TET3 epigenetically controls feeding and stress response behaviors via AGRP neurons
Di Xie, Bernardo Stutz, Feng Li, Fan Chen, Haining Lv, Matija Sestan-Pesa, Jonatas Catarino, Jianlei Gu, Hongyu Zhao, Christopher E. Stoddard, Gordon G. Carmichael, Marya Shanabrough, Hugh S. Taylor, Zhong-Wu Liu, Xiao-Bing Gao, Tamas L. Horvath, Yingqun Huang
Di Xie, Bernardo Stutz, Feng Li, Fan Chen, Haining Lv, Matija Sestan-Pesa, Jonatas Catarino, Jianlei Gu, Hongyu Zhao, Christopher E. Stoddard, Gordon G. Carmichael, Marya Shanabrough, Hugh S. Taylor, Zhong-Wu Liu, Xiao-Bing Gao, Tamas L. Horvath, Yingqun Huang
View: Text | PDF
Research Article Metabolism Neuroscience

TET3 epigenetically controls feeding and stress response behaviors via AGRP neurons

Abstract

The TET family of dioxygenases promote DNA demethylation by oxidizing 5-methylcytosine to 5-hydroxymethylcytosine (5hmC). Hypothalamic agouti-related peptide–expressing (AGRP-expressing) neurons play an essential role in driving feeding, while also modulating nonfeeding behaviors. Besides AGRP, these neurons produce neuropeptide Y (NPY) and the neurotransmitter GABA, which act in concert to stimulate food intake and decrease energy expenditure. Notably, AGRP, NPY, and GABA can also elicit anxiolytic effects. Here, we report that in adult mouse AGRP neurons, CRISPR-mediated genetic ablation of Tet3, not previously known to be involved in central control of appetite and metabolism, induced hyperphagia, obesity, and diabetes, in addition to a reduction of stress-like behaviors. TET3 deficiency activated AGRP neurons, simultaneously upregulated the expression of Agrp, Npy, and the vesicular GABA transporter Slc32a1, and impeded leptin signaling. In particular, we uncovered a dynamic association of TET3 with the Agrp promoter in response to leptin signaling, which induced 5hmC modification that was associated with a chromatin-modifying complex leading to transcription inhibition, and this regulation occurred in both the mouse models and human cells. Our results unmasked TET3 as a critical central regulator of appetite and energy metabolism and revealed its unexpected dual role in the control of feeding and other complex behaviors through AGRP neurons.

Authors

Di Xie, Bernardo Stutz, Feng Li, Fan Chen, Haining Lv, Matija Sestan-Pesa, Jonatas Catarino, Jianlei Gu, Hongyu Zhao, Christopher E. Stoddard, Gordon G. Carmichael, Marya Shanabrough, Hugh S. Taylor, Zhong-Wu Liu, Xiao-Bing Gao, Tamas L. Horvath, Yingqun Huang

×

Figure 4

TET3 promotes the association of a chromatin-modifying complex with the Agrp/AGRP promoters.

Options: View larger image (or click on image) Download as PowerPoint
TET3 promotes the association of a chromatin-modifying complex with the ...
(A and B) STAT3- and FOXO1-binding sites labeled in red and green, respectively. Numbers depict starting and ending positions of nucleotides in chromosomes. PCR primers for the zoomed-in ChIP/hMeDIP regions are underlined. (C) Mice were injected with AAV-sgTet3 or AAV on day 1 (D1). On D21, the mice were fasted 22 hours and treated with leptin or saline 2 hours, followed by ARC isolation. (D) Mice were treated as in C. ARCs from 4 mice/group were pooled for ChIP-qPCR. n = 3/group, technical replicates. *P < 0.05, **P < 0.01, ***P < 0.001, 1-way ANOVA with Tukey post test. (E) SH-SY5Y cells transfected with NT siRNA were maintained in Lept L (NT siRNA/Lept L) or Lept H (NT siRNA/Lept H) or were transfected with TET3 siRNA and maintained in Lept H (TET3 siRNA/Lept H). Cells were collected 36 hours after transfection for ChIP-qPCR. Data presented as percentage of input. n = 3/group, technical replicates. *P < 0.05, **P < 0.01, ***P < 0.001, 1-way ANOVA with Tukey post test. (F) Mice were fasted 22 hours, followed by leptin injection and isolation of ARCs 2 hours after leptin injection. ARCs from 3 mice were pooled for co-IP. Representative immunoblots shown, protein sizes in kDa on right. The band labeled with an asterisk is likely an isoform of TET3. (G) Representative immunoblots of co-IP from GT1-7 cells maintained in Lept H. Samples in lanes 2 and 3 were run on the same gel but were noncontiguous. (H) Representative immunoblots of co-IP from SH-SY5Y cells maintained in Lept H. (I) Mice were treated as in C. ARCs from 2 mice/group were pooled for ChIP-qPCR. n = 3, technical replicates. **P < 0.01, ***P < 0.001, 1-way ANOVA with Tukey post test. (J) SH-SY5Y cells were treated as in E, followed by ChIP-qPCR. n = 3/group, technical replicates. **P < 0.01, ***P < 0.001, 1-way ANOVA with Tukey post test. (K) Mice were treated as in C. ARCs from 2 mice in each group were pooled for hMeDIP-qPCR. n = 3/group, technical replicates. **P < 0.01, 1-way ANOVA with Tukey post test. (L) hMeDIP-qPCR of SH-SY5Y cells transfected with NT siRNA or TET3 siRNA in Lept H 36 hours. n = 3/group, technical replicates. **P < 0.01, 2-tailed Student’s t tests. (M) Representative micrographs and statistical analysis of p-STAT3+ (green) AGRP neurons (red) from fed mice. n = 5/group. Two-tailed Student’s t test. Scale bars: 50 μm. Data represent mean ± SEM.