Nicotinic Acetylcholine Receptor Signaling Activates Beige Adipocytes and Mediates Systemic Metabolism

Abstract

Optimal function of beige adipocytes is essential for energy balance and metabolic homeostasis. We previously identified that cholinergic receptor nicotinic α2 subunit (CHRNA2) mediates a beige fat selective signaling in mice and humans. Here, we investigate the molecular composition of CHRNA2-containing nicotinic acetylcholine receptors (nAChRs) in beige adipocytes and its impact on whole-body metabolism. Expression levels of cholinergic receptor nicotinic β2 subunit (CHRNB2) and CHRNA2 positively correlate within murine and human beige adipocytes, and both nAChRs are regulated by a spectrum of beige fat regulators. CHRNB2 is essential for the response to nAChR agonists in beige adipocytes. CHRNB2 partial agonists, a family of drugs clinically used for smoking cessation, activate both murine and human beige adipocytes. Mice deficient in Chrnb2, with both whole-body knockout or adipocyte-specific deletion, exhibit compromised adaptive thermogenesis in subcutaneous fat and more aggravated metabolic dysfunction after challenge with high-fat-diet feeding compared with control mice, underscoring the importance of the nAChR signaling in maintaining energy balance. This cholinergic signaling declines in subcutaneous fat with aging. These findings indicate CHRNB2 forms a functional receptor with CHRNA2 in beige adipocytes and highlight their potential as therapeutic targets against metabolic disorders.

Article Highlights

• Beige adipocyte activity, mediated by CHRNA2, significantly influences adipose function and systemic metabolism.

• The CHRNB2 subunit forms a functional receptor with CHRNA2 and is essential for the response to nicotinic acetylcholine receptor agonists in beige adipocytes.

• Deletion of Chrnb2 in mice compromises the adaptive response to cold in subcutaneous adipose tissue and renders exacerbated metabolic dysfunction due to diet-induced obesity.

• This cholinergic signaling within subcutaneous adipose tissue declines with aging. CHRNB2 partial agonists, a family of drugs clinically used for smoking cessation, activate both murine and human beige adipocytes.

Introduction

The significant health risks associated with obesity include cardiovascular diseases; metabolic diseases, including type 2 diabetes; and certain types of cancers (1). Notably, adipose tissue is more than just a storage hub for energy surplus, it also serves as an active endocrine and thermogenic organ that plays a vital role in the regulation of whole-body energy homeostasis (2,3). In mammals, various types of adipocytes differ in developmental lineages, morphology, anatomic locations, and function (4). Broadly speaking, thermogenic brown and beige adipocytes dissipate energy as heat, whereas white adipocytes primarily store energy (4). Activation of thermogenic fat, through both uncoupling protein 1 (UCP1)-dependent and -independent mechanisms, have been linked to weight loss and improved insulin sensitivity and metabolic health, thereby counteracting obesity and other metabolic disorders in rodents and humans (5–10).

As a key regulator of whole-body metabolic health, adipose activity is tightly monitored through various mechanisms, such as neuronal signaling from the brain and paracrine communications from adipose tissue resident immune cells (11–13). The well-established neuronal pathway for adipose tissue thermogenesis is through β3-adrenergic receptor (AR) and mirabegron, a β3-AR agonist, and activates thermogenic fat in humans (14,15). However, the systemic activation of β3-AR signaling inevitably elevates blood pressure, which is a significant risk factor for stroke and cardiovascular disease, particularly for individuals with obesity (16). Thus, a better understanding of beige fat regulation independent of the β3-AR signaling may offer an alternative and safe approach to stimulate beige fat thermogenesis. A beige-selective immune-adipose interaction via cholinergic receptor nicotinic α2 subunit (CHRNA2) orchestrates beige fat activation and energy homeostasis after environmental challenges (17–21). Chrna2 in beige adipocytes and choline acetyltransferase (ChAT)-expressing immune cells in subcutaneous fat depot are both activated when beige fat is stimulated upon cold exposure (17,19,21–23). Conversely, deletion of Chrna2 in adipocytes or Chat (the rate-limiting enzyme of acetylcholine [ACh] synthesis) in macrophages render compromised adaptive thermogenic response in subcutaneous fat after cold challenge (18,19).

Nicotinic acetylcholine receptors (nAChRs) have been explored as potential drug targets for various human diseases (24). nAChRs function in ligand-gated ion channels as hetero- or homo-pentamers composed of various combinations from total 16 structurally similar subunits (25). In this study, we examined the molecular composition, regulation, and metabolic function of CHRNA2-containing nAChRs in beige adipocytes. An additional nAChR subunit, cholinergic receptor nicotinic β2 subunit (CHRNB2), whose expression increases during beige adipocyte differentiation, interacts with CHRNA2 to form a functional nAChR. These two subunits positively correlate in both mice and human beige adipocytes. Mice lacking Chrnb2 have a blunted thermogenic response in subcutaneous fat after cold exposure and are more prone to insulin resistance after being challenged with high-fat diets (HFDs). Decline with aging of this cholinergic signaling in subcutaneous fat is observed and CHRNB2 partial agonists stimulate both murine and human beige adipocytes. Collectively, our results support that CHRNA2 and CHRNB2 form a functional nAChR in beige adipocytes and may present new potential targets counteracting metabolic diseases such as type 2 diabetes.

Research Design and Methods

Reagents

Dimethyloxalylglycine (DMOG) (catalog no. 71210) and rosiglitazone (catalog no. 71740) were obtained from Cayman. Acetylcholine chloride (catalog no. A2661), cytisine (catalog no. 712264), epibatidine dihydrochloride hydrate (catalog no. E1145), nicotine (catalog no. N3876), varenicline tartrate (catalog no. PZ0004), isoproterenol (catalog no. I6504), dibutyryl cAMP (catalog no. D0260), and Fura 2-AM (catalog no. F0888) were obtained from Sigma-Aldrich. Dexamethasone (catalog no. D4902), insulin (catalog no. I5500), 3-isobutyl-1-methylxanthine (catalog no. I7018), biotin (catalog no. B4639), and d-pantothenic acid hemicalcium salt (P5155) were obtained from Sigma-Aldrich. Recombinant human BMP4 (catalog no. 314-bp-010) was acquired from R&D Systems. Mecamylamine (catalog no. 826-39-1) was purchased from Cayman Chemical.

Animals and Metabolic Analysis

All mouse experiments followed protocols approved by the Institutional Animal Care and Use Committee at the University of Michigan. Mice of both sexes were used in this study and similar phenotypes were observed. Wild-type (WT) C57BL/6J (strain no. 000664), Pdgfra-Cre (strain no. 013148), and Adipoq-Cre (strain no. 028020) mice were obtained from The Jackson Laboratory. The Chrnb2 knockout (KO) mice and Chrnb2^fl/fl mice on a C57BL/6J genetic background were provided by Michael C. Crair (Yale University) (26,27). For the aging study, we used 28-month-old C57BL/6J mice. Mice were individually housed in an environmental chamber at 15°C for the cold exposure experiments. Body temperature was measured using a RT-3 mouse rectal probe. For the HFD study, intraperitoneally glucose tolerance tests (GTTs) and insulin tolerance tests (ITTs) were performed as previously described (28). For the in vivo insulin signaling study, mice fed an HFD for 18 weeks were fasted for 4 h before being injected intraperitoneally with recombinant human insulin at a dose of 1.5 units/kg body weight. Tissues were collected 15 min after the insulin injection.

Mouse Stromal Vascular Fractions Isolation

Stromal vascular fractions (SVFs) from inguinal white adipose tissue (iWAT) were isolated from male and female WT or Chrnb2 KO mice similarly as previously described (29). Floating mature adipocytes were collected if needed for gene expression analysis. The SVF pellet was resuspended in medium, then filtered through a 40-μm cell strainer and centrifuged again. The resulting pellet was collected for gene expression analysis or adipocyte differentiation. CD81⁺ progenitor cells were isolated per the previously published protocol (30,31). In brief, MACS Non-Adipocyte Progenitor Depletion Cocktail for mice (catalog no. 130-106-639, Miltenyi Biotec) and MACS LS columns (Miltenyi Biotec) were used to deplete lineage-positive cells. The following antibodies were used for the isolation of CD81⁺ cells (Lin⁻:Sca1⁺;CD81⁺): Sca-1-PB (1:800; catalog no. 108120 Biolegend), and CD81-APC (1:50; catalog no. 104910, Biolegend) in autoMACS Rising Solution (Miltenyi Biotec) containing 0.5% BSA in the dark at 4°C for 15 min. The isolated cells were infected with a retrovirus expressing large T antigen (pBabe SV40 Large T antigen; Addgene) and subsequently selected using puromycin (2 μg/mL), as previously described (32).

Cell Culture

HEK293T cells were maintained in DMEM (Gibco) supplemented with 10% FBS and 1% penicillin/streptomycin. SVF cells and CD81⁺ progenitor cells from iWAT were cultured in DMEM/F12 with 10% FBS and 1% penicillin/streptomycin, as previously described (30,31). Human adipose-derived stromal cells (hASCs) were provided by Jeffrey M. Gimble (Tulane University). These cells were cultured in MesenPRO RS medium (Gibco) supplemented with 10% FBS (PeproTech) and 1% penicillin/streptomycin. Adipocyte differentiation was conducted as previously described (17).

Plasmid Construction, Virus Packaging, and Luciferase Reporter Assay

The coding sequences of mouse Chrna2 and Chrnb2 were amplified using PCR and cloned into the pLVX-puro vector (Clontech). Ucp1 (∼6.7 kb), Chrna2 (5 kb), and Chrnb2 (2 kb) reporter plasmids were constructed as previously reported (17). All plasmids were verified by Sanger sequencing. Recombinant lentiviral constructs were packaged into lentiviruses in HEK293T cells, using the packaging plasmid psPAX2 (Addgene) and the envelope plasmid pMD2.G (Addgene). Luciferase assay was performed similarly to what we reported previously (17).

Immunoprecipitation and Western Blotting

The differentiated CD81⁺ cells were treated with 10 µmol/L MG132 for 20 h. Whole-cell lysates were prepared in NP-40 lysis buffer. Equal amounts of proteins were immunoprecipitated with anti-Flag or anti-HA antibody with protein G agarose beads (Santa Cruz Biotechnology, catalog no. sc-2002). The protein samples were collected and subjected to immunoblotting as previously described (33). Antibodies used in this study are listed in Supplementary Table 1.

RNA Extraction and Quantitative PCR

Total RNA was extracted using TRI reagent (Sigma), following the manufacturer’s instructions. An equal amount of RNA was used to synthesize complementary DNA using M-MLV Reverse Transcriptase (Invitrogen). Real-time PCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems). The relative mRNA expression levels were determined using the comparative cycle threshold method. The mouse Tbp gene or human TBP gene was used as an internal control for mouse or human samples, respectively. The sequences of the primers used are listed in Supplementary Table 2.

Chromatin Immunoprecipitation Assay

Cells were fixed with 0.8% formaldehyde and then quenched with glycine. The chromatin immunoprecipitation (ChIP) assay was performed as described previously (17). Chromatin was immunoprecipitated with an anti-CREB or anti–hypoxia-inducible factor 1 subunit α (anti-HIF1A) antibody, purified using a PCR purification kit (Sigma-Aldrich), and used for PCR analysis. The sequences of the primers used are listed in Supplementary Table 2.

Calcium Imaging Assay

Primary mouse inguinal preadipocytes, CD81⁺ cells, or hASCs were seeded on collagen-coated glass-bottom culture dishes and differentiated. Upon full differentiation, calcium imaging assay was carried out similarly to what we reported previously (17,18). The cells were perfused with an nAChR agonist (500 μmol/L nicotine or 100 μmol/L ACh), antagonist (100 μmol/L mecamylamine), and CHRNB2-specific nAChR agonists (100 μmol/L varenicline, 100 μmol/L cytisine, and 10 μmol/L epibatidine) to detect changes in a 340/380 nm ratio. The data were analyzed using MetaFluor software (Molecular Devices).

Mitochondrial DNA Copy-Number Analysis

Total DNA was extracted from mouse iWAT using TRI-reagent (Sigma) following the manufacturer’s instructions. Real-Time PCR was used to quantify mitochondrial DNA copy number relative to nuclear DNA copy number, similarly to what we previously described (34). The sequences of the primers used are listed in Supplementary Table 2.

Histology

Adipose tissue was fixed in 10% formalin overnight at 4°C. After dehydration, paraffin-embedding and hematoxylin and eosin (H&E) staining were performed by the University of Michigan School of Dentistry Histology Core. Images were captured using the LEICA DM2000 fluorescence microscope. Adipocyte size was quantified using ImageJ software, and frequency of cell sizes of indicated ranges was plotted using Graph Pad Prism software.

Statistical Analyses

All data are presented as mean ± SEM. Statistical analyses were conducted using Prism 9 (GraphPad Software). An unpaired Student t test was used for comparisons between two groups. Multiple-group comparisons were performed using a ANOVA with Tukey post hoc test. Pearson correlation was used for linear regression. A P value of less than 0.05 was considered statistically significant.

Data and Resource Availability

Transcriptomic data of human subcutaneous fat were obtained from published data sets in the Gene Expression Omnibus (GEO) under accession numbers GSE29718 and GSE132411. All other data generated and analyzed during this study are available from the correspondence author upon reasonable request.

Results

CHRNB2 and CHRNA2 Positively Correlate in Beige Adipocytes

To evaluate potential involvements of other nAChR subunits in the CHRNA2-containing ion channel in beige adipocytes, we examined the expression of all nAChR subunits throughout murine subcutaneous, inguinal preadipocyte differentiation via qPCR. Among all the subunits of nAChR, Chrna2 demonstrated a clear induction through the beige adipogenesis process, consistent with previous reports (17). It is of note that Chrnb2 expression levels followed the exact same trend as Chrna2, albeit a more modest increase in the trend was observed (Fig. 1A). During the differentiation process of hASCs of subcutaneous fat from multiple donors, the synchronously increasing expression patterns of CHRNA2 and CHRNB2 were detected (Fig. 1B). Collectively, a strong positive association between Chrnb2 and Chrna2 expression was seen in differentiated murine SVF from iWAT and in hASCs (Fig. 1C and D). Examination of publicly available gene expression data sets further revealed that this correlation exists in human subcutaneous adipose tissue (Fig. 1E and F). These results indicate CHRNB2 may act in concert with CHRNA2 in the regulation of beige adipocytes.

Figure 1.

CHRNB2 and CHRNA2 form an nAChR in beige fat. A: Relative mRNA expression levels of nAChR subunits during beige adipogenesis of primary iSVF from WT mice (n = 4–5 for each time point). B: qPCR analysis of the mRNA expression of adipogenic marker PPARG, thermogenic marker UCP1, and CHRNA2 and CHRNB2 during beige adipogenesis of hASCs from two individuals (n = 4–6). C: Pearson correlation of Chrnb2 with Chrna2 mRNA in differentiated mouse inguinal preadipocytes (n = 17). D: Correlation of CHRNB2 with CHRNA2 mRNA in differentiated hASCs (n = 32). E and F: CHRNB2 and CHRNA2 are positively correlated in human subcutaneous adipose tissue. E: The GSE29718 data set includes subcutaneous fat from lean participants (n = 8). F: The GSE132411 data set includes subcutaneous fat from the thigh (n = 8). G and H: Cell lysates of differentiated CD81⁺ preadipocytes stably expressing indicated genes were immunoprecipitated (IP) with anti-Flag antibody (G) or anti-HA antibody (H). The immunoprecipitated proteins were analyzed by immunoblotting with both anti-HA and anti-Flag antibodies. * indicates a nonspecific band. Pearson correlation analysis was conducted for C–F, Pearson R and P values are labeled. P < 0.05 was considered statistically significant.

Next, we investigated whether CHRNA2 and CHRNB2 physically interact in beige adipocytes. Previous research has shown that CD81⁺ adipocyte progenitors give rise to beige adipocytes and are essential for tissue remodeling within the subcutaneous adipose tissue niche upon environmental challenges (30). Both nicotine and ACh exposure prompted calcium influx in differentiated primary and immortalized CD81⁺ preadipocytes isolated from inguinal subcutaneous fat depot, indicating the activation of nAChR signaling in these beige adipocytes (Supplementary Fig. 1A–D). The activation of CD81⁺-derived beige adipocytes via nicotine was further demonstrated by an increase in thermogenic gene expression and the phosphorylation of both CREB protein and p38 mitogen-activated kinase (p38) (Supplementary Fig. 1E and F). Coimmunoprecipitation experiments demonstrated a reciprocal interaction between CHRNB2 and CHRNA2 in CD81⁺ progenitors with stable expression of CHRNA2 and CHRNB2 (Fig. 1G and H). These data strongly support the hypothesis that CHRNB2 and CHRNA2 could work together and form a functional nAChR complex in beige fat.

The Expression of Chrnb2 and Chrna2 Is Modulated by Beige Fat Regulators

We next examined whether Chrnb2 expression is enriched in adipocytes in vivo. We separated the SVF and mature adipocytes after collagenase-mediated digestion of mouse iWAT, confirmed by the enriched expression of Cd45 and Pparg, respectively (Fig. 2A). Notably, Chrna2 and Chrnb2 were expressed at a higher level in mature adipocytes (Fig. 2A). Consistent with the modest increase during beige adipogenesis, Chrnb2 expression was elevated with a high dose of the PPARG agonist rosiglitazone in differentiated preadipocytes from iWAT (Fig. 2B). The direct regulation by PPARG was further confirmed via a luciferase assay, during which PPARG and its cofactor PGC1A increased the activity of the Chrnb2 promoter, although to a lesser extent than effects on the promoters of Ucp1 and Chrna2 (Fig. 2C). This activation of CHRNB2 by PPARG agonist treatment was also observed in differentiated hASCs, indicating this regulatory signaling is conserved in humans (Fig. 2D).

Figure 2.

Chrnb2 is enriched in beige adipocytes. A: qPCR analysis of mRNA levels of Cd45, Pparg, Chrna2, and Chrnb2 in the SVF (n = 3) and mature adipocytes (n = 3) fractionated from inguinal fat of WT mice. B: qPCR analysis of the mRNA expression of Ucp1, Chrna2, and Chrnb2 in differentiated primary preadipocytes from iWAT of WT mice treated with vehicle control or 1 μmol/L or 5 μmol/L rosiglitazone (Rosi) (n = 8). C: Transcriptional activity assay of Ucp1-, Chrna2-, and Chrnb2-promoter luciferase reporter activity in 293T cells transfected with vector, PPARG, and PGC1A (n = 4). D: qPCR analyses of UCP1, CHRNA2, and CHRNB2 in differentiated hASCs treated with vehicle control or 5 μmol/L Rosi for 4 days (n = 6). Unpaired Student t test (A and D) and one-way ANOVA with Tukey multiple comparisons test (B and C). *P < 0.05, **P < 0.01, ***P < 0.001.

Transcription factor CREB protein regulates thermogenic gene expression downstream of β-adrenergic signaling (11,35). Several copies of CREB protein binding consensus sequences were detected within the promoter regions of murine Chrna2 and Chrnb2 (Fig. 3A). As expected, overexpression of CREB protein activated the Ucp1 promoter (Fig. 3B). This effect was also observed in the Chrna2 and Chrnb2 promoters, because their activities were upregulated following CREB protein overexpression (Fig. 3B). The ChIP assays demonstrated that CREB protein directly binds to the Chrna2 and Chrnb2 promoters in differentiated CD81⁺ cells after treatment with isoproterenol, an agonist of the β-ARs, which can activate adipocyte thermogenesis through CREB protein (Fig. 3C). Previous research has indicated that hypoxia signaling is activated in iWAT following cold exposure (36). Hypoxia-responsive elements were identified in the promoter regions of both Chrna2 and Chrnb2 (Fig. 3D), and luciferase assays confirmed the direct activation of Chrna2 and Chrnb2 promoters by HIF1A (Fig. 3E). Additionally, inhibition of the degradation of HIF1A with DMOG resulted in elevated Chrna2 and Chrnb2 transcript levels in differentiated mouse primary inguinal preadipocytes as well as in immortalized CD81⁺ preadipocytes (Fig. 3F–I). The direct binding of HIF1A to the promoters of Chrna2 and Chrnb2 was verified through a ChIP assay (Fig. 3J). Further increases of Chrna2 and Chrnb2 promoter activities were observed when CREB protein and HIF1A were coexpressed, consistent with the notion that these transcriptional regulators bind to different sites within the promoters of Chrna2 and Chrnb2 (Supplementary Fig. 1G).

Figure 3.

Chrna2 and Chrnb2 expression in beige adipocytes is regulated by CREB and HIF1A. A: Schematic of CREB-binding sites in murine Chrna2 and Chrnb2 promoters. B: Transcriptional activity assay of Ucp1-, Chrna2-, and Chrnb2-promoter luciferase reporter activity in 293T cells transfected with vector or CREB (n = 4). C: ChIP assay of the CREB-binding sites in differentiated CD81⁺ preadipocytes treated with 10 µmol/L isoproterenol for 4 h. NC, negative control. D: Schematic of HIF1A binding sites in murine Chrna2 and Chrnb2 promoters. E: Transcriptional activity assay of Ucp1-, Chrna2-, and Chrnb2-promoter luciferase reporter activity in 293T cells transfected with vector, or HIF1A (n = 4). F and G: Differentiated primary preadipocytes from mouse iWAT were treated with 1 mmol/L DMOG for 24 h. F: Cell lysates were subjected to immunoblotting analysis for HIF1A and α-tubulin. G: The mRNA levels of Vegfa, Chrna2, and Chrnb2 were determined by qPCR (control [Ctrl], n = 6; DMOG, n = 5). H and I: Differentiated immortalized CD81⁺ preadipocytes were treated with 1 mmol/L DMOG for 8 h. H: Cell lysates were subjected to immunoblotting analysis for HIF1A and α-tubulin. I: The mRNA levels of Vegfa, Chrna2, and Chrnb2 were determined by qPCR (Ctrl, n = 3; DMOG, n = 4). J: ChIP assay of the HIF1A-binding sites in differentiated CD81⁺ preadipocytes treated with 1 mmol/L DMOG for 6 h. K: The mRNA expression of thermogenic gene Ucp1, hypoxia-activating gene Vegfa, and Chrna2 and Chrnb2 in iWAT from WT mice at room temperature (RT) and 10°C cold exposure (CE) for 3 days (n = 12). Unpaired Student t test (B, E, G, I, and K), *P < 0.05, **P < 0.01, ***P < 0.001.

We also investigated the expression levels of Chrna2 and Chrnb2 in vivo in response to cold exposure. Cold exposure led to the activation of the well-known HIF1A target gene Vegfa as well as thermogenic marker Ucp1 (Fig. 3K). Additionally, expression levels of Chrna2 and Chrnb2 were upregulated in iWAT by cold exposure (Fig. 3K). Collectively, these results demonstrate that the nAChRs Chrna2 and Chrnb2 are closely regulated by established thermogenic modulators, consistent with their potential role mediating an integrative response to various signaling that activates beige fat.

CHRNB2 Is Essential for nAChR Signaling in Beige Adipocytes

To establish whether CHRNB2 is an essential component of the functional nAChR signaling in beige adipocytes, the response to nAChR agonists in differentiated mouse primary inguinal preadipocytes of either WT or Chrnb2 KO mice was examined using calcium imaging assays. In the absence of Chrnb2, calcium influx triggered by ACh or nicotine in differentiated inguinal preadipocytes was completely abolished, similar to the previous observation of no detected response to either agonist in Chrna2 KO beige fat cells (Fig. 4A–C) (17). This finding presents direct evidence that both CHRNA2 and CHRNB2 are indispensable for the functional nAChR signaling in beige adipocytes.

Figure 4.

CHRNB2 is required for nAChR-mediated signaling in beige adipocytes. A: qPCR analysis of Chrnb2 mRNA levels in differentiated primary inguinal preadipocytes from WT or Chrnb2 KO mice (WT, n = 5; KO, n = 6). B and C: Measurement of intracellular calcium influx in response to ACh (100 μmol/L; WT, n = 9; KO, n = 23) (B) or nicotine (Nic) (500 μmol/L; WT, n = 24; KO, n = 23) (C) in differentiated primary inguinal preadipocytes from WT or Chrnb2 KO mice. D: Measurement of intracellular calcium influx in response to the CHRNB2-containing receptor agonist varenicline (Var) in differentiated primary inguinal preadipocytes from WT or Chrnb2 KO mice (100 μmol/L; WT, n = 29; KO, n = 31). E: Immunoblots of phosphorylation of CREB and p38 in differentiated primary inguinal preadipocytes from WT mice treated with 500 μmol/L Var for the indicated time or 250 μmol/L cAMP for 15 min. Total CREB, p38, and HSP90 were used as loading controls. F: Measurement of intracellular calcium influx in response to Var (100 μmol/L) in differentiated immortalized CD81⁺ preadipocytes with or without pretreatment of the nAChR antagonist mecamylamine (100 μmol/L; control [Ctrl], n = 16; mecamylamine, n = 35). G: Immunoblots of phosphorylation of CREB and p38 in differentiated immortalized CD81⁺ preadipocytes treated with 500 μmol/L Var for the indicated time or 250 μmol/L cAMP for 15 min. Total CREB, p38, and HSP90 were used as loading controls. H: Measurement of intracellular calcium influx in response to Var (100 μmol/L) in differentiated hASCs from subcutaneous adipose tissue with or without pretreatment of mecamylamine (100 μmol/L; Ctrl, n = 14; mecamylamine, n = 20). I: Immunoblots of phosphorylation of CREB and p38 in differentiated hASCs treated with 500 μmol/L Var for the indicated time or 250 μmol/L cAMP for 15 min. Total CREB, p38, and HSP90 were used as loading controls. J: Measurement of intracellular calcium influx in response to Var (100 μmol/L) of differentiated hASCs in the presence or absence of 5 μmol/L rosiglitazone (Rosi) (Ctrl, n = 10; Rosi, n = 8). Unpaired Student t test (A). ***P < 0.001.

In addition to nicotine and ACh, various known CHRNB2 partial agonists, including varenicline, cytisine, and epibatidine (37–39), were able to activate CHRNB2-containing nAChR in differentiated murine primary inguinal preadipocytes (Fig. 4D and Supplementary Fig. 2). The varenicline-induced calcium influx and the increased phosphorylation of CREB and p38 downstream were observed in differentiated mouse iWAT preadipocytes, differentiated CD81⁺ beige adipocytes, and differentiated hASCs (Fig. 4E–I). This activation is CHRNB2 dependent, because no response to varenicline was detected in differentiated primary inguinal preadipocytes isolated from Chrnb2 KO mice, or CD81⁺ beige adipocytes or differentiated hASCs pretreated with mecamylamine, an antagonist of nAChRs (40) (Fig. 4D, F and H). It is of note that an enhanced response to varenicline was detected in differentiated hASCs treated with rosiglitazone to increase beige fat activity, further supporting the notion that CHRNB2 agonists selectively mediate beige fat activity (Fig. 4J).

Thermogenic Defects in Subcutaneous Fat in the Absence of Chrnb2

We next examined the physiological significance of CHRNB2 signaling in beige fat in vivo using Chrnb2 KO mice. Deletion of Chrnb2 did not alter the expression of Chrna2 or other nAChR subunits in iWAT (Supplementary Fig. 3). Several lines of whole-body KO of Chrnb2 mice have been independently generated, with overall similar phenotypes in the nervous systems being reported, suggesting a functional involvement of this nAChR in associative memory and visual system development (26,41,42). In the context of metabolic function assessment, when housed at an ambient temperature with regular chow-diet feeding, no gross phenotype was noted due to the absence of Chrnb2 (Supplementary Fig. 4A–G). No significant differences in body weight, food intake, fasting blood glucose level, and adipose tissue mass between WT and Chrnb2 KO mice were noted (Supplementary Fig. 4A–C, E, and F). Additionally, the deletion of Chrnb2 did not affect the histology and expression of thermogenic markers and mitochondrial genes in iWAT (Supplementary Fig. 4D and G). When challenged with housing in a cold environment, no differences in food intake or overall body weight were noted between WT and Chrnb2 KO mice (Supplementary Fig. 4H and I). However, a trends, albeit not statistically significant, toward increases in iWAT mass and clearly blunted thermogenic activation were detected in the iWAT of Chrnb2 KO mice, as indicated by the less-elevated expression of both thermogenic and mitochondrial genes compared with that of control mice similarly cold challenged (Supplementary Fig. 4J and K).

We further probed the adipocyte-specific effects of CHRNB2 in the adaptive response to cold challenges using two cell-type-specific Chrnb2 KO mouse models (Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre) (Figs. 5 and 6 and Supplementary Figs. 5–8). Platelet-derived growth factor receptor α (Pdgfra)-Cre mice were crossed to Chrnb2^fl/fl mice, mediating deletion in adipose progenitor cells, where Chrnb2 is expressed, albeit to a lesser extent than in mature adipocytes (Fig. 1A). Deletion of Chrnb2 was confirmed in the iWAT in Chrnb2^fl/fl;Pdgfra-Cre mice (Fig. 5E). No deletion of Chrnb2 was noted in other key organs except for the brain, consistent with previous reports regarding the off-target deletion of this Cre model (Supplementary Fig. 5A) (43). Mature adipocyte-specific deletion of Chrnb2 was achieved through Adiponectin-Cre, with no observed deletion in other organs (Fig. 6E and Supplementary Fig. 7A). Similar to what was observed in Chrnb2 KO mice, expression levels of other nAChRs were not affected by the cell-type-specific deletion of Chrnb2 (Supplementary Figs. 5B and D and 7B and D). In comparison with littermate control mice, both Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice showed no differences in body weight, food intake, core body temperature, adipose tissue mass, or blood glucose levels (Figs. 5A–D and 6A–D; and Supplementary Figs. 5C, H, and L and 7C, H, and L). Likewise, no morphological differences in iWAT, brown adipose tissue (BAT) or visceral white adipose tissue (vWAT) were seen between the flox and Cre littermate controls (Figs. 5F and 6F and Supplementary Figs. 5J and N and 7J and N), nor were any differences observed in thermogenic gene expression in adipose tissues when mice were housed at ambient temperature (Supplementary Figs. 5E–G, I, K, M, and O and 7E–G, I, K, M, and O). However, when challenged with cold exposure, Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice showed less body weight loss (or body weight loss trended downward) without different food intake and other gross anomaly (Figs. 5G–K and 6G–K).

Figure 5.

Deletion of Chrnb2 in adipose progenitor cells impaired thermogenic response in iWAT after cold exposure. A–F: Chrnb2^fl/fl and Chrnb2^fl/fl;Pdgfra-Cre mice were housed at room temperature and fed a chow diet. A: Body weight (Chrnb2^fl/fl, n = 11; Chrnb2^fl/fl;Pdgfra-Cre, n = 7). B: Food intake (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Pdgfra-Cre, n = 11). C: Body temperature (n = 11). D: iWAT mass (Chrnb2^fl/fl, n = 11; Chrnb2^fl/fl;Pdgfra-Cre, n = 7). E: qPCR analyses of Chrnb2 in iWAT (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Pdgfra-Cre, n = 10). F: Representative images of H&E staining of iWAT sections. Scale bar, 50 μm. G–T: Chrnb2^fl/fl and Chrnb2^fl/fl;Pdgfra-Cre mice were subjected to 15°C cold challenge for 3 days. G: Body weight change after cold challenge (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Pdgfra-Cre, n = 14). H: Food intake (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Pdgfra-Cre, n = 9). I: Body temperature (Chrnb2^fl/fl, n = 11; Chrnb2^fl/fl;Pdgfra-Cre, n = 9). J: iWAT mass (n = 10). K: qPCR analyses of Chrnb2 in iWAT (Chrnb2^fl/fl, n = 14; Chrnb2^fl/fl;Pdgfra-Cre, n = 10). L: Representative images of H&E staining of iWAT sections. Scale bar, 50 μm. M: Mitochondrial content (Chrnb2^fl/fl, n = 14; Chrnb2^fl/fl;Pdgfra-Cre, n = 7). N and O: qPCR analyses of thermogenic markers (N) and mitochondrial DNA-encoded transcripts (O) in iWAT (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Pdgfra-Cre, n = 6). P: Immunoblots of UCP1 and oxidative phosphorylation (OXPHOS) complexes and loading control HSP90 in iWAT. Q: BAT mass (n = 8). R: Representative images of H&E staining of BAT sections. Scale bar, 50 µm. S and T: qPCR analyses of thermogenic markers (S) and mitochondrial DNA-encoded transcripts (T) in BAT (n = 8). Unpaired Student t test (A–E, G–K, M–O, Q, S, and T). *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6.

Deletion of Chrnb2 in mature adipocytes impaired thermogenic response in iWAT after cold exposure. A–F: Chrnb2^fl/fl and Chrnb2^fl/fl;Adipoq-Cre mice were housed at room temperature and fed a chow diet. A: Body weight (n = 6). B: Food intake (Chrnb2^fl/fl, n = 6; Chrnb2^fl/fl;Adipoq-Cre, n = 7). C: Body temperature (Chrnb2^fl/fl, n = 6; Chrnb2^fl/fl;Adipoq-Cre, n = 7). D: iWAT mass (n = 8). E: qPCR analyses of Chrnb2 in iWAT (n = 8). F: Representative images of H&E staining of iWAT. Scale bar, 50 µm. G–T: Chrnb2^fl/fl and Chrnb2^fl/fl;Adipoq-Cre mice were subjected to 15°C cold challenge for 3 days. G: Body weight change after cold challenge (Chrnb2^fl/fl, n = 7; Chrnb2^fl/fl;Adipoq-Cre, n = 12). H: Food intake (Chrnb2^fl/fl, n = 8; Chrnb2^fl/fl;Adipoq-Cre, n = 7). I: Body temperature (Chrnb2^fl/fl, n = 8; Chrnb2^fl/fl;Adipoq-Cre, n = 7). J: iWAT mass (Chrnb2^fl/fl, n = 9; Chrnb2^fl/fl;Adipoq-Cre, n = 11). K: qPCR analyses of Chrnb2 in iWAT from Chrnb2^fl/fl (n = 6) and Chrnb2^fl/fl;Adipoq-Cre mice (n = 12). L: Representative images of H&E staining of iWAT sections. Scale bar, 50 µm. M: Mitochondrial content (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Adipoq-Cre, n = 15). N and O: qPCR analyses of thermogenic markers (N) and mitochondrial DNA-encoded transcripts (O) in iWAT (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Adipoq-Cre, n = 17). P: Immunoblots of UCP1 and oxidative phosphorylation (OXPHOS) complexes and loading control HSP90 in iWAT. Q: BAT mass (Chrnb2^fl/fl, n = 7; Chrnb2^fl/fl;Adipoq-Cre, n = 9). R: Representative images of H&E staining of BAT sections. Scale bar, 50 µm. S and T: qPCR analyses of thermogenic markers (S) and mitochondrial DNA-encoded transcripts (T) in BAT (Chrnb2^fl/fl, n = 8; Chrnb2^fl/fl;Adipoq-Cre, n = 9). Unpaired Student t test (A–E, G–K, M–O, Q, S, and T). *P < 0.05, **P < 0.01, ***P < 0.001.

H&E staining revealed that iWAT of Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice had larger adipocytes than did littermate Chrnb2^fl/fl mice (Figs. 5L and 6L). At the transcriptional level, thermogenic genes, including mitochondrial DNA-encoded transcripts, were less induced in iWAT upon cold exposure when Chrnb2 was deleted in preadipocytes and adipocytes (Figs. 5N and O and 6N and O). Consistently, mitochondrial DNA content, protein expression of UCP1, and mitochondrial oxidative phosphorylation complexes in the iWAT were all significantly less induced in Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice compared with littermate Chrnb2^fl/fl controls (Figs. 5M and P and Fig. 6M and P). On the contrary, no differences after cold challenges were seen in tissue mass, morphology, or thermogenic gene expression in BAT and vWAT between the mutant and control mice (Figs. 5Q–T and 6Q–T, and Supplementary Figs. 6C–F and 8C–F), consistent with the notion that CHRNA2 and CHRNB2 mediate a beige fat selective signaling and render minimal effect in brown fat and white fat. These results indicate CHRNB2-mediated signaling influences adaptive thermogenic response in subcutaneous fat without altering metabolic baseline.

Loss of CHRNB2 Affects Systemic Metabolism

This impact of CHRNB2 signaling on energy balance was further observed when mice were challenged with excessive calorie intake when fed an HFD (Figs. 7 and 8, and Supplementary Fig. 9). Despite similar food intake, Chrnb2 KO, Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice fed the HFD gained significantly more body weight than did respective control mice (Figs. 7A and C and 8A and C, and Supplementary Fig. 9A and B). This increase in body weight was consistent with the higher fat content observed in mice with Chrnb2 deletion: greater masses of BAT, iWAT, and vWAT (Figs. 7D and 8D, and Supplementary Fig. 9C). Histological analysis demonstrated that mice with Chrnb2 deletion had a higher proportion of large adipocytes in iWAT compared with respective control mice (Figs. 7E and 8E, and Supplementary Fig. 9D). HFD feeding resulted in a lower thermogenic gene expression in the iWAT of Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice as compared with that in the iWAT of respective control mice without core body temperature difference (Figs. 7B, F, and G and 8B, F, and G).

Figure 7.

Mice with Chrnb2 deletion in adipose progenitor cells exhibit exaggerated metabolic dysfunction after HFD. A–I: Chrnb2^fl/fl and Chrnb2^fl/fl;Pdgfra-Cre mice were on HFD for 14 weeks. A: Food intake (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Pdgfra-Cre, n = 11). B: Body temperature (Chrnb2^fl/fl, n = 7; Chrnb2^fl/fl;Pdgfra-Cre, n = 13). C: Body weight (Chrnb2^fl/fl, n = 9; Chrnb2^fl/fl;Pdgfra-Cre, n = 10). D: Adipose tissue mass (Chrnb2^fl/fl, n = 9; Chrnb2^fl/fl;Pdgfra-Cre, n = 10). E: Representative images of H&E staining (left) and the distribution of adipocyte size (right) in iWAT. Scale bar, 100 µm. F–I: qPCR analyses of Chrnb2 (F), Ucp1 (G), glucose metabolism genes (H), and creatine metabolism–related genes (I) in iWAT (Chrnb2^fl/fl, n = 7; Chrnb2^fl/fl;Pdgfra-Cre mice, n = 13). J: GTT after HFD feeding for 12 weeks (Chrnb2^fl/fl, n = 12; Chrnb2^fl/fl;Pdgfra-Cre, n = 19), AUC, area under the curve. K: ITT after HFD feeding for 14 weeks (n = 19). Unpaired Student t test (A–D, F–K). *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 8.

Mice with Chrnb2 deletion in mature adipocytes exhibit exaggerated metabolic dysfunction after HFD feeding. A–I: Chrnb2^fl/fl and Chrnb2^fl/fl;Adipoq-Cre mice were on HFD for 14 weeks. A: Food intake (Chrnb2^fl/fl, n = 6; Chrnb2^fl/fl;Adipoq-Cre, n = 9). B: Body temperature (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Adipoq-Cre, n = 12). C: Body weight (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Adipoq-Cre, n = 12). D: Adipose tissue mass (Chrnb2^fl/fl, n = 10; Chrnb2^fl/fl;Adipoq-Cre, n = 12). E: Representative images of H&E staining (left) and the distribution of adipocyte size (right) in iWAT. Scale bar, 100 µm. F–I: qPCR analyses of Chrnb2 (F), Ucp1 (G), glucose metabolism genes (H) and creatine metabolism–related genes (I) in iWAT (Chrnb2^fl/fl, n = 9; Chrnb2^fl/fl;Adipoq-Cre mice, n = 12). J: GTT after HFD feeding for 12 weeks (Chrnb2^fl/fl, n = 6; Chrnb2^fl/fl;Adipoq-Cre, n = 9). AUC, area under the curve. K: ITT after HFD feeding for 14 weeks (Chrnb2^fl/fl, n = 8; Chrnb2^fl/fl;Adipoq-Cre, n = 7). L: qPCR analyses of Chat, Chrna2, and Chrnb2 and cellular senescence gene expression in iWAT of the young (8–10 weeks; n = 10) and old (28 months old; n = 14) mice. Unpaired Student t test (A–D, F–K). *P < 0.05, **P < 0.01, ***P < 0.001.

CHRNA2-mediated signaling in beige adipocytes plays a regulatory role in glycolytic beige (g-beige) fat activation and creatine-dependent thermogenic response (18). G-beige fat arises in the absence of adrenergic signaling (44); therefore, its regulation may become particularly relevant in obesity, due to catecholamine resistance that occurred in iWAT after HFD feeding. Likewise, creatine energetics participate in diet-induced thermogenesis in subcutaneous fat (45). Significantly less expression of both g-beige fat and creatine metabolism genes were detected in iWAT of Chrnb2^fl/fl;Pdgfra-Cre and Chrnb2^fl/fl;Adipoq-Cre mice than that of control mice after HFD feeding (Figs. 7H and I and 8H and I). It is of interest that no difference was observed in the expression levels of g-beige fat and creatine signaling–related genes in iWAT of mice with Chrnb2 deletion compared with the control, either under basal condition or after cold challenge (Fig. 8A and B and Supplementary Figs. 5F and G, 6A and B, and 7F and G), implicating the influence of CHRNA2 and CHRNB2 on g-beige fat and creatine metabolism is contingent upon specific metabolic contexts.

The results from the GTT and ITT revealed that these mice with Chrnb2 deletion had less efficient clearance of blood glucose and worsened insulin sensitivity than respective control mice after HFD challenge (Figs. 7J and K and 8J and K and Supplementary Fig. 9E and F). The blunted insulin signaling in peripheral organs were further evaluated with in vivo insulin stimulation assay. We observed less elevated phosphorylation of AKT after insulin injection in both the liver and iWAT of Chrnb2 KO mice in comparison with those of control mice (Supplementary Fig. 9G). These findings strongly support that CHRNB2 plays a role contributing to metabolic resilience against energy imbalance during diet-induced obesity.

Previous research suggests that adipose tissue aging compromises adipogenesis and thermogenesis of beige adipocytes (46–49). Age-related changes in adipose tissue include reconfigurations of fat distribution and composition, coupled with functional decline of adipocyte progenitors and an increase in senescent cells. We further analyzed the expression of nAChR signaling in iWAT of young and old mice. We found age-related increases in the expression of cell senescence markers Cav1, p15, p16^Ink4a, p19^Arf, and p21 in the iWAT of old mice in comparison with their younger counterparts (Fig. 8L). Conversely, the expression levels of Chat, Chrna2, and Chrnb2 were notably diminished in the iWAT of aged mice (Fig. 8L). These results suggest that targeting nAChR signaling could potentially improve age-related functional decline of iWAT and whole-body metabolic dysfunction.

Discussion

Thorough understanding of molecular regulation of thermogenic beige adipocytes is essential for targeting adipose tissue function against energy imbalance and metabolic disorders associated with obesity and aging (49,50). The nAChRs are pentameric ligand-gated ion channels formed from various subunits, resulting in many different subtypes (24,25). The nAChR subunit CHRNA2 has been reported to selectively function in beige adipocytes; however, the full composition of nAChRs in these cells is not completely understood. Here, we demonstrated that CHRNA2 and CHRNB2 physically interact with each other and functionally regulate nAChR signaling in murine and human beige adipocytes. nAChR signaling in subcutaneous adipose tissue, where beige adipocytes are prominent, has been implicated to be involved in hypermetabolism after burn injury (51). Conversely, adipose tissue undergoes substantial changes with age, contributing to physiological decline and the prevalence of age-associated diseases among the elderly (49). The reduced expression of Chat, Chrna2, and Chrnb2 in the iWAT of older mice suggests diminished nAChR signaling with aging. In comparison with CHRNA2, much more extensive research has been carried out to examine the function of CHRNB2, and many subtype-specific agonists targeting CHRNB2 have been generated and evaluated (52). In particular, some of these partial agonists, such as cytisine and varenicline, have been used clinically for smoking cessation (53,54). Further research will reveal whether these molecules with a relatively safe profile and minimal drug interactions could be evaluated in rejuvenating beige fat activity in obesity-associated metabolic dysfunction or age-related adipose tissue functional decline.

In conclusion, our research shows that CHRNB2 and CHRNA2 form a functional nAChR and collaborate to mediate beige fat activity responding to changes in energy balance. This signaling is conserved between species: activation of this signaling is observed in differentiated primary subcutaneous adipocytes isolated from both mice and humans. Defective thermogenic response to cold challenge in iWAT and more exacerbated metabolic dysfunction after HFD feeding were noted in Chrnb2 KO mice and mice with adipocyte-specific deletion of Chrnb2. Recent investigations have increasingly shed light on the functional significance of the nonneuronal cholinergic signaling in regulating peripheral tissue function in various physiological processes (23,28,55,56). The discovery that CHRNB2 is a part of the functional nAChR in beige adipocytes suggests an attractive possibility that a family of smoking-cession drugs could be repurposed to improve metabolic fitness against obesity and its related disorders.

This article contains supplementary material online at https://doi.org/10.2337/figshare.32055912.

Funding Statement

This work was supported by funding from the National Institute of Diabetes and Digestive and Kidney Disease (grants R01DK107583 to J.W. and R01DK097441 to S.K.), the National Institute on Alcohol Abuse and Alcoholism (grant R01AA028761 to J.W.), and the Howard Hughes Medical Institute (to S.K.).