Neurosecretory Protein GL Accelerates Liver Steatosis in Mice Fed Medium-Fat/Medium-Fructose Diet

Sugar consumption can readily lead to obesity and metabolic diseases such as liver steatosis. We previously demonstrated that a novel hypothalamic neuropeptide, neurosecretory protein GL (NPGL), promotes fat accumulation due to the ingestion of sugar by rats. However, differences in lipogenic efficiency of sugar types by NPGL remain unclear. The present study aimed to elucidate the obesogenic effects of NPGL on mice fed different sugars (i.e., sucrose or fructose). We overexpressed the NPGL-precursor gene (Npgl) in the hypothalamus of mice fed a medium-fat/medium-sucrose diet (MFSD) or a medium-fat/medium-fructose diet (MFFD). Food intake and body mass were measured for 28 days. Body composition and mRNA expression of lipid metabolic factors were measured at the endpoint. Npgl overexpression potently increased body mass with fat accumulation in the white adipose tissue of mice fed MFFD, although it did not markedly affect food intake. In contrast, we observed profound fat deposition in the livers of mice fed MFFD but not MFSD. In the liver, the mRNA expression of glucose and lipid metabolic factors was affected in mice fed MFFD. Hence, NPGL induced liver steatosis in mice fed a fructose-rich diet.


Introduction
Diets with unbalanced nutrients lead to obesity and ectopic fat deposition, resulting in metabolic syndrome [1,2]. Recent studies have revealed that the ingestion of sugars, such as sucrose and fructose, profoundly aggravates type 2 diabetes, fatty liver, and hypertension due to overnutrition and changes in systemic energy metabolism [3][4][5]. Thus, detailed research focusing on the relationship between dietary sugar and energy metabolism is required to treat metabolic syndromes.
Several studies have demonstrated that systemic energy metabolism is controlled by hypothalamic neuropeptides and peripheral hormones. The arcuate nucleus in the hypothalamus, which regulates feeding and energy metabolism, produces potent feeding-related neuropeptides. For instance, neuropeptide Y (NPY) and agouti-related peptide (AgRP) are well-known orexigenic neuropeptides [6][7][8]. Conversely, α-melanocyte-stimulating hormone (α-MSH), derived from proopiomelanocortin (POMC), functions as an anorexigenic neuropeptide via melanocortin receptor type 4 (MC4R) [9]. In contrast, leptin and insulin are peripheral hormones involved in energy metabolism. Leptin is secreted by white adipose tissue (WAT) and acts as an anorexigenic hormone by inhibiting the activity of NPY/AgRP neurons [10]. Insulin stimulates lipid deposition by inhibiting lipolysis [11]. In addition, a few studies have reported that neuropeptides and hormones manipulate feeding behavior and energy metabolism depending on the nutritional composition of diets, especially carbohydrates. Corticotropin-releasing hormone (CRH) exerts palatable effects on carbohydrates in mice [12], whereas fibroblast growth factor 21 (FGF21) inhibits sugar intake [13]. Although there is accumulating evidence, which suggests that energy metabolic regulation via the endocrine system participates in the intake of carbohydrates, there are few observations concerning the relationships between types of sugar consumed and the endocrine system.
To understand the mechanisms involved in energy metabolism, we recently identified a novel gene from the hypothalamus of chickens, rats, mice, and humans [14][15][16]. Since the novel gene produces a small secretory protein whose C-terminus amino acid sequence is Gly-Leu-NH 2 , it was termed neurosecretory protein GL (NPGL) [14]. The primary structure of NPGL is evolutionarily conserved in vertebrates, although the receptor for NPGL has not yet been discovered [17]. To reveal the physiological function of NPGL, we observed its effects on chickens, rats, and mice. Chronic administration of NPGL enhanced lipid metabolism in chickens [18]. Furthermore, NPGL promoted the intake of carbohydrate and fat accumulation via de novo lipogenesis in rats [15,19]. Moreover, overexpression of the NPGL-precursor gene (Npgl) rapidly induced obesity in mice [20]. Thus, we speculate that NPGL exerts obesogenic effects by evoking the palatability of sugars in birds and mammals. Although sucrose and fructose are well-known as representative sugars in human nutrition [5], the sugars that are easily used in lipogenesis by NPGL remain unclear.
To reveal the lipogenic efficiency of NPGL with different types of sugar, we induced hypothalamic overexpression of Npgl in mice fed a medium-fat/medium-sucrose diet (MFSD) or a medium-fat/medium-fructose diet (MFFD). Here, we report that more potent liver steatosis was observed in Npgl overexpressed mice fed MFFD than in mice fed MFSD. This study reveals the effects of Npgl overexpression on food intake, body mass, body composition, and blood parameters in mice.

Effects of NPGL-Precursor Gene Overexpression on Food Intake, Body Mass, and Food Efficiency
To reveal the effects of NPGL on lipid metabolism attributed to the different types of sugars, we overexpressed Npgl using adeno-associated virus (AAV) in mice fed MFSD or MFFD. Two-way repeated-measures ANOVA showed the main effect of time and time × group interaction on food intake ( Figure 1A). One-way ANOVA revealed that food intake was not affected by Npgl overexpression in mice ( Figure 1B). The main effects of time, group, and time × group interaction were significant on body mass as indicated by two-way repeated-measures ANOVA ( Figure 1C). The effects of Npgl overexpression on body mass gain were observed at early phases (day 4 after surgery) in mice fed MFFD ( Figure 1C). One-way ANOVA showed that body mass was increased by Npgl overexpression in mice fed the two diets ( Figure 1D). We subsequently calculated food efficiency, which is an index of the amount of mass gained per unit of food intake [20]. Two-way ANOVA with repeated measures revealed the main effects of time and treatment on food efficiency ( Figure 1E). Bonferroni's test showed that food efficiency was increased by Npgl overexpression under the two diets ( Figure 1E). Food efficiency was significantly augmented by Npgl overexpression at the endpoint ( Figure 1F). Based on these data, Npgl overexpression stimulated body mass gain without changing food intake in mice, especially when fed MFFD. Figure 1. Effects of Npgl overexpression on food intake, body mass, and food efficiency. The panels show the data obtained after injection of AAV-CTL or AAV-NPGL into mice fed MFSD or MFFD for 28 days. (A) Cumulative food intake at all points. (B) Cumulative food intake 28 days after injection. (C) Body mass at all points. (D) Body mass 28 days after injection. (E) Food efficiency expressed as body weight gain per cumulative food intake per week at all points. (F) Food efficiency 28 days after injection. Each value represents the mean ± standard error of the mean (n = 5-6/group). * p < 0.05, *** p < 0.005 AAV-CTL (MFSD) vs. AAV-NPGL (MFSD) in the same period by one-way ANOVA with Tukey's test for multiple comparisons, † p < 0.05, † † p < 0.01, † † † p < 0.005 AAV-CTL (MFFD) vs. AAV-NPGL (MFFD) in the same period by one-way ANOVA with Tukey's test for multiple comparisons, ‡ ‡ ‡ p < 0.005 AAV-CTL (MFSD) vs. AAV-NPGL (MFSD) by two-way ANOVA by  Each value represents the mean ± standard error of the mean (n = 5-6/group). * p < 0.05, *** p < 0.005 AAV-CTL (MFSD) vs. AAV-NPGL (MFSD) in the same period by one-way ANOVA with Tukey's test for multiple comparisons, † p < 0.05, † † p < 0.01, † † † p < 0.005 AAV-CTL (MFFD) vs. AAV-NPGL (MFFD) in the same period by one-way ANOVA with Tukey's test for multiple comparisons, ‡ ‡ ‡ p < 0.005 AAV-CTL (MFSD) vs. AAV-NPGL (MFSD) by two-way ANOVA by repeated measures with Bonferroni's test for multiple comparisons, § § § p < 0.005 AAV-CTL (MFFD) vs. AAV-NPGL (MFFD) by two-way ANOVA by repeated measures with Bonferroni's test for multiple comparisons. NPGL, neurosecretory protein GL; AAV-CTL, AAV-based control vector; AAV-NPGL, AAV-based NPGL-precursor gene vector; MFSD, medium-fat/medium-sucrose diet; MFFD, mediumfat/medium-fructose diet.

Effects of NPGL-Precursor Gene Overexpression on Body Composition and Serum Parameters
To determine why Npgl overexpression induced an increase in body mass, we measured the masses of adipose tissues, muscles, and several organs (Figures 2 and 3). In mice fed MFSD, the masses of epididymal WAT (eWAT) and perirenal WAT (pWAT) were significantly increased compared to those of control mice. Inguinal WAT (iWAT) and retroperitoneal WAT (rWAT) were slightly affected, but the difference was not significant ( Figure 2A). Under MFFD, the masses of iWAT, eWAT, rWAT, and pWAT were increased by Npgl overexpression (Figure 2A). Hematoxylin and eosin staining revealed larger adipocytes in the iWAT of mice fed both diets ( Figure 2B). In contrast, the mass of the gastrocnemius muscle was not affected by Npgl-overexpressing mice under the two dietary conditions ( Figure 3A). Subsequently, we measured the masses of peripheral organs. The mass of the liver was increased in mice fed MFFD but not in mice fed MFSD ( Figure 3B). Oil Red O staining showed steatosis in the liver of Npgl-overexpressing mice ( Figure 3C). Measurement of blood glucose, insulin, and lipid levels in Npgl-overexpressing mice fed MFSD indicated an increase in circulating insulin levels, but no change was observed in the other serum parameters ( Figure 4A-E).

Effects of NPGL-Precursor Gene Overexpression on Body Composition and Serum Parameters
To determine why Npgl overexpression induced an increase in body mass, we measured the masses of adipose tissues, muscles, and several organs (Figures 2 and 3). In mice fed MFSD, the masses of epididymal WAT (eWAT) and perirenal WAT (pWAT) were significantly increased compared to those of control mice. Inguinal WAT (iWAT) and retroperitoneal WAT (rWAT) were slightly affected, but the difference was not significant (Figure 2A). Under MFFD, the masses of iWAT, eWAT, rWAT, and pWAT were increased by Npgl overexpression (Figure 2A). Hematoxylin and eosin staining revealed larger adipocytes in the iWAT of mice fed both diets ( Figure 2B). In contrast, the mass of the gastrocnemius muscle was not affected by Npgl-overexpressing mice under the two dietary conditions ( Figure 3A). Subsequently, we measured the masses of peripheral organs. The mass of the liver was increased in mice fed MFFD but not in mice fed MFSD ( Figure 3B). Oil Red O staining showed steatosis in the liver of Npgl-overexpressing mice ( Figure 3C). Measurement of blood glucose, insulin, and lipid levels in Npgl-overexpressing mice fed MFSD indicated an increase in circulating insulin levels, but no change was observed in the other serum parameters ( Figure 4A-E).

Discussion
Sugar is the root cause of many diseases, such as metabolic syndrome, via dysfunction of systemic energy metabolism. We recently showed that hypothalamic overexpression of Npgl, a novel precursor gene encoding a small protein, induces obesity by enhancing the palatability of carbohydrates in rodents [15,20]. However, the difference in lipogenic efficiency of NPGL with different types of sugars remains unclear. In this study, we induced hypothalamic overexpression of Npgl and subsequently found potent fat accumulation and liver steatosis in mice fed MFFD relative to those consuming MFSD. The present data imply that NPGL exacerbates fatty liver in fructose-fed mice.
We confirmed the presence of fatty liver in Npgl-overexpressing mice fed MFFD by determining the hepatic mass, Oil Red O staining, and upregulation of Pparγ mRNA expression. Excess input of lipids over output in the liver cause steatosis as well as systemic lipid accumulation [21]. Liver steatosis is not only caused by de novo lipogenesis but also by the intake of free fatty acids. A previous study reported that a high-fructose diet enhances the release of fatty acids from adipocytes [22]. In addition, lipid oxidation is involved in the output (i.e., lipids released from the liver) [21]. Exposure to fructose suppresses lipid oxidation in the liver [23,24]. In the present data analyzed by two-way ANOVA, the mRNA expression of factors related to lipid oxidation, such as Cpt1a, Atgl, Hsl, and Pparα were downregulated in the liver of mice fed MFFD. Notably, one-way ANOVA showed that Cpt1a and Pparα were downregulated in Npgl-overexpressing mice fed MFFD compared to those consuming MFSD. Moreover, two-way ANOVA demonstrated that the mRNA expression of Cd36, which is involved in fatty acid intake, was significantly upregulated by Npgl overexpression. Hence, we speculate that the synergistic effects of fructose and NPGL exacerbate liver steatosis.
Even though our previous study reported the orexigenic effects of NPGL [15,16,25], the present study emphasized that Npgl overexpression does not affect sugar-rich food intake. The endocrine system, including neuropeptides and hormones, regulate nutrient preferences. CRH-positive neurons participate in the selection of high-carbohydrate diets [12]. In addition, FGF21, a hormone secreted by the liver, downregulates sweet-seeking behavior and food intake [13]. Our data indicated that the mRNA expression of Fgf21 was upregulated in the livers of Npgl-overexpressing mice. Thus, it is possible that the orexigenic effects of NPGL were masked by the anorexigenic effects of FGF21 against sugar. Recently, we reported that the effects of NPGL on feeding behavior depend on dietary nutritional composition in rats [19]. Further studies are needed to understand the relationship between the regulation of feeding behavior by NPGL and the nutritional components of different diets.
This study revealed that Npgl overexpression increased serum insulin levels in mice fed MFSD, whereas insulin levels hardly increased in mice fed MFFD. Some reports have demonstrated that the effects of sugar on insulin secretion depend on the type of sugar involved. A sucrose-rich diet promotes glucose-induced insulin secretion [26]. Meanwhile, excess intake of high-fructose corn syrup leads to impaired glucose tolerance due to insulin secretion deficiency [27]. We recently reported that Npgl overexpression increases circulating insulin levels in mice fed a high-calorie diet, including medium-sucrose [20]. Hence, the effects of NPGL on insulin secretion were disturbed by ingestion of MFFD.
Further research is required to elucidate the effects of NPGL on insulin secretion under different dietary components involving sugar.
The present study had several limitations. First, we assessed lipid metabolism in iWAT and liver using qRT-PCR. In iWAT, Npgl overexpression had a limited effect on the mRNA expression. Since lipid metabolic factors are controlled at both transcriptional and post-translational levels, NPGL may affect lipid metabolism at the protein level in iWAT. To date, the receptor for NPGL has not been found. Additional studies to measure the enzymatic activities of lipid metabolic factors and to identify the receptor for NPGL will help to understand the lipid metabolic regulation by NPGL. In addition, since previous studies indicate that the lipogenic effects of NPGL are more potent in rodents fed a fat-rich diet [15,20,28], we used MFSD or MFFD, which includes fat as well as sucrose or fructose in this study. However, it is well-known that dietary fat affects energy metabolism [29]. For instance, ChREBP, a key regulator of de novo lipogenesis, is activated in response to glucose and fructose and inactivated by an increase in fatty acids [30,31]. Thus, several studies into low-fat/high-sucrose or high-fructose conditions will enable us to investigate the lipogenic effects of NPGL under different sugar types in detail. Moreover, the present study showed that Npgl overexpression had little effect on serum parameters, although it led to excess adiposity in mice fed MFFD. Several reports have demonstrated that the intake of dietary fructose causes metabolic disorders [4,5]. The present data imply that Npgl overexpression in the moderate term (i.e., 28 days) maintains a steady metabolic state, although it easily provokes fatty liver in mice fed MFFD. Further studies involving long-term analysis will open new avenues into the relationship between types of sugar consumed and metabolic abnormalities.
In conclusion, we showed that NPGL augments the effects of fructose on lipid accumulation, including fatty liver, using AAV-induced overexpression in mice. This is the first report demonstrating differences in the lipogenic efficiency of NPGL according to the type of sugar consumed. Progress in research surrounding the effects of NPGL on lipid metabolism in several nutritional conditions, including types of sugar or fatty acids, will help us to better understand the relationship between hypothalamic regulation and obesityrelated diseases such as nonalcoholic steatohepatitis under different nutritional conditions.

Animals
Male C57BL/6J mice (7 weeks old) were purchased from Nihon SLC (Hamamatsu, Japan) and individually housed under standard conditions (25 ± 1 • C under a 12 h light/12 h dark cycle) with ad libitum access to water and MFSD (32% of calories from fat, 20% of calories from sucrose, D14050401; Research Diets, New Brunswick, NJ, USA) or MFFD (32% of calories from fat, 20% of calories from fructose, D19061101; Research Diets). Nutritional compositions are shown in Table 3. Animal surgery was conducted under isoflurane anesthesia. All animal experiments were performed according to the Guide for the Care and Use of Laboratory Animals prepared by Hiroshima University (Higashi-Hiroshima, Japan), and these procedures were approved by the Institutional Animal Care and Use Committee of Hiroshima University (permit numbers: C19-8, 30 August 2019; C21-1, 19 April 2021).

Production of AAV-Based Vectors
AAV-based vectors were generated following a previously reported method [20,32]. In this study, the primers for mouse NPGL were 5'-CGATCGATACCATGGCTGATCCTGGGC-3' for the sense primer and 5'-CGGAATTCTTATTTTCTCTTTACTTCCAGC-3' for the antisense primer. AAV-based vectors were prepared at a concentration of 1 × 10 9 particles/µL and stored at −80 • C until use.

Stereotaxic Surgery
Npgl overexpression was conducted as previously described [20,32]. Mice were bilaterally injected with 0.5 µL/site (5.0 × 10 8 particles/site) of AAV-based vectors (AAV-NPGL or AAV-CTL) using a Neuros Syringe (7001 KH; Hamilton, Reno, NV, USA) into the mediobasal hypothalamic region with the coordinates 2.2 mm caudal to the bregma, 0.25 mm lateral to the midline, and 5.8 mm ventral to the skull surface. Npgl overexpression was maintained for 28 days in mice fed with MFSD or MFFD. Npgl overexpression was confirmed by qRT-PCR at the endpoint ( Figure S1).

Measurement of Body Mass, Food Intake, and Body Composition
The mice were divided into two groups according to their diet (MFSD or MFFD). Food intake and body mass were measured at the beginning of the light period (9:00). Food efficiency (g/kcal) was calculated as body mass gain (g)/cumulative food intake (kcal) [33]. A total of 28 days after stereotaxic surgery, the mice were decapitated between 13:00 and 15:00. The mediobasal hypothalamus, adipose tissues, organs, and skeletal muscles were collected, weighed, and frozen in liquid nitrogen. Blood was collected at the same time as the mice were sacrificed.

qRT-PCR
Total RNA was extracted using QIAzol lysis reagent (QIAGEN, Venlo, Netherlands) for iWAT or TRIzol reagent (Life Technologies, Carlsbad, CA, USA) for hepatic tissue and the mediobasal hypothalamus, according to the manufacturer's instructions. First-strand cDNA was synthesized from total RNA using a ReverTra Ace kit (TOYOBO, Osaka, Japan).
Sequences of primers used in this study are listed in Table 4. PCR amplifications were conducted with THUNDERBIRD SYBR qPCR Mix (TOYOBO) using the following conditions: 95 • C for 20 s, followed by 40 cycles each consisting of 95 • C for 3 s and 60 • C for 30 s. The PCR products in each cycle were monitored using a Bio-Rad CFX Connect (Bio-Rad Laboratories, Hercules, CA, USA). Relative quantification of each gene was determined by the 2 −∆∆Ct method using ribosomal protein S18 (Rps18) for iWAT or β-actin (Actb) for the liver and mediobasal hypothalamus as an internal control [34]. The expression of internal control genes was verified to be stable across the experimental groups.

Hematoxylin and Eosin Staining
iWAT was soaked in 4% paraformaldehyde at the endpoint of Npgl overexpression, embedded in paraffin, and sectioned to a thickness of 8 µm using a microtome. The sections were then air-dried and deparaffinized in a graded alcohol series. Nuclei and cytoplasm were stained with hematoxylin and eosin (5 min for each stain), and the sections were washed with tap water. After dehydration in a graded alcohol series and clearing with xylene, the sections were mounted on slides and examined under a microscope. Table 4. Sequences of oligonucleotide primers for qRT-PCR.