INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

SUBSTANCES CONSUMED ORALLY are absorbed from the intestine into the blood, circulate in the hepatic vasculature, then enter the systemic circulation. Thus the liver is the ideal place to monitor the quality and quantity of these substances and transduce them into hepatic afferent nerve activity (HANA; 3, 9, 19). A series of studies has demonstrated the existence of a hepatoportal Na⁺-sensor mechanism that projects to the medulla and hypothalamus via the hepatic nerve (11) and plays a significant role in regulating body fluid homeostasis (8, 10, 21). A recent study from our laboratory demonstrated that periarterial HANA increases in response to intraportal hypertonic NaCl infusion and that this response is completely abolished by pretreatment with ouabain, furosemide, or bumetanide, but not by amiloride or 4-acetamido-4'-isothiocyanatostilbene-2,2'-disulfonic acid (7). These results suggest that the hepatoportal Na⁺-sensor mechanism senses the Na⁺ concentration via the bumetanide-sensitive Na⁺-K⁺-2Cl cotransporter. Although the linkage between the Na⁺-K⁺-2Cl cotransporter and the increased HANA is unclear, two possibilities can be considered. First, an increase in hepatocellular volume, induced by Na⁺ influx via the Na⁺-K⁺-2Cl cotransporter, might stimulate the connected nerve terminal. Second, the nerve terminal itself might bear the Na⁺-K⁺-2Cl cotransporter and the increasing Na⁺ concentration might depolarize the nerve terminal. If either of these is true, the possibility exists that the bumetanide-sensitive Na⁺-K⁺-2Cl cotransporter also senses the K⁺ concentration in the portal vein. The goal of the present study was to examine this hypothesis. Urinary K⁺ excretion responses to intraportal KCl infusion were also examined.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The experiments were performed on male Sprague-Dawley rats weighing 300-360 g (n = 43). The animals were maintained in accordance with the Guiding Principles for Care and Use of Animals in the Field of Physiological Science of the Physiological Society of Japan. Rats were maintained at constant humidity (60 ± 5%), temperature (23 ± 1°C), and light cycle (0700-1900) and fed normal rat diet (CE-2, Clea Japan; approximate daily K⁺ intake was 4-5 meq/day). All experiments were performed during the light hours. Before the experiments, rats were fasted overnight, but water was available throughout the food-deprivation period.

HANA recording. The rats (n = 19) were anesthetized with pentobarbital sodium (50 mg/kg ip), and a venous catheter was inserted into the inferior vena cava via the femoral vein for infusion of any required supplemental doses of anesthetic. By means of central laparotomy, a portal venous catheter was placed ~5 mm from the liver via the mesenteric vein. The periarterial hepatic nerve was isolated, and two stainless steel electrodes (model 7901, A-M Systems, Everett, WA) were placed around the nerve. The nerve and electrodes were covered and held together with silicone gel (Semicosil 932 A and B, Wacker Chemie, München, Germany), then the proximal side of the hepatic nerve was ligated, and afferent nerve activity was recorded. The electrical activity recorded from the hepatic nerve was amplified using a 50-Hz to 1-kHz band-pass filter (model AVB-10, Nihon Kohden, Tokyo, Japan). The amplifier output was rectified by an absolute value circuit, and the rectified signal was integrated arithmetically. The original waveform and the rectified signal were sampled using an analog-digital converter (mode MP100, Biopac Systems, Goleta, CA) at a rate of 100 samples/s.

Nineteen rats were divided into three groups. In group 1 (n = 6), while HANA was being measured, hypertonic NaCl and LiCl solutions (both 750 mM) and three isotonic KCl + NaCl solutions (25 mM KCl + 125 mM NaCl, 50 mM KCl + 100 mM NaCl, and 100 mM KCl + 50 mM NaCl) were injected as bolus doses of 0.1, 0.2, 0.5, and 1.0 ml/kg body wt via the portal venous catheter. The order of injection of the solutions was randomized. In group 2 (n = 6), while HANA was being measured, a solution of 50 mM KCl + 100 mM NaCl was injected as a bolus into the portal vein at doses of 0.1, 0.2, 0.5, and 1.0 ml/kg body wt. This set of injections was repeated four times. In group 3 (n = 7), HANA responses to injection of 50 mM KCl + 100 mM NaCl were examined after portal venous infusion of graded doses of bumetanide (Sigma Chemical; 0, 3, 30, and 100 µmol/kg) dissolved in an 80% dimethyl sulfoxide-saline mixture (Nakarai Tesque, Kyoto, Japan); the KCl + NaCl solution was injected 2-5 min after bumetanide infusion. In group 1, to quantitate HANA responses, the peak value induced by injection of 1.0 ml/kg of hypertonic NaCl solution was taken as the 100% level, and the responses were plotted as a percentage. In group 2, the peak value induced by injection of 1.0 ml/kg of 50 mM KCl + 100 mM NaCl during the first set of injections was taken as the 100% level. In group 3, injection of 1.0 ml/kg of 50 mM KCl + 100 mM NaCl after pretreatment with vehicle not containing bumetanide was taken as the 100% level.

Renal K⁺ excretion. The rats (n = 24) were anesthetized with pentobarbital sodium (50 mg/kg ip), and two venous catheters were inserted into the inferior vena cava via the femoral veins for blood sampling, administration of supplemental doses of anesthetic, and continuous infusion of saline. By means of central laparotomy, a portal venous catheter was placed ~5 mm from the liver via the mesenteric vein. A Silastic catheter was inserted into each ureter for urine sampling. The catheters were exteriorized, and the incision was closed. To obtain a constant volume of urine, physiological saline (50 µl · kg¹ · min¹) was continuously infused via the vena caval catheter. Urine was collected via the ureter catheters at 30-min intervals. After a stabilization period of 1-2 h, a 1.5-h control period was started. At the end of the control period, the saline infusion was stopped and 50 mM KCl + 100 mM NaCl was infused at a rate of 50 µl · kg¹ · min¹ for 30 min via the portal venous (n = 8) or vena caval catheter (n = 8), after which saline infusion was started for a 1-h recovery period. Venous blood samples (300 µl) for determination of plasma K⁺, Na⁺, and Cl concentrations were taken at the end of the control and KCl + NaCl infusion periods and after 30 min of recovery. Portal venous infusion experiments were also performed in rats after severing the periarterial hepatic nervous plexus (n = 8).

All values are presented as means ± SE. The data in Figs. 1-3 were analyzed by one-way ANOVA, with the KCl concentration, number of repetitions, and doses of bumetanide as a factor. The data in Figs. 4-6 were analyzed by repeated-measure ANOVA. If the F ratio indicated a statistical significance, a post hoc test was applied to compare between-group and within-group means. The significance level of the post hoc comparisons was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

HANA response. Figure 1, left, shows typical individual changes in HANA in response to intraportal injection of hypertonic NaCl and isotonic KCl + NaCl solutions. An increase in HANA occurred with short latency and lasted a few seconds, then returned to the baseline. The summarized data of peak responses for six rats are presented in Fig. 1, right. The nerve did not respond to hypertonic LiCl, but responded to hypertonic NaCl or isotonic NaCl + KCl solutions in a volume-dependent manner. Furthermore, the increase in HANA was greater with increasing K⁺ concentration [F(2,87) = 24.59, P < 0.0001]. A series of 50 mM KCl + 100 mM NaCl injections, 0.1, 0.2, 0.5, and 1.0 ml/kg, was then repeated four times (Fig. 2). Repetition of injection did not affect the HANA response [F(3,116), P = 0.8222]. Figure 3 shows the effects of bumetanide pretreatment on the HANA response to intraportal injection of 50 mM KCl + 100 mM NaCl. Pretreatment with bumetanide significantly suppressed the HANA response in a dose-dependent manner [F(3,136) = 24.868, P < 0.0001].

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Left: original records illustrating hepatic afferent nerve activity (HANA) responses to intraportal bolus injection (arrowheads) of 750 mM NaCl, 25 mM KCl + 125 mM NaCl, 50 mM KCl + 100 mM NaCl, or 100 mM KCl + 50 mM NaCl. Summarized data for 6 rats are shown at right . , 750 mM NaCl; , 750 mM LiCl; , 25 mM KCl + 125 mM NaCl; , 50 mM KCl + 100 mM NaCl; , 100 mM KCl + 50 mM NaCl. Solutions were injected as a bolus into portal vein. * P < 0.05: responses to other combinations of KCl + NaCl are significantly different from those to injection of 25 mM KCl + 125 mM NaCl.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Summarized data for 6 rats showing effects of repeated injections on HANA responses to intraportal injections of 50 mM KCl + 100 mM NaCl solution. , first series of injections; , second series of injections; , third series of injections; , fourth series of injections.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Left: original records illustrating effects of pretreatment with graded doses of bumetanide (0, 3, 30, or 100 µmol/kg into portal vein) on HANA responses to intraportal bolus injection of 50 mM KCl + 100 mM NaCl (arrowheads). Summarized data for 7 rats are shown at right . , 0 µmol/kg (vehicle); , 3 µmol/kg; , 30 µmol/kg; , 100 µmol/kg of bumetanide. * P < 0.05: responses significantly different from those with vehicle pretreatment.

Urinary K⁺ excretion. Figure 4 shows the effect of infusion with 50 mM KCl + 100 mM NaCl on plasma K⁺, Na⁺, and Cl concentrations. In both portal venous and inferior venous infusion groups, the infusion did not alter the plasma K⁺, Na⁺, or Cl concentration. In the hepatic-denervated group, the infusion did not alter the plasma K⁺ or Na⁺ concentration; however, it elicited a small but significant increase in the plasma Cl concentration at 30 min after the infusion. There was no difference in the plasma K⁺, Na⁺, and Cl concentration responses among the three groups [F(2,42) = 0.198, P = 0.8217; F(2,42) = 0.419, P = 0.6628; and F(2,42) = 0.410, P = 0.6687, respectively] and no significant interaction between groups and time [F(4,42) = 0.369, P = 0.8293; F(4,42) = 0.561, P = 0.6919; and F(4,42) = 1.096, P = 0.3709, respectively].

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Plasma K+, Na+, and Cl concentration and urine volume responses to infusion of 50 mM KCl + 100 mM NaCl solution into portal vein (), inferior vena cava (), or portal vein in hepatic-denervated rats (). * P < 0.05: response is significantly different from that in control.

In the portal venous infusion group, urine volume and urinary K⁺ concentration tended to increase, but the effect did not reach statistical significance [F(3,28) = 1.812, P = 0.1679 and F(3,28) = 2.764, P = 0.0605, respectively]. However, the product of these two variables, urinary K⁺ excretion, significantly increased from 38.9 ± 6.7 to 72.8 ± 9.1 µeq · kg¹ · 30 min¹ in response to infusion and remained elevated until 60 min after infusion (Fig. 5). The response was immediate and powerful, because 45 ± 10% of the loaded K⁺ was excreted during the 30-min infusion period and 154 ± 18% of the loaded K⁺ was excreted during 90 consecutive min (Fig. 6). The urinary K⁺ excretion responses in the inferior vena caval infusion group and hepatic-denervated group were quite different from those in the portal venous infusion group. Kaliuresis during infusion, seen in the portal venous infusion group, was not observed in the other two groups. In these groups, urinary K⁺ excretion tended to increase at 30 or 60 min after infusion, but the effect did not reach statistical significance [F(3,28) = 1.32, P = 0.2876 and F(3,28) = 2.354, P = 0.0934, respectively]. In these groups, only 40-60% of the loaded K⁺ was excreted during the 90 consecutive min.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Top: urinary K+, Na+, and Cl (UK+, UNa+, UCl, respectively) concentration responses to infusion of 50 mM KCl + 100 mM NaCl into portal vein (), inferior vena cava (), or portal vein in hepatic-denervated rats (). Bottom: UK+, UNa+, and UCl excretion responses to infusion of 50 mM KCl + 100 mM NaCl into portal vein (), inferior vena cava (), or portal vein in hepatic-denervated rats (). * P < 0.05: response is significantly different from that in control.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Cumulative K+ excretion as a percentage of total loaded K+. , portal venous infusion; , inferior vena caval infusion; , portal venous infusion in hepatic-denervated rats. * P < 0.05: responses are significantly different from those in inferior vena caval infusion group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The major findings of the present study are as follows. 1) The periarterial hepatic afferent nerve responded to increases in the portal venous K⁺ concentration in a dose-dependent manner. 2) The increase in HANA was attenuated by bumetanide pretreatment. 3) Stimulation of the hepatoportal K⁺-sensor mechanism elicited kaliuresis. These observations support the hypothesis that a hepatoportal bumetanide-sensitive K⁺-sensor mechanism controls renal K⁺ excretion.

We measured HANA from the periarterial hepatic nervous plexus. This nerve responded to intraportal injection of both isotonic KCl + NaCl solution and hypertonic NaCl solution, but not of hypertonic LiCl solution. Furthermore, a previous study from our laboratory demonstrated that this nerve responds to intraportal injection of hypertonic NaHCO₃ but not hypertonic mannitol or isotonic NaCl solution (11). These results suggest that the rat periarterial hepatic nerve is sensitive to the portal venous Na⁺ and K⁺ concentration, but not to the Cl concentration, osmolality, or volume. The hepatoportal Na⁺-sensor mechanism has a significant role in postprandial natriuresis and in maintenance of Na⁺ balance (8, 10, 21). However, the physiological significance of the hepatoportal K⁺-sensor mechanism had been unclear.

The increase in HANA induced by intraportal KCl injection was suppressed by inhibition of the Na⁺-K⁺-2Cl cotransporter by bumetanide. This was not due to a deterioration of the nerve and/or accumulation of injected KCl, because four repeated injections did not affect the HANA response. Bumetanide also suppresses the HANA response induced by intraportal hypertonic NaCl injection (7). These results suggest that the Na⁺-K⁺-2Cl cotransporter might be involved in both the hepatoportal Na⁺- and K⁺-sensor mechanisms, which transduce Na⁺ and K⁺ concentrations into HANA. Several lines of functional evidence support the existence of the Na⁺-K⁺-2Cl cotransporter in the basolateral membrane of the hepatocyte (2, 4, 5, 23). The Na⁺-K⁺-2Cl cotransporter can modulate hepatocellular Na⁺ and K⁺ uptake and regulates the hepatocellular volume. If this is the case for the Na⁺- and K⁺-sensor mechanism(s), it is possible that an increase in hepatocellular volume induced by Na⁺ and/or K⁺ influx via the Na⁺-K⁺-2Cl cotransporter might stimulate the connected nerve terminals. The other possibility is that the nerve terminal itself might bear the Na⁺-K⁺-2Cl cotransporter (12), and the increasing Na⁺ or K⁺ concentration might depolarize the nerve terminal. The Na⁺-K⁺-2Cl cotransporter-mediated cellular depolarization is demonstrated in the macula densa cells (6, 20). These hypotheses raise the question whether different hepatoportal sensor mechanisms sense the Na⁺ and K⁺ concentrations.

Previous studies from our laboratory demonstrated that stimulation of hepatoportal sensor mechanisms by high-NaCl diet or intraportal injection of hypertonic NaCl solution causes natriuresis (8, 21). In the present study, however, natriuresis was not seen after stimulation of the hepatoportal sensor mechanisms by intraportal infusion of KCl. Thus it is possible that different hepatoportal sensor mechanisms can sense the Na⁺ and K⁺ concentration and operate different regulatory mechanisms. Because we did not record single fiber spikes, we cannot say for certain whether different fibers respond to the concentrations of Na⁺ and K⁺. Tyryshkina et al. (22) recorded single fiber spikes from the cat periarterial hepatic plexus and found that KCl-sensitive fibers did not respond to hypertonic NaCl and vice versa. If this is the case in the present study, it is possible that the Na⁺-K⁺-2Cl cotransporter might be involved in the common pathway of signal transduction but not involved in the individual sensor mechanisms themselves. However, species differences have to be considered, because the cat periarterial hepatic nervous plexus contains osmosensitive nerve fibers (22), whereas, in the present study, the rat periarterial hepatic nervous plexus did not respond to hypertonicity. These possibilities should be investigated in future studies.

The conventional theory regarding K⁺ excretion is that K⁺ intake increases the plasma K⁺ concentration, which stimulates aldosterone release, then both the aldosterone and the increased K⁺ concentration independently stimulate renal K⁺ excretion. Actually, aldosterone has a profound kaliuretic effect when the plasma K⁺ concentration is above its normal value; however, when the plasma K⁺ concentration is within the normal range, aldosterone has no significant effect on K⁺ excretion (18). Furthermore, when the plasma K⁺ concentration is within the normal range, very slight changes in plasma K⁺ appear to have no significant effect on K⁺ excretion (24). Calò et al. (1) demonstrated that ingestion of potassium citrate induces a significant kaliuresis without a significant increase in the plasma K⁺ and plasma aldosterone concentrations. To explain the kaliuresis, they suggested an enteric reflex with a K⁺-sensor mechanism located in the gastrointestinal tract or the hepatoportal region. In support of such a mechanism, Rabinowitz et al. (16) found the kaliuresis after rumen infusion of KCl to be several times greater than that seen during intravenous infusion. Furthermore, Rabinowitz (13) developed a mathematical model of K⁺ homeostasis specifically to incorporate an enteric receptor-reflex kaliuresis, a model whose simulations produced results remarkably similar to those of the present study. In the present study, intraportal KCl infusion elicited a greater kaliuresis than intravenous KCl infusion, and kaliuresis was attenuated by severing the periarterial hepatic nervous plexus, suggesting that it was mediated by receptors located in the hepatoportal region.

Finkinstein and co-workers in Russia presented somewhat similar results obtained from experiments on dogs in support of hepatoportal reflex control of renal K⁺ excretion. Their studies were extensively reviewed by Rabinowitz and Aizman (14, 15). They found that brief injection of 0.1 µmol/kg body wt of KCl into the portal vein produced kaliuresis, which was prevented by subdiaphragmatic vagotomy or hypophysectomy. Although efferent pathways of the kaliuresis were not determined in the present study, several candidates for the efferent limb of hepatoportal K⁺ receptor-mediated kaliuresis have been proposed by Rabinowitz and Aizman (14, 15), i.e., vasopressin, oxytocin, ACTH-activated release of aldosterone or cortisol, or undetermined novel kaliuretic factors. These possibilities have to be determined in future studies.

In conclusion, our data support the hypothesis that the hepatoportal bumetanide-sensitive K⁺-sensor mechanism can detect the portal venous K⁺ concentration. When this mechanism is stimulated, kaliuresis occurs via the periarterial hepatic nervous plexus. Thus the hepatoportal bumetanide-sensitive K⁺-sensor mechanism may play an important role in regulating extracellular K⁺ homeostasis.

Perspectives