Microglial UCP2 Mediates Inflammation and Obesity Induced by High-Fat Feeding

Abstract

Microglia play a crucial role in immune responses, including inflammation. Diet-induced obesity (DIO) triggers microglia activation and hypothalamic inflammation as early as 3 days after high-fat diet (HFD) exposure, before changes in body weight occur. The intracellular mechanism(s) responsible for HFD-induced microglia activation is ill defined. Here, we show that in vivo, HFD induced a rapid and transient increase in uncoupling protein 2 (Ucp2) mRNA expression together with changes in mitochondrial dynamics. Selective microglial deletion of Ucp2 prevented changes in mitochondrial dynamics and function, microglia activation, and hypothalamic inflammation. In association with these, male and female mice were protected from HFD-induced obesity, showing decreased feeding and increased energy expenditure that were associated with changes in the synaptic input organization and activation of the anorexigenic hypothalamic POMC neurons and astrogliosis. Together, our data point to a fuel-availability-driven mitochondrial mechanism as a major player of microglia activation in the central regulation of DIO.

Keywords: POMC; diet-induced obesity; hypothalamus; metabolism; microglia; mitochondrial dynamics; mitochondrial respiration; neuroinflammation; synaptic plasticity.

Figure 1.. High-Fat Diet Induces a Rapid and Transient Upregulation of Ucp2 mRNA Levels and a Rapid and Transient Change in Mitochondrial Morphology

A–E) Real-time PCR data showing relative mRNA levels of Ucp2 (A), IL-1β (B), IL-6 (C), Tnfα (D), and Cx3cr1 (E) in isolated hypothalamic microglia (CD11b⁺ cells) from male mice fed on standard chow diet (SD), 3-day high-fat diet (3d HFD), 7-day HFD (7d HFD), or 8-week HFD (8w HFD) (n = 4–6 mice per group). (F) Representative electron micrographs showing mitochondria (asterisks) in microglial cells in the arcuate nucleus of the hypothalamus of male mice exposed to SD, 3d HFD, 7d HFD, or 8w HFD. (G–I) Average mitochondrial area (G), mitochondrial density (H), and mitochondrial coverage (I), in microglial cells from male mice fed on SD (n = 5), 3d HFD (n = 6), 7d HFD (n = 5), or 8w HFD (n = 5). (J) Graph showing the percentage of phosphorylated DRP1 (p-Drp1) at site S616 (activated form) in tomato-positive cells (microglia) in the hypothalamus and cortex of mice exposed to either standard chow diet (SD; n = 3) or 3d HFD (n = 5). Scale bar in (A) (for all panels) represents 500 nm. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. For detailed statistical analyses, see Table S1.

Figure 2.. Validation of the Animal Model

A) Graph showing Ucp2 mRNA levels in isolated hypothalamic microglia (CD11b-positive cells, CD11b⁺) of male control (n = 4) and Ucp2^MGKO mice (n = 4). (B) Graph showing Ucp2 mRNA levels in isolated hypothalamic CD11b-negative (CD11b⁻) cells of control (n = 4) and Ucp2^MGKO mice (n = 4). (C) Graph showing Cx3cr1 mRNA levels in isolated hypothalamic CD11b⁺ and CD11b⁻ cells of control (n = 4) and Ucp2^MGKO mice (n = 4) showing marginal expression of Cx3cr1 mRNA expression in CD11b⁻ cells. (D) Graph showing Ucp2 mRNA levels in peripheral CD11b⁺ cells of the liver, spleen, blood, and brown adipose tissue (BAT) in control (n = 4) and Ucp2^MGKO mice (n = 4). (E) Graphs showing Cx3cr1 mRNA levels in isolated peripheral CD11b⁺ and CD11b⁻ cells of control mice (n = 4). Note the not detectable (ND) or marginal (in spleen) expression of Cx3cr1 mRNA expression in CD11b⁻ cells as validation of the data shown in (D). Data are presented as mean ± SEM. ***p < 0.001. For detailed statistical analyses, see Table S1.

Figure 3.. Selective Deletion of UCP2 Prevents HFD-Induced Mitochondrial Dynamics

(A) Representative electron microscopic pictures showing mitochondria (asterisks) in microglia cells of male Ucp2^MGKO mice fed on SD, 3d HFD, 7d HFD, or 8w HFD. (B–D) Average mitochondrial area (B), mitochondrial density (C), and mitochondrial coverage (D) in microglial cells of Ucp2^MGKO mice fed on SD (n = 4), 3d HFD (n = 6), 7d HFD (n = 6), or 8w HFD (n = 5). (E) Graph showing the percentage of phosphorylated DRP1 at site S616 (activated form) in tomato-positive cells (microglia) in the hypothalamus and cortex of male Ucp2^MGKO mice exposed to either SD (n = 3) or 3d HFD (n = 5). Scale bar in (A) (for all panels) represents 500 nm. Data are presented as mean ± SEM. For detailed statistical analyses, see Table S1.

Figure 4.. Selective Deletion of UCP2 in Microglia Affects Mitochondrial Function

(A–C) Graphs showing the OCR (A and B) and ECAR (C) under L-glutamine (2 mM) incubation (n = 4 for controls and n = 3 for Ucp2^MGKO mice) in primary microglial cell culture of control and Ucp2^MGKO mice. (D–F) Graphs showing the OCR (D and E) and ECAR (F) under L-glutamine (2 mM) and palmitate (200 μM) incubation (n = 5 for controls and n = 5 for Ucp2^MGKO mice) in primary microglial cell culture of control and Ucp2^MGKO mice. (G–I) Graphs showing the OCR (G and H) and ECAR (I) under L-glutamine (2 mM) and glucose (25 mM) incubation (n = 5 for controls and n = 5 for Ucp2^MGKO mice) in primary microglial cell culture of control and Ucp2^MGKO mice. (J–L) Graphs showing the OCR (J and K) and ECAR (L) under L-glutamine (2 mM), glucose (25 mM), and palmitate (200 μM) incubation (n = 5 for controls and n = 5 for Ucp2^MGKO mice) in primary microglial cell culture of control and Ucp2^MGKO mice. (M–O) Graphs showing basal respiration (M), maximal respiration (N), and ATP production (O) of primary microglial cells from control mice. (P–R) Graphs showing basal respiration (P), maximal respiration (Q), and ATP production (R) of primary microglial cells from Ucp2^MGKO mice. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. For detailed statistical analyses, see Table S1.

Figure 5.. Selective Deletion of UCP2 in Microglia Affects HFD-Induced Microglial Activation

A) Representative micrographs of hypothalamic sections showing tomato-expressing microglia in the mediobasal hypothalamus of male control (upper panels) and Ucp2^MGKO mice (lower panels) fed on SD (n = 8 for both groups), 3d HFD (n = 7 for both groups), 7d HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice), and 8w HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice). (B) Graph showing the number of microglia in the ARC of control and Ucp2^MGKO mice fed on SD (n = 8 for both groups), 3d HFD (n = 7 for both groups), 7d HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice), and 8w HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice). (C) Graph showing microglia size in the ARC of control and of Ucp2^MGKO mice fed on SD (n = 8 for both groups), 3d HFD (n = 7 for both groups), 7d HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice), and 8w HFD (n = 8 for controls and n = 7 for Ucp2^MGKO mice). (D) Graph showing microglia number in the VMH of control and Ucp2^MGKO mice fed on SD, 3d HFD, 7d HFD, and 8w HFD (n = 5 mice/group). (E) Graph showing microglia size in the VMH of control and Ucp2^MGKO mice fed on SD, 3d HFD, 7d HFD, and 8w HFD (n = 5 mice/group). Scale bar in (A) (upper left panel and for all panels) represents 100 μm. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. For detailed statistical analyses, see Table S1.

Figure 6.. Selective Deletion of UCP2 Prevents HFD-Induced Obesity in Male Mice

A) Graph showing body weight in male control (n = 8) and Ucp2^MGKO mice (n = 8) fed on HFD for up to 8 weeks. (B and C) Graphs showing fat mass (B) and lean mass (C) in male control and Ucp2^MGKO mice on HFD for 4 weeks. (D–L) Graphs showing energy expenditure (D–F), consumed O₂ (G and H), produced CO₂ (I and J), and the respiratory exchange ratio (RER) (K and L) in male control and Ucp2^MGKO mice on HFD for 4 weeks. (M and N) Graphs showing 24 h food intake (M; average of 3 days) and 48 h locomotor activity (N) of male control and Ucp2^MGKO mice on HFD for 4 weeks. (O and P) Graphs showing glucose tolerance test (O) and insulin tolerance test (P) of male control and Ucp2^MGKO mice on HFD for 4 weeks. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. For detailed statistical analyses, see Table S1.

Figure 7.. Selective Deletion of UCP2 in Microglia Affects POMC Synaptic Input Organization and Neuronal Activation

(A) Graphs showing the total number of synapses onto POMC neurons of male SD-fed controls (11 cells/3 mice), 8w HFD-fed controls (14 cells/3 mice), SD-fed Ucp2^MGKO mice (19 cells/4 mice), and 8w HFD-fed Ucp2^MGKO mice (17 cells/4 mice for HFD). (B) Graphs showing the number of asymmetrical (putative excitatory) synapses onto POMC neurons of male SD-fed controls (11 cells/3 mice), 8w HFD-fed controls (14 cells/3 mice), SD-fed Ucp2^MGKO mice (19 cells/4 mice), and 8w HFD-fed Ucp2^MGKO mice (17 cells/4 mice for HFD). (C) Graphs showing the number of symmetrical (putative inhibitory) synapses onto POMC neurons of male SD-fed controls (11 cells/3 mice), 8w HFD-fed controls (14 cells/3 mice), SD-fed Ucp2^MGKO mice (19 cells/4 mice), and 8w HFD-fed Ucp2^MGKO mice (17 cells/4 mice for HFD). (D and E) Representative electron micrographs showing examples of asymmetrical (D) and symmetrical (E) synapses onto POMC neurons. Arrowheads in (D) indicate postsynaptic density characteristic of asymmetrical excitatory synapsis. Asterisks indicate axon terminals. (F) Representative micrographs showing hypothalamic sections of fed male control (upper panels) and Ucp2^MGKO mice (lower panels) on HFD immunostained for POMC (left panels in green), c-Fos (middle panels in red), and merged (right panels). (G) Graph showing the percentage of c-Fos-positive POMC neurons in fed male control and Ucp2^MGKO mice (n = 4 for both groups) on HFD. (H) Graph showing no difference in total POMC cell number in fed male control and Ucp2^MGKO mice (n = 4 for both groups) on HFD. Scale bar in (E) represents 1 μm. Scale bar in (F) (upper left panel) represents 100 μm (for all panels). Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001. For detailed statistical analyses, see Table S1. 3v, 3^rd ventricle; ARC, arcuate nucleus; ME, median eminence.