Fasting activation of AgRP neurons requires NMDA receptors and involves spinogenesis and increased excitatory tone

Abstract

AgRP neuron activity drives feeding and weight gain whereas that of nearby POMC neurons does the opposite. However, the role of excitatory glutamatergic input in controlling these neurons is unknown. To address this question, we generated mice lacking NMDA receptors (NMDARs) on either AgRP or POMC neurons. Deletion of NMDARs from AgRP neurons markedly reduced weight, body fat and food intake whereas deletion from POMC neurons had no effect. Activation of AgRP neurons by fasting, as assessed by c-Fos, Agrp and Npy mRNA expression, AMPA receptor-mediated EPSCs, depolarization and firing rates, required NMDARs. Furthermore, AgRP but not POMC neurons have dendritic spines and increased glutamatergic input onto AgRP neurons caused by fasting was paralleled by an increase in spines, suggesting fasting induced synaptogenesis and spinogenesis. Thus glutamatergic synaptic transmission and its modulation by NMDARs play key roles in controlling AgRP neurons and determining the cellular and behavioral response to fasting.

Figure 1. Effects of Grin1 Deletion on AMPAR- and NMDAR-Mediated Currents

(A–D) Electrically-evoked (A and C) or spontaneous (B and D) AMPAR- and NMDAR-mediated EPSCs in AgRP neurons (A, B) and POMC neurons (C, D) within brain slices obtained from control (Grin1^lox/lox, Npy-hrGFP or Grin1^lox/lox, Pomc-hrGFP) mice and from Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox, Npy-hrGFP or Pomc-Cre, Grin1^lox/lox, Pomc-hrGFP) mice. B, right panels - quantitative analyses of AMPAR-sEPSC frequency and amplitude for control versus Agrp-ires-Cre, Grin1^lox/lox, Npy-hrGFP mice. Data are from male mice and are expressed as mean ± SEM. The number of GFP-positive neurons studied for each group is shown in parentheses.

Figure 2. Effects of Grin1 Deletion on Energy Balance

(A–D) Body weight (A) (males and females), body composition (B) (males and females), food intake (C) (males, ad lib (left) and 24 hr fasted-refeeding (right)), and respiratory exchange ratio (D) (males) of control (Grin1^lox/lox) mice and AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox) mice. (E) Body weight (males and females) of control (Grin1^lox/lox) mice and POMC neuron-specific, Grin1-deleted (Pomc-Cre, Grin1^lox/lox) mice. Data are expressed as mean ± SEM. * p < 0.05, ** p < 0.01, unpaired t test compared to controls. The number of mice studied for each group is shown in parentheses.

Figure 3. Dendritic Spines on AgRP and POMC neurons

(A–B) Visualization (A) and quantification (B) of spines on dendrites of AgRP and POMC neurons in control (Agrp-ires-Cre or Pomc-Cre, respectively) and neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox or Pomc-Cre, Grin1^lox/lox) mice. (C–D) Determination of head size (C) and neck length (D) of spines on dendrites of AgRP neurons in control (Agrp-ires-Cre) and AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox) mice. Data for each group was collected from n=3 male mice (7–9 neurons per mouse) and are expressed as mean ± SEM. ** p < 0.01, unpaired t test compared to controls.

Figure 4. Fasting-Induced Increases in c-Fos protein and Neuropeptide mRNAs in AgRP Neurons – Dependence on NMDARs

(A) Effects of fasting on c-Fos induction in AgRP neurons of control (Grin1^lox/lox, Npy-hrGFP) mice and AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox, Npy-hrGFP) mice. The number of mice studied for each group is shown in parentheses. (B–D) Effects of fasting on Agrp (B), Npy (C) and Pomc (D) mRNA expression in hypothalami of control (Grin1^lox/lox) mice, AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox) mice and POMC neuron-specific, Grin1-deleted (Pomc-Cre, Grin1^lox/lox) mice. For (B–D), the number of mice studied for each group is greater than or equal to 4. Data are from male mice and are expressed as mean ± SEM. * p < 0.05, ** p < 0.01, *** p < 0.001, one-way ANOVA followed by Tukey’s HSD post hoc test compared to controls as indicated in the figure.

Figure 5. Fasting-Induced Increase in Dendritic Spines – Dependence on NMDARs

Effects of fasting on number of spines on AgRP neuron dendrites in control (Agrp-ires-Cre) mice and in AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox) mice. Data for each group was collected from n=3 male mice (7–9 neurons per mouse) and are expressed as mean ± SEM. ** p < 0.01, unpaired t test compared to controls. Note, data shown in this Figure for the two fed groups was previously presented in Figure 3B (labeled as 1 Agrp-ires-Cre and 2 Agrp-ires-Cre, Grin1^lox/lox in that Figure) and is shown here again to permit comparison with the fasted groups.

Figure 6. Fasting-Induced AMPAR-Mediated EPSCs in AgRP Neurons – Dependence on NMDARs

(A–B) Effects of fasting on AMPAR-mediated EPSC frequencies and amplitudes (spontaneous EPSCs (A) and miniature EPSCs (B)) in AgRP neurons within brain slices obtained from control (Grin1^lox/lox, Npy-hrGFP) mice and from AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox, Npy-hrGFP) mice. Representative traces are shown in the left panels of (A) and (B). Data are from male mice and are expressed as mean ± SEM. * p < 0.05, *** p < 0.001, unpaired t test compared to controls. The number of GFP-positive neurons studied for each group is shown in parentheses.

Figure 7. Fasting-Induced Depolarization and Firing Rates in AgRP Neurons – Dependence on NMDARs

(A–B) Effects of fasting on membrane potential and firing rate of AgRP neurons within brain slices obtained from control (Grin1^lox/lox, Npy-hrGFP) mice and from AgRP neuron-specific, Grin1-deleted (Agrp-ires-Cre, Grin1^lox/lox, Npy-hrGFP) mice. Representative traces are shown in (A). (B) Bar graph summaries of membrane potential (left) and firing rate (right). Data are from male mice and are expressed as mean ± SEM. * p < 0.05, ** p < 0.01, unpaired t test compared to controls. The number of GFP-positive neurons studied for each group is shown in parentheses.

Figure 8. Reversibility of Fasting-Induced Changes in EPSCs and Spines

(A) Daily food intake in male Npy-hrGFP mice which were ad libitum fed, then fasted for 24 hrs, and then refed for 3 days. Data for each group was collected from n≥3 male mice. (B) Effects of fasting and refeeding on AMPAR-mediated spontaneous EPSC frequencies in AgRP neurons within brain slices obtained from male Npy-hrGFP mice. The number of GFP-positive neurons studied ranges from 14–18 in each group. (C) Effects of fasting and refeeding on the number of spines on AgRP neuron dendrites in male Agrp-ires-Cre mice. Spines were counted using methods identical to those employed for data in Figures 3 and 5. Spine data for each group was collected from three mice (7–9 neurons per mouse). Data in A–C are expressed as mean ± SEM with ** p < 0.01, *** p < 0.001, one-way ANOVA followed by Tukey’s HSD post hoc test compared to ad lib fed controls.