*Article* **Angiopoietin-Like Growth Factor Involved in Leptin Signaling in the Hypothalamus**

**Yunseon Jang 1,2,3,†, Jun Young Heo 1,2,3,†, Min Joung Lee 1,2,3, Jiebo Zhu 1,2,3, Changjun Seo 1,2,3 , Da Hyun Go 1,2,3, Sung Kyung Yoon 1,2,3, Date Yukari <sup>4</sup> , Yuichi Oike <sup>5</sup> , Jong-Woo Sohn <sup>6</sup> , Minho Shong 7,\* and Gi Ryang Kweon 1,3,\***

	- Tel.: +82-42-280-7161 (M.S.); +82-42-580-8226 (G.R.K.)

**Abstract:** The hypothalamic regulation of appetite governs whole-body energy balance. Satiety is regulated by endocrine factors including leptin, and impaired leptin signaling is associated with obesity. Despite the anorectic effect of leptin through the regulation of the hypothalamic feeding circuit, a distinct downstream mediator of leptin signaling in neuron remains unclear. Angiopoietinlike growth factor (AGF) is a peripheral activator of energy expenditure and antagonizes obesity. However, the regulation of AGF expression in brain and localization to mediate anorectic signaling is unknown. Here, we demonstrated that AGF is expressed in proopiomelanocortin (POMC) expressing neurons located in the arcuate nucleus (ARC) of the hypothalamus. Unlike other brain regions, hypothalamic AGF expression is stimulated by leptin-induced signal transducers and activators of transcription 3 (STAT3) phosphorylation. In addition, leptin treatment to hypothalamic N1 cells significantly enhanced the promoter activity of AGF. This induction was abolished by the pretreatment of ruxolitinib, a leptin signaling inhibitor. These results indicate that hypothalamic AGF expression is induced by leptin and colocalized to POMC neurons.

**Keywords:** hypothalamus; AGF; leptin; POMC neuron

#### **1. Introduction**

The hypothalamus controls feeding behavior for the maintenance of whole-body energy balance, and dysregulation of energy intake leads to obesity [1]. Leptin is an anorectic hormone to regulate the hypothalamic neural network including proopiomelanocortin (POMC) neurons, producing α- melanocyte-stimulating hormone (MSH) [2]. Insensitivity to leptin signaling is observed in obese patients, despite the increase in leptin produced by adipose tissue [3,4]. It is reported that the combination of recombinant human leptin, metreleptin and pramlintide enhanced the reduction of body weight in a clinical trial [5]. Additionally, celastrol, originated from thunder god vine roots, increased brain leptin sensitivity to promote weight loss and food intake [6]. However, while considerable research

**Citation:** Jang, Y.; Heo, J.Y.; Lee, M.J.; Zhu, J.; Seo, C.; Go, D.H.; Yoon, S.K.; Yukari, D.; Oike, Y.; Sohn, J.-W.; et al. Angiopoietin-Like Growth Factor Involved in Leptin Signaling in the Hypothalamus. *Int. J. Mol. Sci.* **2021**, *22*, 3443. https://doi.org/10.3390/ ijms22073443

Academic Editor: Rosalía Rodríguez-Rodríguez

Received: 26 February 2021 Accepted: 24 March 2021 Published: 26 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

on the beneficial effect of leptin on feeding control and energy expenditure has been done, downstream molecules of leptin signaling in the hypothalamic neuron still remain elusive.

Hypothalamic leptin signaling initiates Janus kinase 2-signal transducers and activators of transcription 3 (JAK2-STAT3) signaling and extracellular signal-regulated kinase (ERK) to regulate food intake [7] and adenosine monophosphate-activated protein kinase (AMPK) signaling for energy homeostasis [8]. The POMC neuronal population is located within the arcuate nucleus (ARC) of the hypothalamus. In postprandial state, a POMC neuron produces an anorectic molecule on leptin binding to leptin receptor (LEP-R) [2]. Leptin promotes POMC expression through the phosphorylation of STAT3. A significant increase in body weight was observed in *Pomc*-Cre, *Leprflox/flox* mice, in which the leptin receptor was deleted in the POMC neuron [9]. Although leptin signaling in the POMC neuron is important for the regulation of appetite, the mediator of leptin signaling in the POMC neuron is only partially identified. In the current study, we identified angiopoietinlike growth factor (AGF) as a downstream of leptin, and AGF expression is enhanced by leptin in the hypothalamus.

AGF has been identified as a peripheral activator of energy expenditure [10,11] that antagonizes obesity and insulin resistance. Whole-body AGF-knockout (KO) mice exhibited severe obesity [12,13]. Adenovirus-mediated hepatic overexpression of AGF showed to reverse the obesity of AGF-KO mice [12]. AGF is a member of the angiopoietin-like proteins (ANGPTLs) family composed of eight members, all of which possess a coiled-coil domain at the N-terminus and a fibrinogen-like domain at the C-terminus [14]. ANGPTLs are orphan ligands that do not bind Tie1 (tyrosine kinase with immunoglobulin-like and epidermal growth factor (EGF)-like domains 1) or Tie2 receptors—targets of angiopoietin [14]. However, the existence of AGF and the regulating factor of AGF in brain is unknown. In the present study, we demonstrate that AGF is induced by the leptin hormone in the hypothalamus and expressed in POMC neurons.

#### **2. Results**

#### *2.1. Angptl6 Is Expressed in the Hypothalamus and Induced after Feeding*

Hypothalamic appetite control is critical for the maintenance of whole-body energy balance, and the impairment of anorectic signaling is associated with obesity [1]. AGF is known as an activator of energy expenditure in the liver and adipocytes [12,13]. Although AGF protein expression and function in the hypothalamus have not been investigated, AGF mRNA has been detected in the brain of C57BL/6 mice [15]. To determine whether AGF is expressed in the hypothalamus, we performed immunohistochemistry on human brain. As shown in Figure 1A, AGF immunoreactivity was detected in the hypothalamus of postmortem human brain. Because the hypothalamus is sensitive to nutritional status and governs whole-body energy homeostasis [1,2,16,17], we tested whether hypothalamic AGF expression is affected by changes in nutritional status. Mice were allowed to freely approach food, and the refed group was fed after overnight fasting. Fasting did not alter AGF mRNA level in the hypothalamus or epididymal white adipose tissue (eWAT), but did decrease hepatic AGF mRNA level (Figure S1A,B). In contrast, AGF mRNA level increased ~2-fold in the hypothalamus after refeeding (Figure 1B) and modestly increased in liver and eWAT, indicating that feeding induces AGF expression in metabolic organs. Fasting and refeeding status of mice were confirmed by measurements of blood glucose level and insulin concentration in plasma (Figure S1C,D).

**Figure 1.** Angiopoietin-like growth factor (AGF) is expressed in the hypothalamus**.** (**A**) Immunohistochemical detection of AGF in the human brain. The boxed area within the image indicates a magnified field. (**B**) *Agf* mRNA expression in the hypothalamus of *ad libitum-fed*, overnight fasted or refed mice, confirmed by qPCR analysis (*n* = 5/group). Data represent means ± SEM (\**p* < 0.05). Scale bars: 50 μm. **Figure 1.** Angiopoietin-like growth factor (AGF) is expressed in the hypothalamus. (**A**) Immunohistochemical detection of AGF in the human brain. The boxed area within the image indicates a magnified field. (**B**) *Agf* mRNA expression in the hypothalamus of *ad libitum-fed*, overnight fasted or refed mice, confirmed by qPCR analysis (*n* = 5/group). Data represent means ± SEM (\* *p* < 0.05). Scale bars: 50 µm.

hypothalamus. As shown in Figure 2A, AGF immunoreactivity was detected in the ARC where the feeding regulatory circuit is localized. AGF immunoreactivity exhibited a 91% overlap with the neuronal marker, NeuN. However, AGF did not colocalize with glial fibrillary acidic protein (GFAP) (Figure 2B), one of a marker of astrocytes (a type of glial cell). Leptin binds to the LEP-R of the POMC-expressing neuron, which is located in the ARC of the hypothalamus and activates anorectic signaling in a postprandial state [18,19]. POMC activation reduces food intake through the release of α-MSH, and *Pomc*-deficient

hypothalamus was induced by feeding (Figure 1B), we verified the spatial expression of AGF in relation to these major appetite-controlling neurons by performing immunofluorescence staining for AGF in the brain of POMC-mCherry reporter mice. We found that 45.0% of AGF-positive neurons in the ARC of the hypothalamus were also POMC-positive (Figure 2C,D). These results indicate that AGF localized to the POMC-expressing neuronal

) showed obesity and hyperphagia [20]. Because AGF in the

.

#### *2.2. Angptl6 Is Expressed in POMC Neurons of the Hypothalamus*

population and it is not colocalized to GFAP-positive cells.

mice in the ARC (Arc*Pomc***−/<sup>−</sup>**

The hypothalamus delivers pivotal anorectic signaling after feeding. Because AGFpositive cells were found in the human hypothalamus (Figure 1A), we performed immunofluorescence staining of AGF in mice to identify the localization and cell type in the hypothalamus. As shown in Figure 2A, AGF immunoreactivity was detected in the ARC where the feeding regulatory circuit is localized. AGF immunoreactivity exhibited a 91% overlap with the neuronal marker, NeuN. However, AGF did not colocalize with glial fibrillary acidic protein (GFAP) (Figure 2B), one of a marker of astrocytes (a type of glial cell). Leptin binds to the LEP-R of the POMC-expressing neuron, which is located in the ARC of the hypothalamus and activates anorectic signaling in a postprandial state [18,19]. POMC activation reduces food intake through the release of α-MSH, and *Pomc*-deficient mice in the ARC (Arc*Pomc*−/−) showed obesity and hyperphagia [20]. Because AGF in the hypothalamus was induced by feeding (Figure 1B), we verified the spatial expression of AGF in relation to these major appetite-controlling neurons by performing immunofluorescence staining for AGF in the brain of POMC-mCherry reporter mice. We found that 45.0% of AGF-positive neurons in the ARC of the hypothalamus were also POMC-positive (Figure 2C,D). These results indicate that AGF localized to the POMC-expressing neuronal population and it is not colocalized to GFAP-positive cells.

**Figure 2.** AGF is colocalized to proopiomelanocortin (POMC) neurons rather than glial fibrillary

#### acidic protein **(**GFAP)-positive cells in the hypothalamus. (**A**,**B**) Immunofluorescence staining of *2.3. AGF Expression Is Associated with Leptin-Induced STAT3 Phosphorylation*

AGF (green) and NeuN (red), a neuronal nuclei marker and GFAP (magenta), an astrocyte marker Insulin and leptin, which are secreted from peripheral organs and act on the hypothalamus, are induced during the postprandial state and serve to inhibit food intake [21,22]. Accordingly, we tested the possibility that this hormone might act through effects on AGF level to modulate feeding behavior. Treatment of the insulin receptorexpressing cell lines, SN4741 and C2C12, with insulin (100 nM) had no effect on AGF protein or mRNA level (Figure S2A–C). Next, we examined the relationship between AGF and leptin, the latter of which dominates feeding behavior through the activation of anorectic signaling in the hypothalamus [18,23]. The leptin receptor (LepRb) is expressed in hypothalamic neurons, and, upon binding leptin, induces the JAK2 (Janus kinase 2)-mediated phosphorylation of STAT3 (signal transducer and activator of transcription 3) [18,19,24]. To examine changes in AGF level induced by leptin in vivo, we injected mice intraperitoneally with leptin (3 mg/kg) and sacrificed mice after 1 h. Leptin induced an increase in AGF protein in the hypothalamus (Figure 3A,B), but not in other brain areas, including the cortex, striatum and midbrain (Figure S3A). Immunostaining for Tyr705–phosphorylated STAT3 (pSTAT3), a leptin signaling marker, was increased in hypothalamus tissue (Figure 3A,C) and the ARC by leptin injection (Figure S3B).

mice in the ARC (Arc*Pomc***−/<sup>−</sup>**

To demonstrate the response of AGF-expressing neurons to leptin, we immunostained for AGF in the ARC of mice injected with leptin or vehicle (Figure 3D). This analysis showed that leptin increased the number of AGF-positive neurons by 56% (Figure 3D,E). Collectively, these data suggest that leptin induces hypothalamic AGF level. AGF in relation to these major appetite-controlling neurons by performing immunofluorescence staining for AGF in the brain of POMC-mCherry reporter mice. We found that 45.0% of AGF-positive neurons in the ARC of the hypothalamus were also POMC-positive (Figure 2C,D). These results indicate that AGF localized to the POMC-expressing neuronal population and it is not colocalized to GFAP-positive cells.

) showed obesity and hyperphagia [20]. Because AGF in the

#### *2.4. AGF Promoter Activity Is Enhanced by Leptin*

hypothalamus. As shown in Figure 2A, AGF immunoreactivity was detected in the ARC where the feeding regulatory circuit is localized. AGF immunoreactivity exhibited a 91% overlap with the neuronal marker, NeuN. However, AGF did not colocalize with glial fibrillary acidic protein (GFAP) (Figure 2B), one of a marker of astrocytes (a type of glial cell). Leptin binds to the LEP-R of the POMC-expressing neuron, which is located in the ARC of the hypothalamus and activates anorectic signaling in a postprandial state [18,19].

hypothalamus was induced by feeding (Figure 1B), we verified the spatial expression of

*Int. J. Mol. Sci.* **2021**, *22*, x FOR PEER REVIEW 3 of 10

Consistent with in vivo experiments, both AGF and pSTAT3 expression were increased after leptin treatment in N1 cells (Figure 4A–C). We further found that AGF mRNA level was 2.5-fold higher in the hypothalamus of leptin-injected mice (Figure 4D) and 2-fold higher in leptin-treated N1 cells compared with vehicle-treated controls (Figure 4E), suggesting that leptin induces AGF transcription. To test this, we first sought to identify transcription factor binding sites in the AGF promoter using an online search program (http://tfbind.hgc.jp/ accessed on 23 May 2020), which predicted three consensus STAT3 binding sequences in a 1020-bp (−1072 to −52 bp) 5<sup>0</sup> untranslated region (Figure 4F). Using a promoter-reporter construct in which this 1020-bp region was placed upstream of the luciferase gene, we tested whether the promoter activity of AGF was affected by leptin treatment. In N1 cells transfected with this promoter-reporter construct, leptin (100 ng/ml) induced a 2.6-fold increase in reporter activity within 45 min (Figure 4G). Notably, induction of AGF promoter activity by exogenous leptin treatment was ablated by a 1 h pre-incubation with ruxolitinib (1 µM), a leptin signaling inhibitor that blocks JAK2-mediated phosphorylation (Figure 4G). These results indicate that AGF transcription is induced by leptin and mediated by phospho-STAT3, a downstream of leptin/JAK2 signaling. . **Figure 1.** Angiopoietin-like growth factor (AGF) is expressed in the hypothalamus**.** (**A**) Immunohistochemical detection of AGF in the human brain. The boxed area within the image indicates a magnified field. (**B**) *Agf* mRNA expression in the hypothalamus of *ad libitum-fed*, overnight fasted or refed mice, confirmed by qPCR analysis (*n* = 5/group). Data represent means ± SEM (\**p* < 0.05). Scale bars: 50 μm.

**Figure 2.** AGF is colocalized to proopiomelanocortin (POMC) neurons rather than glial fibrillary acidic protein **(**GFAP)-positive cells in the hypothalamus. (**A**,**B**) Immunofluorescence staining of AGF (green) and NeuN (red), a neuronal nuclei marker and GFAP (magenta), an astrocyte marker **Figure 2.** AGF is colocalized to proopiomelanocortin (POMC) neurons rather than glial fibrillary acidic protein (GFAP) positive cells in the hypothalamus. (**A**,**B**) Immunofluorescence staining of AGF (green) and NeuN (red), a neuronal nuclei marker and GFAP (magenta), an astrocyte marker in the hypothalamus section. (**C**) POMC neurons in the hypothalamus, detected based on red fluorescence in POMC-specific tandem dimer Tomato (tdTomato) expressing reporter mice, and AGF immunofluorescence staining. Brain slices were prepared by cryosectioning (25 µm/section). Sections were incubated overnight at 4 ◦C with Alexa 488-conjugated mouse anti-AGF antibody (1:100). Fluorescence was visualized by confocal microscopy. A yellow-colored neuron represents the colocalization of AGF with POMC. (**D**) Number of AGF and POMC double-positive neurons in the ARC (between −1.7 and −3.4 mm from bregma) presented as means ± SD (*n* = 10/slide). Scale bars: 50 µm (upper).

bars: 50 μm (upper).

that leptin induces hypothalamic AGF level.

in the hypothalamus section. (**C**) POMC neurons in the hypothalamus, detected based on red fluorescence in POMC-specific tandem dimer Tomato (tdTomato) expressing reporter mice, and AGF immunofluorescence staining. Brain slices were prepared by cryosectioning (25 μm/section). Sections were incubated overnight at 4 °C with Alexa 488-conjugated mouse anti-AGF antibody (1:100). Fluorescence was visualized by confocal microscopy. A yellow-colored neuron represents the colocalization of AGF with POMC. (**D**) Number of AGF and POMC double-positive neurons in the ARC (between −1.7 and −3.4 mm from bregma) presented as means ± SD (*n* = 10/slide). Scale

Insulin and leptin, which are secreted from peripheral organs and act on the hypothalamus, are induced during the postprandial state and serve to inhibit food intake [21,22]. Accordingly, we tested the possibility that this hormone might act through effects on AGF level to modulate feeding behavior. Treatment of the insulin receptor-expressing cell lines, SN4741 and C2C12, with insulin (100 nM) had no effect on AGF protein or mRNA level (Figure S2A–C). Next, we examined the relationship between AGF and leptin, the latter of which dominates feeding behavior through the activation of anorectic signaling in the hypothalamus [18,23]. The leptin receptor (LepRb) is expressed in hypothalamic neurons, and, upon binding leptin, induces the JAK2 (Janus kinase 2)-mediated phosphorylation of STAT3 (signal transducer and activator of transcription 3) [18,19,24]. To examine changes in AGF level induced by leptin in vivo, we injected mice intraperitoneally with leptin (3 mg/kg) and sacrificed mice after 1 h. Leptin induced an increase in AGF protein in the hypothalamus (Figure 3A,B), but not in other brain areas, including the cortex, striatum and midbrain (Figure S3A). Immunostaining for Tyr705–phosphorylated STAT3 (pSTAT3), a leptin signaling marker, was increased in hypothalamus tissue (Figure 3A, C) and the ARC by leptin injection (Figure S3B). To demonstrate the response of AGF-expressing neurons to leptin, we immunostained for AGF in the ARC of mice injected with leptin or vehicle (Figure 3D). This analysis showed that leptin increased the number of AGF-positive neurons by 56% (Figure 3D,E). Collectively, these data suggest

*2.3. AGF Expression Is Associated with Leptin-Induced STAT3 Phosphorylation*

**Figure 3.** AGF expression is induced by leptin-stimulated signal transducer and activator of transcription 3 (STAT3) activation. (**A**) AGF level in the hypothalamus of C57BL/6 mice after intraperi-**Figure 3.** AGF expression is induced by leptin-stimulated signal transducer and activator of transcription 3 (STAT3) activation. (**A**) AGF level in the hypothalamus of C57BL/6 mice after intraperitoneal injection of recombinant mouse leptin (3 mg/kg) or vehicle, determined by Western blotting. AGF, pSTAT3 (Tyr705), STAT3 and α-tubulin were detected in tissue lysates containing equal amounts of protein from the hypothalamus of each group. (**B**,**C**) Band intensities of AGF and pSTAT3 measured by the ImageJ program; values are normalized to vehicle-injected samples. (**D**) AGF and DAPI (40 ,6-diamidino-2-phenylindole) co-staining in brain sections from leptin- or vehicle-injected mice. Immunoreactivity was detected by confocal microscopy. (**E**) Number of AGF-positive neurons in the ARC. Data represent means ± SEM (\* *p* < 0.05, \*\* *p* < 0.01).

**Figure 4.** AGF promoter activity is increased by leptin. (**A**) AGF protein level in N1 cells, determined after treatment with 100 ng/ml leptin for 0, 30, 45 and 60 min. (**B**,**C**) AGF and pSTAT3 band intensities, measured using the ImageJ program. (**D**,**E**) AGF mRNA expression level in the hypothalamus (**D**) and in N1 cells (**E**) after leptin administration, normalized to 18s rRNA level, as determined by qPCR (*n* = 5/group). (**F**) Location of STAT3 binding sequences in the AGF promoter (−1072 to −52 bp). (**G**) AGF promoter activity, measured in N1 cells transfected with an AGF promoter-luciferase reporter plasmid after a 45 min incubation with leptin (100 ng/ml). Cells were pre-incubated in the presence or absence of the JAK2 inhibitor, ruxolitinib (1 µM) for 1 h. N1 cells were harvested and luminescence was measured. Data represent means ± SEM (\* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001).

#### **3. Discussion**

Modulation of appetite by the hypothalamus is pivotal for energy homeostasis and is associated with the regulation of body weight [1,2,17]. Leptin binding to LEP-R on POMC neurons provokes anorectic signaling in the hypothalamus. It is reported that leptin activates ERK1/2 in the ARC for anorectic and sympathetic effect as confirmed by leptin-induced reduction of food intake was reversed by ERK inhibitor U0126 injection [7,8]. However, the downstream mediator of leptin signaling to modulate hypothalamic neuronal activity that is associated with appetite control is still elusive.

AGF, a known peripheral activator of energy expenditure, is determined by leptin in mice hepatocyte, and serum AGF level is paradoxically increased in human patients with metabolic disease due to compensation [25]. In the previous report, whole-body AGF KO mice showed extreme obesity due to a decrease in energy consumption with no change in food intake. The adenovirus-mediated reconstitution of AGF leads to a decrease in body weight and improved glucose tolerance [12], suggesting that peripheral AGF has a role in energy expenditure. However, alteration of brain AGF expression in this model was not examined and cell-type-specific AGF expression in the hypothalamus was unknown. In the current study, we address the fact that AGF is expressed in the hypothalamus and has leptin responsiveness. As we observed that AGF promoter activity was enhanced by the treatment of leptin and ablated this effect by ruxolitinib, JAK1/2 inhibitor in hypothalamic N1 cells (Fig. 4G), it would be supportive to elucidate the involvement of AGF in leptin signaling if reproducible effects similar with cells occur in the hypothalamus of mice.

To assess whether AGF expression is dependent on leptin in the hypothalamus, *ob/ob* and *db/db* mice with leptin or leptin receptor deficiency could be useful. There is a possibility that AGF expression is reduced in ob/ob mice and exogenous leptin may restore AGF level in the hypothalamus. Additionally, feeding-induced hypothalamic increase in AGF expression may be ablated in these mice. Hypothalamic control of feeding by leptin is disrupted in obese subjects due to insensitivity to leptin signaling, despite the elevated level of circulating leptin [26–28]. The alteration of hypothalamic AGF in obesity remains to be elucidated.

AGF immunoreactivity was detectable in the hypothalamus of human postmortem brain around the third ventricle. Neuronal cell type and nucleus where AGF is localized in the human hypothalamus was not identified. The human protein atlas program (https://www. proteinatlas.org/, accessed on 23 May 2020) revealed AGF expression in the hypothalamus, but it does not present specific regions such as the paraventricular nucleus (PVN) and ARC. In the case of mice, AGF immunoreactivity predominantly colocalized to POMC neurons in the ARC (Figures 1A and 2C), not GFAP-positive cells. AGF expression was not restricted to the ARC but observed in dorsomedial hypothalamic nucleus (DMH) and PVN (data not shown). It is known that leptin administration activates NUCB2/nesfatin-1-expressing neurons in PVN to inhibit food intake [29]. Additionally, stimulation of the Thyrotropinreleasing hormone (TRH) neuron by leptin leads to an increase in pSTAT3 and neuronal activation, reducing appetite [30]. Therefore, to determine whether PVN AGF colocalize to neurons related to food intake regulation would give an insight into the role of AGF in appetite control and metabolism. Recently, it has been reported that tanycytes transport leptin into the hypothalamus and it is required for the regulation of normal hypothalamic leptin signaling [31]. To determine distinct neuronal populations which express AGF in the ARC and whether tanycytes co-express AGF could give an insight into AGF as a downstream mediator of leptin signaling.

A physiological role of hypothalamic AGF as a downstream effector of leptin in POMC neurons has not been determined in this study. Based on our finding that 45% of AGFexpressing neurons colocalized with POMC neurons in the ARC, the highly overlapped expression of AGF and POMC gives rise to the possibility that AGF has a role in POMC neuron-mediated metabolic changes such as appetite regulation and energy expenditure, which are promoted by leptin [32]. Both AGF expression and STAT3 phosphorylation were significantly induced after 30 min of leptin treatment. In contrast, pSTAT3 level declined after 60 min, whereas AGF level was sustained (Figure 4A). From this, we infer that, after induction by pSTAT3, AGF could act to regulate leptin downstream signaling, for example, by inducing JAK2 phosphatase, which is a negative regulator of this signaling [33]. It is also implied that leptin-induced AGF interacts with leptin signaling pathways such as ERK and AMPK in leptin receptor-expressing neurons to control appetite and energy balance. Further study for verification of the physiological role of hypothalamic AGF is including use of hypothalamic POMC neuron-specific AGF KO mice via assessment of an alteration of leptin signaling and α-MSH level. In addition, recording the neuronal activity of POMC in a condition of increasing AGF expression by leptin treatment using a patch-clamp study could be a useful approach. Our study suggests that AGF is a downstream molecule of leptin signaling in the hypothalamus.

#### **4. Materials and Methods**

#### *4.1. Animals*

Male C57BL/6 mice were purchased from Harlan Teklad (Indianapolis, IN, USA). Mice were maintained at 22 ◦C under a 12 h light–dark cycle. For diet-induced obesity, 5-weekold C57BL/6 mice were fed a high-fat diet (60% fat; Research Diets Inc., New Brunswick, NJ, USA). Leptin (3 mg/kg, R&D systems, Minneapolis, MN, USA) was dissolved in 0.9% saline and intraperitoneally injected into C57BL/6. Animal experiments were approved

by the Institutional Animal Care and Use Committee of Chungnam National University (Ethical approval number, 201903A-CNU-46, approved on 1 June 2019).

#### *4.2. Cell Culture*

HypoE-N1 mice embryonic hypothalamus cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (Gibco, Waltham, MA, USA), 1% penicillin/streptomycin at 37 ◦C under 5% CO2/21% O<sup>2</sup> condition.

#### *4.3. Human Brain Tissue*

The hypothalamus was obtained from postmortem brains of otherwise healthy individuals at the University of Miyazaki, Japan, with the approval of the institutional review board (approval number, C-0037; 11 May 2018). Written informed consent was acquired from each outpatient, and the study was conducted according to provisions of the Declaration of Helsinki.

#### *4.4. Immunofluorescence Staining and Immunohistochemistry*

Postmortem hypothalamus was fixed in 4% paraformaldehyde for 24 h, embedded in paraffin, and cut into 4-µm–thick sections. Immunohistochemistry was performed using an anti-human AGF antibody, produced by immunizing rabbits with a synthetic peptide corresponding to amino acids 392–408 (NDKPESTVDRDRDSYSG) of the AGF protein. Sections were pretreated with periodic acid (Nichirei, Tokyo, Japan) and incubated with anti-human AGF antibody (1:100) overnight. Then, the sections were washed with phosphate-buffered saline (PBS) and immunostained using EnVision/horseradish peroxidase (Dako, Glostrup, Denmark). After incubating with 3,3-diaminobenzidine solution, images were acquired using an IX70 microscope (Olympus, Tokyo, Japan). Mice whole brain was dipped in the 4% PFA and moved to 30% sucrose solution. The brain section was cut into 25 µm and blocked for 1 hour with 3% donkey serum (Dako, Glostrup, Denmark) and 0.3% triton x-100. Then, brain sections, including the arcuate nucleus region, were obtained from −1.7 to −3.4 mm from bregma and were incubated with primary antibodies, including a mouse anti-AGF-488 alexa conjugated antibody (1:100, Bioss, MA, USA), anti-NeuN (1:500, Abcam, Cambridge, UK), a chicken anti-GFAP (1:1000, Abcam, Cambridge, UK), and a rabbit anti-POMC (1:300, Phoenix, CA, USA) antibody, overnight at 4 ◦C. Sections were washed with PBS and incubated in secondary fluorescence antibody. Fluorescence was visualized using a IX70 fluorescent microscope (Olympus, Tokyo, Japan).

#### *4.5. RNA Iolation and Real-Time PCR*

The hypothalamus was removed from mice brain and homogenized by TissueLyserII (Qiagen, Netherlands). Total RNA was isolated using Isol-RNA lysis reagent (5 PRIME, South San Francisco, CA, USA). cDNA was synthesized by an moloney murine leukemia virus (M-MLV) reverse transcriptase kit (Invitrogen, Carlsbad, CA, USA). For real-time PCR, after mixing cDNA, primers and 10X SYBR mix, mRNA expression was analyzed using a Rotor Gene 6000 system (Corbett Life Science, Venlo, Netherlands) and normalized to 18s rRNA. Gene-specific primers are listed in Table S1.

#### *4.6. Western Blotting*

Proteins of mice tissues were prepped and homogenized by TissueLyserII, and N1 cells were extracted using radioimmunoprecipitation assay (RIPA) buffer (1% Nonidet P-40, 0.1% SDS, 150 mM NaCl, 50 mM Tris–HCl pH 7.5 and 0.5% deoxycholate) with 10% of phosphatase and protease inhibitor (Roche, Basel, Switzerland). A total of 12 µg of protein samples was loaded on SDS-PAGE gel and run by electrophoresis; afterwards, it was transferred to polyvinylidene fluoride (PVDF) membrane, blocked by 5% skim milk. Membranes were incubated with primary antibody including a mouse anti-AGF (R&D systems, Minneapolis, MN, USA), a mouse anti-α-tubulin (Santa Cruz Biotechnology, Dallas, TX, USA), anti-phospho-STAT3 and anti-STAT3 (Cell Signaling, Danvers, MA, USA) antibody at 4 ◦C overnight. Anti-Immunoglobulin G (IgG) horseradish peroxidase antibody (Pierce Biotechnology, MA, USA) corresponds with the host of the primary antibody and was used as a secondary antibody. A protein band was detected by the enhanced chemiluminescence (ECL) system (Thermo Scientific, Waltham, MA, USA)

#### *4.7. Promoter-Luciferase Reporter Assay*

N1 cells were grown in 6-well plates and transfected with 1 µg pAGF prom-Luc DNA mixture using lipofectamin reagents (Invitrogen, Carlsbad, CA, USA). The pLightSwitchprom plasmid encodes RenSP luciferase sequence under control of 1020 bp AGF promoter. After transfection, cells were treated with 100 ng/ml leptin or vehicle for 45 min and harvested for luminiscence reading. Luciferase activity was measured using a Berthold LB9507 luminometer (Berthold Technologies, Black Forrest, Germany).

#### *4.8. Statistical Analysis*

All data are represented as mean ± standard error mean (SEM) from triplicate results. The statistical analysis was determined using Prizm version 5 software (Graphpad, San Diego, CA, USA). The significance of differences between two groups was analyzed by one-tailed student's *t*-test. *p* value < 0.05 was considered statistically significant.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/ijms22073443/s1, Figure S1: AGF mRNA expression is increased after feeding in mice, Figure S2: AGF expression is not affected by insulin, Figure S3: Leptin injection did not induce AGF expression in various brain region excluding hypothalamus.

**Author Contributions:** Y.J., J.Y.H., M.S. and G.R.K. contributed to conceptualization and study design. Y.J. and J.Y.H. were responsible for the data acquisition and performing animal experiment. M.S. and G.R.K. helped with data interpretation. Y.J., J.Y.H., M.J.L., M.S. and G.R.K. wrote the manuscript. J.Z., C.S., J.-W.S., S.K.Y. and D.H.G. contributed to discussion and article revision. D.Y. and Y.O. performed human tissue histology. M.S. and G.R.K. supervised, and all authors approved the final version of the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was funded by Bio & Medical Technology Development Program of the National Research Foundation of Korea (NRF) grant and the Ministry of Science and ICT (MSIT) and the Ministry of Education (2014R1A6A1029617, 2019R1F1A1059586, 2019R1A2C2088345, 2019M3E5D1A02068575, 2017R1A5A201538, 2020R1I1A1A01061771, 2017R1E1A1A01075126), research fund of Chungnam National University and a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (HR20C0025).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki with the approval of institutional review board (approval number, C-0037; May 11, 2018). Animal experiments were approved by the Institutional Animal Care and Use Committee of Chungnam National University (Ethical approval number, 201903A-CNU-46, approved on 1 June 2019).

**Informed Consent Statement:** Written informed consent was acquired from each outpatient involved in the study.

**Data Availability Statement:** The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. No applicable resources were generated or analyzed during the current study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Adiponectin Controls Nutrient Availability in Hypothalamic Astrocytes**

**Nuri Song 1,†, Da Yeon Jeong 1,†, Thai Hien Tu <sup>1</sup> , Byong Seo Park <sup>2</sup> , Hye Rim Yang <sup>1</sup> , Ye Jin Kim <sup>1</sup> , Jae Kwang Kim <sup>1</sup> , Joon Tae Park <sup>1</sup> , Jung-Yong Yeh <sup>1</sup> , Sunggu Yang 3,\* and Jae Geun Kim 1,\***


**Abstract:** Adiponectin, an adipose tissue-derived hormone, plays integral roles in lipid and glucose metabolism in peripheral tissues, such as the skeletal muscle, adipose tissue, and liver. Moreover, it has also been shown to have an impact on metabolic processes in the central nervous system. Astrocytes comprise the most abundant cell type in the central nervous system and actively participate in metabolic processes between blood vessels and neurons. However, the ability of adiponectin to control nutrient metabolism in astrocytes has not yet been fully elucidated. In this study, we investigated the effects of adiponectin on multiple metabolic processes in hypothalamic astrocytes. Adiponectin enhanced glucose uptake, glycolytic processes and fatty acid oxidation in cultured primary hypothalamic astrocytes. In line with these findings, we also found that adiponectin treatment effectively enhanced synthesis and release of monocarboxylates. Overall, these data suggested that adiponectin triggers catabolic processes in astrocytes, thereby enhancing nutrient availability in the hypothalamus.

**Keywords:** adiponectin; astrocyte; energy metabolism; hypothalamus; glycolysis; metabolic diseases

### **1. Introduction**

The brain constitutes a metabolically active organ that requires the highest energy demands in the human body. Although the adult brain represents approximately 2% of the total body weight, it consumes approximately a quarter of the total glucose used for its energy supply [1–4]. Therefore, nutrient availability in the central nervous system (CNS) is directly linked to the maintenance of life. As neurons expend high levels of energy resources, such as glucose and lactate, to initiate and propagate their action potentials [4,5], impairment of the energy supply can lead to perturbation of neuronal excitability. Consistent with these concepts, multiple brain disorders are also deeply associated with abnormalities of energy metabolism in the CNS.

Astrocytes, which comprise the most abundant cell type in the CNS, support normal neuronal functions by regulating the concentration of chemical substances in the synaptic cleft area and providing nutrients between blood vessels and neurons [4–6]. Although astrocytes are responsible for the metabolic processing of glucose absorbed by the brain [4,7], they do not require as much energy as they uptake. Rather, the primary driving factor underlying astrocyte participation in glucose uptake and utilization is the provision of energy sources from astrocytes to neurons.

Adiponectin, an adipokine predominantly secreted from adipocytes, has functional roles in regulating glucose and lipid metabolism along with insulin sensitivity. In particu-

**Citation:** Song, N.; Jeong, D.Y.; Tu, T.H.; Park, B.S.; Yang, H.R.; Kim, Y.J.; Kim, J.K.; Park, J.T.; Yeh, J.-Y.; Yang, S.; et al. Adiponectin Controls Nutrient Availability in Hypothalamic Astrocytes. *Int. J. Mol. Sci.* **2021**, *22*, 1587. https://doi.org/10.3390/ ijms22041587

Academic Editor:

Rosalía Rodríguez-Rodríguez Received: 9 December 2020 Accepted: 1 February 2021 Published: 4 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

lar, adiponectin facilitates systemic glucose and lipid homeostasis by regulating several major metabolic organs, such as adipose tissue, liver, and muscle [7–11]. Accordingly, adiponectin elicits beneficial effects in multiple metabolic diseases and their related secondary complications. For example, adiponectin improves hyperglycemia by alleviating glucose intolerance and insulin resistance [9,10]. Furthermore, it mitigates hepatic steatosis and dyslipidemia through regulation of lipid metabolism [10]. Specifically, adiponectin stimulates glucose uptake by skeletal and cardiac muscle and inhibits glucose production by the liver, consequently decreasing blood glucose levels [7,8]. However, although it is well established that adiponectin dynamically participates in the regulation of peripheral energy metabolism, its impact on nutrient metabolism in astrocytes of the hypothalamus, a central unit for the regulation of energy homeostasis, has not yet been clearly elucidated.

We hypothesized that central adiponectin regulates multiple metabolic processes in hypothalamic astrocytes including glucose uptake, glycolytic activity, fatty acid oxidation and metabolites secretion. To verify this hypothesis, in this study we mainly utilized primary astrocytes extracted from the mouse hypothalamus and determined the active roles of circulating adiponectin on hypothalamic astrocytes coupled to whole body energy homeostasis.

#### **2. Results**

#### *2.1. Central Administration of Adiponectin Results in the Activation of Hypothalamic Astrocytes*

Based on the evidence that astrocytes respond to metabolic alterations and reactive astrocytes display morphological changes [12], we evaluated the number of astrocytes and their pattern of interaction with blood vessels in the hypothalamus assessed by immunohistochemistry with an antibody against Gfap, a molecular maker for the astrocyte after central administration of adiponectin. Icv administration of recombinant adiponectin into the lateral ventricle of mice resulted in an elevated number of astrocytes in the hypothalamus (Figure 1A,B). We further examined the contact ratio between astrocytes and blood vessels by performing fluorescence immunohistochemistry combined with a visualization of blood vessels by cardiac infusion of lectin to speculate whether adiponectin participates in nutrient shuttling between astrocytes and blood vessels. Notably, icv administration of adiponectin led to an increase in astrocytes interaction with the blood vessel (Figure 1C). In order to further confirm whether adiponectin triggers reactive astrogliosis, we examined the mRNA expression involved in the astrocyte activation. Semi-quantitative RT-PCR results showed an elevation of *Gfap* and *Catenin beta 1*(*Ctnnb1)* transcripts involved in the processes of astrocyte activation (Figure 1D,E), indicating that adiponectin induced reactive astrogliosis.

**Figure 1.** Central administration of adiponectin leads to activation of hypothalamic astrocytes. The whole brain was collected from mice that received an intracerebroventricular (icv) injection of adiponectin (ADN, 3 µg/2 µL) and intracardiac injection of lectin. The distribution of astrocytes was examined by immunohistochemistry using an antibody against glial fibrillary acidic protein (Gfap), a molecular marker for astrocytes. (**A**) Representative images of Gfap immunolabeling in the hypothalamus. Icv administration of adiponectin led to increased (**B**) number of astrocytes and (**C**) contact ratio between astrocytes and blood vessels in the hypothalamus of the mouse brain (*n* = 5 for each group). Icv administration of adiponectin elevated mRNA levels of (**D**) *Gfap* and (**E**) *Ctnnb1* as determined by qRT-PCR (*n* = 5 for CTL; *n* = 6 for ADN). Results are presented as mean ± SEM. \* *p* < 0.05. Scale bar = 50 µm.

#### *2.2. Adiponectin Enhances Glucose Uptake in Astrocytes*

To validate the purification of primary astrocytes, we tested the enrichment of GFAP protein in cultured primary astrocytes compared to that in primary microglial cells. GFAP protein was predominantly present in primary astrocytes but not in primary microglia. Additionally, the Iba-1 protein was almost absent in primary astrocytes (Figure 2A). Given the well-known effect of adiponectin on phosphorylation of the AMPK protein, an evolutionarily conserved energy sensor and regulator of energy metabolism, we evaluated the induction of AMPK phosphorylation as determined by immunoblot assay to validate the cellular impact of recombinant adiponectin in primary astrocytes. A treatment of adiponectin resulted in a significant increase in AMPK phosphorylation (Figure 2B–D).

As multiple lines of evidence indicate that adiponectin enhances glucose uptake and utilization in metabolically active peripheral organs, such as muscle and adipose tissue [9], we interrogated whether adiponectin altered glucose uptake in astrocytes using mouse hypothalamic primary astrocytes. We first identified that a significant elevation in glucose uptake was observed in adiponectin-treated astrocytes compared to vehicle-treated astrocytes (Figure 2E). In support of this finding, we observed that exogenous treatment of adiponectin led to increased levels of glucose transporter-1 (Glut-1) protein (Figure 2F,G) and mRNA (Figure 2H), consistent with the effects of adiponectin on glucose metabolism in peripheral organs. To further verify the effect of adiponectin on glucose uptake, we quantified the mRNA expression of *Glut-1* after silencing the expression of adiponectin receptor 1 (AdipoR1) and adiponectin receptor 2 (AdipoR2). We validated the reduced levels of *AdipoR1* and *AdipoR2* mRNAs in response to transfection with siRNAs (Figure 2I,J). An increase of *Glut1* mRNA expression induced by adiponectin treatment was significantly reversed by AdipoR1 and AdipoR2 siRNAs (Figure 2K). These observations suggest that adiponectin has an active role for glucose uptake in hypothalamic astrocytes.

**Figure 2.** Adiponectin enhances glucose uptake in primary astrocytes. The primary astrocytes and microglia were set up at <sup>5</sup> <sup>×</sup> <sup>10</sup><sup>5</sup> cell/well and subjected to western blot analysis by using antibodies for GFAP and Iba-1 proteins. (**A**) GFAP protein was predominantly present in primary astrocytes but Iba-1 protein was almost absent in primary microglia. (**B**–**K**) Primary astrocytes were seeded at 5 <sup>×</sup> <sup>10</sup><sup>5</sup> cell/well, starved overnight, and treated with adiponectin (ADN, 1 µg/mL) for 24 h. Cell lysates were subjected to western blot analysis using antibodies against (**B**–**D**) pAMPK, total AMPK, β-actin and (**F**,**G**) glucose transporter-1 (GLUT-1) proteins and qRT-PCR using a primer set of (**H**,**K**) *Glut-1* gene. Exogenous treatment of adiponectin led to increased levels of (**F**,**G**) Glut-1 protein and (**H**) *Glut-1* mRNA in primary astrocytes. (**E**) Adiponectin treatment enhanced glucose uptake in cultured primary astrocytes as determined by 2DG-glucose uptake assay. AdipoR1 and AdipoR2-specific siRNA were transfected into the primary astrocytes to confirm the mRNA expression of *Glut1* induced by adiponectin treatment. qPCR data showed a significant decrease in (**I**) *AdipoR1* and (**J**) *AdipoR2* mRNAs. Elevated (**K**) *Glut1* mRNA induced by adiponectin was effectively rescued by transfection of AdipoR1 and AdipoR2-specific siRNAs. Results are presented as the means ± SEM. *n*= 4 for (**B**–**D**); *n*= 3 for (**E**); *n*= 4–5 for (**F**,**G**); *n*= 6 for (**H**); *n*= 4–5 for (**I**,**J**); *n*= 5–6 for (**K**). All experiments were performed from at least three different preparations of astrocytes. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001; # *p* < 0.05.

#### *2.3. Adiponectin Enhances Glycolytic Activities in Astrocytes*

We next examined the effect of adiponectin on the glycolytic activities in primary astrocytes. Semi-quantitative RT-PCR results showed that adiponectin treatment elevated the expression of *hexokinase1 (Hk1)*, which catalyzes glucose phosphorylation during glycolysis (Figure 3A). In addition, we further evaluated glycolytic activity utilizing a seahorse XF-24 extracellular flux analyzer. The ECAR in primary astrocytes was significantly elevated by adiponectin treatment (Figure 3B). To verify that enhanced glycolytic activity was triggered by adiponectin, we performed GC-MS to measure the metabolites produced in the glycolytic processes in the hypothalamus of mice injected with adiponectin. Central administration of adiponectin led to an elevation of pyruvate and lactate levels, which correlated with glycolytic activity (Figure 3C,D). Furthermore, icv injection of adiponectin induced increased in levels of fumarate, malate, and citrate generated during the TCA cycle (Figure 3E–G). Notably, elevated glucose levels were observed in the hypothalamus of adiponectin-treated mice (Figure 3H). These data indicated that adiponectin facilitated brain glucose utilization by reinforcing glycolytic activity in astrocytes.

**Figure 3.** Adiponectin enhances glycolytic activity in primary astrocytes. Primary astrocytes were seeded at <sup>5</sup> <sup>×</sup> <sup>10</sup><sup>5</sup> cell/well, starved overnight, and treated with adiponectin (ADN, 1 µg/mL) for 24 h. (**A**) The level of mRNA encoding *hexokinase 1 (Hk1)* was upregulated in adiponectin-treated primary astrocytes compared to that in vehicle-treated primary astrocytes as determined by qRT-PCR (*n* = 6 for each group). (**B**) The extracellular acidification rates (ECAR) were elevated by adiponectin treatment (*n* = 3 for each group). GC-MS results showed increased levels of (**C**) pyruvate, (**D**) lactate, (**E**) fumarate, (**F**) malate, (**G**) citrate, and (**H**) glucose in the hypothalamus of adiponectin-treated mice compared with that of vehicle-treated mice (*n* = 4 for each group). Results are presented as the means ± SEM. qPCR and ECAR experiments were performed from at least three different preparations of astrocytes. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001.

#### *2.4. Adiponectin Promotes Synthesis and Release of Monocarboxylates in Primary Astrocytes*

As the production and release of monocarboxylates from astrocytes are coupled to glycolytic processes [4,13], we next interrogated whether adiponectin alters production of monocarboxylates by performing the analysis of mRNA expression in primary astrocytes treated with recombinant adiponectin. We observed that adiponectin treatment led to increased mRNA level of *lactate dehydrogenase (Ldh)*, a catalytic enzyme that reversibly

induces conversion of lactate to pyruvate (Figure 4A), and upregulated *monocarboxylate transporter-1 (Mct-1)* (Figure 4B). In support of these molecular observations, an exogenous treatment of recombinant adiponectin led to an elevation of lactate release in cultured primary astrocytes (Figure 4C). In addition, adiponectin-treated astrocytes showed increased mRNA levels of genes involved in the biosynthesis of ketone bodies including *hydroxymethylglutaryl-CoA synthase (Hmgcs)* (Figure 4D) and *hydroxymethylglutaryl-CoA lyase (Hmgcl)* (Figure 4E). In accordance with the patterns of mRNA expression, the medium level of β-hydroxybutyrate was significantly elevated in adiponectin-treated primary astrocytes (Figure 4F). To confirm the increased synthesis and release of ketone bodies, we identified that adiponectin-treated astrocytes showed increased mRNA levels of fatty acid transport protein *(FATP)*, a fatty acid transporter and genes involved in fatty acid oxidation, such as peroxisome proliferator-activated receptor-α *(PPAR-α)* and carnitine palmitoyltransferase 1 α *(CTP1-α)* (Figure 4G–I). These findings indicate that adiponectin promotes synthesis and release of monocarboxylates including lactate and β-hydroxybutyrate in the hypothalamic astrocyte through promoting glucose and lipid utilization.

#### *2.5. Central Administration of Adiponectin Leads to an Elevation of Catabolic Processes in the Hypothalamic Astrocytes*

To further specifically confirmed the impacts of adiponectin on the metabolic process in hypothalamic astrocyte, we used Ribo-Tag technique with *Gfap-Cre;Rpl22HA* mice that expressed HA-tagged ribosomal protein Rpl22 in astrocytes. The immunohistochemistry experiment confirmed astrocyte-specific Cre recombination by identifying the immunosignals of HA protein in the Gfap-positive hypothalamic astrocytes (Figure 5A). In addition, we validated the purification of mRNA extracted from the hypothalamic astrocyte by confirming a predominant expression of *Gfap* mRNA, a molecular maker for the astrocytes and minor expression of *Iba-1* mRNA, a molecular marker for microglia and *NeuN* mRNA, a maker for neuron in purified sample compared with the input control sample (Figure 5B). In consistent with mRNA expression data obtained from mouse total hypothalamus and primary astrocytes, we observed that icv administration of adiponectin effectively elevated mRNA levels of *Glut-1* (Figure 5C) and *Hk1* (Figure 5D), indicating enhanced glucose utilization. In addition, Ribo-tag results revealed that central administration of adiponectin resulted in an increase of mRNA levels of genes involved in the synthesis of monocarboxylates including *Ldh* (Figure 5E)*, Hmgcl* (Figure 5F) and *Hmgcs* (Figure 5G), enzymes for the synthesis of lactate and ketone body as well as *Mct-1 gene* (Figure 5H). In support of these observations and in vitro findings, we also identified that central administration adiponectin resulted in elevated levels of mRNAs involved in the fatty acid utilization including *FATP* (Figure 5I)*, PPAR-α* (Figure 5J) and *CTP1-*α (Figure 5K). From these data, we successfully confirmed that adiponectin triggers enhanced the catabolic process of nutrients such as glucose and fatty acids, and monocarboxylates production in hypothalamic astrocytes utilizing a mouse model that enabled the purification of astrocyte-specific mRNA.

**Figure 4.** Adiponectin promotes production of lactate and ketone body in primary astrocytes. Primary astrocytes were seeded at 5 <sup>×</sup> <sup>10</sup><sup>5</sup> cell/well, starved overnight, and treated with adiponectin (1 µg/mL) for 24 h. The elevated levels of mRNA encoding (**A**) *lactate dehydrogenase (Ldh)* and (**B**) *monocarboxylate transporter-1 (Mct-1)* genes were observed in adiponectintreated primary astrocyte as determined by qRT-PCR (*n* = 6 for each group). (**C**) The concentration of medium lactate was increased 24 h after adiponectin treatment (*n* = 5 for each group). The elevated mRNA levels of (**D**) *hydroxymethylglutaryl-CoA synthase (Hmgcs)* and (**E**) *hydroxymethylglutaryl-CoA lyase (Hmgcl)* were observed in adiponectin-treated primary astrocytes as determined by qRT-PCR (*n* = 8 for each group). (**F**) The medium concentration of β-hydroxybutyrate was increased 24 h after adiponectin treatment (*n* = 5 for each group). The increased mRNA levels of (**G**) *fatty acid transport protein (FATP)*, (**H**) *peroxisome proliferator-activated receptor-α (PPAR-α)* and (**I**) *carnitine palmitoyltransferase 1-*α (*CTP1-*α) were observed in the adiponectin-treated astrocytes compared with the that in vehicle-treated astrocytes (*n* = 6 for each group). Results are presented as the means ± SEM. All experiments were performed from at least three different preparations of astrocytes. \* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001.

**Figure 5.** Central administration of adiponectin leads to enhanced glucose utilization and monocarboxylates production in hypothalamic astrocytes of the mice. Ribo-tag analysis were performed to determine alterations in astrocyte-specific mRNA expression in the hypothalamus of *Gfap-CreERT2:Rpl22HA* mice. (**A**) Representative images revealing co-expression of HA and Gfap immunosignals in the hypothalamus of *Gfap-CreERT2:Rpl22HA* mice. (**B**) qRT-PCR data showing enrichment of *Gfap* mRNA, a molecular marker for astrocyte (but not *Iba-1*, a molecular maker for microglia and *NeuN*, a molecular marker for neuron) in the purified RNA immunoprecipitated with HA antibody compared with the input RNA extracted from hypothalamus. Elevated mRNA levels of (**C**) *Glut-1,* (**D**) *Hk1,* (**E**) *Ldh,* (**F**) *Hmgcl*, (**G**) *Hmgcs*, (**H**) *Mct-1*, (**I**) *FATP*, (**J**) *PPAR-α* and (**K**) *CTP1-*α were observed in hypothalamic astrocytes from *Gfap-CreERT2:Rpl22HA* mice that received an icv injection of adiponectin (ADN, 3 µg/2 µL) compared with vehicle-treated control group (CTL). Results are presented as mean ± SEM. *n* = 6 for each group. \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001. Scale bar = 50 µm.

#### *2.6. Adiponectin Rescued 2-DG-Induced Hyperphagia*

Central administration of 2-deoxy-D-glucose (2-DG) led to an increase in appetite, suggesting that lower glucose availability drives feeding behavior in association with hypothalamic circuit activity [14]. Thus, we further evaluated the effect of adiponectin on the hyperphagic response induced by 2-DG treatment to identify the physiological relevance of adiponectin-induced alteration in cellular metabolism of hypothalamic astrocytes. In accordance with previous findings, icv administration of 2-DG led to an increase in food intake compared to that in vehicle-treated mice (Figure 6A,B). Increased 2 h and 18 h cumulative food intake induced by 2-DG administration was effectively rescued by adiponectin treatment (Figure 6A,B). These observations suggest that the enhanced catabolic processes of nutrients in hypothalamic astrocytes triggered by adiponectin may be coupled with the homeostatic feeding behavior under the hypoglycemic conditions.

**Figure 6.** Central administration of adiponectin reversed increased food intake in response to icv injection of 2-DG. Icv injection of adiponectin curbed the increase of (**A**) 2 h and (**B**) 18 h cumulative food intake induced by icv injection of 2-DG. Results are presented as the mean ± SEM. *n* = 4–6 for each group. \*\*\* *p* < 0.001 for the 2-DG-treated versus the CTL groups; *# p* < 0.05, ### *p* <0.001 for the 2-DG + ADN-treated versus 2-DG-treated groups.

#### **3. Discussion**

The current study highlights an active role of adiponectin in the regulation of nutrient availability in hypothalamic astrocytes. In turn, this suggests that the alteration of nutrient availability between the hypothalamic neuron-astrocyte-blood vessel axis controlled by adiponectin might be tightly coupled to the function of the hypothalamic circuit that controls whole-body energy metabolism [4,15].

Consistent with the concretized evidence that adiponectin affects glucose and lipid utilization in peripheral organs, we here verified the active contributions of adiponectin in glucose and lipid metabolism of hypothalamic astrocytes. We identified that adiponectin facilitates the availability of nutrients in the hypothalamus by promoting the utilization of glucose and fatty acids through the multiple in vitro and in vivo strategies. The hypothalamus constitutes the master organ that controls energy homeostasis by mediating afferent signals derived from metabolically active peripheral organs [16]. In particular, adipose tissue dynamically communicates with the hypothalamus through its own chemical messengers termed adipokines [17,18]. Notably, the circulating levels of most adipokines are proportional to adiposity and long-term elevation of adipokines elicits adverse effects, such as inflammation, oxidative stress, and endoplasmic reticulum stress, which might comprise potential pathogenic elements for the development of metabolic disorders [19].

However, the effects of adiponectin on metabolic controls are distinct from those of general adipokines. In particular, it exerts beneficial effects against multiple cellular stresses and the development of metabolic disorders [15]. In accordance with previous studies, which indicate that adiponectin improves inflammatory responses, [20,21], our re-

sults validated the anti-inflammatory properties of adiponectin in hypothalamic astrocytes (Supplementary Figure S1). Moreover, as most adipokines participate in the operation of hypothalamic circuit activity, central adiponectin also acts as an appetite regulator by targeting hypothalamic neurons associated with the brain glucose concentration [22]. Together with our data, these observations suggest that adiponectin-controlled glucose metabolism in hypothalamic astrocytes might be linked to the hypothalamic control of energy homeostasis. Consistent with this, a growing body of evidence has suggested that the nutrient availability between hypothalamic astrocytes and neurons determines the operation of the circuit activity triggering homeostatic feeding behavior [6,23]. Furthermore, recent reports have shown that the metabolic shift induced by fasting or high-fat diet treatment results in alteration of hypothalamic metabolites, such as lactate and ketone bodies [24]. In the present study, we also identified elevated synthesis and release of the monocarboxylates such as lactate and ketone body in primary astrocytes in response to adiponectin treatment. Thus, it is reasonable to hypothesize that adiponectin may affect the excitability of hypothalamic neurons by modulating astrocyte-derived substances, such as gliotransmitters and metabolites. Recently, considerable effort has been paid to unmask the direct contributions of hypothalamic glial cells in regulation of the hypothalamic neuronal circuit [4,25]. However, there is still insufficient information regarding the astrocyte-derived tropic factors that give rise to homeostatic behaviors and responses directly coupled to hypothalamic neuronal functions. In the present study, we proposed that the hypothalamic astrocyte responds to fluctuations in circulating adiponectin and presumably participates in operation of the hypothalamic circuit by modulating the nutrient availability linked to various cellular metabolic processes, including nutrient uptake and release between the extracellular environment and hypothalamic neurons. Notably, a recent study showed that adiponectin evokes the excitation of hypothalamic proopiomelanocortin neurons, which promotes satiety signals under low glucose conditions [22]. In this regard, we identified that central administration of adiponectin effectively curbed hyperphagic behavior induced by 2-DG-mediated limited glucose utilization. As neuron-glia metabolic coupling constitutes a critical cellular event to preserve normal brain functions and the availability of oxygen and energy resources is tightly coupled to neuronal excitability and functions [26], the impairment of nutrient shuttling between astrocytes and neurons may serve as a primary pathological event in the development of multiple neurodegenerative diseases. Thus, our findings raised an open question regarding the potential beneficial effects of central adiponectin in the development of metabolic diseases caused by the degeneration of hypothalamic neurons coupled with the impairment of the neuron-glia metabolic interaction in the hypothalamic circuitry. Therefore, further studies are required to clarify whether disruptions of neuronal functions linked to metabolic disorders can be reversed by adiponectin treatment. Additionally, further investigations are warranted to elucidate whether adiponectin-controlled astrocytes control the activity of hypothalamic neurons linked to energy homeostasis via tropic factors or modulating synaptic inputs. Nevertheless, the current findings collectively suggest that central adiponectin may support normal hypothalamic functions by promoting nutrient availability in hypothalamic astrocytes.

#### **4. Materials and Methods**

#### *4.1. Animals*

Eight-week-old male C57BL/6 mice (Dae Han Bio Link, Seoul, Korea) were housed in a 12-h light-dark cycle at 25 ◦C and 55 ± 5% humidity. The mice were allowed access to normal diet and tap water *ad libitum*. All animal care and experimental procedures were performed in accordance with a protocol approved by the Institutional Animal Care and Use Committee (IACUC) at the Incheon National University (permission number: INU-2016-001).

#### *4.2. Cannula Implantation for Intracerebroventricular (Icv) Injection*

The mice were anesthetized with an intraperitoneal injection of tribromoethanol (250 mg/kg, Sigma-Aldrich, St. Louis, MO, USA) and placed in a stereotaxic apparatus (Stoelting, Wood Dale, IL, USA). The 2.5-mm cannula (26 gauge) was implanted into the lateral ventricle (X: 1 mm, Y: 0.4 mm to the bregma) and secured to the skull with dental cement. Animals were kept warm until they recovered from the anesthesia and then placed in individual cages. After surgery, a recovery period of seven days was allowed prior to the initiation of experiments.

#### *4.3. Icv Injection of Adiponectin*

Mice were matched based on body weight and food intake during the adaptation period and divided into adiponectin and phosphate-buffered saline (PBS)-injected groups. Recombinant globular adiponectin (Lugen Sci, Bucheon, Korea) was dissolved in PBS to a concentration of 1.5 mg/mL. Each solution was freshly prepared on the day of administration and free of any contaminants, such as endotoxin. Mice were administered the first icv injection of adiponectin (3 µg/2 µL) 24 h before sacrifice and hypothalamic tissues were harvested 1 h after the second injection of adiponectin (3 µg/2 µL). For the test of feeding behavior, mice were injected with recombinant globular adiponectin (3 µg/2 µL) 1 h before icv injection of 2-deoxy-D-glucose (2.5 mg/2 µL, Sigma-Aldrich, St. Louis, MO, USA), and their food intake was measured.

#### *4.4. Immunohistochemistry*

Mice were anesthetized, their thoracic cavities were opened, and 30 µg of tomato lectin (Vectorlabs, Burlingame, CA, USA) in 100 µL volume was injected directly into the left ventricle of the heart, over a period of approximately 30 s. The heart continued to beat for approximately 1 min following injection. Subsequently, the animals were transcardially perfused with 0.9% saline (wt/vol), followed by fresh fixative of 4% paraformaldehyde in phosphate buffer (PB, 0.1 M, pH 7.4). Brains were collected and post-fixed overnight before coronal sections (50 µm thickness) were taken by vibratome (5100 mz Campden Instruments, Leicestershire, England). After washing in PB several times, the sections were pre-incubated with 0.3% Triton X-100 (Sigma-Aldrich) in PB for 30 min at room temperature (RT) and incubated overnight with rabbit anti-GFAP antibody (1:1000 dilution, ab7260, abcam, Cambridge, UK) or mouse anti-HA antibody (1:1000; MMS-101R, BioLegend, San Diego, CA, USA) at RT. Immunofluorescence was performed with the secondary antibodies (Alexa Fluor 488-labeled anti-mouse antibody, 1: 500; A11001, Invitrogen, Carlsbad, CA, USA or Alexa Fluor 594-labeled anti-rabbit antibody, 1: 500; A21209, Invitrogen, Carlsbad, CA, USA) for 2 h at RT. The sections were then mounted onto glass slides and covered by coverslips with a drop of mounting medium (Dako North America Inc, Carpinteria, CA, USA). The coverslips were sealed with nail polish to prevent desiccation and movement of the samples under the microscope. The images were recorded using fluorescence microscopy (Axioplan2 Imaging, Carl Zeiss Microimaging Inc., Thornwood, NY, USA) and subjected to analyses.

#### *4.5. IHC Image Capture and Analyses*

Images were acquired by fluorescence microscopy (Axioplan2 Imaging; Carl Zeiss Microimaging Inc.). For IHC analyses, sections were anatomically matched with the mouse brain using atlas50 (hypothalamic region: between 1.46 and −1.82 mm from bregma). Both sides of the bilateral hypothalamic region were analyzed for two brain sections per mouse. The number of GFAP-positive astrocytes was counted using ImageJ 1.47 v software (National Institutes of Health, Bethesda, MD, USA; http://rsbweb.nih.gov/ij/). Hoechst (Sigma-Aldrich, St. Louis, MO, USA) staining was performed to identify cell nuclei. In sections double-labeled with GFAP antibody and tomato lectin, the number of astrocytes in contact with blood vessels was calculated as a percentage of the GFAP-positive cells in contact with blood vessels per total number of GFAP-positive cells.

#### *4.6. Primary Astrocyte Culture*

Following decapitation of five C57BL6 mice (5 days old), the hypothalamic tissues were removed, combined in a sterile dish and triturated in Dulbecco's modified Eagle medium (DMEM) F-12 containing 1% penicillin-streptomycin. The cell suspension was filtered through a 100-µm sterile cell strainer to remove debris and fibrous layers. The suspension was centrifuged and the pellet was resuspended in DMEM F-12 containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cells were grown in this culture medium in 75-cm<sup>3</sup> culture flasks at 37 ◦C and 5% CO2. When cells grew to confluence (about 9 days), the flasks were placed in a 37 ◦C shaking incubator at 240 rpm for 16 h. The cells were then harvested using 0.05% trypsin-ethylenediamine tetraacetic acid, resuspended in DMEM F-12 containing 10% FBS and 1% penicillin-streptomycin, and centrifuged for 5 min at 1000 rpm. Cells were seeded at a concentration of 5 <sup>×</sup> <sup>10</sup><sup>5</sup> cells/mL in culture plates previously treated with poly-L-lysine hydrobromide (50 µg/mL) and grown for 24 h. The medium was changed to DMEM F-12 containing 1% antibiotics without FBS; 24 h later, the same medium plus either 1 µg/mL of recombinant globular adiponectin or vehicle was added. Cells and/or medium were collected 1 or 24 h after treatment. For siRNA transfection, primary astrocytes were seeded (5 <sup>×</sup> <sup>10</sup><sup>5</sup> cells/mL) in 6-well culture plates and transfected with siRNAs specific for AdipoR1 (# 72674–1, Bioneer, Daejeon, South Korea) and AdipoR2 (# 68465–1, Bioneer, Daejeon, South Korea) using Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocols. A scrambled siRNA (Bioneer) was used as a negative control.

#### *4.7. Measurement of Glucose Uptake*

Mouse primary astrocytes were seeded at a concentration of 5 <sup>×</sup> <sup>10</sup><sup>5</sup> cells/well in 6-well plates. After 24 h treatment of globular adiponectin (1 µg/mL), the cells were processed for a glucose uptake assay following the glucose uptake assay kit instructions (Bio Vision Research Products, Milpitas, CA, USA).

#### *4.8. Measurement of Monocarboxylates*

The mouse primary astrocytes were seeded at a concentration of 1 <sup>×</sup> <sup>10</sup><sup>5</sup> cells/well in 12-well plates and treated with PBS (vehicle control) or globular adiponectin (1 µg/mL) for 24 h. After treatment, the medium were collected and centrifuged to eliminate remaining cells and debris, and processed to measure the concentration of monocarboxylates. For the lactate release assay, the concentration of lactate in the collected medium was examined by using the lactate assay kit (Biomedical Research Service Center, Buffalo, NY, USA) according to the manufacturer's instructions. The concentration of β-hydroxybutyrate (ketone body) in the collected medium was measured by using β-hydroxybutyrate colorimetric assay kit (Cayman, 700190, MI, USA) according to the manufacturer's instructions.

#### *4.9. Extraction and Analysis of Hydrophilic Metabolites in the Hypothalamus*

The extraction and analysis of hydrophilic metabolites in the hypothalamus was performed as previously described [27]. Each sample was extracted with 1 mL of methanol: water:chloroform solution (5:2:2, *v*/*v*/*v*) and 0.03 mL of L-2-chlorophenylalanine in distilled water (0.3 mg/mL) as an internal standard (IS). A mixture of approximately 300 mg glass beads (acid-washed, 425–600 µm, G8772, Sigma-Aldrich, St. Louis, MO, USA) was homogenized for 20 s using a bead beater (Mini Beadbeater-96, BioSpec Products, Bartlesville, OK, USA) and sonicated for 10 min. Thereafter, the samples were incubated in a thermomixer (model 5355, Eppendorf AG, Hamburg, Germany) at 1200 rpm for 30 min at 37 ◦C and centrifuged at 16,000× *g* at 4 ◦C for 5 min. The liquid from upper layer (0.8 mL) was transferred into a clean tube and mixed with 0.4 mL distilled water. After centrifugation under the same conditions, 0.9 mL was separated in a new tube and dried using a centrifugal concentrator (VS-802F, Visionbionex, Gyeonggi, Korea) for at least 3 h. The sample was placed in a freeze-dryer (MCFD8512, Ilshin, Gyeonggi-do, Korea) for 16 h to achieve complete concentration. Thereafter, 0.08 mL methoxyamine hydrochloride (20 mg/mL) in

pyridine was added for derivatization and incubated at 1200 rpm at 30 ◦C for 90 min. Next, 0.08 mL *N*-methyl-*N*-(trimethylsilyl) trifluoroacetamide was added and allowed to react at 1200 rpm at 37 ◦C for 30 min. The derivatized hydrophilic metabolites (1 µL) were analyzed using gas chromatography–mass spectrometry (GC–MS). The GC–MS system consisted of an auto-sampler AOC-20i and GCMS-QP2010 Ultra system (both Shimadzu, Kyoto, Japan) equipped with a DB-5 column (30 m × 0.25 mm id, film thickness 1.0 µm, 122–5033, Agilent, Santa Clara, CA, USA). The flow rate of helium as carrier gas was 1.1 mL/min, and split ratio was 1:10. The injector temperature was 280 ◦C. The column oven program was as follows: 100 ◦C for 4 min, increased at a rate of 10 ◦C/min to 320 ◦C, and then maintained at 320 ◦C for 11 min. The ion source temperature was 200 ◦C, and the interface was set at 280 ◦C. The scanned mass range was 45–600 *m*/*z*. Lab solutions GCMS solution software (version 4.11; Shimadzu, Kyoto, Japan) was used to identify the hydrophilic metabolites in the hypothalamus. The results were filtered with their retention times and mass spectra with reference to standard compounds and the in-house library. Quantitative analysis was conducted using the ratio of the analyte peak area to the IS peak area.

#### *4.10. Ribo Tag Analysis*

To specifically evaluated mRNA expression from hypothalamic astrocytes, we utilized the Ribo-Tag translational profiling system [28,29]. In order to generate Ribo-tag mice (*Gfap-CreERT2: Rpl22HA*), *Rpl22HA* mice (Stock No. 011029, Jackson Laboratory) were crossbred with *glial fibrillary acidic protein (Gfap)-CreERT2* mice (Stock No.012849), which specifically expresses *Cre* recombinase in astrocytes. Since the *Gfap-CreERT2* mice expressed *Cre* recombinase under the control of the tamoxifen inducible GFAP promoter, 8-weekold *Gfap-CreERT2: Rpl22HA* mice received daily intraperitoneal injections for 5 days of tamoxifen (100 mg/kg,T5648, Sigma-Aldrich, St. Louis, MO, USA) dissolved in corn oil (C8267, Sigma-Aldrich). RNA isolation with the Ribo-Tag system was performed as described by a previous reporters [29,30]. Briefly, the hypothalamus was harvested and homogenized before RNA extraction. RNA was extracted from 10% of the cleared lysate and used as an input control. The remaining lysate was incubated with mouse anti-HA antibody for 4 h at 4 ◦C followed by the addition of protein G agarose beads (LGP-1018B, Lugen, Gyeonggi-Do, South Korea) and overnight incubation at 4 ◦C. The beads were washed three times in high-salt solution. The bound ribosomes and RNA were separated from the beads with 30 s of vortexing. Total RNA was extracted using a QIAGEN RNeasy Micro Kit (74034, Qiagen, Hilden, Germany), according to the manufacturer's instructions and quantified with NanoDrop Lite (Thermo Scientific, Waltham, MA, USA). To evaluate the levels of ribosome-associated mRNA in astrocytes, we synthesized cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA) and performed quantitative real-time PCR (qRT-PCR).

#### *4.11. Quantitative Real-Time Reverse Transcription-Polymerase Chain Reaction (qRT-PCR)*

Total RNA was extracted from the hypothalamus or cultured cells according to the Tri-Reagent protocol, and cDNA was then synthesized from total RNA using a high-capacity cDNA reverse transcription kit. Real-time PCR amplification of the cDNA was detected using the SYBR Green Real-time PCR Master Mix (Toyobo Co., Ltd., Osaka, Japan) in a Bio-Rad CFX 96 Real-Time Detection System (Bio-Rad Laboratories, Hercules, CA, USA). The results were analyzed by the CFX Manager software and normalized to the levels of *β-actin* and *L19*, housekeeping genes. The primer sequences used are shown in Table 1. All reactions were performed under the following conditions: initial denaturation at 95 ◦C for 3 min, followed by 40 cycles of 94 ◦C for 15 s, 60◦C for 20 s, and 72 ◦C for 40 s.


#### **Table 1.** Real-time PCR primer sequences.

#### *4.12. Immunoblotting*

Primary astrocytes were seeded at 5 <sup>×</sup> <sup>10</sup><sup>5</sup> cells/well in 6-well plates, allowed to attach overnight, and then incubated in DMEM containing 1 µg/mL adiponectin for 24 h. The cells were washed twice with PBS, followed by scraping and resuspension of the cell pellet in RIPA lysis buffer containing protease inhibitors and centrifugation to remove debris, unbroken cells, and cellular nuclei. Protein content was determined by Bradford's method using bovine serum albumin as a standard. Samples containing 10 µg of total protein were separated using 12% sodium dodecyl sulfate acrylamide gels. After electrophoresis, proteins were transferred to a nitrocellulose membrane and then blocked with 5% skim milk in TBS (Tris-buffered saline, 0.1% Tween20) buffer for 2 h and incubated in the primary antibodies, including anti-GFAP (1:1000 dilution, Sigma-Aldrich, USA), anti-Iba-1 (1:1000 dilution, Wako, Osaka, Japan), anti-Glut1 (1:500 dilution, Santa Cruz Biotechnology, Dallas, TX, USA), anti-phosphorylated (p)AMPK, anti-total AMPK (1:1000 dilution, Cell Signaling, Danvers, MA, USA), and anti-β-actin (1:10,000 dilution, Sigma-Aldrich, USA).

#### *4.13. Measurement of Extracellular Acidification Rates (ECAR)*

ECAR measurements were performed using the XF24 Extracellular Flux analyzer (Seahorse Bioscience, North Billerica, MA, USA). Briefly, primary astrocytes were plated into XF24 (V7) polystyrene cell culture plates (Seahorse Bioscience). Primary astrocytes were seeded at 5 <sup>×</sup> <sup>10</sup><sup>4</sup> cells/well in poly-L-lysine hydrobromide-coated XF24 plates. The cells were attached overnight and then incubated with globular adiponectin (1 µg/mL) for 6 h in a humidified 37 ◦C incubator with 5% CO2. Prior to performing an assay, growth medium in the wells of an XF cell plate was exchanged with the appropriate assay medium to achieve a minimum 1:1000 dilution of growth medium. Subsequently, 560 µL of the assay medium was added to cells for an XF assay. While sensor cartridges were calibrated, cell plates were incubated in a 37 ◦C/non-CO<sup>2</sup> incubator for 60 min prior to the start of an assay. All experiments were performed at 37 ◦C. Each measurement cycle consisted of a mixing time of 3 min and a data acquisition period of 3 min (12 data points) for the XF24. All compounds were prepared at appropriate concentrations in desired assay medium and adjusted to pH 7.4. For XF24, 80 µL of compound was added to each injection port. In a typical experiment, three baseline measurements were taken prior to the addition of any compound, and three response measurements were taken after the addition of each compound. ECAR were reported as absolute rates mph/min for ECAR.

#### *4.14. Statistical Analysis*

Data were analyzed using analysis of variance (ANOVA) followed by Student's *t*-test using Prism GraphPad. *p* value ≤ 0.05 was considered statistically significant. The values are represented as the means ± standard error of the mean (SEM).

**Supplementary Materials:** The following are available online at https://www.mdpi.com/1422-0 067/22/4/1587/s1, Figure S1: Adiponectin leads to decreased mRNA levels of proinflammatory cytokines in primary astrocytes.

**Author Contributions:** Conceptualization, N.S., D.Y.J., S.Y. and J.G.K.; investigation, N.S., D.Y.J., T.H.T., B.S.P., H.R.Y., Y.J.K., J.K.K., J.T.P. and S.Y.; data curation, B.S.P., J.T.P. and J.-Y.Y.; writing original draft preparation, N.S., D.Y.J., T.H.T., S.Y. and J.G.K.; writing—review and editing, S.Y. and J.G.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by an Incheon National University (Foreign post-doctor program-2016) Research Grant for Thai Hien Tu and supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI17C1600).

**Institutional Review Board Statement:** This study was approved by the Institutional Animal Care and Use Committee (IACUC) at the Incheon National University (permission number: INU-2016-001).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All data reported in the manuscript and in the Supplementary materials.

**Conflicts of Interest:** The authors declare no competing financial interest.

### **References**

