Newman-Keuls test). Differences between series were evaluated using paired Student's t-test. In all cases, level of significance was 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

The effects of high and low sodium intake on baseline conditions of the subjects are presented in Table 1. Twenty-four-hour sodium excretion was approximately four times higher during the last day of HiNa⁺ than LoNa⁺. On the day of the experiment, hematocrit, Hb, and oncotic pressure were significantly lower with the HiNa⁺ than with the LoNa⁺ diet. Body weight, plasma sodium concentration, and osmolality were not significantly altered by changes in dietary sodium intake. CVP was 2.5 ± 0.7 and 1.3 ± 0.2 mmHg after LoNa⁺ and HiNa⁺, respectively. However, these values were not statistically different. Basal creatinine clearance was increased after HiNa⁺ compared with LoNa⁺ (157 ± 5 vs. 140 ± 3 ml/min). The HiNa⁺ diet markedly suppressed plasma ANG II (78 ± 6%) and increased plasma ANP (+64 ± 20%). Finally, the rate of urinary excretion of urodilatin was significantly higher after HiNa⁺ than LoNa⁺ (5.7 ± 0.7 vs. 4.7 ± 0.6 ng/24 h).

During the acute infusions, plasma sodium concentration and osmolality increased in response to Hypr but remained unchanged during Isot (Table 2). Changes in plasma sodium are presented in Fig. 1. The increases in plasma sodium generated by Hypr after LoNa⁺ and HiNa⁺ were similar (3.8 ± 0.5 and 3.5 ± 0.4 mmol/l, respectively). Accordingly, Hypr increased plasma osmolality by 9 ± 1 and 8 ± 1 mosmol/kgH₂O after LoNa⁺ and HiNa⁺, respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Change in plasma sodium concentration. , Control after low-salt diet; , isotonic saline infusion after low-salt diet; , hypertonic saline after low-salt diet; , control after high-salt diet; , isotonic saline infusion after high-salt diet; , hypertonic saline after high-salt diet. Values are means ± SE. * Significantly different from both preinfusion values, i.e., 1st 2 periods of same series, by ANOVA (P < 0.05) and Newman-Keuls test.

CVP increased in response to all loading procedures but most consistently during Isot (Fig. 2). After LoNa⁺ and HiNa⁺, Isot produced gradual increases in CVP that reached statistical significance 75 min after the start of the infusion. Maximum responses were observed at the end of the infusion (2.2 ± 0.2 and 2.4 ± 0.6 mmHg for LoNa⁺ for HiNa⁺, respectively). During Hypr, statistically smaller increases in CVP were observed (maximum increase 1.2 ± 0.3 and 1.0 ± 0.5 mmHg for LoNa⁺ and HiNa⁺, respectively).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Change in central venous pressure (CVP, A) and percent change in plasma volume calculated from measured hematocrit and Hb concentration (B). See Fig. 1 legend for explanation of symbols. Values are means ± SE. * Significantly different from preinfusion values by ANOVA (P < 0.05) and Newman-Keuls test. § Significantly different from corresponding value during hypertonic infusion by Student's t-test (P < 0.05).

Relative changes in plasma volume were calculated from hematocrit and Hb concentrations. All loadings induced increases in plasma volume, but Isot increased plasma volume significantly more than Hypr (Fig. 2). Oncotic pressure decreased in response to all loading procedures but was unchanged in the time control series (Table 2). Again, maximal decreases tended to be more pronounced during Isot (4.6 ± 0.4 and 4.1 ± 0.3 mmHg for LoNa⁺ and HiNa⁺, respectively) than during Hypr (3.6 ± 0.2 and 3.4 ± 0.3 mmHg for LoNa⁺ and HiNa⁺, respectively), but in the case of oncotic pressure this was not statistically significant.

After LoNa⁺, Isot resulted in a gradual increase in sodium excretion that reached statistical significance in the last period of infusion (Fig. 3). The largest response was observed in the last period, where excretion was more than tripled compared with the preinfusion value (from 42 ± 5 to 152 ± 17 µmol/min). Infusion of the same amount of sodium as a hypertonic solution also increased renal sodium excretion, but the response was delayed, and by the end of the observation, excretion rate had increased from 50 ± 7 to 99 ± 12 µmol/min. This was significantly smaller than the response to Isot. After HiNa⁺, Isot caused a gradual increase in sodium excretion that seemed to stabilize in the postinfusion phase. In absolute values, excretion increased from 177 ± 28 to 318 ± 11 µmol/min. During Hypr, sodium excretion showed a very similar pattern. Excretion rates increased from 179 ± 31 to 318 ± 42 µmol/min.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Renal sodium excretion and urine flow. See Fig. 1 legend for explanation of symbols. Values are means ± SE. * Significantly different from both preinfusion values, i.e., 1st 2 periods of same series, by ANOVA (P < 0.05) and Newmans-Keuls test. § Significantly different from corresponding value during hypertonic infusion by Student's t-test (P < 0.05).

Urine flow remained unchanged during time control and Isot, except for a small increase after Isot in the LoNa⁺ series (Fig. 3). Hypr produced a clear antidiuresis from 8.8 ± 1.3 to 0.7 ± 0.2 ml/min and from 9.7 ± 2.0 to 1.5 ± 0.2 ml/min after LoNa⁺ and HiNa⁺, respectively. This antidiuresis persisted throughout the recovery period (Fig. 3).

The blood pressures and heart rate were unchanged in all experimental series (Table 3). Heart rate tended to decrease with salt loading, but this trend did not reach statistical significance. Creatinine clearance remained unchanged within all experimental series (Fig. 4), although basal levels were increased during high sodium intake (see above).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Urinary creatinine clearance (Clcreatine). See Fig. 1 legend for explanation of symbols. Values are means ± SE.

Basal levels of ANG II were significantly higher after LoNa⁺ than after HiNa⁺, and plasma levels decreased in response to all saline infusions (Fig. 5). Isot and Hypr decreased plasma ANG II by 51 ± 9 and 23 ± 13%, respectively, after LoNa⁺ and by 52 ± 16 and 40 ± 8%, respectively, after HiNa⁺. Plasma AVP concentrations doubled during Hypr (Table 4). After LoNa⁺, Isot significantly decreased plasma AVP in the last period of infusion. In contrast, there were no changes in plasma concentrations of ANP or oxytocin in response to any of the saline infusions (Table 4). Urinary excretion rates of endothelin-1 and urodilatin were not significantly changed in any of the series (Table 5).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Plasma ANG II concentration. See Fig. 1 legend for explanation of symbols. Values are means ± SE. * Significantly different from preinfusion value of same series by ANOVA (P < 0.05) and Newman-Keuls test.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The purpose of this study is to elucidate the role of osmoreceptors in the natriuretic response to acute salt loading. Data from the time control series show that the seated position provides a steady-state situation with regard to hemodynamic, endocrine, and renal excretory variables.

Osmostimulation and natriuresis. Isot induced a selective increase in body fluid volume, whereas Hypr increased plasma tonicity as well as extracellular volume. To evaluate the possible contribution of osmoreceptors in the natriuretic response to Hypr, the simultaneous volumetric stimulus should be quantified and compared with that produced by Isot. The overall changes in plasma sodium concentration and CVP after the start of saline infusion in subjects fed the high-sodium diet are presented in Fig. 6, together with the cumulated sodium excretion. These results strongly indicate that Isot caused a stronger stimulus to volume receptors than Hypr; this finding is also supported by the measurements of oncotic pressure and the calculated change in plasma volume. If renal sodium excretion is controlled primarily by the load to volume receptors, the natriuretic response to Isot should exceed the response to Hypr. This was actually the case after LoNa⁺. However, as presented in Figs. 3 and 6, the natriuretic responses to Isot vs. Hypr were very similar after HiNa⁺, despite the stronger volumetric stimulus after Isot. Thus it is likely that, during Hypr, part of the natriuresis was induced by osmostimulation. We hypothesize that the increase in plasma sodium concentration stimulated sodium excretion sufficiently to counterbalance the smaller increase in volume. That stimulation of osmoreceptors is able to induce natriuresis only in subjects fed the high-sodium diet is supported by earlier findings in dogs in which the natriuretic response to intracarotid infusion of hypertonic saline was inhibited after sodium depletion (14).

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Mean change in plasma sodium concentration (A) and CVP (B) from beginning of saline infusion throughout observation period in subjects fed high-sodium diet. Isot, 0.9% saline; Hypr, 5% saline. C: cumulated sodium excretion over 5 h. * Significantly different from corresponding value during isotonic infusion by Student's t-test (P < 0.05).

Natriuretic mediators. Natriuresis may be mediated by changes in endocrine factors, renal sympathetic nerve traffic, or physical factors. MAP was remarkably constant in all experimental series. Nor did urinary creatinine clearance change significantly during any of the saline infusions. However, it should be noted that creatinine clearance is not a very precise index of glomerular filtration rate in humans, and even if glomerular filtration rate remained unchanged, the filtered sodium load increased in the hypertonic experiments because of the increase in plasma sodium. Although this increase was modest (~2.5%), it cannot be excluded that part of the natriuresis after hypertonic saline loading could be explained by increased filtration of sodium.

1

1

Summary and conclusion. Osmoregulatory control of renal sodium seems to be present in subjects with a high sodium intake. Isot and Hypr produced identical natriuretic responses after HiNa⁺, despite a significantly stronger volumetric stimulus after Isot, and we suggest that stimulation of osmoreceptors accounted for a part of the sodium excretion after the combination of high sodium intake and hypertonic saline infusion. However, osmostimulated natriuresis could not be demonstrated in subjects with a low sodium intake. Furthermore, the sensitivity of osmoregulatory control of sodium excretion may not be as high as in other species. In all series the natriuresis was preceded by a decrease in plasma concentration of ANG II, and we conclude that suppression of ANG II was responsible for at least part of the natriuretic response to Isot as well as Hypr. Whether an additional humoral component was involved in the natriuresis after Hypr is uncertain, but plasma ANP, plasma oxytocin, and urinary excretion rates of urodilatin and endothelin-1 remained unchanged.

Perspectives