Prolactin Receptor Knockdown in the Rat Paraventricular Nucleus by a Morpholino-Antisense Oligonucleotide Causes Hypocalcemia and Stress Gas
http://www.100md.com
内分泌学杂志 2005年第8期
Department of Biochemistry and Proteomics, Institute of Genomics and Regenerative Biology, Mie University Graduate School of Medicine (T.F., K.T., T.K., Y.M., M.O.), Mie 514-8507, Japan; Department of Psychiatry, University of Cincinnati Medical Center (T.F., R.R.S.), Cincinnati, Ohio 45267-0559; Department of Legal Medicine, Kyoto Prefectural University of Medicine (K.S.), Kyoto 602-8566, Japan; Department of Integrated Human Sciences, School of Dentistry, Health Sciences University of Hokkaido (A.Y.), Hokkaido 061-0293, Japan; Laboratory of Exercise Biochemistry, Institute of Health and Sport Sciences, University of Tsukuba (H.S.), Ibaraki 305-8574, Japan; and Faculty of Human Health Science, Tokai Gakuen University (K.N.), Aichi 468-8514, Japan
Address all correspondence and requests for reprints to: Dr. Takahiko Fujikawa, Department of Biochemistry, Mie University School of Medicine, 2-174 Edobashi, Tsu, Mie 514-8507, Japan. E-mail: t-fuji@doc.medic.mie-u.ac.jp.
Abstract
Under acute stress conditions in the rat, there is rapid and transient increase in circulating prolactin (PRL). This leads to an elevated expression of the long form of PRLR (PRLR-L) first in the hypothalamus and the choroid plexus. This increase in PRL is involved in the inhibition of stress-induced hypocalcemia and gastric erosion. In this study we used rat PRL and a PRLR morpholino-antisense oligonucleotide to elucidate the mechanism by which hypothalamic PRLR mediates the inhibition of restraint stress in water (RSW)-induced hypocalcemia and gastric erosion. We found that this effect is largely mediated by PRLRs in the paraventricular nucleus (PVN), medial preoptic nucleus, and ventromedial hypothalamus. We also show that when measured after 7 h of RSW, microinjection of the PRLR antisense oligonucleotide into these areas down-regulates RSW-enhanced expression of PRLR-L protein in the PVN and increases the plasma PRL level, but does not affect plasma levels of another hormone, GH. Furthermore, our experiments demonstrated that under nonstress conditions, knockdown of the PRLR in the PVN significantly lowers circulating Ca2+ levels, but does not affect gastric erosion. These results suggest that PRL acting on the PRLR-L in the PVN is one of the critical pathways for regulating circulating Ca2+ levels under both acute stress and nonstress conditions.
Introduction
Ca2+ PLAYS A fundamental role in biological processes, including intracellular signal transduction, bone mineralization, and muscle contraction. In the rat, circulating Ca2+ levels are regulated by PTH (1, 2), calcitonin (3), 1,25-dihydroxyvitamin D, and a blood calcium-lowering peptide (4). Extracellular circulating levels of Ca2+ are about 10,000-fold higher than intracellular levels (5, 6, 7, 8). Stress causes a decrease in circulating Ca2+ levels in rats (5, 6), and the imbalance of Ca2+ raises the likelihood of psychological and physical disorders, such as anxiety, depression, and gastric ulcers (5, 9, 10). In a human study, it was reported that mental stress increases plasma catecholamine levels, and adrenaline decreases plasma levels of ionized and total calcium, but there was also a small increase in plasma PTH levels (11). The hypothalamus, which is a pivotal target for both calciotropic and stress hormones, regulates circulating Ca2+ homeostasis via the vagus nerve, which innervates the stomach and parathyroid glands (12). In particular, the hypothalamic paraventricular (PVN) and ventromedial (VMH) nuclei are thought to regulate the development of hypocalcemia and gastric erosion during stress (5, 6, 7).
The hypothalamus is a target of prolactin (PRL), and it contains abundant PRL receptor (PRLR) mRNA (13) and protein (14, 15). The PRL system in the brain is closely associated with the stress response. A striking increase in serum PRL has been observed during acute stress (16), and an active receptor-mediated transport of PRL from the circulation into the brain compartment has been detected in the choroid plexus (CP) (17). In the rat, two forms of the PRLR, long (PRLR-L) and short, are encoded by the respective mRNAs. We have previously reported that the mRNA expression of the short form of PRLR in the rat brain is very low compared with the expression of the PRLR-L before or after restraint stress in water (RSW) (16), and RSW in the rat causes the up-regulation of only the PRLR-L in the CP (16) and PVN (12). Specifically, after 0.5 h of RSW, there is a rapid and transient increase in circulating PRL, followed 1.5 h later by enhanced expression of PRLR-L in the CP (16). It is thought that circulating PRL controls brain function during stress via activation of the PRLR-L. We have also shown that in the rat, the RSW-induced increase in PRL enhances the expression of PRLR protein and corticotropin-releasing factor mRNA in the PVN and that this is followed by an increase in PRLR-L mRNA in the CP, which, in turn, protects against RSW-induced hypocalcemia and gastric erosion (12).
Histochemical analyses (1) have revealed that PRLR is widely expressed in the brain, including the CP, PVN, medial preoptic nucleus (MPO), VMH, arcuate nucleus, hippocampus, and cortex. Intracerebroventricular injection of rat PRL before RSW suppresses the generation of hypocalcemia and gastric ulcers (12, 18). This is followed by induction of PRLR protein and expression of corticotropin-releasing factor mRNA in the PVN (12). Therefore, although the exact function of hypothalamic PRLR in the stress response is not completely clear, this result suggests that it plays a role in counteracting stress-induced hypocalcemia and gastric erosion.
Several reports show that PRL is a novel neuromodulator of emotionality and hypothalamo-pituitary-adrenal (HPA) axis reactivity in the rat (19). However, little is known about where PRL acts in the central nervous system to cause stress tolerance. Thus, we hypothesized that PRLRs in the hypothalamus, including the PVN, MPO, and VMH, participate in the control of RSW-induced hypocalcemia and gastric erosion. We tested this hypothesis using an animal model of acute stress, RSW. We used rat PRL and PRLR morpholino-antisense oligonucleotides to examine whether PRLR in the hypothalamus and extrahypothalamic regions mediates the regulation of RSW-induced hypocalcemia and gastric erosion. We also used immunohistochemical analysis to examine the effects of pretreating rats with microinjections of rat PRL or PRLR morpholino-antisense oligonucleotides into the hypothalamus and extrahypothalamic regions before RSW.
First, to examine whether rat PRL prevents RSW-induced hypocalcemia and gastric erosions via activation of PRLR in the hypothalamus, we microinjected rats with rat PRL into the hypothalamic and extrahypothalamic regions before exposing them to RSW for 7 h. In a second set of studies, we showed that intracerebroventricular injection of PRLR morpholino-antisense oligonucleotides causes a partial down-regulation of PRLR in the hypothalamus or extrahypothalamus. This result and the immunohistochemical analysis of the PRLR in the CP and PVN showed that the preventive effect of rat PRL on RSW-induced hypocalcemia and gastric erosion was mainly mediated by activation of PRLR in the PVN. Finally, we carried out intra-PVN injections of a PRLR morpholino-antisense oligonucleotide to examine whether PRLR knockdown in the PVN decreases circulating levels of Ca2+ and reduces gastric erosion under nonstressed conditions. We found that elevation of PRLR in the PVN is largely related to the control of circulating levels of Ca2+ under nonstressed conditions. These results demonstrate that PRL helps protect against stress-induced gastric erosion by acting on its receptor in the PVN and that it is one of the critical factors regulating circulating Ca2+ levels under both acute stress and nonstressed conditions.
Materials and Methods
Animals
Male adult (8-wk-old) rats (Sprague Dawley) were purchased from SLC, Inc. (Shizuoka, Japan) and group-housed (three or four per cage) with food (CE-2 rat chow, CLEA, Tokyo, Japan) and drinking water available ad libitum. All animals were handled daily for 2 wk before the start of the experiment. The animals were housed in a temperature- and humidity-controlled room maintained at 23–25 C and 50–60% humidity with a 12-h light, 12-h dark cycle (lights on at 0710 h). The institutional animal care and use committees at Mie University Faculty of Medicine and the University of Cincinnati approved the animal facilities and protocols. All procedures were in accordance with the National Institute of Health’s guidelines regarding the principles of animal care (1996).
RSW
The RSW experiments were initiated at midnight when the rats were highly active, as previously described (12, 16, 20 , 21). Adult male rats were placed in individual wire-mesh restraint cages and immersed tail first in water up to chest level. The water temperature was maintained at 23 ± 0.5 C. After 7 h of RSW, the animals were removed from the cages and killed. The nonstressed rats were killed at the same time as rats exposed to 7 h of RSW. Because dietary restriction was not carried out before the start of RSW, there was food in the stomachs of all animals. For immunohistochemical analyses, animals were decapitated, and their brains were quickly removed, immediately frozen on powdered dry ice, and stored at –80 C until sectioning.
RSW-induced hypocalcemia and gastric erosion
Ca2+ concentrations and pH in whole blood were measured in duplicate using ion-selective electrodes (643 Ca2+/pH analyzer, Ciba Corning, Chiba, Japan) in the presence of balanced heparin (20 IU/ml total blood; Ciba Corning) (6, 7). The blood was adjusted to pH 7.4, and using an equation previously described (22), the pH-independent changes in circulating Ca2+ were determined. At death, stomachs were removed and inflated by injection of 1% formalin (10 ml), then immersed in 1% formalin solution for 30 min. The stomachs were then opened along the greater curvature and examined for gastric lesions. The gastric erosion obtained from this acute stress model was in the shape of a point or a line, and wide areas were not observed. Regardless of the size of the hemorrhage point, we defined all points to be 1 mm, and the gastric lesion index was calculated as the cumulative length (millimeters) of gastric lesions (12, 20). For example, when there were 35 hemorrhage points and a 15-mm hemorrhage line, the gastric lesion index was 50 mm.
Intracerebroventricular injection of PRLR morpholino-antisense oligonucleotides
Rats were anesthetized with sodium pentobarbital [20 mg/kg body weight (bw), ip injection], stereotaxically positioned, and subjected to bilateral intracerebroventricular injection with PRLR morpholino-, control 1, or control 2 antisense oligonucleotides (300 pmol/μl) or saline (1 μl) as described in our previous reports (12, 23). Thirty minutes after an ip injection of sodium pentobarbital, the antisense oligonucleotides or saline was administered bilaterally into the ventricles of the animals. Only the animals that woke from the anesthesia and were standing within 0.5 h after these microinjections were subjected to RSW. After 7 h of RSW, the whole blood Ca2+ level, plasma level of PRL and GH, and the index of gastric erosion were measured. The animals were separated from their original cagemates after anesthesia.
Microinjection of rat PRL or PRLR morpholino-antisense oligonucleotides into the hypothalami and extrahypothalamic regions
Rats were anesthetized with sodium pentobarbital (20 mg/kg bw, ip), stereotaxically positioned, and subjected to bilateral microinjection of rat PRL (rPRL; 500 pg/0.2 μl or 5 ng/0.2 μl), heat-denatured rat PRL (D-rPRL; made by heating at 100 C and pipetting several times over 1 h, 5 pg/0.2 μl), morpholino-antisense oligonucleotides (PRLR, control 1, or control 2; 30 pmol/0.2 μl), or saline (0.2 μl) into the MPO [anterior, 8.08; lateral, 0.3; height (H) 1.6], PVN (anterior, 7.20; lateral, 0.5; H2.0–2.2; upper site), VMH (anterior, 5.86; lateral, 0.9; H0.0), arcuate nucleus (anterior, 5.40; lateral, 0.3; H-0.1), hippocampus (anterior, 5.40; lateral, 1.1; H6.0), or cortex (anterior, 5.40; lateral, 1.2; H8.6) according to the atlas of Paxinos and Watson (24). Thirty minutes after an ip injection of sodium pentobarbital, rPRL, D-rPRL, the antisense oligonucleotides, or saline was administered into the different nuclei of the animals. Only the animals that woke from the anesthesia and were standing within 0.5 h after the microinjection of the antisense oligonucleotides or saline were subjected to RSW, but the animals that were standing within 1 h after the microinjection of rPRL or D-rPRL were subjected to RSW. After 7 h of RSW, the whole blood Ca2+ level, plasma levels of PRL and GH, and the index of gastric erosion were measured.
In contrast, in the study in which antisense oligonucleotides were administered into the PVN under nonstress conditions, all rats (treated with or without saline, control 1 antisense, and PRLR morpholino-antisense) were anesthetized with sodium pentobarbital (20 mg/kg bw, ip), stereotaxically positioned, and subjected to bilateral intra-PVN injection with morpholino-antisense oligonucleotides (PRLR or control 1; 30 pmol/0.2 μl) or saline (0.2 μl) as described in our previous reports (12, 23). Only the animals that woke from the anesthesia and were standing within 1 h after an ip injection of sodium pentobarbital were decapitated (0 h). In addition, 30 min after an ip injection of sodium pentobarbital, the antisense oligonucleotides were administered into the PVN of the animals. Only animals that woke from the anesthesia and were standing were decapitated at 0.5, 2, 4, or 7 h after the intra-PVN injection. Samples of brain and blood were used for the measurement of PRLR expression, whole blood Ca2+ level, and plasma levels of PRL and GH. The animals were separated from their original cagemates after anesthesia and were not restrained.
Immunohistochemical analysis of PRLR-L expression
Brains were taken for immunohistochemistry and sectioned (16 μm thickness) from adult male Sprague Dawley rats after fixation with 4% paraformaldehyde in 0.1 M phosphate buffer. Immunohistochemical analysis used a polyclonal antiserum against a synthetic peptide corresponding to amino acids 29–46 of PRLR-L (provided by F. Talamantes, Department of Molecular, Cell and Developmental Biology, University of California, Santa Cruz, CA) (25). Nonspecific protein binding was blocked by incubating the sections for 3 h with PBS containing 10% normal goat serum. Tissue sections were incubated overnight at room temperature in PBS containing 1:500 PRLR1 antiserum, followed by 1 h at room temperature with PBS containing 0.5% biotinylated goat antirabbit serum and 3% normal goat serum. After washing, the brain sections were incubated for 30 min at room temperature in PBS containing avidin-biotin-peroxidase complex reagent (1:25; Vectastain ABC Kit, Vector Laboratories, Inc., Burlingame, CA). Finally, the tissue sections were stained with 3,3'-diaminobenzidine (Sigma Fast DAB, Sigma-Aldrich Corp., St. Louis, MO) in the presence of hydrogen peroxide, producing a light brown immunostaining.
Histology
At the completion of RSW in all experiments, brains were removed from rats and sectioned (16 μm thickness). The sections were stained with cresyl violet for histological identification of the local injections. Only data obtained from animals that had appropriate sites of microinjection were used for the analyses.
Membrane preparation after dissection of hypothalamus
Each microsomal membrane was prepared from the pool of hypothalami dissected from the brains of two rats as described previously (26). Briefly, tissues were homogenized on ice in homogenization buffer (300 mM sucrose and 50 mM HEPES with protease inhibitors, pH 8.0) using a homogenizer. The homogenates were centrifuged at 20,000 x g for 30 min. The resulting supernatant was centrifuged at 10,000 x g for 1 h to pellet microsomal membrane fractions. The pellets were then washed in a buffer containing 50 mM HEPES, 10 mM EDTA, and protease inhibitors (pH 7.5) and recentrifuged. Membrane proteins were then solubilized by vigorous agitation of the pellet in solubilization buffer (10 mM EDTA, 150 mM NaCl, and 2% Triton X-100, pH 7.5). An aliquot of each membrane preparation was assayed for total protein using a bicinchoninic acid protein assay kit (Pierce Chemical Co., Rockford, IL). Samples were stored at –80 C for later use in Western blotting.
Western blots for hypothalamic PRLR-L
Solubilized membrane proteins from the rat hypothalamus (50 mg protein/lane for each hypothalamus dissection sample) were separated on a 10% sodium dodecyl sulfate-polyacrylamide gel under reducing conditions and transferred to polyvinylidenedifluoride membranes (Millipore Corp., Bedford, MA). Prestained standards (Bio-Rad Laboratories, Hercules, CA) were used as molecular weight markers. The membranes were incubated with 1 μg/ml of the mouse antirat monoclonal antibody U5 (MA1-610, Affinity Bioreagents, Inc., Golden, CO) diluted in PBS containing 2% normal horse serum for 2 h, followed by 30-min incubation with horse antimouse IgG horseradish peroxidase conjugate (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted (1:1:50,000) in PBS containing 2% normal horse serum. The PRLR bands were detected using ECL Plus chemiluminescence reagents and ECL Hyperfilm (Amersham Biosciences, Piscataway NJ) exposure. Densitometric units per square micron of area were measured using image analysis software (NIH Image 1.61, National Institutes of Health, Bethesda, MD).
Measurement of plasma PRL
Plasma PRL concentrations were assayed by ELISA using a rat PRL ELISA kit from Sakamoto Bio Co. Ltd. (Akita, Japan) according to the manufacturer’s instructions.
Measurement of plasma GH
Plasma GH concentrations were assayed by a double-antibody RIA using kits supplied by the National Hormone and Pituitary Program, National Institute of Diabetes and Digestive and Kidney Diseases. The serum samples were incubated with [125I]GH (obtained by GH iodination with chloramine-T), and the antibody-bound hormones were separated by the second antibody technique using commercially available antiserum before radioactivity in the pellets was counted with a -counter. The sensitivity of the GH assay was 0.6 ng/ml. Intra- and interassay variances were less than 3% and 8%, respectively.
Morpholino-antisense oligonucleotides
The morpholino-antisense oligonucleotides were synthesized by Gene Tools, Inc. (Corvallis, OR). The PRLR antisense oligonucleotide corresponded to bases 1–25 from the translation initiation site of rat PRLR mRNA (5'-GGACGAAAGCAAGTGCAGATGGCAT-3'). Control 1 oligonucleotide corresponded to the inverse of the PRLR antisense oligonucleotide sequence (5'-TACGGTAGACGTGAACGAAAGCAGG-3'), and control 2 was a commercial standard control sequence corresponding to a splicing abnormality at nucleotide 705 of erythrocyte ?-globulin mRNA that is found only in thalamia (5'-CCTCTTACCTCAGTTACAATTTATA-3').
Results
Determination of the main site of action of PRL in brain
Microinjections of rPRL (500 pg/0.2 μl and 5 ng/0.2 μl) into the PVN, MPO, or VMH dose-dependently attenuated RSW-induced hypocalcemia and gastric erosion (Table 1). In addition, the intrahippocampal injection of rat PRL dose-dependently suppressed the development of hypocalcemia and gastric erosions, but its antistress effect was weak even compared with that of microinjection into the hypothalamus. In contrast, intra-arcuate nucleus (AN) and intracortex injections of rPRL were without effect (Table 1). There was no great difference in RSW-induced hypocalcemia and gastric erosion between microinjections of heat-denatured rPRL and saline into different regions. Compared with rats injected with rPRL into the MPO or VMH, injection into the PVN had the most powerful effect on RSW-induced hypocalcemia and gastric erosion. The antistress effect on hypocalcemia and gastric erosion was observed when the tip of the injector was within the PVN area (Fig. 1B). Misplaced cannulas in the reunions thalamus nucleus and the ventral reunions thalamus nucleus failed to produce the antistress effect (data not shown; n = 2, respectively).
TABLE 1. Microinjection of PRL in the MPO, PVN, VMH, arcuate nucleus (AN), hippocampus (HIPP) and cortex inhibits hypocalcemia and gastric erosion
FIG. 1. Identification of sites of microinjection of rat PRL in the MPO (A), PVN (B), VMH (C), arcuate nucleus (D), and hippocampus and cortex (E). , Saline (0.2 μl); , D-rPRL (500 ng/0.2 μl); , low rPRL (50 ng/0.2 μl); , high rPRL (500 ng/0.2 μl). All microinjections were bilateral, but for illustration purposes, identification of the local injections is only depicted as unilateral (n = 4–6).
Determination of the main site of action of PRL in brain by microinjection of a PRLR morpholino-antisense oligonucleotide
To examine whether the PVN, MPO, and VMH are the main targets of PRL, we administered a PRLR antisense oligonucleotide into the lateral ventricle, PVN, MPO, VMH, hippocampus, or cortex. We compared the effects of this oligonucleotide with those of two negative control antisense oligonucleotides (C-AS1 and C-AS2) or saline. Intracerebroventricular administration of the receptor antisense oligonucleotide enhanced RSW-induced hypocalcemia and gastric erosion (Fig. 2, A and B). In addition, 7 h after intracerebroventricular administration, the receptor antisense oligonucleotide significantly decreased the expression level of the hypothalamic PRLR-L by 51% compared with the effect of saline and caused obvious down-regulation of the expression level of PRLR-L protein in the CP and PVN (Fig. 2, C and D). The influence of intracerebroventricular administration of C-AS1 or C-AS2 was the same as that of saline administration on RSW-induced hypocalcemia, gastric erosion, and the expression level of hypothalamic PRLR-L. Also, hypocalcemia and gastric erosion elicited by RSW were aggravated when the PRLR antisense oligonucleotide was administered into the PVN, MPO, or VMH, but not when it was injected into the hippocampus or cortex (Fig. 3, A and B). Furthermore, microinjection of the receptor antisense oligonucleotide into the MPO, PVN, or VMH decreased the elevated level of expression of hypothalamic PRLR-L by 50.7%, 49.0%, and 50.8%, respectively, at 7 h of RSW (Fig. 3C), and plasma PRL increased, but plasma GH did not (Fig. 4, A and B). Neither of the control antisense oligonucleotides or saline affected gastric erosion, circulating Ca2+ levels, or the expression level of hypothalamic PRLR-L (Fig. 3). Finally, microinjection of the receptor antisense oligonucleotide into the PVN clearly caused a greater down-regulation of PRLR-L protein in the PVN than microinjection into the cortex, and neither treatment caused down-regulation of PRLR-L protein in the CP (Fig. 5). The expression of PRLR-L protein in the different regions of the hypothalamus, except in the PVN, was not able to be compared after microinjection of the antisense oligonucleotide into the PVN or cortex, because of lower expression levels of PRLR-L compared with those in the PVN. Cannulas misplaced into the bed nucleus of stria terminals, medial division, posterolateral part (n = 3), and the reunions thalamus nucleus (n = 2) did not affect RSW-induced hypocalcemia or gastric erosion (data not shown).
FIG. 2. RSW-induced hypocalcemia (A), gastric erosion (B), hypothalamic PRLR-L expression (C), and PRLR immunoreactivity (D) in the CP and PVN after an intracerebroventricular injection of a morpholino-antisense oligonucleotide against the PRLR. A and B, Intracerebral administration of the PRLR antisense oligonucleotide enhanced RSW-induced hypocalcemia and gastric erosion. C and D, Intracerebral administration of the PRLR antisense oligonucleotide suppressed the expression of hypothalamic PRLR-L, especially in the PVN, and PRLR expression in the CP. Magnification, x50 (PVN) or x100 (CP). Saline, 1 μl; C-AS1, control 1 antisense oligonucleotide, 300 pmol/μl; C-AS2, control 2 antisense oligonucleotide, 300 pmol/μl; PRLR-AS, PRLR antisense oligonucleotide, 300 pmol/μl. Data represent the mean ± SEM from six animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A and B: ***, P < 0.005 vs. saline; ###, P < 0.005 (vs. C-AS1 or C-AS2). C: ****, P < 0.0001 (vs. saline, C-AS1, or C-AS2).
FIG. 3. Effect of microinjection of the PRLR-L morpholino-antisense oligonucleotide into the MPO, PVN, VMH, hippocampus, and cortex on RSW-induced hypocalcemia (A) and gastric erosion (B) and the elevated levels of hypothalamic PRLR-L expression (C). A–C, Administration of the PRLR antisense oligonucleotide into the MPO, PVN, and VMH further enhanced RSW-induced hypocalcemia and gastric erosion and decreased the elevated levels of hypothalamic PRLR-L expression at 7 h of RSW (C). Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; C-AS2, control 2 antisense oligonucleotide, 30 pmol/0.2 μl); PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. The microinjection of C-AS1 and C-AS2 had approximately the same effect as saline on RSW-induced hypocalcemia and gastric erosion. Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A: **, P < 0.01; ***, P < 0.005 (vs. saline, C-AS1, or C-AS2). B: ***, P < 0.005; ****, P < 0.0001 (vs. saline, C-AS1, or C-AS2). C: ***, P < 0.005; ****, P < 0.001 (vs. saline or C-AS1).
FIG. 4. Effect of microinjection of the PRLR-L morpholino-antisense oligonucleotide into the MPO, PVN, VMH, hippocampus, and cortex on plasma levels of PRL (A) and GH (B). A and B, Administration of the PRLR antisense oligonucleotide into the MPO, PVN, and VMH increased plasma PRL, but did not affect plasma GH. Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A: *, P < 0.0005; **, P < 0.0001 (vs. saline or C-AS1). B: ##, P < 0.0001 (vs. non-RSW).
FIG. 5. Immunohistochemical analysis of PRLR expression in the CP and PVN after administration of the PRLR morpholino-antisense oligonucleotide into PVN and cortex. C-AS1, Control 1 antisense oligonucleotide, 30 pmol/0.2 μl); PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. Only the effects of C-AS1 in the CP and PVN are shown as a negative control, because the effects of microinjection of C-AS1 and C-AS2 on receptor expression in both regions were nearly the same as that of saline. Magnification, x50 (PVN) or x100 (CP).
Effect of microinjection of PRLR morpholino-antisense oligonucleotide into the PVN on circulating Ca2+ levels under non-RSW conditions
To determine whether PRLR in the PVN helps regulate circulating Ca2+ levels under non-RSW conditions, we microinjected the receptor antisense oligonucleotide into the PVN of rats under anesthesia. Two hours after the receptor microinjection, there was a significant decrease in circulating Ca2+ levels compared with those in rats treated with or without saline and C-AS1 (Fig. 6A), and the levels continued to decrease after an additional 5 h under non-RSW conditions (Fig. 6B). When examined 7 h after the microinjection, the receptor antisense oligonucleotide significantly decreased the expression level of hypothalamic PRLR-L by 53.4% (Fig. 6C) and caused down-regulation of the expression of PRLR-L protein in the PVN, but did not affect expression in the CP (Fig. 6D). Also, neither microinjection of the control antisense oligonucleotides or saline into the PVN affected circulating Ca2+ levels or PRLR-L protein expression in the PVN compared with that in the nontreated rats within 1 h after administration of anesthesia (data of C-AS2 not shown; Fig. 6, B–D). Furthermore, intra-PVN injection of the receptor antisense oligonucleotide did not cause gastric erosion under non-RSW conditions (data not shown).
FIG. 6. Effect of administration of the PRLR morpholino-antisense oligonucleotide into the PVN on circulating Ca2+ levels under non-RSW conditions. A, Identification of the location of microinjection in the upper area of the PVN. B, Circulating Ca2+ levels before and after intra-PVN microinjections. Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. C, Expression level of hypothalamic PRLR-L 7 h after the microinjection. NT, Nontreatment after anesthesia; S, saline microinjection after anesthesia; C-AS1, C-AS1 microinjection after anesthesia; P-AS, PRLR antisense oligonucleotides, microinjection after anesthesia. D, Expression of PRLR immunoreactivity in the CP and PVN 7 h after microinjections of antisense oligonucleotides. Magnification, x50 (PVN) or x100 (CP). Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. B: *, P < 0.01; **, P < 0.005 (vs. nontreatment, saline, or C-AS1); #, P < 0.05; ##, P < 0.01; ###, P < 0.005 (vs. time zero in rats treated with PRLR-AS). C: ****, P < 0.0001 (vs. nontreatment, saline, or C-AS1).
In rats under non-RSW conditions, intra-PVN injection of the receptor antisense oligonucleotide significantly and time-dependently increased plasma PRL 4 h after the microinjection, but did not affect plasma GH compared with the effect of microinjection of C-AS1 (Table 2).
TABLE 2. Effect of the PRLR morpholino-antisense oligonucleotide into the PVN on circulating levels of PRL and GH in rats before or after the intra-PVN administration under non-RSW conditions
Discussion
In the current study we used local microinjection of a PRLR morpholino-antisense oligonucleotide or rPRL to study the role of PRL in the rat PVN. Our studies clearly show that PRL activation of its receptor in the PVN is largely involved in the regulation of circulating Ca2+ levels under both acute stress and nonstress conditions. We also showed that RSW-induced hypocalcemia and gastric erosion are inhibited by microinjection of rPRL into the PVN, MPO, or VMH. In contrast, administration of a PRLR morpholino-antisense oligonucleotide in the same area aggravated these symptoms. These results suggest that the PVN, MPO, and VMH are the main targets through which endogenous PRL elicits its antistress effects.
Our previous study revealed that PRL is released into the circulation during RSW and that most of its antistress effects on hypocalcemia and gastric erosion are mediated at the hypothalamic level (22). Also, during acute stress, including RSW, peptide fragments of PRL in the circulation appear to enter the cerebrospinal fluid through stress-induced, PRLR-L-mediated transport in the CP (16, 17). PRL in the circulation is thought to mediate a majority of the antistress effects. However, because PRL has been detected in the hypothalamus (13, 14, 26, 27, 28), it is possible that PRL produced by cerebral parenchymal cells has a direct role in its antistress effect.
Several investigations have implicated the hypothalamus, including the PVN, VMH, and lateral hypothalamic area, as well as the limbic system (29), including the amygdala and hippocampus (28, 29, 30), in the development of hypocalcemia and gastric erosion. Moreover, PRLR mRNA and its protein are detected in hypothalamic and extrahypothalamic regions of the brain. PRL acting within the hypothalamus is largely involved in the modulation of physical and emotional stress responses and defensive, maternal, feeding, sexual, and anxiolytic behaviors (19, 31). In our current studies of the various hypothalamic regions, microinjection into the PVN exhibited the most powerful effect on RSW-induced hypocalcemia and gastric erosion. In contrast, intrahippocampal injection only slightly suppressed the development of hypocalcemia and gastric erosion. Because forebrain limbic sites connect with the PVN via -aminobutyric acid-containing neurons in the bed nucleus of the stria terminals, MPO, and hypothalamus (including the VMH) (30), this finding suggests that the antistress effects of PRL occur after activation of the hippocampus and are mediated by receptors in the hypothalami, especially the PVN. Moreover, administration of rPRL into the cortex or arcuate nucleus, which includes dopaminergic neurons that participate in the control of PRL secretion during stress (32), did not suppress RSW-induced hypocalcemia or gastric erosion. These results suggest that PRL activation of hypothalamic PRLRs, mainly in the PVN, MPO, and VMH, is closely involved in the inhibition of RSW-induced hypocalcemia and gastric erosion. The findings also indicate that the weak antistress effect of PRL in the hippocampus might be mediated by inhibitory -aminobutyric acid neurons (29, 30) that project into the PVN, MPO, and VMH. In contrast to the hypothalamic PRLRs, which work together to mediate the antistress effect of PRL, the receptors in the MPO and VMH seem to have opposite effects on the induction of maternal behavior by PRL (33, 34).
Intracerebral administration of PRLR antisense oligonucleotides caused an obvious aggravation of RSW-induced hypocalcemia and gastric erosion, and it down-regulated the expression level of hypothalamic PRLR-L and PRLR-L protein expression in the CP and PVN. In addition, RSW-induced hypocalcemia and gastric erosion were enhanced when PVN expression of PRLR-L protein was down-regulated by microinjection of the receptor antisense oligonucleotide into the PVN, MPO, or VMH. However, there was no effect on PRLR-L protein expression in the PVN or on hypocalcemia or gastric erosion when it was injected into the hippocampus or cortex. Consistent with the results of microinjecting rat PRL, these findings suggest that the PVN, MPO, and VMH are the main target sites through which PRL elicits its antistress effects.
As proof of a successful PRLR antisense strategy, microinjection of the receptor antisense oligonucleotide into the PVN, MPO, or VMH produced an increase in plasma PRL, but did not elevate plasma GH under RSW when expression levels of the hypothalamic PRLR were down-regulated by about 50%, respectively. This provides evidence of the efficacy of the receptor antisense oligonucleotides. Such efficacy was seen after intra-PVN injection under non-RSW conditions.
This result is also in agreement with our previous finding that intracerebroventricular injection of rPRL significantly elevates the expression of PRLR protein and corticotropin-releasing factor mRNA in the PVN (12). Antagonists of the corticotropin-releasing factor type 1 receptor have been shown to almost completely inhibit the antistress effects of rPRL at the brain level, suggesting that neurons expressing corticotropin-releasing factor type 1 receptor are intimately involved in the neuronal pathways downstream of PRLR activation in the brain (35). This supports the idea that PRL mainly acts in the PVN during receptor-mediated activation of the HPA axis. There is also in vitro evidence that PRL stimulates the secretion of corticotropin-releasing factor from the hypothalamus and ACTH secretion from the pituitary (36). The stress-induced increase in PRL has been shown to be closely related to the activity of the HPA axis. Ibotenic acid lesions in the bed nucleus of the stria terminals not only inhibited the increase in HPA activity, but also increased circulating PRL levels under stress (37). The bed nucleus of the stria terminals receives input from the hippocampus, amygdala, and limbic cortex and sends heavy axonal projections to the PVN. The findings and previous results (12) suggest that the rapid release of PRL into the circulation during acute stress activates the HPA axis; as a result, increased PRL increases the release of corticotropin-releasing factor. In contrast, a previous study (19) reported that chronic intracerebroventricular administration of ovine PRL seems to inhibit the HPA axis response to chronic emotional stress. These results suggest that PRL has an opposing effect at the receptor level in the brain hypothalamic area on the response of the HPA axis to acute and chronic stress.
Several studies have indicated that the release of corticotropin-releasing factor (23, 38, 39, 40) and corticosterone (40, 41) during stress exerts a gastroprotective effect. This suggests that during acute stress, activation of the HPA axis as a result of PRL action in the PVN partially contributes to the inhibition of RSW-induced gastric erosion. It is also possible that this pathway prevents hypocalcemia caused by acute stress.
In conclusion, the present study demonstrates that PRL, acting preferentially at receptors in the PVN, MPO, or VMH, is involved in the suppression of hypocalcemia and gastric erosion during acute stress. Our studies of successful down-regulation of the PRLR by its antisense oligonucleotide suggest that PRLRs in the PVN play a critical role in the regulation of circulating Ca2+ not only in rats exposed to acute stress, but also in nonstressed rats. Furthermore, together with our previous studies, our current findings suggest that PRL acting on the hypothalami, particularly in the PVN, is one of the key circulating stress hormones that inhibit gastric erosion.
Acknowledgments
We thank Dr. F. Talamantes for providing the PRLR1 antiserum, and Dr. A. F. Parlow (National Institute of Diabetes and Digestive and Kidney Diseases) for providing rat PRL.
References
Fourman P, Rawnsley K, Davis RH, Jones KH, Morgan DB 1967 Effect of calcium on mental symptoms in partial parathyroid insufficiency. Lancet 2:914–915
Matsui H, Aou J, Ma J, Hori T 1995 Central actions of parathyroid hormone on blood calcium and hypothalamic neuronal activity in the rats. Am J Physiol 268:R21–R27
Gotzman D, Tannenbaum GS 1987 Induction of hypocalcemia by intra-cerebroventricular injection of calcitonin: evidence for control of blood calcium by the nervous system. Brain Res 416:1–6
Persson P, Hakanson R, Axelson J, Sundler F 1989 Gastric releases a blood calcium-lowering peptide from the acid-producing part of the rat stomach. Proc Natl Acad Sci USA 86:2834–2838
Aou S, Ma, J, Hori T, Tashiro N 1994 Hypothalamic linkage in stress-induced hypocalcemia, gastric damage, and emotional behavior in rats. Am J Physiol 267:R38–R43
Ma J, Aou S, Hori T, Ding J 1994 The linkage of stress-induced hypocalcemia, gastric lesions and emotional behavior in Wistar Kyoto rats. Am J Physiol 266:R960–R965
Aou S, Ma J, Shiramine K, Hori T 1993 The stomach is the etiological organ for immobilization-induced hypocalcemia in rats. Am J Physiol 265: R1376–R1379
Keep RF, Ulanski II LJ, Xiang J, Ennis SR, Betz AL 1999 Blood-brain barrier mechanisms involved in brain calcium and potassium homeostasis. Brain Res 815:200–205
Hofmann P, Schwille PO, Thun R 1979 Hypocalcemia during restraint stress in rats: indication that gastric ulcer prophylaxis by exogenous calcium interferes with calcitonin release. Res Exp Med 175:159–168
Linder J, Brismar K, Beck-Friis J, S??f J, Wetterberg L 1989 ?. Calcium and magnesium concentrations in affective disorder: difference between plasma and serum in relation to symptoms. Acta Psychiatr Scand 80:527–5371
Joborn H, Hjemdahl P, Larsson PT, Lithell H, Olsson G, Wide L, Bergstrom R, Ljunghall S 1990 Effects of prolonged adrenaline infusion and of mental stress on plasma minerals and parathyroid hormone. Clin Physiol 10:37–53
Fujikawa T, Soya H, Tamashiro KL, Sakai RR, McEwen BS, Nakai N, Ogata M, Suzuki I, Nakashima K 2004 Prolactin prevents acute stress-induced hypocalcemia and ulcerogenesis by acting in the brain of rat. Endocrinology 145:2006–2013
Crumeyrolle-Arias M, Latouche J, Jammes H, Djiane J, Kelly PA, Reymond HJ, Haour F 1993 Prolactin receptor in the rat hypothalamus: autoradiographic localization and characterization. Neuroendocrinology 57:457–466
Chiu S, Wise PM 1994 Prolactin receptor mRNA localization in the hypothalamus by in situ hybridization. J Neuroendcrinol 6:191–199
Pi X-J, Grattan DR 1998 Differential expression of the two forms of prolactin receptor mRNA within microdissected hypothalamic nuclei of the rat. Mol Brain Res 59:1–12
Fujikawa T, Soya H, Yoshizato H, Sakaguchi K, Doh-Ura K, Tanaka M, Nkashima K 1995 Restraint stress enhances the gene expression of prolactin receptor long form at the choroid plexus. Endocrinology 36:5608–5613
Walsh RJ, Slaby FJ, Posner BI 1987 A receptor-mediated mechanism for the transport of prolactin from blood to cerebrospinal fluid. Endocrinology 120:1846–1850
Drago F, Continella G, Conforto G, Scapagnini U 1985 Prolactin inhibits the development of stress-induced ulcers in the rats. Life Sci 36:191–197
Torner L, Toschi N, Pohlinger A, Landgraf R, Neumann ID 2001 Anxiolytic and anti-stress effects of brain prolactin: improved efficacy of antisense targeting of the prolactin receptor by molecular modeling. J Neurisci 21:3207–3214
Fujikawa T, Yamaguchi A, Morita I, Takeda H, Nishibe S 1996 Protective effects of Acanthopanax senticosus HARMS from Hokkaido and its components on gastric ulcer in restrained cold water stressed rats. Biol Pharm Bull 19:1227–1230
Fujikawa T, Fukuoka H, Alam KSM, Yoshizato H, Higashimoto Y, Soya H, Tanaka M, Nakashima K 2000 Subcutaneously administered prolactin and 20K hGH, but not rGH or 22K hGH, prevent restraint stress-induced gastric ulcers in rats. Endocr J 47:S49–S52
Fogh-Andersen N 1981 Ionized calcium analyzer with a built-in pH correction. Clin Chem 27:1264–1267
Taché Y, Goto Y, Gunion MW, Vale W, Rivier J, Brown M 1983 Inhibition of gastric acid secretion in rats by intracerebral injection of corticotropin-releasing factor. Science 222:935–937
Paxinos G, Watoso C 1986. The rat brain in stereotaxic coordinates. 2nd ed. San Diego: Academic Press
Camarillo IG, Thordarson G, Moffat JG, Van Horn KM, Binart N, Kelly PA, Talamantes F 2001 Prolactin receptor expression in the epithelia and stroma of the rat mammary gland. J Endocrinol 171:85–95
Paut-Pagano L, Roky R, Valatx JL, Kitahama K, Jouvet M 1993 Anatomical distribution of prolactin-like immunoreactivity in the rat brain. Neuroendocrinology 58:682–695
DeVito WJ, Stone S, Avakian C 1991 Stimulation of hypothalamic prolactin release by veratridine and angiotensin II in the female rat: effect of ovariectomy and estradiol administration. Neuroendocrinology 54:391–398
Glavin GB, Murison R, Overmier JB, Pare WP, Bakke HK, Henke PG, Hernandez D 1991 The neurobiology of stress ulcers. Brain Res Rev 16:301–34338
Shiramine K, Aou S, Hori T 1997 Lateral hypothalamic injection of GABAA antagonist induces gastric vagus-mediated hpocalcemia in the rats. Am J Physiol 273:R1492–R1500
Herman JP, Cullinan WE 1997 Neurocircuitry of stress: central control of the hypothalamo-pituitary-adrenocortical axis. Trends Neurosci 20:78–84
Dutt A, Kaplitt MG, Kow LM, Pfaff DW 1994 Prolactin, central nervous system and behavior: a critical review. Neuroendocrinology 59:413–419
Moore KE, Demarest KT, Lookingland KJ 1987 Stress, prolactin and hypothalamic dopaminergic neurons. Neuropharmacology 26:801–808
Bridges RS, Numan M, Ronsheim PM, Mann PE, Lupini CE 1990 Central prolactin infusions stimulate maternal behavior in steroid-treated, nulliparous female rats. Proc Natl Acad Sci USA 87:8003–8007
Bridges RS, Mann PE, Coppeta JS 1999 Hypothalamic involvement in the regulation of maternal behavior in the rat: inhibitory roles for the ventromedial hypothalamus and the dorsal/anterior hypothalamic areas. J Neuroendocrinol 11:259–266
Fujikawa T, Soya H, Alam KSM, Fukuoka H, Yoshizato H, Nakashima K. 2000 Prolactin induces CRF mRNA and blocks stress-induced hypocalcemia and gastric injuries. Gordon Conference Res Abs
Weber RFA, Calogero AE 1983 Prolactin stimulates rat hypothalamic corticotropin-releasing hormone and pituitary adrenocorticotropin secretion in vitro. Neuroendocrinology 36:165–186
Gray TS, Piechowski RA, Yracheta JM, Rittenhouse PA, Bethea CL, Van de Kar LD 1993 Ibotenic acid lesions in the bed nucleus of the stria terminals attenuate conditioned stress-induced increases in prolactin, ACTH and corticosterone. Neuroendocrinology 57:517–524
Kooy A, de Greef WJ, Vreeburg JTM, Hackeng WHL, Ooms MP, Lamberts SWJ, Weber RFA 1990 Evidence for the involvement of corticotropin-releasing factor in the inhibition of gonadotropin release induced by hyperprolactinemia. Neuroendocrinology 51:261–266
Shibasaki T, Yamauchi N, Hotta M, Masuda A, Oono H, Wakabayashi I, Ling N, Demura H 1990 Brain corticotropin-releasing factor acts as inhibitor of stress-induced gastric erosion in rats. Life Sci 47:925–932
Filaretova LP, Filaretov AA, Makara GB 1998 Corticosterone increase inhibits stress-induced gastric erosions in rats. Am J Physiol 274:G1024–G1030
Filaretova L, Maltcev N, Bogdanov A, Levkovich Y 1999 Role of gastric microcirculation in the gastroprotection by glucocorticoids released during water-restraint stress in rats. Chin J Physiol 42:145–152(Takahiko Fujikawa, Kenji )
Address all correspondence and requests for reprints to: Dr. Takahiko Fujikawa, Department of Biochemistry, Mie University School of Medicine, 2-174 Edobashi, Tsu, Mie 514-8507, Japan. E-mail: t-fuji@doc.medic.mie-u.ac.jp.
Abstract
Under acute stress conditions in the rat, there is rapid and transient increase in circulating prolactin (PRL). This leads to an elevated expression of the long form of PRLR (PRLR-L) first in the hypothalamus and the choroid plexus. This increase in PRL is involved in the inhibition of stress-induced hypocalcemia and gastric erosion. In this study we used rat PRL and a PRLR morpholino-antisense oligonucleotide to elucidate the mechanism by which hypothalamic PRLR mediates the inhibition of restraint stress in water (RSW)-induced hypocalcemia and gastric erosion. We found that this effect is largely mediated by PRLRs in the paraventricular nucleus (PVN), medial preoptic nucleus, and ventromedial hypothalamus. We also show that when measured after 7 h of RSW, microinjection of the PRLR antisense oligonucleotide into these areas down-regulates RSW-enhanced expression of PRLR-L protein in the PVN and increases the plasma PRL level, but does not affect plasma levels of another hormone, GH. Furthermore, our experiments demonstrated that under nonstress conditions, knockdown of the PRLR in the PVN significantly lowers circulating Ca2+ levels, but does not affect gastric erosion. These results suggest that PRL acting on the PRLR-L in the PVN is one of the critical pathways for regulating circulating Ca2+ levels under both acute stress and nonstress conditions.
Introduction
Ca2+ PLAYS A fundamental role in biological processes, including intracellular signal transduction, bone mineralization, and muscle contraction. In the rat, circulating Ca2+ levels are regulated by PTH (1, 2), calcitonin (3), 1,25-dihydroxyvitamin D, and a blood calcium-lowering peptide (4). Extracellular circulating levels of Ca2+ are about 10,000-fold higher than intracellular levels (5, 6, 7, 8). Stress causes a decrease in circulating Ca2+ levels in rats (5, 6), and the imbalance of Ca2+ raises the likelihood of psychological and physical disorders, such as anxiety, depression, and gastric ulcers (5, 9, 10). In a human study, it was reported that mental stress increases plasma catecholamine levels, and adrenaline decreases plasma levels of ionized and total calcium, but there was also a small increase in plasma PTH levels (11). The hypothalamus, which is a pivotal target for both calciotropic and stress hormones, regulates circulating Ca2+ homeostasis via the vagus nerve, which innervates the stomach and parathyroid glands (12). In particular, the hypothalamic paraventricular (PVN) and ventromedial (VMH) nuclei are thought to regulate the development of hypocalcemia and gastric erosion during stress (5, 6, 7).
The hypothalamus is a target of prolactin (PRL), and it contains abundant PRL receptor (PRLR) mRNA (13) and protein (14, 15). The PRL system in the brain is closely associated with the stress response. A striking increase in serum PRL has been observed during acute stress (16), and an active receptor-mediated transport of PRL from the circulation into the brain compartment has been detected in the choroid plexus (CP) (17). In the rat, two forms of the PRLR, long (PRLR-L) and short, are encoded by the respective mRNAs. We have previously reported that the mRNA expression of the short form of PRLR in the rat brain is very low compared with the expression of the PRLR-L before or after restraint stress in water (RSW) (16), and RSW in the rat causes the up-regulation of only the PRLR-L in the CP (16) and PVN (12). Specifically, after 0.5 h of RSW, there is a rapid and transient increase in circulating PRL, followed 1.5 h later by enhanced expression of PRLR-L in the CP (16). It is thought that circulating PRL controls brain function during stress via activation of the PRLR-L. We have also shown that in the rat, the RSW-induced increase in PRL enhances the expression of PRLR protein and corticotropin-releasing factor mRNA in the PVN and that this is followed by an increase in PRLR-L mRNA in the CP, which, in turn, protects against RSW-induced hypocalcemia and gastric erosion (12).
Histochemical analyses (1) have revealed that PRLR is widely expressed in the brain, including the CP, PVN, medial preoptic nucleus (MPO), VMH, arcuate nucleus, hippocampus, and cortex. Intracerebroventricular injection of rat PRL before RSW suppresses the generation of hypocalcemia and gastric ulcers (12, 18). This is followed by induction of PRLR protein and expression of corticotropin-releasing factor mRNA in the PVN (12). Therefore, although the exact function of hypothalamic PRLR in the stress response is not completely clear, this result suggests that it plays a role in counteracting stress-induced hypocalcemia and gastric erosion.
Several reports show that PRL is a novel neuromodulator of emotionality and hypothalamo-pituitary-adrenal (HPA) axis reactivity in the rat (19). However, little is known about where PRL acts in the central nervous system to cause stress tolerance. Thus, we hypothesized that PRLRs in the hypothalamus, including the PVN, MPO, and VMH, participate in the control of RSW-induced hypocalcemia and gastric erosion. We tested this hypothesis using an animal model of acute stress, RSW. We used rat PRL and PRLR morpholino-antisense oligonucleotides to examine whether PRLR in the hypothalamus and extrahypothalamic regions mediates the regulation of RSW-induced hypocalcemia and gastric erosion. We also used immunohistochemical analysis to examine the effects of pretreating rats with microinjections of rat PRL or PRLR morpholino-antisense oligonucleotides into the hypothalamus and extrahypothalamic regions before RSW.
First, to examine whether rat PRL prevents RSW-induced hypocalcemia and gastric erosions via activation of PRLR in the hypothalamus, we microinjected rats with rat PRL into the hypothalamic and extrahypothalamic regions before exposing them to RSW for 7 h. In a second set of studies, we showed that intracerebroventricular injection of PRLR morpholino-antisense oligonucleotides causes a partial down-regulation of PRLR in the hypothalamus or extrahypothalamus. This result and the immunohistochemical analysis of the PRLR in the CP and PVN showed that the preventive effect of rat PRL on RSW-induced hypocalcemia and gastric erosion was mainly mediated by activation of PRLR in the PVN. Finally, we carried out intra-PVN injections of a PRLR morpholino-antisense oligonucleotide to examine whether PRLR knockdown in the PVN decreases circulating levels of Ca2+ and reduces gastric erosion under nonstressed conditions. We found that elevation of PRLR in the PVN is largely related to the control of circulating levels of Ca2+ under nonstressed conditions. These results demonstrate that PRL helps protect against stress-induced gastric erosion by acting on its receptor in the PVN and that it is one of the critical factors regulating circulating Ca2+ levels under both acute stress and nonstressed conditions.
Materials and Methods
Animals
Male adult (8-wk-old) rats (Sprague Dawley) were purchased from SLC, Inc. (Shizuoka, Japan) and group-housed (three or four per cage) with food (CE-2 rat chow, CLEA, Tokyo, Japan) and drinking water available ad libitum. All animals were handled daily for 2 wk before the start of the experiment. The animals were housed in a temperature- and humidity-controlled room maintained at 23–25 C and 50–60% humidity with a 12-h light, 12-h dark cycle (lights on at 0710 h). The institutional animal care and use committees at Mie University Faculty of Medicine and the University of Cincinnati approved the animal facilities and protocols. All procedures were in accordance with the National Institute of Health’s guidelines regarding the principles of animal care (1996).
RSW
The RSW experiments were initiated at midnight when the rats were highly active, as previously described (12, 16, 20 , 21). Adult male rats were placed in individual wire-mesh restraint cages and immersed tail first in water up to chest level. The water temperature was maintained at 23 ± 0.5 C. After 7 h of RSW, the animals were removed from the cages and killed. The nonstressed rats were killed at the same time as rats exposed to 7 h of RSW. Because dietary restriction was not carried out before the start of RSW, there was food in the stomachs of all animals. For immunohistochemical analyses, animals were decapitated, and their brains were quickly removed, immediately frozen on powdered dry ice, and stored at –80 C until sectioning.
RSW-induced hypocalcemia and gastric erosion
Ca2+ concentrations and pH in whole blood were measured in duplicate using ion-selective electrodes (643 Ca2+/pH analyzer, Ciba Corning, Chiba, Japan) in the presence of balanced heparin (20 IU/ml total blood; Ciba Corning) (6, 7). The blood was adjusted to pH 7.4, and using an equation previously described (22), the pH-independent changes in circulating Ca2+ were determined. At death, stomachs were removed and inflated by injection of 1% formalin (10 ml), then immersed in 1% formalin solution for 30 min. The stomachs were then opened along the greater curvature and examined for gastric lesions. The gastric erosion obtained from this acute stress model was in the shape of a point or a line, and wide areas were not observed. Regardless of the size of the hemorrhage point, we defined all points to be 1 mm, and the gastric lesion index was calculated as the cumulative length (millimeters) of gastric lesions (12, 20). For example, when there were 35 hemorrhage points and a 15-mm hemorrhage line, the gastric lesion index was 50 mm.
Intracerebroventricular injection of PRLR morpholino-antisense oligonucleotides
Rats were anesthetized with sodium pentobarbital [20 mg/kg body weight (bw), ip injection], stereotaxically positioned, and subjected to bilateral intracerebroventricular injection with PRLR morpholino-, control 1, or control 2 antisense oligonucleotides (300 pmol/μl) or saline (1 μl) as described in our previous reports (12, 23). Thirty minutes after an ip injection of sodium pentobarbital, the antisense oligonucleotides or saline was administered bilaterally into the ventricles of the animals. Only the animals that woke from the anesthesia and were standing within 0.5 h after these microinjections were subjected to RSW. After 7 h of RSW, the whole blood Ca2+ level, plasma level of PRL and GH, and the index of gastric erosion were measured. The animals were separated from their original cagemates after anesthesia.
Microinjection of rat PRL or PRLR morpholino-antisense oligonucleotides into the hypothalami and extrahypothalamic regions
Rats were anesthetized with sodium pentobarbital (20 mg/kg bw, ip), stereotaxically positioned, and subjected to bilateral microinjection of rat PRL (rPRL; 500 pg/0.2 μl or 5 ng/0.2 μl), heat-denatured rat PRL (D-rPRL; made by heating at 100 C and pipetting several times over 1 h, 5 pg/0.2 μl), morpholino-antisense oligonucleotides (PRLR, control 1, or control 2; 30 pmol/0.2 μl), or saline (0.2 μl) into the MPO [anterior, 8.08; lateral, 0.3; height (H) 1.6], PVN (anterior, 7.20; lateral, 0.5; H2.0–2.2; upper site), VMH (anterior, 5.86; lateral, 0.9; H0.0), arcuate nucleus (anterior, 5.40; lateral, 0.3; H-0.1), hippocampus (anterior, 5.40; lateral, 1.1; H6.0), or cortex (anterior, 5.40; lateral, 1.2; H8.6) according to the atlas of Paxinos and Watson (24). Thirty minutes after an ip injection of sodium pentobarbital, rPRL, D-rPRL, the antisense oligonucleotides, or saline was administered into the different nuclei of the animals. Only the animals that woke from the anesthesia and were standing within 0.5 h after the microinjection of the antisense oligonucleotides or saline were subjected to RSW, but the animals that were standing within 1 h after the microinjection of rPRL or D-rPRL were subjected to RSW. After 7 h of RSW, the whole blood Ca2+ level, plasma levels of PRL and GH, and the index of gastric erosion were measured.
In contrast, in the study in which antisense oligonucleotides were administered into the PVN under nonstress conditions, all rats (treated with or without saline, control 1 antisense, and PRLR morpholino-antisense) were anesthetized with sodium pentobarbital (20 mg/kg bw, ip), stereotaxically positioned, and subjected to bilateral intra-PVN injection with morpholino-antisense oligonucleotides (PRLR or control 1; 30 pmol/0.2 μl) or saline (0.2 μl) as described in our previous reports (12, 23). Only the animals that woke from the anesthesia and were standing within 1 h after an ip injection of sodium pentobarbital were decapitated (0 h). In addition, 30 min after an ip injection of sodium pentobarbital, the antisense oligonucleotides were administered into the PVN of the animals. Only animals that woke from the anesthesia and were standing were decapitated at 0.5, 2, 4, or 7 h after the intra-PVN injection. Samples of brain and blood were used for the measurement of PRLR expression, whole blood Ca2+ level, and plasma levels of PRL and GH. The animals were separated from their original cagemates after anesthesia and were not restrained.
Immunohistochemical analysis of PRLR-L expression
Brains were taken for immunohistochemistry and sectioned (16 μm thickness) from adult male Sprague Dawley rats after fixation with 4% paraformaldehyde in 0.1 M phosphate buffer. Immunohistochemical analysis used a polyclonal antiserum against a synthetic peptide corresponding to amino acids 29–46 of PRLR-L (provided by F. Talamantes, Department of Molecular, Cell and Developmental Biology, University of California, Santa Cruz, CA) (25). Nonspecific protein binding was blocked by incubating the sections for 3 h with PBS containing 10% normal goat serum. Tissue sections were incubated overnight at room temperature in PBS containing 1:500 PRLR1 antiserum, followed by 1 h at room temperature with PBS containing 0.5% biotinylated goat antirabbit serum and 3% normal goat serum. After washing, the brain sections were incubated for 30 min at room temperature in PBS containing avidin-biotin-peroxidase complex reagent (1:25; Vectastain ABC Kit, Vector Laboratories, Inc., Burlingame, CA). Finally, the tissue sections were stained with 3,3'-diaminobenzidine (Sigma Fast DAB, Sigma-Aldrich Corp., St. Louis, MO) in the presence of hydrogen peroxide, producing a light brown immunostaining.
Histology
At the completion of RSW in all experiments, brains were removed from rats and sectioned (16 μm thickness). The sections were stained with cresyl violet for histological identification of the local injections. Only data obtained from animals that had appropriate sites of microinjection were used for the analyses.
Membrane preparation after dissection of hypothalamus
Each microsomal membrane was prepared from the pool of hypothalami dissected from the brains of two rats as described previously (26). Briefly, tissues were homogenized on ice in homogenization buffer (300 mM sucrose and 50 mM HEPES with protease inhibitors, pH 8.0) using a homogenizer. The homogenates were centrifuged at 20,000 x g for 30 min. The resulting supernatant was centrifuged at 10,000 x g for 1 h to pellet microsomal membrane fractions. The pellets were then washed in a buffer containing 50 mM HEPES, 10 mM EDTA, and protease inhibitors (pH 7.5) and recentrifuged. Membrane proteins were then solubilized by vigorous agitation of the pellet in solubilization buffer (10 mM EDTA, 150 mM NaCl, and 2% Triton X-100, pH 7.5). An aliquot of each membrane preparation was assayed for total protein using a bicinchoninic acid protein assay kit (Pierce Chemical Co., Rockford, IL). Samples were stored at –80 C for later use in Western blotting.
Western blots for hypothalamic PRLR-L
Solubilized membrane proteins from the rat hypothalamus (50 mg protein/lane for each hypothalamus dissection sample) were separated on a 10% sodium dodecyl sulfate-polyacrylamide gel under reducing conditions and transferred to polyvinylidenedifluoride membranes (Millipore Corp., Bedford, MA). Prestained standards (Bio-Rad Laboratories, Hercules, CA) were used as molecular weight markers. The membranes were incubated with 1 μg/ml of the mouse antirat monoclonal antibody U5 (MA1-610, Affinity Bioreagents, Inc., Golden, CO) diluted in PBS containing 2% normal horse serum for 2 h, followed by 30-min incubation with horse antimouse IgG horseradish peroxidase conjugate (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted (1:1:50,000) in PBS containing 2% normal horse serum. The PRLR bands were detected using ECL Plus chemiluminescence reagents and ECL Hyperfilm (Amersham Biosciences, Piscataway NJ) exposure. Densitometric units per square micron of area were measured using image analysis software (NIH Image 1.61, National Institutes of Health, Bethesda, MD).
Measurement of plasma PRL
Plasma PRL concentrations were assayed by ELISA using a rat PRL ELISA kit from Sakamoto Bio Co. Ltd. (Akita, Japan) according to the manufacturer’s instructions.
Measurement of plasma GH
Plasma GH concentrations were assayed by a double-antibody RIA using kits supplied by the National Hormone and Pituitary Program, National Institute of Diabetes and Digestive and Kidney Diseases. The serum samples were incubated with [125I]GH (obtained by GH iodination with chloramine-T), and the antibody-bound hormones were separated by the second antibody technique using commercially available antiserum before radioactivity in the pellets was counted with a -counter. The sensitivity of the GH assay was 0.6 ng/ml. Intra- and interassay variances were less than 3% and 8%, respectively.
Morpholino-antisense oligonucleotides
The morpholino-antisense oligonucleotides were synthesized by Gene Tools, Inc. (Corvallis, OR). The PRLR antisense oligonucleotide corresponded to bases 1–25 from the translation initiation site of rat PRLR mRNA (5'-GGACGAAAGCAAGTGCAGATGGCAT-3'). Control 1 oligonucleotide corresponded to the inverse of the PRLR antisense oligonucleotide sequence (5'-TACGGTAGACGTGAACGAAAGCAGG-3'), and control 2 was a commercial standard control sequence corresponding to a splicing abnormality at nucleotide 705 of erythrocyte ?-globulin mRNA that is found only in thalamia (5'-CCTCTTACCTCAGTTACAATTTATA-3').
Results
Determination of the main site of action of PRL in brain
Microinjections of rPRL (500 pg/0.2 μl and 5 ng/0.2 μl) into the PVN, MPO, or VMH dose-dependently attenuated RSW-induced hypocalcemia and gastric erosion (Table 1). In addition, the intrahippocampal injection of rat PRL dose-dependently suppressed the development of hypocalcemia and gastric erosions, but its antistress effect was weak even compared with that of microinjection into the hypothalamus. In contrast, intra-arcuate nucleus (AN) and intracortex injections of rPRL were without effect (Table 1). There was no great difference in RSW-induced hypocalcemia and gastric erosion between microinjections of heat-denatured rPRL and saline into different regions. Compared with rats injected with rPRL into the MPO or VMH, injection into the PVN had the most powerful effect on RSW-induced hypocalcemia and gastric erosion. The antistress effect on hypocalcemia and gastric erosion was observed when the tip of the injector was within the PVN area (Fig. 1B). Misplaced cannulas in the reunions thalamus nucleus and the ventral reunions thalamus nucleus failed to produce the antistress effect (data not shown; n = 2, respectively).
TABLE 1. Microinjection of PRL in the MPO, PVN, VMH, arcuate nucleus (AN), hippocampus (HIPP) and cortex inhibits hypocalcemia and gastric erosion
FIG. 1. Identification of sites of microinjection of rat PRL in the MPO (A), PVN (B), VMH (C), arcuate nucleus (D), and hippocampus and cortex (E). , Saline (0.2 μl); , D-rPRL (500 ng/0.2 μl); , low rPRL (50 ng/0.2 μl); , high rPRL (500 ng/0.2 μl). All microinjections were bilateral, but for illustration purposes, identification of the local injections is only depicted as unilateral (n = 4–6).
Determination of the main site of action of PRL in brain by microinjection of a PRLR morpholino-antisense oligonucleotide
To examine whether the PVN, MPO, and VMH are the main targets of PRL, we administered a PRLR antisense oligonucleotide into the lateral ventricle, PVN, MPO, VMH, hippocampus, or cortex. We compared the effects of this oligonucleotide with those of two negative control antisense oligonucleotides (C-AS1 and C-AS2) or saline. Intracerebroventricular administration of the receptor antisense oligonucleotide enhanced RSW-induced hypocalcemia and gastric erosion (Fig. 2, A and B). In addition, 7 h after intracerebroventricular administration, the receptor antisense oligonucleotide significantly decreased the expression level of the hypothalamic PRLR-L by 51% compared with the effect of saline and caused obvious down-regulation of the expression level of PRLR-L protein in the CP and PVN (Fig. 2, C and D). The influence of intracerebroventricular administration of C-AS1 or C-AS2 was the same as that of saline administration on RSW-induced hypocalcemia, gastric erosion, and the expression level of hypothalamic PRLR-L. Also, hypocalcemia and gastric erosion elicited by RSW were aggravated when the PRLR antisense oligonucleotide was administered into the PVN, MPO, or VMH, but not when it was injected into the hippocampus or cortex (Fig. 3, A and B). Furthermore, microinjection of the receptor antisense oligonucleotide into the MPO, PVN, or VMH decreased the elevated level of expression of hypothalamic PRLR-L by 50.7%, 49.0%, and 50.8%, respectively, at 7 h of RSW (Fig. 3C), and plasma PRL increased, but plasma GH did not (Fig. 4, A and B). Neither of the control antisense oligonucleotides or saline affected gastric erosion, circulating Ca2+ levels, or the expression level of hypothalamic PRLR-L (Fig. 3). Finally, microinjection of the receptor antisense oligonucleotide into the PVN clearly caused a greater down-regulation of PRLR-L protein in the PVN than microinjection into the cortex, and neither treatment caused down-regulation of PRLR-L protein in the CP (Fig. 5). The expression of PRLR-L protein in the different regions of the hypothalamus, except in the PVN, was not able to be compared after microinjection of the antisense oligonucleotide into the PVN or cortex, because of lower expression levels of PRLR-L compared with those in the PVN. Cannulas misplaced into the bed nucleus of stria terminals, medial division, posterolateral part (n = 3), and the reunions thalamus nucleus (n = 2) did not affect RSW-induced hypocalcemia or gastric erosion (data not shown).
FIG. 2. RSW-induced hypocalcemia (A), gastric erosion (B), hypothalamic PRLR-L expression (C), and PRLR immunoreactivity (D) in the CP and PVN after an intracerebroventricular injection of a morpholino-antisense oligonucleotide against the PRLR. A and B, Intracerebral administration of the PRLR antisense oligonucleotide enhanced RSW-induced hypocalcemia and gastric erosion. C and D, Intracerebral administration of the PRLR antisense oligonucleotide suppressed the expression of hypothalamic PRLR-L, especially in the PVN, and PRLR expression in the CP. Magnification, x50 (PVN) or x100 (CP). Saline, 1 μl; C-AS1, control 1 antisense oligonucleotide, 300 pmol/μl; C-AS2, control 2 antisense oligonucleotide, 300 pmol/μl; PRLR-AS, PRLR antisense oligonucleotide, 300 pmol/μl. Data represent the mean ± SEM from six animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A and B: ***, P < 0.005 vs. saline; ###, P < 0.005 (vs. C-AS1 or C-AS2). C: ****, P < 0.0001 (vs. saline, C-AS1, or C-AS2).
FIG. 3. Effect of microinjection of the PRLR-L morpholino-antisense oligonucleotide into the MPO, PVN, VMH, hippocampus, and cortex on RSW-induced hypocalcemia (A) and gastric erosion (B) and the elevated levels of hypothalamic PRLR-L expression (C). A–C, Administration of the PRLR antisense oligonucleotide into the MPO, PVN, and VMH further enhanced RSW-induced hypocalcemia and gastric erosion and decreased the elevated levels of hypothalamic PRLR-L expression at 7 h of RSW (C). Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; C-AS2, control 2 antisense oligonucleotide, 30 pmol/0.2 μl); PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. The microinjection of C-AS1 and C-AS2 had approximately the same effect as saline on RSW-induced hypocalcemia and gastric erosion. Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A: **, P < 0.01; ***, P < 0.005 (vs. saline, C-AS1, or C-AS2). B: ***, P < 0.005; ****, P < 0.0001 (vs. saline, C-AS1, or C-AS2). C: ***, P < 0.005; ****, P < 0.001 (vs. saline or C-AS1).
FIG. 4. Effect of microinjection of the PRLR-L morpholino-antisense oligonucleotide into the MPO, PVN, VMH, hippocampus, and cortex on plasma levels of PRL (A) and GH (B). A and B, Administration of the PRLR antisense oligonucleotide into the MPO, PVN, and VMH increased plasma PRL, but did not affect plasma GH. Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. A: *, P < 0.0005; **, P < 0.0001 (vs. saline or C-AS1). B: ##, P < 0.0001 (vs. non-RSW).
FIG. 5. Immunohistochemical analysis of PRLR expression in the CP and PVN after administration of the PRLR morpholino-antisense oligonucleotide into PVN and cortex. C-AS1, Control 1 antisense oligonucleotide, 30 pmol/0.2 μl); PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. Only the effects of C-AS1 in the CP and PVN are shown as a negative control, because the effects of microinjection of C-AS1 and C-AS2 on receptor expression in both regions were nearly the same as that of saline. Magnification, x50 (PVN) or x100 (CP).
Effect of microinjection of PRLR morpholino-antisense oligonucleotide into the PVN on circulating Ca2+ levels under non-RSW conditions
To determine whether PRLR in the PVN helps regulate circulating Ca2+ levels under non-RSW conditions, we microinjected the receptor antisense oligonucleotide into the PVN of rats under anesthesia. Two hours after the receptor microinjection, there was a significant decrease in circulating Ca2+ levels compared with those in rats treated with or without saline and C-AS1 (Fig. 6A), and the levels continued to decrease after an additional 5 h under non-RSW conditions (Fig. 6B). When examined 7 h after the microinjection, the receptor antisense oligonucleotide significantly decreased the expression level of hypothalamic PRLR-L by 53.4% (Fig. 6C) and caused down-regulation of the expression of PRLR-L protein in the PVN, but did not affect expression in the CP (Fig. 6D). Also, neither microinjection of the control antisense oligonucleotides or saline into the PVN affected circulating Ca2+ levels or PRLR-L protein expression in the PVN compared with that in the nontreated rats within 1 h after administration of anesthesia (data of C-AS2 not shown; Fig. 6, B–D). Furthermore, intra-PVN injection of the receptor antisense oligonucleotide did not cause gastric erosion under non-RSW conditions (data not shown).
FIG. 6. Effect of administration of the PRLR morpholino-antisense oligonucleotide into the PVN on circulating Ca2+ levels under non-RSW conditions. A, Identification of the location of microinjection in the upper area of the PVN. B, Circulating Ca2+ levels before and after intra-PVN microinjections. Saline, 0.2 μl; C-AS1, control 1 antisense oligonucleotide, 30 pmol/0.2 μl; PRLR-AS, PRLR antisense oligonucleotide, 30 pmol/0.2 μl. C, Expression level of hypothalamic PRLR-L 7 h after the microinjection. NT, Nontreatment after anesthesia; S, saline microinjection after anesthesia; C-AS1, C-AS1 microinjection after anesthesia; P-AS, PRLR antisense oligonucleotides, microinjection after anesthesia. D, Expression of PRLR immunoreactivity in the CP and PVN 7 h after microinjections of antisense oligonucleotides. Magnification, x50 (PVN) or x100 (CP). Data represent the mean ± SEM from five animals. The significance of difference between the values was analyzed by a Bartlett test, followed by a Fisher protected least significant difference post hoc test. P < 0.05 was considered significant. B: *, P < 0.01; **, P < 0.005 (vs. nontreatment, saline, or C-AS1); #, P < 0.05; ##, P < 0.01; ###, P < 0.005 (vs. time zero in rats treated with PRLR-AS). C: ****, P < 0.0001 (vs. nontreatment, saline, or C-AS1).
In rats under non-RSW conditions, intra-PVN injection of the receptor antisense oligonucleotide significantly and time-dependently increased plasma PRL 4 h after the microinjection, but did not affect plasma GH compared with the effect of microinjection of C-AS1 (Table 2).
TABLE 2. Effect of the PRLR morpholino-antisense oligonucleotide into the PVN on circulating levels of PRL and GH in rats before or after the intra-PVN administration under non-RSW conditions
Discussion
In the current study we used local microinjection of a PRLR morpholino-antisense oligonucleotide or rPRL to study the role of PRL in the rat PVN. Our studies clearly show that PRL activation of its receptor in the PVN is largely involved in the regulation of circulating Ca2+ levels under both acute stress and nonstress conditions. We also showed that RSW-induced hypocalcemia and gastric erosion are inhibited by microinjection of rPRL into the PVN, MPO, or VMH. In contrast, administration of a PRLR morpholino-antisense oligonucleotide in the same area aggravated these symptoms. These results suggest that the PVN, MPO, and VMH are the main targets through which endogenous PRL elicits its antistress effects.
Our previous study revealed that PRL is released into the circulation during RSW and that most of its antistress effects on hypocalcemia and gastric erosion are mediated at the hypothalamic level (22). Also, during acute stress, including RSW, peptide fragments of PRL in the circulation appear to enter the cerebrospinal fluid through stress-induced, PRLR-L-mediated transport in the CP (16, 17). PRL in the circulation is thought to mediate a majority of the antistress effects. However, because PRL has been detected in the hypothalamus (13, 14, 26, 27, 28), it is possible that PRL produced by cerebral parenchymal cells has a direct role in its antistress effect.
Several investigations have implicated the hypothalamus, including the PVN, VMH, and lateral hypothalamic area, as well as the limbic system (29), including the amygdala and hippocampus (28, 29, 30), in the development of hypocalcemia and gastric erosion. Moreover, PRLR mRNA and its protein are detected in hypothalamic and extrahypothalamic regions of the brain. PRL acting within the hypothalamus is largely involved in the modulation of physical and emotional stress responses and defensive, maternal, feeding, sexual, and anxiolytic behaviors (19, 31). In our current studies of the various hypothalamic regions, microinjection into the PVN exhibited the most powerful effect on RSW-induced hypocalcemia and gastric erosion. In contrast, intrahippocampal injection only slightly suppressed the development of hypocalcemia and gastric erosion. Because forebrain limbic sites connect with the PVN via -aminobutyric acid-containing neurons in the bed nucleus of the stria terminals, MPO, and hypothalamus (including the VMH) (30), this finding suggests that the antistress effects of PRL occur after activation of the hippocampus and are mediated by receptors in the hypothalami, especially the PVN. Moreover, administration of rPRL into the cortex or arcuate nucleus, which includes dopaminergic neurons that participate in the control of PRL secretion during stress (32), did not suppress RSW-induced hypocalcemia or gastric erosion. These results suggest that PRL activation of hypothalamic PRLRs, mainly in the PVN, MPO, and VMH, is closely involved in the inhibition of RSW-induced hypocalcemia and gastric erosion. The findings also indicate that the weak antistress effect of PRL in the hippocampus might be mediated by inhibitory -aminobutyric acid neurons (29, 30) that project into the PVN, MPO, and VMH. In contrast to the hypothalamic PRLRs, which work together to mediate the antistress effect of PRL, the receptors in the MPO and VMH seem to have opposite effects on the induction of maternal behavior by PRL (33, 34).
Intracerebral administration of PRLR antisense oligonucleotides caused an obvious aggravation of RSW-induced hypocalcemia and gastric erosion, and it down-regulated the expression level of hypothalamic PRLR-L and PRLR-L protein expression in the CP and PVN. In addition, RSW-induced hypocalcemia and gastric erosion were enhanced when PVN expression of PRLR-L protein was down-regulated by microinjection of the receptor antisense oligonucleotide into the PVN, MPO, or VMH. However, there was no effect on PRLR-L protein expression in the PVN or on hypocalcemia or gastric erosion when it was injected into the hippocampus or cortex. Consistent with the results of microinjecting rat PRL, these findings suggest that the PVN, MPO, and VMH are the main target sites through which PRL elicits its antistress effects.
As proof of a successful PRLR antisense strategy, microinjection of the receptor antisense oligonucleotide into the PVN, MPO, or VMH produced an increase in plasma PRL, but did not elevate plasma GH under RSW when expression levels of the hypothalamic PRLR were down-regulated by about 50%, respectively. This provides evidence of the efficacy of the receptor antisense oligonucleotides. Such efficacy was seen after intra-PVN injection under non-RSW conditions.
This result is also in agreement with our previous finding that intracerebroventricular injection of rPRL significantly elevates the expression of PRLR protein and corticotropin-releasing factor mRNA in the PVN (12). Antagonists of the corticotropin-releasing factor type 1 receptor have been shown to almost completely inhibit the antistress effects of rPRL at the brain level, suggesting that neurons expressing corticotropin-releasing factor type 1 receptor are intimately involved in the neuronal pathways downstream of PRLR activation in the brain (35). This supports the idea that PRL mainly acts in the PVN during receptor-mediated activation of the HPA axis. There is also in vitro evidence that PRL stimulates the secretion of corticotropin-releasing factor from the hypothalamus and ACTH secretion from the pituitary (36). The stress-induced increase in PRL has been shown to be closely related to the activity of the HPA axis. Ibotenic acid lesions in the bed nucleus of the stria terminals not only inhibited the increase in HPA activity, but also increased circulating PRL levels under stress (37). The bed nucleus of the stria terminals receives input from the hippocampus, amygdala, and limbic cortex and sends heavy axonal projections to the PVN. The findings and previous results (12) suggest that the rapid release of PRL into the circulation during acute stress activates the HPA axis; as a result, increased PRL increases the release of corticotropin-releasing factor. In contrast, a previous study (19) reported that chronic intracerebroventricular administration of ovine PRL seems to inhibit the HPA axis response to chronic emotional stress. These results suggest that PRL has an opposing effect at the receptor level in the brain hypothalamic area on the response of the HPA axis to acute and chronic stress.
Several studies have indicated that the release of corticotropin-releasing factor (23, 38, 39, 40) and corticosterone (40, 41) during stress exerts a gastroprotective effect. This suggests that during acute stress, activation of the HPA axis as a result of PRL action in the PVN partially contributes to the inhibition of RSW-induced gastric erosion. It is also possible that this pathway prevents hypocalcemia caused by acute stress.
In conclusion, the present study demonstrates that PRL, acting preferentially at receptors in the PVN, MPO, or VMH, is involved in the suppression of hypocalcemia and gastric erosion during acute stress. Our studies of successful down-regulation of the PRLR by its antisense oligonucleotide suggest that PRLRs in the PVN play a critical role in the regulation of circulating Ca2+ not only in rats exposed to acute stress, but also in nonstressed rats. Furthermore, together with our previous studies, our current findings suggest that PRL acting on the hypothalami, particularly in the PVN, is one of the key circulating stress hormones that inhibit gastric erosion.
Acknowledgments
We thank Dr. F. Talamantes for providing the PRLR1 antiserum, and Dr. A. F. Parlow (National Institute of Diabetes and Digestive and Kidney Diseases) for providing rat PRL.
References
Fourman P, Rawnsley K, Davis RH, Jones KH, Morgan DB 1967 Effect of calcium on mental symptoms in partial parathyroid insufficiency. Lancet 2:914–915
Matsui H, Aou J, Ma J, Hori T 1995 Central actions of parathyroid hormone on blood calcium and hypothalamic neuronal activity in the rats. Am J Physiol 268:R21–R27
Gotzman D, Tannenbaum GS 1987 Induction of hypocalcemia by intra-cerebroventricular injection of calcitonin: evidence for control of blood calcium by the nervous system. Brain Res 416:1–6
Persson P, Hakanson R, Axelson J, Sundler F 1989 Gastric releases a blood calcium-lowering peptide from the acid-producing part of the rat stomach. Proc Natl Acad Sci USA 86:2834–2838
Aou S, Ma, J, Hori T, Tashiro N 1994 Hypothalamic linkage in stress-induced hypocalcemia, gastric damage, and emotional behavior in rats. Am J Physiol 267:R38–R43
Ma J, Aou S, Hori T, Ding J 1994 The linkage of stress-induced hypocalcemia, gastric lesions and emotional behavior in Wistar Kyoto rats. Am J Physiol 266:R960–R965
Aou S, Ma J, Shiramine K, Hori T 1993 The stomach is the etiological organ for immobilization-induced hypocalcemia in rats. Am J Physiol 265: R1376–R1379
Keep RF, Ulanski II LJ, Xiang J, Ennis SR, Betz AL 1999 Blood-brain barrier mechanisms involved in brain calcium and potassium homeostasis. Brain Res 815:200–205
Hofmann P, Schwille PO, Thun R 1979 Hypocalcemia during restraint stress in rats: indication that gastric ulcer prophylaxis by exogenous calcium interferes with calcitonin release. Res Exp Med 175:159–168
Linder J, Brismar K, Beck-Friis J, S??f J, Wetterberg L 1989 ?. Calcium and magnesium concentrations in affective disorder: difference between plasma and serum in relation to symptoms. Acta Psychiatr Scand 80:527–5371
Joborn H, Hjemdahl P, Larsson PT, Lithell H, Olsson G, Wide L, Bergstrom R, Ljunghall S 1990 Effects of prolonged adrenaline infusion and of mental stress on plasma minerals and parathyroid hormone. Clin Physiol 10:37–53
Fujikawa T, Soya H, Tamashiro KL, Sakai RR, McEwen BS, Nakai N, Ogata M, Suzuki I, Nakashima K 2004 Prolactin prevents acute stress-induced hypocalcemia and ulcerogenesis by acting in the brain of rat. Endocrinology 145:2006–2013
Crumeyrolle-Arias M, Latouche J, Jammes H, Djiane J, Kelly PA, Reymond HJ, Haour F 1993 Prolactin receptor in the rat hypothalamus: autoradiographic localization and characterization. Neuroendocrinology 57:457–466
Chiu S, Wise PM 1994 Prolactin receptor mRNA localization in the hypothalamus by in situ hybridization. J Neuroendcrinol 6:191–199
Pi X-J, Grattan DR 1998 Differential expression of the two forms of prolactin receptor mRNA within microdissected hypothalamic nuclei of the rat. Mol Brain Res 59:1–12
Fujikawa T, Soya H, Yoshizato H, Sakaguchi K, Doh-Ura K, Tanaka M, Nkashima K 1995 Restraint stress enhances the gene expression of prolactin receptor long form at the choroid plexus. Endocrinology 36:5608–5613
Walsh RJ, Slaby FJ, Posner BI 1987 A receptor-mediated mechanism for the transport of prolactin from blood to cerebrospinal fluid. Endocrinology 120:1846–1850
Drago F, Continella G, Conforto G, Scapagnini U 1985 Prolactin inhibits the development of stress-induced ulcers in the rats. Life Sci 36:191–197
Torner L, Toschi N, Pohlinger A, Landgraf R, Neumann ID 2001 Anxiolytic and anti-stress effects of brain prolactin: improved efficacy of antisense targeting of the prolactin receptor by molecular modeling. J Neurisci 21:3207–3214
Fujikawa T, Yamaguchi A, Morita I, Takeda H, Nishibe S 1996 Protective effects of Acanthopanax senticosus HARMS from Hokkaido and its components on gastric ulcer in restrained cold water stressed rats. Biol Pharm Bull 19:1227–1230
Fujikawa T, Fukuoka H, Alam KSM, Yoshizato H, Higashimoto Y, Soya H, Tanaka M, Nakashima K 2000 Subcutaneously administered prolactin and 20K hGH, but not rGH or 22K hGH, prevent restraint stress-induced gastric ulcers in rats. Endocr J 47:S49–S52
Fogh-Andersen N 1981 Ionized calcium analyzer with a built-in pH correction. Clin Chem 27:1264–1267
Taché Y, Goto Y, Gunion MW, Vale W, Rivier J, Brown M 1983 Inhibition of gastric acid secretion in rats by intracerebral injection of corticotropin-releasing factor. Science 222:935–937
Paxinos G, Watoso C 1986. The rat brain in stereotaxic coordinates. 2nd ed. San Diego: Academic Press
Camarillo IG, Thordarson G, Moffat JG, Van Horn KM, Binart N, Kelly PA, Talamantes F 2001 Prolactin receptor expression in the epithelia and stroma of the rat mammary gland. J Endocrinol 171:85–95
Paut-Pagano L, Roky R, Valatx JL, Kitahama K, Jouvet M 1993 Anatomical distribution of prolactin-like immunoreactivity in the rat brain. Neuroendocrinology 58:682–695
DeVito WJ, Stone S, Avakian C 1991 Stimulation of hypothalamic prolactin release by veratridine and angiotensin II in the female rat: effect of ovariectomy and estradiol administration. Neuroendocrinology 54:391–398
Glavin GB, Murison R, Overmier JB, Pare WP, Bakke HK, Henke PG, Hernandez D 1991 The neurobiology of stress ulcers. Brain Res Rev 16:301–34338
Shiramine K, Aou S, Hori T 1997 Lateral hypothalamic injection of GABAA antagonist induces gastric vagus-mediated hpocalcemia in the rats. Am J Physiol 273:R1492–R1500
Herman JP, Cullinan WE 1997 Neurocircuitry of stress: central control of the hypothalamo-pituitary-adrenocortical axis. Trends Neurosci 20:78–84
Dutt A, Kaplitt MG, Kow LM, Pfaff DW 1994 Prolactin, central nervous system and behavior: a critical review. Neuroendocrinology 59:413–419
Moore KE, Demarest KT, Lookingland KJ 1987 Stress, prolactin and hypothalamic dopaminergic neurons. Neuropharmacology 26:801–808
Bridges RS, Numan M, Ronsheim PM, Mann PE, Lupini CE 1990 Central prolactin infusions stimulate maternal behavior in steroid-treated, nulliparous female rats. Proc Natl Acad Sci USA 87:8003–8007
Bridges RS, Mann PE, Coppeta JS 1999 Hypothalamic involvement in the regulation of maternal behavior in the rat: inhibitory roles for the ventromedial hypothalamus and the dorsal/anterior hypothalamic areas. J Neuroendocrinol 11:259–266
Fujikawa T, Soya H, Alam KSM, Fukuoka H, Yoshizato H, Nakashima K. 2000 Prolactin induces CRF mRNA and blocks stress-induced hypocalcemia and gastric injuries. Gordon Conference Res Abs
Weber RFA, Calogero AE 1983 Prolactin stimulates rat hypothalamic corticotropin-releasing hormone and pituitary adrenocorticotropin secretion in vitro. Neuroendocrinology 36:165–186
Gray TS, Piechowski RA, Yracheta JM, Rittenhouse PA, Bethea CL, Van de Kar LD 1993 Ibotenic acid lesions in the bed nucleus of the stria terminals attenuate conditioned stress-induced increases in prolactin, ACTH and corticosterone. Neuroendocrinology 57:517–524
Kooy A, de Greef WJ, Vreeburg JTM, Hackeng WHL, Ooms MP, Lamberts SWJ, Weber RFA 1990 Evidence for the involvement of corticotropin-releasing factor in the inhibition of gonadotropin release induced by hyperprolactinemia. Neuroendocrinology 51:261–266
Shibasaki T, Yamauchi N, Hotta M, Masuda A, Oono H, Wakabayashi I, Ling N, Demura H 1990 Brain corticotropin-releasing factor acts as inhibitor of stress-induced gastric erosion in rats. Life Sci 47:925–932
Filaretova LP, Filaretov AA, Makara GB 1998 Corticosterone increase inhibits stress-induced gastric erosions in rats. Am J Physiol 274:G1024–G1030
Filaretova L, Maltcev N, Bogdanov A, Levkovich Y 1999 Role of gastric microcirculation in the gastroprotection by glucocorticoids released during water-restraint stress in rats. Chin J Physiol 42:145–152(Takahiko Fujikawa, Kenji )