Effects o f Oral Salt Load on Arginine-Vasopressin Secretion in N orm al Subjects*

Transcription

1 ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 17, No. 5 Copyright 1987, Institute for Clinical Science, Inc. Effects o f Oral Salt Load on Arginine-Vasopressin Secretion in N orm al Subjects* LETIZIA SPIN ELLI, M.D., PAOLO G O LINO, M.D., FE D E R IC O PISC IO N E, M.D., MASSIMO CH IA RIELLO, M.D., AM ELIA FO CACCIO, M.D., G IU SEPPE AMBROSIO, M.D., and MARIO CO N D O R ELLI, M.D. Department of Internal Medicine, Division o f Cardiology, 2nd School o f Medicine, University o f Naples, Naples, Italy ABSTRACT Arginine-vasopressin (AVP) plays an im portant role in regulating water balance in humans. Its secretion is under control of several mechanisms, some of which are not completely understood. The purpose of the present study was to evaluate the effects of an acute oral salt load on AVP secretion in normal subjects. Six normal volunteers received 350 m E q of NaCl per os. Pulm onary capillary wedge pressure and right atrial pressure, plasma AVP, plasm a sodium and potassium concentration, plasm a osmolality, hem atocrit, urinary sodium and potassium excretion, and urinary flow were m easured at baseline and every 30 m inutes for two hours after the salt load. H em odynam ics as well as urinary sodium and potassium excretion did not change over the study. Ninety m inutes after the salt load, plasma AVP increased from the basal value of 6.0 ± 0.9 pg per ml to 10.1 ± 1.2 pg p er ml (mean ± SE, p < 0.005) and a significant reduction in diuresis of about 50% was observed. However, plasm a osmolality and plasma sodium concentration increased significantly only 120 min after the salt load, from the initial value of ± 2.2 mosm per kg and ± 1.4 m E q p er 1 (mean ± SE) to ± 2.5 mosm per kg and ± 1.5 m E q per 1, respectively (p < 0.01). Ninety m inutes after the salt load, no correlation was found betw een plasm a osmolality and plasm a AVP concentration, indicating that AVP secretion was independent of changes in system ic blood osmolality. This phenom enon suggests that the observed increase in AVP concentration at 90 m inutes was not due to a stimulation of the hypothalamic osmoreceptors since it was not accompanied by any d etectable change both in plasm a osmolality and plasm a sodium concentration. These data suggest that osmoreceptors located in areas other than the hypothalam us (probably the portal vascular bed) may play an im portant role in regulating AVP secretion in response to osmotic stimuli of dietary origin. * Address reprint requests to: Paolo Golino, M.D. Via Rocco lem m a 8, Napoli, Italy /87/ $01.20 Institute for Clinical Science, Inc.

2 AVP SECRETION AND PORTAL OSMORECEPTORS 351 In troduction A num ber of physiologic mechanisms are involved in m aintaining the balance betw een intake and excretion of w ater and salts. Am ong these m echanism s, several horm ones are known to play an im portant role in regulating body fluid homeostasis in man. Arginine-vasopressin (AVP) is a nonapeptide whose main biological action is to conserve w ater and to concentrate urine. This action of AVP contributes in maintaining constant the osmolality and v o lu m e,o f body fluids. U nder norm al conditions, AVP release is prim arily regulated by osmoreceptors located in the hypothalamus. It has been dem onstrated that the infusion in norm al subjects of hypertonic saline solutions at a constant rate leads to a release of AVP p ro p o r tional to the increase in plasm a osmolality.10 It is also w ell know n th a t th e reduction in plasm a volum e stim ulates the release of AVP by reducing the tonic inhibitory impulses from the left atrium to the hypothalam us.5,14,17 It is believed that this action is the result of the effect of hypovolem ia on stre tc h recep to rs located in the left atrium and perhaps in the pulm onary veins, as well as of the activation of carotid and aortic baroreceptors in response to hypotension.5,14,17 However, it is not com pletely defined how osm otic stim uli of dietary origin affect AVP secretion. In a recent review, Bie2 suggested th at the hypothesis of hypothalam ic osm oreceptors could not account for all of the results obtained in stu d ie s of th e re la tio n sh ip b e tw e e n plasm a osmolality and AVP release. In a n e sth e tiz ed dogs, re c e p to rs located outside the head were reported to play an im portant role in the osmotic control of plasma AVP concentration,3 and there is som e evidence of hepatic or portal o sm o re c ep to rs in rats and d o g s.1,4,6 O ther investigators, however, could not confirm th eir existence.8,12 The goal of the present study was to test the hypothesis of w hether or not an oral hyperosm otic stimulus would affect AVP release in man before any change in system ic plasm a osm olality, th u s suggesting the presence of osm oreceptors lo c a te d in th e p o rta l v a scu lar b e d. Accordingly, plasma AVP concentration, plasm a osmolality, and plasm a sodium concentration w ere m onitored after an a c u te o ral sa lt load. F u rth e rm o re, p roper consideration was given to the changes in sodium and potassium excretion as well as right atrial and pulmonary capillary w edge p ressu re because in terferen ce from recep to rs controlling system ic sodium co n cen tratio n or from stretch receptors could com plicate in terpretation of th e results. M ethods Six normal volunteers (ages 31 to 42) entered the study. All of them gave consent after being inform ed of the procedure. The present study was approved by th e local E thical C om m ittee. The patients were not taking any drugs. Five days before the study began a diet was observ ed w ith a stan d ard c o n te n t of sodium (150 m E q p er day). No food or w ater w ere allow ed starting from the night before the experim ent. The study was conducted in a quiet room with the patients lying in th e supine position. A Sw an-ganz c a th e te r was in serted percutaneously through the right fem o ral vein (using 10 ml of a two percent lidocaine solution to induce local anesthesia) and advanced into the pulmonary artery under fluoroscopic control. A urinary c a th eter of appropriate size was inserted via the uretra into the urinary bladder. After 60 m inutes of rest, right atrial pressure and pulmonary capillary w edge pressure (Statham P23 db transducer) and blood and urine samples were obtained to assess plasm a AVP, N a+, and K + concentration, diuresis, and u ri

3 3 5 2 SPINELLI, GOLINO, PISCIONE, CHIARIELLO, FOCACCIO, AMBROSIO, AND CONDORELLI nary N a+ and K + excretion. These m easurem en ts w ere rep e a te d 15 m inutes later, and the two values w ere averaged and considered as baseline. The patients w ere then given an oral salt load (350 m E q of NaCl divided in seven tablets of 50 m E q each) w ith a small am ount of w ater (30 to 50 ml). Haem odynam ics w ere recorded at 30- m in intervals over two hours. At th e same tim e intervals, 12 ml of blood were w ithdraw n to m easure AVP, N a+, and K + concentration. In addition, all the urine in the bladder was drained through the catheter, and th e quantity was carefully m easured; the urine was later used to assay its N a+ and K + content. 500 gamma counter. The lower lim it of sensitivity in this assay was 0.63 pg per ml. T he intraassay and interassay coefficients of variation were 6.9 percent and 10.4 p e rc e n t, respectively. T he a n tiserum exhibited less than 0.01 percent cross-reactivity w ith oxytocin, angiotensin I, and angiotensin II, or bradykinin betw een concentrations of 0 to 1000 pg p e r ml; how ever, a 42 p e rc e n t crossreactivity was observed w ith lysine vasopressin. Urinary and plasma concentrations of N a+ and K + w ere determ ined by flam e photom etry. $ Plasm a osm olality was m easured by freezing point depression with a Fiske osmometer. A ssa y s Blood sam ples w ere collected into plastic tubes containing 0.1 ml of a six percent sodium ethylenediam ine tetraacetic acid (EDTA) solution and im m ediately placed into ice bath. At the end of th e e x p e rim e n t, blood sam ples w ere centrifuged and plasma and urine stored frozen at 70 C. P la s m a A V P c o n c e n tr a tio n w as assayed in triplicate by a commercially available radioimmunoassay (RIA) kit.* T he AVP was ex tracted from plasm a sam ples by p ro tein precipitation w ith ab so lu te ethanol. A ir-dryed extracts were then reconstituted with one ml of phosphate buffer for assay. 125I-arginine vasopressin was used as a tracer, and synthetic arginine vasopressin was used for th e c a lib ra tio n curve. A nti-a V P serum was obtained from rabbits sensitized against AVP. After seven days of incubation at 4 C, with the tracer added after two days, bound AVP was separa te d by th e addition of goat gam m a globulin anti-rabbit gamma globulin, and centrifugation. Bound and free fraction were counted! with a Beckman Gamma S t a t is t ic s D ata are expressed as m ean ± stand ard erro r of the m ean. Statistical comparisons w ere made using the one-way analysis of variance w ith a design for repeated m easurem ents. Results The oral salt load was well tolerated by the patients; none of them had gastric pain, nausea, or thirst during th e observ a tio n p e rio d. P u lm o n a ry c a p illa ry w edge p ressu re (reflecting left atrial pressure) and mean right atrial pressure did not change during the experim ent (table I). Blood hem atocrit decreased significantly 120 m inutes after the salt load from the baseline of 42 ± 3 percent to 37 ± 3 percent (p < 0.01, table I). P lasm a K + c o n c e n tra tio n d id not change 30, 60, and 90 m inutes after the salt load, b ut it decreased significantly from the initial value of 4.9 ± 0.1 to 3.9 ± 0.1 m E q per 1 at the end of the experim ent, p < (figure 1). No signifi- * Sorin Biomedica. t Beckman, Gamma 500 gamma counter. $ Beckman, model E2 A.

4 Hemodynamics and Blood Hematocrit After Administration of 350 meq of NaCl p&n oi Wedge capillary pressure (mmhg) Mean right atrial pressure (mmhg) Haematocrit (%) AVP SECRETION AND PORTAL OSMORECEPTORS ± 1 4 ± 1 3 ± 1 3 ± ± 4 42 ± * *p < 0.01 Data expressed as mean ± standard error cant changes in plasm a N a + concentratio n w e re o b s e rv e d 30, 60, an d 90 m inutes after the oral salt load as com p ared to the baseline of ± 1.4 m Eq per 1 (figure 1). Only 120 m inutes after the salt load, plasm a N a + concentra tio n in creased significantly, up to ± 1.5 m E q per I, p < 0.01). This represents an increase of about two percent as com pared to the initial value. Similarly, plasma osmolality showed no significant differences at 30, 60, and 90 m inutes after the salt load respect to the basal value of ± 2.2 mosm per kg, whereas a significant increase in plasma osmolality of about 2.5 percent (284.8 ± 2.5 mosm per kg) was observed at 120 m inutes after the salt load (p < 0.01, figu re 2). In te r e s tin g ly, a s ig n ific a n t increase in plasm a AVP concentration occurred before any detectable changes in plasma osmolality; in fact, 90 m inutes after the oral salt load, w hen plasm a osmolality was not different from th e initial value, plasma AVP increased about 60 percent, from 6.0 ± 0.9 to 10.1 ± 1.2 pg per ml (p < 0.001, figure 2). A close correlation was found betw een p lasm a AVP and p lasm a o sm o lality before, 30, and 60 m inutes after the salt load (r = 0.94, p < 0.01). At 90 minutes, how ever, plasm a AVP increased significantly without a simultaneous increase in plasm a osmolality, and the correlation betw een the two param eters was not significant (r = 0.71, p = NS). In figure 3, plasm a AVP and osmolality are plotted at 30, 60, and 90 m inutes for each patient. It is evident that, at 90 m inutes, virtually no changes in plasma osm olality occurred, w hereas a sharp increase in plasm a AVP concentration was observed in all patients. Plasma AVP further increased at 120 m inutes, up to 10.9 ± 1.3 pg per ml, (p < 0.001, figure 2). The increase in AVP co n c en tra tio n at 90 and 120 m inutes after the salt load was accompanied by a significant reduction in urine excretion of about 50 percent (table II). The conseq u en t haem odilution may explain the significant decrease in plasm a K + conc e n tr a t i o n a n d b lo o d h e m a to c r it observed at the end of the study , J PLASMA Na+ (m E q /l) MINUTES ** F ig u r e 1. T im e-course o f plasm a sodium and potassium concentration after administration of an oral salt load (350 m Eq of NaCl) in normal human subjects (n = 6). Values are expressed as mean ± standard error of the mean. *p < 0.01 and **p <

5 3 5 4 SPINELLI, GOLINO, PISCIONE, CHIARIELLO, FOCACCIO, AMBROSIO, AND CONDORELLI F ig u r e 2. Time course of plasma osmolality (close circles) and plasma AVP concentration (open circles) after administration of the salt load. Note that no difference was observed in plasma osmolality 90 min after the salt load, whereas plasma AVP increased significantly from the initial value. Each point represents the mean ± standard error of the mean for six subjects. *p < 0.01 and **p < S odium a n d p o ta ssiu m e x c re tio n rem ain ed u n chan g ed th ro u g h o u t the study (table II). hypothalam oneurohypophyseal tra c t.1 However, it should be pointed out that th e osm otic stim uli used in th e p r e viously m entioned studies are somehow unphysiologic. In fact, changes in the composition and osmolality of the portal blood normally represent the net result of intestinal absorption and secretion of flu id s and so lu tes. P o rtal infusions m erely serve as a way of imitating these processes under more precisely defined conditions. U nder these circumstances, the actual physiologic contribution of the entero-hepatic circulation per se may be questioned. Furtherm ore, very little is known about this issue in humans. W ith that in mind, the hyperosm otic stimulus used in the present study, i.e., 350 m Eq of N acl p e r os, is to be considered a m ore physiologic one, since it was given Discussion The goal of the present study was to test the hypothesis that AVP release in m an may also b e regulated by osm oreceptors located in areas other than the hypothalam us, namely, th e hepatic-portal vascular bed. The possibility that the portal vascular bed may contribute in regulating water balance was first suggested by H aberich6 in This investigator showed that in trap o rtal infusion of hypotonic solutio n s le d to an e a r lie r a n d h ig h e r increase in diuresis as compared to infusion into the inferior vena cava.6 Subseq u e n tly, o th e r in v e s tig a to rs h a v e d e s c r ib e d e v id e n c e for p e r ip h e r a l o sm o re c ep to rs lo cated in th e p o rta l area.1,4 The evidence for these receptors is largely based upon experim ents in which intraportal infusions of hypotonic or hypertonic solutions w ere linked to changes in urine flow,6 plasma AVP conc e n tra tio n,4 or n eu ral activity of the o E «g 290- MIN. _I_ 60 MIN. F ig u r e 3. Plasm a osm olality and plasm a AVP concentration plotted for each subject 30, 60, and 90 min after the salt load. Note that although plasma osmolality did not change, a sharp increase in plasma AVP was observed at 90 min. *p <

6 AVP SECRETION AND PORTAL OSMORECEPTORS TABLE I I Urine Flow, Na+ and K+ Urinary Excretion After Administration of 350 meq of NaCl pea 04 B a s e U r i n e f l o w (ml/min) ± ± ± ± ± 0.11* N a + U r i n a r y e x c r e t i o n (meq/min) ± ± ± ± ± K + U r i n a r y e x c r e t i o n (meq/min) ± ± ± ± *p < D a t a e x p r e s s e d a s m e a n ± s t a n d a r d e r r o r per os and, therefore, takes into account the intestinal absorption and secretion of fluids and solutes. In discussing the data of the present study, proper consideration should be given to two findings. First, plasma AVP concentration sharply increased at 90 m inutes after the salt load; this increase occurred in all the subjects w hen no d e te c ta b le changes both in system ic plasm a osmolality and N a+ concentration w ere observed. It is im portant to em phasize that, as expected,9 a close correlation was found betw een plasm a AVP and osmolality before, 30, and 60 m in u tes after th e salt load, w hereas these param eters were not correlated at 90 m inutes, indicating that the increase in AVP concentration at this tim e was independent of changes in plasma osmolality. Furtherm ore, the increase in AVP concentration at 90 m inutes was m ost likely independent of volume expansion of th e in tra v asc u la r flu id, since no changes in pulm onary capillary wedge pressure and right atrial pressure and hem atocrit w ere observed at this tim e of the study. Second, system ic blood osm olality rem ained unaltered for a relatively long period of tim e. In fact, in the present study, plasma osmolality increased only 120 m inutes after the salt load. These data indicate that the entero-hepatic circulation plays an im portant role in buffering ab ru p t changes in system ic blood osmolality as a consequence of dietary intake of fluids and solutes. A lim itation of the p rese n t study is represented by the fact that, owing to its intrinsic difficulties, portal osm olality was not m easured. Consequently, direct ev id en ce is n ot available th a t p o rtal osm olality increased earlier than system ic osmolality after the salt load. However, it has been dem onstrated in dogs that a solute load related to food intake produced a larger and earlier increase in p o r ta l b lo o d o s m o la lity th a n th a t observed in the systemic venous blood.4 It is, therefore, very likely that portal osmolality increased in a similar fashion in our patients. A n o th er p o te n tia l p ro b lem of this study is th at the salt load could have stim ulated the release of gastrointestinal horm ones which could have interfered with the AVP assay. Although data are not available regarding th e cross-reactivity of gastrointestinal hormones with the anti-avp serum, the fact that urine flow d e c re a sed by 50 p e rc e n t at 90 m in strongly suggests th a t th e substance m easured in plasma at that tim e period was actually AVP. The im portance of osmotic pressure in regulating AVP release has b een recognized for many years, and several studies have been conducted to evaluate the changes in AVP secretion after various osmotic stimuli. It is well known that the secretion of vasopressin is controlled by

7 3 5 6 SPINELLI, GOLINO, PISCIONE, CHIARIELLO, FOCACCIO, AMBROSIO, AND CONDORELLI a very sensitive mechanism that is able to d e te c t and resp o n d to changes in plasm a osm olality as little as one p e r cent.9 From pioneering studies of Verney16 and other evidence, it has been concluded that the osm oreceptors are located in the anterior part of the brain, presum ably in the hypothalam us. The AVP secretio n is also in flu e n c e d by changes in blood volum e,5,13,15 and this interacts w ith blood osmolality in reg u lating plasm a AVP concentration.7 Evidence has accum ulated over the past decade indicating that signals which modify AVP release may arise not only in the hypothalam ic receptors, as first p ro posed by Verney,16 b u t also in an area located som ew here within the portal vascular bed. It has, in fact, been suggested th a t th e hypothesis of hypothalam ic osmoreceptors could not entirely explain the results obtained in studies on the relationship betw een plasma osmolality and AVP re le a s e.2 H ab erich d em o n strated in rats that the diuretic effects of hypotonic fluid loading w ere m ore pronounced if the fluids were infused into the portal vein com pared to system ic vein infusion.6 In anesthetized dogs, receptors located outside the head were reported to play an im portant role in the osmotic control of plasm a AVP concentration.3 Recently, portal osmoreceptors have b een dem o n strated in rats1 and conscious dogs.4 In particular, th e infusion of hypertonic saline into the portal vein of conscious dogs was associated with a prom pt increase in plasma AVP concentration in absence of changes in systemic blood osmolality.4 The maximal increase was observed five m inutes after starting the infusion; subsequently, AVP concentration decreased rapidly despite continued infusion and was back at the control level after 10 m inutes. Infusion of a m ore concentrated solution p ro duced a n o th e r tra n sie n t elevation of plasm a AVP concen tratio n. In te re s t ingly, no changes in plasm a AVP w ere observed if the same experim ent was rep eated after chronic surgical denervation of the liver, thus indicating a reflex nature of the observed phenom enon.4 In conclusion, data from the present study suggest the presence of osm oreceptors located in areas other than the hypothalam us (probably the portal vascular bed) and that they can be activated by an oral salt load. The physiological role of such osm oreceptors in responding to osmotic stimuli of dietary origin may be relevant. References 1. B a e r t s c h i, A. J. and V a l l e t, P. G.: Osmosensitivity o f the hepatic portal vein area and vasopressin release in rats. J. Physiol. 315:217, B i e, P.: Osmoreceptors, vasopressin, and control of renal water excretion. Physiol. Rev. 60:961, B i e, P.: Sustained water diuresis in anesthetized dogs: Antidiuresis in response to intravenous and bilateral intracarotid infusion o f hyperosmolar solutions of sodium chloride. Acta Physiol. Scand. 101:446, C h w a l b in s k a -M o n e t a, J.: Role of hepatic portal osmoreception in the control o f ADH release. Am. J. Physiol. 263:E603, D e T o r r e n t e, A., R o b e r t s o n, G. L., M c D o n a l d, K. M., and S c h r ie r, R. W. Mechanism of diuretic response to increased left atrial pressure in the anesth etized dog. K idney Int. 8:355, H a b e r i c h, F. J.: O sm oreception in the portal circulation. Fed. Proc. 27:1137, M ic h e l a k is, A. M.: The relationship between plasma renin and aldosterone in normal man. Circ. Res. 26(supp. I.):185, P o t k a y, S. and G i l m o r e, J. P.: Renal responses to vena caval and portal infusion of sodium chloride in unanesthetized dogs. Clin. Sei. 39:13, R o b e r t s o n, G. L.: The im portance of plasma osm olality in regulating antidiuretic hormone secretion in man. J. Clin. Invest. 51:79, R o b e r t s o n, G. L.: The regulation of vasopressin function in health and disease. Reg. Prog. Hormone Res. 33:333, R o b e r t s o n, G. L. and A t h a r, S.: The interaction of blood osmolality and blood volume in regulating plasm a vasopressin in man. J. Clin. Endocrinol. Metab. 42:613, S c h n e i d e r, E. G., D a v is, J. O., R o b b, C. A.,

8 AVP SECRETION AND PORTAL OSMORECEPTORS B a u m b e r, J. S., J o h n s o n, J. A., and W r i g h t, F. S.: Lack o f evidence for a hepatic osmoreceptor mechanism in conscious dogs. Am. J. Physiol. 218:42, S c h r ie r, R. W. and B e r l, T.: Non-osmolar factors affecting renal water excretion. N ew Engl. J. Med. 292:81, S k o r e c k i, K. L. a n d B r e n n e r, B. M.: B o d y f lu i d h o m e o s ta s is in c o n g e s tiv e h e a r t fa ilu re a n d c irro s is w ith a s c ite s. A m. J. M e d. 72:323, S k o r e c k i, K. L. a n d B r e n n e r, B. M.: B o d y f lu i d h o m e o s ta s is in m a n. A m. J. M e d. 70:77, V e r n e y, E. B.: The antidiuretic hormone and the factors which determ ine its release. Proe. R. Soc. London, Ser. B 135:25, W e it z m a n, R. E., R e v ic z k y, A., O d d i e, T. H., and F is h e r, D. A.: Effect of osmolality on arginine vasopressin and renin release after hemorrage. Am. J. Physiol. 23S:E62, 1980.