Prolyl carboxypeptidase in Agouti-related Peptide neurons modulates food intake and body weight

Abstract

Objective: Prolyl carboxypeptidase (PRCP) plays a role in the regulation of energy metabolism by inactivating hypothalamic α-melanocyte stimulating hormone (α-MSH) levels. Although detected in the arcuate nucleus, limited PRCP expression has been observed in the arcuate POMC neurons, and its site of action in regulating metabolism is still ill-defined.

Methods: We performed immunostaining to assess the localization of PRCP in arcuate Neuropeptide Y/Agouti-related Peptide (NPY/AgRP) neurons. Hypothalamic explants were then used to assess the intracellular localization of PRCP and its release at the synaptic levels. Finally, we generated a mouse model to assess the role of PRCP in NPY/AgRP neurons of the arcuate nucleus in the regulation of metabolism.

Results: Here we show that PRCP is expressed in NPY/AgRP-expressing neurons of the arcuate nucleus. In hypothalamic explants, stimulation by ghrelin increased PRCP concentration in the medium and decreased PRCP content in synaptic extract, suggesting that PRCP is released at the synaptic level. In support of this, hypothalamic explants from mice with selective deletion of PRCP in AgRP neurons (Prcp^AgRPKO) showed reduced ghrelin-induced PRCP concentration in the medium compared to controls mice. Furthermore, male Prcp^AgRPKO mice had decreased body weight and fat mass compared to controls. However, this phenotype was sex-specific as female Prcp^AgRPKO mice show metabolic differences only when challenged by high fat diet feeding. The improved metabolism of Prcp^AgRPKO mice was associated with reduced food intake and increased energy expenditure, locomotor activity, and hypothalamic α-MSH levels. Administration of SHU9119, a potent melanocortin receptor antagonist, selectively in the PVN of Prcp^AgRPKO male mice increased food intake to a level similar to that of control mice.

Conclusions: Altogether, our data indicate that PRCP is released at the synaptic levels and that PRCP in AgRP neurons contributes to the modulation of α-MSH degradation and related metabolic control in mice.

Keywords: Energy metabolism; Food intake; NPY/AgRP; Prolyl carboxypeptidase; α-MSH.

Figure 1

PRCP expression in NPY/AgRP neurons and its secretion. (A–C) Representative photographs of the hypothalamic arcuate nucleus showing GFP (NPY-GFP mouse; A) and PRCP immunolabeling (B). Analysis of double labeled cells (C) showed that 77.8 ± 2.39% of NPY neurons were positive for PRCP (arrowheads). (D) Graph showing the result of the PRCP measurements in hypothalamic explants of mice containing the PVN and the anterior arcuate nucleus. Ghrelin addition (100 nM) to the basal (8 mM glucose) aCSF induced a significant increase of PRCP in the medium compared to either basal aCSF or 15 mM glucose-containing aCSF (n = 7 per treatment). (E) Graph showing the result of the PRCP measurements in hypothalamic explants of Prcp^AgRPKO and control mice (n = 5 per group) containing the PVN and the anterior arcuate nucleus. Compared to controls, PRCP levels in the medium after ghrelin incubation were significantly lower in Prcp^AgRPKO mice. (F and G) Western blot images (F) and quantification of the density (G) of PRCP in cytoplasmic extracts from hypothalamic explants (n = 4 per group) incubated with either 8 mM Glucose, 15 mM Glucose or 8 mM Glucose plus 100 nM Ghrelin for PRCP, β-actin and GluR2. PRCP density was normalized to β-actin density. (H and I) Western blot images (H) and quantification of the density (I) of PRCP in synaptic extracts from hypothalamic explants (n = 4 per group) incubated with either 8 mM Glucose, 15 mM Glucose or 8 mM Glucose plus 100 nM Ghrelin for PRCP, β-actin and GluR2. PRCP density was normalized to β-actin density. 3v = third ventricle; ARC = arcuate nucleus; ME = median eminence. Bar scale in A (for all panels) represents 50 μm. Data represent the mean ± SEM.* = P < 0.05; ** = P < 0.01; *** = P < 0.001.

Figure 2

Selective PRCP deletion in AgRP neurons affects metabolic phenotype in male mice. (A–C) Graphs showing body weight (A), fat mass (B), and lean mass (C) of 3-month-old male Prcp^flox/flox-AgRP-cre negative control mice (n = 5), Prcp^+/+-AgRP-cre control mice (n = 10), and Prcp^AgRPKO male mice (n = 10). * = P < 0.05 compared to Prcp^+/+-AgRP-cre mice; ** = P < 0.01 compared to Prcp^+/+-AgRP-cre mice; # = P < 0.05 compared to Prcp^flox/flox-AgRP-cre negative mice; ## = P < 0.01 compared to Prcp^flox/flox-AgRP-cre negative mice. (D and E) Graphs showing food intake in control mice (n = 9) and Prcp^AgRPKO mice (n = 8). Results of food intake as total in the 24-h cycle and in the dark and light phases of the cycle. Gray area represents dark phases. (F) Graph showing food intake in male Prcp^AgRPKO and control mice (n = 4/5 per group) at 1 h, 2 h and 3 h after ip ghrelin administration. (G–L) Graphs showing locomotor activity (G), energy expenditure (H and I), O₂ consumed (J), CO₂ produced (K), and their ratio (L) of three months old male control (n = 9) and Prcp^AgRPKO mice (n = 8). All data are represented as mean ± SEM. * = P < 0.05; ** = P < 0.01; *** = P < 0.001; **** = P < 0.0001 compared to control mice.

Figure 3

Effects of PRCP deletion in AgRP neurons on circulating hormone levels in male mice. (A) Graph showing serum total ghrelin levels (A) of 3-month-old male control (n = 5), and Prcp^AgRPKO mice (n = 5) on fed and fasted state. (B) Graph showing serum active ghrelin levels of male control (n = 10), and Prcp^AgRPKO mice (n = 5) on fed and fasted states. (C) Graph showing circulating corticosterone levels in male control (n = 5) and Prcp^AgRPKO mice (n = 8) on fed and fasted state. (D and E) Graphs showing the analysis of serum free T₃ (D) and free T₄ (E) levels in male control (n = 5), and Prcp^AgRPKO mice (n = 5) on fed state. (F and G) Graphs showing Real Time PCR data for Ucp1 (F) and Dio2 (G) in the brown adipose tissue of 3 months old male control and Prcp^AgRPKO mice (n = 5/4 per group). All data are represented as mean ± SEM. * = P < 0.05 compared to control mice; ** = P < 0.01 compared to control mice; **** = P < 0.0001 compared to control mice.

Figure 4

Selective PRCP deletion in AgRP neurons improves glucose metabolism in male mice. (A) Graph showing glucose tolerance test in 3-month-old male control and Prcp^AgRPKO mice (n = 10 per group). (B and C) Graphs showing circulating insulin (B) and glucagon (C) levels during glucose tolerance test in control male control (n = 5) and Prcp^AgRPKO mice (n = 5). (D and E) Graphs showing mRNA levels of liver Pepck (D) and G6Pase (E) in male control (n = 5) and Prcp^AgRPKO mice (n = 5) in fed state. All data are represented as mean ± SEM. * = P < 0.05 compared to control mice; ** = P < 0.01 compared to control mice.

Figure 5

α-MSH and â-endorphin levels in Prcp^AgRPKO mice. (A) Graph showing the α-MSH levels (expressed as fmol α-MSH/mg protein) in the hypothalamus of male Prcp^AgRPKO mice (n = 9) compared to control mice (n = 9). (B) Graph showing the β-endorphin levels in the hypothalamus of male Prcp^AgRPKO mice (n = 8) compared to controls (n = 9). (C and D) Graphs showing α-MSH and β-endorphin levels in the hypothalamus of female Prcp^AgRPKO (n = 6) compared to controls mice (n = 6). (E and F) Graphs showing PC1 (E) and PC2 (F) mRNA levels in the hypothalamus of male Prcp^AgRPKO (n = 5) compared to control mice (n = 5). Data represent the mean ± SEM. * = P < 0.05 compared to control mice.

Figure 6

Role of PRCP deletion in AgRP neurons in the paraventricular nucleus of the hypothalamus. (A and B) Representative hypothalamic sections from a control male mouse (A) and a Prcp^AgRPKO male mouse (B) fasted overnight and immunostained for cfos (red) in the hypothalamic paraventricular nucleus (PVN). (C) Quantification of cfos expression in hypothalamic paraventricular nucleus neurons of fasted control male mice (n = 3) and fasted Prcp^AgRPKO male mice (n = 3). (D and E) Representative micrographs of hypothalamic sections showing immunostaining for cfos from a fasted control female mouse (n = 3) and a fasted Prcp^AgRPKO female mouse (n = 3) in the PVN. (F) Graph showing the quantification in hypothalamic paraventricular nucleus neurons immunostained for cfos in fasted control female mice (n = 3) and fasted Prcp^AgRPKO female mice (n = 3). (G and H) Representative light micrographs showing immunostained á-MSH fibers in the PVN of a fasted control male mouse (n = 4) and a fasted Prcp^AgRPKO male mouse (n = 4). (I) Graph showing the integrated density quantification of á-MSH fibers in the PVN of fasted control and Prcp^AgRPKO male mice (n = 4 per group). (J and K) Representative micrographs of hypothalamic sections from a fasted control female mouse (n = 4) and a fasted Prcp^AgRPKO female mouse (n = 4) immunostained for á-MSH fibers in the PVN. (L) Graph showing the quantification of the fluorescent density of á-MSH fibers in the PVN of fasted control female mice (n = 4) and fasted Prcp^AgRPKO female mice (n = 4). 3v = third ventricle; PVN = paraventricular nucleus of the hypothalamus. Bar scale in E (for A, B, D, G, H, J and K) represents 100 ìm. All data are represented as mean ± SEM. ** = P < 0.01compared to control mice; *** = P < 0.001 compared to control mice.

Figure 7

Selective SHU9119 administration in the PVN increases feeding and neuronal activation in male Prcp^AgRPKO mice. (A) Graph showing feeding response after bilateral PVN injections of SHU9119 in 3-month-old male control and Prcp^AgRPKO mice (n = 5 per group). (B and C) Representative hypothalamic sections from a control male mouse (B) and a Prcp^AgRPKO male mouse (C) injected with SHU9119 immunostained for cfos (red) in the hypothalamic paraventricular nucleus (PVN). (D) Quantification of cfos expression in hypothalamic paraventricular nucleus neurons of control male mice (n = 5) and Prcp^AgRPKO male mice (n = 5) injected with SHU9119. 3v = third ventricle; PVN = paraventricular nucleus of the hypothalamus. Bar scale in C (for B) represents 100 ìm. All data are represented as mean ± SEM. *** = P < 0.001; **** = P < 0.0001.