Low Liver-Derived IGF-1 Drives the Alterations in Growth Hormone Secretion in Food-Restricted Male Mice

Abstract

Chronic food restriction (FR) changes growth hormone (GH) secretion from a pulsatile pattern, observed in ad libitum–fed mice, to a tonic secretion, in which basal (nonpulsatile) GH secretion prevails. However, the physiological mechanisms driving this alteration are not fully understood. We hypothesize that suppressed liver-derived insulin-like growth factor-1 (IGF-1) production may be a key underlying mechanism responsible for changing the GH secretion pattern in FR mice. To test this possibility, GH secretion patterns were investigated in ad libitum–fed hepatocyte-specific GH receptor (GHR) knockout (KO) (Albumin^ΔGHR) male mice and compared to those of ad libitum–fed and FR control male mice. As expected, serum IGF-1 and liver Igf1 messenger RNA (mRNA) expression were similarly suppressed in Albumin^ΔGHR-fed and FR wild-type (WT) mice. Plasma ghrelin did not differ between ad libitum–fed control and Albumin^ΔGHR mice, but increased in FR control mice. Like the results observed in FR animals, Albumin^ΔGHR-fed mice exhibited increases in total and basal (nonpulsatile) GH secretion without alterations in GH pulse amplitude compared to control mice. Although Albumin^ΔGHR-fed and FR WT mice both exhibited suppressed Ghr mRNA levels in the liver, there were significant differences in the hepatic expression of sexually dimorphic genes and those regulating GH sensitivity. Hepatocyte-specific adeno-associated virus–induced expression of IGF-1 increased circulating IGF-1 levels and prevented most changes in the pattern of GH secretion in FR WT mice. In conclusion, suppressed liver-derived IGF-1 is the primary mechanism behind the changes in the GH secretion pattern observed in FR male mice.

Growth hormone (GH) is a 191-amino acid, single-chain polypeptide secreted predominantly through a pulsatile pattern by the anterior pituitary gland (1). GH secretion is regulated by neuropeptides produced by neurons in the hypothalamus. Hypothalamic GH-releasing hormone (GHRH) stimulates GH secretion, whereas somatostatin (SST) has an inhibitory role (2-4). The stomach-derived hormone ghrelin also regulates GH secretion by activating the GH secretagogue receptor (GHS-R, also known as the ghrelin receptor). Ghrelin is mainly secreted during food restriction, whereas food intake inhibits ghrelin secretion (5-10). The antimicrobial peptide 2 (LEAP2) was identified as an endogenous ghrelin receptor antagonist (11). Thus, LEAP2 can control ghrelin-induced GH secretion (11, 12).

GH secretion is a complex process, and the mechanisms that regulate it are not entirely understood. For example, the GH pulse generator has not yet been identified, although some studies suggest an interplay between SST and GHRH peptides (13-21). GH secretion is strongly influenced by negative feedback loops (2-4). In this regard, GH receptor (GHR) ablation in the hypothalamus or specific neuronal populations increases GH secretion in mice, indicating that brain GHR signaling restrains pituitary GH release (22-26), representing a critical short-loop negative feedback mechanism. GH secretion regulates hepatic insulin-like growth factor 1 (IGF-1) expression and circulating IGF-1 levels (27). In turn, IGF-1 inhibits GH secretion by activating IGF-1 receptors (IGF1Rs) in the pituitary gland (28-30) and the hypothalamus (23, 31), closing a long-loop negative feedback that regulates the GH/IGF-1 axis (2-4).

GH secretion regulates body and tissue growth, cell proliferation, protein synthesis, and carbohydrate and lipid metabolism (32). The metabolic effects of GH, including its action in specific neuronal populations (33), suggest that the GH/IGF-1 axis has a critical role in contributing to survival in situations of metabolic stress (34, 35), particularly during prolonged food restriction (FR) (22, 25, 36-42). In the absence of ghrelin-induced GH secretion, mice subjected to several days of FR develop severe hypoglycemia (36, 38). GHR ablation in neurons that produce the agouti-related protein (AgRP) or γ-aminobutyric acid impairs energy-saving adaptations that prolong survival during FR (22, 25). Furthermore, GHR signaling in the lateral hypothalamic area regulates food-seeking behavior in energy-restricted mice (41). Thus, the central actions of GH, together with its classic metabolic effects—increasing lipolysis in white adipose tissue and hepatic glucose production, as well as insulin resistance, and glycemia—help the organism maintain homeostasis even in the absence of sufficient food (34, 43). Importantly, GH secretion must increase during FR to trigger its metabolic effects (36, 38, 42). However, the underlying mechanisms that drive the alterations in GH secretion during FR are not entirely understood. The rising ghrelin levels have been considered the primary mechanism explaining the increase in GH secretion in FR animals (36-38, 42). Nevertheless, previous studies have used only a single GH measurement (36-38, 42), and analyzing the GH pulsatile pattern is necessary for a more detailed evaluation. In addition, circulating GH levels indeed increase in FR mice that cannot produce ghrelin compared to fed animals (36, 42), indicating that a ghrelin-independent mechanism also stimulates GH secretion during FR.

It is well known that hepatic IGF-1 production is drastically reduced in FR animals (36, 39, 44). For this reason, low IGF-1 levels can be considered an indicator of malnutrition (45-47). This occurs through the development of a liver-specific GH resistance state, which is essential for suppressing the anabolic effects of the GH/IGF-1 axis and, therefore, restraining growth in undernourished animals (48, 49). Considering the vital role of circulating IGF-1 levels in regulating GH secretion, the present study investigated whether low liver-derived IGF-1 is a key underlying mechanism responsible for altering the GH secretion pattern in FR mice. To test this hypothesis, GH secretion patterns were evaluated in ad libitum–fed hepatocyte-specific GHR knockout (KO) (Albumin^ΔGHR) mice and compared to ad libitum–fed and FR control mice. Additionally, we used an adeno-associated virus (AAV) to manipulate hepatic IGF-1 production and thereby increase circulating IGF-1 levels in FR wild-type (WT) mice. Overall, our results suggest that decreased IGF-1 levels are a key factor affecting GH secretion patterns in mice after extended FR.

Materials and Methods

Animals

The present study used WT C57BL/6J male mice (RRID: IMSR_JAX:000664, Jackson Laboratory) and male mice with congenital KO of the hepatocyte GHR. To induce hepatocyte-specific GHR KO mice, Ghr^flox/flox mice (27) were crossed with Alb^Cre mice (RRID: IMSR_JAX:003574, Jackson Laboratory). Mice carrying both mutations were crossed again with Ghr^flox/flox mice until generating Ghr^flox/flox::Alb^Cre mice (named Albumin^ΔGHR) and their respective control animals (Ghr^flox/flox). The LF50 body composition mouse analyzer (Bruker) was used to determine body fat and lean mass in approximately 8-week-old mice. The ethics committee approved the experimental procedures on the use of animals at the Institute of Biomedical Sciences, University of São Paulo.

Food Restriction

Approximately 8-week-old mice were single-housed, and basal food intake was measured to determine the amount of food that should be provided for each animal during the FR period. Then, the pattern of GH secretion was analyzed in ad libitum–fed conditions. Subsequently, the same mice were subjected to 4 days of FR, during which each mouse received 40% of their basal food intake at 17:00 hours (3 hours before lights off) for 4 consecutive days. Body weight was recorded daily. Serial blood collection started the fourth day of food restriction at 13:00 hours, and mice were humanely killed after the evaluation of the pulsatile GH secretion at approximately 20:00 hours (Fig. 1A).

Figure 1.

Food restriction (FR) causes a notable rise in nonpulsatile growth hormone (GH) secretion. A, Illustration of the experimental design, in which 2-month-old wild-type C57BL/6J male mice were subjected to 4 days of FR (40% of ad libitum food intake), followed by serial blood collections to determine the pattern of GH secretion. The arrows indicate when food was provided (3 hours before lights off; lights off at 20:00). Mice were humanely killed just after the serial blood collections (at ∼20:00). B, Changes in body weight in relation to initial weight during the 4-day FR period. C and D, Representative examples of the pattern of GH secretion in 2 mice in C, ad libitum–fed condition and the same animals after D, FR. The shaded areas represent basal (nonpulsatile) GH secretion, whereas the lines show total GH secretion. The arrows point to the pulse peaks. E to L, Total GH secretion (area under the curve), median blood GH, pulsatile GH secretion, pulse frequency, GH pulse amplitude, ratio between pulse amplitude and average GH values, basal GH secretion (area under the curve of nonpulsatile GH secretion), and percentage of contribution of basal secretion to total secretion in the same mice in ad libitum–fed condition and following FR (n = 10). Differences between groups were analyzed by paired 2-tailed t test. *P less than .05; **P less than .01; ***P less than .001; ****P less than .0001.

Evaluation of Growth Hormone Secretion

Before blood collection, mice were acclimated daily to the tail-tip blood sampling procedure for 30 days. Serial blood collection started at 13:00 hours in FR and ad libitum–fed 8-week-old mice by removing a small portion of the tail tip (1 mm) using a surgical blade. For each mouse, 36 sequential tail-tip blood collections were performed at 10-minute intervals. Each blood sample (5 μL) was transferred to a tube containing 105 μL of phosphate-buffered saline with 0.05% Tween-20. Samples were immediately placed on dry ice and then stored at −80 °C. After each blood collection, gentle fingertip pressure was applied to the tail tip to stop the bleeding. Mice were allowed to move freely in their home cages with ad libitum access to water. Food was available during blood collection only for the ad libitum–fed groups.

GH pulses were identified using the DynPeak pulse detection algorithm (50). The area under the curve (AUC) of blood GH levels was used to calculate total GH secretion. Median GH values and the number of GH pulses per hour (frequency) were also calculated for each mouse. To calculate pulsatile GH secretion, we initially identify the start and end of each pulse. If GH pulses started and ended with values above zero, the average of the initial and final values was subtracted from each point of the GH pulse to remove the values considered as basal secretion. If only the initial or final value of the GH pulse was above zero, this value was subtracted from the following or preceding point, respectively. Points that did not belong to GH pulses were set to zero. Pulsatile GH secretion was obtained by calculating the AUC of the values considered only as pulsatile secretion. GH pulse amplitude was the average of all GH peaks for each animal. GH pulse amplitude was divided by mean GH levels to calculate the ratio between pulse amplitude and average GH values. Basal GH secretion was calculated by subtracting the AUC of pulsatile secretion from the AUC of total GH levels. Finally, the percentage of contribution of basal GH secretion to total secretion was determined by dividing the AUC of basal GH secretion by the AUC of total GH secretion × 100. Prism software version 8.4.3 (GraphPad) was used to calculate the AUC. In all representative examples of the GH secretion pattern, the shaded areas indicate basal (nonpulsatile) GH secretion. In contrast, the lines show the actual values obtained in GH measurements (total GH secretion). The arrows point to the pulse peaks.

Hormone Measurement

An “in-house” enzyme-linked immunosorbent assay was used to measure blood GH levels (1:22 dilution), as described previously (22-24, 31, 44, 51, 52). Commercially available enzyme-linked immunosorbent assay kits were used to measure the concentrations of serum IGF-1 (catalog No. MG100; RRID: AB_2827989; R&D Systems), plasma acylated ghrelin (catalog No. A05317; RRID: AB_2619624; Bertin Pharma), and plasma LEAP2 (catalog No. EK-075-40; RRID: AB_2909436; Phoenix Pharmaceuticals). Blood samples used to measure ghrelin were collected in tubes containing EDTA (MilliporeSigma) and 4-(2-aminoethyl) benzene sulfonyl fluoride hydrochloride (AEBSF; MilliporeSigma), at a final concentration of 2 mg of AEBSF per milliliter of whole blood. After centrifugation, plasma was acidified to a final concentration of 0.1 M HCl to minimize ghrelin degradation. Acidified plasma samples were stored at −80 °C, and ghrelin was measured within 1 week of sample collection.

Gene Expression

Quantitative real-time polymerase chain reaction (PCR) was used to determine the expression of genes in the liver and hypothalamus. RNA was extracted using TRIzol (Invitrogen) and then incubated with RNase-free DNase I (MilliporeSigma). Complementary DNA was synthesized by reverse transcription using 2 µg of total RNA, SuperScript II Reverse Transcriptase (Invitrogen), and random primers p(dN)6 (MilliporeSigma). Real-time PCR was performed using the 7500 Real-Time PCR System (Applied Biosystems), SYBR Green Gene Expression PCR Master Mix (Applied Biosystems), and specific primers for target genes (Table 1). Data were normalized to the geometric average of Actb and Gapdh. Relative quantification of messenger RNA (mRNA) was calculated by 2^−ΔΔCt.

Table 1.

Primer sequences

Target gene	Forward primer	Reverse primer
Abcd2	tgggatttccctactctcca	gcgactccaacttttgcttt
Acot3	gcacgcatttaccacccaga	gctggtcacaaaggcaacca
Actb	gctccggcatgtgcaaag	catcacaccctggtgccta
Agrp	ctttggcggaggtgctagat	aggactcgtgcagccttacac
Cis	tcgggaatctgggtggtact	taggaatgtaccctccggca
Cyp17a1	ggactcacctcctcatgctg	cagtgcccagagattgatga
Cyp2b9	ggaaaccaggccattggta	gtggactccagaaatcaaaacag
Cyp2d9	cattctcagcaggccgca	accatagactccagagttgct
Cyp7b1	tgcgtgacgaaattgacagt	tgctggagtatgagcacagc
Gapdh	gggtcccagcttaggttcat	tacggccaaatccgttcaca
Ghr	atcaatccaagcctggggac	acagctgaatagatcctgggg
Ghrh	tatgcccggaaagtgatccag	atccttgggaatccctgcaaga
Ghrl	agaggaggagctggagatca	gctggcgcctctttgacc
Igf1	ccacactgacatgcccaaga	gtacttcctttccttctcctttgc
Igfbp1	attagctgcagcccaacaga	gacacaccagcagagtccag
Igfbp3	gcagcctaagcacctacctc	acttggaatcggtcactcgg
Leap2	gaccccattttggagagggg	gggaacagcgtctttttctgc
Leprot	cgggcgttaaagctctcgtg	agacgaacaggggccagtaa
Mup1	acctatccaatgccatggacc	ctggaggcagcgatctgtag
Npy	ccgcccgccatgatgctaggta	ccctcagccagaatgcccaa
Pomc	atagacgtgtggagctggtgc	gcaagccagcaggttgct
Prlr	cagtaaatgccaccaacgaa	gaggaggctctggttcaaca
Serpina6	tctgatttcgcagacaccac	aggaagatgaagggcctgtt
Sirt1	taccttggagcaggttgcag	gcaccgaggaactacctgat
Slco1a1	atctacggtgacgcacacac	tcagtagaattgaatgctgtttcag
Socs2	cgcgagctcagtcaaacag	aagttccttctggagcctctt
Sst	ctgtcctgccgtctccagt	ctgcagaaactgacggagtct

Adeno-Associated Virus Administration

WT C57BL/6J male mice (n = 17) were acclimated daily to the tail-tip blood sampling procedure for at least 30 days. When these animals reached approximately age 8 weeks, a blood sample was collected to measure IGF-1 concentration under ad libitum–feeding conditions. Some of these mice underwent serial blood collection to assess pulsatile GH secretion (fed group; n = 6). Subsequently, the control group (n = 7) received a single injection into the retro-orbital plexus of 100 µL of pAAV8.TBG.PI.Null.bGH (hereafter named AAV8-TBGp-Null; 2 × 10^11 genome copies per mouse, Addgene), which is an inert vector. The remaining mice (n = 10) received a single injection of 100 µL of AAV8.TBG.PI.ratIGF1.WPRE.bGH (hereafter named AAV8-TBGp-IGF1; 2 × 10^11 genome copies per mouse) into the retroorbital plexus. This construct induces a hepatocyte-specific expression of rat IGF-1 (53, 54) and was used here to increase circulating IGF-1 levels in FR mice. After a 4-day recovery period with ad libitum food intake, all mice were subjected to FR as previously described. From 1 to 7 Pm on the fourth day of FR, serial blood samples were collected. Mice were humanely killed between 7 and 8 Pm, and serum was collected to measure IGF-1 levels in FR mice. The IGF-1 levels in FR mice that received AAV8-TBGp-IGF1 determined group classification: the low IGF-1 group (AAV injection was ineffective at increasing IGF-1 levels in FR mice; n = 4) and high IGF-1 group (partial recovery of IGF-1 levels; n = 6).

Statistical Analysis

A 2-tailed t test was used to analyze the differences between the 2 groups. When 3 groups were compared simultaneously, one-way analysis of variance (ANOVA) followed by Tukey or Newman-Keuls multiple comparisons test was used. Normality and homogeneity of variance were assessed using Shapiro-Wilk and Bartlett tests, respectively. Data that did not have a normal distribution or showed differences in the variance were analyzed using Brown-Forsythe and Welch ANOVA tests, followed by Dunnett T3 multiple comparison test. Correlation analyses were determined using the Pearson correlation coefficient. Statistical analyses were performed using Prism software. All results are expressed as mean ± SEM, and only P values less than .05 were considered statistically significant.

Results

Food Restriction Causes a Notable Rise in Nonpulsatile Growth Hormone Secretion

To compare the pattern of GH secretion between fed and FR mice, serial blood collections were performed in 2-month-old ad libitum–fed WT. In the following week, these mice were subjected to 4 days of FR (40% of their average daily intake) to deplete body fat stores, followed by serial blood collections starting approximately 20 hours after the last meal (Fig. 1A). Thus, blood GH levels were evaluated in fasted mice with low energy reserves, a situation in which the GH/IGF-1 axis is necessary to maintain glycemia and produce energy-saving adaptations (22, 25, 34, 36-40). A 4-day FR resulted in an approximately 30% weight loss compared to initial body mass (Fig. 1B). Fig. 1C and 1D show the pattern of GH secretion in 2 representative mice in ad libitum–fed conditions (see Fig. 1C) and the same animals after FR (see Fig. 1D). Basal (nonpulsatile) GH secretion is represented as the shaded areas in Fig. 1C and 1D, while the lines indicate total GH secretion. In ad libitum–fed mice, GH was mainly secreted through large pulses (see Fig. 1C). FR increased total GH secretion compared to fed mice (Fig. 1E). Median blood GH was zero (or near zero) in fed mice, reflecting the very low GH secretion during interpulse intervals, which comprised the majority of the analyzed points (see Fig. 1C and 1F). Median blood GH robustly increased in FR mice, indicating that GH secretion was maintained tonically elevated in this condition (see Fig. 1F). When pulsatile GH secretion was calculated, we observed an increased pulsatile secretion in FR mice compared to fed animals (Fig. 1G). However, the higher pulsatile GH secretion in FR mice is explained by increases in pulse frequency (Fig. 1H) rather than changes in GH pulse amplitude (Fig. 1I) or mean GH secretion per pulse (data not shown). Consequently, the ratio between pulse amplitude and average GH values was substantially reduced in FR mice (Fig. 1J). Higher values of this ratio indicate more significant differences between peak and nadir values, a key feature of pulsatile GH secretion. The most remarkable change in the pattern of GH secretion in the FR mice was an increase in basal (nonpulsatile) GH secretion (Fig. 1K). Thus, the contribution of basal GH secretion to total secretion robustly increased in FR mice compared to fed animals (Fig. 1L). Altogether, FR male mice showed significant increases in total and mainly nonpulsatile GH secretion.

Growth Hormone Receptor Ablation in Hepatocytes Recapitulates Low Circulating Insulin-like Growth Factor-1 Levels Observed in Food-Restricted Mice

FR reduces hepatic IGF-1 secretion, leading to low circulating IGF-1 levels (36, 39, 44). The present study confirmed that prolonged FR drastically reduced hepatic Igf1 mRNA expression and serum IGF-1 levels by approximately 75% and 90%, respectively, compared to control fed mice (Fig. 2A and 2B). To investigate whether low circulating IGF-1 is behind the changes in the pattern of GH secretion observed during prolonged FR, mice with a hepatocyte-specific GHR ablation were generated (ad libitum–fed Albumin^ΔGHR mice). The control group consisted of ad libitum–fed Ghr^flox/flox littermates. Additionally, the same parameters were analyzed in a group of FR WT mice. Albumin^ΔGHR mice had only 1.2% of the hepatic Ghr mRNA expression compared to control animals (Fig. 2C). Notably, FR caused a 6-fold reduction in hepatic Ghr mRNA expression in comparison to control fed mice (see Fig. 2C). As previously demonstrated (27, 55), ad libitum–fed Albumin^ΔGHR mice showed very low hepatic Igf1 mRNA expression (Fig. 2A). Additionally, serum IGF-1 levels in ad libitum–fed Albumin^ΔGHR mice were similar to those observed in FR WT animals (Fig. 2B). Ghrelin is a GH secretagogue secreted during food restriction (5, 9, 10). Thus, plasma ghrelin levels were also measured to investigate whether GHR ablation in hepatocytes affects the ghrelin/GH axis. Ad libitum–fed Albumin^ΔGHR mice showed similar plasma ghrelin concentrations compared to ad libitum–fed control mice (Fig. 2D). As expected (36-39), ghrelin levels were significantly increased in FR mice compared to the fed groups (Fig. 2D). The plasma concentration of the ghrelin receptor antagonist LEAP2 was also analyzed. Albumin^ΔGHR mice showed higher LEAP2 levels than ad libitum–fed control animals (Fig. 2E). In contrast, FR resulted in a robust suppression of plasma LEAP2 concentrations (see Fig. 2E). Hepatic Leap2 mRNA expression tended to recapitulate the changes observed in plasma LEAP2 levels, although the difference between ad libitum–fed control and Albumin^ΔGHR mice was not statistically significant (Fig. 2F). Thus, these findings suggest that Albumin^ΔGHR mice can be used as a model to investigate whether low serum IGF-1 levels influence the pattern of GH secretion independently of FR or variations in ghrelin levels.

Figure 2.

Growth hormone receptor (GHR) ablation in hepatocytes recapitulates the low circulating insulin-like growth factor-1 (IGF-1) levels observed in food-restricted (FR) mice. A, Hepatic Igf1 messenger RNA (mRNA) expression in 2-month-old ad libitum–fed control (Ghr^flox/flox littermates; n = 6), fed Albumin^ΔGHR (n = 9), and FR wild-type (WT) C57BL/6J male mice (n = 6). B, Serum IGF-1 levels in 2-month-old ad libitum–fed control (Ghr^flox/flox littermates; n = 6), fed Albumin^ΔGHR (n = 9), and FR WT male mice (n = 9). (C) Hepatic Ghr mRNA expression. D and E, Plasma ghrelin and LEAP2 levels in ad libitum–fed control (Ghr^flox/flox littermates; n = 6-11), fed Albumin^ΔGHR (n = 7-8), and FR WT male mice (n = 5-7). F, Hepatic Leap2 mRNA expression. Differences were analyzed by one-way analysis of variance (ANOVA) and Tukey multiple comparisons test or Brown-Forsythe and Welch ANOVA tests, followed by Dunnett T3 multiple comparison test if the data did not show a normal distribution. *P less than .05; **P less than .01; ***P less than .001; ****P less than .0001.

Albumin^ΔGHR Male Mice Mimic the Pattern of Growth Hormone Secretion Observed in Food-Restricted Animals

Before investigating the pattern of GH secretion, the body composition of Albumin^ΔGHR mice was characterized. Two-month-old Albumin^ΔGHR mice exhibited reduced body growth, as evidenced by a significant decrease in body and lean mass compared to control animals (Fig. 3A and 3B). Fat mass was unaffected in Albumin^ΔGHR male mice (Fig. 3C). Next, the pattern of GH secretion was investigated in ad libitum–fed 2-month-old control (Ghr^flox/flox littermates) and Albumin^ΔGHR male mice (Fig. 4A and 4B). Remarkably, Albumin^ΔGHR mice showed similar alterations in the pattern of GH secretion as observed in FR WT mice. In this regard, Albumin^ΔGHR mice exhibited increased total GH secretion and median blood GH compared to control mice (Fig. 4C and 4D). However, unlike FR mice (Fig. 1G and 1H), fed Albumin^ΔGHR mice did not show alterations in pulsatile GH secretion or pulse frequency (Fig. 4E and 4F). GH pulse amplitude (Fig. 4G) and mean GH secretion per pulse (data not shown) were not different compared to control fed mice. Like FR mice (Fig. 1J), the ratio between pulse amplitude and average GH values was substantially suppressed in fed Albumin^ΔGHR mice (Fig. 4H). Notably, basal GH secretion and the percentage of contribution of basal secretion to total secretion were robustly increased in Albumin^ΔGHR mice compared to controls (Fig. 4I and 4J). Thus, the pattern of GH secretion in Albumin^ΔGHR-fed mice is relatively similar to that observed in FR mice, with a prevailing basal/tonic secretion over the pulsatile secretion.

Figure 3.

Albumin^ΔGHR mice show reduced body growth. A to C, Body weight, lean body mass, and fat mass in 2-month-old control (n = 6) and Albumin^ΔGHR (n = 9) male mice. Data were analyzed using unpaired 2-tailed t test. **P less than .01.

Figure 4.

Albumin^ΔGHR male mice mimic the pattern of growth hormone (GH) secretion observed in food-restricted (FR) mice. A and B, Representative examples of the pattern of GH secretion in A, 2 control and B, Albumin^ΔGHR-fed male mice. The shaded areas represent basal (nonpulsatile) GH secretion, and the arrows indicate the pulse peaks. C to J, Total GH secretion (area under the curve), median blood GH, pulsatile GH secretion, pulse frequency, GH pulse amplitude, ratio between pulse amplitude and average GH values, basal GH secretion (area under the curve of nonpulsatile GH secretion), and percentage of contribution of basal secretion to total secretion in control (n = 6) and Albumin^ΔGHR-fed (n = 9) male mice. Differences between groups were analyzed by unpaired 2-tailed t test. *P less than .05; **P less than .01; ****P less than .0001.

Hypothalamic Gene Expression in Ad Libitum–Fed Albumin^ΔGHR and Food-Restricted Wild-Type Male Mice

The hypothalamic gene expression was analyzed in male mice to investigate how the expression of neuropeptides that control the somatotropic axis could explain the alterations in GH secretion. Despite the marked changes in GH secretion observed in Albumin^ΔGHR or FR mice, hypothalamic Ghrh mRNA expression did not differ between the experimental groups (Fig. 5A). Hypothalamic Sst mRNA expression increased in Albumin^ΔGHR mice compared to control animals, whereas FR did not affect Sst mRNA levels (Fig. 5B). The expression of some neuropeptides involved in regulating energy balance was also assessed. As previously described (56, 57), FR robustly increased the hypothalamic expression of Agrp and Npy (Fig. 5C and 5D). Compared to ad libitum–fed control animals, a 1.7-fold increase in Agrp mRNA expression was observed in the hypothalamus of Albumin^ΔGHR mice. However, this difference did not reach statistical significance (P = .1475; see Fig. 5C). On the other hand, Albumin^ΔGHR mice showed a significant increase in Npy mRNA expression compared to ad libitum–fed control animals (see Fig. 5D). No significant differences were observed in the hypothalamic expression of Pomc mRNA (Fig. 5E). Hypothalamic ghrelin expression was also analyzed, and FR mice showed increased Ghrl mRNA expression compared to other groups (Fig. 5F). Thus, the hypothalamic expression of neuropeptides cannot explain the alterations in the GH secretion pattern observed in Albumin^ΔGHR or FR mice.

Figure 5.

Hypothalamic gene expression in ad libitum-fed Albumin^ΔGHR and food-restricted (FR) wild-type (WT) male mice. A to F, Hypothalamic messenger RNA (mRNA) expression in 2-month-old ad libitum–fed control (Ghr^flox/flox littermates; n = 6), fed Albumin^ΔGHR (n = 7-8), and FR WT C57BL/6J male mice (n = 4-6). Differences were analyzed by one-way analysis of variance (ANOVA) and Tukey multiple comparisons test or Brown-Forsythe and Welch ANOVA tests, followed by Dunnett T3 multiple comparison test if the data did not show a normal distribution. *P less than .05; **P less than .01.

Differences in Hepatic Expression of Genes That Regulate Growth Hormone Sensitivity Between Ad Libitum–Fed Albumin^ΔGHR and Food-Restricted Wild-Type Male Mice

Albumin^ΔGHR and FR WT mice both share a phenotype of hepatic GH resistance. Reduced hepatic Ghr expression is the primary cause of GH resistance in the liver of Albumin^ΔGHR mice and, to a lesser extent, in FR WT mice (see Fig. 2C). However, additional mechanisms regulate hepatic GH sensitivity, including the leptin receptor overlapping transcript (LEPROT), the protein deacetylase sirtuin 1 (SIRT1), and members of the suppressor of cytokine signaling (SOCS) family, like SOCS2 and CIS (48, 49, 58-60). Thus, we compared the mRNA expression of these genes between ad libitum–fed control, ad libitum–fed Albumin^ΔGHR, and FR WT male mice. FR mice exhibited increased hepatic expression of Leprot and Socs2, whereas Cis mRNA levels were reduced compared to ad libitum–fed control mice (Fig. 6A-6C). Albumin^ΔGHR mice showed no changes in Leprot mRNA levels and a decreased expression of Socs2 and Cis compared to ad libitum–fed control animals (see Fig. 6A-6C). Sirt1 expression was reduced in Albumin^ΔGHR mice compared to fed and FR control animals (Fig. 6D). Insulin-like growth factor-binding proteins (IGFBPs) transport IGF-1 and regulate IGF-1 half-life and its capacity to activate IGF1R. Considering the key role played by IGF-1 in regulating GH secretion via long-loop negative feedback (2-4, 28-31), the expression of some IGFBPs was analyzed. FR caused a 50-fold increase in hepatic Igfbp1 mRNA levels (Fig. 6E). Igfbp3 expression also increased in FR animals compared to Albumin^ΔGHR mice (Fig. 6F). GHR ablation in the liver did not affect Igfbp1 and Igfbp3 mRNA expression compared to ad libitum–fed control mice (see Fig. 6F). Thus, these findings reveal very distinct alterations in the expression of genes that control hepatic GH sensitivity between ad libitum–fed Albumin^ΔGHR and FR WT male mice.

Figure 6.

Hepatic expression of genes that regulate growth hormone (GH) sensitivity in ad libitum-fed Albumin^ΔGHR and food-restricted (FR) wild-type (WT) male mice. A to F, Hepatic messenger RNA (mRNA) expression in 2-month-old ad libitum–fed control (Ghr^flox/flox littermates; n = 6), fed Albumin^ΔGHR (n = 8-9), and FR WT C57BL/6J male mice (n = 5-6). Differences were analyzed by one-way analysis of variance (ANOVA) and Tukey multiple comparisons test or Brown-Forsythe and Welch ANOVA tests, followed by Dunnett T3 multiple comparison test if the data did not show a normal distribution. *P less than .05; **P less than .01; ***P less than .001; ****P less than .0001.

Ad Libitum–Fed Albumin^ΔGHR and Food-Restricted Wild-Type Male Mice Show Differences in Hepatic Expression of Sexually Dimorphic Genes

GH secretion pattern and GH sensitivity regulate the hepatic expression of numerous sexually dimorphic genes (23, 31, 61-64). Thus, we investigated in male mice how FR or hepatocyte-specific GHR ablation affects the expression of sexually regulated genes. Prolonged FR suppressed all analyzed genes that show a predominant expression in females, including Cyp2b9, Prlr, Abcd2, Cyp17a1, Acot3, and Serpina6 (Fig. 7A-7F). In contrast, only part of the male-predominant genes were affected by FR (Fig. 7G-7J). In this sense, FR mice exhibited decreased Slco1a1 and Mup1 expression in the liver (see Fig. 7G and 7H), compared to ad libitum–fed control animals, while no changes were observed in the mRNA levels of Cyp2d9 and Cyp7b1 (Fig. 7I and 7J). Albumin^ΔGHR mice showed increased Cyp2b9 mRNA levels compared to other groups (see Fig. 7A), whereas Prlr expression was as reduced as in FR mice (see Fig. 7B). Hepatocyte-specific GHR inactivation also suppressed the expression of Slco1a1 and Mup1 compared to ad libitum–fed control animals (see Fig. 7G and 7H). The expression of the other analyzed genes was not altered in Albumin^ΔGHR male mice. Thus, except for Prlr, the expression of all female-predominant genes analyzed here was differentially regulated in Albumin^ΔGHR and FR male mice. In contrast, most male-dominant genes we assessed showed similar expressions when comparing ad libitum–fed Albumin^ΔGHR and FR WT mice.

Figure 7.

Hepatic expression of sexually regulated genes in ad libitum–fed Albumin^ΔGHR and food-restricted (FR) wild-type (WT) male mice. A to J, Hepatic messenger RNA (mRNA) expression in 2-month-old ad libitum–fed control (Ghr^flox/flox littermates; n = 6), fed Albumin^ΔGHR (n = 8-9), and FR WT C57BL/6J male mice (n = 6). Differences were analyzed by one-way analysis of variance (ANOVA) and Tukey multiple comparisons test or Brown-Forsythe and Welch ANOVA tests, followed by Dunnett T3 multiple comparison test if the data did not show a normal distribution. *P less than .05; **P less than .01; ***P less than .001; ****P less than .0001.

Adeno-Associated Virus–Induced Increase in Insulin-like Growth Factor-1 Levels Prevents Most Changes in Growth Hormone Secretion Pattern Caused by Food Restriction

Although the pattern of GH secretion in fed Albumin^ΔGHR mice resembles that in FR WT animals, their metabolic states differ significantly, which introduces a major confounding factor when comparing these groups. Notable variations in gene expression analysis of the hypothalamus and liver highlighted these differences. To address this issue, we used systemically delivered AAV to specifically increase hepatocyte Igf1 expression in FR WT mice. Initially, we validated this vector in adult IGF-1–deficient, ad libitum–fed mice. A single systemic injection of AAV8-TBGp-IGF1 increased circulating IGF-1 levels by about 11 times in dwarf Ghrhr^{little/little} mice (baseline: 24.1 ± 4.3 ng/mL; 8 days after injection: 265.5 ± 29.6 ng/mL; n = 3) and by 3.4 times in Albumin^ΔGHR mice (baseline: 75.2 ± 30.2 ng/mL; 8 days after injection: 254.4 ± 59.9 ng/mL; n = 2). This treatment was also sufficient to increase the body weight of Ghrhr^{little/little} mice by 13.6% after 13 days (baseline: 12.0 ± 0.7 g; 13 days after injection: 13.7 ± 0.6 g; n = 3). Following successful tests with this vector, adult WT C57BL/6J male mice received a single systemic injection of AAV8-TBGp-IGF1 or AAV8-TBGp-Null (control group). After 4 days, these animals underwent FR for 4 days, and the pattern of GH secretion was measured. A fed group was included to represent the typical GH secretion observed in mice fed ad libitum. Fig. 8A to 8D shows representative examples of GH secretion patterns for each group. We observed a varied response to AAV8-TBGp-IGF1, with 4 mice showing low levels of IGF-1, similar to FR AAV8-TBGp-Null controls (Fig. 8E). Since an increase in IGF-1 levels is the necessary condition to test the hypothesis of this study, these animals were classified into the low IGF-1 group. However, 6 mice injected with AAV-TBGp-IGF1 showed partial recovery of their IGF-1 levels and were therefore referred to as the high IGF-1 group (see Fig. 8E).

Figure 8.

Adeno-associated virus (AAV)-induced increase in insulin-like growth factor-1 (IGF-1) levels prevents most of the changes in growth hormone (GH) secretion pattern caused by food restriction (FT) in adult wild-type (WT) C57BL/6J male mice. A to D, Representative examples of GH secretion patterns in a mouse A, from the fed group; B, control group; C, low IGF-1 group; and D, high IGF-1 group. The shaded areas represent basal (nonpulsatile) GH secretion, whereas the lines show total GH secretion. The arrows point to the pulse peaks. E, Serum IGF-1 concentrations in fed (n = 17), control (n = 7), low IGF-1 (n = 4), and high IGF-1 (n = 6) groups. F to L, Total GH secretion (area under the curve), median blood GH, pulsatile GH secretion, pulse frequency, the ratio between pulse amplitude and average GH levels, basal GH secretion (area under the curve of nonpulsatile GH secretion), and the percentage contribution of basal secretion to total secretion across fed (n = 6), control (n = 7), low IGF-1 (n = 4), and high IGF-1 (n = 6) groups. M, Significant negative correlation between serum IGF-1 concentration and nonpulsatile GH secretion across all experimental groups (n = 23). Differences between groups were analyzed by one-way analysis of variance followed by the Newman-Keuls multiple comparisons test. Different letters on the bars indicate statistically significant differences between groups (P < .05).

Similar to the findings seen in FR WT mice (see Fig. 1) and fed Albumin^ΔGHR mice (see Fig. 4), FR control and low IGF-1 mice showed significant increases in total GH secretion (Fig. 8F), median GH levels (Fig. 8G), basal (nonpulsatile) GH secretion (Fig. 8K), and in the contribution of basal GH secretion to total secretion (Fig. 8L) compared to fed mice. Remarkably, a partial restoration of circulating IGF-1 levels (in the high IGF-1 group) was sufficient to prevent all these changes (see Fig. 8). Pulsatile GH secretion (Fig. 8H) and GH pulse amplitude (Fig. 8J) were similarly suppressed in the control, low IGF-1, and high IGF-1 groups, compared to fed animals. No statistically significant differences in GH pulse frequency were observed between groups (Fig. 8I). Considering the data from all groups in this experiment, serum IGF-1 levels showed a significant negative correlation with basal (nonpulsatile) GH secretion (Fig. 8M). Overall, these findings provide clear and strong evidence that reduced IGF-1 levels are a key factor influencing GH secretion patterns in FR mice, particularly promoting an increase in basal nonpulsatile GH secretion.

Discussion

The role of IGF-1 in regulating GH secretion has been described since the early 1980s (65-67). IGF-1 inhibits GH secretion via the hypothalamus (23, 31) and pituitary gland (28, 29). Accordingly, IGF1R ablation, alone or combined with other receptors, in GHRH or SST neurons and somatotropes leads to GH oversecretion (23, 28, 29, 31). The present study demonstrates a particularly significant role of IGF-1 in regulating GH secretion during chronic FR, when circulating IGF-1 levels are severely reduced. Our findings further indicate that the negative feedback loop mediated by IGF-1 prevails over GH-mediated negative feedback. In this sense, GH oversecretion cannot restrain its own secretion under conditions of low circulating IGF-1 levels. Furthermore, low liver-derived IGF-1 not only increases total GH secretion but also changes the pattern of secretion. Importantly, IGF1R ablation in GHRH neurons or combined deletions of IGF1R and GHR in SST neurons increase total GH secretion by elevating GH pulse amplitude without affecting basal secretion (23, 31). Thus, the absence of IGF-1 negative feedback increases basal secretion, likely via extrahypothalamic cells. Unfortunately, there is no information about the GH pulsatile pattern in somatotrope-specific IGF1R KO mice (28, 29). Therefore, future studies should investigate whether IGF1R signaling in the pituitary gland modulates tonic/basal GH secretion. Nonetheless, somatotrope-specific IGF1R KO mice show increased GH levels when using a single (random) blood collection (28, 29). In ad libitum–fed animals, a random blood measurement has a higher chance of detecting interpulse intervals, as the median GH values observed in our study are near zero in this condition. In contrast, single, random blood collections will probably detect higher GH levels in animals whose secretion is tonically elevated (higher median GH levels), as is the case with prolonged FR mice (36, 38), and likely in somatotrope-specific IGF1R KO mice (28, 29).

Ghrelin can stimulate GH secretion by activating GHS-R either in GHRH neurons (68, 69) or somatotropes (70, 71). Previous studies demonstrated that increases in ghrelin partially explain the rise in GH levels in prolonged FR mice (36, 38, 42). However, GH levels still increase during FR in mice unable to produce ghrelin (36, 42). Thus, the present findings do not rule out the critical role of ghrelin in inducing GH secretion during FR. Instead, we uncovered an additional mechanism explaining the increase in basal GH secretion. The administration of a GHS-R antagonist decreases GH secretion by reducing peak amplitude without affecting GH nadir values, an indicator of basal secretion (72). Another study found that although 24-hour ghrelin infusion increases basal GH secretion in adult humans, the increase in pulsatile secretion is much more potent (73). Ghrelin KO mice exhibit reduced pulsatile GH secretion without alterations in basal secretion (74). GHS-R KO mice also show reduced GH pulse amplitude (75). These results indicate that ghrelin primarily regulates pulsatile GH release rather than basal secretion (76). This explains why FR WT mice also exhibited increased pulsatile GH secretion, an effect not observed in Albumin^ΔGHR mice. In the case of FR animals, the consequences of low IGF-1 levels, which increase predominantly basal GH secretion, were combined with the secretagogue effects of ghrelin, which primarily stimulates pulsatile GH secretion.

LEAP2 was described as an endogenous GHS-R antagonist (11). Accordingly, LEAP2 infusion prevents ghrelin-induced increases in GH secretion and food intake (11, 12). Conversely, Leap2 KO mice show enhanced responsiveness to ghrelin (77). Our findings are in accordance with previous studies, which have shown reduced plasma LEAP2 and hepatic Leap2 mRNA expression in FR mice (12, 78 ). Thus, both increased ghrelin secretion and decreased LEAP2 levels act in concert to increase GHS-R signaling during FR. Ad libitum–fed Albumin^ΔGHR mice exhibited a significant increase in plasma LEAP2 concentrations compared to control fed mice. The causes of this increase are unknown, but this result suggests reduced ghrelin responsiveness in Albumin^ΔGHR mice (11, 12). However, future studies are necessary to evaluate this possibility.

Despite the robust alterations in GH secretion, ad libitum–fed Albumin^ΔGHR and FR control mice showed no modifications in hypothalamic Ghrh expression. This fact contrasts with evidence indicating that alterations in GH or IGF-1 secretion frequently affect hypothalamic Ghrh mRNA levels (22, 24, 28, 31, 79, 80). Increases in GH secretion can also upregulate hypothalamic SST expression (80-85), closing a well-described negative feedback loop (2-4). In this case, increased GH secretion may have led to higher hypothalamic Sst expression in Albumin^ΔGHR mice. However, FR mice exhibited hypothalamic Sst expression similar to that of fed animals. It is unknown whether, during FR, specific neuronal populations may develop GH resistance, similar to that observed in the liver (48, 49). The present study confirmed previous findings showing upregulation of Agrp and Npy expression in the hypothalamus during FR (56, 57). Approximately 95% of neuropeptide Y (NPY)-expressing neurons that corelease AgRP in the arcuate nucleus of the hypothalamus (ARH) express the GHR (86, 87). GH induces the activation of ARH^NPY/AgRP neurons (25, 86, 88). So, the increased GH secretion in Albumin^ΔGHR and FR mice may have contributed to the upregulation of these hypothalamic neuropeptides. Accordingly, GHR or STAT5 ablation in ARH^NPY/AgRP neurons partially blunts the upregulation of Npy and Agrp in the hypothalamus of FR mice (25, 40). Previous studies have shown that NPY inhibits GH secretion during fasting (2, 89). However, GHR ablation in ARH^NPY/AgRP neurons does not affect GH secretion in ad libitum–fed conditions or during acute fasting or chronic FR (44). Thus, it seems unlikely that hypothalamic Npy or Agrp upregulation is behind the increased GH secretion observed in Albumin^ΔGHR or FR mice.

It is currently unclear whether the hypothalamus can produce ghrelin in physiologically relevant amounts (90). Using real-time PCR, a highly sensitive technique, we observed a 2-fold increase in ghrelin mRNA levels in the hypothalamus of FR mice compared to the other groups. Considering that ghrelin stimulates the activity of GHRH neurons (68, 69), this could help explain the more pronounced increase in GH secretion in FR mice. However, as reported previously (90), we observed very low levels of ghrelin mRNA expression in the hypothalamus (with cycle threshold values averaging >31), which makes it challenging to determine whether this result has a physiologically relevant meaning. Altogether, the hypothalamic expression of neuropeptides does not comprehensively explain the observed alterations in the GH secretion pattern of Albumin^ΔGHR or FR mice.

Albumin^ΔGHR mice exhibit hepatic GH resistance due to the genetic ablation of GHR in hepatocytes. Part of the hepatic GH resistance of FR WT mice is also caused by reduced liver GHR expression. However, FR also induces hepatic GH resistance by increasing the expression of LEPROT, SOCS2, and SIRT1 proteins (48, 49, 58-60). Interestingly, ad libitum–fed Albumin^ΔGHR and FR mice showed striking differences in the hepatic expression of transcripts controlling GH sensitivity. The same occurred in relation to IGFBPs, which were altered only in FR mice. A possible mechanism that explains the differentiated expression of these genes between these groups is insulin (91). Insulin secretion is suppressed in FR animals (36) but increases in Albumin^ΔGHR mice (27). Insulin action in the liver or differentiated hepatoma cells downregulates LEPROT, SIRT1, and IGFBP1 (48, 92-94) and stimulates GHR expression (91, 95, 96).

Hepatic gene expression exhibits a well-documented sexual dimorphism, with a significant contribution attributed to sex differences in GH secretion (23, 31, 61-64). Our findings indicate that prolonged FR has a general inhibitory effect on the transcription of genes whose expression shows sex differences. Except for the Prlr gene, all female-predominant genes analyzed showed reduced expression in FR animals compared to Albumin^ΔGHR mice. This finding is intriguing because Albumin^ΔGHR and FR mice have a similar phenotype of hepatic GH resistance and GH secretion. Thus, the differences observed in hepatic gene expression are unlikely to be mediated by GH action in the liver, which appears to be equally blunted in these 2 experimental groups. Instead, profound metabolic differences between fed Albumin^ΔGHR mice and FR animals likely played a significant role in the observed differences in hepatic gene expression. While ad libitum–fed Albumin^ΔGHR mice exhibit insulin resistance, hyperinsulinemia, hyperglycemia, and increased lipogenesis (27, 97), the livers of FR WT mice exhibit a very distinct metabolic status. FR also causes marked alterations in the levels of several hormones that control hepatic gene expression, including insulin, glucagon, corticosterone, and fibroblast growth factor 21, among others. Additionally, increased GH secretion, as seen in SST KO mice (63, 98, 99) or in mice with constitutively active STAT5b in the liver (64, 100), feminizes hepatic gene expression. However, neither Albumin^ΔGHR nor FR mice showed overall feminization in the analyzed genes, further indicating that factors other than GH action on the liver may have influenced the results. Finally, it is essential to note that we examined only a relatively small number of genes that display sexual dimorphism in the liver. These genes were chosen based on previous research demonstrating the role of hepatic GHR signaling in their regulation (63, 98, 99). Therefore, our conclusions should be limited to this specific group of genes. Additionally, our work was performed on males. As a result, we cannot extend our findings to females, either in terms of hepatic expression of sexually dimorphic genes or the role of IGF-1 in regulating GH secretion during FR.

As mentioned earlier, comparing groups with different metabolic states represents a confounding factor not only in analyzing gene expression in the hypothalamus and liver but also in the pattern of GH secretion. Therefore, in our final experiment, a group of WT, FR mice received an injection of an AAV to increase hepatocyte-specific expression of IGF-1 and restore circulating IGF-1 levels closer to those of fed animals. Although our pilot experiment in IGF-1–deficient mice and previous studies using this vector (53, 54) showed a nearly complete recovery in circulating IGF-1 levels in ad libitum–fed mice, serum IGF-1 levels in the high IGF-1 group were 60% of those found in fed mice. However, the construct used was able to elevate serum IGF-1 levels more than 2-fold compared to the control and low IGF-1 groups. Multiple mechanisms likely contribute to the suppression of Igf1 gene expression in the liver during starvation, affecting both endogenous and transgene vector expression. It is also possible that FR affects the bioavailability and half-life of circulating IGF-1, preventing the transgene vector from fully restoring levels of this hormone in FR animals. Despite the partial recovery in IGF-1 levels, the FR high IGF-1 group showed normalization in total GH secretion, median GH levels, nonpulsatile GH secretion, and the contribution of basal GH secretion to total secretion. Including the low IGF-1 group was essential to demonstrate that even when receiving the same viral vector, a partial increase in IGF-1 levels is necessary to decrease total and basal GH secretion. Notably, the results from Albumin^ΔGHR mice and the experiment with the viral vector were consistent in showing that changes in IGF-1 levels seem to have little effect on pulsatile GH secretion and the amplitude and frequency of pulses. Therefore, the increases in pulsatile GH secretion and pulse frequency in FR mice depend on other mechanisms, possibly involving the ghrelin/LEAP2/GHS-R system, as discussed earlier.

In conclusion, our findings confirmed our initial hypothesis that suppressed hepatic IGF-1 production is a critical mechanism that alters the GH secretion pattern during prolonged FR, primarily by increasing basal, nonpulsatile secretion. Thus, low liver-derived IGF-1 explains the increase in GH secretion observed during starvation. Increasing GH secretion produces numerous beneficial effects during FR because GHR signaling mobilizes fatty acids as a fuel source (101), increases hepatic glucose production to maintain glycemia (36, 38, 42, 43, 102), and stimulates hunger, energy-saving adaptations, and food-seeking behavior via the hypothalamus (25, 40, 41, 103-105). These GH-induced metabolic and behavioral effects contribute to survival during prolonged FR (34, 35), while low hepatic IGF-1 production hinders growth as food availability remains scarce.

Abbreviations

AAV

adeno-associated virus

AgRP

agouti-related protein

ANOVA

analysis of variance

ARH

arcuate nucleus of the hypothalamus

AUC

area under the curve

FR

food restriction

GH

growth hormone

GHR

growth hormone receptor

GHRH

growth hormone–releasing hormone

GHS-R

growth hormone secretagogue receptor

IGF-1

insulin-like growth factor-1

IGF1R

insulin-like growth factor-1 receptor

IGFBP

insulin-like growth factor-binding protein

KO

knockout

LEAP2

antimicrobial peptide 2

LEPROT

leptin receptor overlapping transcript

mRNA

messenger RNA

NPY

neuropeptide Y

PCR

polymerase chain reaction

SIRT1

sirtuin 1

SOCS

suppressor of cytokine signaling

SST

somatostatin

WT

wild-type

Funding

This work was supported by the Fundação de Amparo à Pesquisa do Estado de São Paulo (grant No. 2020/01318-8 to J.D., and No. 2023/11833-5 to L.M.M.S.), Conselho Nacional de Desenvolvimento Científico e Tecnológico (306024/2023-3 to J.D.), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (Finance Code 001 to M.E.S.), National Institutes of Health (National Institute on Aging grant No. R01AG059779 to J.J.K. and E.O.L.), and the US Department of Veterans Affairs (grant No. BX004448 to R.D.K.).

Disclosures

J.D. and R.D.K. are editorial board members for Endocrinology and played no role in the journal’s manuscript evaluation.

Data Availability

The data supporting this study’s findings are available from the corresponding author on reasonable request.