Peroxisome proliferation-associated control of reactive oxygen species sets melanocortin tone and feeding in diet-induced obesity

Abstract

Previous studies have proposed roles for hypothalamic reactive oxygen species (ROS) in the modulation of circuit activity of the melanocortin system. Here we show that suppression of ROS diminishes pro-opiomelanocortin (POMC) cell activation and promotes the activity of neuropeptide Y (NPY)- and agouti-related peptide (AgRP)-co-producing (NPY/AgRP) neurons and feeding, whereas ROS-activates POMC neurons and reduces feeding. The levels of ROS in POMC neurons were positively correlated with those of leptin in lean and ob/ob mice, a relationship that was diminished in diet-induced obese (DIO) mice. High-fat feeding resulted in proliferation of peroxisomes and elevated peroxisome proliferator-activated receptor γ (PPAR-γ) mRNA levels within the hypothalamus. The proliferation of peroxisomes in POMC neurons induced by the PPAR-γ agonist rosiglitazone decreased ROS levels and increased food intake in lean mice on high-fat diet. Conversely, the suppression of peroxisome proliferation by the PPAR antagonist GW9662 increased ROS concentrations and c-fos expression in POMC neurons. Also, it reversed high-fat feeding-triggered elevated NPY/AgRP and low POMC neuronal firing, and resulted in decreased feeding of DIO mice. Finally, central administration of ROS alone increased c-fos and phosphorylated signal transducer and activator of transcription 3 (pStat3) expression in POMC neurons and reduced feeding of DIO mice. These observations unmask a previously unknown hypothalamic cellular process associated with peroxisomes and ROS in the central regulation of energy metabolism in states of leptin resistance.

Fig. 1. Honokiol suppresses while H2O2 increases anorexigenic melanocortin tone and related feeding

a: Fluorescence double labeling for c-fos (red nuclei) and GFP (green cell bodies) shows colocalization of c-fos with GFP (white arrows) in vehicle- and honokiol-treated animals. Bar scale on upper right panel of a indicates 100 μm for all 4 panels. b: Honokiol treatment elevated the percent of c-fos/NPY-GFP double labeled neurons compared to vehicle controls. Honokiol treatment decreased the percentage of c-fos/POMC-GFP double labeled cells. * indicates P<0.05. c: Food intake was significantly induced by icv. honokiol treatment.* P<0.05. d: Upper panel: H2O2 significantly (*P<0.05) depolarized membrane potential of POMC-GFP neurons an event that was rapidly reversed by washout. Control values indicate firing of POMC neuron when exposed to vehicle. Lower panel: H2O2 significantly increased (P<0.05) the firing frequency of arcuate nucleus POMC-GFP neurons, and event that was reversed by washout. Control values indicate firing of POMC neuron when exposed to vehicle. e: Icv. administration of H202 significantly reduced food intake of mice 1h (*P<0.05), 2h (*P<0.05), 4h (*P<0.05) and 8h (*P<0.05) subsequent to injections. Control animals were injected with equivalent amount of vehicle. f: DHE (red fluorescence) in POMC neurons (green fluorescence) of ob/ob, fasted lean, fed lean and diet-induced obese (DIO) mice. g: Quantification of DHE in POMC neurons indicated the highest level of ROS production in POMC neurons of DIO and fed lean mice compared to fasted lean and ob/ob mice. *indicates significant (P<0.05) difference compared to ob/ob values. # indicates significant (P<0.05) difference compared to fasted lean values. There was no significant difference between fed and DIO values. h: Leptin levels were not detectable in ob/ob mice. Leptin levels were significantly higher in fed animals compared to fasted values (#: P<0.05). DIO animals had leptin levels that were significantly higher compared to fed ($: P<0.05) and fasted values (#: P<0.05). * indicates significant (P<0.05) difference compared to ob/ob values.

Fig. 2. Peroxisome proliferation in POMC neurons

a: Electron micrographs showing representative section of POMC-GFP perykarya in the arcuate nucleus from fasted lean (upper left panel), ob/ob (lower left panel), fed lean (upper middle panel) and DIO (lower middle panel). White arrows on the middle panels point to peroxisomes. The upper right and lower right panels are high power magnifications of peroxisomes from the fed and DIO panels, respectively. Bar scale on upper left panel indicate 1 μm for left and middle panels. Bar scale on lower right panel indicates 500 nm for the right panels. b: Mitochondria number was the highest in POMC neurons of fed and DIO mice compared to fasted and ob/ob values. *: significantly different (P<0.05) from ob/ob values. #: significantly (P<0.05) different from fasted values. c: Highest number of peroxisomes in POMC neurons was in DIO animals. Significantly lower numbers of peroxisomes were detected in fed and fasted animals, while no peroxisomes were detected in POMC neurons of ob/ob animals. *: significantly different (P<0.05) from ob/ob values. #: significantly (P<0.05) different from fasted values. $: significantly (P<0.05) different from fed values. d: PCR analyzes of PPARα, δ and γ showes that absolute levels of PPARγ is several fold higher than transcripts of PPARα and δ in the hypothalamus. e: Real time PCR analyzes of PPARα, δ and γ and various other gene transcripts related to PPAR signaling and cellular metabolism in DIO hypothalamus relative to lean control values. LPL: lipoprotein lipase, FASN: fatty acid synthase, LXRa: liver X receptor alpha, FABP1: fatty acid binding protein 1, CIDEA: cell death-inducing DFFA-like effector a, ADFP/PLIN2: adipose differentiation related protein/perilipin 2, GCK: glucokinase. Results are shown as mean ± SEM. The comparison of different groups was carried out using two-tailed unpaired Student’s t-test. * indicates P<0.05. f: PPARγ mRNA is expressed in POMC neuronal cell line mHypoA-2/28, and its relative expression level did not change after pioglitazone treatment (pio). g: Pioglitazone induced the expression of the PPARγ target gene, Gpd1 in the POMC-expressing hypothalamic cell line. h: Pioglitazone induced the expression of the PPARγ target gene, Fabp4 in the AgRP-expressing hypothalamic cell line. i: Pioglitazone induced the expression of the PPARγ target gene, Pin2 in the AgRP-expressing hypothalamic cell line. j: Real time PCR analyzes of PPARα, δ and γ and various other gene transcripts related to PPAR signaling and cellular metabolism in ob/ob hypothalamus relative to wild type controls. LPL: lipoprotein lipase, FASN: fatty acid synthase, LXRa: liver X receptor alpha, FABP1: fatty acid binding protein 1, CIDEA: cell death-inducing DFFA-like effector a, ADFP/PLIN2: adipose differentiation related protein/perilipin 2, GCK: glucokinase. Results are shown as mean ± SEM. The comparison of different groups was carried out using two-tailed unpaired Student’s t-test. * indicates P<0.05.

Fig. 3. Peroxisome proliferation in POMC neurons is associated with altered feeding

a-d: Electron micrographs with fluorescence inserts of POMC neurons from lean vehicle-treated (upper left panel), lean rosiglitazone-treated (upper right panel), DIO-vehicle treated (lower left panel) and DIO GW9662-treated (lower right panel) mice. On the electron micrographs, blue arrows point to peroxisomes. On the fluorescent inserts, red labeling indicated DHE in green POMC-GFP neurons. Bar scale on the upper left electron micrograph indicates 1 μm for all EM panels; bar scale on the fluorescent insert of the upper left panel indicate 10 μm for all inserts. e: Rosiglitazone induced increase in peroxisome number in POMC neurons of lean animals on high fat diet. On the other hand, treatment of DIO mice with the PPARγ antagonist, GW9662, decreased peroxisome number in POMC neurons. *: P<0.05). f: DHE levels were decreased in lean animals treated with rosiglitazone, while DIO animals treated with GW9662 had elevated DHE counts. *: P<0.05). g: Treatment of lean mice consuming normal chow with rosiglitazone, daily food intake of animals was significantly higher compared to vehicle-treated values. GW9662 treatment of DIO mice consuming high fat diet resulted in a decline in daily food intake. *: P<0.05. h: Double immunofluorescence labeling for c-fos (red) and POMC (green) from control DIO (upper left panel), rosigliatzone-treated (upper right panel), GW9662-treated (lower left panel) and H2O2-treated (lower right panel) high fat-fed animals. Bar scale on lowe left panel represents 100 μm for all panels of h. i: Bar graphs showing the percentage of c-fos immunolabeled POMC neurons in the different experimental groups. * indicates significant (P<0.05) difference compared to DIO control values; # indicates significant (P<0.05) differences compared to values of rosiglitazone treated animals.

Fig. 4. GW9662 reverses DIO-triggered electric activity of the melanocortin system and ROS regulates feeding behavior in DIO animals

a: Firing rate of DIO NPY/AgRP neurons is significantly decreased (*:P<0.05) by GW9662 treatment. b: Firing rate of DIO POMC neurons is significantly increased (*:P<0.05) by GW9662 treatment. c: GW9662 treatment increased the percentage of silent DIO NPY/AgRP neurons. d: GW9662 treatment decreased the percentage of silent DIO POMC neurons. e: Rosiglitazone induces food intake in lean animals. Left column indicates daily food intake after 5 days on high fat diet. Middle column shows mean daily food intake after 5 days treatment with rosiglitazone. Right column indicates daily food intake after 7 days rosiglitazone treatment. Red bar graph on right indicates daily food intake of animals with 7 day rosiglitazone treatment with H202 in the last 2 days of the 7 day treatment. * indicates significant (P<0.05) difference relative to values before rosiglitazone treatment. # indicates significant (P<0.05) differences between daily food intake values after 5 day treatment with rosiglitazone treatments. $ indicates significant (P<0.05) differences between 7 day treatment values. f: GW9662 induces suppression of food intake in DIO animals fed a high fat diet. Left column indicates daily food intake at the beginning of treatment (day 0). Middle column shows mean food intake after 5 days treatment with GW9662. Right column indicates daily food intake after 7 days DW9662 treatment. Green bar graph on right indicates daily food intake of animals with 7 day GW9662 treatment with honokiol in the last 2 days of the 7 day treatment. * indicates significant (P<0.05) difference relative to values before GW9662 treatment. # indicates significant (P<0.05) differences between daily food intake values after 5 day treatment with GW9662 treatments. $ indicates significant (P<0.05) differences between 7 day treatment values. g: 2 days icv. H202 treatment alone resulted in significantly (P<0.05) decreased daily food intake of DIO animals compared to vehicle treated controls. h: Photomicrographs of pStat3 (red; asterisks) and POMC double immuno-labeled hypothalamic sections from leptin treated DIO mice co-treated with vehicle or H2O2 after peripheral leptin injections. There was significantly (P<0.05) higher percentage of POMC neurons with pStat3 labeled nucleus (arrow) in the arcuate nucleus of H2O2 treated mice compared to vehicle-treated controls. Bar scale represents 100 μm for both panels.