Norepinephrine innervation of the supraoptic nucleus contributes to increased copeptin and dilutional hyponatremia in male rats

Abstract

Dilutional hyponatremia associated with liver cirrhosis is due to inappropriate release of arginine vasopressin (AVP). Elevated plasma AVP causes water retention resulting in a decrease in plasma osmolality. Cirrhosis, in this study caused by ligation of the common bile duct (BDL), leads to a decrease in central vascular blood volume and hypotension, stimuli for nonosmotic AVP release. The A1/A2 neurons stimulate the release of AVP from the supraoptic nucleus (SON) in response to nonosmotic stimuli. We hypothesize that the A1/A2 noradrenergic neurons support chronic release of AVP in cirrhosis leading to dilutional hyponatremia. Adult, male rats were anesthetized with 2–3% isoflurane (mixed with 95% O₂/5% CO₂) and injected in the SON with anti-dopamine β-hydroxylase (DBH) saporin (DSAP) or vehicle followed by either BDL or sham surgery. Plasma copeptin, osmolality, and hematocrit were measured. Brains were processed for ΔFosB, dopamine β-hydroxylase (DBH), and AVP immunohistochemistry. DSAP injection: 1) significantly reduced the number of DBH immunoreactive A1/A2 neurons (A1, P < 0.0001; A2, P = 0.0014), 2) significantly reduced the number of A1/A2 neurons immunoreactive to both DBH and ΔFosB positive neurons (A1, P = 0.0015; A2, P < 0.0001), 3) reduced the number of SON neurons immunoreactive to both AVP and ΔFosB (P < 0.0001), 4) prevented the increase in plasma copeptin observed in vehicle-injected BDL rats (P = 0.0011), and 5) normalized plasma osmolality and hematocrit (plasma osmolality, P = 0.0475; hematocrit, P = 0.0051) as compared with vehicle injection. Our data suggest that A1/A2 neurons contribute to increased plasma copeptin and hypoosmolality in male BDL rats.

INTRODUCTION

Hyponatremia is a common electrolyte imbalance in patients hospitalized for liver cirrhosis (1, 2). When patients with cirrhosis have a plasma sodium concentration below 135 mEq/L, it is associated with increased risk of complications including ascites (2, 3), hepatic encephalopathy (4, 5), and spontaneous bacterial infections (4, 5). These patients also have longer hospital length of stay (6), higher health care costs (6), and increased mortality (7). The cause of hyponatremia in liver cirrhosis has been linked to an inappropriate release of arginine vasopressin (AVP) (8). Patients with cirrhosis have improved plasma sodium, reduced ascites, and increased urine output after administration of AVP antagonists (9–12). Blocking AVP receptors in rats also improves hyponatremia in syndrome of inappropriate secretion of antidiuretic hormone (13).

AVP is synthesized by the magnocellular neurosecretory cells (MNCs) of the supraoptic and paraventricular nucleus (SON and PVH) and certain accessory nuclei in the anterior hypothalamus and released into systemic circulation from nerve endings in the posterior pituitary (14–16). AVP synthesis begins with a precursor polypeptide called preprovasopressin (17). During axonal transport to the posterior pituitary, preprovasopressin is processed into AVP, neurophysin II, and copeptin (18). When activated by osmotic or hemodynamic stimuli, vasopressin-releasing MNCs secrete all three peptides in equimolar concentrations into systemic circulation (19). The plasma concentration of copeptin correlates with plasma AVP and responds similarly to changes in plasma osmolality and blood volume (20, 21). In addition, copeptin has a longer half-life and is more stable as compared with AVP. This makes it a suitable and sensitive surrogate marker for AVP in various clinical settings (19, 22, 23). However, copeptin may not be an ideal diagnostic marker for hyponatremia in some cases of syndrome of inappropriate antidiuretic hormone (SIADH) secretion since other factors including somatic stress could lead to nonspecific increase in plasma copeptin concentration (23). A previous study from our group demonstrated similar elevations of copeptin and AVP associated with hyponatremia and hypoosmolality in male bile-duct ligated (BDL) rats (24). For these reasons, plasma copeptin concentrations were used as a surrogate marker for plasma AVP in this study.

A decrease in blood pressure and blood volume or increased plasma osmolality can increase AVP release, which increases water reabsorption via V2 receptors in the kidneys. Normally, a decrease in plasma osmolality inhibits AVP release (25–27). However, in cirrhosis, there is an increase in plasma AVP concentration in the face of a decreased plasma osmolality that contributes to excessive water retention and dilutional hyponatremia (28). The mechanisms supporting the inappropriate release of AVP in cirrhosis have not been determined. In advanced cirrhosis, there is an increase in vascular resistance in the cirrhotic liver resulting in portal hypertension, ascites formation, mesenteric vasodilation due to release of vasodilators, and increased pooling of blood in the splanchnic system (28–31). These changes in fluid distribution lead to a decrease in effective central vascular blood volume (32, 33) even though total blood volume is expanded (31, 34).

The SON receives afferents from noradrenergic neurons in the A1 region of the caudal ventrolateral medulla (CVLM) and the A2 region of the nucleus tractus solitaris (NTS) (35–43). These regions regulate AVP and oxytocin release to a number of visceral physiological challenges as well as pain (27). Pressure and volume receptors found in the aortic arch and heart sense a decrease in blood pressure and volume that is relayed to the A1 neurons in the CVLM and A2 neurons in the dorsal brainstem (26, 44). Changes in body fluid homeostasis including a decrease in central blood volume associated with cirrhosis could activate A1 and A2 neurons which would, in turn, contribute to AVP release in the face of decreasing plasma osmolality.

The aim of this study was to investigate the role of noradrenergic innervation of the SON from A1 and A2 neurons in the inappropriate secretion of AVP in an animal model of liver cirrhosis. In these studies, liver cirrhosis is modeled in rats using the bile-duct ligation (BDL) method. In this model, the common bile duct that drains bile from the liver to the intestines is ligated leading to obstructive cholestasis and cirrhosis (45, 46). The rats develop ascites, decreased plasma osmolality, and an elevated circulating AVP, comparable with that seen in patients with cirrhosis (28, 47, 48). They also have systemic hypotension (34) despite increased cardiac output (34) and expanded blood volume (31).

Previous studies from our laboratory showed that female BDL rats do not show an increase in copeptin secretion or an increase in SON AVP neuron activation as has been observed in male BDL rats (49). Instead, there was a significant increase in plasma oxytocin levels and there was an absence of dilutional hyponatremia (49). These findings indicate a sex difference in the effect of BDL surgery on plasma AVP concentration. Because the goal of this study was to determine the contributions of norepinephrine innervation of the SON in AVP or copeptin release, it focused on males. Previous studies have suggested A2 noradrenergic neurons from the NTS project mainly to the oxytocin neurons (50) and could play a role in the increased plasma oxytocin levels in female BDL rats. Further investigation will be required to determine if the A1 and A2 neurons play a role in the changes seen in female BDL rats.

We hypothesize that activation of A1 and A2 neurons support the increased stimulation of AVP secretion in liver cirrhosis leading to dilutional hyponatremia. This hypothesis was tested using anti-dopamine β-hydroxylase (DBH) saporin (DSAP), a cytotoxin that specifically targets noradrenergic neurons, to lesion the noradrenergic innervation of the SON that is primarily derived from A1 and A2 neurons in the brainstem (35) to investigate their effect on plasma copeptin, plasma osmolality, and activation of AVP SON neurons. Also, immunohistochemistry for the transcription factor ΔFosB, which is a splice product of FosB with a longer half-life, was used as a marker for neuronal activation (51–53).

MATERIALS AND METHODS

Animals

In these studies, 6–12 adult male Sprague–Dawley rats (250–300 g, Charles River Laboratories) per group were used. The rats were individually housed in a temperature-controlled (25°C) room on a 12:12 light/dark cycle with light onset at 0700. Food and drinking water were available ad libitum. Rats that received bile duct ligation surgery were monitored for the development of infections or ascites. A nonsteroidal anti-inflammatory drug carprofen (Rimadyl 2 mg) was given orally postsurgery for the management of pain and inflammation. Experiments were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the University of North Texas Health Science Center Institutional Animal Care and Use Committee.

Stereotaxic Surgery

Rats were anesthetized with 2–3% isoflurane (mixed with 95% O₂-5% CO₂), and their scalps were shaved and disinfected with iodine and alcohol. Each rat was secured in a stereotaxic apparatus (David Kopf, Tujunga, CA) using ear bars. Isoflurane anesthesia was maintained using a commercially available vaporizer (Kent Scientific, Torrington, CT) that was connected to a nose holder adapter on the stereotaxic frame. The skull surface was exposed and leveled between bregma and lambda as previously described (54). Each rat was bilaterally injected into the SON using the following coordinates from bregma: 1.4 mm posterior, ±1.4 mm lateral, and 9.1 mm ventral (55). A hole was drilled into the skull at these coordinates and a 30-gauge injector lowered to the SON. At each site, 200–300 nL of anti-DBH saporin (50 ng/500 nL, Advanced Targeting Systems, San Diego, CA) or Vehicle (0.14 M sodium chloride, 0.003 M potassium chloride, 0.002 M potassium phosphate, 0.01 M sodium phosphate, pH 7.4) was injected. After the injection, the injector was left in place for 5 min then slowly withdrawn. A piece of gel foam was used to fill the hole in the skull. The incision site was closed with absorbable sutures. Each animal was allowed to recover fully from anesthesia before being returned to their home cage.

Bile Duct Ligation Surgery

Two weeks after the stereotaxic surgery, the rats were anesthetized with 2–3% isoflurane mixed with oxygen using a commercially available vaporizer (Kent Scientific). For each rat, the abdomen was shaved, cleaned, and disinfected with iodine followed by alcohol. A midline incision was made to access the abdominal cavity. The common bile duct was located and cauterized between two ligatures as previously described (56, 57). Sham rats received the same surgical procedure except their bile duct was neither ligated nor cauterized. The rats were allowed to fully recover before they were returned to their home cages. All rats were monitored for signs of jaundice and ascites. Rats that showed signs of morbidity, ascites greater than 10% of the body weight, or lost more than 10% of their presurgery weight were euthanized with inactin (100 mg/kg ip, St. Louis, MO). At the time of euthanasia, livers were removed and weighed to calculate their liver weight-to-body weight ratio (Table 1).

Table 1.

Effects of bile duct ligation and SON injections of vehicle or anti-DBH saporin on plasma osmolality, hematocrit, and liver weight-to-body weight ratio

	Veh/Sham (7)	Veh/BDL (6)	DSAP/Sham (7)	DSAP/BDL (8)
Osmolality, mosmol/kgH2O	301.9 ± 2.8*	290.9 ± 2.0	301.8 ± 2.8*	300.4 ± 1.1*
Hematocrit, %	43.5 ± 0.8*	39.8 ± 1.6	43.7 ± 0.5*	45.8 ± 1.2*φ
Liver weight, g/Body wt, g	0.038 ± 0.001*	0.078 ± 0.005	0.044 ± 0.003*	0.060 ± 0.005*+

Data are represented as means ± SE. Total number of rats for each group are in parenthesis (φ indicates an n of 6 rats). Data were analyzed by separate two-way ANOVAs and Student–Newman–Keuls post hoc tests. BDL, bile duct ligation; DBH, dopamine β-hydroxylase; DSAP, DBH saporin; SON, supraoptic nucleus.

*P < 0.05 vs. Veh/BDL; +P < 0.05. DSAP/sham by Student–Newman–Keuls tests.

Immunohistochemistry

Four weeks after BDL or sham surgery, rats were deeply anesthetized with inactin (thiobutabarbital sodium salt hydrate; 100 mg/kg ip) and perfused transcardially with 0.1 M phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (4% PFA) in 0.1 M PBS. After the perfusion, brains were carefully removed from the skulls, placed in 4% PFA for 24 h then transferred to 30% sucrose until dehydrated. The brains were sectioned at −20°C at a thickness of 40 µm using a cryostat. They were grouped into forebrain sections containing the SON and hindbrain sections that contained the NTS and CVLM.

Forebrain sections were processed for ΔFosB and AVP using rabbit monoclonal anti-ΔFosB antibody (1:1,000; Cat. No. D3S8R, Cell signaling Technology, Danvers, MA) and guinea pig polyclonal anti-(Arg8)-Vasopressin antibody (1:5,000; Cat. No. T-5048, Peninsula Labs, San Carlos, CA) respectively. Hindbrain sections were processed for ΔFosB as described earlier and DBH using a mouse monoclonal anti-dopamine β-hydroxylase antibody (1:1,000; Cat. No. MAB 308, Millipore Sigma, Temecula, CA). Both forebrain and hindbrain sections were washed with 0.1 M PBS and incubated with horse serum. After rinsing, the sections were placed in solution containing ΔFosB primary antibodies for 2 days at 4°C. At the end of the incubation period, sections were rinsed with 0.1 M PBS and incubated with the secondary antibody, biotinylated goat anti-rabbit IgG antibody (1:100; Cat. No. BA-1000, Vector Laboratories, Burlingame, CA). Vectastain ABC-HRP Kit Peroxidase (Cat. No. PK-4000, Vector Laboratories) and 3,3′-diaminobenzidine tetrahydrochloride (DAB, Cat. No. D5905, Sigma Aldrich) were used to visualize ΔFosB. The sections were rinsed and incubated for 2 days at 4°C in the second primary antibody against AVP for the forebrains and DBH for the hindbrains. After the incubation period, Cy-3 conjugated anti-mouse IgG (1:1,000, Cat. No. 715-165-151, Jackson ImmunoResearch Laboratories, PA) was used to visualize dopamine β-hydroxylase and a Cy-2 conjugated anti-guinea pig IgG (1:1,000, Cat. No. 706-225-148, Jackson ImmunoResearch Laboratories, PA) for AVP. Sections were rinsed again in PBS and mounted on gelatin-coated slides. Permount mounting medium (Cat. No. SP15-100, Fisher Scientific) was applied to the slides and they were cover slipped. Slides were allowed to dry for 2 days. Sections were examined using light and immunofluorescence microscopy. Forebrain sections were examined for cells immunoreactive to AVP using immunofluorescence microscopy and ΔFosB DAB staining using light microscopy. For hindbrain sections, DBH immunofluorescence and ΔFosB DAB staining were examined. All colocalization analysis and counts were done using the ImageJ software. For the forebrain sections, the optic chiasm was used as a visual landmark. For each rat, five forebrain sections between 0.92 mm and 1.60 mm caudal to bregma (55) were collected and the number of AVP and ΔFosB immunoreactive cells was counted. The central canal and pyramidal tract were used as visual landmarks for the hindbrain sections. Five caudal hindbrain sections per rat between 14.08 mm and 14.60 mm caudal to bregma (55) were selected and the number of DBH and ΔFosB immunoreactive A1 and A2 cells per section was counted. The counts from each brain region were totaled per rat, with 6–12 rats/group. For colocalization analysis, images were converted to 8-bit color and uniformly adjusted for brightness and contrast. The images of ΔFosB staining were pseudo-colored green whereas images of DBH and AVP staining images were pseudo-colored red. Composite images for the SON sections were created by merging AVP and ΔFosB images and that for the hindbrain sections by merging DBH and ΔFosB images. Colocalization was determined by identifying profiles containing either DBH or AVP, with ΔFosB in the composite images. The presence of a pseudo-colored green nuclear staining surrounded by a pseudo-colored red cytosolic staining corresponding to previously identified immunopositive profiles were counted as colocalization. The numbers of profiles with colocalization were totaled for each section and combined for each rat.

Plasma Measurements

Before each rat was transcardially perfused, a syringe was used to collect blood from the left ventricle of the heart into a 2-mL centrifuge tube and Vacutainer EDTA tube containing 12 mg of EDTA. A pair of heparinized capillary tubes (Fisher Scientific, Hampton, NH) was filled with blood from the 2-mL centrifuge tube and centrifuged using a microcapillary centrifuge (Adams Micro-Hematocrit II Centrifuge, Clay Adams, Parsippany, NJ). The hematocrit was then measured with a Micro-Hematocrit capillary tube reader (Lancer, St. Louis, MO). The remaining blood was centrifuged at 1,600 g for 5 min at room temperature and the plasma was collected. The plasma osmolality was measured using a vapor pressure osmometer (Wescor, Logan, UT).

Aprotinin, a protease inhibitor (0.6 TIU/mL of blood; Phoenix Pharmaceuticals, Inc., Burlingame, CA) was added to the blood in the EDTA tubes followed by centrifugation at 4,000 rpm for 20 min at 4°C. The resulting plasma was collected, and aliquots made for measuring copeptin using a rat ELISA kit according to the manufacturer’s instructions (MyBiosource, San Diego, CA). A four-parametric logistic analysis was used to quantify the copeptin concentration. Copeptin measurements varied from one batch of ELISA kits to the other. To address this issue, the copeptin concentrations were standardized to the respective controls.

Experimental Groups and Statistics

In all experiments for this study, rats were divided into four groups as follows: 1) BDL rats injected with anti-DBH saporin (DSAP/BDL, n = 12 rats), sham rats injected with anti-DBH saporin (DSAP/sham, n = 8 rats), BDL rats injected with vehicle (Veh/BDL, n = 6 rats), and Sham rats injected with vehicle (Veh/Sham, n = 8 rats). Forebrain sections from each rat were examined to identify the injector canula track and this analysis was used to determine the accuracy of the SON injections. Based on this, four rats were eliminated from the DSAP/BDL group and one from the DSAP/Sham group (Fig. 1).

Figure 1.

Light microscopy images showing cannula tracks for injection that hit the supraoptic nucleus (SON; A), injection that missed the SON (B). Cannula tracks are shown with yellow arrows. OC, optic chiasm.

Statistical analyses and graphs were prepared using GraphPad Prism (v.9.4.0) (GraphPad Software, San Diego, CA). For the immunohistochemistry, the number of AVP, DBH, and ΔFosB-positive cells was manually counted using ImageJ software. Images were counted by a second experimenter blinded to the treatment conditions to confirm the results. Where significant differences were obtained, the counts were repeated. Counts were totaled for each rat. Data were reported as means ± SE. All data were analyzed using two-way ANOVA with injection (vehicle vs. DSAP) as the first factor and surgery (sham vs. BDL) as the second factor along with Student–Newman–Keuls multiple-comparison post hoc tests (SNK) using GraphPad Prism (v.9.4.0). Copeptin data from BDL rats were standardized to their respective controls. The number of animals per group was determined by power analysis and effect size calculated from our previously published work (24, 49, 57, 58) and preliminary data using SigmaPlot 12.0 (Systat Software, Inc., San Jose, CA). For all tests, a P value less than 0.05 was considered statistically significant. For all tables and graphs, the n represents the number of animals used in the analysis.

RESULTS

Liver Weight-to-Body Weight Ratio

BDL was used to model liver cirrhosis in rats. In this model, an increase in liver weight-to-body weight ratio is a widely used marker to confirm successful liver cirrhosis (57). The liver weight-to-body weight ratio was significantly increased in BDL rats as compared with their respective sham-ligated groups. Two-way ANOVA showed a significant interaction indicating that liver weight-to-body weight ratio was affected by the injection and BDL [injection × surgery: F(1,24) = 10.20, P = 0.0039; Table 1]. Student–Newman–Keuls post hoc comparisons between the factors revealed that BDL significantly increased liver-to-body weight ratio in all BDL rats compared with their respective sham controls (vehicle: P < 0.00001, DSAP: P = 0.0477). This was consistent with previous reports that BDL increases liver weight-to-body weight ratio as compared with sham (57). Also, BDL rats that received DSAP had significantly lower liver weight-to-body weight ratios as compared with those that received vehicle injection (SNK, P < 0.0014).

Plasma Osmolality, Hematocrit, and Plasma Copeptin Measurements

Dilutional hyponatremia is normally associated with a decrease in hematocrit values suggesting plasma volume expansion and a decrease in plasma osmolality and hematocrit. Both plasma osmolality and hematocrit were affected by injection and BDL as indicated by significant interaction terms [plasma osmolality: F(1,24) = 4.635, P = 0.0416; hematocrit F(1,23) = 8.327, P = 0.0083]. Plasma copeptin concentrations also were affected by both injection and BDL [injection × surgery: F(1,24) = 8.119, P = 0.0089]. Student–Newman–Keuls post hoc analysis showed that DSAP injection in BDL rats prevented the increase in copeptin (SNK, P = 0.0005, Fig. 2) and normalized plasma osmolality (SNK, P = 0.0064) and hematocrit values (SNK, P = 0.0021) as compared with vehicle-injected BDL rats (Table 1).

Figure 2.

Plasma copeptin concentrations graphed as ratios of controls for each group. Data were analyzed by two-way, which demonstrated a significant interaction [F(1,24) = 8.119, P = 0.0089]. P values on the graphs indicate significant group differences based on Student–Newman–Keuls follow-up tests. Data are represented as means ± SE. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

DBH and ΔFosB Expression in A1 Cells

DSAP was injected into the SON to lesion the noradrenergic innervation (A1 and A2 cells) to the SON. To determine the efficacy of the lesions, the number of DBH immunoreactive A1 cells was counted. Also, the number of ΔFosB immunoreactive cells was counted since an increase in ΔFosB is associated with increased neuronal activation (59, 60). Two-way ANOVA revealed a significant difference in the number of DBH immunoreactive A1 cells in response to the injections [F(1,24) = 21.95, P < 0.0001] that was not influenced by BDL (Fig 3A and Fig. 4). In contrast, ΔFosB staining in the same region of the medulla was significantly affected by both injection and surgery [F(1,24) = 4.396, P = 0.0467]. Post hoc analysis showed that in vehicle-injected rats, BDL increased the number of ΔFosB immunoreactive A1 cells as compared with sham (SNK, P = 0.0003). Also, BDL rats that received DSAP had significantly reduced ΔFosB expression as compared with the vehicle BDL group (SNK, P = 0.0004, Fig. 3B and Fig. 4). To determine the number of activated A1 cells following BDL and DSAP injection, the number of DBH immunoreactive cells that was also immunoreactive to ΔFosB was counted. The results showed a significant interaction suggesting the activation of A1 cells was affected by both injection and surgery [injection × surgery: F(1,24) = 10.76, P = 0.0032]. Post hoc analysis revealed that in vehicle-injected rats, BDL surgery was associated with an increase in ΔFosB-positive A1 cells (SNK, P = 0.0003) and this effect was prevented by DSAP injections in the SON (SNK, P = 0.0004, Figs. 3C and Fig. 4).

Figure 3.

Graphs of the total numbers of dopamine β-hydroxylase (DBH; A), ΔFosB (B), and DBH and ΔFosB (C) positive cells in the A1 region of the caudal ventrolateral medulla. Data were analyzed by separate two-way ANOVAs. The P value on A represents a significant main effect of injection [F(1,24) = 21.95, P < 0.0001], P values on B and C indicate significant group differences based on Student–Newman–Keuls post hoc tests based on significant interactions [B: F(1,24) = 4.396, P = 0.0467; C: F(1,24) = 10.79, P = 0.0032]. Data are represented as means ± SE. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

Figure 4.

Representative immunofluorescence (dopamine β-hydroxylase, DBH), light microscopy (ΔFosB), and merged images of A1 neurons. Cells immunoreactive to both DBH and ΔFosB are shown with yellow arrows. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

DBH and ΔFosB Expression in A2 Cells

To determine the effects of BDL and DSAP injection on A2 cells, the number of ΔFosB and DBH immunoreactive cells was counted. The results of the two-way ANOVA demonstrated a significant main effect of DSAP on the numbers of DBH immunoreactive A2 cells [F(1,24) = 12.94, P = 0.0014] but no effect of BDL. Also, significant differences in the numbers of ΔFosB immunoreactive A2 cells were observed due to a significant main effect of injection [F(1,24) = 5.404, P = 0.0289 F; Fig. 5B and Fig. 6]. For both DBH staining and ΔFosB staining, DSAP appears to significantly decrease the numbers of cells independent of BDL. In contrast, the numbers of DBH immunoreactive cells that was also ΔFosB positive was affected. Two-way ANOVA showed a significant interaction of injection and BDL [F(1,24) = 21.84, P < 0.0001]. Post hoc analysis revealed a significant increase in the number of A2 cells immunoreactive to both ΔFosB and DBH in vehicle-injected BDL rats as compared with sham rats (SNK, P < 0.0001) and this was reduced by DSAP (SNK, P < 0.0001, Fig. 5C and Fig. 6).

Figure 5.

Total numbers of dopamine β-hydroxylase (DBH; A), ΔFosB (B), and DBH and ΔFosB (C) positive cells in the caudal nucleus tractus solitaris (NTS)/A2 region. Data were analyzed by separate two-way ANOVAs. The P values on A and B represent significant main effects of injection [A: F(1,24) = 12.94, P = 0.0014; B: F(1,24) = 5.404, P = 0.0289]. The P values on C indicate significant group differences based on Student–Newman–Keuls post hoc tests based on a significant interaction [F(1,24) = 21.84, P < 0.0001]. Data are represented as means ± SE. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

Figure 6.

Representative immunofluorescence (dopamine β-hydroxylase, DBH), light microscopy (ΔFosB), and merged images of A2 neurons. Cells immunoreactive to both DBH and ΔFosB are shown with yellow arrows. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

Anti-DBH Saporin Injection Reduces ΔFosB and AVP/ΔFosB Colocalization in the SON of BDL Rats

An increase in ΔFosB expression in the SON has been shown previously to contribute to a decrease in plasma osmolality in BDL rats (57). The results of the two-way ANOVA demonstrated a significant interaction suggesting the number of ΔFosB positive cells in the SON was affected by both injection and surgery [F(1,23) = 9.836, P = 0.0046]. Student–Newman–Keuls post hoc analysis showed that in vehicle-injected rats, BDL resulted in a significant increase in the number of ΔFosB immunoreactive cells as compared with sham surgery (SNK, P = 0.0007). This was significantly attenuated by DSAP injection in the SON (SNK, P = 0.0001, Fig. 7A and Fig. 8). The numbers of SON cells that were immunoreactive for both ΔFosB and AVP were examined to determine the effects of DSAP injection on SON AVP cell activation in association with the increased plasma copeptin concentration and hypoosmolality observed in BDL rats. Two-way ANOVA analysis showed that SON AVP cell activation was significantly influenced by surgery and the type of injection the rat received [injection × surgery: F(1,23) = 36.67, P < 0.0001]. Post hoc analysis showed that the numbers of ΔFosB positive SON AVP cells were significantly higher in vehicle-injected BDL rats as compared with the respective sham controls (SNK, P < 0.0001). However, injection of DSAP attenuated the number of ΔFosB positive AVP neurons associated with BDL (SNK, P < 0.0001, Fig. 7B and Fig. 8).

Figure 7.

Total numbers of ΔFosB (A) and arginine vasopressin (AVP; B) and ΔFosB-positive cells in supraoptic nucleus (SON). Data were analyzed by separate two-way ANOVA. P values on the graphs indicate significant group differences based on Student–Newman–Keuls post hoc tests following significant interactions [A: F(1,23) = 9.836, P = 0.0046; B: F(1,23) = 36.67, P < 0.0001]. Data are represented as means ± SE. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

Figure 8.

Representative light microscopy images of ΔFosB and merged images [arginine vasopressin (AVP) and ΔFosB] of supraoptic nucleus (SON) neurons. Cells immunoreactive to both AVP and ΔFosB are shown with yellow arrows. BDL, bile duct ligation; DSAP, dopamine β-hydroxylase saporin.

DISCUSSION

In normal physiology, the maintenance of body fluid homeostasis requires the coordination of autonomic and endocrine responses. The neurohypophyseal hormone, AVP plays an important role in maintaining blood volume by trafficking aquaporin-2 channels to the membranes of epithelia cells in the distal nephron to increase water reabsorption (61). An increase in plasma osmolality or a decrease in blood volume or blood pressure stimulates the secretion of AVP. In humans, a 10% decrease in plasma volume is enough to increase plasma AVP (16, 62). In patients with cirrhosis, there is increased pooling of blood in the splanchnic system that decreases central blood volume, and this is associated with increased plasma AVP concentration, water retention, and dilutional hyponatremia (28, 29).

In this study, we tested the contribution of the noradrenergic innervation of the SON to increased plasma copeptin concentration (a surrogate marker for plasma AVP) and hypoosmolality in a rat model of liver cirrhosis by injecting anti-DBH saporin, a noradrenergic-specific cytotoxin into the SON. There was a significant increase in plasma copeptin concentrations in vehicle-injected BDL rats as compared with sham operated controls. Also, decreases in plasma osmolality and hematocrit were observed in vehicle-injected BDL rats that are consistent with plasma volume expansion and hemodilution. Our results demonstrated that injections of DSAP in the SON of male rats prevented significant fluid retention associated with BDL as compared with those injected with vehicle. The BDL rats injected with anti-DBH saporin had significantly lower plasma copeptin concentrations as compared with BDL rats injected with vehicle. Furthermore, the decreases in plasma osmolality and hematocrit related to fluid retention normally seen in this model were also prevented by injections of DSAP in the SON of BDL rats. Thus, the DSAP injections in the SON of BDL rats prevented the changes in plasma copeptin, osmolality, and hematocrit. However, plasma copeptin concentration did not return to control levels. A reason for this result could be that in addition to the SON, copeptin is also released from the PVH and other accessory nuclei in the hypothalamus (14–16), and these nuclei were not targeted in our experiments.

AVP has been shown to be proinflammatory and profibrotic via activation of AVP receptors on immune cells and hepatocytes (63, 64). In one study, rats that were subjected to chronic AVP deficiency after portacaval anastomosis (a model of liver damage and fibrosis) showed significantly decreased signs of liver damage and fibrosis (63). These findings could explain why in our studies, BDL rats that received DSAP had significantly lower liver weight-to-body weight ratios as compared with BDL-rats injected with vehicle. This suggests that activation of the A1/A2 cells could support the increased plasma AVP concentration which may contribute to the progression of cirrhosis. In addition, increased AVP secretion increases fluid retention that could worsen the existing hemodynamic changes contributing to further increases in portal hypertension and increased vascular resistance in the cirrhotic liver. Finally, it may be that the attenuation of the increased liver weight-to-body weight ratios was not related to any of the physiological effects of the DSAP lesions of SON but another effect of DSAP injection in the hypothalamus that does not require the SON. Further investigations will be required to help explain this observed effect. Together, these results are consistent with our working hypothesis that the noradrenergic projections to the SON contribute to the fluid retention and dilutional hyponatremia associated with BDL. Although plasma sodium concentration was not directly measured, it could be inferred that injection of DSAP prevented the development of dilutional hyponatremia in male BDL rats.

Dopamine β-hydroxylase (DBH) is an enzyme located in secretory vesicles that converts dopamine to norepinephrine. Cells expressing DBH release norepinephrine as a neurotransmitter. Our results show that the numbers of A1 and A2 cells were reduced by DSAP injections in the SON as compared with vehicle-injected rats. This shows that DSAP was effective at lesioning the noradrenergic innervation of the SON. BDL did not appear to affect the numbers of DBH-positive cells in either A1 or A2.

ΔFos B is a splice variant of FosB, a member of the Fos family of transcription factors. These proteins dimerize with a member of the Jun family of transcription factors to form an Activator Protein-1 (AP-1) complex that binds to AP-1 regulatory domains of various genes to regulate transcription (65). Increased expression of ΔFosB in the rat brain has been observed in response to stress (60), vagal nerve stimulation (59), drug addiction (66), and hyponatremia associated with cirrhosis (57). ΔFosB shows a delayed activation but persists longer than other Fos family proteins like c-Fos, as such it is regarded as a marker for chronic neuronal activation (59, 66). This evidence supports the use of ΔFosB as a marker of chronic neuronal activation in our studies. Our results show that, in vehicle-injected rats, BDL surgery increased the activation of A1 cells as seen in the increase in ΔFosB immunoreactive cells. Injection of anti-DBH saporin significantly decreased the number of ΔFosB immunoreactive A1 cells in BDL rats. However, the number of ΔFosB immunoreactive A2 cells was not significantly affected by BDL but was significantly decreased by DSAP. Finally, anti-DBH saporin injection in BDL rats significantly reduced the number of A1 and A2 cells that was positive for both ΔFosB and DBH. Increased expression of ΔFosB in neurons is associated with increased neuronal activity (49, 51, 57). This suggests that SON projecting noradrenergic A1 and A2 neurons were activated during the hemodynamic changes that occurred in BDL and were lesioned by the anti-DBH saporin injections. The A1 and A2 neurons project to other areas in addition to the SON (67, 68). The failure to observe an effect of BDL on the numbers of ΔFosB immunoreactive cells in NTS could be because it is sensitive to changes in afferent input associated with sham surgery in addition to changes in hemodynamics produced by BDL. Nonetheless, a decrease in the number of ΔFosB-positive A1 and A2 cells was associated with a decrease in plasma copeptin concentration and normalization of plasma osmolality and hematocrit. This suggests that SON projecting A1/A2 neurons contribute to elevated plasma copeptin, hypoosmolality, and fluid expansion associated with BDL. Furthermore, BDL rats showed an increased number of SON neurons positive for both ΔFosB and AVP and this was reduced with DSAP injection. This is consistent with a decrease in activation of the AVP neurons due to decreased stimulation from the A1/A2 neurons and lower plasma copeptin levels. The decrease in SON AVP neuron activation was associated with a decrease in plasma copeptin concentration. These findings further support a role for A1/A2 neurons in AVP secretion in male BDL rats. It is not clear whether the increase in ΔFosB directly translates to increased firing frequency, increased neurotransmitter synthesis, or some other effect that supports the observed changes in plasma copeptin, osmolality, and hematocrit. However, the decrease in A1/A2 ΔFosB and DBH colocalization was associated with a decrease in plasma copeptin and a decrease in SON AVP neurons positive for ΔFosB.

The most potent regulator of AVP release is plasma osmolality (25). Synaptic inputs from osmosensitive circumventricular organs interact with intrinsic mechanisms in SON neurons to increase AVP secretion in response to increased plasma osmolality and to inhibit AVP release when plasma osmolality decreases (25, 69, 70). The secretion of AVP is also regulated by peripheral baroreceptors and cardiopulmonary volume receptors that sense changes in blood pressure and volume such as occurs during liver cirrhosis. When blood volume and pressure is normal, afferent impulses from these receptors tonically inhibit the release of AVP (71, 72). Reduced blood volume and/or blood pressure causes release of AVP partially through a withdrawal of the inhibitory effects of baroreceptors and volume receptors (16, 73, 74). Previous studies showed that BDL causes a decrease in mean arterial pressure and systemic vascular resistance and increases cardiac output, stroke volume, and portal pressure as compared with control (34, 75). These point to significant changes in hemodynamic parameters during liver cirrhosis. There is also a significant increase in both plasma and total blood volumes in BDL rats as compared with controls (31). However, studies have shown that patients with cirrhosis have an altered blood volume distribution with decreased central blood volume and an increased splanchnic blood volume due to increased mesenteric vasodilation (32, 33, 76). Reduced blood volume and pressure are known stimuli for the activation of AVP secretion (25, 26). The reductions in blood volume and blood pressure observed in liver cirrhosis could explain in part the elevated plasma AVP concentrations.

Peripheral pressure and volume receptors send sensory inputs to the brainstem that are then relayed to the SON. Neuroanatomic and electrophysiological studies have shown that there are direct projections of A1 cells from the CVLM to the SON (38, 40–43). Evidence for a direct connection between the A2 cells in the NTS and the SON also exist (35–38) although this may not play a crucial role in normal physiology. Day and Sibbald (42) reported that stimulation of the SON by the NTS may be mediated via the A1 cell group and lesioning of the A1 abolished this effect. Our results indicate that in cirrhosis, activation of both A1 and A2 neurons that project directly to the SON may play a role in the pathogenesis of increased plasma AVP concentration and dilutional hyponatremia although the A1 cells could be of greater significance. BDL increased the number of ΔFosB-positive A1 cells and activation of the A1 and A2 cell group. This was attenuated with the injection of anti-DBH saporin and was associated with a decrease in the plasma copeptin concentration and normalization of plasma osmolality and hematocrit values. These findings suggest that the decrease in central blood volume and blood pressure that occurs in cirrhosis activates the A1 and A2 cells that project to the SON, contributing to increased activation of the AVP neurons in the SON, inappropriate AVP release, and dilutional hyponatremia.

In addition to norepinephrine, A1 and A2 neurons corelease other neurotransmitters that could play an important role in its effects on the magnocellular neurons. Since A1 and A2 neurons are catecholaminergic, lesioning of these neurons with anti-DBH saporin provides useful information on the population of neurons that may be involved in increasing AVP secretion during cirrhosis. However, this does not provide conclusive evidence on the neurotransmitters involved since in addition to norepinephrine these neurons corelease ATP (77–79), substance P (78, 80), neuropeptide Y (81), and other neuropeptides (78), which play an important role in the regulation of AVP release from the SON. Further studies using techniques that differentiate between the various neurotransmitters will be required to determine the contributions of other neurotransmitters to the pathophysiological effects seen in BDL rats.

Perspectives Significance

It is well known that hemodynamic changes, dilutional hyponatremia, and increased plasma AVP are associated with liver cirrhosis. However, the pathogenesis of these events and the neural pathways supporting these changes are not completely understood. This study indicates that the noradrenergic innervation of the SON is implicated in the increased stimulation and inappropriate release of AVP leading to dilutional hyponatremia during cirrhosis. Figure 9 summarizes the main findings of this study and other previous studies. Our results suggest that decreased effective central vascular blood volume leads to activation of the A1 and A2 neurons projecting to the SON. It has been shown that BDL causes an increase in brain-derived neurotrophic factor (BDNF) in the SON and this is associated with an increase in plasma AVP and hypoosmolality (24, 82). Increased BDNF/TrkB signaling could result from A1/A2 activation of the SON leading to a change in the phosphorylation status of potassium chloride cotransporter 2 (KCC2), a transporter that maintains a low intracellular chloride concentration (24). A resulting increase in intracellular chloride concentrations could decrease GABA_A-mediated inhibition of AVP neurons (83). GABA_A is a ligand-gated ion channel that allows the flow of chloride ions based on chloride’s concentration gradient. An increase in intracellular chloride concentration will lead to an efflux of chloride when the channel opens decreasing inhibition or causing cell activation. The possible role of A1/A2 neurons in the increase in BDNF and the decrease in GABA_A-mediated inhibition after BDL surgery will be investigated in future studies. Understanding the mechanisms supporting increased AVP secretion in cirrhosis will help in developing a better understanding of the pathogenesis of hyponatremia associated with liver cirrhosis that could lead to new treatment modalities.

Figure 9.

Summary figure. Increased activation of A1 and A2 neurons supports the chronic release of arginine vasopressin (AVP) in male bile duct ligation (BDL) rats. Previous studies showed an increase in brain-derived neurotrophic factor (BDNF) levels in the supraoptic nucleus (SON), an increase in tyrosine kinase receptor type B (TrkB) phosphorylation and a decrease in phosphorylated potassium chloride cotransporter 2 (KCC2). The effects of BDL on intracellular chloride concentration and GABA-mediated inhibition will be investigated in future studies. [Cl⁻], intracellular chloride concentration; GABA, γ-aminobutyric acid; GABA_A, GABA receptor type A; NE, norepinephrine; NKCC1, sodium potassium chloride cotransporter 1; PNZ, perinuclear zone; pTrkB, phosphorylated tyrosine kinase receptor type B.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grant R01 HL142341 (to J. T. Cunningham).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.O.A. and J.T.C. designed research; A.O.A., J.T.L., and N.R. performed experiments; A.O.A., N.R., and J.T.C. analyzed data; A.O.A., N.R., and J.T.C interpreted results of experiments; A.O.A. prepared figures, A.O.A. and J.T.C. drafted manuscript; A.O.A. and J.T.C. edited and revised manuscript; A.O.A., J.T.L., N.R., and J.T.C. approved final version of manuscript.