Acts Phy~iolS c a d 1990, 140, 119-127
Increase in plasma sodium enhances natriuresis in response t o a sodium load unable t o change plasma atrial peptide concentration C. E M M E L U T H , H. J. S C H U T T E N , U . K N I G G E , J. W A R B E R G and P. B I E Department of Medical Physiology C, Panum Institute, University of Copenhagen, Denmark
C., SCHUTTEN, H. J., KNIGGE,U., WARBERG, J. & BIE,P. 1990. Increase in EMMELUTH, plasma sodium enhances natriuresis in response to a sodium load unable to change plasma atrial peptide concentration. Actu Physiol Scnnd 140, 119-127. Received 8 September 1989, accepted 29 March 1990. ISSN 0001-6772. Department of Medical Physiology C, Panum Institute, University of Copenhagen, Denmark. The influence of plasma sodium concentration in the control of sodium excretion was investigated in conscious, water-diuretic dogs. NaCl was infused for 60 min as a hypertonic or isotonic solution at a rate of 60 pmol NaCl min-' kg-' body wt. Plasma sodium concentration rose only during hypertonic infusion (P < 0.05). Sodium excretion increased markedly with both infusions (hypertonic, from 2.4 f0.6 to 105 +27 pmol min-'; isotonic, from 3.95- 1.3 to 58+ 17 pmol min-'). Fractional sodium excretion increased more during hypertonic than during isotonic infusion. Hypertonic infusion decreased diuresis from 3.1 k0.5 to 1.3f0.6 ml rnin-l, while isotonic infusion elicited an increase from 3.9 f0.5 to 7.2 f0.7 ml min-'. Plasma renin activity and plasma aldosterone decreased markedly in both series (P < 0.05), the relative changes in the two series being very similar. Central venous pressure increased (2.8 2 0.7 to 4.5 f 1.O mmHg) during isotonic infusion but not significantly during hypertonic infusion. Arterial pressure, heart rate and plasma levels of atrial natriuretic peptide and catecholamines did not change measurably in either series. It is concluded that simultaneous increases in extracellular volume and sodium concentration cause a larger natriuretic response than a change in volume alone, and that a 40-fold increase in sodium excretion may occur without measurable changes in plasma atrial natriuretic peptide concentration. Key words : atrial natriuretic peptide, catecholamines, conscious dogs, natriuresis, renin-angiotensin-aldosterone system, volume expansion.
I t is well documented that humoral, nervous and physical factors, among the last-mentioned notably renal perfusion pressure, participate in the regulation of sodium excretion. However, the relative importance of these factors is not well determined. T h e renin-angiotensin-aidosterone system is of primary importance. Angiotensin I1 influences vascular as well as tubular functions and participates in the tubuloglomerular feedback Correspondence : Claus Emmeluth MD, Department of Medical Physiology C, Blegdarnsvej 3C, DK2200 Copenhagen, Denmark.
mechanism (Navar ct al. 1987). T h e effect of aldosterone is primarily on the distal part of the nephron (Young 1985). ANP has recently been intensively studied (Lang et al. 1987, Goetz 1988), however its precise role in the regulation of renal sodium excretion is still a matter of controversy (Atlas & Laragh 1987). Proximal reabsorption and renal renin release have been shown to be influenced by adrenergic stimulation (via a- and ,&receptors respectively) (DiBona 1985), but the role of dopamine in the regulation of renal sodium excretion is still uncertain (Lee 1982, Levinson et al. 1985). T h e 'pressure natriuresis ' (Hall et ul. 1986,
119
120
C. Emmeluth et
a/.
Guqton 1987) a n d substances inhibiting Na',K-ATPase activitj- (DeWardener 1985) are other factors which should h e considered during investigations into renal sodium excretion. T h e multifactorial regulation of sodium and water excretion calls for simultaneous measurements of a n u m b e r of potentially important regulatory components under precisely defined conditions. We were interested in the roles of volume, sodium concentration and atrial natriuretic peptide in sodium excretion, and the experiments therefore included simultaneous recordings of haemodynamic, humoral and renal excretory variables in connection with transition from a state of relative excess of water t o a state of excess of water and Na'. It was considered important that modest amounts of N a C l were administered rather sloa ly t o conscious animals i n order t o mimic naturally occurring circumstances, e.g. the load of sodium provided by t h e usual diet.
14ATERIAI.S AND M E T H O D S ..?nimii/s. Experiments were performed in seven conscious, trained female dogs (Beagles) weighing 11-16 kg fed on a fixed diet of commercial dog food (Latz Komplett mixed with Latz canned food). T h e dogs received one meal a daj-, usuall>-around 14.00 h, and had free access to tap water. The intake of Naand K- was measured regularly and averaged 2.7 kO.2 and 2.5+.0.2 mmol daj- kg-' body wt respectively. Prior to the present stud!- the common carotid arteries were placed in skin loops using an aseptic technique and general anaesthesia. Prrpururion. Each dog was used for all three tl-pes of experiments with intervals between experiments of at least 1 week. Prior to the experiments the dogs were fasted for 18 h as part of the daily routine. Overnight the! did not haw access to nater ; this M-as an attempt
CHANGE I N S O D I U M C O N T E N T INFUSION
isotonic
to minimize variations between animals. A sterile catheter (Intracath) was introduced via an external jugular vein and the tip was placed in or near the right atrium for measurement of central venous pressure using a Statham P.50 transducer. The pressure profile served as an indicator of the proper position: when a ventricular pressure curve was obtained the catheter was withdrawn a few centimetres. I n addition, catheters were placed in both saphenous veins and in a common carotid artery. The venous catheters were used for blood sampling and infusion, while the arterial one was used to measure the arterial blood pressure (Statham P50). .4 modified silicone Foley catheter was inserted into the bladder. Heart rate was measured by registration of the ECG. Protocol. The dog was hydrated via a stomach tube with a volume of water equivalent to 2";) of body weight 1 h before urine collection was started. Thereafter body weight was kept constant, i.e. ~ 0 . l o , , by use of a servo mechanism (Bie 1976) infusing a hypotonic solution of glucose and urea, 40 mM and 25 mM respectively. Clearances of paraaminohippurate (PAH) and exogenous creatinine were measured as previously described (Emmeluth et al. 1987) and used as measures of renal plasma flow and glomerular filtration rate respectively. When a steadystate water diuresis was achieved, urine sampling at 10-min intervals was initiated. Thirty minutes later an infusion of isotonic or hypertonic saline was started. Three series of NaC1 infusion were carried out. Srrmuli. In one series of experiments 60pmol NaCl kg-'min-' was infused for 60 min as a hypertonic solution (780 mmol I ') at a rate of 0.08 ml kg-' min-'. Due to the servo mechanism there was no net increase in the weight of the dog and therefore the load resulted in an increase in the sodium content of the dog but no change in total fluid r-olume (Fig. I). I n another series of experiments NaCl was infused a t the same rate but in an isotonic 290 mosmol kg ') and solution (156 mmol I-', therefore in a larger volume (0.39 ml kg-' min '). The saline was placed as part of the load of the weight cell prior to infusion so that the infusion was not registered
-
C H A N G E I N WATER C O N T E N T 400.
INFUSION
isotonic
7 100 hypertonic I
.
0
30
.
,
.
.
,
60 90 120150180 Time
(min)
-1001 0
'
30
'
60
Time
'
'
'
.
90 120150180 (rnin)
Fig. 1. :I schematic illustration of the changes in (a) sodium and (b) water content during hypertonic and isotonic infusion.
Circulatory and renal responses to N a C l by the servo mechanism. In this situation both the sodium and the water contents of the dog increased, in contrast to the hypertonic infusion. The increase in the water content was linear as the response to volume expansion was compensated by the action of the servo system (Fig. 1). Control experiments included an infusion of isotonic saline at the low rate of 0.08 ml kg-' min-'. Four blood samples (14 ml each) were obtained at designated points in time during the experiments. After immediate centrifugation at 4 "C the erythrocytes were resuspended in isotonic saline, and reinfused no later than 15 min after the blood sample had been taken. Chemical analyses. Na+ and K' were measured by flame photometry (IL 243, Instrumentation Laboratory). Concentrations of PAH and creatinine in plasma and urine were measured by spectrophotometry (Spectronic 1001, Bausch & Lomb). The osmolality of plasma and urine was estimated by freezing point depression (Digimatic Osmometer Model 3DI1, Advanced Instruments).
+
121
Hormone analyses. Aldosterone was determined on dichloromethane-extracted plasma. The recovery of added [3H]aldosterone was 89 f 1yo (mean fSE, n = 10). The antiserum was a generous gift from Dr W . Vetter (Vetter et al. 1974). The immunoreactive aldosterone was determined by radioimmunoassay. The least detectable limit was 2-3 pg per tube, and half-maximal binding of [3H]aldosterone was at 20-30 pg per tube. Inter- and intra-assay coefficients of variance were 12.3 and 6.1 yo respectively. Plasma renin activity was determined by use of the antibodytrapping method described by Poulsen & J~rgensen (1974). Inter- and intra-assay coefficients of variance were 7.5 and 1.0% respectively. Catecholamines were determined by a radioenzymatic assay described by Ben-Jonathan & Porter (1976). The sensitivity of the assay was 10-20 pg mi-' and the intra-assay coefficients of variations for noradrenaline, adrenaline and dopamine were 6.8, 4.3 and 5.0% respectively. The corresponding interassay coefficients of variation were 14.8, 12.3 and 7.5% respectively (Knigge et al.
Table 1. Plasma concentration of' Na+ (p-Na+) and plasma osmolality (p-osm), mean arterial blood pressure (MABP), heart rate (HR), and central venous pressure (CVP) -
~~
Time (min) ~
15
45
85
145
p-Na' (mmol I-') Control Hypertonic Isotonic
141 f 1 142f 1 1 4 2 11
141 f 1 142f1 141 f 1
141f 1 146f 1" 141 f 1
141 f 1 144f 1" 140f 1
p-osm (mosmol kg-') Control Hypertonic Isotonic
291 1 291 -t 1 293 f 1
292 1 294+ 1 293 & 1
294f 1 301 1" 293 f 1
293 f 1 30Of 1" 294f 1
MABP (mmHg) Control Hypertonic Isotonic
102 f2 100 1 4 11629
102f 1 101+2 115f5
101f2 100f3 119k6
102f2 101f3 118f6
IIR (beats min-') Control Hypertonic Isotonic
80+3 86k4 9517
78f4 78+6 98+3
7312 86f6 107 f8
80+5 82f5 96+6
CVP (mmHg) Control Hypertonic Isotonic
3.2f 1.2 4.5 f0.7 2.8 0.7
3.3f 1.1 5.0k0.8 3.210.7
3.4f 1.2 5.0+ 1.0 4.5f 1.0"
2.8-1- 1.2 4.8 k 0.7 6.0 f 1.4"
+
+
*
NaCl was infused from min 30 to min 90. MABP and HR were measured throughout the experiments; there was no significant change from the data shown. Values significantly different from the preinfusion level are marked with an asterisk (").
122
C. Emmeluth et al.
1988). The concentration of atriul natriuretrc prptide (ANP) in plasma was measured as described by Schutten et al. (1987). Briefly, acidified plasma was extracted using C-18 Sep-Pak cartridges and immunoreacrive ANP was measured bj- a radioimmunoassay. The detection limit was less than 3 pg per tube and the half-maximal binding nas observed at 40 pg per tube. The inter- and intra-assay coefficients of variation, determined at the middle of the sensitive range (50 pg per tube) of the standard curve, were 5.6 and 4.0°,, respectively.. Crossreactivity with synthetic rat a-atrial natriuretic peptide was 10O0,, (Schutten et al. 198i). Statisrics. Results are given as means f SE. Data were subjected to one-way analysis of variance for repeated measures (Winer 1971). In the case of significantly large F-values, all possible differences \\ere evaluated systematically by the Newman-Keuls
test (Winer 1971). The level of significance was 0.05 unless otherwise indicated. In the figures and tables, values significantly different from the preinfusion level are marked with an asterisk (*).
RESULTS During the control studies plasma osmolality (posm) and plasma sodium (p-Na') did not change significantly (Table 1). Likewise, detectable changes did not occur in mean arterial blood pressure (MABP), heart rate (HR), central venous pressure (CVP) (Table l), excretion of sodium (Fig. 2) and potassium (Fig. 4), diuresis (Fig. 3), free water clearance (Fig. 3) or the clearances of creatinine and PAH (C(CREA) and
Table 2. Plasma renin activit!- (PRA, ng ml-' h-'), aldosterone (pg ml-'), atrial natriuretic peptide (ANP, pg ml l ) , noradrenaline, adrenaline and dopamine (ng ml-') Time (min) ~
15
45
145
85
PRA 1..5+0.2 1 . jf 0 . 3 2.7 & 0.7
16i0.4 1.3f0.4 1.6 k 0.4"
0.8 0.2*
1.7f0.3 0.7 f0.2* 0.6 k 0.2*
53k 19 57k 19 142 k 25
68+. 18 53 & 20 138f23
5 3 i 15 28 f 12" 85 & 13'
62 f22 23 f 10* 75 f 12*
72f10 53f9 59f5
65i9 62$7 7555
53 i8* 50i8 79+ 10
62f7 51 +4 78k 13
Noradrenaline Control Hypertonic Isotonic
0.34 f 0.0 0.38 f0.04 0.41 i0.03
0 30 f0.02 0.38 i0.04 0.26 i0.02
0.34 f0.0 0.39 f0.04 0.36 f 0.01
0.29 & 0.02 0.39 f0.04 0.37 & 0.02
Adrenaline Control Hypertonic Isotonic
0.14i0.01 0.21 f0.03 0.18 & 0.06
0.11 fO.O1" 0.21 0.03 0.14 0.02
*
0. I 1 & 0.01* 0.21 f0.03 0.12 0.02
*
0.11 f0.01* 0.21 f 0.02 0.13 f O . O 1
Dopamine Control Hypertonic Isotonic
0.52 f0.03 0.52f-0.05 0.47 +_ 0.03
0.47 k 0.03" 0.52 i0.05 0.47 f0.04
0.46 k 0.03* 0.45 0.06 0.45 f0.03
0.40f 0.02+ 0.48 f0.05 0.46 f 0.03
Control €{)pertonic Isotonic Aldosterone Control Hypertonic Isotonic
1.7f0.5 0.6 i0.2*
:IN P Control Hypertonic Isotonic
NaCl was infused from min 30 to min 90. Values significantly different from preinfusion level are marked with an asterisk (*).
Circulatory and renal responses t o NaCl
1
INFUSIOTT
123
INFUSION
\a, 8 lo]
...
1
.5
*.
- *
I
0 0
INFUSION
.
30 60 90 120150180 Time
(min)
-2-
0
30 60 90 120150180 Time
(min)
Fig. 2. Sodium excretion in absolute (upper part) and fractional figures (lower part). NaCI was infused from
Fig. 3. Diuresis (upper part) and free water clearance (lower part). NaCl was infused from min 30 to min 90.
min 30 to min 90. Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantly different (P< 0.05) from the preinfusion level of the same series are marked with an asterisk (").
Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantly different (P < 0.05) from the preinfusion level of the same series are marked with an asterisk (").
C(PAH)), control values of the last two being 45 f3 and 118 f 14 ml min-l respectively. In addition, there were no detectable changes in plasma renin activity (PRA) (Table 2), plasma aldosterone (Table 2) or plasma noradrenaline (Table 2). However, the plasma concentration of ANP (Table 2) decreased marginally but significantly (from 72 10 to 53 f 8 pg ml-l). Likewise the plasma concentrations of adrenaline (Table 2) and dopamine (Table 2) decreased slightly (from 0.14f0.01 to 0.11 fO.01 ng ml-' and from 0.52 f0.03 to 0.46 f0.03 ng ml-' respectively). For technical reasons (probably extraction errors) results on the concentration of vasopressin were not obtained. During the infusion of hypertonic NaCl p-osm rose significantly from 291 f 1 to 301 f 1 mosmol 1-' (Table l), and p-Na+ showed a comparable change from 142 & 1 to 146 & 1 mmol 1-1 (Table 1). There was no detectable change in MABP, HR, CVP (Table 1) or in C(CREA) or C(PAH), the control values of the last two being 40+3 and 114f 10 mi min-' respectively. Sodium excretion increased late in
the infusion period from a preinfusion level of to a maximum of 2.4 0.6 pmol min-' 105 f27 pmol min-' (Fig. 2) and remained above the preinfusion level for the rest of the experiment. Potassium (Fig. 4) showed similar changes (from 8.8 & 1.1 to 39 5 4 pmol min-'), however the increase was smaller, i.e. three- to fourfold in contrast to the 40-fold increase in sodium excretion, and shorter lasting. The large natriuresis occurred without significant changes in plasma concentrations of ANP (Table 2), although there was a tendency towards an increase. However this trend was not related in time to the increase in sodium excretion, as it took place at the beginning of the infusion period, while the peak increase in sodium excretion occurred in the period following termination of the saline infusion. During the saline infusion there was a statistically insignificant trend towards an increase in diuresis (Fig. 3) and free water clearance (Fig. 3). Subsequently diuresis declined significantly from the preinfusion level of 3.1 f0.5 ml min-' to 1.3k0.6 ml min-l. Similarly, free water
124
C. Emmeluth et al.
-
I
1
INFUSION
W a J
Q 10 UL 4O4 0
x E
baJ3
25-
decreased by 41°0 (from 143+25 to 85 k 13 pg ml-'). There were no significant changes in the plasma levels of the catecholamines (Table 2). T h e increase in fractional sodium excretion in the h! pertonic series was significantly larger than during the isotonic infusion ( P < 0.05 by Student's I-test.
"0 3 0 6 0 90 120150180
Time (min)
Fig. 4. Potassium excretion. NaCl was infused from min 30 to min 90. Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantl! different (P < 0.05) from the preinfusion level ofthe same series are marked with an asterisk (").
clearance decreased from 2.9 F 0.5 to 0.1 i0.6 ml min- '.PR;2 and plasma aldosterone (Table 2) were reduced by 6O0,, (from 1.5kO.3 to 0.6 & 0.2 ng ml-.' h-') and 51 O o (from 57 & 19 to 28 2 12 pg ml-') respectively. There were no significant changes in the plasma levels of catecholamines (Table 2). During infirsion of isotonic. salint, there was no change in p-osm, py a ' (Table l ) , 41.4RP or HR (Table 1). .A significant increase in CVP (Table 1) from 2.8 i O . 7 to 4.5 1.0 mmHg occurred during the infusion. This increase persisted, as CVP H ~ S 6.0-t 1.4 mmHg 55 min after the infusion. Statistica1l)- there was no change in C(CREA) or C(P.4H), the control values being 48 $ 3 and 131 k 12 ml min-' respectively. Sodium excretion (Fig. 2) increased from .3.9+ 1.3 to 58S_17 pmolmin ' (b>- a factor of 15) and remained above the preinfusion level for the rest of the experiments. Potassium excretion (Fig. 4) showed a similar pattern (from 12$ 1 to 3 4 F 4 pnol min..'). Diuresis (Fig. 3) increased from 3.9t_0.5 to 7.2k0.7 ml min-' and remained above the preinfusion level for the rest of the experiments, i.e. 90 min. Free water clearance (Fig. 3) increased during the infusion from 3.1 1 0 . 4 to 5.8_+0.6ml min-' and also remained elevated for the rest of the experiments. T h e plasma concentration of ANP (Table 2 ) showed a tendencl- towards an increase from 5 9 F 5 to 7 9 F 10 pg ml-' at the end of the infusion. This was, however, not significant. PRA (Table 2 ) decreased by 5Oo0 (from 2 . 4 k 0 . 7 3 to 0.8010.22 ng mi-' h-'). T h e plasma concentration of aldosterone (Table 2)
DISCUSSION T h e results thus show that simultaneous increases in extracellular volume and tonicity cause a larger increase in Na+ excretion than an increase in volume alone. This larger increase occurred without any measurable increases in I\lABP, HR, CVP, G F R , clearance of PAH or in the concentrations of catecholamines and atrial natriuretic peptide in plasma. Given reasonable assumptions (extracellular 1-olumeequals 20°6 of the body weight [Gaudino & Levitt 19491, infused sodium remains in the extracellular space for the time of the experiment), the cumulated increases in extracellular volume can be calculated to be 19.2 ml lig-' body wt (hypertonic infusion) and 23.4 ml kg-' body wt (isotonic infusion). As the estimated changes in extracellular volume were similar in the two series - or, if anything, smaller with hypertonic saline - any exaggeration of the response in the hypertonic experiments seems to be best explained by the concurrent increase in osmolality/sodium concentration occurring in this series. High-pressure volume/pressure receptors have been described to affect sodium excretion (Zambraski et a[. 1976), although this conclusion seems to disagree with earlier findings (Gilmore & Weisfeldt 1965). These studies were, however, performed in anaesthetized dogs. During our experiments it was not possible to detect any increases in MABP (Table 1). This indicates that in the present circumstances high-pressure receptors cannot be a principal factor in the natriuretic response. -4s expected, the volume expansion during the hypertonic as well as during the isotonic infusion was accompanied by decreases in PRA and the plasma concentration of aldosterone (Table 2 ) . During the isotonic infusion the significant increase in CVP and the tendency towards an increase in the plasma concentration of ANP might be associated with the decrease in PRA. It
Circulatory and renal responses to N a C l has been demonstrated that infusion of ANP causes simultaneous decreases in the plasma levels of renin activity and aldosterone concentration (Cody et al. 1986). Under different conditions, however, where the plasma levels of renin and aldosterone were low prior to infusion of ANP, there were no detectable changes (Goetz et al. 1986). CVP and plasma ANP did not change significantly during the hypertonic series, and the decrease in PRA and therefore plasma aldosterone concentration could hardly be triggered via these variables. As changes in arterial blood pressure were not observed, it does not seem possible that the acute changes in renal sodium excretion were due to an increase in renal perfusion pressure. Sympathetic nervous activity was monitored by measurements of venous plasma catecholamine concentrations ; these did not change. Consequently, the natriuresis after sodium loading could have been elicited either by changes in sympathetic activity not influencing venous plasma catecholamine concentration or by specific mechanisms of unknown character. Atrial distension causes release of atrial natriuretic peptide (for review see Goetz 1988). In our experiments we observed an increase in CVP during isotonic infusion without any detectable changes in plasma ANP. A venous pressure increase could have provided an appropriate stimulus for the activation of lowpressure baroreceptors. However, as plasma ANP did not change measurably, it may be concluded that the renal effects were mediated through mechanisms other than ANP, leaving a vagorenal sympathetic reflex as a possible explanation. This is in line with the findings of Goetz et al. (1988) that an increase in plasma ANP is not necessary for an increase in sodium excretion in these circumstances. Most remarkably, however, during our hypertonic infusion the excretion of sodium increased by a factor of more than 40 without detectable changes in CVP or plasma ANP, suggesting that other receptors and mediators are operating effectively as well, and that these are able to exert their effects despite no measurable stimulus for the lowpressure stretch receptors. These results are in line with the notion that plasma ANP within the physiological range may not play an important role in the precise regulation of renal sodium excretion, at least in the short term, in normal conscious dogs (Bie et al. 1988).
125
One possible explanation for our results is that the increase in p-Na+ or p-osm is able to elicit appropriate changes in renal excretory function. The sensor for these changes could be located in the central nervous system, as several authors have demonstrated that a hypothalamic structure may be involved in the regulation of the extracellular volume (Keeler 1975, Mouw et al. 1980, Lichardus et al. 1987). Of these investigations, two were performed in conscious animals, one in sheep (Lichardus et al. 1987) and one in rats (Mouw et al. 1980). The results from the latter indicated that the periventricular nuclei are involved in the regulation of renal sodium excretion. The study performed in sheep (Lichardus et al. 1987) with lesions of the anterior wall of the third cerebral ventricle demonstrated that these animals were not able to respond normally to a hypertonic load of saline. It is therefore possible that the larger increase in sodium excretion that we observed during the hypertonic infusion was caused by a stimulation of periventricular nuclei in the hypothalamus. The efferent pathway might be secretion of AVP (Mouw et al. 1980, Pierce & Mouw 1984), although other results are not consistent with this suggestion (Park et al. 1985, Merrill et a/. 1986). Another possible pathway is a reduction in renal nerve activity (Morita & Vatner 1985), or even a shift in renal nerve activity from noradrenergic neurons to dopaminergic neurons. However, results supporting this explanation were not obtained in the present study. Intrarenal mechanisms might also be activated by the sodium loading. Expansion of the extracellular volume has been shown to cause a resetting of the glomerulotu bular balance, thereby allowing GFR to increase concomitant with a fall in fractional proximal reabsorption. This could be elicited by a decrease in plasma oncotic pressure (Mend& & Brenner 1990). In our experiments an increase in G F R was not observed, which of course does not exclude a slight increase not detectable by the method applied by us. The present protocol did not allow evaluation of proximal reabsorption ; the use of Li+ clearance as a measure of proximal reabsorption was discouraged, as what appears to be unrealistically small values for distal delivery were obtained in preliminary experiments during water diuresis. I n addition, it is reasonable to assume that comparable decreases in plasma protein concentration oc-
126
C. Emmeluth et nl.
curred during both series of Na' loading. T h i s means that the change in plasma protein might well have contributed to the natriuresis observed, but the difference between the natriuretic responses is not explained by this mechanism. Other intrarenal mechanisms might have been operating as well, i.e. prostaglandins and kallikrein. I t has been demonstrated that increases in medullary tonicity can cause release of PGE,, from interstitial cells, inhibiting tubular reabsorption of sodium (Koeppen 1990). O n the other hand, oral administration of isotonic saline has been shown to cause increases in sodium and water excretion without changes in kallikrein and PGE, excretion (Filep &- Foldes-Filep 1987). Further experiments are required to resolve this issue. T h e present results indicate that several principally different regulatory mechanisms are required to explain the natriuretic response to physiological increases in the total body sodium. One set of receptors is sensitive to an increase in extracellular volume (volume/pressure) and other receptors mediate natriuresis on the basis of increases in Na' concentration of the extracellular volume. Complex interactions between relatively small changes in known regulatory systems or the action of an as yet unidentified natriuretic substance remain the most likely explanations for the present results. The authors thank Inge Pedersen, Sigurd Kramer Hansen, Karen Klausen, Elsa Larsen, Kirsten Larsen, Jytte Oxbel and Trine Eidsvold for expert technical assistance. Inge Pedersen and Sigurd Kramer Hansen are thanked for secretary assistance. Dr W. Vetter kindly provided the antibody for the aldosterone assay. This work was supported by grants from the Danish Medical Research Council, the Nordisk Insulin Foundation, the Velux Foundation and the Brdr Hartman's Foundation.
REFERENCES .kri..ks, S.A. & LARAGH, J.H. 1987. Physiological actions of atrial natriuretic factor. In: P.J. Mulrow & R. Schrier (eds.) Atrial Hormones and Other .Vutririretic Factors, pp. 53-76. Clinical Physiology Series, American Physiological Society, Bethesda, %ID. BEN-JONATHAN, N.& PORTER, J.C. 1976. .I sensitive radioenzymatic assay for dopamine, norepinephrine, and epinephrine in plasma and tissues. Enducrinolon 98., 1497-1507. BIE, P. 1976. Studies of cerebral osmoreceptors in
Increase in plasma sodium enhances natriuresis in response t o a sodium load unable t o change plasma atrial peptide concentration C. E M M E L U T H , H. J. S C H U T T E N , U . K N I G G E , J. W A R B E R G and P. B I E Department of Medical Physiology C, Panum Institute, University of Copenhagen, Denmark
C., SCHUTTEN, H. J., KNIGGE,U., WARBERG, J. & BIE,P. 1990. Increase in EMMELUTH, plasma sodium enhances natriuresis in response to a sodium load unable to change plasma atrial peptide concentration. Actu Physiol Scnnd 140, 119-127. Received 8 September 1989, accepted 29 March 1990. ISSN 0001-6772. Department of Medical Physiology C, Panum Institute, University of Copenhagen, Denmark. The influence of plasma sodium concentration in the control of sodium excretion was investigated in conscious, water-diuretic dogs. NaCl was infused for 60 min as a hypertonic or isotonic solution at a rate of 60 pmol NaCl min-' kg-' body wt. Plasma sodium concentration rose only during hypertonic infusion (P < 0.05). Sodium excretion increased markedly with both infusions (hypertonic, from 2.4 f0.6 to 105 +27 pmol min-'; isotonic, from 3.95- 1.3 to 58+ 17 pmol min-'). Fractional sodium excretion increased more during hypertonic than during isotonic infusion. Hypertonic infusion decreased diuresis from 3.1 k0.5 to 1.3f0.6 ml rnin-l, while isotonic infusion elicited an increase from 3.9 f0.5 to 7.2 f0.7 ml min-'. Plasma renin activity and plasma aldosterone decreased markedly in both series (P < 0.05), the relative changes in the two series being very similar. Central venous pressure increased (2.8 2 0.7 to 4.5 f 1.O mmHg) during isotonic infusion but not significantly during hypertonic infusion. Arterial pressure, heart rate and plasma levels of atrial natriuretic peptide and catecholamines did not change measurably in either series. It is concluded that simultaneous increases in extracellular volume and sodium concentration cause a larger natriuretic response than a change in volume alone, and that a 40-fold increase in sodium excretion may occur without measurable changes in plasma atrial natriuretic peptide concentration. Key words : atrial natriuretic peptide, catecholamines, conscious dogs, natriuresis, renin-angiotensin-aldosterone system, volume expansion.
I t is well documented that humoral, nervous and physical factors, among the last-mentioned notably renal perfusion pressure, participate in the regulation of sodium excretion. However, the relative importance of these factors is not well determined. T h e renin-angiotensin-aidosterone system is of primary importance. Angiotensin I1 influences vascular as well as tubular functions and participates in the tubuloglomerular feedback Correspondence : Claus Emmeluth MD, Department of Medical Physiology C, Blegdarnsvej 3C, DK2200 Copenhagen, Denmark.
mechanism (Navar ct al. 1987). T h e effect of aldosterone is primarily on the distal part of the nephron (Young 1985). ANP has recently been intensively studied (Lang et al. 1987, Goetz 1988), however its precise role in the regulation of renal sodium excretion is still a matter of controversy (Atlas & Laragh 1987). Proximal reabsorption and renal renin release have been shown to be influenced by adrenergic stimulation (via a- and ,&receptors respectively) (DiBona 1985), but the role of dopamine in the regulation of renal sodium excretion is still uncertain (Lee 1982, Levinson et al. 1985). T h e 'pressure natriuresis ' (Hall et ul. 1986,
119
120
C. Emmeluth et
a/.
Guqton 1987) a n d substances inhibiting Na',K-ATPase activitj- (DeWardener 1985) are other factors which should h e considered during investigations into renal sodium excretion. T h e multifactorial regulation of sodium and water excretion calls for simultaneous measurements of a n u m b e r of potentially important regulatory components under precisely defined conditions. We were interested in the roles of volume, sodium concentration and atrial natriuretic peptide in sodium excretion, and the experiments therefore included simultaneous recordings of haemodynamic, humoral and renal excretory variables in connection with transition from a state of relative excess of water t o a state of excess of water and Na'. It was considered important that modest amounts of N a C l were administered rather sloa ly t o conscious animals i n order t o mimic naturally occurring circumstances, e.g. the load of sodium provided by t h e usual diet.
14ATERIAI.S AND M E T H O D S ..?nimii/s. Experiments were performed in seven conscious, trained female dogs (Beagles) weighing 11-16 kg fed on a fixed diet of commercial dog food (Latz Komplett mixed with Latz canned food). T h e dogs received one meal a daj-, usuall>-around 14.00 h, and had free access to tap water. The intake of Naand K- was measured regularly and averaged 2.7 kO.2 and 2.5+.0.2 mmol daj- kg-' body wt respectively. Prior to the present stud!- the common carotid arteries were placed in skin loops using an aseptic technique and general anaesthesia. Prrpururion. Each dog was used for all three tl-pes of experiments with intervals between experiments of at least 1 week. Prior to the experiments the dogs were fasted for 18 h as part of the daily routine. Overnight the! did not haw access to nater ; this M-as an attempt
CHANGE I N S O D I U M C O N T E N T INFUSION
isotonic
to minimize variations between animals. A sterile catheter (Intracath) was introduced via an external jugular vein and the tip was placed in or near the right atrium for measurement of central venous pressure using a Statham P.50 transducer. The pressure profile served as an indicator of the proper position: when a ventricular pressure curve was obtained the catheter was withdrawn a few centimetres. I n addition, catheters were placed in both saphenous veins and in a common carotid artery. The venous catheters were used for blood sampling and infusion, while the arterial one was used to measure the arterial blood pressure (Statham P50). .4 modified silicone Foley catheter was inserted into the bladder. Heart rate was measured by registration of the ECG. Protocol. The dog was hydrated via a stomach tube with a volume of water equivalent to 2";) of body weight 1 h before urine collection was started. Thereafter body weight was kept constant, i.e. ~ 0 . l o , , by use of a servo mechanism (Bie 1976) infusing a hypotonic solution of glucose and urea, 40 mM and 25 mM respectively. Clearances of paraaminohippurate (PAH) and exogenous creatinine were measured as previously described (Emmeluth et al. 1987) and used as measures of renal plasma flow and glomerular filtration rate respectively. When a steadystate water diuresis was achieved, urine sampling at 10-min intervals was initiated. Thirty minutes later an infusion of isotonic or hypertonic saline was started. Three series of NaC1 infusion were carried out. Srrmuli. In one series of experiments 60pmol NaCl kg-'min-' was infused for 60 min as a hypertonic solution (780 mmol I ') at a rate of 0.08 ml kg-' min-'. Due to the servo mechanism there was no net increase in the weight of the dog and therefore the load resulted in an increase in the sodium content of the dog but no change in total fluid r-olume (Fig. I). I n another series of experiments NaCl was infused a t the same rate but in an isotonic 290 mosmol kg ') and solution (156 mmol I-', therefore in a larger volume (0.39 ml kg-' min '). The saline was placed as part of the load of the weight cell prior to infusion so that the infusion was not registered
-
C H A N G E I N WATER C O N T E N T 400.
INFUSION
isotonic
7 100 hypertonic I
.
0
30
.
,
.
.
,
60 90 120150180 Time
(min)
-1001 0
'
30
'
60
Time
'
'
'
.
90 120150180 (rnin)
Fig. 1. :I schematic illustration of the changes in (a) sodium and (b) water content during hypertonic and isotonic infusion.
Circulatory and renal responses to N a C l by the servo mechanism. In this situation both the sodium and the water contents of the dog increased, in contrast to the hypertonic infusion. The increase in the water content was linear as the response to volume expansion was compensated by the action of the servo system (Fig. 1). Control experiments included an infusion of isotonic saline at the low rate of 0.08 ml kg-' min-'. Four blood samples (14 ml each) were obtained at designated points in time during the experiments. After immediate centrifugation at 4 "C the erythrocytes were resuspended in isotonic saline, and reinfused no later than 15 min after the blood sample had been taken. Chemical analyses. Na+ and K' were measured by flame photometry (IL 243, Instrumentation Laboratory). Concentrations of PAH and creatinine in plasma and urine were measured by spectrophotometry (Spectronic 1001, Bausch & Lomb). The osmolality of plasma and urine was estimated by freezing point depression (Digimatic Osmometer Model 3DI1, Advanced Instruments).
+
121
Hormone analyses. Aldosterone was determined on dichloromethane-extracted plasma. The recovery of added [3H]aldosterone was 89 f 1yo (mean fSE, n = 10). The antiserum was a generous gift from Dr W . Vetter (Vetter et al. 1974). The immunoreactive aldosterone was determined by radioimmunoassay. The least detectable limit was 2-3 pg per tube, and half-maximal binding of [3H]aldosterone was at 20-30 pg per tube. Inter- and intra-assay coefficients of variance were 12.3 and 6.1 yo respectively. Plasma renin activity was determined by use of the antibodytrapping method described by Poulsen & J~rgensen (1974). Inter- and intra-assay coefficients of variance were 7.5 and 1.0% respectively. Catecholamines were determined by a radioenzymatic assay described by Ben-Jonathan & Porter (1976). The sensitivity of the assay was 10-20 pg mi-' and the intra-assay coefficients of variations for noradrenaline, adrenaline and dopamine were 6.8, 4.3 and 5.0% respectively. The corresponding interassay coefficients of variation were 14.8, 12.3 and 7.5% respectively (Knigge et al.
Table 1. Plasma concentration of' Na+ (p-Na+) and plasma osmolality (p-osm), mean arterial blood pressure (MABP), heart rate (HR), and central venous pressure (CVP) -
~~
Time (min) ~
15
45
85
145
p-Na' (mmol I-') Control Hypertonic Isotonic
141 f 1 142f 1 1 4 2 11
141 f 1 142f1 141 f 1
141f 1 146f 1" 141 f 1
141 f 1 144f 1" 140f 1
p-osm (mosmol kg-') Control Hypertonic Isotonic
291 1 291 -t 1 293 f 1
292 1 294+ 1 293 & 1
294f 1 301 1" 293 f 1
293 f 1 30Of 1" 294f 1
MABP (mmHg) Control Hypertonic Isotonic
102 f2 100 1 4 11629
102f 1 101+2 115f5
101f2 100f3 119k6
102f2 101f3 118f6
IIR (beats min-') Control Hypertonic Isotonic
80+3 86k4 9517
78f4 78+6 98+3
7312 86f6 107 f8
80+5 82f5 96+6
CVP (mmHg) Control Hypertonic Isotonic
3.2f 1.2 4.5 f0.7 2.8 0.7
3.3f 1.1 5.0k0.8 3.210.7
3.4f 1.2 5.0+ 1.0 4.5f 1.0"
2.8-1- 1.2 4.8 k 0.7 6.0 f 1.4"
+
+
*
NaCl was infused from min 30 to min 90. MABP and HR were measured throughout the experiments; there was no significant change from the data shown. Values significantly different from the preinfusion level are marked with an asterisk (").
122
C. Emmeluth et al.
1988). The concentration of atriul natriuretrc prptide (ANP) in plasma was measured as described by Schutten et al. (1987). Briefly, acidified plasma was extracted using C-18 Sep-Pak cartridges and immunoreacrive ANP was measured bj- a radioimmunoassay. The detection limit was less than 3 pg per tube and the half-maximal binding nas observed at 40 pg per tube. The inter- and intra-assay coefficients of variation, determined at the middle of the sensitive range (50 pg per tube) of the standard curve, were 5.6 and 4.0°,, respectively.. Crossreactivity with synthetic rat a-atrial natriuretic peptide was 10O0,, (Schutten et al. 198i). Statisrics. Results are given as means f SE. Data were subjected to one-way analysis of variance for repeated measures (Winer 1971). In the case of significantly large F-values, all possible differences \\ere evaluated systematically by the Newman-Keuls
test (Winer 1971). The level of significance was 0.05 unless otherwise indicated. In the figures and tables, values significantly different from the preinfusion level are marked with an asterisk (*).
RESULTS During the control studies plasma osmolality (posm) and plasma sodium (p-Na') did not change significantly (Table 1). Likewise, detectable changes did not occur in mean arterial blood pressure (MABP), heart rate (HR), central venous pressure (CVP) (Table l), excretion of sodium (Fig. 2) and potassium (Fig. 4), diuresis (Fig. 3), free water clearance (Fig. 3) or the clearances of creatinine and PAH (C(CREA) and
Table 2. Plasma renin activit!- (PRA, ng ml-' h-'), aldosterone (pg ml-'), atrial natriuretic peptide (ANP, pg ml l ) , noradrenaline, adrenaline and dopamine (ng ml-') Time (min) ~
15
45
145
85
PRA 1..5+0.2 1 . jf 0 . 3 2.7 & 0.7
16i0.4 1.3f0.4 1.6 k 0.4"
0.8 0.2*
1.7f0.3 0.7 f0.2* 0.6 k 0.2*
53k 19 57k 19 142 k 25
68+. 18 53 & 20 138f23
5 3 i 15 28 f 12" 85 & 13'
62 f22 23 f 10* 75 f 12*
72f10 53f9 59f5
65i9 62$7 7555
53 i8* 50i8 79+ 10
62f7 51 +4 78k 13
Noradrenaline Control Hypertonic Isotonic
0.34 f 0.0 0.38 f0.04 0.41 i0.03
0 30 f0.02 0.38 i0.04 0.26 i0.02
0.34 f0.0 0.39 f0.04 0.36 f 0.01
0.29 & 0.02 0.39 f0.04 0.37 & 0.02
Adrenaline Control Hypertonic Isotonic
0.14i0.01 0.21 f0.03 0.18 & 0.06
0.11 fO.O1" 0.21 0.03 0.14 0.02
*
0. I 1 & 0.01* 0.21 f0.03 0.12 0.02
*
0.11 f0.01* 0.21 f 0.02 0.13 f O . O 1
Dopamine Control Hypertonic Isotonic
0.52 f0.03 0.52f-0.05 0.47 +_ 0.03
0.47 k 0.03" 0.52 i0.05 0.47 f0.04
0.46 k 0.03* 0.45 0.06 0.45 f0.03
0.40f 0.02+ 0.48 f0.05 0.46 f 0.03
Control €{)pertonic Isotonic Aldosterone Control Hypertonic Isotonic
1.7f0.5 0.6 i0.2*
:IN P Control Hypertonic Isotonic
NaCl was infused from min 30 to min 90. Values significantly different from preinfusion level are marked with an asterisk (*).
Circulatory and renal responses t o NaCl
1
INFUSIOTT
123
INFUSION
\a, 8 lo]
...
1
.5
*.
- *
I
0 0
INFUSION
.
30 60 90 120150180 Time
(min)
-2-
0
30 60 90 120150180 Time
(min)
Fig. 2. Sodium excretion in absolute (upper part) and fractional figures (lower part). NaCI was infused from
Fig. 3. Diuresis (upper part) and free water clearance (lower part). NaCl was infused from min 30 to min 90.
min 30 to min 90. Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantly different (P< 0.05) from the preinfusion level of the same series are marked with an asterisk (").
Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantly different (P < 0.05) from the preinfusion level of the same series are marked with an asterisk (").
C(PAH)), control values of the last two being 45 f3 and 118 f 14 ml min-l respectively. In addition, there were no detectable changes in plasma renin activity (PRA) (Table 2), plasma aldosterone (Table 2) or plasma noradrenaline (Table 2). However, the plasma concentration of ANP (Table 2) decreased marginally but significantly (from 72 10 to 53 f 8 pg ml-l). Likewise the plasma concentrations of adrenaline (Table 2) and dopamine (Table 2) decreased slightly (from 0.14f0.01 to 0.11 fO.01 ng ml-' and from 0.52 f0.03 to 0.46 f0.03 ng ml-' respectively). For technical reasons (probably extraction errors) results on the concentration of vasopressin were not obtained. During the infusion of hypertonic NaCl p-osm rose significantly from 291 f 1 to 301 f 1 mosmol 1-' (Table l), and p-Na+ showed a comparable change from 142 & 1 to 146 & 1 mmol 1-1 (Table 1). There was no detectable change in MABP, HR, CVP (Table 1) or in C(CREA) or C(PAH), the control values of the last two being 40+3 and 114f 10 mi min-' respectively. Sodium excretion increased late in
the infusion period from a preinfusion level of to a maximum of 2.4 0.6 pmol min-' 105 f27 pmol min-' (Fig. 2) and remained above the preinfusion level for the rest of the experiment. Potassium (Fig. 4) showed similar changes (from 8.8 & 1.1 to 39 5 4 pmol min-'), however the increase was smaller, i.e. three- to fourfold in contrast to the 40-fold increase in sodium excretion, and shorter lasting. The large natriuresis occurred without significant changes in plasma concentrations of ANP (Table 2), although there was a tendency towards an increase. However this trend was not related in time to the increase in sodium excretion, as it took place at the beginning of the infusion period, while the peak increase in sodium excretion occurred in the period following termination of the saline infusion. During the saline infusion there was a statistically insignificant trend towards an increase in diuresis (Fig. 3) and free water clearance (Fig. 3). Subsequently diuresis declined significantly from the preinfusion level of 3.1 f0.5 ml min-' to 1.3k0.6 ml min-l. Similarly, free water
124
C. Emmeluth et al.
-
I
1
INFUSION
W a J
Q 10 UL 4O4 0
x E
baJ3
25-
decreased by 41°0 (from 143+25 to 85 k 13 pg ml-'). There were no significant changes in the plasma levels of the catecholamines (Table 2). T h e increase in fractional sodium excretion in the h! pertonic series was significantly larger than during the isotonic infusion ( P < 0.05 by Student's I-test.
"0 3 0 6 0 90 120150180
Time (min)
Fig. 4. Potassium excretion. NaCl was infused from min 30 to min 90. Control experiments are marked with triangles, the hypertonic infusion with circles and the isotonic infusion with squares. Values significantl! different (P < 0.05) from the preinfusion level ofthe same series are marked with an asterisk (").
clearance decreased from 2.9 F 0.5 to 0.1 i0.6 ml min- '.PR;2 and plasma aldosterone (Table 2) were reduced by 6O0,, (from 1.5kO.3 to 0.6 & 0.2 ng ml-.' h-') and 51 O o (from 57 & 19 to 28 2 12 pg ml-') respectively. There were no significant changes in the plasma levels of catecholamines (Table 2). During infirsion of isotonic. salint, there was no change in p-osm, py a ' (Table l ) , 41.4RP or HR (Table 1). .A significant increase in CVP (Table 1) from 2.8 i O . 7 to 4.5 1.0 mmHg occurred during the infusion. This increase persisted, as CVP H ~ S 6.0-t 1.4 mmHg 55 min after the infusion. Statistica1l)- there was no change in C(CREA) or C(P.4H), the control values being 48 $ 3 and 131 k 12 ml min-' respectively. Sodium excretion (Fig. 2) increased from .3.9+ 1.3 to 58S_17 pmolmin ' (b>- a factor of 15) and remained above the preinfusion level for the rest of the experiments. Potassium excretion (Fig. 4) showed a similar pattern (from 12$ 1 to 3 4 F 4 pnol min..'). Diuresis (Fig. 3) increased from 3.9t_0.5 to 7.2k0.7 ml min-' and remained above the preinfusion level for the rest of the experiments, i.e. 90 min. Free water clearance (Fig. 3) increased during the infusion from 3.1 1 0 . 4 to 5.8_+0.6ml min-' and also remained elevated for the rest of the experiments. T h e plasma concentration of ANP (Table 2 ) showed a tendencl- towards an increase from 5 9 F 5 to 7 9 F 10 pg ml-' at the end of the infusion. This was, however, not significant. PRA (Table 2 ) decreased by 5Oo0 (from 2 . 4 k 0 . 7 3 to 0.8010.22 ng mi-' h-'). T h e plasma concentration of aldosterone (Table 2)
DISCUSSION T h e results thus show that simultaneous increases in extracellular volume and tonicity cause a larger increase in Na+ excretion than an increase in volume alone. This larger increase occurred without any measurable increases in I\lABP, HR, CVP, G F R , clearance of PAH or in the concentrations of catecholamines and atrial natriuretic peptide in plasma. Given reasonable assumptions (extracellular 1-olumeequals 20°6 of the body weight [Gaudino & Levitt 19491, infused sodium remains in the extracellular space for the time of the experiment), the cumulated increases in extracellular volume can be calculated to be 19.2 ml lig-' body wt (hypertonic infusion) and 23.4 ml kg-' body wt (isotonic infusion). As the estimated changes in extracellular volume were similar in the two series - or, if anything, smaller with hypertonic saline - any exaggeration of the response in the hypertonic experiments seems to be best explained by the concurrent increase in osmolality/sodium concentration occurring in this series. High-pressure volume/pressure receptors have been described to affect sodium excretion (Zambraski et a[. 1976), although this conclusion seems to disagree with earlier findings (Gilmore & Weisfeldt 1965). These studies were, however, performed in anaesthetized dogs. During our experiments it was not possible to detect any increases in MABP (Table 1). This indicates that in the present circumstances high-pressure receptors cannot be a principal factor in the natriuretic response. -4s expected, the volume expansion during the hypertonic as well as during the isotonic infusion was accompanied by decreases in PRA and the plasma concentration of aldosterone (Table 2 ) . During the isotonic infusion the significant increase in CVP and the tendency towards an increase in the plasma concentration of ANP might be associated with the decrease in PRA. It
Circulatory and renal responses to N a C l has been demonstrated that infusion of ANP causes simultaneous decreases in the plasma levels of renin activity and aldosterone concentration (Cody et al. 1986). Under different conditions, however, where the plasma levels of renin and aldosterone were low prior to infusion of ANP, there were no detectable changes (Goetz et al. 1986). CVP and plasma ANP did not change significantly during the hypertonic series, and the decrease in PRA and therefore plasma aldosterone concentration could hardly be triggered via these variables. As changes in arterial blood pressure were not observed, it does not seem possible that the acute changes in renal sodium excretion were due to an increase in renal perfusion pressure. Sympathetic nervous activity was monitored by measurements of venous plasma catecholamine concentrations ; these did not change. Consequently, the natriuresis after sodium loading could have been elicited either by changes in sympathetic activity not influencing venous plasma catecholamine concentration or by specific mechanisms of unknown character. Atrial distension causes release of atrial natriuretic peptide (for review see Goetz 1988). In our experiments we observed an increase in CVP during isotonic infusion without any detectable changes in plasma ANP. A venous pressure increase could have provided an appropriate stimulus for the activation of lowpressure baroreceptors. However, as plasma ANP did not change measurably, it may be concluded that the renal effects were mediated through mechanisms other than ANP, leaving a vagorenal sympathetic reflex as a possible explanation. This is in line with the findings of Goetz et al. (1988) that an increase in plasma ANP is not necessary for an increase in sodium excretion in these circumstances. Most remarkably, however, during our hypertonic infusion the excretion of sodium increased by a factor of more than 40 without detectable changes in CVP or plasma ANP, suggesting that other receptors and mediators are operating effectively as well, and that these are able to exert their effects despite no measurable stimulus for the lowpressure stretch receptors. These results are in line with the notion that plasma ANP within the physiological range may not play an important role in the precise regulation of renal sodium excretion, at least in the short term, in normal conscious dogs (Bie et al. 1988).
125
One possible explanation for our results is that the increase in p-Na+ or p-osm is able to elicit appropriate changes in renal excretory function. The sensor for these changes could be located in the central nervous system, as several authors have demonstrated that a hypothalamic structure may be involved in the regulation of the extracellular volume (Keeler 1975, Mouw et al. 1980, Lichardus et al. 1987). Of these investigations, two were performed in conscious animals, one in sheep (Lichardus et al. 1987) and one in rats (Mouw et al. 1980). The results from the latter indicated that the periventricular nuclei are involved in the regulation of renal sodium excretion. The study performed in sheep (Lichardus et al. 1987) with lesions of the anterior wall of the third cerebral ventricle demonstrated that these animals were not able to respond normally to a hypertonic load of saline. It is therefore possible that the larger increase in sodium excretion that we observed during the hypertonic infusion was caused by a stimulation of periventricular nuclei in the hypothalamus. The efferent pathway might be secretion of AVP (Mouw et al. 1980, Pierce & Mouw 1984), although other results are not consistent with this suggestion (Park et al. 1985, Merrill et a/. 1986). Another possible pathway is a reduction in renal nerve activity (Morita & Vatner 1985), or even a shift in renal nerve activity from noradrenergic neurons to dopaminergic neurons. However, results supporting this explanation were not obtained in the present study. Intrarenal mechanisms might also be activated by the sodium loading. Expansion of the extracellular volume has been shown to cause a resetting of the glomerulotu bular balance, thereby allowing GFR to increase concomitant with a fall in fractional proximal reabsorption. This could be elicited by a decrease in plasma oncotic pressure (Mend& & Brenner 1990). In our experiments an increase in G F R was not observed, which of course does not exclude a slight increase not detectable by the method applied by us. The present protocol did not allow evaluation of proximal reabsorption ; the use of Li+ clearance as a measure of proximal reabsorption was discouraged, as what appears to be unrealistically small values for distal delivery were obtained in preliminary experiments during water diuresis. I n addition, it is reasonable to assume that comparable decreases in plasma protein concentration oc-
126
C. Emmeluth et nl.
curred during both series of Na' loading. T h i s means that the change in plasma protein might well have contributed to the natriuresis observed, but the difference between the natriuretic responses is not explained by this mechanism. Other intrarenal mechanisms might have been operating as well, i.e. prostaglandins and kallikrein. I t has been demonstrated that increases in medullary tonicity can cause release of PGE,, from interstitial cells, inhibiting tubular reabsorption of sodium (Koeppen 1990). O n the other hand, oral administration of isotonic saline has been shown to cause increases in sodium and water excretion without changes in kallikrein and PGE, excretion (Filep &- Foldes-Filep 1987). Further experiments are required to resolve this issue. T h e present results indicate that several principally different regulatory mechanisms are required to explain the natriuretic response to physiological increases in the total body sodium. One set of receptors is sensitive to an increase in extracellular volume (volume/pressure) and other receptors mediate natriuresis on the basis of increases in Na' concentration of the extracellular volume. Complex interactions between relatively small changes in known regulatory systems or the action of an as yet unidentified natriuretic substance remain the most likely explanations for the present results. The authors thank Inge Pedersen, Sigurd Kramer Hansen, Karen Klausen, Elsa Larsen, Kirsten Larsen, Jytte Oxbel and Trine Eidsvold for expert technical assistance. Inge Pedersen and Sigurd Kramer Hansen are thanked for secretary assistance. Dr W. Vetter kindly provided the antibody for the aldosterone assay. This work was supported by grants from the Danish Medical Research Council, the Nordisk Insulin Foundation, the Velux Foundation and the Brdr Hartman's Foundation.
REFERENCES .kri..ks, S.A. & LARAGH, J.H. 1987. Physiological actions of atrial natriuretic factor. In: P.J. Mulrow & R. Schrier (eds.) Atrial Hormones and Other .Vutririretic Factors, pp. 53-76. Clinical Physiology Series, American Physiological Society, Bethesda, %ID. BEN-JONATHAN, N.& PORTER, J.C. 1976. .I sensitive radioenzymatic assay for dopamine, norepinephrine, and epinephrine in plasma and tissues. Enducrinolon 98., 1497-1507. BIE, P. 1976. Studies of cerebral osmoreceptors in
