Neurons of the organum vasculosum of the lamina terminalis (OVLT) are necessary for thirst and vasopressin secretion during hypersmolality in rodents. Recent evidence suggests the osmosensitivity of these neurons is mediated by a gene product encoding the transient receptor potential vanilloid-1 (TRPV1) channel. The purpose of the present study was to determine whether mice lacking the TRPV1 channel had blunted thirst responses and central Fos activation to acute and chronic hyperosmotic stimuli. Surprisingly, TRPV1−/− vs. wild-type mice ingested similar amounts of water after injection (0.5 ml sc) of 0.5 M NaCl and 1.0 M NaCl. Chronic increases in plasma osmolality produced by overnight water deprivation or sole access to a 2% NaCl solution for 48 h produced similar increases in water intake between wild-type and TRPV1−/− mice. There were no differences in cumulative water intakes in response to hypovolemia or isoproterenol. In addition, the number of Fos-positive cells along the lamina terminalis, including the OVLT, as well as the supraoptic nucleus and hypothalamic paraventricular nucleus, was similar between wild-type and TRPV1−/− mice after both acute and chronic osmotic stimulation. These findings indicate that TRPV1 channels are not necessary for osmotically driven thirst or central Fos activation, and thereby suggest that TRPV1 channels are not the primary ion channels that permit the brain to detect changes in plasma sodium concentration or osmolality.
- water intake
- organum vasculosum of the lamina terminalis
changes in the osmotic pressure of extracellular fluid are sensed by neurons in the central nervous system and stimulate the ingestion of water, modulate neuroendocrine function, and alter the activity of the sympathetic nervous system to maintain body fluid homeostasis (2, 7, 14, 34). The most influential set of osmosensitive neurons in the control of body fluid homeostasis is located within the forebrain lamina terminalis (2, 7, 14). This structure contains the median preoptic nucleus and two circumventricular organs, the subfornical organ, (SFO) and the organum vasculosum of the lamina terminalis (OVLT). Lesions of the lamina terminalis completely abolish osmotically induced thirst and vasopressin secretion (3, 8, 15, 16, 37, 38). While more discrete lesions of these various structures have yielded conflicting results across species (15, 16, 29, 37–39), several lines of evidence suggest that the OVLT plays an important role in osmoregulatory responses. First, both acute and chronic increases in systemic osmolality increase Fos immunoreactivity, a marker of neuronal activation, in OVLT neurons (9, 10, 22, 30). Second, in vivo and in vitro electrophysiological studies indicate that OVLT neurons are excited by increases in osmolality (6, 21, 27). In fact, recent evidence indicates that OVLT neurons are intrinsically osmosensitive (5). Lastly, functional studies have demonstrated that localized lesions of the OVLT in dogs largely attenuate thirst and the secretion of vasopressin during plasma hypernatremia (37, 38).
Recently, Ciura and Bourque (5) reported that the osmosensitivity of OVLT neurons is mediated by a gene product encoding the transient receptor potential vanilloid-1 (TRPV1) channel. TRPV channels contain six transmembrane spanning segments and a pore-loop domain that are sensitive to temperature, chemical, and mechanical/osmotic stimuli (20, 24). Although TRPV1 channels in heterogolously expressed systems are not osmosensitive (11, 36), an N-terminal variant of the TRPV1 gene that is insensitive to capsaicin has been proposed to mediate the osmosensitivity of these cells (5). First, OVLT neurons in hypothalamic explants from mice lacking the TRPV1 channel failed to show an increase in neuronal discharge during bath hypertonicity (5). Moreover, the changes in membrane conductance of OVLT neurons during osmotic stimulation were blocked by the broad-spectrum TRPV blocker ruthenium red or absent in TRPV1−/− mice (5). The same authors also reported that TRPV1−/− vs. wild-type mice drank less water in response to an acute osmotic challenge (5). These observations suggest that a gene product encoding TRPV1 may be the putative osmosensory element linking plasma hypernatremia to activation of central osmoregulatory pathways.
The present study sought to confirm and extend these previous observations (5) that a gene product of the TRPV1 channel significantly contributes to osmotically induced thirst and activation of neural circuits involved in body fluid homeostasis. Water intake was stimulated in wild-type and TRPV1−/− mice by acute and chronic sodium loads, as well as nonosmotic stimuli. If a product of the TRPV1 gene is the primary osmotic transduction mechanism of neurons in the forebrain lamina terminalis, we hypothesized that TRPV1−/− mice should ingest less water than wild-type mice in response to osmotic challenges but drink normally in response to nonosmotic stimuli. In contrast to the findings of Ciura and Bourque (5), we unexpectedly observed that TRPV1−/− vs. wild-type mice drank similar amounts of water in response to varying degrees of acute sodium loads and in response to overnight water deprivation or chronic sodium loading produced by sole access to a 2% NaCl drinking solution. Subsequent experiments demonstrate that both acute and chronic sodium loads still lead to activation of neurons in the forebrain lamina terminalis, including the OVLT as well as vasopressinergic and oxytocinergic neurons of the supraoptic nucleus (SON) and hypothalamic paraventricular nucleus (PVH).
Male C57BL/6 mice (8 wk) were obtained from either Charles River Laboratories (Wilmington, MA) (n = 6) or The Jackson Laboratory (Bar Harbor, ME) (n = 8). TRPV1−/− mice (8 wk, n = 12, B6.129X1-TRPV1tm1Jul/J) were obtained from The Jackson Laboratory. The original TRPV1−/− mice were made by deletion of the exon encoding the fifth and sixth transmembrane domains and the pore-loop domain (4). These mice exhibit no TRPV1 mRNA in dorsal root ganglion cells or TRPV1 immunoreactivity in the lumbar dorsal horn (4). Animals were housed individually in a temperature-controlled room (22–23°C) and maintained on a 12:12-h light-dark cycle (lights on at 6 AM) for at least 1 wk before experiments. Experiments began at 8 AM. Standard laboratory chow (Harlan Teklad Global Diet #2018) and deionized water were provided ad libitum except where noted. Baseline body weights of wild-type vs. TRPV1−/− mice did significantly differ (23.6 ± 0.6 vs. 20.9 ± 0.4 g, respectively; P < 0.01), and this difference was maintained throughout the experimental paradigms. All of the experimental procedures were approved by the University of Kentucky Institutional Animal Care and Use Committee.
To determine whether osmotically stimulated thirst depends upon TRPV1 channels, C57BL/6 and TRPV1−/− mice received various treatments to increase water intake, in part, due to hypernatremia or hyperosmolality. In the first set of experiments, mice received an injection (0.5 ml sc) of either 0.15 M, 0.5 M, or 1.0 M NaCl. To account for any differences in body weight, an additional experiment injected mice with 1.0 M NaCl as a function of body weight (10 μl/g body wt sc). Animals were returned to home cages without food and provided immediate access to water. Cumulative intakes (±0.05 ml) were measured at 0, 30, 60, 120, and 180 min. All drinking tests were separated by a minimum of 3 days.
Additional experiments were conducted to determine whether TRPV1 channels contributed to thirst stimulated by chronic increases in osmolality. Both C57BL/6 and TRPV1−/− mice were deprived overnight of water, but not food. In a second set of experiments, mice were given access to food and a 2% NaCl solution, but not water, for 48 h. Then, food was removed, and mice were provided with sole access to water. Cumulative water intakes (±0.05 ml) were measured at 0, 30, 60, 120, and 180 min. Control experiments were also performed in which water intakes were measured in untreated mice over the same time period.
A final set of experiments determined whether TRPV1 channels contribute to thirst stimulated by nonosmotic treatments. C57BL/6 and TRPV1−/− mice were made hypovolemic by two successive injections of the diuretic furosemide (LASIX, 0.5%, 0.25 ml sc) separated by 2 h. Mice were denied access to food throughout and water until 2 h after the second injection. In a separate set of experiments, water intake was stimulated by an injection of the β-adrenergic agonist DL-isoproterenol hydrochloride (0.04 μg/0.25 ml sc). Cumulative water intakes were monitored as described above.
Central Fos activation and immunocytochemistry.
To determine whether TRPV1 channels are the putative osmosensory element that link hypernatremia to the activation of central neural pathways involved in body fluid homeostasis, a final set of experiments examined Fos immunoreactivity in forebrain structures of both C57BL/6 and TRPV1−/− mice after acute and chronic increases in plasma osmolality. At least 2 wk after the completion of all thirst experiments, mice received an injection of 1 M NaCl (0.5 ml sc; n = 4 for both strains) and were returned to home cages without access to water for 90 min. A second group of mice (n = 4 for both strains) were given access to food and a 2% NaCl solution, but not water, for 48 h. Control mice in each strain received an injection of 0.15 M NaCl (n = 2 for each strain) or given access to food and water for 2 days (n ≥ 2 for each strain). Then, animals were anesthetized with isoflurane (3% in 100% O2). Blood was collected via intracardiac puncture and 25-gauge needle into microcentrifuge tubes containing heparin (5 units) and centrifuged (10,000 g). Then, mice were perfused transcardially with isotonic saline (2 ml) followed by 4% paraformaldehyde in 0.1 PBS (10 ml, 4°C). Brains were postfixed in 4% paraformaldehyde overnight at 4°C and transferred to 30% sucrose for 48 h. Forebrains were sectioned at 40 μm using a vibratome. Sections were collected into three serially adjacent sets and stored in cryoprotectant (41) at −25°C. Plasma osmolality was determined in duplicate using a freezing-point depression osmometer (Advanced Instruments, Norwood, MA), hematocrit was measured in duplicate capillary tubes, and plasma protein concentration was measured by protein refractometry (Refractometer Veterinary ATC, VWR International).
Sections were processed for Fos immunoreactivity as described previously by our laboratory (32, 33). Briefly, sections incubated in 0.5% sodium borohydride dissolved in 0.1 M PBS for 30 min prior to incubation with a rabbit polyclonal anti-Fos antibody (PC-38, 1:10,000; EMD Biosciences, San Diego, CA) at 4°C for 72 h. Then, sections were incubated with biotinylated donkey anti-rabbit IgG (1:250; Jackson Immunoresearch, West Grove, PA) for 1.5 h followed by an incubation in an avidin-peroxidase conjugate (ABC Vectastain Kit, Vector Laboratories, Burlingame, CA) for 1 h at room temperature. Then, the tissue was reacted for 4 min in Tris buffer (pH 7.2) containing 0.05% 3,3′-diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO), 2.5% nickel sulfate hexahydrate (Sigma), and 0.003% hydrogen peroxide. The reaction was terminated by several rinses in 0.1 M PBS. Once tissue sections were processed for Fos immunoreactivity, the tissue was incubated with a monoclonal antioxytocin neurophysin (PS38, 1:25; Dr. Hal Gainer, National Institutes of Health) and a polyclonal rabbit anti-vasopressin (T-4563, 1:3,000; Bachem, Torrance, CA) for 48 h at 4°C. Then, sections were incubated overnight at 4°C in donkey anti-mouse AlexaFluor 488 and donkey anti-rabbit AlexaFluor 594 (1:250; Invitrogen, Carlsbad, CA). Sections were mounted onto glass slides, briefly dehydrated through graded concentrations of alcohol, cleared in xylene and coverslipped with Cytoseal 60 (Fisher Scientific, Pittsburgh, PA). All incubations were performed in 0.1 M PBS containing 1% donkey serum (Jackson Immunoresearch), and sections were rinsed between incubations with 0.1 M PBS. The Fos antibody recognizes the ∼55 kDa c-Fos and 62 kDa v-Fos proteins but not ∼39 kDa Jun protein. Specificity was verified by preabsorption with the synthetic peptide (EMD Biosciences, San Diego, CA). The specificity of the oxytocin neurophysin antibody has been verified previously in our laboratory by the absence of immunoreactivity in oxytocin knockout mice (35). The specificity of the vasopressin antibody was confirmed by the absence of immunoreactivity after preabsorption with vasopressin peptide.
Sections were examined under brightfield and fluorescent microscopy using a Nikon E600 Microscope and Nikon DS-Qil Digital Camera with NIS-Elements Imaging Software 2.3 (Nikon, Melville, NY). The number of Fos-positive nuclei was quantified from a representative section in OVLT, SFO, and dorsal and ventral median preoptic nucleus (MnPO). Counts for the dorsal and ventral MnPO were combined. Additional cell counts were performed bilaterally at two different rostrocaudal levels in the SON (−0.58 and −0.82 mm to bregma) and PVH (−0.70 and −0.94 mm to bregma). Coordinates are in reference to bregma from the atlas of Paxinos and Franklin (23). In SON and PVH sections, the number of vasopressin, oxytocin, and double-labeled cells (Fos plus vasopressin or oxytocin) was also analyzed. The number of Fos or double-labeled cells was summed for all SON or PVH sections in each animal.
Genotyping and functional confirmation of TRPV1−/− mice.
To confirm the presence or absence of the TRPV1 allele in C57BL/6 and TRPV1−/− mice, animals were genotyped by PCR, according to the protocol provided by The Jackson Laboratory (http://jaxmice.jax.org/strain/003770.html). Genomic DNA was isolated from tail snips ∼0.5 cm by incubation in 200 μl of 50 mM NaOH for 1 h at 100°C. The resulting digest was buffered with 20 μl 1 M Tris (pH 7.4) and centrifuged (10,000 g, 5 min) to pellet undigested particulate. The supernatant containing genomic DNA was transferred to a new tube and diluted with water (10×). PCR was performed using AccuPrime Taq with buffer II (Invitrogen) and 2 μl of the diluted supernatant. The primers used in the PCR reaction were as specified by The Jackson Laboratory (oIMR0297, 5′-CACGAGACTAGTGAGACGTG-3′; oIMR1561, 5′-CCTGCTCAACATGCTCAT TG-3′; oIMR1562, 5′-TCCTCATGCACTTCAGGAAA-3′) under the following conditions for 40 cycles; 94°C, 30 s; 64°C, 1 min; 72°C, 1 min. Upon completion, the PCR reaction (15 μl) was run on a 1.5% agarose gel containing 0.5 μg/ml ethidium bromide for 1 h at 70 V, and the PCR products were visualized with a UV transilluminator. The wild-type and null alleles generated either a 984-bp or a 600-bp product, respectively, and were compared with a DNA ladder (1 kb Plus DNA Ladder; Invitrogen) that contained 400, 500, 650, 850, 1,000, 1,650, and 2,000 bp.
An additional experiment was performed to confirm that TRPV1−/− mice functionally lacked the TRPV1 channel. A drop of the TRPV1 agonist capsaicin (0.025 mg/ml) was applied topically to the surface of the eye to produce an eye-wiping response. This response consists of the animal wiping the treated eye with its forepaws and typically persists for 1–2 min. Isotonic saline was applied to the contralateral eye on a separate day. The number of eye swipes was counted over a 2-min period in response to application of capsaicin or isotonic saline in every wild-type and TRPV1−/− mice.
Values are expressed as means ± SE. Cumulative water intakes of wild-type and TRPV1−/− mice were analyzed by a group t-test at each time, and multiple comparisons were corrected by a layered Bonferroni analysis (Systat 10.2; Systat Software, San Jose, CA). A similar analysis was performed when water intakes were normalized to body weight (ml/20 g body wt). All other data, including cell counts, eye-wiping response, plasma osmolality, plasma protein, and hematocrit were analyzed by a two-way ANOVA. When significant F values were obtained, independent t-tests with a layered Bonferroni analysis was performed.
Genotyping and functional confirmation of TRPV1−/− mice.
The PCR reaction product generated from tail-snips of wild-type mice contained the expected 984-bp fragment corresponding to the TRPV1 allele (Fig. 1A). In contrast, the 984-bp fragment was absent in TRPV1−/− mice. Instead, the PCR reaction product from TRPV1−/− mice contained the expected 600-bp fragment corresponding to the disrupted allele (4). An additional experiment was performed to directly test whether TRPV1−/− mice lacked a functional TRPV1 channel. Topical application of the TRPV1 agonist capsaicin to the surface of the eye produced an obvious response in wild-type mice that consisted of the forepaws immediately wiping the treated eye (Fig. 1B). This eye-wiping response was completely absent in TRPV1−/− mice. In fact, the number of forepaw wipes to capsaicin in TRPV1−/− mice was not different from those produced by application of isotonic saline to the contralateral eye.
Effect of acute and chronic hyperosmolality on thirst in wild-type and TRPV1−/− mice.
A major goal of the present study was to determine whether mice lacking the TRPV1 channel had attenuated thirst responses to hyperosmolality. Both wild-type and TRPV1−/− mice ingested significant amounts of water in response to 0.5 M NaCl (Fig. 2A) and 1.0 M NaCl (Fig. 2B). Indeed, the water intakes of wild-type and TRPV1−/− mice in response to these treatments displayed a dose-dependent relationship. In contrast to our hypothesis, injection of 0.5 M and 1.0 M NaCl produced similar increases in water intake between TRPV1−/− vs. wild-type mice. When water intake was expressed as a function of body weight (ml/20 g body wt), the cumulative amount of ingested water was again not attenuated in TRPV1−/− mice. Rather, TRPV1−/− vs. wild-type mice drank significantly more water in response to injection of 0.5 M NaCl (180 min: 0.52 ± 0.03 vs. 0.41 ± 0.03 ml/20 g body wt, respectively; P < 0.05) and 1.0 M NaCl (180 min: 1.38 ± 0.06 vs. 1.03 ± 0.06 ml/20 g body wt, respectively; P < 0.01). An additional set of experiments administered the sodium load as a function of body weight (Fig. 2C), and wild-type and TRPV1−/− mice again ingested similar amounts of water in response to 1.0 M NaCl (10 μl/g body wt). When the water intake was expressed as a function of body weight for this test, the amount of ingested water was not different between TRPV1−/− and wild-type mice (180 min: 0.67 ± 0.03 vs. 0.64 ± 0.03 ml/20 g, respectively). The lack of a blunted thirst response to TRPV1−/− mice cannot be attributed to an elevated basal water intake as injection of 0.15 M NaCl did not cause any significant increase in water intake during the 180-min test period in either wild-type (0.07 ± 0.02 ml) or TRPV1−/− mice (0.07 ± 0.02 ml).
In a second set of experiments, thirst was stimulated by chronic elevations in plasma osmolality due to overnight water deprivation or sole access to a 2% NaCl solution for 48 h. Water deprivation stimulated a significant increase in water intake of wild-type and TRPV1−/− mice (Fig. 2D). However, TRPV1−/− mice ingested similar amounts of water as wild-type mice. The only difference between strains was observed at 180 min, and TRPV1−/− mice ingested significantly more water. When water intake was expressed as a function of body weight, TRPV1−/− mice drank significantly more, not less, water than wild-type mice (180 min: 1.19 ± 0.05 vs. 0.83 ± 0.05 ml/20 g body wt, respectively; P < 0.01). Access to 2% NaCl for 48 h significantly increased water intake in both wild-type and TRPV1−/− mice (Fig. 2E). Again, there were no significant differences in the cumulative water intakes (Fig. 2E) or water intake expressed as a function of body weight between TRPV1−/− and wild-type mice (180 min: 0.62 ± 0.06 vs. 0.57 ± 0.04 ml/20 g, respectively). In addition, time control experiments indicate that baseline water intakes were not different between wild-type and TRPV1−/− mice (0.05 ± 0.02 vs. 0.07 ± 0.02 ml, respectively).
Effect of nonosmotic stimuli on thirst in wild-type and TRPV1−/− mice.
To evaluate whether deletion of the TRPV1 channel altered thirst responses to nonosmotic stimuli, wild-type and TRPV1−/− mice were treated with furosemide or isoproterenol. Both treatments significantly increased water intake in both TRPV1−/− and wild-type mice (Fig. 3). However, there were no statistical differences between groups except at 180 min during the furosemide test. Here, TRPV1−/− mice drank less than wild-type mice (Fig. 3A), but this difference was no longer observed when water intake was normalized as a function of body weight (180 min: 0.57 ± 0.03 vs. 0.63 ± 0.04 ml/20 g, respectively). In a similar manner, isoproterenol produced similar increases in water intake of TRPV1−/− vs. wild-type mice regardless of whether the amount of ingested water was analyzed by absolute values (Fig. 3B) or normalized to body weight (180 min: 0.24 ± 0.01 vs. 0.24 ± 0.02 ml/20 g, respectively).
Effect of acute and chronic hyperosmolality on Fos immunoreactivity in the OVLT, SFO, and MnPO.
A second major goal of the present study was to determine whether the presence of the TRPV1 channel was necessary for activation of central neural circuits involved in body fluid homeostasis. Fos immunoreactivity was analyzed in wild-type and TRPV1−/− mice after injection of 1.0 M NaCl (0.5 ml sc) or sole access to a 2% NaCl solution for 48 h. Examples of Fos immunoreactivity in the OVLT, SFO, and dorsal MnPO of wild-type and TRPV1−/− mice in the control treatment group or after injection of 1 M NaCl are illustrated in Fig. 4. Summary data are presented in Fig. 5. Both wild-type and TRPV1−/− mice receiving control treatments expressed low levels of Fos-immunoreactive neurons in the OVLT, SFO, and MnPO. As expected, injection of 1.0 M NaCl or sole access to 2% NaCl for 48 h significantly increased the number of Fos-immunoreactive nuclei in the OVLT, SFO, and MnPO of wild-type mice. Surprisingly, the same treatments significantly increased the number of Fos-positive nuclei in the OVLT, MnPO, and SFO of TRPV1−/− mice. In fact, the number of Fos-immunoreactive cells in the OVLT, SFO, or MnPO after injection of 1 M NaCl or sole access to 2% NaCl was not different between wild-type and TRPV1−/− mice.
Effect of acute and chronic hyperosmolality on Fos immunoreactivity in the SON and PVH.
In addition to the structures along the forebrain lamina terminalis, we analyzed Fos immunoreactivity in the SON and PVH of wild-type and TRPV1−/− mice after injection of 1.0 M NaCl or sole access to 2% NaCl for 48 h. Examples of Fos immunoreactivity in wild-type and TRPV1−/− mice in the SON and PVH are presented in Fig. 6. Summary data are illustrated in Fig. 7. As expected, both treatments significantly increased the number of Fos-positive nuclei in the SON and PVH of wild-type mice. Similar to the findings observed in structures along the lamina terminalis, injection of 1.0 M NaCl and 2% NaCl also significantly increased the number of Fos-positive nuclei in TRPV1−/− mice. Again, the number of Fos-immunoreactive cells in the SON and PVH was not different between wild-type and TRPV1−/− mice. A number of these Fos-positive nuclei were colocalized to either vasopressin- or oxytocin- immunoreactive neurons in the SON (Fig. 6C, Fig. 7A) and PVH (Fig. 6F, Fig. 7B). There were no statistical differences in the number of Fos + vasopressin or Fos + oxytocin double-labeled neurons between wild-type and TRPV1−/− mice in the SON or PVH. In addition, wild-type and TRPV1−/− mice displayed similar numbers of vasopressin (SON: 228 ± 9 vs. 225 ± 13 cells; PVH: 168 ± 9 vs. 184 ± 18 cells, respectively) or oxytocin (SON: 80 ± 5 vs. 71 ± 5 cells; PVH: 175 ± 7 vs. 181 ± 10 cells, respectively) immunoreactive cells.
Effect of acute and chronic hyperosmolality on plasma osmolality, plasma protein, and hematocrit.
Baseline plasma osmolality was not different between wild-type and TRPV1−/− mice (Table 1). Injection of 1.0 M NaCl and sole access to 2% NaCl for 48 h significantly raised plasma osmolality in both groups. However, there were no significant differences in plasma osmolality between wild-type and TRPV1−/− mice after either treatment. As expected, both plasma protein and hematocrit decreased in both strains after injection of 1.0 M NaCl.
Previous studies have suggested that a gene product of the TRPV1 channel contributes to thirst stimulated by an acute sodium load (5). The present study sought to confirm and extend these findings by evaluating thirst responses in wild-type and TRPV1−/− mice during both acute and chronic osmotic stimulation. In contrast to previous report (5), we unexpectedly observed that TRPV1−/− and wild-type mice drank similar amounts of water to varying degrees of acute sodium loads and in response to overnight water deprivation or chronic sodium loading produced by sole access to a 2% NaCl drinking solution. Consistent with these findings, TRPV1−/− and wild-type mice displayed similar numbers of Fos-positive nuclei in the OVLT and several other brain regions in response to acute or chronic osmotic stimulation. These findings indicate that a TRPV1 gene product is not necessary for osmotically driven thirst or central Fos activation and further suggest that TRPV1 is not a candidate ion channel that permits the brain to detect changes in plasma sodium concentration or osmolality.
The OVLT plays a pivotal role in the regulation of water and electrolyte balance through its ability to detect changes in extracellular fluid osmolality and initiate thirst and alter neuroendocrine function (7, 14). In this regard, discrete lesions of the OVLT in dogs significantly attenuate these responses to systemic hypernatremia (37, 38). Recently, Ciura and Bourque (5) suggested that the intrinsic osmosensitivity of OVLT neurons is mediated by TRPV1 as bath hypertonicity evoked an increase in neuronal discharge of OVLT neurons from hypothalamic explants of wild-type but not TRPV1−/− mice. Furthermore, the changes in membrane conductance of OVLT neurons during osmotic stimulation were blocked by the broad-spectrum TRPV blocker ruthenium red or absent in TRPV1−/− mice (5). While TRPV1 channels expressed in heterologous systems are not osmosensitive, Ciura and Bourque (5) suggested that an N-terminal splice variant of the TRPV1 gene that is capsaicin insensitive mediates the osmosensitivity of OVLT neurons.
Altogether, the above observations indicate that a product of the TRPV1 gene is the putative cellular link between systemic hypertonicity and activation of brain osmoregulatory pathways. In fact, Ciura and Bourque (5) reported that TRPV1−/− vs. wild-type mice display an attenuated thirst response to an acute sodium load. The initial goal of the present study was to further characterize the role of TRPV1 in osmotically stimulated thirst under both acute and chronic conditions. However, in marked contrast to the findings of Ciura and Bourque (5), we unexpectedly did not observe a difference in cumulative water intakes at any time between wild-type and TRPV1−/− mice in response to varying degrees of acute osmotic loads. Moreover, the amount of ingested water was not different after water deprivation or chronic sodium loading produced by access to 2% NaCl for 48 h—both treatments are known to stimulate thirst, in part, through cellular dehydration (25, 26). The lack of any differences between wild-type and TRPV1−/− mice cannot be attributed to differences in the intensity of the osmotic stimulus as acute and chronic osmotic stimulation produced similar increases in plasma osmolality between strains. Furthermore, both stimuli evoked similar increases in Fos immunoreactivity of structures associated with water and electrolyte balance, including the OVLT. Collectively, these findings suggest that increases in systemic osmolality can activate brain circuits involved in water and electrolyte balance to initiate thirst despite disruption of the TRPV1 gene.
An explanation for the discrepancy between the findings of Ciura and Bourque (5) and those of the present study is not readily apparent. The previous study administered the sodium load as a function of body weight and the resultant increases in water intake were presented as a function of body weight (5). However, we did not detect any differences in the amount of ingested water between wild-type and TRPV1−/− mice during acute osmotic loads regardless of several factors, including 1) the amount of the sodium load, 2) the sodium load as a function of body weight, or 3) analysis of cumulative water intakes as absolute values or as a function of body weight. Although the sodium loads were administered differently between these studies (ip vs. sc), it is not clear how this difference would affect water intake over the duration of the present experiment. One noteworthy observation regarding the previous findings of Ciura and Bourqe (5) is that deletion of the TRPV1 channel did not eliminate the increase in water intake in response to the acute sodium load. Rather, TRPV1−/− mice drank ∼75% of the water ingested by wild-type animals. This observation seems at odds with the finding that a product of the TRPV1 gene mediates the osmosensitivity of OVLT neurons and lesions of the OVLT significantly attenuate osmotically stimulated thirst (37, 38). However, the contribution of the OVLT to osmotically induced thirst may differ across species. For example, lesion of the OVLT in sheep decreases (15) or has no effect (16) on osmoregulatory thirst, whereas similar lesions in dogs significantly blunts this response (37, 38). On the other hand, lesion of the ventral lamina terminalis largely abolishes such responses in sheep, dogs, rats, and mice (3, 8, 16, 37, 38). Therefore, we cannot exclude the possibility that the lack of a difference in the osmotic responses between TRPV1−/− and wild-type mice reflects either the presence of a different osmosensory transducer in OVLT neurons or the contribution of other structures in the forebrain lamina terminalis with distinct osmosensory transduction mechanisms. Clearly, the deletion of a gene product from development may promote the contribution of other cellular mechanisms that normally do not play an important role. Altogether, the present findings suggest that disruption of the TRPV1 gene does not impact osmoregulatory thirst, but these data do not exclude a role for TRPV1 in the osmosensory transduction of OVLT neurons.
In addition to osmotically-stimulated thirst, OVLT neurons play a pivotal role in the secretion of vasopressin and oxytocin during systemic hyperosmolality or hypernatremia (7, 14). Lesions of the OVLT and/or adjacent areas largely attenuate osmotically stimulated vasopressin secretion and Fos activation in magnocellular neurons of the SON and PVH (9, 13, 18, 31, 37). Interestingly, magnocellular neurons of the SON display some intrinsic osmosensitivity that may be mediated by a gene product of TRPV1−/− (28). In this regard, a previous study reported that TRPV1−/− mice have an elevated basal osmolality and attenuated levels of circulating vasopressin during salt loading (28). While we did not measure circulating vasopressin levels, both acute and chronic osmotic stimulation increased Fos immunoreactivity in vasopressin and oxytocin neurons of wild-type and TRPV1−/− mice. In fact, the number of Fos-positive nuclei was not different between strains regardless of brain region, cell type, or osmotic stimulus. One limit to the interpretation of our findings is that Fos immunoreactivity does not indicate the level of cellular activation in these neurons. Therefore, it remains possible that TRPV1−/− mice have an impaired secretion of vasopressin (and possibly oxytocin) during hyperosmolality. In contrast to the previous findings in TRPV1−/− mice (28), plasma osmolality in our study was not different between strains at baseline or during osmotic stimulation. Our observations are consistent with another report that 24-h water intake and urine osmolality are not different between wild-type and TRPV1−/− mice (1). Differences would be expected if TRPV1−/− mice had impaired osmoregulatory responses.
The mice used in the present study were confirmed through genotyping to contain the appropriate TRPV1 or disrupted allele. In addition, TRPV1−/− mice failed to show a physiological response to topical application of TRPV1 agonist capsaicin to the eye, thereby indicating that the TRPV1 channel was deleted or disrupted in our knockout mice. These mice did show normal thirst responses to nonosmotic stimuli, including hypovolemia produced by furosemide or isoproterenol. These findings suggest that disruption of the TRPV1 gene does not affect hypovolemic thirst.
The present findings indicate that the TRPV1 channel or a product of the TRPV1 gene is not necessary for osmotically stimulated thirst or activation of central pathways involved in water and electrolyte balance. This suggests that the cellular mechanism(s) linking plasma hypernatremia to the central osmoregulatory pathways remain elusive. However, recent data suggest that other members of the TRPV family may be candidate ion channels. For example, the TRPV4 channel is activated by hypotonicity in heterogously expressed systems (11, 36). Interestingly, TRPV4 mRNA and immunoreactivity have been found in neurons of the lamina terminalis (11, 12). However, disruption of the TRPV4 gene yields conflicting data regarding its impact on osmoregulatory responses. Liedtke and Friedman (12) suggest TRPV4-/- mice have impaired thirst and vasopressin secretion, whereas Mizuno and colleagues (17) indicate that TRPV4-/- mice show enhanced vasopressin secretion during a hyperosmotic/hypovolemia challenge. Intracerebroventricular administration of a TRPV4 agonist did not affect thirst after water deprivation or intragastric administration of hypertonic saline (40). On the other hand, TRPV2 ion channels are activated by hypotonicity in heterologous systems and in myocytes (19). If TRPV2 or TRPV4 is the cellular link between plasma hypernatremia and activation of brain osmoregulatory pathways, one unanswered question is how such channels that are activated during hypotonicity lead to an increase in neuronal discharge of osmosensory neurons in the OVLT when many electrophysiological studies suggest that hypertonicity, not hypotonicity, excites these neurons (5, 6, 21, 27). One possibility is that hypertonicity inhibits an OVLT interneuron to increase the discharge of neurons projecting to the MnPO, SON, or PVH. Clearly, further investigation is necessary to identify these important cellular mechanisms.
This research was supported by an American Heart Association Scientist Development Grant 0630202N (S. D. Stocker), National Institutes of Arthritis and Musculoskeletal and Skin Diseases Grant AR053641 (J. J. McCarthy), an American Physiological Society Undergraduate Research Fellowship (A. C. Taylor), and the University of Kentucky College of Medicine.
The authors would like to thank Dr. Hal Gainer (National Institutes of Health) for the generous gift of the OT antibody, Dr. Lu-Yuan Lee for insightful discussions regarding this work, and Dennis Silcox for technical assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society