White Adipocyte Adiponectin Exocytosis Is Stimulated via β3-Adrenergic Signaling and Activation of Epac1: Catecholamine Resistance in Obesity and Type 2 Diabetes
Abstract
This comprehensive study embarked on a meticulous investigation into the physiological regulation of adiponectin exocytosis, a crucial process governing the release of this beneficial hormone, both in states of health and in the context of metabolic disease. To achieve this, a synergistic combination of advanced electrophysiological techniques, specifically membrane capacitance patch-clamp recordings, and precise biochemical measurements of short-term adiponectin secretion, typically observed over 30-minute incubation periods, was employed.
Our initial experiments revealed that stimulation with epinephrine, a key physiological catecholamine, or with CL 316,243 (CL), a selective β3-adrenergic receptor (AR) agonist, robustly enhanced adiponectin exocytosis and subsequent secretion. This effect was consistently observed in both widely used cultured 3T3-L1 adipocytes and in primary subcutaneous mouse adipocytes, demonstrating a broad applicability across different adipocyte models. Crucially, this stimulated adiponectin release was significantly inhibited by ESI-09, a specific antagonist of Epac (Exchange Protein directly Activated by cAMP), strongly implicating the Epac pathway in this adrenergic-mediated response. Further characterization of adrenergic receptors in these cells showed that the β3AR was expressed at high levels in both cultured and primary adipocytes, while other adrenergic receptor subtypes were detected at considerably lower levels. Complementing this, molecular analysis confirmed the expression of Epac1 in both 3T3-L1 and primary adipocytes, whereas Epac2, another isoform, was found to be undetectable, further narrowing the focus to Epac1 as the key mediator.
A critical aspect of this study involved examining the impact of metabolic disease on adiponectin secretion. We found that adiponectin secretion could not be effectively stimulated by either epinephrine or CL in adipocytes specifically isolated from obese and type 2 diabetic mice. This indicated a profound impairment in adrenergic responsiveness in the diseased state. Interestingly, while stimulated secretion was abrogated, the basal (unstimulated) adiponectin release level from these diseased adipocytes was found to be significantly elevated, approximately twofold higher than that observed in healthy controls. Further molecular investigations revealed a crucial link to gene expression. The gene expression levels of both β3AR and Epac1 were markedly reduced in adipocytes obtained from obese animals. This reduction at the transcriptional level corresponded to a respective approximately 35% reduction in β3AR protein expression and an approximately 30% reduction in Epac1 protein expression, confirming translational consequences. To directly test the functional significance of these reduced expression levels, small interfering RNA (siRNA)-mediated knockdown experiments were performed. A targeted reduction of β3AR expression by approximately 60% and Epac1 expression by approximately 50% in healthy adipocytes was directly associated with the complete abrogation of catecholamine-stimulated adiponectin secretion, functionally mimicking the impaired response observed in diseased adipocytes.
In conclusion, based on these comprehensive findings, we propose a refined model wherein adiponectin exocytosis is physiologically stimulated via adrenergic signaling pathways that predominantly involve β3-adrenergic receptors and are critically mediated by the Epac1 pathway. Furthermore, our study strongly suggests that adrenergically stimulated adiponectin secretion is significantly disturbed and impaired in conditions of obesity and type 2 diabetes. This disturbance is a direct consequence of the markedly reduced expression of both β3ARs and Epac1 within adipocytes in the diseased state, a condition that we herein define as “catecholamine resistance.” This novel understanding of adiponectin dysregulation in metabolic disease provides crucial insights for potential therapeutic interventions.
Introduction
Adiponectin levels are consistently found to be reduced in individuals with type 2 diabetes, a metabolic disorder characterized by insulin resistance and hyperglycemia. Conversely, high circulating levels of adiponectin have been demonstrably shown to confer significant protection against the development of this complex disease, highlighting its beneficial metabolic roles. The intricate control mechanisms governing adiponectin release from adipocytes, the primary source of this hormone, have predominantly been investigated under longer-term experimental conditions, typically spanning several hours. Shorter-term regulation, specifically within periods of 30 to 60 minutes, has received considerably less attention, with only a limited number of studies exploring this rapid dynamic. These few investigations have indicated that adiponectin release can be acutely induced by insulin and involves a crucial calcium-dependent component.
Moreover, our own recent groundbreaking work has provided further insights, demonstrating that white adipocyte exocytosis, which directly corresponds to adiponectin secretion, is robustly stimulated via an elevation of intracellular cyclic adenosine monophosphate (cAMP) levels. This stimulation is mediated through the activation of Epac (Exchange Protein directly Activated by cAMP), which are a family of cAMP sensors. Epac proteins uniquely couple cAMP production to protein kinase A (PKA)-independent signaling pathways in a wide variety of cell types, providing an alternative route for cAMP-mediated cellular responses. Our findings specifically showed that this cAMP-stimulated adiponectin exocytosis is independent of calcium, but remarkably, it can be potently augmented by a synergistic combination of calcium and ATP. Given the central and established roles of both cAMP and calcium in the precise regulation of short-term adiponectin secretion, it is highly conceivable and biologically logical that adrenergic signaling, a major physiological regulatory system, actively participates in the intricate control of adiponectin exocytosis. The pivotal role of adrenergic signaling in regulating white adipocyte lipolysis, the breakdown of fats, is already well-established. Catecholamine binding to adrenergic receptors (ARs) leads to complex intracellular signaling. Activation of α1 receptors typically elevates cytoplasmic calcium levels, while activation of β1, β2, and β3 receptors, particularly β3ARs in adipocytes, primarily increases intracellular cAMP production. Conversely, the activation of α2ARs results in a decrease in cAMP production. Consequently, a dynamic and finely tuned functional balance involving these different adrenergic receptor subtypes can be readily envisaged to profoundly influence the intricate process of adiponectin exocytosis, orchestrating its release in response to physiological cues.
In this comprehensive study, we embarked on a detailed investigation into the effects of adrenaline, a key physiological catecholamine, and CL 316,243 (CL), a specific β3 adrenergic agonist, on white adipocyte adiponectin exocytosis. To achieve a multi-faceted understanding, we employed a synergistic combination of advanced biophysical techniques, such as membrane capacitance measurements, and biochemical methods for quantifying adiponectin secretion. Our investigations were conducted in vitro, utilizing cultured 3T3-L1 adipocytes, which serve as a commonly employed and well-characterized adipocyte model. Furthermore, we extended our studies *ex vivo* by using primary subcutaneous adipocytes meticulously isolated from both lean and obese/type 2 diabetic mice, providing crucial physiological relevance. Our findings represent a significant advancement, demonstrating for the first time that adiponectin exocytosis is indeed stimulated via adrenergic signaling pathways, with a predominant involvement of β3ARs. Moreover, we further conclusively demonstrate that adrenergically stimulated adiponectin secretion is significantly disturbed and impaired in adipocytes isolated from mice fed a high-fat diet (HFD-fed mice), a model for diet-induced obesity and type 2 diabetes. This disturbance is attributed to a crucial reduction in the expression of both β3ARs and Epac1, a condition that we herein formally denote as “catecholamine resistance,” providing a novel mechanistic insight into adiponectin dysregulation in metabolic disease.
Research Design and Methods
Cell and Animal Work
For the cell-based experiments, 3T3-L1 cells, obtained from ZenBio, were meticulously maintained and differentiated into mature adipocytes following previously established protocols. Studies were specifically conducted using mature 3T3-L1 adipocytes, typically 8 to 9 days after the initiation of differentiation, to ensure a fully differentiated phenotype. All cell culture reagents used throughout the study were procured from reputable suppliers, either Life Technologies or Sigma-Aldrich, guaranteeing high quality and consistency.
For the animal-derived studies, inguinal white adipose tissue (IWAT) was precisely isolated from 8-week-old male C57BL/6J mice. The tissue isolation was performed in Hank’s Balanced Salt Solution supplemented with 2% Bovine Serum Albumin (BSA) to maintain tissue viability. Tissue samples were then meticulously minced and subsequently subjected to enzymatic digestion using Collagenase type II (at a concentration of 1 mg/mL) for a duration of 45 to 60 minutes at 37 degrees Celsius, facilitating the dissociation of adipocytes from the tissue matrix. Following digestion, the suspension was carefully poured through a 100 µm nylon mesh to separate adipocytes from undigested tissue and debris. The floating adipocytes were then thoroughly washed with Krebs-Ringer glucose buffer containing 1% BSA. These isolated primary adipocytes were either immediately utilized for acute experiments or snap-frozen in liquid nitrogen and stored at -80 degrees Celsius for future use.
To model diet-induced obesity and type 2 diabetes, 5-week-old male mice were divided into two dietary groups: one group was fed a regular chow diet (Global Diet #2016, Harlan-Teklad), serving as the lean control, while the other group was fed a high-fat diet (D12492, Research Diets Inc.), providing 60% of its calories from fat, for a period of 8 weeks. All animal work procedures, including housing, feeding, and tissue collection, were conducted under strict ethical guidelines and received explicit approval from the Regional Ethical Review Board, ensuring animal welfare and compliance.
Electrophysiology and [Ca2+]i Imaging
For electrophysiological measurements, 3T3-L1 adipocytes were cultured in either plastic (Sarstedt) or glass (IBL) 35mm Petri dishes, depending on the specific experimental requirement. During the experiments, cells were continuously superfused with an extracellular solution (EC) precisely formulated to mimic physiological conditions. This EC contained (in mM): 140 NaCl, 3.6 KCl, 2 NaHCO3, 0.5 NaH2PO4, 0.5 MgSO4, 5 HEPES (adjusted to pH 7.4 with NaOH), 2.6 CaCl2, and 5 glucose. Exocytosis, specifically measured as increases in membrane capacitance, was performed using the standard whole-cell patch-clamp configuration, with cells clamped at a holding potential of -70 mV, as previously described. The intracellular pipette-filling solutions were carefully composed to control the internal milieu of the cells. These solutions consisted of (in mM): 125 K-glutamate, 10 KCl, 10 NaCl, 1 MgCl2, and 3 Mg-ATP, with 5 Hepes (adjusted to pH 7.15 with KOH). Various internal solutions were further supplemented to achieve specific intracellular conditions: IC1 included 0.1 mM cAMP and 10 mM EGTA (a calcium chelator); IC2 contained only 10 mM EGTA; IC3 had 0.05 mM EGTA; and IC4 contained 9 mM CaCl2 along with 10 mM EGTA, allowing for precise control of intracellular calcium.
Intracellular Ca2+ concentrations ([Ca2+]i) were quantitatively recorded using dual-wavelength ratio imaging in cells that had been loaded with Fura-2 AM (Life Technologies), a fluorescent calcium indicator, as previously described. The excitation wavelengths employed were 340 and 380 nm, and the emitted light was meticulously collected at 510 nm. The absolute intracellular calcium concentration was calculated using a standard equation, employing a Kd value of 224 nM for Fura-2. All electrophysiological and imaging measurements were consistently carried out at a controlled temperature of 32 degrees Celsius, optimizing cellular responses.
Adiponectin Secretion in Cultured and Primary Adipocytes
Adiponectin secretion was quantitatively measured in both cultured 3T3-L1 adipocytes, grown on 12-well plates (Sarstedt), and in mouse primary adipocytes. These measurements were conducted during short-term, 30-minute incubation periods at a controlled temperature of 32 degrees Celsius, allowing for the assessment of rapid release dynamics. For the 3T3-L1 adipocytes, cells were pre-incubated in a glucose-depleted extracellular solution (EC), and, as indicated for specific experiments, supplemented with the membrane-permeable Ca2+ chelator BAPTA-AM (Life Technologies) for 30 minutes to reduce intracellular calcium. Primary adipocytes were diluted to a 25% V/V concentration in their incubation medium. Secretion was measured in EC containing the test substances as indicated in the experimental design. Primary adipocyte incubations were meticulously terminated by centrifugation through a layer of diisononyl phthalate (Sigma-Aldrich), which separates cells from the aqueous medium. Following centrifugation, the tubes were carefully cut through the oil layer at two precise points, effectively separating the cell pellet from the supernatant media. Cells were then lysed in PBS containing 2% SDS and a protease inhibitor cocktail (1 tablet/10 ml; cOmplete™ Mini; Roche Diagnostics) to prevent protein degradation. Both EC aliquots (containing secreted adiponectin) and cell homogenates (for total protein normalization) were snap-frozen and stored at -80 degrees Celsius. Secreted adiponectin levels were quantified using mouse ELISA DuoSets (R&D Systems) and expressed in relation to total protein content, determined by the Bradford protein assay, providing a standardized measure of secretion.
Quantitative Real-Time RT-PCR
For quantitative real-time RT-PCR (qRT-PCR) analysis, RNA was extracted and purified using QIAzol (Qiagen) and the ReliaPrep RNA Cell Miniprep System (Promega), ensuring high-quality RNA isolation. The total RNA was then reverse transcribed into complementary DNA (cDNA) using the QuantiTect Reverse Transcription Kit (Qiagen), which allows for efficient synthesis of cDNA from RNA templates. SYBR Select Master Mix (Life Technologies) was employed for the quantitative RT-PCR reactions, a method that uses a fluorescent dye to detect and quantify amplified DNA. Gene expression levels were meticulously normalized against the expression of beta-actin (*Actb*), a commonly used housekeeping gene, using the relative ∆Ct method. Primers for the target genes were used at a concentration of 500 nM, and the PCR efficiencies for each primer pair were precisely determined from the slope of their respective standard curves, ensuring accurate quantification of gene expression.
siRNA Transfection
To investigate the functional roles of specific genes, small interfering RNA (siRNA) transfection was performed. Cells were transfected with Silencer Select siRNA (Ambion) on differentiation day 6. The transfection was carried out in Opti-MEM medium, which is devoid of antibiotics, utilizing Lipofectamine 2000 (Life Technologies) as the transfection reagent. The specific siRNAs used had IDs s62085 (targeting *Adrb3*) and s104656 (targeting *Rapgef3*, which encodes Epac1). A concentration of 80 nM siRNA was employed for transfection. Eight hours after the transfection, the medium was carefully changed to antibiotic-free DMEM supplemented with 10% FBS, allowing cells to recover. Experiments were conducted 60 hours post-transfection, providing sufficient time for gene knockdown. The efficiency of gene knockdown was rigorously validated by qRT-PCR, confirming reduced mRNA levels of the target genes. Following successful knockdown, adiponectin secretion was measured as described previously, allowing for the assessment of the functional consequences of reduced β3AR and Epac1 expression.
Serum Glucose, Insulin, and Adiponectin Levels
Blood samples were collected at the termination of the animal experiments, specifically from the axillary vessels (subclavian artery and vein), ensuring a consistent method of collection. Serum glucose concentrations were precisely measured using a glucose oxidase-peroxidase enzyme assay (No. P7119; Sigma-Aldrich) in combination with o-Dianisidine dihydrochloride (No. F5803; Sigma-Aldrich), providing accurate quantification of blood sugar levels. The insulin concentration in the serum was quantitatively analyzed using a dedicated ELISA Mouse insulin kit (No 10-1247-01; Mercodia), allowing for the assessment of pancreatic beta cell function. Total adiponectin and high-molecular-weight (HMW) adiponectin levels, representing different biologically active forms of the hormone, were both measured by ELISA (EZMADP-60K; EMD Millipore and MBS028367; MyBiosource, respectively), providing a comprehensive profile of circulating adiponectin.
Lipolysis and Measurements of cAMP Content
To assess the rate of lipolysis, glycerol release into the culture medium was quantitatively measured using a free glycerol assay (G7793 and F6428; Sigma-Aldrich), strictly adhering to the manufacturer’s instructions. Glycerol is a product of triglyceride breakdown and serves as a reliable indicator of lipolytic activity. Intracellular cAMP levels, a crucial second messenger in adrenergic signaling, were determined in cell homogenates using the Cyclic AMP XP Assay Kit (No. 4339; Cell Signaling), providing a measure of adenylyl cyclase activity.
Immunocytochemistry
Immunocytochemistry experiments were meticulously performed to visualize the expression and localization of specific proteins. Incubations were carried out using an antibody diluent solution, precisely composed of PBS containing 0.1% Saponin and 5% donkey serum, to minimize non-specific binding. The primary antibody used was an Anti-beta 3 Adrenergic Receptor antibody (ab94506, Abcam), and the secondary antibody was AlexaFluor 488 (#711-545-152; Jackson ImmunoResearch Labs), which provides a fluorescent signal for visualization. Prior to antibody incubation, adipocytes were fixed in PBS containing 4% paraformaldehyde (PFA) for 6 minutes to preserve cellular structures. Subsequently, cells were incubated for 60 minutes with PBS containing 0.1 M glycine (Sigma-Aldrich) to quench any remaining aldehyde groups. Cells were then incubated with the primary antibody at a dilution of 1:500 overnight, followed by incubation with the secondary antibody at a dilution of 1:500 for 1 hour. A negative control was included, incubated with only the secondary antibody, to account for non-specific fluorescence.
Image acquisition was performed using a total internal reflection fluorescence (TIRF) Observer Z1 microscope equipped with an alpha Plan-Apochromat 100x/1.46 Oil objective (both from Zeiss) and an EMCCD camera (Evolve 512 delta, Photometrics), enabling high-resolution imaging of membrane-proximal events. Images were acquired using Zen blue 2012 software (Zeiss). The excitation wavelength was 488nm, and emission was collected between 500-540nm.
Data analysis
The rate of membrane capacitance increase (∆C/∆t), a direct measure of exocytosis, was precisely quantified by applying linear fits to the recorded data, as previously described. ∆C/∆t was determined both at its maximal rate after substance application (∆C/∆tmax) and at indicated later time-points, providing kinetic profiles of vesicle release. The statistical significance of variance between two means was rigorously calculated using OriginPro software (OriginLab Corporation) and Student’s t-test, applied as paired or unpaired as appropriate. One-way ANOVA was utilized when comparisons involved more than two groups. The free intracellular calcium concentration ([Ca2+]) was calculated using the MAXCHELATOR software.
TIRF images, captured from individual cells, were meticulously analyzed using ImageJ software (National Institutes of Health). Fluorescence intensity was determined by defining a circular region-of-interest (ROI) encompassing the entire cell and then normalized to the area of the defined region, providing a standardized measure of signal. These ROIs were further corrected by subtracting the average fluorescence value of a background ROI defined outside the cell. Relative image quantifications were performed to compare raw fluorescence intensity between cells from chow-fed mice, high-fat diet (HFD-fed) mice, and negative control groups. To ensure comparability, identical acquisition parameters, including exposure time and gain settings, were consistently used across all images. Images from chow and HFD cells were selected randomly to avoid bias. All fits, plots, and statistical analyses (Student’s t-test) were performed using Origin Pro software. All quantitative data are consistently presented as mean values ± SEM for the designated number of independent experiments, ensuring clarity and transparency in reporting.
Results
Adrenergic stimulation of short-term adiponectin secretion in cultured adipocytes
Our prior research, which established that an elevation in intracellular cAMP levels triggers white adipocyte exocytosis and subsequent adiponectin secretion in both 3T3-L1 adipocytes and human primary subcutaneous white adipocytes, strongly suggested that short-term adrenergic stimulation could act as a crucial physiological regulator of adiponectin exocytosis. To rigorously test this hypothesis, 3T3-L1 adipocytes were meticulously incubated for 30 minutes with either 5 µM adrenaline (epinephrine), a key physiological catecholamine, or 1 µM of CL 316,243 (CL), a selective β3-adrenergic receptor (AR) agonist. As depicted, both adrenaline and CL significantly stimulated adiponectin secretion by approximately 1.8-fold. To further explore the potential involvement of intracellular calcium, which can be elevated via the activation of α1ARs, 3T3-L1 adipocytes were pre-incubated with the Ca2+ chelator BAPTA-AM prior to adrenergic stimulation. Interestingly, adrenaline/CL-stimulated adiponectin secretion remained intact in BAPTA-treated adipocytes, indicating that this particular calcium-dependent pathway was not essential for the acute adrenergic stimulation of adiponectin release. In fact, CL-induced secretion was observed to be slightly elevated in Ca2+-depleted cells compared to non-treated adipocytes, suggesting that intracellular calcium might, under certain circumstances, exert a subtle modulatory role rather than a direct stimulatory one in this specific context.
To confirm the presence of key components within the hypothesized adrenergically stimulated adiponectin exocytosis pathway, we systematically investigated the expression profiles of adrenergic receptors (ARs), Epac (Exchange Protein directly Activated by cAMP), and adiponectin in 3T3-L1 adipocytes. Our analysis revealed that β3ARs (*Adrb3*) were readily and abundantly expressed, whereas β1 (*Adrb1*), β2 (*Adrb2*), and α1D (*Adra1d*) receptors were expressed at considerably lower levels, clearly pointing to β3ARs as the predominant adrenergic receptor subtype involved in this process in 3T3-L1 adipocytes. There are two known isoforms of Epac: Epac1 (*Rapgef3*) and Epac2 (*Rapgef4*). While the expression of Epac2 is typically restricted to neuroendocrine cell types, Epac1 is known to be more generically expressed across a wider range of cell types. Consistent with findings in human preadipocytes, Epac1 was found to be abundantly expressed in undifferentiated 3T3-L1 cells and, although its expression was reduced during differentiation, it remained readily detectable in mature adipocytes. In agreement with previous observations, Epac2 was not expressed in these cells. As anticipated, adiponectin (*Adipoq*) was completely absent in undifferentiated cells but became highly expressed in fully differentiated 3T3-L1 adipocytes, confirming their capacity for adiponectin synthesis.
Furthermore, CL still stimulated adiponectin release by approximately 2-fold even in 3T3-L1 adipocytes that had been pre-treated with the protein synthesis inhibitor cycloheximide (10 µg/mL). This crucial finding supports the concept that short-term adiponectin secretion occurs primarily due to the release of pre-stored adiponectin-containing vesicles, rather than requiring *de novo* protein synthesis. Moreover, adiponectin gene expression itself remained unaffected by CL exposure, further reinforcing that the observed acute secretion was due to exocytosis of pre-existing stores, not increased synthesis.
Epac-dependent adrenergic stimulation of adipocyte exocytosis
To gain a detailed and real-time understanding of adrenergic effects on adiponectin exocytosis, we precisely measured vesicle release as increases in membrane capacitance in 3T3-L1 adipocytes using patch-clamp electrophysiology. Adrenaline or CL was applied extracellularly to the continuously superfused cell dish during the recordings, enabling online registrations of the dynamic release events. We have previously demonstrated that membrane capacitance increases in 3T3-L1 adipocytes predominantly represent the release of adiponectin-containing vesicles. Consistent with this, exocytosis was robustly triggered by a pipette-filling solution (IC1) that contained cAMP but was depleted of Ca2+. We then infused cells with the same solution but lacking cAMP (IC2). In accordance with our own previous results, exocytosis was not triggered by this cAMP-depleted pipette solution, confirming the cAMP dependence. However, extracellular addition of adrenaline or CL during the recording, even with the cAMP-depleted intracellular solution, stimulated exocytosis at a maximal rate similar to the ∆C/∆t induced by intracellular cAMP, highlighting the extracellular adrenergic effect.
Our previous research has firmly established that cAMP stimulates adiponectin exocytosis primarily via the activation of Epac. To rigorously investigate the critical role of this cAMP-binding protein in adrenergically stimulated 3T3-L1 adipocyte exocytosis, we pre-treated cells with ESI-09, a novel and specific Epac inhibitor. Guided by studies on Epac function in pancreatic β-cell insulin secretion, we utilized a concentration of 10 µM ESI-09 during 30-minute incubations. The compelling result was that adrenaline-stimulated exocytosis was completely abolished in ESI-09 pre-treated cells, and this inhibition was consistent across all investigated time points. This definitively demonstrated that adrenergically stimulated adiponectin exocytosis is unequivocally Epac-dependent, positioning Epac as a central mediator in this physiological process.
The Role of cAMP and Cytoplasmic Ca2+ in Adrenergically Stimulated Exocytosis
We undertook a detailed investigation of intracellular cAMP content in 3T3-L1 adipocytes exposed to adrenaline or CL 316,243 (CL) for a period of 30 minutes. As comprehensively illustrated, adrenaline treatment led to a significant elevation of cAMP levels, exceeding a 7.5-fold increase. This magnitude of increase was remarkably similar to that produced by a combination of the adenylyl cyclase activator forskolin (10 µM) and the phosphodiesterase inhibitor IBMX (200 µM; commonly denoted as FSK/IBMX), which is known to potently elevate intracellular cAMP. Intriguingly, the adrenaline-induced elevation of cAMP was largely attenuated, by approximately 60%, when intracellular calcium was chelated, suggesting a calcium-dependent component in this process. In contrast, CL elevated cytoplasmic cAMP to a similar extent as adrenaline in BAPTA pre-treated (calcium-chelated) cells, and importantly, this increase remained unaffected by calcium buffering. Our collective results strongly indicate the involvement of calcium-dependent adenylyl cyclases (ACs), which are enzymes responsible for converting ATP to cAMP, in the adrenaline-induced production of cAMP within adipocytes.
We further investigated the direct effect of adrenaline on adipocyte intracellular Ca2+ concentrations ([Ca2+]i). Adrenaline (5 µM) was observed to elevate [Ca2+]i in approximately 60% of the investigated cells, presumably through the activation of α1-adrenergic receptors (α1ARs). The pattern of this [Ca2+]i response exhibited heterogeneity, even among cells within the same culture dish. It was characterized by distinct response types: high-amplitude oscillations, representative of 20% of responsive cells with a peak amplitude of 275 ± 15 nM; a single, transient peak, observed in 15% of responsive cells; or a slow, sustained increase in [Ca2+]i, occurring over several minutes, observed in 65% of responsive cells, with an increase from a basal 104 ± 3 nM to a stable 156 ± 4 nM. This observed heterogeneous [Ca2+]i response pattern is comparable to results previously reported in human adipocytes exposed to noradrenaline, and it has been linked to an irregular inter-cellular distribution of adrenergic receptor subtypes. As anticipated, CL (1 µM), being a selective β3AR agonist, had no discernible effect on [Ca2+]i, confirming its specific receptor target.
To precisely elucidate the contribution of [Ca2+]i to adrenaline-stimulated exocytosis, cells were infused with a cAMP-depleted pipette solution containing a low concentration of EGTA (50 µM) to allow for subtle [Ca2+]i fluctuations. In agreement with our secretion data, adrenaline-triggered exocytosis was not significantly augmented under these conditions. In fact, the maximal rate of membrane capacitance increase (∆C/∆tmax) was, if anything, slightly reduced compared to conditions where Ca2+ was completely chelated by 10 mM EGTA (~11 fF/s at 50 µM EGTA vs. ~19 fF/s with 10 mM EGTA; P=0.1). We next added CL to 3T3-L1 adipocytes infused with a cAMP-depleted solution containing approximately 1.5 µM free Ca2+. Consistent with previous findings, exocytosis was not triggered by this solution alone. However, the subsequent addition of CL potently stimulated exocytosis, with a ∆C/∆tmax that was significantly higher than the rate achieved with IC3 and showed a tendency towards significance when compared to the CL effect using Ca2+-free IC2. Adrenaline, when added to cells already infused with a cAMP-containing IC1 solution, had no additional effect on exocytosis; ∆C/∆t averaged 9.4 ± 2.5 fF/s before and 9.4 ± 2.4 fF/s one minute after adrenaline application, further confirming the predominant role of cAMP in adrenaline-stimulated exocytosis.
Collectively, our results unequivocally indicate that adrenergic stimulation primarily triggers adiponectin exocytosis via the activation of β3ARs, leading to a consequent elevation of cytoplasmic cAMP. While adrenergically stimulated adiponectin exocytosis can be augmented by the presence of intracellular Ca2+, the primary source of this augmenting Ca2+ appears to originate from Ca2+-generating pathways other than α1ARs, suggesting a complex interplay of signaling.
Adrenergic stimulation of adiponectin secretion in primary subcutaneous mouse adipocytes
To verify the physiological importance and generalizability of our findings obtained using cultured 3T3-L1 adipocytes, we extended our investigation to study adrenergically stimulated adiponectin secretion in primary mouse inguinal white adipose tissue (IWAT) adipocytes. In 30-minute incubations, both adrenaline (5 µM) and CL (1 µM) effectively stimulated adiponectin secretion by approximately 2-fold. This magnitude of stimulation was highly comparable to that observed in 3T3-L1 adipocytes, confirming the relevance of the *in vitro* model. Adrenaline/CL-stimulated adiponectin secretion remained unaffected by calcium chelation (P=0.3 vs. adrenaline alone and P=0.5 vs. CL alone), reinforcing the calcium-independent nature of the primary stimulatory pathway. Crucially, however, this stimulated secretion was completely abolished by pre-incubation with ESI-09, a specific Epac inhibitor. These findings collectively demonstrate that adiponectin exocytosis and secretion are regulated in a remarkably similar manner in both cultured 3T3-L1 adipocytes and in primary subcutaneous mouse adipocytes, underscoring the physiological significance of the observed mechanisms.
Impaired adiponectin secretion in IWAT adipocytes isolated from obese/type 2 diabetic mice
We conducted a critical investigation into short-term (30 minutes) adiponectin secretion in IWAT adipocytes isolated from mice that had been fed either a regular chow diet or a high-fat diet (HFD) for eight weeks. The HFD-fed mice developed pronounced obesity, evidenced by their significantly higher average weight (46.3 ± 0.8 g) compared to the chow-fed animals (31.8 ± 0.6 g; P<0.001). Furthermore, these HFD-fed mice also exhibited a diabetic phenotype, as indicated by elevated serum glucose and insulin levels, confirming the success of the diet-induced metabolic dysfunction model. While serum total adiponectin levels were found to be similar in both chow and HFD-fed mice, a more nuanced analysis of adiponectin forms was performed. Adiponectin is known to circulate in various forms, and reduced levels of specifically the high-molecular weight (HMW) adiponectin have been suggested to be associated with metabolic aberrations. Consistent with this, we found that levels of HMW adiponectin were approximately 50% lower in HFD-fed animals compared to chow-fed animals, highlighting a qualitative defect in adiponectin composition despite similar total levels.
Intriguingly, basal (unstimulated) adiponectin secretion was more than 2-fold elevated in adipocytes derived from HFD-fed mice when compared to adipocytes from chow-fed animals. In adipocytes from chow-fed mice, incubation with FSK/IBMX (forskolin/IBMX), which potently elevates cAMP, resulted in a 2.5-fold increase in adiponectin secretion, whereas both CL and adrenaline stimulated adiponectin release by approximately 2-fold. However, a stark difference emerged in HFD adipocytes. As shown, FSK/IBMX stimulated secretion in HFD adipocytes by only approximately 1.5-fold, which was a significantly reduced response compared to chow-fed adipocytes. More strikingly, adiponectin release triggered by adrenaline or CL was essentially abolished in adipocytes isolated from the obese mice, indicating a profound unresponsiveness to adrenergic stimulation.
Mechanisms underlying the blunted adiponectin secretion in adipocytes from obese/type 2 diabetic mice
To thoroughly investigate the intricate mechanisms underlying the blunted adiponectin secretion observed in response to adrenergic stimulation in adipocytes from obese/type 2 diabetic mice, we systematically examined the expression profiles of adrenergic receptors (ARs) and Epac (isoforms 1 and 2) in adipocytes from both chow- and high-fat diet (HFD)-fed mice. Similar to our findings in 3T3-L1 adipocytes, the β3ARs were amply expressed in adipocytes derived from chow-fed animals, whereas α1D, β1, and β2 adrenergic receptors were present at considerably lower levels. However, a critical observation was the pronounced downregulation of β3ARs, by approximately 5-fold, in HFD adipocytes, indicating a significant reduction in the primary adrenergic receptor responsible for adiponectin stimulation. Furthermore, the expression of α1DARs and β1ARs was also significantly reduced in these diseased adipocytes. Consistent with findings in 3T3-L1 adipocytes, Epac1 was confirmed as the predominant Epac isoform expressed in primary adipocytes, and its expression was notably downregulated by approximately 40% in adipocytes from HFD mice.
We further investigated whether a reduced cellular adiponectin content could account for the decreased stimulated secretion observed in HFD adipocytes. As shown, the total cellular adiponectin content was not significantly decreased in HFD adipocytes compared to chow-fed adipocytes, suggesting that the primary defect was not a lack of stored adiponectin. However, a detailed analysis of the percentage of released adiponectin demonstrated that HFD adipocytes secreted a larger fraction of their total adiponectin content under basal (unstimulated) conditions compared to chow-fed adipocytes. Moreover, while chow adipocytes secreted approximately 3% of their adiponectin content upon stimulation with FSK/IBMX, adrenaline, or CL, the fraction secreted by HFD adipocytes tended to be smaller upon adrenaline-stimulation and was markedly reduced, amounting to only approximately 0.5% in response to CL. These results strongly reinforce the notion that impaired regulated adiponectin exocytosis, rather than reduced total content, underlies the diminished adiponectin secretion observed in adipocytes from obese/type 2 diabetic mice. The approximately 3% adiponectin released in stimulated chow adipocytes is in reasonably good agreement with the fraction of insulin (2-2.5%) released during a 30-minute glucose-stimulation of pancreatic β-cells, indicating a similar efficiency in regulated exocytosis.
Measurements of intracellular cAMP levels revealed that basal cAMP levels were equivalent in both HFD and chow adipocytes. FSK/IBMX or adrenaline elevated cAMP to a similar extent in both chow and HFD adipocytes, suggesting that the initial cAMP generation pathways were largely intact (P=0.3 for FSK/IBMX and P=0.1 for adrenaline in chow vs. HFD adipocytes). However, the CL-induced cAMP elevation was significantly reduced in HFD adipocytes, indicating a specific impairment in the β3AR-mediated cAMP production pathway. To investigate the preservation of other cAMP signaling pathways, distinct from adiponectin secretion, we measured lipolysis. FSK/IBMX, adrenaline, or CL stimulated lipolysis in chow adipocytes. While FSK/IBMX remained capable of stimulating lipolysis in HFD adipocytes, the lipolytic response to adrenaline or CL was notably blunted (P=0.09 for adrenaline vs. control and P=0.1 for CL vs. control). However, adrenergically stimulated lipolysis appeared more intact than adiponectin secretion in HFD adipocytes, where adiponectin release was completely abrogated. This disparity may be explained by the fact that lipolysis is primarily stimulated via the activation of protein kinase A (PKA), and cAMP has been shown to exhibit a lower affinity for Epac1 (which mediates adiponectin release) than for PKA (which mediates lipolysis). Consequently, higher cAMP concentrations are expectedly required to maintain adrenergically stimulated adiponectin release, suggesting that the reduced cAMP production in HFD adipocytes specifically impairs the Epac-dependent pathway more severely.
To further confirm the reduced expression of β3ARs at the protein level, we performed immunocytochemistry labeling in both chow and HFD adipocytes. Quantitative TIRF (Total Internal Reflection Fluorescence) imaging unequivocally confirmed a significantly lower abundance of β3ARs in HFD adipocytes compared to chow adipocytes. Moreover, ELISA measurements provided quantitative evidence that Epac1 protein levels were also decreased by approximately 30% in HFD adipocytes. To definitively verify the crucial importance of β3ARs and Epac1 in catecholamine-stimulated adiponectin secretion, we carried out siRNA (small interfering RNA)-mediated knockdown experiments in 3T3-L1 adipocytes. In five different experimental series, where β3AR expression was effectively reduced by approximately 60%, catecholamine-stimulated secretion was completely blunted, whereas scramble siRNA-transfected cells (control for non-specific effects) remained fully responsive. Similarly, specific silencing of Epac1 expression, by approximately 50%, also resulted in the complete abrogation of adrenaline-stimulated adiponectin secretion. These knockdown experiments provide compelling functional evidence, directly demonstrating that reduced expression of β3ARs and Epac1 is a direct cause of impaired catecholamine-stimulated adiponectin secretion.
Discussion
In this comprehensive study, our primary aim was to precisely define the physiological regulation of adiponectin exocytosis and to elucidate how this intricate regulatory process may be perturbed in the prevalent conditions of obesity and type 2 diabetes. The compelling fact that adiponectin exocytosis is triggered by cAMP highlights the existence of an unconventional physiological regulatory mechanism for adiponectin secretion. This mechanism appears quite distinct from the well-characterized pathways controlling archetypical calcium-stimulated hormone exocytosis, which typically rely on a direct calcium influx or release to trigger vesicle fusion. Below, we discuss the most important and transformative findings of our study, along with their profound physiological and pathophysiological implications.
Adiponectin vesicle exocytosis is stimulated via β3ARs and activation of Epac1
Our extensive studies consistently demonstrate that adiponectin exocytosis and short-term secretion are robustly stimulated by the catecholamine adrenaline, as well as by the selective β3AR agonist CL 316,243. Crucially, this stimulation is definitively dependent on Epac (Exchange Protein directly Activated by cAMP). These findings are entirely consistent with our previous detailed characterization of a novel cAMP-stimulated adiponectin exocytosis pathway. Furthermore, comprehensive gene expression analysis rigorously identified Epac1 as the specific Epac isoform predominantly present in both cultured 3T3-L1 adipocytes and in primary subcutaneous mouse adipocytes, thereby pinpointing the key intracellular mediator. Based on these convergent lines of evidence, we propose a refined model wherein catecholamines trigger the release of adiponectin-containing vesicles. Our model explicitly suggests that catecholamine binding to β3ARs, leading to a subsequent elevation of cytoplasmic cAMP and the direct activation of Epac1, constitutes the primary and main regulatory pathway for acute adiponectin exocytosis.
The secretory response specifically triggered via β3ARs appears to be largely unaffected by concomitant signaling through α1ARs. This is evidenced by the observation that adrenaline-stimulated secretion remains remarkably intact even in the presence of calcium chelation. Furthermore, adrenaline is unable to stimulate exocytosis at a higher rate when intracellular calcium fluctuations are permitted, suggesting that calcium, while potentially playing a modulatory role, is not the primary or rate-limiting factor for this rapid adrenergic stimulation. Our precise calcium measurements consistently show that the effects of adrenaline on intracellular calcium concentrations are relatively small, observed in only about 60% of responsive cells with a maximal calcium peak of 275 nM. This finding is entirely consistent with the low expression levels of α1ARs observed in our adipocyte models. The observation that calcium chelation nonetheless decreases adrenaline-induced cAMP levels may initially appear conflicting, but this merely indicates that calcium-dependent adenylyl cyclases (ACs), enzymes that produce cAMP, are indeed involved in adipocyte cAMP production. It is well-established that cAMP signaling is highly compartmentalized within cells due to the targeted localization of essential signaling proteins, such as ACs and phosphodiesterases (which break down cAMP). Consequently, cAMP can differentially regulate diverse cellular processes within discrete intracellular domains. Therefore, the calcium-dependent cAMP production might be crucial for other specific signaling processes within the adipocyte; for instance, calcium has been proposed to influence lipolysis. However, the calcium-mediated augmentation of adiponectin exocytosis is clear, suggesting that other pathways leading to increased intracellular calcium must exist. Our own preliminary observations strongly indicate the occurrence of store-operated Ca2+ channels in white adipocytes, which are plasma membrane-bound ion channels known to function within cAMP-generating microdomains. Additionally, several studies have suggested the existence of voltage-dependent Ca2+ (Cav) channels in white adipocytes. However, functional evidence for the precise presence and role of Cav channels in adipocytes is currently unavailable, warranting further investigation.
Catecholamine resistance in obesity and diabetes leads to defect adiponectin exocytosis
Our study provides compelling evidence demonstrating a significant disruption of adrenergically stimulated adiponectin secretion in adipocytes meticulously isolated from obese/type 2 diabetic mice. As comprehensively summarized in the corresponding model, we propose that adipocytes in an obese and metabolically compromised state are fundamentally unable to respond to catecholamine stimulation with adequate and potent adiponectin release. This critical impairment is primarily attributed to a markedly low abundance of β3ARs on the adipocyte surface, diminishing their capacity to sense adrenergic signals. Furthermore, the disturbance in intracellular post-receptor signaling is further compounded by decreased levels of Epac1, a crucial downstream mediator in the cAMP pathway. We therefore propose that this obesity-induced reduction in both β3ARs and Epac1 collectively results in blunted catecholamine-stimulated adiponectin secretion, a condition that can be appropriately referred to as “catecholamine resistance.” This concept of catecholamine resistance, leading to impaired lipolytic noradrenaline sensitivity, has been previously described in human obesity and attributed to a low density of β2ARs. Moreover, a recent study has further demonstrated that obesity-induced adipose tissue inflammation can contribute to catecholamine resistance by reducing cAMP production, a process linked to the induction of non-canonical IκB kinases.
The magnitude of FSK/IBMX-stimulated adiponectin secretion was observed to be reduced by a similar magnitude as the decrease in Epac1 protein levels (approximately 30%), strongly suggesting that decreased Epac1 expression represents the principal post-receptor disruption affecting adiponectin exocytosis. It may appear surprising that a quantified approximately 35% reduction of β3ARs in HFD adipocytes resulted in a completely abrogated adiponectin secretion. However, this level of decrease may indeed be sufficient to profoundly disturb the function of the highly organized signaling microdomains critically involved in adiponectin exocytosis. Furthermore, the concurrent lower Epac1 level in these cells may significantly aggravate this disturbance, creating a synergistic impairment. Moreover, our functional knockdown experiments provide direct evidence: a mere 60% knockdown of β3ARs alone was clearly sufficient to entirely abolish catecholamine-stimulated adiponectin secretion, unequivocally demonstrating the critical role of these receptors.
The obesity-induced perturbation of adiponectin exocytosis appears not to result from a reduced cellular adiponectin content, but rather arises from a fundamental defect in cAMP/catecholamine-stimulated adiponectin exocytosis, as evidenced by the significantly smaller proportion of adiponectin secreted in stimulated HFD adipocytes. It is expected that a very large reduction of adiponectin cell content would be required to physically affect the exocytosis of the adipokine. For instance, lipoglucotoxicity is known to decrease β-cell insulin content by 75% without affecting the number of hormone-containing releasable vesicles; in that case, the secretory defect is instead due to disturbances in the insulin vesicle release process itself. The precise reason for the observed higher basal adiponectin release in HFD adipocytes is currently unclear but could potentially represent a compensatory mechanism attempting to offset the disrupted stimulated adiponectin exocytosis. It remains to be definitively determined whether different adiponectin forms (high, medium, and low molecular weight) are differentially secreted under basal and cAMP/catecholamine-stimulated conditions, an area for future research.
Our study strongly suggests that a fundamental secretory defect underlies the reduced capacity of catecholamine-resistant adipocytes to respond with adequate adiponectin release. Moreover, it can be envisaged that decreased catecholamine levels, which are often observed in patients with type 2 diabetes, could further exacerbate this impaired adiponectin secretion, creating a vicious cycle that contributes to the progression of metabolic dysfunction.
Pathophysiological implications and future directions
The intricate involvement of adrenergic signaling in regulating the metabolic function of white adipose tissue is a concept that is already well-established in scientific literature. However, our study represents a significant advancement, as it is the first to conclusively demonstrate that adiponectin exocytosis, the process of its release, is directly triggered via specific adrenergic signaling pathways. Considering that it has now been two decades since adiponectin was initially discovered, it is indeed surprising that the precise physiological regulation of adiponectin exocytosis has not been more extensively investigated and fully elucidated.
Several previous studies have consistently shown that long-term activation of adrenergic receptors, sustained over numerous hours or even days, leads to a subsequent reduction in the expression and/or secretion of the adipokine. We propose that this observed long-term detrimental effect of catecholamines on adiponectin secretion is primarily attributable to a gradual depletion of adiponectin-containing vesicles within the adipocytes. This mechanism is analogous to how pancreatic beta-cells, when exposed to extended periods of stimulation for insulin release, eventually experience exhaustion of their secretory capacity. Given the essential and profound metabolic effects attributed to adiponectin, this adipokine unequivocally stands as a promising and highly attractive target in contemporary drug discovery efforts aimed at combating metabolic diseases. The reported half-life of adiponectin, approximately 75 minutes, indicates that secreted adiponectin is capable of exerting its beneficial effects not only on nearby organs (paracrine action) but also systemically throughout the body (endocrine action). Adiponectin is indeed an unconventional hormone in several respects, as it is secreted from a ubiquitously distributed organ (adipose tissue) and its circulating levels are incessantly high. It is clear that much fundamental knowledge remains to be defined regarding the precise regulation of adiponectin release. This includes a deeper understanding of the distinct roles played by different adipose tissue depots in regulated adiponectin secretion, as well as the intricate relationship between visceral versus subcutaneous adiposity and differential adiponectin secretion patterns. Nonetheless, the compelling findings presented in this study constitute one crucial and important piece of the complex puzzle. This puzzle must be fully assembled and understood for the successful future pharmacological adjustment of adiponectin secretory defects, ultimately paving the way for more effective treatments for metabolic diseases.
Competing interests
None of the authors involved in this study have any conflicts of interest to declare. They solely bear responsibility for the content and writing of this paper, ensuring impartiality and scientific integrity.
Author contributions
The conception and design of the experiments were collaboratively undertaken by A.M.K., S.M., I.W.A., and C.S.O. The meticulous collection, analysis, and interpretation of the data were performed by A.M.K., S.M., E.P., M.F.E., A.A., M.J., and C.S.O. The initial drafting of the manuscript was carried out by C.S.O. The manuscript then underwent thorough drafting and revision by A.M.K., S.M., E.P., M.F.E., A.A., M.J., I.W.A., and C.S.O., ensuring accuracy and clarity. All authors meticulously read and approved the final version of the manuscript, signifying their complete endorsement of the work. All experiments were conducted at the Department of Physiology/Metabolic Physiology, Gothenburg University. C.S.O. serves as the guarantor of this work, and as such, had full access to all the data generated in the study and assumes complete responsibility for the integrity of the data and the accuracy of its analysis.
Funding
This study received essential financial support from several prestigious organizations. Funding was provided by the Åke Wiberg Foundation (awarded to C.S.O. and I.W.A.), the Swedish Diabetes Foundation (grant IDs: DIA2013-070, DIA2014-074, and DIA2015-062), the Novo Nordisk Foundation Excellence project grant (awarded to I.W.A.), the Diabetes Wellness Research Foundation (grant ID: 8349/2014SW), and the Swedish Medical Research Council (grant IDs: 2010-2656, 2012-2994, 2012-1601, and 2013-7107).
Acknowledgements
We express our sincere gratitude to Birgit Linder and Ann-Marie Alborn from the Department of Physiology/Metabolic Physiology for their invaluable assistance with adiponectin release measurements, adipocyte isolation procedures, and cell culturing, which were essential for the successful completion of this study.