G protein-coupled lactate receptor GPR81 control of ventrolateral ventromedial hypothalamic nucleus glucoregulatory neurotransmitter and 5′-AMP-activated protein kinase expression

Abstract

Astrocytes store glycogen as energy and promote neurometabolic stability through supply of oxidizable l-lactate. Whether lactate regulates ventromedial hypothalamic nucleus (VMN) glucostatic function as a metabolic volume transmitter is unknown. Current research investigated whether G protein-coupled lactate receptor GPR81 controls astrocyte glycogen metabolism and glucose-regulatory neurotransmission in the ventrolateral VMN (VMNvl), where glucose-regulatory neurons reside. Female rats were pretreated by intra-VMN GPR81 or scramble siRNA infusion before insulin or vehicle injection. VMNvl cell or tissue samples were acquired by laser-catapult- or micropunch microdissection for Western blot protein or uHPLC-electrospray ionization-mass spectrometric glycogen analyses. Data show that GPR81 regulates eu- and/or hypoglycemic patterns of VMNvl astrocyte glycogen metabolic enzyme and 5′-AMP-activated protein kinase (AMPK) protein expression according to VMNvl segment. GPR81 stimulates baseline rostral and caudal VMNvl glycogen accumulation but mediates glycogen breakdown in the former site during hypoglycemia. During euglycemia, GPR81 suppresses the transmitter marker neuronal nitric oxide synthase (nNOS) in rostral and caudal VMNvl nitrergic neurons, but stimulates (rostral VMNvl) or inhibits (caudal VMNvl) GABAergic neuron glutamate decarboxylase_65/67 (GAD)protein. During hypoglycemia, GPR81 regulates AMPK activation in nitrergic and GABAergic neurons located in the rostral, but not caudal VMNvl. VMN GPR81 knockdown amplified hypoglycemic hypercorticosteronemia, but not hyperglucagonemia. Results provide novel evidence that VMNvl astrocyte and glucose-regulatory neurons express GPR81 protein. Data identify neuroanatomical subpopulations of VMNvl astrocytes and glucose-regulatory neurons that exhibit differential reactivity to GPR81 input. Heterogeneous GPR81 effects during eu- versus hypoglycemia infer that energy state may affect cellular sensitivity to or postreceptor processing of lactate transmitter signaling.

INTRODUCTION

Brain astrocytes are a vital contributor to neuronal metabolic stability as these neuroglia take up glucose from the circulation, amass this energy substrate in the form of the complex polysaccharide glycogen, and convert glucose to the oxidizable metabolic fuel l-lactate (1). The astrocyte glycogen shunt, which involves sequential glucose incorporation into and liberation from this complex carbohydrate polymer before entry into the glycolytic pathway, is an active process that accounts for a significant fraction of glucose catabolism in these glial cells (2, 3). The mediobasal hypothalamus (MBH) is a critical component of the brain glucose-regulatory network (4, 5). Lactate acts within the MBH to shape counter-regulation hormone secretion as plasma glucagon and corticosterone profiles are normalized in insulin-induced hypoglycemic (IIH) rats by exogenous lactate administration to the MBH (6). Although lactate is assumed to exert such influence, in part, in its capacity as an energy source, the alternative prospect that lactate may control neural regulation of glucose homeostasis by extracellular signaling mechanisms is intriguing. Lactate is the sole endogenous physiological ligand for the G protein-coupled plasma membrane lactate receptor GPR81, e.g., hydroxycarboxylic acid receptor-1 (HCAR1) (7, 8). GPR81 expression in brain is documented (9, 10), but insight on whether lactate functions as a metabolic volume transmitter within the MBH to control glucose-regulatory function is lacking. Our work shows that lactate governs glucose-regulatory transmitter signaling and counter-regulatory hormone secretion in adult female rats (11, 12). Studies described here used a characterized ovariectomized (OVX) female rat model involving re-establishment of circulating estradiol levels at estrous cycle-like concentrations (13) to investigate the role of GPR81 in these regulatory actions in this sex.

The ventromedial hypothalamic nucleus (VMN), a principal neuroanatomical component of the MBH, is an evident source of metabolic substrate-controlled signaling as IIH regulates phosphorylation, i.e., activation of the ultrasensitive energy sensor 5′-AMP-activated protein kinase (AMPK) in VMN neuron phenotypes that stimulate (nitric oxide; NO) or suppress (γ-aminobutyric acid; GABA) glucose counter-regulatory hormone secretion (14, 15). Neurons in the ventrolateral division of the VMN (VMNvl) exhibit increased or decreased synaptic firing in reaction to neuroglucopenia and are implicated in the regulation of circulating glucose levels (16). The current project used in situ immunocytochemistry, high-resolution single-cell microdissection techniques, and high-sensitivity immunoblotting to investigate the premise that VMNvl nitrergic and/GABAergic neurons express hypoglycemia-responsive GPR81 protein. We reported that nNOS and GAD protein responses to IIH vary between rostral versus caudal segments of the VMNvl (17). Therefore, it was of interest here to examine whether region-specific effects of hypoglycemia on VMNvl NO or GABA nerve cell function involve GPR81-dependent mechanisms. Here, pure VMNvl neurotransmitter cell samples acquired by laser-catapult-microdissection from rostral versus caudal levels of the VMNvl were analyzed by Western blot to examine GPR81 knockdown effects on eu- and hypoglycemic patterns of neurotransmitter marker [nitric oxide synthase (nNOS); glutamate decarboxylase_65/67 (GAD)] and AMPK protein expression in these neuron populations.

Glycogen metabolism shapes VMN glucose-regulatory neurotransmission (12). Glycogen synthesis and glycogenolysis are, respectively, controlled by glycogen synthase (GS) or glycogen phosphorylase (GP) enzyme action. GP variant proteins expressed in brain, i.e., GP-muscle (GPmm) type versus GP-brain (GPbb) type are characterized by dissimilar cellular localization and regulation (18). Both GP isoforms are expressed in astrocytes, yet GPbb is also detectable at comparatively lower levels in neurons. Phosphorylation causes complete (GPmm) or partial (GPbb) activation of these variants, whereas AMP more potently activates GPbb than GPmm and is required for optimal GPbb Km and function. GPmm and GPbb correspondingly mediate noradrenergic or glucoprivic induction of astrocyte glycogen glycogenolysis (19), which infers that these isoforms confer physiological stimulus-specific regulation of glycogen mobilization in those cells. Recent studies document distinctive effects of GPmm versus GPbb on VMN nitrergic and GABAergic nerve cell transmitter marker and AMPK protein expression (20). Cortical astrocytes express GPR81 mRNA and exhibit augmented glycolysis and intracellular lactate accumulation following GPR81 agonist treatment (21). Present studies sought to determine if VMNvl astrocytes express GPR81 and, if so, whether this receptor controls astrocyte glycogen metabolism in a VMNvl segment-specific manner.

MATERIALS AND METHODS

Animals

Adult female Sprague-Dawley rats (230–280 g body wt) were housed in groups of 2–3 animals per cage, under a 14-h light/10-h dark cycle (lights on at 0500). Animals had unrestricted access to standard laboratory chow (Harlan Teklad LM-485; Harlan Industries, Madison, WI) and water and were acclimated to daily handling. All surgical and experimental protocols were carried out in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals (8th ed.), under approval by the ULM Institutional Animal Care and Use Committee.

Experimental Design

Animals were randomly assigned to four treatment groups (n = 4/group). A priori R-software power analysis showed a large size effect and significance level of 0.05, a sample size of n = 4 was predicted to yield power of 0.80. On study day 1, animals were anesthetized with ketamine-xylazine (0.1 mL/100 g body wt; 90 mg ketamine-10 mg xylazine/mL; Henry Schein Animal Health, Dublin, OH), then bilaterally ovariectomized (OVX) and implanted with a subcutaneous (sc) 17β estradiol-3-benzoate (30 µg/mL safflower oil)-filled Silastic capsule (10 mm/100 g body wt; 0.062 inches internal diameter 0.125 inches outer diameter). This steroid replacement regimen yields approximate plasma estradiol concentrations of 22 pg/mL (13), which replicate circulating hormone levels characteristic of 4-day estrous cycle metestrus in ovary-intact female rats (22). Although under anesthesia, animals in groups 1 (n = 4) and 3 (n = 4) were bilaterally infused into the VMN with Accell Control Pool Non-Targeting siRNA [Scramble (Scr); 500 pmol; Prod. No. D-001910-10-20; Horizon Discovery, Waterbeach, UK] using a NeuroStar GmbH (Tübingen, Germany) Drill Injection robot, as described (23). At the same time, groups 2 (n = 4) and 4 (n = 4) were administered Accell Rat GPR81 siRNA set of 4 (500 pmol; Prod. No. A-108404-01-04–0010; Horizon Discovery). Postsurgery rats were injected with enrofloxacin (Enroflox 2.27%; 10 mg/kg body wt im) and ketoprofen (3 mg/kg body wt sc), received topical application of 0.25% bupivacaine (Hospira, Inc.) to closed incisions, then transferred to individual cages. On study day 7, animals were injected subcutaneously at 0900 with the vehicle sterile insulin diluent (V; groups 1 and 2; Eli Lilly & Co., Indianapolis, IN) or neutral protamine Hagedorn insulin [INS; groups 3 and 4; 10 U/kg body wt; Eli Lilly (24)]. Animals were euthanized at 1000 on day 7 by rapid decapitation for brain and trunk blood collection. Dissected brains were snap-frozen in dry ice-cooled isopentane for storage at −80°C. Plasma was stored at −20°C. The experiment was repeated involving an identical set of four treatment groups 1–4 (n = 4 rats/group), with the exception that animals were euthanized by microwave fixation (1.45-s exposure, 4.5 kW energy) for optimum preservation of brain tissue glycogen (In Vivo Microwave Fixation System, Prod. No. 50037; Stoelting Co., Wood Dale, IL) (25).

VMNvl Cell Laser-Catapult Microdissection for Western Blot Analysis

Forebrains from the first experiment were cut into serial 10-μm thick frozen sections over the 1,500-µm rostrocaudal length of the VMN, i.e., between −1.8 mm and −3.3 mm relative to bregma. Sections cut through rostral (−1.8 to − 2.3 mm) or caudal (−2.8 to −3.3 mm) levels of the VMN were mounted on polyethylene naphthalate membrane-coated slides (Prod. No. 415190-9041-000; Carl Zeiss Microscopy, LLC, White Plains, NY) and processed by immunocytochemistry to identify rostral or caudal VMNvl glial fibrillary acidic protein (GFAP)-ir-positive astrocytes, nNOS-ir-positive neurons, or GAD-ir-positive neurons, as described (11, 12, 23, 26), using the following primary antisera: mouse anti-GFAP (Prod. No. 3670S, 1:500; Cell Signaling Technology, Danvers, MA), rabbit anti GAD_65/67 (Prod. No. ABN904, 1:1,000; Millipore Sigma, Burlington, MA), or rabbit anti-nNOS (Prod. No. NBP1-39681, 1:1,000; Novus Biologicals, Littleton, CO). Immunolabeled cells from rostral or caudal VMNvl were individually removed from tissue sections using a Zeiss P.A.L.M. UV-A Microlaser-IV System (Carl Zeiss Microscopy), and collected, for each animal, into lysis buffer [2% sodium dodecyl sulfate (SDS), 0.05 M dithiothreitol, 10% glycerol, 1 mM EDTA, 60 mM Tris·HCl, pH 7.2]. For each protein of interest, three or four separate cell lysate pools were created within each treatment group (n = 50 cells/pool/group) by combining lysate aliquots from individual subjects. Sample pools across treatments were electrophoresed in Bio-Rad TGX 10% stain-free gels (Prod. No. 1610183, Bio-Rad Laboratories, Inc., Hercules, CA). After protein separation, gels were activated (1 min) by UV light in a Bio-Rad ChemiDoc MP Imaging System to acquire measures of total protein present in individual lanes, as described (27). Proteins were then transblotted (30 V, overnight at 4°C) to 0.45-mm PVDF-Plus membranes (Prod. No. 1212639; Data Support, Co., Panorama City, CA). Membranes containing rostral or caudal VMNvl astrocyte proteins were blocked with either Tris-buffered saline, pH 7.4, containing 0.1% Tween-20 and 2% bovine serum albumin or with SuperBlock Blocking Buffer (Prod. No. 37537; Thermo Fisher Scientific, Waltham, MA) before overnight incubation (4°C) with primary antisera raised in rabbit against AMPK_α1/2 (AMPK) (Prod. No. 2532S, 1:2,000; Cell Signaling Technol.), phosphoAMPK_α1/2 (pAMPK) (Prod. No. 2535S, 1:2,000; Cell Signaling Technol.), glycogen phosphorylase-brain form (GPbb) (Prod. No. NBP1-32799, 1:2,000; Novus Biol.), glycogen phosphorylase-muscle form (GPmm) (Prod. No. NBP2-16689, 1:2,000; Novus Biol.), or GPR81 (Prod. No. SAB1300793, 1:2,000; Sigma-Aldrich, St. Louis, MO), as reported (25–27). Proteins from nNOS- or GAD-ir-positive VMNvl neurons were analyzed using a rabbit primary antibody against nNOS (Prod. No. NBP1-39681, 1:2,000; Novus Biol.) or GAD (Prod. No. ABN904, 1:10,000; Millipore Sigma), respectively (12, 20, 23, 28), as well as anti-AMPK, -pAMPK, and -GPR81 antisera described above for analysis of astrocyte proteins. Validation of antibodies used in this study involves determination of protein target specificity, including linear detection of a single band of expected molecular weight, and reproducibility. An essential aspect of establishing the Western blot protocol used here involves the characterization of conditions that allow discriminative detection of optical density readouts that correspond to variable target protein quantities. We observe consistent repeatability of antibody performance on samples representative of a single treatment group, on samples collected from different treatment groups in an individual experiment, and on samples obtained during separate studies carried out by different investigator teams. After incubation (1 h) with horseradish peroxidase-labeled goat anti-rabbit antibody (Prod. No. NEF812001EA, 1:5,000; PerkinElmer, Boston, MA), membranes were exposed to SuperSignal West Femto maximum-sensitivity chemiluminescence substrate (Prod. No. 34096; Thermo Fisher Sci.). Membrane blocking, buffer washes, and antibody incubations were carried out by Freedom Rocker Blotbot VR automation (Next Advance, Inc., Troy, NY). Target protein band optical density (O.D.) signals were detected and quantified using the Bio-Rad ChemiDoc MP Imaging System and ImageLab 6.0.0 software, build 25, 2017, and normalized to in-lane total protein. This superior method for Western blot normalization markedly reduces data variability through improved measurement accuracy and precision (29, 30). Precision plus protein molecular weight dual-color standards (Prod. No. 161–0374, Bio-Rad) were included in each Western blot analysis. Supplemental Fig. S1 (see https://doi.org/10.6084/m9.figshare.21312150.v1) depicts the migration pattern of molecular weight marker protein bands in a representative full Western blot, and correspondence between ladder proteins and molecular weight markers inserted at left as horizontal bars.

uHPLC-Electrospray Ionization-Mass Spectrometry Analysis of Rostral versus Caudal VMNvl Glycogen Content

Brains collected after microwave fixation were cut into 100-μm-thick frozen sections over the rostrocaudal length of the VMN, as described (20). For each animal, VMNvl tissue was bilaterally micropunch-dissected from rostral or caudal VMN sections (defined by anterior-posterior coordinates described above) using a 0.50-μm calibrated hollow needle (Prod. No. 57401; Stoelting), combined within each segment in 0.02 M Tris buffer, pH 7.2, then heat-denatured and homogenized by ultrasonification (25). For each treatment group, separate triplicate sample pools were created within the rostral and the caudal VMNvl by combining aliquots from individual subjects. Sample supernatants (10 μL) were hydrolyzed by incubation (2 h) with 0.5 mg/mL amyloglucosidase (10 μL) and 0.1 M sodium acetate, pH 5.0 (10 μL), followed by sequential heating (100°C; 5 min) and cooling to room temperature. Glycogen was measured in a Thermo Fisher Scientific Vanquish UHPLC + System equipped with Thermo Scientific Dionex Chromeleon 7 Chromatography Data System software, as described (12, 25, 31–33). Column and autosampler temperatures were 35°C and 10°C, respectively. The autosampler needle was washed with 10% (vol/vol) methanol (10 s). Hydrolyzed and nonhydrolyzed samples were derivatized with 100-μL 0.5 M 1-phenyl-3-methyl-5-pyrazolone (PMP) reagent supplemented with 0.3 M NaOH. After acidification with 400-μL 0.75% formic acid and extraction with chloroform, supernatant aliquots (400 μL) were vacuum-concentrated, frozen at −80°C, and lyophilized. Lyophilization products were diluted to 1.0 mL with 10 mM ammonium acetate, bath-sonicated (30 s), and centrifuged. Supernatant aliquots (250 μL) were transferred to 350-μL inserts, which were placed into 2-mL Surestop vials in an autosampler tray. d-(+)-glucose-PMP was resolved using the Shodex Asahipak NH2P-40 3E column with a mobile phase (75:25 vol/vol), acetonitrile:10 mM ammonium acetate (0.2 mL/min). d-(+)-glucose-PMP ion chromatograms were obtained at m/z 510.2 to generate area-under-the curve data. Critical uHPLC-electrospray ionization-mass spectrometry (LC-ESI-MS) parameters, such as sheath gas pressure (SGP; 25 psig), auxiliary gas pressure (AGP; 4.6 psig), sweep gas pressure (SWGP; 0.5 psig), vaporizer temperature (VT; 150°C), ion transfer tube temperature (ITT; 150°C), source voltage (−2,000 V), foreline pressure (1.76 Torr; auto-set by instrument and variable), source gas (nitrogen; Genius NM32LA 110 V, 10–6520; Peak Scientific, Inchinnan, Scotland), and mass peak area detection algorithm (ICIS/Genesis) were each maintained at optimum (25). Sample protein content was determined using a Thermo Fisher Scientific NanoDrop One^c microvolume UV-Vis spectrophotometer (840–274200). Glycogen concentrations were expressed as ug/mg protein.

Plasma Glucose and Counterregulatory Hormone Analyses

Circulating glucose concentrations were measured in duplicate for each subject using an ACCU-CHECK Aviva-plus glucometer (Roche Diagnostic Corporation, Indianapolis, IN) (24). Plasma corticosterone (Cat. No. ADI-900-097; Enzo Life Sciences, Inc., Farmingdale, NY) and glucagon (Cat. No. EZGLU-30K, EMD Millipore, Billerica, MA) levels were determined in duplicate using commercial ELISA kit reagents, as described (27).

Statistical Analyses

Mean-normalized VMNvl protein O.D., VMNvl tissue glycogen measures, and plasma glucose and counter-regulatory hormone concentrations were evaluated by two-way analysis of variance and Student-Newman-Keuls post hoc test. Differences of P < 0.05 were considered significant. In each figure, statistical differences between specific pairs of treatment group means are denoted as follows: *P < 0.05; **P < 0.01; ***P < 0.001. Post hoc statistical power analysis showed that each ANOVA test had a 1-β value of 0.80 or greater.

RESULTS

Figure 1 depicts effects of VMN GPR81 gene knockdown on rostral VMNvl nitrergic neuron GPR81 (Fig. 1A) nNOS (Fig. 1B), total AMPK (Fig. 1C), and pAMPK (Fig. 1D) protein expression profiles and mean pAMPK/AMPK ratio (Fig. 1E). Data in Fig. 1A show that GPR81 siRNA administration caused significant downregulation of GPR81 protein expression in rostral VMNvl nitrergic neurons from V- or INS-injected rats [F(3,8): 95.31, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 222.10; P < 0.0001; INS main effect: F(1,8): 0.22; P = 0.65; Knockdown/INS interaction: F(1,8): 63.59; P < 0.001]. IIH stimulated GPR81 protein in rostral VMNvl NO neurons. As shown in Fig. 1B, GPR81 knockdown elevated nNOS levels in cells collected from V-injected euglycemic rats [GPR81 siRNA/V (diagonal-striped white bar) vs. Scr siRNA/V (solid white bar)] [F(3,8): 20.74. P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 14.32; P = 0.005; INS main effect: F(1,8): 47.36; P < 0.001; Knockdown/INS interaction: F(1,8): 0.535; P = 0.485]. INS administration stimulated nNOS expression in these neurons [Scr siRNA/INS (solid gray bar) vs. Scr siRNA/V (solid white bar)]; this stimulatory response to hypoglycemia was not affected by GPR81 siRNA pretreatment [GPR81 siRNA/INS (diagonal-striped gray bar) vs. Scr siRNA/INS (solid gray bar)]. Data in Fig. 1C indicate that GPR81 gene suppression did not alter AMPK protein expression in euglycemic animals, but amplified inhibitory effects of IIH on this protein profile [F(3,8): 45.37, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 6.86; P = 0.031; INS main effect: F(1,8): 106.47; P < 0.001; Knockdown/INS interaction: F(1,8): 22.78; P < 0.001]. Rostral VMNvl NO neurons exhibit upregulated pAMPK protein content in response to VMN GPR81 gene knockdown (Fig. 1D) [F(3,8): 27.33, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 7.83; P = 0.023; INS main effect: F(1,8): 17.03; P = 0.003; Knockdown/INS interaction: F(1,8): 57.13; P < 0.001]. However, this pretreatment had a contrary effect on pAMPK expression during IIH, as this profile was significantly lower in GPR81 siRNA- versus Scr siRNA-pretreated INS-injected animals. As seen in Fig. 1E, GPR81 knockdown and IIH both increased the mean pAMPK-to-AMPK ratio, yet the latter response was augmented by GPR81 siRNA pretreatment [F(3,8): 33.05; P < 0.001; Knockdown main effect: F(1,8): 45.02, P < 0.001, 1-β = 1.00; INS main effect: F(1,8): 53.08; P < 0.001; Knockdown/INS interaction: F(1,8): 1.05; P = 0.337]. Results show that GPR81 signaling inhibits rostral VMNvl nitrergic neuron nNOS protein expression in euglycemic rats, but does not affect this protein profile during hypoglycemia. Lactate receptor signaling imposes a stimulatory tone on AMPK protein expression during hypoglycemia but lacks influence on this profile during euglycemia. GPR81 regulation of rostral VMNvl NO neuron pAMPK content varies according to metabolic state, causing suppression during euglycemia, yet upregulating this protein during hypoglycemia. GPR81 signaling inhibits mean pAMPK/AMPK in this cell type during eu- and hypoglycemia. GPR81 knockdown-associated suppression of gene product profiles in rostral VMNvl NO neurons supports the inference that the effects of this genetic manipulation on target protein expression reflect, in part, diminished lactate receptor input to these cells.

Figure 1.

GPR81 gene knockdown effects on eu- and hypoglycemic patterns of nitric oxide (NO) synthase enzyme and 5′-AMP-activated protein kinase (AMPK) activity in rostral ventrolateral ventromedial hypothalamic nucleus (VMNvl) nitrergic neurons. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused into the VMN with GPR81 or scramble (SCR) siRNA before subcutaneous vehicle (V) or insulin (INS; 10.0 U/kg body wt) injection. Neuronal nitric oxide synthase (nNOS)-immunoreactive (-ir)-positive located in the rostral VMNvl were collected by laser-catapult microdissection; in each treatment group, aliquots of NO nerve cell lysate from individual subjects were combined to create triplicate samples for Western blot analysis of individual target proteins. Results depict mean rostral VMNvl nitrergic neuron GPR81 (A), nNOS (B), AMPK (C), and pAMPK (D) protein optical density (O.D.) values ± SE for the following treatment groups: SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); GPR81 siRNA/INS (diagonal-striped gray bars). Effects of GPR81 siRNA pretreatment with or without INS injection on the mean ratio of pAMKP/AMPK protein expression in rostral VMNvl nitrergic neurons are presented in E. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2 presents the effects of VMN GPR81 gene knockdown on caudal VMNvl nitrergic nerve cell GPR81 (Fig. 2A), nNOS (Fig. 2B), AMPK (Fig. 2C), and pAMPK (Fig. 2D) protein expression and mean pAMPK/AMPK ratios (Fig. 2E). Figure 2A shows that GPR81 siRNA administration decreased GPR1 protein expression in caudal VMNvl nitrergic neurons acquired from eu- or hypoglycemic animals [F(3,8): 36.49, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 95.89; P < 0.0001; INS main effect: F(1,8): 0.09; P = 0.772; Knockdown/INS interaction: F(1,8): 13.50; P = 0.005]. Data in Fig. 2B show that GPR81 siRNA upregulated nNOS expression in this cell type [F(3,8): 183.33; P < 0.001; Knockdown main effect: F(1,8): 489.07, P < 0.001, 1-β = 1.00; INS main effect: F(1,8): 3.81; P = 0.087; Knockdown/INS interaction: F(1,8): 57.11; P < 0.001]. In this VMNvl segment, NO neurons exhibit downregulated nNOS protein due to IIH; this inhibitory response was reversed by GPR81 RNA pretreatment. Data in Fig. 2, C and D, show that GPR81 siRNA pretreatment did not alter patterns of AMPK [F(3,8): 82.84, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 40.42; P < 0.001; INS main effect: F(1,8): 185.31; P < 0.001; Knockdown/INS interaction: F(1,8): 22.79; P = 0.001] or pAMPK [F(3,8): 15.27, P = 0.001, 1-β = 0.996; Knockdown main effect: F(1,8): 22.06; P = 0.002; INS main effect: F(1,8): 20.04; P = 0.002; Knockdown/INS interaction: F(1,8): 3.69; P = 0.091] protein expression in euglycemic animals, but exacerbated IIH-associated upregulation of AMPK. Figure 2E shows that caudal VMNvl nitrergic neuron pAMPK/AMPK values were refractory to GPR81 knockdown [ratio: F(3,8): 6.04, P = 0.019, 1-β = 0.990; Knockdown main effect: F(1,8): 2.39; P = 0.161; INS main effect: F(1,8): 14.91; P = 0.005; Knockdown/INS interaction: F(1,8): 0.82; P = 0.392]. Outcomes show that GPR81 input is inhibitory to caudal VMNvl NO nerve cell nNOS protein during eu- and hypoglycemia and suppresses total AMPK and pAMPK protein profiles under the latter conditions.

Figure 2.

Impact of l-lactate receptor signaling on caudal VMNvl nitrergic nerve cell nNOS protein levels and pAMPK/AMPK protein ratio during eu- versus hypoglycemia. Groups of ovariectomized, estradiol-implanted adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Aliquots of laser-microdissected caudal VMNvl neuronal nitric oxide synthase (nNOS)-immunoreactive (-ir)-positive neuron lysates were combined within treatment groups to create triplicate samples for Western blot analysis of individual target proteins. Data show mean caudal VMNvl nitrergic neuron GPR81 (A), nNOS (B), AMPK (C), and pAMPK (D) protein O.D. measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Pretreatment and treatment effects on mean pAMKP/AMPK protein content are illustrated in E. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001. AMPK, 5′-monophosphate-activated protein kinase; nNOS, neuronal nitric oxide synthase; pAMPK, phosphoAMPK; VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Figure 3 illustrates GPR81 gene knockdown effects on rostral VMNvl GABAergic nerve cell GPR81 (Fig. 3A), GAD (Fig. 3B), AMPK (Fig. 3C), and pAMPK (Fig. 3D) protein expression and mean pAMPK/AMPK protein ratio (Fig. 3E) in V- versus INS-injected rats. Data in Fig. 3A show that GPR81 gene suppression reduced GPR81 protein profiles in both V- and INS-injected rats [F(3,8): 39.79, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 22.33; P < 0.001; INS main effect: F(1,8): 96.94; P < 0.001; Knockdown/INS interaction: F(1,8): 0.08; P = 0.779]. IIH significantly reduced GPR81 protein in this neuron population, an inhibitory effect that was intensified by GPR81 siRNA pretreatment. Data in Fig. 3B show that GPR81 siRNA suppressed GAD protein expression in V- or INS-injected animals [F(3,8): 13.66; P = 0.0021-β = 0.991; Knockdown main effect: F(1,8): 32.62; P < 0.001; INS main effect: F(1,8): 8.37; P = 0.020; Knockdown/INS interaction: F(1,8): 0.007; P = 0.936]. GPR81 signaling did not alter AMPK protein expression in these cells under euglycemic conditions (Fig. 3C) but exerted an inhibitory tone on this protein profile during hypoglycemia [F(3,8): 6.16, P = 0.018, 1-β = 0.812; Knockdown main effect: F(1,8): 12.01; P = 0.009; INS main effect: F(1,8): 1.92; P = 0.203; Knockdown/INS interaction: F(1,8): 4.55; P = 0.065]. GPR81 gene repression downregulated rostral VMNvl GABA nerve cell pAMPK expression during eu- and hypoglycemia (Fig. 3D) [F(3,12): 29.20, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,12): 55.33; P < 0.001; INS main effect: F(1,12): 31.92; P < 0.001; Knockdown/INS interaction: F(1,12): 0.34; P = 0.568]. In INS-injected rats, GPR81 siRNA pretreatment exacerbated IIH-associated diminution of pAMPK expression compared with the Scr-pretreated group. Data in Fig. 3E show that GPR81 gene knockdown downregulated mean pAMPK/AMPK protein ratio values in eu- and hypoglycemic rats [F(3,12): 69.94, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,12): 178.48; P < 0.001; INS main effect: F(1,12): 25.52; P < 0.001; Knockdown/INS interaction: F(1,12): 7.82; P = 0.016]. Results show that GPR81 signaling stimulates GAD and pAMPK protein profiles and increases the ratio of pAMPK/AMPK expression in rostral VMNvl GABA neurons during glucose homeostasis or imbalance. Validation of GPR81 knockdown-induced diminution of GPR81 protein in these cells infers that outcomes of this manipulation involve, in part, decreased receptor input to these cells.

Figure 3.

Effects of GPR81 siRNA on eu- and hypoglycemic patterns of γ-aminobutyric acid (GABA) synthetic enzyme and AMPK activity expression in glutamate decarboxylase (GAD)-immunopositive rostral VMNvl neurons. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Aliquots of laser-microdissected rostral VMNvl GAD-ir-positive neuron lysates were combined within treatment groups to create quadruplicate sample pools for Western blot analysis of individual target proteins. Data show mean rostral VMNvl GABA nerve cell GPR81 (A), GAD (B), AMPK (C), and pAMPK (D) protein O.D. measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Pretreatment and treatment effects on mean ratio of pAMKP/AMPK protein expression in rostral VMNvl GABAergic neurons are presented in E. *P < 0.05, **P < 0.01, ***P < 0.001. AMPK, 5′-monophosphate-activated protein kinase; pAMPK, phosphoAMPK; VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Data presented in Fig. 4 illustrate the effects of GPR81 siRNA pretreatment on caudal VMNvl GABA nerve cell GPR81 (Fig. 4A), GAD (Fig. 4B), AMPK (Fig. 4C), and pAMPK (Fig. 4D) protein expression and in mean pAMPK/AMPK protein ratio (Fig. 4E) in V- versus INS-injected animals. Data in Fig. 4A show that GPR81 gene suppression significantly decreased GPR81 protein content in caudal VMNvl GABA neurons [F(3,8): 6.62, P = 0.015, 1-β = 0.841; Knockdown main effect: F(1,8): 19.61; P = 0.002; INS main effect: F(1,8): 0.24; P = 0.637; Knockdown/INS interaction: F(1,8): 0.01; P = 0.938]. In contrast to the rostral VMNvl, GABAergic neurons in the caudal segment of this structure exhibit upregulated GAD protein in response to IIH (Fig. 4B) [F(3,8): 20.09, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 7.33; P = 0.027; INS main effect: F(1,8): 2.31; P = 0.167; Knockdown/INS interaction: F(1,8): 50.63; P < 0.001]. GPR81 gene repression increased GAD protein expression in euglycemic rats but inhibited this protein profile in hypoglycemic animals. Data in Fig. 4C indicate that GPR81 does not regulate AMPK protein in this cell population under either eu- or hypoglycemia [F(3,8): 5.96, P = 0.019, 1-β = 0.798; Knockdown main effect: F(1,8): 3.30; P = 0.107; INS main effect: F(1,8): 14.54; P = 0.005; Knockdown/INS interaction: F(1,8): 0.04; P = 0.842]. Figure 4D depicts GPR81 suppression of pAMPK expression in INS-injected rats [F(3,8): 6.11, P = 0.018, 1-β = 0.809; Knockdown main effect: F(1,8): 9.47; P = 0.015; INS main effect: F(1,8): 7.92; P = 0.023; Knockdown/INS interaction: F(1,8): 0.95; P = 0.359]. Results in Fig. 4E show that the mean pAMPK/AMPK protein ratio was refractory to GPR81 siRNA and/or INS [F(3,8): 3.35, P = 0.076, 1-β = 0.820; Knockdown main effect F(1,8): 7.12; P = 0.028; INS main effect: F(1,8): 2.75; P = 0.136; Knockdown/INS interaction: F(1,8): 0.19; P = 0.674]. Data show that GPR81 signaling inhibits caudal VMNvl GABA neuron GAD protein levels during glucose homeostasis but yet is required for hypoglycemia-associated upregulation of this protein profile. Unlike GABAergic cells acquired from rostral VMNvl, GPR81 protein expression in VMNvl GABA neurons is insensitive to IIH.

Figure 4.

Lactate receptor regulation of caudal VMNvl GABAergic nerve cell GAD profiles and pAMPK/AMPK protein ratio. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Aliquots of laser-microdissected caudal VMNvl GAD-ir-positive neuron lysates were combined within treatment groups to create triplicate samples for Western blot analysis of individual target proteins. Data depict mean caudal VMNvl GABA nerve cell GPR81 (A), GAD (B), AMPK (C), and pAMPK (D) protein O.D. measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Pretreatment and treatment effects on mean ratio of pAMKP/AMPK protein expression in caudal VMNvl GABAergic neurons are presented in E. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, ***P < 0.001. AMPK, 5′-monophosphate-activated protein kinase; GAD, glutamate decarboxylase; pAMPK, phosphoAMPK; VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Figure 5 illustrates the effects of GPR81 gene knockdown on energy sensor and glycogen metabolic protein expression in GFAP-ir-positive laser-microdissected rostral VMNvl astrocytes. Data in Fig. 5A indicate that rostral VMNvl astrocytes express GPR81 protein and that this profile is upregulated in response to IIH [F(3,8): 46.71, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 20.23; P = 0.002; INS main effect: F(1,8): 116.37; P < 0.001; Knockdown/INS interaction: F(1,8): 3.54; P = 0.097]. Data in Fig. 5B indicate that GPR81 siRNA decreased AMPK protein in astrocytes acquired from V- or INS-injected rats [F(3,8): 55.37, P < 0.001,1-β = 1.00; Knockdown main effect: F(1,8): 31.23; P < 0.001; INS main effect: F(1,8): 132.86; P < 0.001; Knockdown/INS interaction: F(1,8): 2.03; P = 0.192]. Hypoglycemia-associated downregulation of AMPK was exacerbated by GPR81 siRNA pretreatment. GPR81 gene repression had inverse effects on astrocyte pAMPK protein content as this genetic manipulation upregulated this protein in euglycemic animals but reduced pAMPK expression during hypoglycemia (Fig. 5C) [F(3,8): 76.64, P = 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 35.41; P < 0.001; INS main effect: F(1,8): 51.91; P < 0.001; Knockdown/INS interaction: F(1,8): 142.59; P < 0.001]. The latter inhibitory effect reversed hypoglycemic stimulation of pAMPK levels. Data in Fig. 5D show that GPR81 siRNA increased mean pAMPK/AMPK protein content during euglycemia but did not affect this ratio during hypoglycemia [F(3,8): 34.19; P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 29.13; P = 0.001; INS main effect: F(1,8): 2.94; P = 0.125; Knockdown/INS interaction: F(1,8): 70.51; P < 0.001]. GPR81 gene knockdown did not affect rostral VMNvl astrocyte GS protein levels in euglycemic rats but reversed IIH-induced suppression of this protein [F(3,12): 16.84, P = 0.001, 1-β = 0.998; Knockdown main effect: F(1,12): 15.76; P = 0.004; INS main effect: F(1,12): 0.31; P = 0.596; Knockdown/INS interaction: F(1,12): 38.56; P < 0.001]. Figure 5F shows that GPR81 gene knockdown altered GPbb protein expression in INS-, but not V-injected rats [F(3,8): 6.60, P = 0.015, 1-β = 0.839; Knockdown main effect: F(1,8): 2.82; P = 0.131; INS main effect: F(1,8): 0.66; P = 0.441; Knockdown/INS interaction: F(1,8): 16.31; P = 0.004]. As shown in Fig. 5G, rostral VMNvl astrocytes collected after V injection showed lower GPmm protein levels after pretreatment with GPR81 siRNA [F(3,8): 22.43, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 14.39; P = 0.005; INS main effect: F(1,8): 50.23; P < 0.001; Knockdown/INS interaction: F(1,8): 2.68; P = 0.140]. Hypoglycemic downregulation of GPmm protein expression was not further exacerbated by GPR81 gene suppression. Results here show GPR81 signaling stimulates rostral VMNvl astrocyte total AMPK regardless of metabolic state, yet imposes bidirectional control of pAMPK protein profiles during eu- versus hypoglycemia. Data also show that GPR81 control of astrocyte metabolic enzyme protein profiles is metabolic state specific.

Figure 5.

Effects of GPR81 siRNA on eu- and hypoglycemic patterns of rostral VMNvl astrocyte AMPK activity and glycogen metabolic enzyme protein expression. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Aliquots of laser-microdissected caudal VMNvl glial fibrillary acidic protein (GFAP)-ir-positive astrocyte lysates were combined within treatment groups to create triplicate sample pools for Western blot analysis of individual target proteins. Data show mean rostral VMNvl astrocyte GPR81 (A), AMPK (B), pAMPK (C), glycogen synthase (E), glycogen phosphorylase (GP)-brain type (GPbb) (F), or GP-muscle type (GPmm) (G) protein O.D. measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Pretreatment and treatment effects on rostral VMNvl astrocyte pAMKP/AMPK protein ratio are presented in D. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001. AMPK, 5′-monophosphate-activated protein kinase; pAMPK, phosphoAMPK; VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Figure 6 depicts GPR81 gene knockdown regulation of caudal VMNvl astrocyte protein expression. Figure 6A shows that caudal VMNvl astrocytes express GPR81 protein and that IIH-associated upregulation of this profile is prevented by GPR81 gene suppression [F(3,8): 34.81, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 23.49; P = 0.001; INS main effect: F(1,8): 79.09; P < 0.001; Knockdown/INS interaction: F(1,8): 1.84; P = 0.212]. Data in Fig. 6B indicate that AMPK protein was affected by GPR81 siRNA pretreatment in INS-, but not V-injected rats [F(3,8): 18.20, P = 0.001, 1-β = 0.999; Knockdown main effect: F(1,8): 42.72; P < 0.001; INS main effect: F(1,8): 0.01; P = 0.925; Knockdown/INS interaction: F(1,8): 11.88; P = 0.009]. Hypoglycemia-associated downregulation of AMPK expression was prevented by GPR81 gene repression. GPR81 gene suppression amplifies pAMPK protein in V-, but not INS-injected rats (Fig. 6C) [F(3,8): 13.53, P = 0.002, 1-β = 0.991; Knockdown main effect: F(1,8): 9.43; P = 0.015; INS main effect: F(1,8): 31.04; P = 0.001; Knockdown/INS interaction: F(1,8): 0.122; P = 0.736]. IIH-associated stimulation of this protein profile was unaffected by GPR81 siRNA pretreatment. Results shown in Fig. 6D indicate that mean pAMPK/AMPK ratio values in euglycemic animals were unaffected by GPR81 siRNA, but IIH-associated augmentation of this ratio was averted by GPR81 knockdown [F(3,8): 39.68, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 33.17; P < 0.001; INS main effect: F(1,8): 50.71; P < 0.001; Knockdown/INS interaction: F(1,8): 35.17; P < 0.001]. Data in Fig. 6E indicate that GPR81 knockdown affected caudal VMNvl astrocyte GS protein content in INS- but not V-injected animals [F(3,12): 5.95, P = 0.020, 1-β = 0.797; Knockdown main effect: F(1,12): 13.50; P = 0.006; INS main effect: F(1,12): 0.29; P = 0.600; Knockdown/INS interaction: F(1,12): 4.04; P = 0.079]. Figure 6F illustrates the preventive effects of GPR81 gene knockdown on hypoglycemic upregulation of VMNvl astrocyte GPbb protein expression [F(3,8): 7.45, P = 0.011, 1-β = 0.883; Knockdown main effect: F(1,8): 16.42; P = 0.004; INS main effect: F(1,8): 1.16; P = 0.312; Knockdown/INS interaction: F(1,8): 4.77; P = 0.060]. Data in Fig. 6G indicate that GPR81 gene repression also inhibited GPmm protein in INS-injected rats [F(3,8): 13.97, P = 0.002, 1-β = 0.992; Knockdown main effect: F(1,8): 22.41; P = 0.001; INS main effect: F(1,8): 18.13; P = 0.003; Knockdown/INS interaction: F(1,8): 1.36; P = 0.278]. Results show that GPR81 exerts differential control of total AMPK versus pAMPK protein under each metabolic condition and affects the ratio of pAMPK/AMPK expression during hypoglycemia only. Lactate receptor signaling regulates caudal VMNvl astrocyte glycogen enzyme protein profiles during hypoglycemia, not euglycemia.

Figure 6.

Lactate receptor regulation of caudal VMNvl astrocyte AMPK activity and glycogen metabolic enzyme protein expression. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Aliquots of laser-microdissected caudal GFAP-ir-positive astrocyte lysates were combined within treatment groups to create triplicate samples for Western blot analysis of individual target proteins. Data show mean caudal VMNvl astrocyte GPR81 (A), AMPK (B), pAMPK (C), glycogen synthase (E), glycogen phosphorylase (GP)-brain type (GPbb) (F), or GP-muscle type (GPmm) (G) protein O.D. measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Pretreatment and treatment effects on rostral VMNvl astrocyte pAMKP/AMPK protein ratio are presented in D. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001. AMPK, 5′-monophosphate-activated protein kinase; pAMPK, phosphoAMPK; VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Effects of VMN GPR81 siRNA on rostral versus caudal VMNvl tissue glycogen are shown in Fig. 7. Figure 7A [F(3,8): 32.42, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 13.41; P = 0.006; INS main effect: F(1,8): 19.96; P = 0.002; Knockdown/INS interaction: F(1,8): 63.89; P < 0.001] and B [F(3,8): 40.82, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,8): 0.21; P = 0.657; INS main effect: F(1,8): 66.16; P < 0.001; Knockdown/INS interaction: F(1,8): 56.09; P < 0.001], shows that gene knockdown caused a significant reduction in rostral and caudal VMNvl glycogen content, respectively, in euglycemic animals. IIH decreased glycogen levels in the rostral, but not caudal VMNvl; GPR81 siRNA pretreatment prevented this suppressive effect on rostral VMNvl glycogen (Fig. 7A). Rostral and caudal VMNvl glycogen concentrations were higher in GPR81 siRNA-pretreated, INS-injected rats compared with the SCR-pretreated, V-injected controls. Outcomes show that VMN GPR81 signaling augments rostral and caudal VMNvl glycogen amassment during glucose homeostasis, yet conversely inhibits glycogen accumulation in each site during IIH.

Figure 7.

Lactate receptor regulation of rostral versus caudal VMNvl tissue glycogen content. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Micropunch-dissected rostral VMNvl tissue lysates were pooled for triplicate sample analysis of glycogen content for each treatment group by ultrahigh performance liquid chromatograph-electrospray-mass spectrometry (uHPLC-ESI-MS). Data illustrate mean rostral (A) or caudal (B) VMNvl tissue glycogen measures ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001. VMN, ventromedial hypothalamic nucleus; VMNvl, ventrolateral VMN; SCR, scramble.

Data presented in Fig. 8 depict GPR81 gene silencing effects on eu- and hypoglycemic plasma glucose (Fig. 8A), glucagon (Fig. 8B), and corticosterone (Fig. 8C) profiles. As shown in Fig. 8A, VMN GPR81 gene knockdown caused a slight but significant reduction in plasma glucose levels [F(3,12): 183.48, P < 0.001, 1-β = 1.00; Knockdown main effect: F(1,12): 2.54; P = 0.137; INS main effect: F(1,12): 543.93; P < 0.001; Knockdown/INS interaction: F(1,12): 3.97; P = 0.069]. Groups pretreated with SCR versus GPR81 siRNA showed equivalent reductions in circulating glucose at 1 h post-INS treatment. Basal plasma glucagon (Fig. 8B) and corticosterone (Fig. 8C) concentrations were unaffected by GPR81 gene silencing. INS treatment significantly increased both counter-regulatory hormone profiles. GPR81 siRNA pretreatment had no effect on hypoglycemic patterns of glucagon release [F(3,12): 8.57, P < 0.0011-β = 0.986; Knockdown main effect: F(1,12): 0.43; P = 0.519; INS main effect: F(1,12): 24.91; P < 0.001; Knockdown/INS interaction: F(1,12): 0.38; P = 0.543], but intensified corticosterone secretion [F(3,12): 14.71, P < 0.001, 1-β = 0.994; Knockdown main effect: F(1,12): 2.97; P = 0.125; INS main effect: F(1,12): 36.79; P < 0.001; Knockdown/INS interaction: F(1,12): 4.37; P = 0.070] in INS-injected animals. Current data show that VMN GPR81 signaling stimulates baseline plasma glucose levels and opposes IIH-induced augmentation of corticosterone secretion.

Figure 8.

Effects of VMN GPR81 gene knockdown on plasma glucose and counter-regulatory hormone profiles during eu- versus hypoglycemia. Groups of ovariectomized, estradiol-replaced adult female rats (n = 4/group) were infused with GPR81 or SCR siRNA into the VMN before subcutaneous injection of vehicle (V) or insulin (INS) (10.0 U/kg body wt). Plasma samples were analyzed for glucose (glucometer), glucagon (ELISA), or corticosterone (ELISA) concentrations. Data show mean plasma glucose (A; n = 4 animals/group), glucagon (B; n = 4 animals/group), and corticosterone (C; n = 4 animals/group) concentrations ± SE for SCR siRNA/V (solid white bars); GPR81 siRNA/V (diagonal-striped white bars); SCR siRNA/INS (solid gray bars); and GPR81 siRNA/INS (diagonal-striped gray bars) treatment groups. Circles depict individual independent data points. Data were analyzed by two-way ANOVA and Student-Neuman-Keuls post hoc test, using GraphPad Prism, Vol. 8 software. *P < 0.05, **P < 0.01, ***P < 0.001. VMN, ventromedial hypothalamic nucleus.

DISCUSSION

Current studies provide unique confirmation that female rat VMNvl glucose-regulatory neurons and astrocytes are direct targets of lactate extracellular signaling and show that hypoglycemia affects receptivity of these cell types to lactate transmitter input. GPR81 controls nitrergic and GABAergic neurotransmission in rostral and caudal segments of the VMNvl, yet this receptor governs AMPK activity in these nerve cell types in the rostral VMNvl only. Data demonstrate GPR81 regulation of VMNvl astrocyte glycogen metabolism and accumulation during both glucose homeostasis and imbalance, and that astrocytes present in rostral versus caudal VMNvl segments exhibit heterogeneous reactivity to receptor input. Observations of bidirectional change and gain-or-loss of GPR81 impact on distinctive neuron or astrocyte protein profiles during eu- versus hypoglycemia infer that energy state may affect cellular receptivity to or postreceptor processing of lactate extracellular signaling. Further research is needed to characterize factors that govern GPR81 expression in VMNvl neurotransmitter nerve cell types and astrocytes, elucidate signal transduction mechanisms that mediate lactate receptor effects described here, and determine if GPR81 regulation of VMNvl neuron and astrocyte target proteins varies by sex. Present outcomes implicate VMN GPR81 in neural control of hypoglycemic hypercorticosteronemia. One or more VMN cell types characterized here by GPR81-dependent protein responses to IIH may be an effector of that regulatory action.

Current evidence for GPR81 gene knockdown-associated suppression of nNOS protein profiles in rostral and caudal VMNvl nitrergic neurons infers lactate receptor-dependent inhibition of local glucose-stimulatory NO release during glucose homeostasis. Divergent IIH effects on rostral versus caudal VMNvl nNOS expression align with prior reports [17]. Data here show that GPR81 signaling does not impede IIH-associated upregulation of nNOS in rostral VMNvl, yet is involved in diminution of caudal VMNvl nNOS. It is unclear if this divergent control reflects, in part, dissimilar lactate signal strength. Although both NO nerve cell subpopulations exhibit elevated GPR81 protein due to hypoglycemia, the possibility that IIH may differentially affect postreceptor signaling mechanisms in each VMNvl segment cannot be excluded. Interestingly, IIH was found here to increase the ratio of pAMPK/AMPK expression in rostral but not caudal VMNvl nitregic neurons. Moreover, GPR81 input evidently negatively controls eu- and hypoglycemia-associated AMPK activity in these rostral cells but does not control this sensor in the caudal subpopulation. Present outcomes infer that lactate extracellular signaling may regulate euglycemic patterns of NO transmission in rostral VMNvl nitrergic neurons through AMPK-dependent mechanisms, yet does not control NO during hypoglycemia despite inhibition of AMPK activity. An important concept that emerges here is the functional heterogeneity of neuroanatomically distinct subsets of NO neurons in the VMNvl, as GPR81 evidently imposes dissimilar control of NO release by these cell populations during hypoglycemia. Lack of correlated lactate receptor regulation of hypoglycemic patterns of nNOS expression and AMPK activity in rostral and caudal VMNvl nitrergic nerve cells supports the possibility that lactate controls this glucose-stimulatory neurochemical during glucose deficiency through gliotransmitter action.

GPR81 imposes divergent control of GAD protein profiles in rostral versus caudal VMNvl GABA neurons from euglycemic rats, inferring that lactate extracellular signaling may stimulate or inhibit glucose-inhibitory GABA transmission in these respective locations during glucose sufficiency. It is unclear if one or both antagonistic GPR81-controlled GABA signals affect local regulation of blood glucose. Although these segment-specific GABA nerve cell populations exhibit dissimilar changes in GAD protein content in reaction to hypoglycemia, this protein profile is repressed in each location by GPR81. Lactate receptor signaling may thus limit diminution of rostral VMNvl GABA transmission during hypoglycemia, yet coincidently drive caudal VMNvl GABA release. IIH downregulates GABAergic neuron GPR81 protein expression in the rostral VMNvl but had no impact on this protein in the caudal VMNvl, indicative of site-specific change in receptivity to lactate. The switch between GPR81 inhibition (euglycemic) versus stimulation (hypoglycemia) of caudal VMNvl GABA is evidently metabolic state dependent and may thus involve regulation of postreceptor signaling. It remains to be determined if and how lactate signal volume is affected by IIH in each VMNvl segment and whether adjustment in ligand concentration contributes to hypoglycemic patterns of GAD expression in each location. Data show that GPR81 stimulates AMPK activity in rostral VMNvl GABA neurons during eu- and hypoglycemia, but does not control sensor activity in this cell type in the caudal VMNvl. Thus, bidirectional lactate receptor regulation of caudal VMNvl GABA may likely not involve AMPK.

Results indicate that rostral and caudal VMNvl astrocytes are direct targets for lactate transmitter regulation and show that GPR81 protein in each astrocyte population is upregulated during hypoglycemia. Rostral and caudal VMNvl astrocytes exhibited elevated AMPK activity in reaction to IIH, an outcome that differs from glucoprivic downregulation of sensor activation in primary hypothalamic astrocyte cell cultures in vitro (32). These discrepant outcomes likely reflect, in part, differences between astrocyte integration of nutrient and systemic metabolic cues in vivo versus reaction in vitro to the former signal alone. Current data show that GPR81 regulation of rostral versus caudal VMNvl astrocyte AMPK activity is metabolic state-specific as this receptor inhibits sensor activity in the rostral site in subjects with euglycemia, yet stimulates activation in caudal VMNvl astrocytes during hypoglycemia. Further effort is warranted to establish postreceptor signaling mechanisms that operate in these distinctive astrocyte populations to control AMPK activity and to identify respective downstream targets of AMPK activation. GPR81 upregulated rostral VMNvl astrocyte GPmm during euglycemia but enhanced rostral and caudal VMNvl GPbb protein profiles during hypoglycemia. GS is active in the nonphosphorylated state and is allosterically activated by glucose 6-phosphate, whereas GP is activated by phosphorylation or AMP allosteric effects. It should be noted that current work does not disclose if or how GS and GP protein-specific activities in rostral versus caudal VMNvl astrocytes are controlled by GPR81 signaling. There is justification for additional research to investigate whether GPR81 exerts VMNvl segment-specific effects on expression of activated GS, GPbb, and GPmm relative to total astrocyte glycogenic enzyme protein profiles, as such information could shed light on lactate receptor regulation of glycogen turnover aside from mass. The issue concerning whether astrocyte glycogen in its entirety is a common substrate for GPbb- versus GPmm-mediated breakdown, or alternatively if this fuel reserve is organized into spatially distinct pools that are disassembled by a single GP variant has not yet been addressed. In the current context, it would be useful to learn if GPR81 elicits similar or disparate changes in GPbb versus GPmm enzyme activity, and how changes in distinct GP isoform-specific activities may affect net glycogen mass. It is noteworthy to comment that efforts to resolve this important question are currently impeded by the lack of available antibody-based analytical tools for quantification of GP-variant phosphorylation status. LC-ESI-MS analysis of rostral and caudal VMNvl tissue glycogen content disclosed here reductions in glycogen levels in each location due to genetic repression of GPR81 signaling in euglycemic animals. Yet, during hypoglycemia, GPR81 was observed to exert an opposite effect, e.g., inhibition of glycogen accumulation in the rostral VMNvl. Thus, GPR81 signaling evidently serves to increase glycogen amassment in each VMNvl segment during glucose homeostasis, yet promotes glycogen disassembly in a selective location within the VMNvl during IIH. Further work will be required to examine whether this shift in direction of lactate receptor control of glycogen mass between states of eu- versus hypoglycemia reflects, in part, dissimilar GPR81 receptor expression levels and/or change in receptor-ligand concentration. GPR81 evidently functions to regulate neuro-metabolic stability by controlling volume of glycogen-derived ligand for GPR81 expressed by glucose-regulatory neurons.

An important outcome of current research is the novel implication of VMN GPR81 in neural control of endocrine counter-regulatory outflow. Data here show that this targeted delivery exacerbates hypoglycemic hypercorticosteronemia, inferring that GPR81 signaling blunts this hormone response to systemic glucose insufficiency. Observed inefficacy of VMN GPR81 knockdown to affect plasma glucose levels in INS-injected rats suggests that effects of enhanced corticosterone release due to that pretreatment may be mitigated by a lack of change in outflow volume of other critical counter-regulatory hormones, namely, glucagon and adrenomedullary epinephrine.

As present work centered on the adult female rat as experimental model, it is imperative for future research to investigate whether sex differences exist regarding GPR81 protein expression patterns and regulation of target proteins in neurotransmitter-characterized VMNvl neuron types and astrocytes during glucose sufficiency versus deficiency. It is unclear if estradiol, a key modulator of VMN nitrergic and GABAergic nerve cell responses to hypoglycemia (28), may be an important factor in potential sex-dimorphic GPR81 control of these nerve cell types. As plasma estradiol secretion varies significantly over the 4- to 5-day course of the rat estrous cycle, the current project used an ovariectomized, estradiol-treated animal model to standardize plasma estradiol levels in study subjects at an equivalent physiological level corresponding to a specific stage of the estrous cycle (13), to circumvent the significant potential for confounding effects of dissimilar patterns of endogenous estradiol secretion that occur between various stages of the estrous cycle in ovary-intact adult female rats. Further studies are needed to examine whether estrous cycle stage-specific estradiol output in the female may have a variable impact on GPR81 regulation of VMNvl neuron and astrocyte protein profiles and glycogen accumulation under conditions of eu- and hypoglycemia. It should be noted that the current strategy of normalizing circulating estradiol concentrations among subjects is impeded by deviation from the normal physiological circumstance of exposure to dynamic day-to-day fluctuations in endogenous steroid secretion over the course of the estrous cycle. Thus, the prospect that constant versus transient exposure to metestrus-like plasma estradiol levels may result in discrepant effects of FD on A2 nerve cell gene profiles investigated here cannot be overlooked.

The experimental design implemented here investigated the role of GPR81 in VMN cell type-specific transmitter biomarker and energy sensor protein, VMN tissue glycogen, and systemic counter-regulatory hormone responses to neuroglucopenia due to IIH as well as sensory input from peripheral metabolic detectors on cellular glucose uptake and energy production. There should be consideration given here to the prospect that insulin therapy alone might directly affect, in either a regulatory or modulatory capacity, one or more of the experimental endpoints evaluated here, independent of lactate control as a GRP81 ligand or as a substrate fuel. For example, VMN nitrergic or GABAergic neurons are a plausible target for insulin regulation of biosynthetic enzyme protein expression in the presence and/or absence of receptor-mediated lactate volume transmitter input.

Current studies underscore the utility of single-cell laser-catapult-microdissection for verification of efficacy of gene product knockdown in individual brain cell types. Outcomes identify neuroanatomical populations of VMNvl astrocytes and glucose-regulatory neurons that exhibit dissimilar target protein and glycogen reactivity to GPR81 control. Differential GPR81 regulation of discrete target proteins during eu- versus hypoglycemia infers that energy state may modulate cellular sensitivity to or postreceptor processing of lactate extracellular signaling. It is noted that definitive understanding of whether cell type-specific responses to GPR81 gene knockdown result from downregulated intracellular expression of that receptor or, alternatively, may reflect in part altered GPR81-controlled signaling from other cells will require an experimental approach that involves selective genetic manipulation of GPR81 expression in distinctive brain cell populations. There is also a need for verification that GPR81-mediated adjustments in AMPK activity state affect target cell function, namely, whether augmentation of pAMPK protein expression affects glucose-regulatory transmitter marker protein in distinctive VMN neuron populations and/or glycogen metabolism in VMN astrocytes. Current work identifies several key cell type-specific (astrocyte glycogen; glucose-regulator neuron transmitter marker and energy sensor protein profiles) and systemic (plasma counter-regulatory hormone levels) parameters that exhibit evident sensitivity to GPR81 control. A future challenge is to establish whether a causal relationship links GPR81 regulation of astrocyte glycogen metabolism with glucose-regulatory neuron energy stability and neurotransmitter production and whether GPR81 control of nitrergic and/or GABAergic signaling mediates control of counter-regulatory hormone release by this lactate receptor. Ongoing research seeks to characterize factors that govern VMNvl neuronal and astrocyte GPR81 expression, identify postreceptor signaling mechanisms that mediate lactate receptor actions during eu- versus hypoglycemia, and determine if GPR81 regulation of VMNvl neuron and astrocyte target protein expression and astrocyte glycogen accumulation is sex-dimorphic. Present outcomes implicate VMN GPR81 in neural control of hypoglycemic hyperglucagonemia and hypercorticosteronemia. One or more VMNvl cell types characterized here by GPR81-dependent protein responses to IIH may be an effector of that regulatory action.

Although Western blotting is a very powerful analytical technique for protein separation and detection, a relative drawback concerns limitation of number of samples that can be processed simultaneously, under identical conditions, by electrophoresis. For complex experimental designs involving multiple treatment groups, this methodological constraint precludes concurrent analysis of samples from all subjects across all treatment groups. The current study involved laser-catapult-microdissection of individual VMNvl astrocytes or phenotypically characterized neurons from each subject in each treatment group. Here, limitation imposed by number of electrophoresis gel lanes was addressed by creating, for each target protein, three or four separate cell lysate sample pools within each treatment group, by combining cell lysate aliquots from individual animal subjects. Thus, averaged cell-type target protein levels were compared between treatment groups in three or four independent Western blot runs. It should be noted, however, that this approach does not disclose potential cell-to-cell variation in target protein expression in individual subjects in a given treatment group.

Perspectives and Significance

Outcomes here affirm a physiological role for lactate gliotransmitter signaling in astrocyte-nerve cell metabolic coupling and VMN regulation of glucose homeostasis. Results identify a promising new path of investigation centered on the characterization of signal transduction pathways and genomic versus nongenomic molecular mechanisms that underlie lactate volume control of brain cells and systemic indicators of metabolic stability, with emphasis on potential sex disparities. This novel research focus is expected to yield valuable insight on if and how lactate receptor signaling may coordinate activity within and between diverse neurochemical nerve cell populations that function within neural networks that govern and link critical physiological functions and behaviors under recognized VMN control.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Fig. S1: https://doi.org/10.6084/m9.figshare.21312150.

GRANTS

This study was supported by National Institutes of Health Grant DK109382 to K. P. Briski.

DISCLAIMERS

The authors verify that this work has not been published previously, is not under consideration for publication elsewhere, that it is approved by all authors, and that if accepted for publication, it will not be published elsewhere in the same form without written consent of the copyright-holder.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.C.R., K.B., and K.P.B. conceived and designed research; S.C.R., P.R.N., M.P., and K.B. performed experiments; S.C.R. and P.R.N. analyzed data; S.C.R., P.R.N., and K.P.B. interpreted results of experiments; S.C.R. and K.P.B. prepared figures; K.P.B. drafted manuscript; S.C.R., P.R.N., K.B., and K.P.B. edited and revised manuscript; S.C.R., P.R.N., M.P., K.B., and K.P.B. approved final version of manuscript.