https://orcid.org/0000-0001-9526-0638
Abstract
Energetic status often affects reproductive function, glucose homeostasis, and feeding in mammals. Malnutrition suppresses pulsatile release of the gonadotropin-releasing hormone (GnRH)/luteinizing hormone (LH) and increases gluconeogenesis and feeding. The present study aims to examine whether β-endorphin-μ-opioid receptor (MOR) signaling mediates the suppression of pulsatile GnRH/LH release and an increase in gluconeogenesis/feeding induced by malnutrition. Ovariectomized female rats treated with a negative feedback level of estradiol-17β (OVX + low E2) receiving 2-deoxy-D-glucose (2DG), an inhibitor of glucose utilization, intravenously (iv) were used as a malnutrition model. An administration of D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP), a selective MOR antagonist, into the third ventricle blocked the suppression of the LH pulse and increase in gluconeogenesis/feeding induced by iv 2DG administration. Histological analysis revealed that arcuate Kiss1 (kisspeptin gene)-expressing cells and preoptic Gnrh1 (GnRH gene)-expressing cells co-expressed little Oprm1 (MOR gene), while around 10% of arcuate Slc17a6 (glutamatergic marker gene)-expressing cells co-expressed Oprm1. Further, the CTOP treatment decreased the number of fos-positive cells in the paraventricular nucleus (PVN) in OVX + low E2 rats treated with iv 2DG but failed to affect the number of arcuate fos-expressing Slc17a6-positive cells. Taken together, these results suggest that the central β-endorphin-MOR signaling mediates the suppression of pulsatile LH release and that the β-endorphin may indirectly suppress the arcuate kisspeptin neurons, a master regulator for GnRH/LH pulses during malnutrition. Furthermore, the current study suggests that central β-endorphin-MOR signaling is also involved in gluconeogenesis and an increase in food intake by directly or indirectly acting on the PVN neurons during malnutrition in female rats.
Energy availability is an essential cue for regulation of reproductive function in mammals. Malnutrition often suppresses reproductive function in various mammalian species, such as cows (1), sheep (2), goats (3), rats (4), mice (5), and monkeys (6). In fact, fasting for 24 to 48 h reduces pregnancy and fertility rates (7), and long-term dietary restrictions delay the reproductive cycle in female rats (8). In addition, 48-h fasting suppresses luteinizing hormone (LH) pulses in female rats.
Previous studies reported that fasting increased levels of β-endorphin, an endogenous opioid peptide produced by processing from the precursor, proopiomelanocortin, in the hypothalamus of male rats (9) and subcutaneous administration of β-endorphin increased feeding in male rats (10). These results suggest that β-endorphin and its receptor, the μ-opioid receptor (MOR), are involved in malnutrition-induced physiological changes, such as feeding. A previous study showed that an intracerebroventricular (icv) administration of naloxone, an opioid receptor antagonist, blocked 48-h fasting-induced suppression of LH pulses in female rats (11). This suggests that β-endorphin-MOR signaling mediates LH pulse suppression during malnutrition, because naloxone shows the highest affinity to MOR (12). On the other hand, naloxone also binds to other opioid receptors, such as the κ-opioid receptor (KOR) and the δ-opioid receptor (DOR), whose major endogenous ligands are dynorphin (Dyn) and enkephalin, respectively. Thus, recovery of LH pulses in the fasted female rats by a central administration of naloxone may be due to the antagonism of KOR and/or DOR. Indeed, our recent study showed that hypothalamic Dyn-KOR signaling mediates LH pulse suppression induced by peripheral or central glucoprivation (13). Therefore, it should be clarified whether β-endorphin-MOR signaling is specifically involved in the suppression of LH pulses during malnutrition.
Acute glucoprivation induced by administration of 2-deoxy-D-glucose (2DG), an inhibitor of glucose utilization, is a useful model to investigate malnutrition-induced changes in LH secretion, feeding, and glucose metabolism. Female rats peripherally or centrally administered with 2DG showed suppression of pulsatile LH release and increases in feeding and blood glucose levels (14-16). It has been suggested that kisspeptin neurons located in the arcuate nucleus (ARC) of the hypothalamus are the master regulator for pulsatile release of gonadotropin-releasing hormone (GnRH)/LH pulses in mammals (17-21). Our recent study suggested that Dyn neurons originating from the hypothalamic paraventricular nucleus (PVN) specifically mediate the glucoprivic suppression of LH pulses by directly acting on KOR expressed in the ARC kisspeptin neurons but do not mediate the glucoprivic induction of food intake and gluconeogenesis in female rats (13). The study showed that icv administration of nor-binaltorphimine, a KOR antagonist, blocked the glucoprivic LH pulse suppression but failed to block an increase in food intake and blood glucose levels. Further, the study indicated that dynorphin receptors are expressed in a majority of the ARC kisspeptin neurons, and that the glucoprivation specifically induced c-fos expression in the PVN Dyn neurons. On the other hand, it is not yet known whether MOR is involved in the glucoprivic suppression of LH pulses and, if so, whether β-endorphin directly or indirectly, via interneurons such as glutamatergic neurons, suppresses the ARC kisspeptin neurons. In addition, it is possible that, unlike KOR signaling, β-endorphin-MOR signaling may mediate malnutrition-induction of food intake and/or gluconeogenesis.
The present study thus aimed to investigate whether β-endorphin-MOR signaling mediates the suppression of pulsatile LH release and increases in feeding and gluconeogenesis during malnutrition using acute glucoprived female model rats. Further, the present study explored the action site(s) of β-endorphin by histological investigation of MOR-expressing cells in the rat hypothalamus. To address the issues, we examined whether third ventricle (3V) administration of D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP), a selective MOR antagonist, blocks the suppression of pulsatile LH release induced by intravenous (iv) administration of 2DG in ovariectomized female rats treated with a negative feedback level of estradiol-17β (OVX + low E2). This was done to determine the involvement of central MOR signaling in the suppression of LH pulse and the induction of gluconeogenesis and food intake by 2DG administration. In addition, we performed double in situ hybridization (ISH) for the MOR gene (Oprm1) and the kisspeptin gene (Kiss1), the GnRH gene (Gnrh1), or the glutamatergic marker gene (Slc17a6) in the hypothalamus of female rats to investigate the action site(s) of β-endorphin. Furthermore, we also investigated the effect of 3V CTOP treatment on fos messenger RNA (mRNA) expression in the Slc17a6-expressing cells in the ARC of iv 2DG-treated OVX + low E2 rats to determine if the ARC glutamatergic neurons are activated by MOR antagonism. In addition, we examined the effect of CTOP administration on fos expression in the PVN of the iv 2DG-treated OVX + low E2 rats to explore whether central MOR antagonism diminishes an activation of PVN neurons induced by malnutrition.
Materials and Methods
Animals
Adult Wistar-Imamichi strain female rats (8-13 weeks old, 200-250 g; Institute for Animal Reproduction, Kasumigaura, Japan) were housed in a controlled environment (14-h light, 10-h darkness with lights on at 5 am, 22 ± 3°C) and had free access to food (CE-2; CLEA Japan, Tokyo, Japan) and water until the day of brain or blood sampling. Female rats having shown at least 2 consecutive estrous cycles were bilaterally ovariectomized. The rats were implanted with subcutaneous Silastic tubing (1.57 mm inner diameter; 3.18 mm outer diameter; 25 mm in length; Dow Corning, Midland, MI, USA) containing E2 (Sigma-Aldrich, St. Louis, MO, USA) dissolved in peanut oil (Sigma-Aldrich) at 20 µg/mL to serve as OVX + low E2 rats 7 days before the blood or brain sampling. The E2 treatment was confirmed to show negative feedback to LH pulses but not to induce LH surge (22). These surgical procedures, if not otherwise specified, were performed under ketamine (27 mg/kg)/xylazine (5.3 mg/kg) mixture and inhalant 1% to 2% isoflurane (Pfizer Japan, Tokyo, Japan) anesthesia. The present study was approved by the Committee on Animal Experiments of the Graduate School of Bioagricultural Sciences, Nagoya University.
Blood Sampling and Food Intake Test to Determine the Effect of Central MOR Antagonism on Peripheral Glucoprivation
OVX + low E2 rats (n = 3-5 in each group) were stereotaxically implanted with a stainless steel guide cannula (22G; Plastics ONE, Roanoke, VA, USA) into the 3V according to the rat brain atlas (23) as follows: at 0.8 mm posterior and 7.5 mm ventral to the bregma at midline 7 days before blood sampling (on the same day of the OVX + low E2 treatment). Then, the OVX + low E2 rats were inserted with a silicon cannula (inner diameter 0.5 mm; outer diameter 1.0mm; Shin-Etsu Polymer, Tokyo, Japan) into the right atrium through the jugular vein on the day before blood sampling. The animals were deprived of their food for 3 h during the blood sampling period to ensure that plasma LH and glucose concentrations would not be affected by food intake. The free-moving conscious rats were administrated with CTOP (Sigma-Aldrich), dissolved in ultrapure water at 5 nmol/µL, into the 3V at the flow rate of 1 µL/min for 2 min by a microsyringe pump (EICOM, Kyoto, Japan) through an internal cannula (28G; Plastics ONE) immediately after the first blood sampling collected via the jugular vein cannula. The dose of CTOP was chosen according to a previous study (24), showing that the administration of the same dose of CTOP into the lateral ventricle (LV) blocked the suppression of LH release induced by 5-thioglucose (5TG) administration into the fourth ventricle (4V). The CTOP was dissolved in ultra-pure water (UPW) and stored -20°C, and UPW was used in vehicle-treated control rats. Immediately after the CTOP administration, 2DG (400 mg/kg BW, Sigma-Aldrich) dissolved at 200 mg/mL in saline or equimolar hypertonic concentration of xylose (366 mg/kg BW, Katayama Chemical Industries, Osaka, Japan), an indigestible sugar for rats, dissolved at 183 mg/mL in saline were intravenously injected and 100 µL of blood samples were collected every 6 min for 3 h. The dose of 2DG and xylose were chosen according to a previous study (14), showing that LH pulses were significantly suppressed by iv 2DG (400 mg/kg BW) treatment in OVX + low E2 rats. Red blood cells taken from donor rats and washed with saline were replaced at each blood collection to keep the hematocrit level constant. Plasma samples (50 µL) were obtained by immediate centrifugation and stored at −20°C until assayed for LH. Plasma glucose concentrations were measured in an additional volume (50 µL) of blood samples obtained every 12 min during the first hour and every 30 min during the last 2 h of the blood sampling period, as previously described (25). Immediately after the end of the blood sampling, animals were allowed to access food for 30 min, and then the amount of food intake was measured to determine the effect of 2DG administration on feeding. Note that the xylose (iv)-CTOP (3V)-treated animals were healthy and their behavior was normal as the xylose (iv)-vehicle (3V)-treated control animals during the 3-h blood sampling period after the CTOP administration and subsequent 30 min observation period for feeding behavior.
Assay for Plasma LH and Glucose Concentrations
Plasma LH concentrations were measured by a double-antibody radioimmunoassay (RIA) as previously described (26) using a rat LH RIA kit provided by the National Hormone and Peptide Program (Harbor-UCLA Medical Center, Torrance, CA, USA) and were expressed in terms of National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) rat LH-RP-3. The least detectable level in 50-µL plasma samples was 0.078 ng/mL, and the intra- and interassay coefficients of variation were 4.0 % and 8.8 % at 1.098 ng/mL, respectively. Plasma glucose concentrations were measured by the glucose oxidase method using a commercial kit (Glucose C-test; Wako Pure Chemical Corporation, Osaka, Japan), as previously described (13). The least detectable level in 1.5-µL plasma samples was 0.25 mg/mL, and the intra- and interassay coefficients of variation were 5.8% and 11.2% at 0.79 mg/mL, respectively.
Analysis for LH Pulse
LH pulses and the baseline LH concentration were identified by the PULSAR computer program (27) as previously described (28). The criteria used to identify LH pulses were chosen so that a single LH concentration had to be 3 SD greater than the baseline LH concentration, 2 consecutive LH concentrations each had to be 1 SD greater, and 3 consecutive LH concentrations each had to be 0.4 SD greater to be considered part of a LH pulse. The SD for each LH concentration was determined by assaying in a single assay 10 duplicates each of 4 plasma pools that had LH concentrations distributed across the whole concentration range of the assay. The equation for the linear regression line of the SDs for the pools was then calculated. For 50 μL plasma, this equation was y = (6.98x + 2.87)/100, where x is the LH concentration and y is the SD of the LH concentration. The mean LH was determined by averaging all of the LH concentration for each animal. The frequency was determined by the number of peaks for the sampling period. The amplitude was determined by the difference between the peak LH concentration and baseline LH level at the peak. The mean LH concentrations and the frequency and amplitude of LH pulses for the 3-h blood sampling period were calculated for each individual and then for the group.
Brain Sampling for Histological Analysis
The other cohort of OVX + low E2 rats (n = 3) were deeply anesthetized with sodium pentobarbital (40 mg/kg; Kyoritsu, Tokyo, Japan) and perfused with 4% paraformaldehyde (Sigma-Aldrich) in 0.05M phosphate buffer. The brain was immediately removed from the skull and postfixed with the same fixative at 4oC overnight then immersed in 30% sucrose in 0.05M phosphate buffer at 4oC. Frozen sections (50-µm thickness) of the hypothalamus including the ARC and preoptic area (POA) were obtained with a cryostat. The brain sections taken from OVX + low E2 rats were subjected to a double ISH for Kiss1/Oprm1, Slc17a6/Oprm1, and Gnrh1/Oprm1.
To determine the effects of the 3V injection of CTOP on fos expression in the ARC glutamatergic neurons in iv 2DG-treated rats, some OVX + low E2 rats (n = 6) were implanted with a stainless steel guide cannula into the 3V and then inserted with iv cannula into the right atrium as described in the previous discussion. Seven days after the brain surgery, ovariectomy, and E2 implantation, free-moving conscious OVX + low E2 rats were administered with CTOP (5 nmol/µL, 1µL/min for 2 min) or vehicle into the 3V and then acutely injected with iv 2DG (400 mg/kg BW) through iv cannula attached on the previous day (2DG/CTOP, n = 3; 2DG/vehicle, n = 3). Animals were deprived of their food after the iv 2DG administration until the brain sampling to avoid the effect of feeding on the expression of the previously mentioned genes. One hour after the iv 2DG administration, the brain was fixed in the same manner as previously mentioned, and frozen sections of the hypothalamus including the ARC and PVN were subjected to a double ISH for Slc17a6/fos.
Double ISH for Kiss1/Oprm1, Slc17a6/Oprm1, Gnrh1/Oprm1, or Slc17a6/fos in rat hypothalamus
Free-floating double ISH for Kiss1/Oprm1, Slc17a6/Oprm1, Gnrh1/Oprm1, or Slc17a6/fos was performed using every fourth section (= every 200 µm) of the hypothalamus including the ARC (for Kiss1/Oprm1, Slc17a6/Oprm1, and Slc17a6/fos) and the PVN (Slc17a6/fos) and every second section (= every 100µm) of the hypothalamus including the POA (for Gnrh1/Oprm1) as previously described (29). The Oprm1-specific digoxigenin (DIG)-labeled complementary RNA (cRNA) probe (position 33-208; GenBank accession number NM_013071.2), and fos-specific DIG-labeled cRNA probe (position 573-1193; GenBank accession number NM_022197.2) was designed and synthesized from the rat whole hypothalamic cDNA by using a DIG-labeling kit (Roche Diagnostics, Basel, Switzerland) as described before (13). The Kiss1-specific fluorescein isothiocyanate (FITC)-labeled cRNA probe (position 33-348; GenBank accession number NM_181692.1), Gnrh1-specific FITC-labeled cRNA probe (position 57-360; GenBank accession number NM_012767.2), and Slc17a6-specific FITC-labeled cRNA probe (position 935-2053; GenBank accession number NM_053427.1) were designed and synthesized by using a FITC-labeling kit (Roche Diagnostics). The brain sections were hybridized with 1 µg/mL antisense cRNA probes at 60oC overnight. Hybridized sections were incubated with a peroxidase (POD)-conjugated anti FITC antibody (Roche Diagnostics) (30) and Tyramide Signal Amplification (TSA) Plus FITC Kit (1:100; Perkin Elmer, Waltham, MA) to detect the hybridized Kiss1, Gnrh1 and Slc17a6 probes. After the inactivation of the POD by incubating the sections in 0.1 N hydrochloric acid for 30 min, the hybridized Oprm1 and fos probes was detected by using the POD-conjugated anti-DIG antibody (31), TSA Plus Biotin Kit (1:100) and Dylight 594-conjugated streptavidin. No positive signals for Slc17a6, Gnrh1, and Oprm1 mRNA were detected in the brain sections hybridized with the corresponding sense probes as negative controls (data not shown), and we previously confirmed that no Kiss1 or fos signal was detected in the brain sections hybridized with the corresponding sense probe (13,32).
Statistical Analysis
Statistical differences in the mean LH concentrations, the frequency and amplitude of LH pulses, the area under the curve (AUC) of plasma glucose level and amount of food intake between the 2DG (iv)- and xylose (iv)-treated groups administered with CTOP (3V) or vehicle (3V) were determined by 2-way (2DG and CTOP treatments as main effects) analysis of variance (ANOVA) followed by analysis of simple main effects. Statistical differences in the number of the Slc17a6-expressing cells, fos-expressing cells, and Slc17a6-expressing cells co-expressing fos in the ARC and PVN between 3V CTOP- and vehicle-treated iv 2DG-injected rats were determined by the Student’s t-test. Statistical differences in plasma glucose level between the groups were determined by 3-way repeated measures (2DG, CTOP treatments, and time as main effects) ANOVA followed by analysis of simple main effects with Bonferroni correction. All statistical analysis used SAS University Edition (https://www.sas.com/).
Results
Central MOR Antagonism Blocked the Suppression of LH Secretion Induced by iv 2DG Injection
Figure 1A shows LH profiles in representative OVX + low E2 rats receiving an iv injection of 2DG or xylose and 3V injection of CTOP, an MOR specific antagonist, or vehicle. Pulsatile LH release was apparent in the xylose (iv)-vehicle (3V)- or xylose (iv)-CTOP (3V)-injected control rats, while LH pulses in the 2DG (iv)-vehicle (3V)-injected group were profoundly suppressed in comparison with the control group. The 3V CTOP injection blocked the 2DG-induced suppression of pulsatile LH release and restored apparent LH pulses in 2DG (iv)-injected group. Two-way ANOVA revealed that there were significant interactions between main effects (CTOP and 2DG treatments) on the mean LH concentration [F (1, 12) = 6.86, P = 0.0225] and the frequency of LH pulses [F (1, 12) = 14.26, P < 0.01]. Specifically, the mean LH and frequency in the 2DG (iv)-vehicle (3V) group was significantly lower than those in the xylose (iv)-vehicle (3V) group [mean LH, P = 0.0423 (Fig. 1B); frequency, P < 0.01, (Fig. 1C)], whereas the mean LH and frequency in the 2DG (iv)-CTOP (3V) group was significantly higher than those in the 2DG (iv)-vehicle (3V) group [mean LH, P = 0.0123 (Fig. 1B); frequency, P < 0.01 (Fig. 1C)], and the mean LH levels and frequency in the 2DG (iv)-CTOP (3V) group were comparable to the xylose (iv)-vehicle (3V)-treated control group. There were no significant main effects and interaction on the amplitude of LH pulses between any groups (Fig. 1D).
Effects of central administration of D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP), a selective antagonist of μ-opioid receptors, on the suppression of luteinizing hormone (LH) pulses induced by intravenous (iv) injection of 2-deoxy-D-glucose (2DG) in ovariectomized female rats treated with a negative feedback level of estradiol-17β (OVX + low E2). Plasma LH profiles in representative OVX + low E2 rats, which were treated with the third ventricle (3V) injection of CTOP (or vehicle) and iv injection of 2DG (or xylose) (A). Βlood samples were collected every 6 min for 3 h. Immediately after the first blood sampling, CTOP or vehicle (timing indicated by arrows) was injected in to the 3V and then 2DG (or xylose) was injected intravenously. Arrowheads indicate the peaks of LH pulses identified by the PULSAR computer program. The mean plasma LH concentrations (B), the frequency (C), and amplitude (D) of LH pulses in each group. *Significant difference in the mean LH concentration and LH pulse frequency between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated groups (P < 0.05, the simple main effect of the 2-way analysis of variance). †Significant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups (P < 0.05, the simple main effect of the 2-way analysis of variance). Values are means ± standard error of the mean. The numbers in each column indicate the number of animals used.
Exploration of Oprm1 (MOR Gene) mRNA Expression in the Kiss1-, Slc17a6-, and Gnrh1-expressing Neurons in the Female Rat Hypothalamus
Oprm1 mRNA expression was found in a few Kiss1-expressing cells, some Slc17a6-expressing cells, and some Kiss1/Slc17a6-negative cells in the ARC of OVX + low E2 rats (Fig. 2A and B). Quantitative analysis revealed that 0.39% of Kiss1- and 13.5% of Slc17a6-expressing cells showed Oprm1 expression in the ARC of OVX + low E2 rats. Oprm1 mRNA expression was not found in the Gnrh1-expressing cells but was found in some Gnrh1-negative cells in the POA of OVX + low E2 rats (Fig. 2C).
Expression of Kiss1 (kisspeptin gene)/Oprm1 (μ-opioid receptor gene), and Slc17a6 (glutamatergic marker gene)/Oprm1 in the arcuate nucleus (ARC), and Gnrh1 (GnRH gene)/Oprm1 in the preoptic area (POA) in the ARC of ovariectomized female rats treated with a negative feedback level of estradiol-17β (OVX + low E2) determined by double in situ hybridization. Kiss1-expressing (green) and Oprm1-expressing cells (magenta) in (A) and Slc17a6-expressing (green) and Oprm1-expressing cells (magenta) in the ARC of representative OVX + low E2 rats (B). Gnrh1-expressing (green) and Oprm1-expressing cells (magenta) in the POA of a representative OVX + low E2 rat (C). The insets indicate the cells pointed by the closed white (merge cells) and opened white (nonmerge cells) arrowheads at higher magnification. Scale bars, 100 µm. The columns indicate the number of Kiss1-, Slc17a6-, or Gnrh1-expressing cells with (green and magenta column) or without (green column) Oprm1 expression in the ARC (for Kiss1 and Slc17a6) or POA (for Gnrh1) of the OVX + low E2 rats. Values are means ± standard error of the mean. The number in each column indicates the number of animals used.
Effect of Central CTOP Administration on fos mRNA Expression in the ARC Slc17a6-expressing Cells and on fos mRNA Expression in the PVN of iv 2DG-treated Female Rats
Figure 3 shows the expression of Slc17a6 and fos mRNA in the ARC and PVN in representative iv 2DG-injected OVX + low E2 rats treated with CTOP or vehicle in the 3V. In the ARC, fos signals were found in a part of Slc17a6-positive cells. Statistical analysis showed that there was no significant difference in the number of fos-expressing Slc17a6-positive cells and the total number of Slc17a6-expressing cells and fos-expressing cells in the ARC between 2DG (iv)-CTOP (3V)-injected and 2DG (iv)-vehicle (3V)-injected groups (Fig. 3A). The percentage of fos-co-expressing cells out of the Slc17a6-expressing cells in the ARC was 14% ± 5.0% in 2DG (iv)-CTOP (3V)-injected rats and 12% ± 5.6% in 2DG (iv)-vehicle (3V)-injected rats. In the PVN, fos signals were found in a part of Slc17a6-positive cells. In 2DG (iv)-vehicle (3V)-injected rats, fos-expressing cells were found in both medial and lateral regions throughout the PVN. On the other hand, in 2DG (iv)-CTOP (3V)-injected rats, fos-expressing cells were mainly found in the lateral region, and few fos-expressing cells were found in the medial region of the PVN. Statistical analysis revealed that the total number of fos-expressing cells of 2DG (iv)-CTOP (3V)-injected rats was significantly lower compared with that in 2DG (iv)-vehicle (3V)-injected controls [P = 0.0188; Student’s t-test (Fig. 3B)]. There was no significant difference in the number of fos-expressing Slc17a6-positive cells and the total number of Slc17a6-expressing cells in the PVN between 2DG (iv)-CTOP (3V)-injected and 2DG (iv)-vehicle (3V)-injected groups (Fig. 3B). The percentage of fos-co-expressing cells out of the Slc17a6-expressing cells in the PVN was 15% ± 4.2% in 2DG (iv)-CTOP (3V)-injected rats and 13% ± 1.1% in 2DG (iv)-vehicle (3V)-injected rats.
Effects of central D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP) administration on expression of Slc17a6 (glutamatergic marker gene)/fos in the arcuate nucleus (ARC) and paraventricular nucleus (PVN) in intravenous (iv) 2-deoxy-D-glucose (2DG)-treated female rats determined by double in situ hybridization. Slc17a6-expressing (green) and fos-expressing cells (magenta) in the ARC (A) and PVN (B) of representative ovariectomized female rats treated with a negative feedback level of estradiol-17β at 1 h after the intravenous 2DG and third ventricle (3V) CTOP or vehicle injection. The insets indicate fos expression in Slc17a6-expressing cells indicated by white arrowheads at higher magnification. Scale bars, 100 µm. The columns indicate the number of Slc17a6--expressing cells with (green and magenta column) or without (green column) fos expression and the number of fos-expressing cells with (green and magenta column) or without (magenta column) Slc17a6-expression in the ARC (A) and PVN (B). Values are means ± standard error of the mean. The number in each column indicates the number of animals used. *Significant difference in the total number of fos-expressing cells between 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated rats (P < 0.05, Student’s t-test).
Central MOR Antagonism Blocked the Glucoprivic Increase in Plasma Glucose Concentrations
Figure 4A shows changes in plasma glucose concentrations for the 3-h blood sampling period in OVX + low E2 rats receiving an iv injection of 2DG or xylose and 3V injection of CTOP or vehicle. Plasma glucose concentrations acutely increased and maintained at high levels in the 2DG (iv)-vehicle (3V)-treated group, while plasma glucose concentrations in the 2DG (iv)-CTOP (3V) group transiently increased by iv 2DG administration and then decreased to similar levels in the xylose (iv)-vehicle (3V) and xylose (iv)-CTOP (3V) groups. Three-way ANOVA analysis revealed that there are significant main effects [2DG, F (1, 12) = 73.89, P < 0.01; CTOP, F (1, 12) = 26.95, P < 0.01; time, F (9, 108) = 3.65, P < 0.01], the secondary interaction [2DG × CTOP × time, F (9, 108) = 6.11, P < 0.01], and the simple interaction [2DG × CTOP, F (1, 12) = 9.5-59.89, P < 0.02 at 24 min and 48-180 min]. Specifically, at 24 min after the iv 2DG administration, the glucose level in the 2DG (iv)-vehicle (3V) group was significantly higher than those in the xylose (iv)-vehicle (3V) control and the 2DG (iv)-CTOP (3V) groups (P < 0.01) (Fig. 4A), and the glucose level in the 2DG (iv)-CTOP (3V) group was significantly higher than that in the xylose (iv)-CTOP (3V) control group (P < 0.01) (Fig. 4A). From 48 to 180 min after the iv 2DG administration, the glucose levels in the 2DG (iv)-vehicle (3V) group were significantly higher (P < 0.01) than those in the xylose (iv)-vehicle (3V) and 2DG (iv)-CTOP (3V) groups, whereas the glucose levels in the 2DG (iv)-CTOP (3V) groups were comparable to that in the xylose (iv)-CTOP (3V)-treated group.
![Effects of third ventricle (3V) D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP) administration on plasma glucose levels after the intravenous treatment of 2-deoxy-D-glucose (2DG) in female rats. Changes in the mean plasma glucose concentrations in female rats treated with a negative feedback level of estradiol-17β receiving the 3V CTOP (or vehicle) and intravenous (iv) 2DG (or xylose) injection (A). *Significant difference between 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated rats and between 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups [P < 0.05, the simple-simple main effect of the 3-way analysis of variance (ANOVA)]. †Significant difference between the 2DG (iv)-CTOP (3V)- and xylose (iv)-CTOP (3V)-treated groups (P < 0.05, the simple-simple main effect of the 3-way ANOVA). § Significant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated rats and between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated rats (P < 0.05, the simple-simple main effect of the 3-way ANOVA). The area under the curve of plasma glucose concentrations for 3 h after iv injection of 2DG or xylose in the rats treated with 3V CTOP or vehicle (B). *Significant difference between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated groups (P < 0.05, the simple main effect of the 2-way ANOVA). †Significant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups (P < 0.05, the simple main effect of the 2-way ANOVA). Values are means ± standard error of the mean. The numbers in each column indicate the number of animals used.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/endo/162/10/10.1210_endocr_bqab140/2/m_bqab140_fig4.jpeg?Expires=1748401865&Signature=NPN9xjb0FgrxNqZ4lAS0HzrDMvzVeZqGkltu-~rdx4Ao9xfxRf~~zbZOfWu1XDy72y2Gq4bppwcFN9ntcRquzKExgX2LcBN4axmT8-YV5wGMAulGt3vxsvKnMzDp~h0N-mSYxAbx60~KMP-9Q1XwdSZkRI5q99-MQYAiSDS0-hvxKpH31SNXE8W50rNkLCYWH6wWjGneS4QUa2ir4v7Kz3jvtfwkkpE6fm70aLm~cw21Aodn~Its3MCk5OUK92CltEahOFIx4bW-ifFW-5Ei~IjhWP313udKvsnKfmxspTVTTg19SxCnvj-0ArxCsFsvJxGmitW3yka5fqXojFYeAA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Effects of third ventricle (3V) D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP) administration on plasma glucose levels after the intravenous treatment of 2-deoxy-D-glucose (2DG) in female rats. Changes in the mean plasma glucose concentrations in female rats treated with a negative feedback level of estradiol-17β receiving the 3V CTOP (or vehicle) and intravenous (iv) 2DG (or xylose) injection (A). *Significant difference between 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated rats and between 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups [P < 0.05, the simple-simple main effect of the 3-way analysis of variance (ANOVA)]. †Significant difference between the 2DG (iv)-CTOP (3V)- and xylose (iv)-CTOP (3V)-treated groups (P < 0.05, the simple-simple main effect of the 3-way ANOVA). § Significant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated rats and between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated rats (P < 0.05, the simple-simple main effect of the 3-way ANOVA). The area under the curve of plasma glucose concentrations for 3 h after iv injection of 2DG or xylose in the rats treated with 3V CTOP or vehicle (B). *Significant difference between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated groups (P < 0.05, the simple main effect of the 2-way ANOVA). †Significant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups (P < 0.05, the simple main effect of the 2-way ANOVA). Values are means ± standard error of the mean. The numbers in each column indicate the number of animals used.
Two-way ANOVA analysis revealed that there was a significant interaction between the main effects [2DG and CTOP, F (1, 12) = 18.50, P < 0.01] on the AUC of the plasma glucose concentrations. Specifically, the AUC in the 2DG (iv)-vehicle (3V) group was significantly higher than those in the xylose (iv) vehicle (3V) (P < 0.01) and the 2DG (iv)-CTOP (3V) groups (P < 0.01) (Fig. 4B).
Central MOR Antagonism Blocked the Glucoprivic Increase in Food Intake
The iv administration of 2DG increased food intake, while the 3V CTOP treatment blocked the increase in OVX + low E2 rats. Two-way ANOVA analysis revealed that there was a significant interaction between the main effects on the amount of food intake for 30 min after the 3-h blood sampling [F (1, 12) = 13.08, P < 0.01]. Specifically, the amount of food intake in the 2DG (iv)-vehicle (3V) group was significantly higher than those in the xylose (iv)-vehicle (3V) (P < 0.01) and 2DG (iv)-CTOP (3V) groups (P < 0.01) (Fig. 5).
Effects of the third ventricle (3V) D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2 (CTOP) administration on food intake after the iv treatment of 2-deoxy-D-glucose (2DG) in female rats. Amount of food intake in ovariectomized female rats treated with a negative feedback level of estradiol-17β receiving the 3V CTOP (or vehicle) and iv 2DG (or xylose) treatment. Food intake of each group was measured for 30 min just after the end of the 3-h blood sampling period for plasma LH and glucose analysis. *Significant difference between the 2DG (iv)-vehicle (3V)- and xylose (iv)-vehicle (3V)-treated groups (P < 0.05, the simple main effect of the 2-way analysis of variance). †ignificant difference between the 2DG (iv)-vehicle (3V)- and 2DG (iv)-CTOP (3V)-treated groups (P < 0.05, the simple main effect of the 2-way analysis of variance). Values are means ± standard error of the mean. The numbers in each column indicate the number of animals used.
Discussion
The present study demonstrated that central β-endorphin-MOR signaling, at least partly, mediates LH pulse suppression during malnutrition, because an administration of CTOP, a selective MOR antagonist, into the 3V blocked the suppression of pulsatile LH release induced by acute peripheral glucoprivation in female rats. It is unlikely that the ARC kisspeptin and POA GnRH neurons are action sites of β-endorphin because the current study showed that little MOR gene expression was found in the kisspeptin and GnRH neurons. Taken together, these findings suggest the possibility that the β-endorphin-MOR signaling activated by glucoprivation could be transmitted to the ARC kisspeptin neurons via some interneurons and consequently suppress GnRH/LH pulses. It has been reported that MOR is widely distributed in several hypothalamic nuclei, such as the ARC, PVN, and POA as well as parabrachial nucleus in the mid brain in rats (33). This indicates that the current administration of CTOP to the 3V may have acted on the MOR in these hypothalamic and mid brain nuclei and blocked the suppression of reproductive function during acute glucoprivation. The present finding is consistent with a previous report showing that 3V administration of naloxone reversed 48-h fasting-induced suppression of LH pulses in female rats (22). The advantage of the current study is that, CTOP, unlike naloxone, specifically antagonizes MOR, and CTOP reportedly did not interact with DOR or KOR (12). Thus, the present study using CTOP suggests that MOR mediates the suppression of reproductive function during malnutrition. The present results are also consistent with a previous study demonstrating that the LV CTOP pretreatment blocked the suppression of LH release induced by the 4V administration of 5TG, another glucose antimetabolite in female rats (24). The current CTOP administration significantly reduced the number of fos-expressing cells in the PVN of female rats treated with 2DG, suggesting that an inactivation of the PVN neurons may partly contribute to the blockade of glucoprivic suppression of LH pulses. This notion is consistent with the previous studies suggested that β-endorphin is also involved in the stress-induced suppression of reproductive function: Central administration of β-funaltrexamine, a selective MOR antagonist, blocked footshock stress-induced suppression of LH secretion in rats (34). Further, immobilization stress upregulates the expression of corticotropin-releasing hormone (CRH) mRNA of the PVN in female rats (35), and the suppression of LH release induced by icv administration of CRH was blocked by icv administration of β-funaltrexamine in male rats (34).These studies and our current study suggest that β-endorphin-MOR signaling mediates the suppression of LH release by stress as well as malnutrition via activating the PVN CRH neurons. Furthermore, a central administration of CRH antagonist blocked the iv 2DG-induced suppression of LH pulses in female rats (36), peripheral 2DG treatment induced noradrenaline release in the PVN (37), and the 3V injection of CRH antagonist blocked LH pulse suppression induced by the noradrenalin injection into the PVN in female rats (38). Furthermore, our recent study showed that 2DG administration into the 4V decreased plasma testosterone levels and induced c-Fos expression in the PVN CRH-immunopositive cells and hindbrain noradrenergic nuclei in male rats (39). Taken together, the following neuronal cascade could be proposed: malnutrition and/or stress activate noradrenergic neurons projecting to the PVN and then activate the PVN CRH neurons, resulting in GnRH/LH suppression via β-endorphin-MOR signaling.
The present study demonstrates that central β-endorphin-MOR signaling also mediates malnutrition-induced gluconeogenesis because the central administration of CTOP blocked the glucoprivic increase in blood glucose levels. The notion is consistent with a previous study showing that a central administration of CTOP reversed an increase in plasma glucose levels induced by an administration of 5TG into the 4V in female rats (24). In addition, an icv administration of Tyr-D-Ala-Gly-N(Me)Phe-Gly-ol (DAMGO), an MOR agonist, increased plasma glucose levels, while blockade of α2A adrenergic receptors weakened the effect of DAMGO on glucose metabolism in male mice (40). A central administration of β-endorphin increased plasma levels of adrenaline, noradrenaline, and cortisol in dogs (41). Furthermore, icv administration of β-endorphin increases CRH mRNA levels in the PVN in male rats (42), suggesting that central β-endorphin activates CRH neurons in the PVN. Taken together, these findings suggest that β-endorphin positively regulates sympathetic nerve system and then hypothalamic-pituitary-adrenal axis to increase blood glucose levels. Indeed, MOR mRNA expression is evident in the PVN in male rats (33). Further, our recent study demonstrated that hindbrain glucoprivation activated the PVN CRH neurons in male rats (39) and the current MOR antagonism reduced fos-expressing cells in the medial region of the PVN, where CRH cell bodies are largely located, in glucoprived female rats. These findings imply that there is a short loop positive feedback system between the PVN CRH neurons and β-endorphin-MOR signaling. Therefore, the current CTOP administration may, at least in part, have blocked the glucoprivic activation of PVN CRH neurons and then the hypothalamic-pituitary-adrenal axis and resulted in the blockade of an increase in plasma glucose levels induced by iv 2DG treatment.
The present study demonstrates that β-endorphin-MOR signaling also mediates malnutrition-induced feeding because the 3V administration of CTOP reversed an increase in food intake induced by iv administration of 2DG. In this context, the β-endorphin-MOR is quite different from Dyn-KOR signaling, which specifically mediates glucoprivic suppression of LH pulses but not glucoprivic induction of feeding and gluconeogenesis (13). The involvement of β-endorphin-MOR signaling in inducing feeding is consistent with previous reports suggesting that β-endorphin is an orexigenic neurotransmitter: food intake is induced by a subcutaneous injection of β-endorphin in male rats (10); a subcutaneous injection of naloxone decreased the amount of food intake in male mice (43); administration of the antisense oligodeoxynucleotide of the MOR gene to the LV resulted in the suppression of food intake and weight loss in male rats (44); administration of β-funaltrexamine, a selective MOR antagonist, into the lateral hypothalamus suppressed food intake in male mice (45); and the PVN administration of DAMGO, an MOR selective agonist, induced food intake in male rats (46). These and current studies indicate that the central β-endorphin-MOR signaling positively regulates food intake and the PVN is an action site of β-endorphin to control feeding.
The present study suggests that ARC glutamatergic neurons might be an action site of β-endorphin to suppress GnRH/LH release. This is because this study showed that MOR gene expression was found in some ARC glutamatergic neurons, and our previous study showed that most of the ARC kisspeptin neurons express NR1, a subunit of the N-methyl-D-aspartate (NMDA)-type glutamate receptor in female rats (47). Our previous in vitro study showed that glutamate treatment to the hypothalamic tissues consist of the ARC and median eminence of female rats enhanced in vitro GnRH release from the tissue (48). Furthermore, a central administration of glutamate promotes LH release in female rats (49) and the central administration of NMDA, a selective agonist of the NMDA-type glutamate receptor, increased LH release in wild-type female rats but failed to induce LH release in global Kiss1 knockout rats (19). These findings suggest that glutamatergic neurons induce GnRH and then LH release via activating ARC kisspeptin neurons. In addition, icv administration of β-endorphin suppressed LH release, but the suppression was restored by iv administration of NMDA in male mice (50). Nevertheless, it is unclear whether the ARC glutamatergic neurons mediate the β-endorphin-MOR signaling activated by peripheral glucoprivation, because the current CTOP treatment failed to affect the number of fos mRNA-expressing glutamatergic cells in the ARC of female rats treated with iv 2DG. Further studies are required to address this issue.
In conclusion, the present study demonstrated that central β-endorphin-MOR signaling mediates the suppression of pulsatile LH release during malnutrition in female rats and that β-endorphin may indirectly suppress ARC kisspeptin neurons by possibly acting on some interneurons. Furthermore, the current study suggests that β-endorphin-MOR signaling in the hypothalamus is also involved in gluconeogenesis and increases in food intake during malnutrition.
Acknowledgments
We thank the National Hormone and Peptide Program for the rat LH RIA kit. The RIA and LH pulse analysis were performed at the Nagoya University Radioisotope Center and the Information Technology Center, respectively. We thank Dr. Nicola Skoulding for editorial assistance.
Financial Support: This work was supported in part by a Grant-in-Aid for a Research Fellow of the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 21H05031 (to H. Tsukamura), 19H03103 (to N I.) and 20H03127 (to Y.U.).
Additional Information
Disclosure Summary: The authors have nothing to disclose.
Data Availability
All data generated or analyzed during this study are included in this published article or in the data repositories listed in references.
