Suppression of TRPV4 channels ameliorates anti-dipsogenic effects under hypoxia in the subfornical organ of rats

The phenomenon of water intake reduction during the 1st day of hypobaric hypoxia has been known for a long time. However, the reason for the same is yet unknown. The transient receptor potential vanilloid (TRPV) channels, including TRPV1 and TRPV4, are located in the subfornical organ (SFO). These are calcium permeable cationic channels gated by various stimuli such as cell swelling, low pH, and high temperature, and participate in anti-dipsogenic effects when activated. We aimed to explore the drinking behavior of rats and the mechanism of TRPVs under hypoxia. Chemical TRPV4 inhibitors (HC-067047 and Gadolinium) or TRPV4 knockout, but not TRPV1 inhibitor SB-705498, could restore the water intake under hypoxia. Hypoxia-mediated direct activation of TRPV4 may be the reason of anti-dipsogenic effects because the serum sodium, pH, and intracranial temperature are unaltered. Interestingly, we found that hypoxia immediately increased the intracellular Ca2+ concentration ([Ca2+]i) in HEK293-TRPV4 cells and primary neurons from SFO region, but not in the HEK293-TRPV1 cells. Moreover, hypoxia-induced [Ca2+]i increase depended on the indispensable hemeoxygenase-2 (HO-2) and TRPV4. HO-2 and TRPV4 were also confirmed to form a complex in SFO neurons. These results demonstrated that SFO cells sense hypoxia and activate via the HO-2/TRPV4 multiple channels, which are associated with anti-dipsogenic effects.

TRPV1 and TRPV4 are considered as osmoreceptors, including central 11 and peripheral 19 , that exist in the brain and liver, respectively. Although TRPV4 was revealed to be activated by hypertonic stimuli in mammalian cells 20 , there are two contradictory reports. One of the studies demonstrated that hypertonicity sensing is a mechanical process requiring TRPV1, but not TRPV4 13 . The other study indicated that TRPV1 and TRPV4 channels are not the primary mechanisms by which the central nervous system responds to hypertonic stimuli and increasing thirst 21 . Although the reasons for this discrepancy are not clear, these studies indicate the diverse roles of TRPV1 and TRPV4 under different stimuli (hypoxia and hypertonicity) in the body fluid homeostasis. Clarification of the molecular mechanism responsible for hypoxia sensing in CVOs neurons is a prerequisite for our understanding of the anti-dipsogenic effects. In the present study, we explored the underlying reason of anti-dipsogenic effects under hypoxia and examined the molecular basis for HEK293-TRPV4 cells and hypoxic sensing primary neurons of SFO, which may participate in the regulation of thirst sensation.

Results
TRPV4 suppressants, but not TRPV1, restore water intake under hypoxia. We first validated the anti-dipsogenic effects of hypoxia at simulated 6000 m altitude. The rats did not exhibit any abnormal behavior including water drinking 1 week after the surgery. Consistent with the previous findings 2, 3 , our data showed that water intake was decreased more than 78% at 6 h as compared to the normoxia + saline group (n = 12, P < 0.01, Fig. 1C). Interestingly, we found that TRPV4 suppressants, such as gadolinium, increased the water intake by 2.7 times at 6 h as compared to the saline control group under hypoxia (n = 12, P < 0.05, Fig. 1C). Another TRPV4 inhibitor, HC-067047, was used because of low-specificity of gadolinium, and similar results were obtained. However, TRPV1 inhibitor, SB-705498, did not affect the accumulation of water intake (n = 12, Fig. 1C). To further affirm the TRPV4 channel on anti-dipsogenic effects of hypoxic, TRPV4 knockout mice was used. Although the knockout did not affect the water drinking under normal oxygen, the intake was increased 3 times at 6 h as compared to the WT under hypoxia (n = 6, P < 0.05, Fig. 2). These results indicated that TRPV4 channel was responsible for the anti-dipsogenic effect of hypoxia. In addition, immunofluorescence data showed that TRPV4 was identified by co-staining with TRPV1 ( Fig. 1D) and MAP-2 (neuronal maker of a neuron) (Fig. 1E) in SFO area that is a characteristic protuberance nearby the third ventricle of the cerebrum (Fig. 1A).

Serum sodium, pH value, and intracranial temperature were unaltered under hypoxia in rats.
We found that hematocrit increased after 6 and 24 h exposure to hypoxia as compared to normoxia (sea level) (n = 12, Fig. 3A). This suggests that the blood is concentrated in the early hypoxia along with less water intake. It is well-known that TRPV4 can be regulated within the microenvironment by different physical and chemical factors, such as temperature, pH, and hypertonic stimuli 9,22,23 . To explore the potential mechanism of TRPV4 in anti-dipsogenic effects under hypoxia, we detected the plasma [Na + ], pH value, and intracranial temperature. Although oxygen saturation declined by 20% at 24 h under hypoxia (n = 12, P < 0.05, Fig. 3A), plasma [Na + ] and pH value did not vary between normoxia and hypoxia groups (n = 12, Fig. 3A). Also, the intracranial temperature was unaltered in rats (n = 12, Fig. 3B,C). These results strongly implied that hypoxia may be involved in TRPV4 activation and reduction of water intake.

Hypoxia-induced [Ca 2+ ] i increase and antagonist of TRPV4 channel inhibited this effect.
To determine whether hypoxia activates the TRPV4 protein, TRPV4 overexpressed HEK293 cells were established. The protein expression was detected with the anti-V5-FITC antibody in the transfected cells, but not in the control (untransfected cells) (Fig. 4A,B). TRPV4 is a calcium channel that has been shown to be highly expressed in the cytomembrane and the perinuclear region of the transfected cells, reflecting expression in the endoplasmic reticulum,Golgi apparatus, and the membrane 22 . Moreover, extracellular Ca 2+ rush intracellularly when activated. In the medium of constant [Ca 2+ ] o (1.5 mM) and temperature maintained at 37 °C by a sealed heating apparatus, hypoxia increased [Ca 2+ ] i in the overexpression cells (n = 60, P < 0.05, Fig. 3C), which was completely abolished or significantly inhibited by Gad (n = 60, P < 0.05, Fig. 4D). However, hypoxia could not increase [Ca 2+ ] i in control or TRPV1 overexpressed cells (n = 60, P < 0.05, Fig. 4E,G). 4α -PDD, an agonist of TRPV4, increased [Ca 2+ ] i in TRPV4 overexpressed cells (n = 60, P < 0.05, Fig. 4H), which did not occur in the control cells (n = 60, P < 0.05, Fig. 4I). We also observed the same phenomenon in primary neurons from SFO region. TRPV4 knockdown predominantly suppressed the intracellular pulse (n = 60, P < 0.05, Fig. 5A). These evidence suggest that TRPV4 calcium ion channel is immediately opened under hypoxia, and this effect within SFO is potentially associated with anti-dipsogenic effects. Moreover, TRPV1 inhibitor SB-705498 could reduce [Ca 2+ ] i under normoxia conditions (n = 60, P < 0.05, Fig. 5C), which implies that only a portion of TRPV1 is opened in normoxia.

Hypoxia could not induce [Ca 2+ ] i increase after the HO-2 knockdown in HEK239-TRPV4 cells.
To explore whether TRPV4 is an oxygen sensitive calcium ion channel, we established a HO-2 knockdown model in HEK293 cells. HO-2 is highly and constitutively expressed in neuronal and chemosensing tissues 24,25 , and co-localizes with the BK channel to sense oxygen concentration 26 . HO-2 is ectopically expressed in the membrane and cytoplasm of the cells and co-localizes with TRPV4 (Fig. 6A). We knocked down HO-2 by specific siRNA in HEK239-TRPV4 cells (TRPV4 overexpression) (Fig. 6B) and then performed live cell calcium imaging. Interestingly, the calcium ion pulse induced by hypoxia was barely observed during HO-2 deficiency when compared with the scrambled siRNA (n = 60, P < 0.05, Fig. 6C). These results imply that HO-2 is a precursor for hypoxia-induced TRPV4 activation.

Hypoxia-induced [Ca 2+
] i increase in an HO-2 dependent manner in primary neurons of SFO region.
In an attempt to verify the synergistic effect of HO-2 and TRPV4 in the SFO region under hypoxia, primary (B) Schematic showing that a guide cannula was implanted into the third ventricle and rats were recovered at 1 week after the surgery. Drugs were microinjected into the ventriculus Tertius of rats at 10 min before hypoxia. Normoxia + saline indicated the sham group. (C) The anti-dipsogenic effects of hypoxia (6000 m) were segmentally reversed by 1 μ g TRPV4 inhibitors, gadolinium (Gad) or HC-067047, but not TRPV1 inhibitor (10 mg SB-705498) (n = 10 for each group; * P < 0.05, * * P < 0.01). Immunocytochemistry examined the colocalization of TRPV4 (green) with (D) TRPV1 (red) and (E) MAP2 (red) in the SFO at 24 h after hypoxia (6000 m). Nuclei were counterstained with DAPI (blue). Scale bar, 30 μ m. The rats in (A), (D), and (E) lack the 3 rd ventricular catheter.

Discussion
The phenomenon that cumulative water intake of the animal significantly reduced in the early stage of hypoxia, anti-dipsogenic effect, has been discovered for a long time 2,3 . However, the reason is yet ill-understood. In the present study, we attempted to explore the underlying mechanism and deduced three features; First, hypoxia could reduce water intake, which was irrelevant to osmotic pressure, pH, and intracranial temperature. Second, these effects are related to TRPV4 in an HO-2 dependent manner on the central nervous system. Third, the neurons in the SFO region may be responsible for these effects.
The transient receptor potential (TRP) channels, including TRPV1 and TRPV4, expressed in the neurons of the SFO area lacking blood-brain barriers, which are regarded as osmosensors in the brain 12,15,24,27,28 . To investigate the molecular mechanism of the anti-dipsogenic effects, a rat model for the third ventricular catheter was established. Interestingly, our data indicated that TRPV4 inhibitor, but not a TRPV1 inhibitor, can regain the water intake under hypoxia. Concurrently, the anti-dipsogenic effects were suppressed in TRPV4 knockout mice. This evidence strongly suggests that TRPV4 was associated with hypoxia-induced anti-dipsogenic implications in the brain. However, two complications required resolving. First, irrespective of TRPV4 inhibitor or TRPV4 knockout, anti-dipsogenic effects could not be eliminated, which indicates that other mechanisms are potentially involved. Second, both TRPV4 inhibitor and TRPV4 knockout cannot increase the water intake under normoxia, which illustrates that TRPV4 is not the key factor generating thirst perception in normoxia.
TRPVs have been reported to be activated by hypotonic stimuli, putatively by the increase of intracellular calcium ion ([Ca 2+ ] i ). However, other factors can lead to TRPVs' activation, such as temperature and acid-base balance. TRPV4 responds to hypertonic stimuli in a temperature-dependent manner and are activated as with escalating temperatures 9,19 . Although the cerebrospinal fluid could not be harvested, we found that serum sodium concentrations, pH value, and intracranial temperature were not altered during hypoxia. This implied that the hypoxia-induced anti-dipsogenic effects were not secondary and that TRPV4 may be activated directly by hypoxia. In order to substantiate it, TRPV4-transfected HEK293 cells were used, which showed that [Ca 2+ ] i was increased by hypoxia or by a TRPV4 agonist while TRPV4 inhibitor reduced [Ca 2+ ] i . Previous studies have revealed that TRPV4 is activated by hypoxia in a pulmonary vascular endothelial cell that is involved in hypoxic pulmonary hypertension 21,29 . This evidence suggested that TRPV4 is an oxygen sensitive calcium ion channel.
Therefore, it was imperative to understand how the SFO area cells containing TRPV4 change during hypoxia and achieve oxygen signal perception. Consistent with the TRPV4-HEK293 cells, we found that the SFO area primary neuron could be excited in hypoxia and inhibited by the TRPV4 antagonist. In fact, not only the partial pressure of blood oxygen (bPaO 2 ) but also that of the cerebrospinal fluid (cPaO 2 ) decreased when the animal remained in a hypoxic environment. The SFO area cells lacking the blood brain barrier and proximal to the third ventricle may respond to cPaO 2 changes easily. As a result, the TRPV4 channel is opened, neurons excited and projected to the senior central cortex. However, TRPV4 channel could not sense PaO 2 directly; we found that the activation of TRPV4 channels by hypoxia was dependent on HO-2 expression while the manipulation of TRPV4 by Gad or 4α -PDD was independent of HO-2 expression. Thus, it can be inferred that HO-2 and TRPV4 form a, ion channel complex in cytomembrane and organelles, wherein HO-2 accepts TRPV4 and react to the oxygen signal. HO-2 is an enzyme that degrades heme into biliverdin and carbon monoxide 30,31 . A previous study found that the knockdown of HO-2 expression reduced the calcium-sensitive potassium channel activity, and carbon monoxide, which is a product of HO-2 activity 26 . Moreover, it can be presumed that HO-2/TRPV4 complex is a key to oxygen sensing, water, and sodium balance adjustment in SFO region at high altitude.
Thirst and fluid intakes are a response to osmotic pressure hoist and body fluid deficits 11,32 . Though disputed, some of the brain regions and neural circuits, such as CVOs, which contains two circumventricular organs, SFO, and OVLT, participate in the physiological regulation of the fluid intake when animals exhibit body fluid deficiencies 18,19 . Such homeostatic regulation of fluid intake is controlled by the thirst drive that can arise when the body is dehydrated, or when the CVOs' neurons are activated 32,33 . Conversely, after adequate or excess ingestion of water, inhibitory influences occur on the thirst, and fluid intake is interrupted in a regulatory manner 1,2 . Water requirement at high altitude theoretically increases due to increased water loss at low ambient water vapor pressure 1 . A tendency of decreased intake during ascent to higher altitudes is observed. Since the anti-dipsogenic effect is significant, it is highly possible that the reduction of water intake would help human and animal to acclimatize with the ambient environment under hypoxia. Additionally, the water intake is significantly changed in a self-regulatory manner, which leads to a decrease in plasma volumes in subjects at high altitudes, which is consistent with that reported by Surks et al. 6,10 . This improves the density of red blood cells and hypoxia tolerance but is not associated with the osmotic pressure, which is essential for maintaining the homeostatic balance 2,6,10,32 . Thus, our data are consistent with the hypothesis that the reduction of water intake may decrease the plasma volumes and elevate the red cell volume relatively.
On the other hand, non-brain factors may exist. We believed that the brain was a key element in the anti-dipsogenic effects. The central nervous system is sensitive to hypoxia. Hence, it is not surprising that the impairment of the neuropsychological functional occurs at high altitudes. However, a clear understanding of the effect of hypoxia on the brain remains elusive 34,35 . Our results only appear to scrape the tip of the iceberg while addressing the hypoxic influence on the brain function and behavior, including drinking, eating, and sleeping. The current study about water intake may provide some insights into these issues.
In summary, we show that HO-2/TRPV4 channels are involved in the increase of [Ca 2+ ] i signaling occurring after hypoxia in the SFO areas of adult rats. Moreover, it sheds light on the anti-dipsogenic effects and sodium water and electrolyte balance in the new migrated plateau population.  free access to water and food. The rats were anesthetized with an i.p. injection of chloral hydrate (100 mg/kg body weight) and placed in a stereotaxic apparatus. A guide cannula (AG-8; Eicom, Tokyo, Japan) was implanted into  (TRPV1, TRPV4, HO-2, and HO-1) for Western blot assay as control (input). (B) 300 μ g from each group was subjected to Co-IP using anti-TRPV4 and anti-TRPV1 antibodies, and then immunoprecipitated with beads. The IP lysates (100 μ g) were analyzed by Western blotting using anti-HO-2 and anti-HO-1 antibodies. (D) 300 μ g was subjected to Co-IP using anti-HO-2. 100 μ g immunoprecipitated lysates () were incubated with anti-TRPV4 and anti-TRPV1 antibodies. (C,E) Relative protein expression was calculated. All the experiments were repeated three times. Hypoxia was induced by a rapid superfusion of neurons with a hypoxic medium for 2 min. the third ventricle (coordinates from bregma: 2.2 mm posterior, 0.9 mm lateral, 8.4 mm below skull surface angled at 5° vertical towards the midline) 29 and fixed to the skull with dental cement and small screws according to the coordinates provided by Klippel's atlas. During the 1 week postoperative recovery period, rats were acclimated to handling and the experimental cage used for drug administration and thermistor probe. A total of 93 rats were cannulated, 13 of which were removed from the study because of weight loss in three rats after the operation and because of cannula displacement in 10 rats.

Methods
Genotyping. TRPV4 was obtained from Dr. Zha Ke-xin (Nanjin University, China). TRPV4− /− mice, were previously made by the deletion of exon 12, which encodes the pore-loop and adjacent transmembrane domains 36  Thirst studies. All the rats were prohibited from drinking water one day before the experiment and were randomly divided into 8 groups (10 rats in each group) as follows: 1) Normoxia + saline; 2) Normoxia + 1 μ g gadolinium (Sigma, St. Louis, MO, USA); 3) Normoxia + 1 μ g HC-067047 (Sigma); 4) Normoxia + 10 mg SB-705498 (MedchemExpress, CA, USA); 5) Hypoxia + saline; 6) Hypoxia + 1 μ g gadolinium; 7) Hypoxia + 1 μ g HC-067047; 8) Hypoxia + 10 mg SB-705498. Normoxic and hypoxic animals were placed in an environment of sea level or a hypobaric chamber simulating an elevation of 6000 m for 6 h. 10 min before the experiment, 50 μ L drug solutions were injected into the ventricle by using a microinjection cannula inserted into the guide cannula. The microinjection cannula was connected via a polyethylene tube to a microsyringe containing different drug solutions. The control groups that did not receive drugs were consecutively administered with sterile saline of the same volume. 5 min after the injection, the microinjection cannula was switched to a dummy cannula, and the rats were placed in the normoxic or hypoxic environment. Blood samples were withdrawn from a femoral artery of the normoxia + saline and hypoxia + saline group animals and analyzed at the appropriate time. TRPV4− /− mice and WT mice were also used to observe the water intake under normoxia (sea level) or hypoxia (6000 m) (6 rats in each group). To ensure the accuracy of measurements, gas was eliminated from the drinking water and filled in the bottles. The consumption of water was measured as (the original water weight) − (the current water weight) at 6 h after hypoxia.
Cranial thermometry. The intracranial temperature was measured by High Accuracy Handheld Thermistor Thermometer (Omega, HH41, USA) and Precision Thermistor (Omega, ON-403-PP tubular stainless, USA) which was inserted into the third ventricle by guide cannula. This set of devices has interchangeable sensors with an accuracy up to ± 0.1 °C. The thermal imaging of rats' head was carried out with an infrared thermometer (Fluke, vt04, USA). The external environment was continuously maintained at 25 °C.
Cell culture and overexpression. TRPV4 and TRPV1 overexpressed HEK293 cell lines were set up as described previously 22 . The cells were grown in standard Dulbecco's modified Eagle's medium at 37 °C, pH 7.4, containing 10% fetal bovine serum (Sigma). The pcDNA3.1/TRPV4-V5 plasmid was transiently transfected into HEK293 cells using the Lipofectamine-2000 kit (Invitrogen). The pcDNA3.1/V5-His vector containing a LacZα cDNA insert with the fused V5 epitope was regarded as a positive control while the negative control had no insert. These verified the cloning and transfection efficiencies. When transiently transfected for 48-72 h, 30-60% cells expressed the V5 epitope, as assessed by immunocytochemistry. SFO dissociation. Dissociated SFO neurons were prepared as previously described 37 . Briefly, male SD rats (275-300 g) were decapitated, and their brains were removed and immediately immersed in ice-cold Ca 2+, and Mg 2+ -free HBSS supplemented with 0.05 M sucrose. A tissue block containing the hippocampal commissure and SFO was freed from the brain, using a dissecting microscope (Olympus, SZ61), to separate the SFO from the surrounding tissue. The isolated organ was immersed in 1 mg/mL trypsin and incubated at 37 °C in 5% CO 2 for 15 min. The cells were then suspended in ice-cold Ca 2+ containing HBSS (Invitrogen), filtered through a 100 μ m nylon cell strainer (BD Falcon, NY, USA), and centrifuged at 900 × g for 5 min. The supernatant was removed, cells re-suspended, and again centrifuged, following which the cells were resuspended and plated n glass dishes for imaging (MatTek, Ashland, MA, USA). The cells were further cultured in neurobasal medium (Invitrogen) at 37 °C for a minimum of 24 h before Ca 2+ imaging and immunocytochemistry. siRNA transfection of cells. Cells were grown in 24-well plates to 65-70% confluency. Transfection of HO-2 siRNA (sense: 5′-GGACAUGGAGUAUUUCUUUTT-3′, antisense: 5′-AAAGAAAUACUCCAUGUCCTT-3′) or TRPV4 siRNA (sense: 5′-ACGAGACTAGTGAGACGTG-3′, antisense: 5′-CCTGCTCAACATGCTCATTG-3′) were carried out according to the manufacturer's instructions (Santa Cruz Biotech, Dallas, TX, USA)). Briefly, the cells were incubated with a transfection solution containing a mixture of siRNA (100 nM) and siRNA transfection reagent (5 mL) for 6 h. Subsequently, the transfection solution was replaced with neurobasal medium. After an additional 24 h, the cells were harvested to measure [Ca 2+ ] i . The cells were also consecutively transfected with scrambled siRNA (100 nM) to ensure specific gene silencing. Transfection efficiency was assessed by Western blot and immunofluorescence analysis. Immunofluorescence and hematoxylin-eosin staining. After 24 h of hypoxia, brain tissue was removed, fixed in 4% formaldehyde, cryoprotected in a 30% sucrose solution, sliced into 20 μ m sections using Leica Microsystems Nussloch GmbH (D-69226, Germany), and stained overnight with primary antibodies at 4 °C. Primary SFO neurons and HEK293 cells were cultured on glass coverslips. The treatments' cells were fixed with 4% paraformaldehyde solution and permeabilized using 0.1% Triton X-100. Following blocking with 10% normal goat serum, the cells were incubated with primary antibodies: anti-TRPV4 (1:200), anti-TRPV1 (1:300), anti-MAP-2 (1:300), anti-HO-2 (1:200), and anti-F-actin (1:200) (Abcam, Cambridge, UK).The samples were washed and probed with the appropriate secondary antibodies (Jackson Immunoresearch, West Grove, PA, USA). Micrographs were randomly selected and captured under a fluorescent microscope and analyzed using MagnaFire SP 2.1B software (Olympus, Melville, NY, USA). Additionally, brain sections were also stained with hematoxylin and eosin to observe the SFO region microscopically.
Western blotting. Total protein was extracted and concentrations estimated using a bicinchoninic acid (BCA) protein assay. The following antibodies were used: anti-TRPV4 (1:1000), and anti-HO-2 (1:1000) from Abcam, and anti-β -actin (1:1000, Santa Cruz Biotechnologies). The immunoreactive bands were visualized using enhanced chemiluminescence (Amersham Biosciences, Arlington Heights, IL, USA) according to the manufacturer's instructions. The expression of target proteins was detected on a bio-imaging system (VersaDoc MP 4000; Bio-Rad, Hercules, CA, USA) and ImageJ software analyzed the densitometric values. β -actin was used as an internal control.
Statistical analysis. Data were analyzed by SPSS13.0 software and presented as means ± standard deviation (SD). A one-way analysis of variance (ANOVA) with repeated measures was used to estimate the significance of water intake at different times. The Student's t-test and ANOVA were carried out for two or multiple group comparisons, respectively. Results were statistically significant at P < 0.05.