Saturated fatty acids stimulate cytokine production in tanycytes via the PP2Ac-dependent signaling pathway

Abstract

The hypothalamic tanycytes are crucial for free fatty acids (FFAs) detection, storage, and transport within the central nervous system. They have been shown to effectively respond to fluctuations in circulating FFAs, thereby regulating energy homeostasis. However, the precise molecular mechanisms by which tanycytes modulate lipid utilization remain unclear. Here, we report that the catalytic subunit of protein phosphatase 2 A (PP2Ac), a serine/threonine phosphatase, is expressed in tanycytes and its accumulation and activation occur in response to high-fat diet consumption. In vitro, tanycytic PP2Ac responds to palmitic acid (PA) exposure and accumulates and is activated at an early stage in an AMPK-dependent manner. Furthermore, activated PP2Ac boosts hypoxia-inducible factor-1α (HIF-1α) accumulation, resulting in upregulation of an array of cytokines. Pretreatment with a PP2Ac inhibitor, LB100, prevented the PA-induced elevation of vascular endothelial growth factor (VEGF), fibroblast growth factor 1 (FGF1), hepatocyte growth factor (HGF), and dipeptidyl peptidase IV (DPPIV or CD26). Our results disclose a mechanism of lipid metabolism in tanycytes that involves the activation of PP2Ac and highlight the physiological significance of PP2Ac in hypothalamic tanycytes in response to overnutrition and efficacious treatment of obesity.

Introduction

Tanycytes, comprising approximately 2.5% of the total hypothalamic cells, ³ communicate within the hypothalamus and respond to systemic factors to maintain energy homeostasis by regulating feeding, ⁴ thermogenesis, ⁵ and nutrition metabolism. ¹ Anatomically, tanycytes are specialized, unciliated ependymal cells lining the wall and the ventrolateral floor of the third ventricle (3 V). ² Their cell somata are embedded in the ependymal layer and are in direct contact with the cerebrospinal fluid (CSF). The elongated foot processes of the β2 tanycytes extend to the fenestrated capillaries or the pia surface, forming a tanycytic barrier that is key to the blood-brain barrier (BBB) or blood-CSF barrier (BCB) in the median eminence (ME). The β1, α2, and α1 tanycytes respectively stretch into the arcuate nucleus (ARC), ventromedial nucleus (VMH), and dorsomedial nucleus (DMH) of the hypothalamus, ⁶ facilitating the monitoring and transporting of nutrients and metabolic hormones between the brain and the blood.^{1,7 –11} Diverse evidence suggests that ablation or functional alteration of tanycytes increases susceptibility to obesity and body fat content in mice.^2,12,13 Selective ablation of the ARC and ME tanycytes in mice leads to a significant increase in adiposity accompanying insulin insensitivity. ¹² Furthermore, tanycytes can act as progenitor cells, generating neurons and astrocytes in the adult hypothalamus. ¹⁴ The newborn neurons from β2-tanycytes can restore the neuron loss induced by a high-fat diet (HFD). ¹⁵

Free fatty acids (FFAs) in the blood circulation undergo significant fluctuations influenced by the systemic energy state. FFAs have been shown to traverse the BBB and their levels are enormously elevated in the CSF of obese individuals. It is well known that the hypothalamic astrocytes uptake the unsaturated fatty acids, which are then metabolized by β-oxidation, generating precursors for ketogenesis under physiological conditions. On the other hand, tanycytes uptake and store both saturated and unsaturated lipids and then transport them to astrocytes. Aside from transporting the lipid droplets (LDs) via endocytosis and exocytosis, tanycytes also metabolize lipids via β-oxidation or the adenosine 5’-monophosphate (AMP)-activated protein kinase (AMPK)-induced lipophagy to generate ATP in mitochondria.^1,16 These capacities of tanycytes are increased in obese mice fed HFD with elevated levels of circulating lipids and hypothalamic ketone bodies. ¹⁷ Since the metabolism of LDs requires a high amount of oxygen consumption, elevated LDs metabolism leads to relative hypoxia in the cytoplasm. The hypoxia-inducible factors (HIFs) act as oxygen sensors and are activated under low oxygen conditions. HIF-1α, one of the subunits of HIFs, orchestrates transcriptional activation of approximately 100-200 genes that encode proteins responsible for regulating nutrient uptake and metabolism, angiogenesis, and cell proliferation and survival. ¹⁸ Moreover, HFD consumption can trigger an inflammatory response in the hypothalamus. Studies have shown that tanycytes secrete fibroblast growth factor 21 (FGF21) and chemerin during overnutrition, which are known to modulate anti-inflammatory and proinflammatory responses, respectively.^1,19,20 Although many efforts have shown the undeniable importance of tanycytes on lipid metabolism, the documentation of underlying mechanisms is rare and controversial.

Protein phosphatase 2 A (PP2A), a serine/threonine phosphatase, is responsible for most of the serine/threonine dephosphorylation of proteins. PP2A is widely present in many types of cells in various brain regions and is highly expressed in developing and adult brains. ²¹ The composition of PP2A holoenzymes comprises a scaffolding subunit A (PP2Aa), a regulatory subunit B (PP2Ab), and a catalytic subunit C (PP2Ac). ²² The core enzyme consists of PP2Ac and PP2Aa, to which various PP2Abs are bound. Given its significant role in diverse physiological functions and its dysfunction in many diseases, numerous efforts have been made to control the phosphatase activity and/or expression of PP2A. ²³ Prior evidence of PP2A expression and function in neurons and non-neurons in the central nervous system (CNS) has implicated its role in neuroinflammation ²⁴ and tauopathy. ²⁵ Furthermore, it has been noticed that high levels of fat or glucose in the peripheral system can significantly boost the expression and activation of PP2A in the liver, ²⁶ adipose tissue, ²⁷ skeletal muscle, ²⁸ heart, ²⁹ pancreas, ³⁰ and artery. ³¹ Thus, PP2A may play a vital role in metabolic phenotypes such as insulin resistance, glucose intolerance, type 2 diabetes, and obesity. Additionally, various genetically modified mouse models targeting PP2A have deepened our insights into the metabolic regulation function of PP2A.^{26,32 –34} For instance, liver-specific deletion of Ppp2cα, which encodes the Cα subunit, enhances glucose metabolism and insulin sensitivity. ²⁶ In addition, a whole-body, inducible knockout of Ppp2r1a (encoding the Aα subunit) in mice leads to a low body weight phenotype. ³² Overall, these findings reveal the undeniable importance of PP2A in regulating metabolism, especially in the periphery. However, there is a lack of evidence on whether PP2A is also involved in central control of metabolism. Likewise, no information is available regarding the role PP2A plays in the neuroendocrine system under overnutrition situations.

The current study shows that PP2Ac in tanycytes is critical in modulating lipid metabolism by regulating cytokine expression in an AMPK-dependent manner. Our findings propose that tanycytic PP2Ac may uniquely function as a lipid metabolism regulator, providing a new target for the efficacious treatment of obesity.

Material and methods

Animals

All procedures were carried out in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (Permit Number: 2022-842). Studies are reported after the ARRIVE guidelines (Animal Research: Reporting In Vivo Experiments). ³⁵ Forty male C57BL/6J mice were used in this study, with twenty-four of them being adults (six-week-old), and sixteen being pups (8-day-old). The twenty-four adults were divided randomly into the Chow or HFD groups, while the sixteen pups were used for tanycytes cultures. No mice were excluded, and every effort was made to minimize the harm and suffering of the animals.

Diet-induced obesity differs in male and female mice, likely due to the influence of sex hormones such as estrogen.^36,37 To avoid the potential impact of sex steroids, male mice were used in our study. Six-week-old C57BL6/J male mice (stock no. 000664) were obtained from Lingchang Bio-Tech and housed in the animal facility of Zhejiang University. The mice were maintained under a 12-hour (h) light-dark cycle and had free access to a normal chow and water. The mice were randomly assigned to either the Chow or HFD group. The feeding regimen was consistent with a previously described one. ³⁸ It comprised a standard low-fat chow diet (Chow) (16 kcal% fat, 5053, LabDiet) and a high-fat diet (60 kcal% Fat, D12492, Research Diet) containing approximately 10% sucrose.

Primary tanycyte cultures and cell treatments

Tanycytes were extracted from the hypothalamus region close to the ARC/ME in 8-day-old male pups, as described. ³⁹ The brain tissue underwent a 15-minute (min) incubation with 0.25% trypsin (15090-057, ThermoFisher Scientific) at 37 °C. The trypsin digestion was halted by adding a medium containing DMEM (11965-092, ThermoFisher Scientific) and 10% FBS (10099-141, ThermoFisher Scientific). The hypothalamic glial cells were then transferred to a culture dish pre-coated with poly-L-lysine (PDL, P6407, Sigma-Aldrich). The cells were allowed to culture for four weeks, during which the medium was replaced every three days. Upon reaching confluence, contaminating cells were removed from the dish by shaking it at 250 rpm for 6 h. The medium was then changed, and the dish was shaken again for 18 h at 250 rpm at 37 °C. The isolated tanycytes were subsequently seeded onto coverslips or plates coated with PDL.

Tanycytes were treated with palmitate (100 μM, P9767, Sigma-Aldrich) coupled to a fatty acid-free BSA (10775835001, Sigma-Aldrich) vehicle for 6 h or a specific duration as outlined in the figure legend. In addition, cells were pre-treated for 0.5 h before the palmitate stimulation with a PP2A inhibitor (LB100, 2.5 μM, S7537, Selleck), ⁴⁰ an AMPK inhibitor (Compound C, HY-13418A, 10 μM), ⁴¹ a HIF-1α inhibitor (LW6, HY-13671, 10 μM), ⁴² and the controls were incubated with the vehicle.

Tanycytes transfection with siRNA

Tanycytes were transfected with siRNA (75 nM) to determine the effect of reduced PP2Ac expression on lipid metabolism. The following siRNA sequences were designed and purchased for the OBiO technology: siRNA-PP2Ac forward 5′-GGAAUUAGAUGACACUUUAAATT-3′ and reverse 5′-UUUAAAGUGUCAUCUAAUUCCTT-3′, and siRNA-Vector forward 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse 5′-ACGUGACACGUUCGGAGAATT-3′. siRNAs were transfected using Lipofectamine 2000 (Invitrogen, 11668-019) according to the manufacturer’s instructions. After 36 h, tanycytes were harvested for protein and immunostaining analysis.

Immunofluorescence

Immunostaining was performed as previously described. ⁴³ Briefly, mouse brains were fixed by perfusion and immersion with 4% PFA and then were equilibrated in 30% sucrose in PBS. Coronal sections (30-μm) were cut using a freezing microtome (CM30503, Leica) and washed thrice in 0.1 M TBS. These sections were then incubated in a blocking solution (10% donkey serum, 10% BSA, and 0.3% Triton X100 in TBS) at room temperature (RT) for 1 h. Next, the sections were incubated in rabbit anti-Iba1 (01919741, Wako), rat anti-NG2 (MA5-24247, Invitrogen), rabbit anti-GFAP (Z0334, Dako), rabbit anti-HuC/D (ab232416, Abcam), chick anti-vimentin (AB5733, Merck), and mouse anti-PP2A C subunit (05-421, Merck) in the blocking solution for 24 h at 4 °C. After being washed with 0.1 M TBS, the sections were exposed to suitable secondary antibodies (Alexa Fluor 488 anti-mouse IgG, A10026; Alexa Fluor 546 anti-rabbit IgG, A10040; Alexa Fluor 647 anti-chick IgG, A78952; Alexa Fluor 647 anti-rat IgG, A48272; Invitrogen) and DAPI (4083, Cell signaling technology) for 1 h at RT. After being washed in TBS three times, the sections were coverslipped for fluorescent image acquisition. A Nikon A1 laser-scanning confocal microscope with ×10/0.45 NA, ×20/0.75 NA, and ×60/1.4 NA objectives was utilized to capture 1024 × 1024 Z-stacked images. All image acquisition processes kept the gain, threshold, and black levels constant. The maximal intensity projection images were generated and analyzed using the ImageJ software (version 1.53t) to determine the mean area and integrated intensity of the fluorescence within the region of interest (ROI). The image analysis was blind to the experimental condition. The colocalization analysis was assessed in ImageJ using the JACoP plugin. ⁴⁴ The Intensity Correlation Quotient (ICQ) based on Li's approach was employed for evaluating the degree of colocalization. ⁴⁵ The ICQ indicates whether the intensity of staining for the two proteins varies in synchrony. An ICQ value of 0.5 means colocalization, whereas an ICQ value of −0.5 implies exclusion. Random staining and images marred by noise will produce an ICQ score close to zero.

The cell culture was fixed with 4% PFA for 10 min. The subsequent steps, including washing, blocking, and antibody incubation, were performed as described above. Additional primary antibodies used are listed as follows: rabbit anti-pP65 (Ser536) (3033, Cell signaling technology), rabbit anti-Cleaved Caspase-3 (9664, Cell signaling technology), rabbit anti-pPLCγ1 (Tyr783) (14008, Cell signaling technology), rabbit-pAMPKα (Thr172) (2535, Cell signaling technology), mouse anti-HIF-1α (66730-1-lg, Proteintech) and mouse anti-HIF-1α (66730-1-Ig, Proteintech). The mean fluorescence intensity of each cell was determined using the cellular segmentation method in ImageJ software via the Cellpose plugin. ⁴⁶

Intracellular Ca²⁺ signals

Measurements of intracellular Ca²⁺ were identical to those previously described. ⁴⁷ The tanycytes were loaded with the Ca²⁺-sensitive fluorescent dye Fluo-4 AM (1 μM, F14201, Invitrogen) and incubated for 30 min at 37 °C. Before imaging, cells were washed with dye-free DMEM medium and then incubated for 20 min. Images were obtained using a confocal microscope and completed within 10 min. The cells were kept under 5% CO2/95% air conditions throughout the imaging process at 37 °C.

PP2A activity

The enzymatic activity of PP2A in tanycytes was assessed using the PP2A Immunoprecipitation Phosphatase Assay Kit (17-313, Merck) according to the manufacturer’s instructions. The relative PP2A activity data are presented as the total PP2A activity normalized to the total PP2A protein level.

Western blotting and proteome profiler antibody array

The lysate of tanycytes was collected with RIPA buffer containing a protease inhibitor cocktail. The total protein concentration in the lysate supernatant was determined using a BCA Protein Assay Kit (23225, ThermoFisher Scientific) after centrifuging the lysate at 12,000 × g for 20 min at 4 °C. Two hundred and 20 μg of total protein were used for the Proteome Profiler antibody array and Western blot analyses, respectively. Tanycytic cytokines were analyzed using the Proteome Profiler antibody array (ARY013, R&D Systems) according to the provided protocol. In the western blot assay, the protein samples were separated using a 4–12% gradient SDS-PAGE gel (ET15412Gel, ACE) and were transferred onto polyvinylidene fluoride (PVDF) membranes (IPVH00010, Millipore) at a constant current of 300 mA for a specific duration (approximately 1 min/kDa). The membranes were then blocked in 5% fatty acid-free milk for 2 h at RT and incubated with the primary antibody overnight at 4 °C. The primary antibody used are listed as follows: rabbit anti-AMPKα Antibody (2532, Cell signaling technology), rabbit anti-AMPKα (2793, Cell signaling technology), Rabbit anti-pAKT (Ser473) (4060, Cell signaling technology), Rabbit anti-AKT (4691, Cell signaling technology), mouse anti-β-actin (3700, Cell signaling technology), PP2Ac (2259, Cell signaling technology), rabbit anti-DPPIV (ab187048, Abcam), rabbit anti-HGF (26881-1-AP, Proteintech), rabbit anti-VEGF (ab46154, Abcam), and rabbit anti-FGF1 (17400-1-AP, Proteintech). The next day, the membranes were washed thrice and incubated with either anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibody (31430, ThermoFisher Scientific) or anti-rabbit HRP-conjugated secondary antibody (31460, ThermoFisher Scientific) for 1 h at RT. After being washed thrice, positive signals were visualized using ECL-Plus Western blotting detection reagents (1863096, 1863097, ThermoFisher Scientific). The intensity of the positive signals was quantified using the ImageJ software, and the results were normalized to reference antibodies within the array or to the β-actin bands.

Statistical analysis

The results are presented as mean ± SD, and the number of independent experiments is specified in the figure legends. Statistical analysis and display were done using GraphPad Prism (version 9.0.0 (121)). The data normality and homogeneity of variances were assessed using the Shapiro-Wilk and Brown-Forsythe tests, respectively. One-way followed by Tukey’s post hoc multiple comparison, two-way ANOVA followed by Sidak’s post hoc comparison, or two-tailed Student’s t-test was used for normal data. Kruskal-Wallis one-way ANOVA test with Dunn's multiple comparisons test or two-tailed Mann Whitney Rank sum test was used for testing non-normal data and non-homogeneous variances.

Results

We first scrutinized the expression pattern of PP2Ac in the mediobasal hypothalamus of adult mice. Immunostaining analysis revealed a substantial and expansive presence of PP2Ac immunoreactivity in neuronal and glial cells, with a notable distribution along the ventricular region that lines the 3 V and the ventral hypothalamic areas such as ARC and ME (Figure 1). The cellular localization of PP2Ac was further investigated through double immunostaining with cell type-specific markers, namely Vimentin (a tanycytic marker), Iba1 (a microglial marker), GFAP (an astrocytic marker), NG2 (an oligodendrocyte precursor cell (OPCs) marker), and HuC/D (a neuronal marker). The results showed that PP2Ac was expressed in Iba1-positive microglia, NG2-positive OPCs, and HuC/D-positive neurons in the ME, but it was only minimally expressed in the ARC (Figure 1(b) and (c)). PP2Ac was also detected in Vimentin-positive ependymal cells and tanycytes, specifically in both α- and β-tanycytic bodies along the 3 V and their basal processes that penetrate into the ME. PP2Ac, however, was rarely seen in GFAP-positive, Vimentin-negative astrocytes (Figure 1(d)). These findings indicate that PP2Ac is expressed in hypothalamic cells close to the 3 V, including tanycytes.

Figure 1.

PP2Ac is highly expressed in the hypothalamus. (a) Vimentin (red) and GFAP (green) immunoreactivity in a coronal section of the mediobasal hypothalamus in adult mice. The DMH (dorsomedial hypothalamus), VMH (ventromedial hypothalamus), ARC (arcuate nucleus), ME (median eminence), and 3 V (the third ventricle) are labeled. (b) Immunoreactivity of PP2Ac (red), Iba1 (cyan), and NG2 (green) immunostaining in the hypothalamus. The middle panels show a high magnification of the yellow dotted box in the left panel, and the solid white boxes in the middle were further magnified in the right panel. (c) Low magnification and high magnification of PP2A (red) and HuC/D (neuronal marker, green) in the hypothalamus and (d) Immunostaining for PP2Ac (gray), Vimentin (red), and GFAP (green) in the wall of the 3 V (the top panel) and the ME (the bottom panel).

Abundant evidence has highlighted the role of tanycytes in sensing nutritional status to regulate energy homeostasis. To evaluate whether the tanycytic PP2Ac is involved in energy metabolism, we first analyzed its expression in overnutrition. Interestingly, we observed a discrepancy in the expression profile of PP2Ac in the VMH and ME regions of mice fed chow or HFD (Figure 2(a)). Specifically, the mean fluorescence density of PP2Ac exhibited a transient increase in the VMH but decreased in the ARC and ME after one month of HFD intake. These alterations vanished after 3 months of HFD exposure (Figure 2(b) and (c)). We further analyzed the tanycytic PP2Ac using the intensity correlation quotient (ICQ) of Vimentin and PP2Ac. The results indicated that the colocalization of PP2Ac and α-tanycytes was significantly enhanced after feeding an HFD for 1 and 3 months, compared to chow-fed mice (Figure 2(d)). On the other hand, HFD intake led to a significant rise in the expression of PP2Ac in β-tanycytes located in the ARC/ME after one month. However, the colocalization decreased slightly after three months (Figure 2(e)). It is noteworthy that the consumption of a HFD caused morphological remodeling of α-tanycytes and β-tanycytes, as their processes shortened or disappeared (Figure 2(f) and (g)). After feeding with HFD, mice exhibited a marked elevation in body weight compared to chow-fed mice (Figure 2(h)). We further assessed the activity of AMPK in the tanycytes. AMPK, a cellular energy sensor, substantially impacts the regulation of food intake and body weight, and its activation requires phosphorylation of the Thr172 residue in the α-subunit. ⁴⁸ As shown in Figure 2(i) and (j), the colocalization of p-AMPK and vimentin in the ME significantly increased after 1 month of HFD consumption compared to the chow-fed mice.

Figure 2.

Tanycytic PP2Ac levels respond to overnutrition. (a) Representative micrographs show PP2Ac and Vimentin along the 3 V and the ME of mice fed chow or HFD. (b) and (c) Mean fluorescence intensity of PP2Ac distribution in VMH, ME, and ARC of mice fed chow or HFD. (b), n = 32-45 slices from 5 mice per group. (c), n = 51-58 slices from 5 mice per group. (d) and (e) Colocalization analysis of PP2Ac with α-tanycyte or β-tanycyte based on the Intensity Correlation Quotient (ICQ) in mice fed chow or HFD. (d), n = 64–75 α-tanycytes from 4 mice per group. (e), n = 72–94 β-tanycytes from 5 mice per group. (f) and (g) The average length of tanycytic processes in the ROIs (solid white boxes in (a)) in the wall of the 3 V of ME and VMH from mice fed with chow or HFD. (f), n = 31 ROIs from 5 mice per group; (g), n = 39–41 ROIs from 5 mice per group. (h) Body weight of mice fed chow or HFD. n = 12 mice per group. (i) and (j) Representative micrographs and colocalization analysis of pAMPK with β-tanycyte based on the ICQ in mice fed chow or HFD, n = 46 ROIs from 5 mice per group. Values are mean ± SD. ***p < 0.001, **p < 0.01, *p < 0.05. Mean statistics are calculated using two-way ANOVA with post hoc Šídák's multiple comparisons test in the (b)–(h) and Welch’s t-test in the (j).

These data suggest a potential link between tanycytic PP2Ac and energy metabolism. Nonetheless, due to the complexity of the signaling pathways involving PP2A and the interference of the neurohormonal system in vivo, comprehending the interplay between energy metabolism and tanycytic PP2Ac is a challenge.

To determine the direct effects of PP2Ac on lipid metabolism in tanycytes, we conducted an in vitro mouse model of tanycytes. The immunochemistry of tanycytic markers and the absence of astrocytic marker expression verified the enrichment of the culture (Figure 3(a) to (c)). Cultured tanycytes expressed PP2Ac, as displayed by immunolabeling PP2Ac in a filamentous arrangement around the nucleus. PP2Ac was also detected within the nucleus, as shown in Figure 3(d). We examined the activity and expression of PP2Ac in cultured tanycytes in response to PA stimulation. The activity of PP2Ac was increased at 2 h after PA addition (Figure 3(g)). However, the expression of PP2Ac was not altered until 6 h of treatment (Figure 3(e) and (f)). PP2Ac kinase activity and protein levels remained elevated to the end of the analytic window (24 h). These results indicate that the activation of tanycytic PP2Ac is rapidly intensified upon short-term exposure to FFAs.

Figure 3.

PA stimulation results in the accumulation and activation of PP2A in tanycytes. (a) Photomicrographs showing the morphology of primary tanycytes using differential interference contrast (DIC) imaging by a confocal microscope. (b) Representative micrographs of vimentin (tanycyte marker, red) and GFAP (astrocyte marker, green) with DAPI (blue) staining. (c) The relative cell number of tanycytes and astrocytes in primary cultures of murine tanycytes, n = 62 images from 4 independent experiments. (d) PP2Ac in cultured tanycytes, the white circle in the right panel is the outline of the DAPI, a label for the nucleus. (e and f) Representative fluorescence images and quantification of PP2Ac and PP2Ac activity (g) in cultured tanycytes exposed to PA for 2–24 h. (b), n = 132–178 cells/group from three independent experiments; (c), n = 4 independent experiments per group. Values indicate mean ± SD. ***p < 0.001, **p < 0.01, *p < 0.05. Mean statistics are calculated using the two-tailed Mann-Whitney Rank sum test in (f) and the two-tailed unpaired t-test in (c) and (g).

To decipher the PP2Ac-associated lipid metabolism during the early phase of PA exposure, we investigated the canonical signaling pathways induced by PA in tanycytes. Lipids may increase the intracellular Ca²⁺ concentration in tanycytes by inducing Ca²⁺ release from the endoplasmic reticulum (ER) in a PLCγ1-dependent manner. ¹ Indeed, we found that PA increased Ca²⁺-dependent Fluo-4 fluorescence, along with a significant upregulation of the levels of Tyr783-phosphorylated PLCγ1 (Figure 4(a) and (b)). We then assessed whether AMPK, a Ca²⁺-dependent metabolic “master switch” that accelerates catabolism by controlling autophagy and enhancing fatty acid β-oxidation, was activated by PA treatment. Immunolabeling confirmed that PA significantly elevated the pAMPK level in tanycytes 6 h post PA addition (Figure 4(c) and (d)). The expression of cleaved-caspase 3 (CC3) and the translocation of p-P65 to the nucleus were not affected by PA treatment for 6 h, in comparison to cells treated with a vehicle (Figure 4(e) to (g)). This evidence suggests that PA triggers AMPK-related responses but not apoptosis or inflammation at an early stage.

Figure 4.

AMPK signaling is activated by PA exposure. (a) Representative fluorescence images and quantification (b) of Fluo-4 AM and pPLCγ1 in primary tanycytes after 6 h of PA treatment. Fluo-4 AM, n = 369–376, pPLCγ1, n = 224–242. (c) Representative fluorescence images and quantification (d) of pAMPK in primary tanycytes after 6 h of PA treatment. n = 194–210. (e) Representative photomicrographs and quantification of the CC3 (cleaved-caspase 3) (f), and the ratio of pP65 in the nucleus (g), CC3, n = 231–251, pP65, n = 195–207. Data are presented as mean ± SD. ***p < 0.001, ns, not significant. Mean statistics are calculated using the two-tailed Mann Whitney Rank sum test.

It has been reported that AMPK can suppress AKT activity in a PP2A-dependent manner in response to energy depletion and adiponectin treatment. ⁴⁹ The Western blotting analysis showed that PA treatment led to an elevation of phosphorylation of AMPK at Thr172 (Figure 5(a) and (b)) and a simultaneous downregulation of phosphorylation of AKT at Ser473 (Figure 5(a) and (d)), while it did not affect the levels of AMPK (Figure 5(c)) and AKT (Figure 5(e)). Pretreatment with LB100, an antagonist of PP2A, effectively blocked the increase in PP2Ac activity induced by PA (Figure 5(f)). Similarly, the expression of p-AKT was restored by LB100 (Figure 5 (a) and (d)). However, LB100 did not affect the expression of p-AMPK at 6 h post-PA treatment (Figure 5(a) and (b)). These results indicate that PA-induced activation of PP2Ac can modulate the phosphorylation of AKT but not of AMPK. Subsequently, we observed that compound C, an inhibitor of AMPK, effectively blocked PA-induced activation of AMPK (Figure 5(g), (h), (k) and (m)) yet did not alter the protein levels of AMPK (Figure 5(n)). Surprisingly, compound C significantly decreased the expression (Figure 5(i), (j), (k), and (l)) and enzymatic activity (Figure 5(o)) of PP2Ac in tanycytes during PA exposure, suggesting that PA-activated AMPK enhances the activity of PP2A.

Figure 5.

PA induces the activation of PP2Ac through AMPK activation in tanycytes. (a) Immunoblotting analysis and quantification of p-AMPK (b), AMPK (c), p-AKT (d), and AKT (e) in primary tanycytes treated with BSA, PA, and LB100 for 6 h; n = 8 independent experiments per group. (f) PP2Ac activity in primary tanycytes treated with BSA, PA, and LB100 for 6 h; n = 5 independent experiments per group. (g) Representative photomicrographs and quantification (h) of pAMPK in primary tanycytes treated with BSA, PA, and Compound C for 6 h; n = 578–608 cells/group from four independent experiments. (i) Representative fluorescence images and quantification (j) of PP2Ac in primary tanycytes treated with BSA, PA, and Compound C for 6 h; n = 532–584 cells/group from four independent experiments. (k) Immunoblotting analysis and quantification of PP2Ac (l), pAMPK (m), and AMPK (n) in primary tanycytes treated with BSA, PA, and Compound C for 6 h; n = 4 independent experiments per group. (o) PP2Ac activity in primary tanycytes treated with BSA, PA, and Compound C for 6 h; n = 5 independent experiments per group. Data are presented as mean ± SD. ***p < 0.001, **p < 0.01, *p < 0.05, ns, not significant. Mean statistics are calculated using one-way ANOVA with post hoc Tukey's multiple comparisons test in (b), (e), (f), (l), (m), (n) and (o), Kruskal-Wallis one-way ANOVA test with Dunn's multiple comparisons test in (c), (d), (h), and (j).

We next investigated the effects of activating tanycytic PP2Ac after exposure to PA. Our analysis of the cytokine arrays showed a significant difference in tanycytic cytokines in the presence or absence of PA (Supplemental Figure 1). Notably, the cytokines, including dipeptidyl peptidase IV (DPPIV), fibroblast growth factor 1 (FGF1), hepatocyte growth factor (HGF), and vascular endothelial growth factor (VEGF), were elevated upon PA treatment. However, this elevation was suppressed by LB100 (Figure 6(a), Supplemental Figure 1 (d)-(g)). These findings suggest that the activation of PP2Ac is essential for lipid metabolism within tanycytes by influencing the expression of cytokines, such as VEGF, DPPIV, FGF1, and HGF.

Figure 6.

The activation of PP2Ac induces cytokines by elevating the expression of HIF-1α. (a and b) Representative western blots and quantification of VEGF in primary tanycytes treated with PA and LB100 for 6 h, n = 4 independent experiments per group. (c–e) Representative photomicrographs and quantification of HIF-1α in whole cell (d) and nucleus (e) of primary tanycytes treated with PA and LW6, and LB100 for 6 h, n = 507–578 cells/group in the (d), and n = 316–325 cells/group in the (e) from four independent experiments. (f and g) Representative western blots and quantification of VEGF in primary tanycytes treated with PA and LW6 for 6 h, n = 4 independent experiments per group. (h and i) Representative micrographs and colocalization analysis of HIF-1α with β-tanycyte based on the ICQ in mice fed chow or HFD, n = 47 ROIs from 5 mice per group. (j–l) Representative photomicrographs and quantification of HIF-1α in whole cell (k) and nucleus (l) of primary tanycytes transfected with vector or siPP2Ac and treated with vehicle or PA for 6 h, the white circle (j) is the outline of the DAPI. n = 259–278 cells/group in the (k), and n = 245–270 cells/group in the (l) from four independent experiments. (m–o) Immunoblotting analysis and quantification of HIF-1α and VEGF in primary tanycytes transfected with vector or siPP2Ac and treated with vehicle or PA for 6 h, n = 6 per group. Data indicate mean ± SD. ***p < 0.001, **p < 0.01, *p < 0.05. Mean statistics are calculated using one-way ANOVA with post hoc Tukey's multiple comparisons test in (b), (g), (n), and (o), Kruskal-Wallis one-way ANOVA test with Dunn's multiple comparisons test in (d), (e), (k), and (l), and Welch’s t-test in (i).

We further explored the mechanisms underlying the enhanced production of cytokines following activating PP2Ac in tanycytes. It has been documented that various cytokines are transcriptional targets of the hypoxia-inducible factor-1α (HIF-1α). ⁵⁰ Activation of PP2A leads to increased HIF-1α protein levels and activity under hypoxia. ⁵¹ Therefore, we presumed that PA-induced activation of PP2Ac might potentially regulate the expression of HIF-1α, resulting in increased levels of cytokines. Immunostaining of HFD-fed mice showed an increase in HIF-1α immunoreactivity within the β-tanycytes of the ME (Figure 6(h)). Since HIF-1α is rapidly degraded under normoxic conditions via ubiquitination mechanisms by the prolyl-hydroxylase 2 (PHD2), tanycytes treated with vehicle only displayed weak HIF-1α immunoreactivity (Figure 6(c)). Consistent with our hypothesis, the accumulation of HIF-1α in both the whole cell and the nucleus was elevated upon PA exposure. In contrast, pretreatment with LB100 reduced this effect (Figure 6(c) to (e)). When PP2Ac was activated, LW6, an inhibitor of HIF-1α, decreased the level of HIF-1a (Figure 6(c) to (e)) and VEGF (Figure 6(f) and (g)) in the tanycytes (Figure 6(c) to (g)). The expression profile of HIF-1α reflected the expression patterns of VEGF in the tanycytes. To confirm the effects of pharmacological modulation of PP2Ac, we utilized small interfering RNAs (siRNAs) to deactivate PP2Ac genetically. Transfection of tanycytes with PP2Ac siRNAs decreased the level of PP2Ac compared to cells transfected with the siRNA vector (Supplemental Figure 2). Consistent with the effect of LB100, the loss of PP2Ac diminished the levels of HIF-1α (Figure 6(j) to (n)) and VEGF (Figure 6(m) and (o)) compared to cells transfected with the vector upon PA treatment. These data indicate that primary tanycytes show a PP2Ac-dependent increase in HIF-1α following PA treatment.

These results indicate that the activation of PP2Ac triggers the HIF-1α signaling pathway, inducing the production of cytokines in response to PA stimulation.

Discussion

Evidence to date has indicated that the influence of PP2A on lipid and glucose metabolism is predominantly in the liver and adipose tissue. However, obesity is caused by a prolonged disruption of energy homeostatic equilibrium, which relies on the coordination between the CNS and the peripheral system. One study has discovered that intraperitoneal injection of LB100 for 6 weeks has a significant anti-obesity effect. ⁵² As LB100 can freely cross the BBB, this hints at the potential involvement of PP2A in the central control of metabolism. Although the expression of PP2Ac is seen in OPCs, astrocytes, microglia, and other cells in the hypothalamus, and these cells are sensitive to FFAs,^{53 –56} excessive FFAs-induced PP2A activation in tanycytes is not negligible as tanycytes play a significant role in regulating lipid metabolism in the hypothalamus. Tanycytes, a peculiar radial glia found in hypothalamic regions, are responsible for various physiological processes, including establishing the blood-hypothalamic barrier and controlling the neuroendocrine axis.^8,57 Interestingly, our results reveal a fast and sustained response of PP2Ac in tanycytes following exposure to FFAs. Of note, PP2A activation coordinates lipid metabolism in tanycytes, resulting in the upregulation of an array of cytokines that are key to lipid metabolism. Although further studies are required to understand the related regulatory mechanisms and the impact of tanycytic PP2Ac on energy balance and metabolism in vivo, our results unveil a previously unknown role of tanycytic PP2Ac in cellular function and lipid metabolism in response to overnutrition.

The effects characteristic of PP2A on energy metabolism are linked to its specific biochemical and structural features. As a vital component of the network of phospho-regulation, PP2A controls and balances the Ser/Thr phosphorylation in most cell types. ²¹ The functional complexity and diversity of PP2A were determined by the B subunits encoded by fifteen genes from four families. These genes determine the subcellular localization and substrate specificity of individual PP2A heterotrimers and display different functions in lipid metabolism in peripheral tissues.^24,33 As protein phosphorylation and dephosphorylation are involved in various phases of lipid metabolism, including the uptake, storage, and utilization of FFAs, multifunctional PP2A is thus crucial to regulating tanycytic lipid metabolism.

Firstly, PP2A may regulate the uptake of long-chain fatty acids (LCFAs). Transporting LCFAs requires membrane-associated proteins, including the plasma membrane-associated fatty acid-binding protein (FABPpm), fatty acid transport proteins (FATPs, CD36), and the vesicular transporting system. ⁵⁸ Multiple cells, including ependymoglia, ⁵⁹ take up LCFAs through CD36. By binding with the extracellular FABPpm, CD36 acts as a transporter and/or a regulator of fatty acid transport. CD36 phosphorylation at Thr92 and Ser237, mediated by protein kinase A (PKA) and C (PKC), significantly reduces PA uptake. ⁶⁰ On the other hand, CD36 can be dephosphorylated by PP2A, ⁶¹ indicating that PP2A may promote PA transmembrane transporting, although there is limited information regarding whether CD36 is expressed in tanycytes. Furthermore, it is widely recognized that clathrin- or caveolin-mediated endocytosis participates in the internalization of LCFAs. The α- and β-tanycytes express clathrin and caveolin, and their endocytosis of LCFAs is also regulated by phosphorylation and dephosphorylation of key proteins. Specifically, PP2A dephosphorylates the β1 subunit of the adaptor protein complex-1 (AP-1) and initiates clathrin assembly. ⁶² Alternatively, caveolae-mediated endocytosis in the tanycytes may involve caveolin-1, which has been shown to interact with PP2A in caveolae, thereby augmenting endocytosis and the subsequent signaling cascade. ⁶³

Secondly, PP2Ac kinase activity is related to the formation of lipid droplets and the subsequent β-oxidation in tanycytes. Tanycytic lipid droplets can be detected as early as two hours following FFA treatment. ¹ PP2Ac kinase activity was elevated at 2 h after PA addition, earlier than the protein levels. One critical issue is how PP2Ac is activated. It is noteworthy that the intracellular Ca²⁺ concentration in tanycytes is increased by PA treatment, which results from the influx of extracellular Ca²⁺ and the release of Ca²⁺ from the intracellular store. ⁶⁴ PP2Ac can be precisely activated by the elevation of intracellular Ca^{2+. 65} Moreover, it has been found that PP2Ac is activated in tanycytes post-PA treatment, while phosphorylation of AMPK is also increased at the same time. It is confusing because AMPK, a serine/threonine protein kinase, is usually dephosphorylated by PP2Ac. Furthermore, inhibiting PP2Ac does not reduce the phosphorylation of AMPK, while inhibiting AMPK decreases the expression and activity of PP2Ac. ⁶⁶ These data suggest that PP2Ac may have been activated during PA exposure, at least in part, by AMPK. This is consistent with a previous study showing that AMPK directly interacts with PP2A and promotes the auto-dephosphorylation of PP2Ac at Tyr307, thereby increasing the activity of PP2A. ⁴⁹ In turn, the increased expression of PP2Ac induced by acute nutrient stress upregulates genes involved in autophagy and lipid metabolism. ⁶⁷ However, AMPK activation may also rely on the inactivation or downregulation of PP2Ac.^68,69 Inhibiting PP2A overactivation induced by long-term PA treatment has been shown to activate AMPK, thus accelerating lipid metabolism. ⁵² Thus, PP2Ac might differentially regulate lipid metabolism in response to short-term vs. long-lasting PA exposures. Our data imply that the feedback regulation of AMPK by PP2Ac activation in tanycytes may be the link between HFD-induced hypothalamic dysfunction and whole-body energy homeostasis.

Compound C, an inhibitor of AMPK, has been widely used to confirm AMPK-dependent mechanisms.^49,66,70 Our results showed that it hindered the activation of PP2Ac in tanycytes under PA treatment in vitro. In addition, it has also been reported that compound C suppresses the bone morphogenetic protein (BMP) type 1 receptor, thereby blocking osteogenic differentiation in vivo. PP2A dephosphorylates the BMP signaling, resulting in an amplification of the pathway. ⁷¹ Further research is necessary to determine if this process is related to lipid metabolism in tanycytes.

Once inside the tanycytes, β-oxidation of lipids remains sustained for six hours. ¹ However, excessive oxidation of FFAs increases cellular oxygen consumption, sufficient to trigger HIF-1α induction. ⁷² The prolyl-hydroxylases 2 (PHD2) strictly control the degradation of HIF-1α under physiological oxygen conditions. After hydroxylation by PHD, the HIF-1α subunit is ubiquitinated by the E3 ubiquitin ligase complex that contains the Von Hippel-Lindau protein (VHL) and is ultimately degraded by the 26S proteasome. LW6, an inhibitor of HIF-1α, stimulates the expression of VHL and accelerates the proteasomal degradation of HIF-1α. ⁴² LW6 was also identified as a potent breast cancer resistance protein (BCRP) inhibitor and could be useful in reducing multidrug resistance during the resistance of cancer cells. ⁷³ Nevertheless, BCRP inhibition does not seem to be responsible for the PA-induced production of VEGF in our study, because both pharmacological experiments and PP2A knockdown confirm that PA-induced PP2Ac activation increases the level of VEGF through the elevation of HIF-1α. PP2A directly dephosphorylates PHD2 at Ser125, reducing PHD2 activity that ultimately boosts HIF-1α accumulation. ⁵¹ Thus, the profile of HIF‐1α expression under PA stimulation was well correlated to PP2Ac activity. ⁷⁴

Thirdly, the tanycytic metabolism of FFAs, in which PP2A plays a critical role, may potentially convey the metabolic information from tanycytes to hypothalamic neurons.^{2,75 –77} Tanycytes monitor and respond to the fluctuations of circulating FFAs and induce the production and secretion of intermediate metabolites, hormones, and cytokines.^1,19,20 These factors, such as DPPIV, FGF1, HGF, and VEGF, are transcriptionally regulated by HIF-1α ⁷⁸ and are involved in energy expenditure, insulin signal transmission, and vascular homeostasis.^{79 –81} In addition, tanycytes release ATP and lactate via monocarboxylate transporters (MCTs), purinergic receptors, and connexin 43 (CX43) channels. ATP and lactate can fuel the neighboring neurons and participate in neuronal signaling pathways.^82,83 Indeed, nutrients and hormones evoke a considerable Ca²⁺ surge in tanycytes. This Ca²⁺ wave, propagated through tanycytic processes, can transmit the information to CX43-expressing neurons and non-neurons.^84,85 The interaction between PP2A and CX43 in astrocytes, endothelial cells, and myocardial cells is critical for the expression and distribution.^{86 –88}

Interestingly, the cytokine levels in the LB100 group are significantly lower than in the vehicle group. We speculate that other PP2A-dependent signaling events may account for the action of LB100 in parallel. It has been shown that PA treatment activates the Wnt/β-catenin pathway, which is essential for VEGF transcription. ⁸⁹ PP2A can positively regulate Wnt signaling by dephosphorylating β-catenin and increasing its level. ⁹⁰ In addition, v-Src is known to trigger the transcription of VEGF, ⁹¹ and PP2A further amplifies this process. ⁹² Therefore, it is possible that LB100 may reduce VEGF transcription by blocking the Wnt/β-catenin pathway and the v-Src-dependent process.

Finally, PP2Ac activation following short-term PA exposure is not involved in an inflammatory reaction in our experimental condition in vitro, as illustrated by the pP65 immunostaining and the cytokine profiles. However, HFD can cause an inflammatory response in the hypothalamus, which may result in structural and functional damage to the BBB at the ME-ARC interface. This abnormal permeability of the BBB is associated with changes in the arrangement and spread of β-tanycytes. ⁹³ Moreover, systemic inflammatory cytokines such as IL-1β can activate NF-κB in α-tanycytes, thus inducing the expression of Cox-2 and the release of anorexigenic prostaglandins from tanycytes. ⁹⁴ PP2Ac, however, may participate in anti-inflammation under persistent PA exposure or high-fat diet conditions. PP2A has been shown to play a role in anti-inflammation by regulating the Toll-like receptors/NF-κB signaling pathway at multiple levels. For example, a macrophage-specific disruption of the mouse Ppp2ca gene enhanced the susceptibility to LPS-induced inflammation, whereas an activator of PP2A, FTY720, decreased the production of pro-inflammatory mediators.^24,95,96

Limitations of the study

Although our work established a critical role for PP2Ac in tanycytic lipid metabolism, there are areas for improvement due to technological restrictions. We did not examine mice with the tanycyte-specific PP2Ac knockdown. An animal model that allows for flexible, reversible, and cell-specific modulation of PP2Ac activity would strengthen our conclusion. In addition, while we have endeavored to dissect the tanycytes, specifically those of the β type, from the hypothalamic ARC/ME regions, accurately distinguishing the subtypes of β-tanycytes in culture presents a challenge. As different subtypes of β-tanycytes exhibit different physiological functions, it is essential to document the metabolic characteristics of each subtype and the impact of tanycytic PP2Ac on BBB and neuronal functions in future research.

Conclusion

In this study, we describe the role of tanycytic PP2Ac in response to short-term lipid stress. Excessive FFAs induced accumulation and activation of tanycytic PP2Ac, leading to elevated cytokine levels crucial for lipid metabolism by increasing the expression of HIF-1α. Our results suggest that increasing PP2Ac activity might provide a means to treat obesity.

Supplemental Material

sj-pdf-1-jcb-10.1177_0271678X231219115 - Supplemental material for Saturated fatty acids stimulate cytokine production in tanycytes via the PP2Ac-dependent signaling pathway

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X231219115 for Saturated fatty acids stimulate cytokine production in tanycytes via the PP2Ac-dependent signaling pathway by Danyang Liu, Tao Wang, Xingqi Zhao, Juan Chen, Tianqi Yang, Yi Shen and Yu-Dong Zhou in Journal of Cerebral Blood Flow & Metabolism