The nonapeptide ADH is synthesized by the hypothalamic paraventricular and supraoptic nuclei and is stored and released from the posterior pituitary gland under a serum osmolality rise, a relevant volume depletion (> 7%), and a circadian rhythmicity (10, 11). Once released into the bloodstream, ADH has a very short half-life (around 20 min) and can bind to platelets, making its plasma measurements highly variable (12); it exerts its functions via G-protein–coupled receptors V1 (a and b subtypes) and V2 receptors. V1a receptor induces vasoconstriction in most of the vascular beds; it is expressed in vascular smooth muscle cells, cardiac myocytes, kidney vasa recta, and medullary interstitial cells; and it participates in the regulation of glucocorticoids by inducing the release of adrenocorticotropic hormone from the anterior pituitary gland (10). V2 receptor activation induces water absorption and sodium retention through aquaporin 2 (AQP2) gene expression stimulation and protein translocation to the apical membrane of the principal cells of the kidney collecting duct (11, 13). The process of AQP2-bearing vesicles docking and fusion to the apical plasma membrane as well as the exocytosis/endocytosis balance are tightly regulated by ADH through posttranslational modifications such as phosphorylation, ubiquitylation, and degradation (14); other factors such as prostaglandin E2, estrogen, and interstitial medullary hypertonicity per se can also affect AQP2 expression and trafficking independently from ADH (15–19). In addition to direct transcellular water transport, ADH contributes to interstitial intramedullary hyperosmolality and allows gradient-driven water absorption by increasing sodium absorption in several other tubular segments [thick-ascending limb of Henle-luminal Na-K-2Cl cotransporter (NKCC2) and basolateral alpha-1-Na-K ATPase, distal convoluted tubulesNa-Cl co-transporter (NCC), and collecting ducts-epithelial sodium channel (ENaC)] and by favoring urea cycling and interstitial urea medullary absorption (increased expression of urea medullary transporters A1 and A3) (20, 21). Sodium and urea are the major interstitial medullary osmolytes, whereas organic osmolytes in renal intracellular fluid to balance the hypertonic interstitium include polyols (sorbitol), neutral free amino acids, and the combination of urea and methylamines (22). One of the principal organic osmolytes is represented by betaine, a neutral free amino acid, and a methylamine. Betaine is synthesized by proximal tubule cells and transported by a specific transporter, the betaine/GABA transporter 1 (BTG-1) to the medulla. Interstitial hypertonicity causes an increase in basolateral membrane BGT-1, which couples the transport of betaine to that of chloride and sodium, allowing betaine to enter the cell from the extracellular fluid. Under normotonic conditions, BGT-1, which is mainly located in the cytoplasm, translocates to basolateral plasmalemmas when interstitial tonicity increases. Intracellular accumulation of betaine and the higher expression of BGT-1 in the tubule reflect the higher interstitial tonicity that increases urine osmolarity in the presence of ADH.

Contrasting results on ADH plasma and urinary levels have been reported in experimental models of hypertension as well as in human essential hypertension (10, 23–25). The slight increase of urinary ADH observed both in benign and malignant essential hypertensive adults and in adolescent patients does not unambiguously and explicitly indicate a relevant contribution of ADH to the development or maintenance of hypertension (26–29). Increased circulating ADH levels have also been reported in prehypertensive young and hypertensive middle-aged spontaneously hypertensive rats (SHR), the most widely used animal model of human essential hypertension (30), as compared to their normotensive controls Wistar Kyoto (WKY), but not in the SHR-stroke-prone strain, where ADH levels were markedly reduced (31, 32). Discordant results on neuron ADH content have been reported as well in this strain: cultured brain neurons from newborn SHR (1-day-old) contained fewer ADH levels than age-matched WKY, as in paraventricular hypothalamic neurons at 6 and 12 weeks of age and in the neurons from the brain stem of adult SHR. When compared with WKY, ADH content was similar in prehypertensive 3-week-old SHR neurons from the hypothalamic paraventricular nucleus but increased in the neurohypophyseal neurons (33–36). In another model of experimental hypertension related to blood volume expansion (deoxycorticosterone-salt hypertensive rat), ADH levels were elevated not only in plasma but also in 24-h urinary excretion, whose increase paralleled the rise in BP levels (26, 37). The importance of ADH in this model was further demonstrated by the fact that hypertension development is accelerated by treating the rats with small doses of argipressin (Pitressin) and that surgically induced diabetes insipidus prevents the development of hypertension (38). Another observation supporting the role of ADH in the development of hypertension in this model relates the fact that Brattleboro rats with hereditary hypothalamic diabetes insipidus did not develop hypertension when treated with deoxycorticosterone-salt (39). In deoxycorticosterone-salt hypertensive rat, it seems that ADH differently contributes to hypertension with a prevalent antidiuretic effect observed during the early phases of deoxycorticosterone-salt hypertension (40) and a direct vasopressor effect in the established phase of hypertension (41). Again, non-univocal results on ADH’s role in hypertension development were reported in salt-sensitive and salt resistant Dahl rats where even if the former showed increases in ADH levels, ADH receptor antagonism failed to reduce BP levels (26). Similar results were observed in the experimental model of partial nephrectomy-salt hypertension after a 70% reduction in renal mass followed by drinking a 1% NaCl solution. Treatment with a vasopressin pressor antagonist in this model had only a small effect on arterial pressure (42). On the contrary, a sustained hypertension could not be produced with the partial nephrectomy-salt protocol in rats with hereditary hypothalamic diabetes insipidus (26, 42).

Although an increased vascular sensitivity to ADH has been reported in several experimental models of hypertension, such as the spontaneously hypertensive (SHR) (43), the deoxycorticosterone (DOC)-salt hypertensive (44), the Dahl salt-sensitive hypertensive (45), the New Zealand genetically hypertensive rats (46), ADH V1a receptor-selective antagonism lowers BP in mineralocorticoid-related hypertension (47) but not in renovascular hypertension (48). Furthermore, it has been reported that chronic infusion of intravenous and renal intramedullary of a selective vasopressin V1a receptor agonist resulted in sustained hypertension, implying that the V1a receptor in the renal medulla is a mediator of the hypertensive effect (49–52); these results appear to be supported by the fact that V1a receptor knockout mice are resistance to salt-induced hypertension (53). Furthermore, while the role of ADH in the development of hypertension via V1a receptor stimulation was shown by the hypotensive effect of OPC-21268, a non-peptide ADH V1a antagonist, in young SHR, with a more pronounced effect in male than in female rats and the interesting observation of a persistence of low BP levels after the drug withdrawal (54), other reports were unable to demonstrate such a hypotensive effect by the chronic treatment with specific V1 receptor antagonism (55). If the evidence could indicate a certain role in the development of hypertension, this seems not to be the case in the maintenance of arterial hypertension because no effect of OPC-21268 was observed in the established hypertensive phase of adult SHR (54, 56). The involvement of V2 receptors in ADH’s potential role in the development of hypertension was also conflicting and inconclusive, observing in fact both a reduction in BP after the administration of V2 receptor antagonist administration OPC 31260 to young male SHR in the prehypertensive phase (57), but also the lack of a hypotensive effect in young SHR but a significant hypertensive effect in the adult SHR (58).

With these promises, this study investigated the abnormalities in water and sodium balance and osmoregulation in female SHRs from their prehypertensive phase until the adult age when arterial hypertension is established and organ damage has already developed. We also evaluated the effects of a short period of oral treatment with selective V1a and V2 receptor antagonists administered very early in the prehypertensive phase to measure the contribution of ADH with its pressor and aquaretic effects on the future development of arterial hypertension in SHR, the most commonly used experimental model of human essential polygenic hypertension.

Rats included in this study were treated in accordance with the European Council on Animal Care guidelines and the protocols were approved by the local institutional animal ethics committee. Two experimental studies with a large number of 3-week-old female SHR (n = 86) and Wistar-Kyoto (WKY) rats (n = 58) were set up to evaluate: (i) (first protocol) the time course of water imbalances measured as ADH renal sensitivity and AQP2 expression and excretion from the prehypertensive phase to the established hypertension and organ damage development in SHR compared to age-matched WKY and (ii) (second protocol) the effect of a very early treatment with ADH V1 and V2 receptor antagonists on future development of arterial hypertension in SHR strain.

Four batches for a total of 104 3-week-old female SHR (n = 58) and Wistar Kyoto (WKY, n = 46) rats were purchased over a period of 5 years from three different European colonies to avoid potential phenotype differences associated with breeding or alimentary factors (Charles River Laboratories, Calco, Italy; Charles River France L’Arbresle, France; Charles River Nederland, Maastricht, Netherlands). Rats were housed, fed a standard laboratory rat diet (Basal Purified Diet 5755C, 0.54% sodium chloride, Purina Mills Inc., St. Louis, MO, United States), had free access to drinking water, and were maintained on a 12-h light/dark cycle, at constant humidity (40–46%) and temperature (22–24°C) (59). After a few days of acclimatization, conscious rats from both groups underwent indirect tail-cuff measurements of systolic BP by plethysmography (Harvard Apparatus Ltd., South Natick, Massachusetts, United States) and recorded on a Maclab/8 system (AD Instruments Ltd., Castle Hill, New South Wales, Australia) as we previously described (60): four to five measurements were taken in each rat and then averaged. Young rats were then individually kept for 3 days in metabolic cages to allow daily 24-h urine collection and food and water intake measurements. The measurement of these parameters refers to the second and third days of metabolic cage stay. Urine albumin, electrolytes, creatinine, osmolality, and AQP2 were evaluated and averaged; both the urine albumin to creatinine ratio as well as the AQP2 to creatinine ratio were calculated. A quantitative sandwich enzyme immunoassay was used to evaluate urine AQP2 and was performed according to manufacturer instructions and done in triplicate (Rat Aquaporin 2 ELISA Kit, MBS703718, MBS San Diego, California, United States). The intra- and interassay coefficients of variation were 7 and 8%, respectively. At the end of the 3-day time in the metabolic cage, 23 SHR and 21 WKY rats underwent blood sampling from the inferior vena cava and basal laboratory tests, including plasma electrolytes, creatinine, and osmolality were performed. Venous blood samples for the determination of plasma ADH were collected using a prechilled heparin-coated tube and immediately centrifuged for 30 min at 3,000 g at 4°C; plasma samples were then stored at –80°C until assays were performed without any freeze-thaw cycles. The enzyme-linked immunosorbent assay was performed according to manufacturer instructions and was done in triplicate (Rat Vasopressin ELISA Kit, MBS3808966, MBS San Diego, California, United States). The intra- and interassay coefficients of variation were 9 and 12%, respectively, and the analytical sensitivity was 0.95 pg/ml.

Following the experiment period in the metabolic cage, a group of young female SHR (n = 28) and age-matched WKY (n = 23) were euthanized with diethyl ether. As we previously described (61), the rat abdomen was opened and the vena cava cannulated with polyethylene tubing (PE-50) and perfused by a roller pump for 30 s with phosphate buffer saline, the right renal artery clamped, and the kidney was rapidly removed from its attachments (renal artery and vein, and fat and adrenal glands). After excision, the right kidney was repeatedly rinsed with 15% and 30% sucrose solutions at 4°C and then frozen embedded in OCT compound for immunofluorescence labeling of cryosections as previously described (62). The left kidney, still attached to vessels, was initially perfused by a roller pump for 5 min with a solution containing 4% of paraformaldehyde to fix the tissues and then excised and embedded in paraffin for subsequent immunohistochemical investigations (immunoperoxidase reaction) (61). After being dewaxed and rehydrated, sections (5–7 μm) underwent microwave antigen retrieval in citrate buffer (citric acid, 10 mM, pH = 6) followed by the blockade of endogenous peroxidase with hydrogen peroxide solution (3%). The sections were then incubated with primary rabbit polyclonal anti-betaine γ-amino-n-butyric acid transporter-1 (BGT-1) (Anti-BGT-1, 1:250 dilution, HPA034973, Sigma-Aldrich antibodies), anti-Aquaporin2-phospho S256 (Anti-AQP2-phospho S256, 1:200 dilution, ab111346), mouse monoclonal anti-alpha 1 Na-K ATPase (1:400 dilution, ab211130), and rabbit monoclonal anti-Uromodulin-Tamm-Horsfall (Anti-UMOD, 1:250 dilution, ab256473) (Abcam, Cambridge Biomedical Campus, Cambridge, United Kingdom; Prodotti Gianni S.p.A., Milano, Italy). Biotin-conjugated anti-rabbit and anti-mouse antibodies (1:30 dilution, Dako Corporation, Carpinteria, California, United States) were used as secondary antibodies for avidin-biotin amplified 3,3′-diaminobenzidine staining as well as fluorescein isothiocyanate (FITC) and Texas Red conjugated secondary antibodies for cryosection immunostaining. To specifically localize BGT-1, AQP2, and alpha-1-Na-K ATPase, cryosections were tested with anti-Uromodulin-Tamm-Horsfall that labels the diluting segment (medullary thick ascending limbs of Henle’s loop, m-TAL) and not the collecting duct.

A separate pilot experiment was performed to evaluate the relationship between AQP2 renal medullary expression and urine AQP2/creatinine ratio during low and high water intake. A small group of young female WKY (n = 8) had a prolongation of metabolic cage stay for 48 h when they received a low water intake (water to drink limited to 20 or 30 ml/kg/day, n = 4) after the 3 days of urine collection by protocol, or a high water intake (water to drink limited to 70 or 90 ml/kg/day, rat free to drink and the difference to reach the established amount by gavage, n = 2) compared to normal water intake (50 ml/kg/day, n = 2). Rats had measurements of urine output at 24 and 48 h and were rapidly euthanized with diethyl ether. Both kidneys were rapidly excised, washed with cold phosphate buffer saline, and the cortex and medulla areas were separated and immediately frozen in liquid nitrogen for immunoblotting analysis. Frozen samples were homogenized (Ultraturrax T8, Janke and Kunke, Staufen, Germany) and further sonicated (four times at 4°C for 30 s) in the extraction buffer (pH 7.4) containing 138 mM NaCl, 1.5 mM KH₂PO₄, 2.7 mM KCl, 0.05% Triton X-100, 2 mmol/L protease inhibitor solution [4-(2-aminoethyl) benzenesulphonyl fluoride (AEBSF), 4 μmol/L aprotinin, 0.1 mmol/L leupeptin, 200 μmol/L bestatin, 50 μmol/L pepstatin A, 15 μmol/L E-64], 0.44 mmol/L O-phenanthroline, 1 mmol/L EGTA, and 0.1% sodium dodecyl sulfate. Homogenates were centrifuged at 12,000 rpm for 10 min at 4°C. The protein concentration of the supernatant was determined using the Bio-Rad protein assay (Bio-Rad, Hercules, California, United States). Renal medulla lysates were subjected to standard SDS-PAGE and Western blotting techniques using 5% for stacking and 8–12% polyacrylamide separating gel. Equal amounts of proteins (75 μg/well) were loaded. Separated proteins were transferred to polyvinylidene difluoride (PVDF) membranes (Immun-Blot PVDF Membrane, Bio-Rad Laboratories, Hercules, California, United States) at 4 mA/cm² for 45 min. The membranes were then incubated with primary antibodies anti-AQP2 (Anti-Aquaporin 2-phospho S264, 1:1,500 dilution, ab254071) and anti β-actin (Anti-β Actin, 1:4,000 dilution, ab115777, Abcam, Cambridge Biomedical Campus, Cambridge, United Kingdom), and the signal was amplified using an Opti-4CN Detection Kit (goat anti-rabbit and anti-mouse HRP-conjugated secondary antibodies, Bio-Rad Laboratories, Hercules, California, United States). Medullary expression of AQP2 (Figures 1A1, A2), urine AQP2/creatinine ratio (Figure 1B), and their relationship at different amounts of water intake in young female WKY (Figure 1C) indicate a good level of correlation (determination coefficient r² = 0.78, p = 0.004, n = 8) suggesting that measurement of urine AQP2/creatinine ratio could be a valuable marker to evaluate water balance during the lifetime course in these rats.

At the end of the prehypertensive phase experiments, as indicated above, the remaining rats (30 SHR and 15 WKY) were periodically evaluated for body weight and hemodynamic parameters (BP and heart rate) at 8, 11, 15, 18, 21 24, and 27 weeks of age. At the age of 14–15 weeks, rats from both groups underwent a 3-day metabolic stay to measure urine albumin, electrolytes, creatinine, osmolality, and AQP2. Finally, at the age of 28–30 weeks, SHR and WKY repeated the same procedures performed in the prehypertensive phase with the euthanasia of all the rats following the metabolic cage stay period.

In the second set of experiments, after a few days of acclimatization, a group of 28 female 25-day-old (3.5 weeks of age) SHR (Charles River France L’Arbresle, France) were divided into three groups. (i) Untreated SHR (vehicle by gavage) (n = 8), (ii) treated from the 25th to 49th day of age with the selective ADH V1 receptor antagonist (OPC 21268-treated SHR, n = 10, N-[3-[4-[4-(2-oxo-3,4-dihydroquinolin-1-yl)piperidine-1-carbonyl]phenoxy]propyl]acetamide, 5 mg/kg/day by gavage, 3924/10, Tocris Bioscience, Bristol, United Kingdom), and (iii) treated from the 25th to 49th day of age with selective ADH V2 receptor antagonist (OPC 41061-treated SHR, n = 10, N-[4-[(7-Chloro-2,3,4,5-tetrahydro-5-hydroxy-1H-1-benzazepin-1-yl)carbonyl]-3-methylphenyl]-2-methylbenzamide, 0.5 mg/kg/day by gavage, Otsuka Pharmaceutical, Osaka, Japan). A group of 12 female untreated WKY rats were also used as controls. All these rats had their tail-cuff systolic BP monitored non-invasively every week starting from the 3rd week of age before the initiation of the active drug treatment period until the 8 weeks of age and then with an interval of 10 days from the 8th to 27th week of age. All of these rats were subjected four times to a 3-day metabolic cage stay as described above (period 1: 23–25 days of age before the initiation of active drug treatment; period 2: 45–49 days of age just before the end of active drug treatment, period 3: 96–106 days of age stable hypertensive phase in untreated SHR; and period 4: 168–186 days of age stable hypertensive phase and organ damage in untreated SHR).

All drugs, except those specifically indicated in the text, and chemical components of solutions were of analytical grade and were purchased from Sigma-Aldrich Chemical (St. Louis, MO, United States).

All values are expressed as means ± SD. The statistical comparison of results was done by one-factor repeated measures analysis of variance or by Student’s paired or unpaired t-test when appropriate. A Bonferroni post hoc analysis was used when appropriate. Relationships between urine AQP2/creatinine ratio and renal medulla AQP2 expression, and urine osmolality and m-TAL BGT-1 were analyzed by linear regression analysis using Pearson or Spearman correlation coefficients and reported as determination coefficient. P < 0.05 was considered to be statistically significant.

Table 1 indicates that young 3–5-week-old prehypertensive female SHRs and age-matched WKYs have similar body weight and systolic BP. No differences were measured in food and water intake at this age (Table 1). Plasma creatinine level and creatinine clearance were comparable between the two strains, as were plasma levels of sodium, potassium, and osmolality. Daily urine volume was significantly reduced in young SHR as compared with WKY (–45%, P < 0.05, Table 1), associated with a marked increase in urinary osmolality (P < 0.01, Table 1) indicating an increased urinary concentration. There was no significant difference in 24-h urinary sodium excretion, even if a mild trend toward a reduced sodium excretion was observed in young SHR; fractional excretion of sodium (FENa) was not changed. All these observations were consistent in the various batches of young SHR and WKY obtained from different colonies except for the batch 3, where a significant daily urinary sodium excretion was significantly reduced in SHR when compared to WKY (Figure 2). Daily excretion of albumin was similar between the two strains at this young age.

Plasma ADH was not different between young SHR and WKY, but a marked rise was observed in urine AQP2/creatinine ratio in young SHR (+ 163% vs. young WKY, P < 0.01, Table 1). Young SHR also showed a parallel increase in immunostaining for AQP2-phospho-serine-256 in the collecting duct (identified by the absent colocalization with Tamm-Horsfall) (Figure 3C) vs. WKY (Figure 3D). Another interesting observation indicating an increased tonicity of the medullary interstitium in young SHR (Figures 3A, 4A,C) as compared to WKY (Figures 3B, 4B,D) is represented by a renal medulla increase in immunostaining (by immunoperoxidase and immunofluorescence) for m-TAL BGT-1 (+ 121% vs. WKY, Table 1) that colocalizes with Tamm-Horsfall protein. In young SHR, BGT-1 expression in m-TAL was directly related to urinary osmolality (r² = 0.28, P = 0.009, n = 23, Figure 5A) and to AQP2/creatinine ratio (r² = 0.19, P = 0.038, n = 23). No significant relationship was observed in young WKY between tissue m-TAL BGT-1 and urinary osmolality (r² = 0.07, P = 0.240, n = 21, Figure 5A) or AQP2/creatinine ratio (r² = 0.02, P = 0.494, n = 21). A mild increase of alpha-1-Na-K ATPase immunostaining in m-TAL (by immunofluorescence) that colocalizes with Tamm-Horsfall protein was observed in young SHR (+ 48%) vs. age-matched WKY (Figures 3E,F). All these results can support an increased sensitivity to ADH in this prehypertensive phase in female SHR.

Table 2 indicates that 28–30-week-old female SHRs showed, as expected, higher systolic BP levels but lower body weight vs. age-matched WKYs. No differences were measured in food and water intake at this age (Table 2). When compared to age-matched WKY, SHR had higher plasma creatinine levels and lower creatinine clearance, indicating the development of mild reduction in renal function; sodium, potassium and plasma osmolality were similar (Table 2). Daily urine volume was mildly but significantly reduced in old SHR as compared with WKY (–23%, P < 0.05, Table 2), without differences in urinary osmolality (Table 1). There was no significant difference in 24-h urinary sodium excretion between the two strains at this age. Daily excretion of albumin was increased in old SHR (+ 70%, P < 0.01) vs. age-matched WKY. Plasma ADH was higher in old vs. young SHR but also increased as compared with old WKY. There were no differences in urine AQP2/creatinine ratio between old SHR and WKY; urine AQP2/creatinine ratio was significantly lower in old SHR vs. young SHR. No difference between old SHR and WKY in tubule immunostaining (by immunoperoxidase and immunofluorescence) for m-TAL BGT-1 (Table 2 and Figures 4E–H). In old SHR, BGT-1 expression in m-TAL was not correlated with urinary osmolality (r² = 0.01, P = 0.795, n = 13, Figure 5B) or AQP2/creatinine ratio (r² = 0.03, P = 0.553, n = 13). No significant relationship was observed in the old WKY between tissue m-TAL BGT-1 and urinary osmolality or AQP2/creatinine ratio. No difference was observed in alpha-1-Na-K ATPase immunostaining in m-TAL or in collecting duct AQP2-phospho-serine-256 in old SHR vs. age-matched WKY. All these results indicate a reduced renal tubule sensitivity to ADH in the hypertensive phase of female SHR with organ damage.

Figure 6 depicts the systolic BP profile from the prehypertensive phase up to the 27th week of age of the four studied groups: untreated SHR, OPC 21268-treated SHR, OPC 41061-treated SHR, and untreated WKY. ANOVA analysis of the systolic BP time course revealed an increased BP over time (P < 0.001), with a significant effect of the treatment (P < 0.01) and also a significant interaction between the time and the treatment (P < 0.001) (Figure 6). In untreated female SHRs, systolic BP levels rose around the 7th to 8th week and reached a plateau around the 13th to 14th week of age. In SHRs, the early active treatment with OPC 21268 slightly (but not significantly) decreased systolic arterial pressure, but it did not interfere with the future rise of BP after its withdrawal. On the contrary, OPC 41061 instead showed an attenuation of the rise in systolic BP, delaying the development of hypertension by 5.5 weeks (Figure 6). As expected, an increased volume of hypotonic urine observed during the active treatment in OPC 41061-treated SHR compared with untreated SHR and OPC 21268-treated SHR (P < 0.01, Figures 7A,B). No significant differences were observed in 24-h urinary sodium excretion rate among OPC 21268-treated or OPC 41061-treated SHR and untreated SHR (Figure 7D). Urine AQP2/creatinine ratio was markedly reduced by active treatment with OPC 41061 indicating a profound decrease in AQP2 shedding from apical tubule membrane into the urine (Figure 7C).