Postnatal hyperosmolality alters development of hypothalamic feeding circuits with context-specific changes in ingestive behavior

Abstract

Drinking and feeding are tightly coordinated homeostatic events and the paraventricular nucleus of the hypothalamus (PVH) represents a possible node of neural integration for signals related to energy and fluid homeostasis. We used TRAP2;Ai14 mice and Fos labeling to visualize neurons in the PVH and median preoptic nucleus (MEPO) responding to both water deprivation and feeding signals. We determined that structural and functional development of dehydration-sensitive inputs to the PVH precedes those of agouti-related peptide (AgRP) neurons, which convey hunger signals and are known to be developmentally programmed by nutrition. Moreover, we found that osmotic hyperstimulation of neonatal mice led to enhanced AgRP inputs to the PVH in adulthood, as well as disruptions to ingestive behaviors during high-fat diet feeding and dehydration-anorexia. Thus, development of feeding circuits is impacted not only by nutritional signals, but also by early perturbations to fluid homeostasis with context-specific consequences for coordination of ingestive behavior.

Keywords: Developmental biology; Developmental neuroscience; Neuroscience.

Figure 1

Drinking and feeding signals converge in the MEPO and PVH of adult mice (A and B) Representative confocal images showing TRAPped (tdTomato+) cells in the MEPO of both EUH-TRAP (A) and Thirst-TRAP (B) mice. A region of interest (ROI) is indicated by the white rectangle. (C) Quantification of number of cells expressing tdTomato+ in the MEPO ROI. (D and E) Representative confocal images showing TRAPped (tdTomato+) cells in the PVH of both EUH-TRAP (D) and Thirst-TRAP (E) mice. ROIs within the magnocellular and parvocellular compartments are indicated by the white rectangles. (F) Quantification of cells expressing tdTomato+ in the magnocellular ROI of the PVH. (G) Quantification of cells expressing tdTomato+ in the parvocellular ROI of the PVH. (H and I) Representative confocal images showing Fos+ cells in the MEPO in response to a fast-refeed stimulus (Hunger-Fos) in both EUH-TRAP (H) and Thirst-TRAP (I) mice. (J) Quantification of immunohistochemical (IHC) analysis of cells expressing Fos+ in the MEPO ROI. (K and L) Representative confocal images showing Fos+ cells in the PVH in response to a fast-refeed stimulus (Hunger-Fos) in both EUH-TRAP (K) and Thirst-TRAP (L) mice. (M) Quantification of IHC analysis of cells expressing Fos+ in the magnocellular ROI of the PVH. (N) Quantification of IHC analysis of cells expressing Fos+ in the parvocellular ROI of the PVH. (O and P) Representative confocal images showing co-labeled tdTomato+/Fos+ cells in the MEPO in both EUH-TRAP (O) and Thirst-TRAP (P) mice. (Q) Quantification of IHC analysis of co-labeled tdTomato+/Fos+ cells in the MEPO ROI. (R and S) Representative confocal images showing co-labeled tdTomato+/Fos+ cells in the PVH in both EUH-TRAP (R) and Thirst-TRAP (S) mice. (T) Quantification of IHC analysis co-labeled tdTomato+/Fos+ cells in the magnocellular compartment of the PVH. (U) Quantification of IHC analysis of co-labeled tdTomato+/Fos+ cells in the parvocellular compartment of the PVH. All data are represented as mean ± SEM and data points are quantified across 1-2 sections for individual animals. Thirst-TRAP (n = 3), and EUH-TRAP (n = 3), where n represents the number of mice; Unpaired t-test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Scale bars, 100μm. See also Figures S1 and S2.

Figure 2

Postnatal developmental projections from the MEPO to the PVH (A) Schematic of experiment design, showing brains from P8 male mice were sectioned from rostral to caudal so as to expose the MEPO and single DiI crystals were implanted in the MEPO using an insect needle, the brains were then placed back into fixative and incubated for 5–6 weeks, and then sectioned on a vibratome. (B–D) Confocal images showing Hoechst Dye and implant location of DiI crystal in the MEPO. (E–G) Confocal images showing Hoechst Dye and downstream robust DiI axonal labeling in the PVH. Images are collected from one animal. Scale bars, 200μm.

Figure 3

Ontogeny of circuits responding to osmotic stimuli (A–D) Representative confocal images showing Fos+ cells in the MEPO of neonatal mice across ages P1 (A), P8 (B), P16 (C), and P30 (D) treated with a subcutaneous (s.c.) administration of 0.1mL/10g BW of 0.9% NaCl (isotonic saline, IS) as a control. (E–H) Representative confocal images showing Fos+ cells in the MEPO of neonatal mice at ages P1 (E), P8 (F), P16 (G), and P30 (H) treated with s.c. administration of 0.1mL/10g BW of 2.0M NaCl as a dehydration stimulus (hypertonic saline, HS). (I) Quantification of IHC analysis of Fos+ cells in the MEPO at P1. IS (n = 3), and HS (n = 4). (J) Quantification of IHC analysis of Fos+ cells in the MEPO at P8. IS (n = 5), and HS (n = 6). (K) Quantification of IHC analysis of Fos+ cells in the MEPO at P16.IS (n = 8), and HS (n = 9). (L) Quantification of IHC analysis of Fos+ cells in the MEPO at P30. IS (n = 6), and HS (n = 7). (M–P) Representative confocal images showing Fos+ cells in the PVH of neonatal mice across ages P1 (M), P8 (N), P16 (O), and P30 (P) treated with an s.c. administration of 0.1mL/10g BW of 0.9% NaCl as a control (IS). (Q–T) Representative confocal images showing Fos+ cells in the PVH of neonatal mice at ages P1 (Q), P8 (R), P16 (S), and P30 (T) treated with s.c. administration of 0.1mL/10g BW of 2.0M NaCl as a dehydration stimulus (HS). (U) Quantification of IHC analysis of Fos+ cells in the PVH at P1. IS (n = 5) and HS (n = 4). (V) Quantification of IHC analysis of Fos+ cells in the PVH at P8. IS (n = 6) and HS (n = 6). (W) Quantification of IHC analysis of Fos+ cells in the PVH at P16. IS (n = 8) and HS (n = 9). (X) Quantification of IHC analysis of Fos+ cells in the PVH at P30. IS (n = 6) and HS (n = 7). (Y) Comparison of Fos+ cells in the MEPO of HS mice across ages. One-way ANOVA with multiple comparisons. (Z) Comparison of Fos+ cells in the MEPO of HS mice across ages. One-way ANOVA with multiple comparisons. All data are represented as mean ± SEM and data points are quantified across 1 section for individual animals; n represents the number of animals; Unpaired t-test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Scale bars, 100μm.

Figure 4

Postnatal perturbation to fluid homeostasis impacts development of AgRP projections to the MEPO and PVH (A and B) Representative confocal images showing AgRP fiber densities in the MEPO of adult mice treated daily from P5-P15 with either s.c administration of 0.9% NaCl as a control (IS^PN; A) or 2.0M NaCl as a dehydration stimulus (HS^PN, B). Regions of interest (ROIs) are indicated by the white rectangles. (C) Quantification of IHC analysis of AgRP fiber densities in the MEPO ROI of IS^PN and HS^PN mice using the spots feature in Imaris software. IS^PN (n = 13) and HS^PN (n = 11). (D) Quantification of IHC analysis of AgRP fiber densities in the MEPO ROI of IS^PW and HS^PW mice using the spots feature in Imaris software. IS^PW (n = 3) and HS^PW (n = 3). (E and F) Representative confocal images showing AgRP fiber densities in the PVH of IS^PN (E) and HS^PN (F) mice. (G) Quantification of IHC analysis of AgRP fiber densities in the PVH ROI using the spots feature in Imaris software. IS^PN (n = 13), and HS^PN (n = 12). (H) Quantification of IHC analysis of AgRP fiber densities in the PVH ROI of IS^PW and HS^PW mice using the spots feature in Imaris software. IS^PW (n = 3) and HS^PW (n = 3). (I and J) Representative confocal images showing β-endorphin fiber densities in the MEPO of adult IS^PN (I) and HS^PN (J) mice. Regions of interest (ROIs) are indicated by the white rectangles. (K) Quantification of IHC analysis of β-endorphin fiber densities in the MEPO ROI using the spots feature in Imaris software. IS^PN (n = 8), and HS^PN (n = 7). (L) Quantification of IHC analysis of β-endorphin fiber densities in the MEPO ROI of IS^PW and HS^PW mice using the spots feature in Imaris software. IS^PW (n = 3) and HS^PW (n = 3). (M and N) Representative confocal images showing β-endorphin fiber densities in the PVH of IS^PN (M) and HS^PN (N) mice. (O) Quantification of IHC analysis of β-endorphin fiber densities in the PVH ROI using the spots feature in Imaris. IS^PN (n = 8), and HS^PN (n = 7). (P) Quantification of IHC analysis of β-endorphin fiber densities in the PVH ROI of IS^PW and HS^PW mice using the spots feature in Imaris software. IS^PW (n = 3) and HS^PW (n = 3). All data are represented as mean ± SEM and data points are quantified across 1 section for individual animals; n represents the number of animals; Unpaired t-test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Scale bars, 50μm. See also Figure S3.

Figure 5

Hyperstimulation of drinking circuits during postnatal development increases water intake in the context of an HFD (A) Quantification of cumulative food intake during the 168h (7 days) of HFD exposure in IS^PN (n = 6) and HS^PN (n = 6) animals. two-way ANOVA: no significant effect of treatment over time (p = 0.7423). Data are presented as group mean values ±SEM. (B) Comparison of final cumulative food intake during HFD exposure. IS^PN (n = 6) and HS^PN (n = 6). (C) Quantification of cumulative water intake during the HFD exposure. IS^PN (n = 6) and HS^PN (n = 6). two-way ANOVA: a significant main effect of treatment and time (p < 0.0001). Data are presented as group mean values ±SEM. (D) Comparison of final cumulative water intake during HFD exposure. IS^PN (n = 6) and HS^PN (n = 6). Data in (B, D) are represented as mean ± SEM and data points are individual animals; n represents the number of animals; Unpaired t-test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. See also Figure S4.

Figure 6

Hyperstimulation of drinking circuits during postnatal development leads to a sustained dehydration-anorexic response and a decrease in water intake after a fast-refeed in adult mice (A) Quantification of daily food intake during the inactive light cycle of IS^PN mice (n = 8) and HS^PN mice (n = 7) for the acclimation period, dehydration period (DEH), and rehydration period (REH). two-way ANOVA: no significant effect of treatment over time. Data are presented as group mean values ±SEM. (B) Quantification of nightly food intake during the active dark cycle of IS^PN (n = 8) and HS^PN (n = 7) for the acclimation period, dehydration period (DEH), and rehydration period (REH). two-way ANOVA: no significant effects of treatment over time. Data are presented as group mean values ±SEM. (C) Quantification of food intake 48h before the dehydration period (acclimation). IS^PN (n = 8) and HS^PN (n = 7). (D) Quantification of average food intake for the 48h dehydration period (DEH). IS^PN (n = 8) and HS^PN (n = 7). (E) Quantification of food intake for the first 48h following rehydration (REH). IS^PN (n = 8) and HS^PN (n = 7). (F) Quantification of water intake during the inactive light cycle before a fast, during the fast, and after refeeding. (IS^PN (n = 7) and HS^PN (n = 8). (G) Quantification of water intake during the active dark cycle before a fast, during the fast, and after refeeding. (IS^PN (n = 7) and HS^PN (n = 8). Data for (C–G) are represented as mean ± SEM and data points are individual animals; n represents the number of animals; Unpaired t-test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. See also Figure S5.