Restriction of food intake by PPP1R17-expressing neurons in the DMH

Abstract

Leptin-deficient ob/ob mice eat voraciously, and their food intake is markedly reduced by leptin treatment. In order to identify potentially novel sites of leptin action, we used PhosphoTRAP to molecularly profile leptin-responsive neurons in the hypothalamus and brainstem. In addition to identifying several known leptin responsive populations, we found that neurons in the dorsomedial hypothalamus (DMH) of ob/ob mice expressing protein phosphatase 1 regulatory subunit 17 (PPP1R17) constitutively express cFos and that this is suppressed by leptin treatment. Because ob mice are hyperphagic, we hypothesized that activating PPP1R17 neurons would increase food intake. However, chemogenetic activation of PPP1R17 neurons decreased food intake and body weight of ob/ob mice while inhibition of PPP1R17 neurons increased them. Similarly, in a scheduled feeding protocol that elicits increased consumption, mice also ate more when PPP1R17 neurons were inhibited and ate less when they were activated. Finally, we found that pair-feeding of ob mice reduced cFos expression to a similar extent as leptin and that reducing the amount of food available during scheduled feeding in DMH^Ppp1r17 neurons also decreased cFos in DMH^Ppp1r17 neurons. Finally, these neurons do not express the leptin receptor, suggesting that the effect of leptin on these neurons is indirect and secondary to reduced food intake. In aggregate, these results show that PPP1R17 neurons in the DMH are activated by increased food intake and in turn restrict intake to limit overconsumption, suggesting that they function to constrain binges of eating.

Keywords: DMH; body weight; food intake; leptin; metabolism.

Fig. 1.

Differential enrichment of genes after leptin treatment. PhosphoTRAP was performed on the samples indicated below, and the RNA abundance determined by the number of RNA-seq reads was plotted in pair-wise comparisons showing (A) differential enrichment of genes in pS6 immunoprecipitates determined by RNA-seq in 14 d leptin treatment of ob/ob mice vs. PBS in (A) hypothalamus and (B) brainstem. TaqMan assays were then performed for the pS6-precipitated polysomes after 14 d PBS treatment of ob/ob mice after leptin treatment (for 2, 4, 7, and 14 d) of ob mice and after 14 d PBS treatment of WT mice. The genes shown are those whose level of enrichment had changed in the comparison between 14 d of leptin vs. PBS (specific genes are shown in A and B): (C) Ppp1r17, (D) POMC, (E) AGRP, (F) Tph2, (G) Npw, and (H) CRH. Data are expressed as the ratio of fold enrichment (IP per input) for each group of mice divided by the fold enrichment (IP per input) for ob/ob mice treated with PBS for 14 d. Homogenates of pooled hypothalami from 15 to 20 mice are used for each group.

Fig. 2.

pS6 expression after leptin treatment of ob mice. Dual ISH and IHC in hypothalamus and brainstem was performed for the marker genes and pS6 (244 and 247) as indicated in sections from ob mice treated with leptin (14 d) vs. PBS. (A) POMC: ISH for Pomc and IHC for pS6 in arcuate nucleus. Pomc neurons in Arc are activated by 14 d leptin treatment of ob/ob mice. (B) Tph2: ISH for Tph2 and IHC for pS6 in dorsal raphe. Tph2 neurons in dorsal raphe are activated by 14 d leptin treatment of ob/ob mice. (C) CRH: ISH for Crh and IHC for pS6 in inferior colliculus. Crh neurons in inferior colliculus are activated by 14 d leptin treatment of ob/ob mice. (D) Ppp1r17: ISH for Ppp1r17 and IHC for pS6 in DMH. Ppp1r17 neurons in DMH are inhibited by 14 d leptin treatment of ob/ob mice. (Scale bars, 50 μm.)

Fig. 3.

Effect of chemogenetic modulation of DMH^Ppp1r17 neurons on food intake and body weight in WT mice. (A) Chemogenetic activation of DMH^Ppp1r17 neurons using AAV8-hSyn-DIO-hM3D(Gq)-mCherry. DREADD-induced activation of DMH^Ppp1r17 neurons significantly decreases food intake over the course of 4 h. (B) Chemogenetic inhibition of DMH^Ppp1r17 neurons using AAV8-hSyn-DIO-hM4D(Gi)-mCherry. DREADD-induced inhibition of DMH^Ppp1r17 neurons does not affect food intake over the course of 4 h. (C) Body weight of Ppp1r17-Cre mice is significantly reduced after chronic activation of DMH^Ppp1r17 neurons by i.p. injection of CNO two times per day. (D) Body weight of Ppp1r17-Cre mice does not significantly change after chronic inhibition of DMH^Ppp1r17 neurons by i.p. injection of CNO two times per day (*P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, two-tailed unpaired t test; n = 5 to 6 mice). All error bars are mean ± SD.

Fig. 4.

Effect of chemogenetic modulation of DMH^Ppp1r17 neurons on food intake and body weight in ob/ob mice. (A) Chemogenetic inhibition of DMH^Ppp1r17 neurons in ob/ob mice. DREADD-induced inhibition of DMH^Ppp1r17 neurons significantly increases food intake over the course of 24 h. (B) Chemogenetic activation of DMH^Ppp1r17 neurons in ob/ob mice. DREADD-induced activation of DMH^Ppp1r17 neurons significantly decreases food intake over the course of 24 h. (C) Body weights of ob/ob mice are significantly increased after chronic inhibition of DMH^Ppp1r17 neurons by i.p. injection of CNO two times per day. (D) Body weights of ob/ob mice are significantly reduced after chronic activation of DMH^Ppp1r17 neurons by i.p. injection of CNO two times per day at the end of 12 d (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, two-tailed unpaired t test; n = 5 mice). All error bars are mean ± SD.

Fig. 5.

Effect of chemogenetic modulation of DMH^Ppp1r17 neurons on food intake, body weight, and FAA during scheduled feeding. (A) DREADD-induced activation (treatment P < 0.0001) and inhibition (treatment P < 0.0001) of DMH^Ppp1r17 neurons significantly alters food intake during scheduled feeding after i.p. injection of CNO 4 h before food presentation. Two-way repeated-measures ANOVA comparing treated and control groups. (B) DREADD-induced inhibition of DMH^Ppp1r17 neurons does not affect FAA. (C) DREADD-induced activation of DMH^Ppp1r17 neurons does not affect FAA. (D) DREADD-induced activation or inhibition does not alter oxygen consumption during scheduled feeding. Two-tailed unpaired t test comparing treated (Ppp1r17-cre CNO) and control group (Ppp1r17-cre PBS; E). DREADD-induced activation or inhibition does not alter body temperature during scheduled feeding. Two-tailed unpaired t test comparing treated (Ppp1r17-cre CNO) and control group (Ppp1r17-cre PBS; n = 8 mice; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001). All error bars are mean ± SD.

Fig. 6.

cFos expression in Ppp1r17 neurons in ob mice and during scheduled feeding. (A) IHC for Ppp1r17 and TdTomato in ObRb TdTomato mice. DMH^Ppp1r17 neurons do not express Leprb. (B) cFos expression in DMH^Ppp1r17 neurons in ob/ob mice is reduced by 14 d pair-feeding. (C) cFos expression in DMH^Ppp1r17 neurons in scheduled feeding during the 3-h feeding window in response to different amounts of chow. Food is provided to mice between Circadian Time (CT)4 and CT7: (Upper) 4 g of food, the amount normally consumed during the 3-h feeding window; (Middle) 2 g of food, the amount mice consume in 3 h after a fast; and (Lower) 1 g, the amount consumed during the dark phase within 3 h. (Scale bars, 50 μm.)