Immunohistochemical Analysis of Neurotransmitters in Neurosecretory Protein GL-Producing Neurons of the Mouse Hypothalamus

We recently discovered a novel neuropeptide of 80 amino acid residues: neurosecretory protein GL (NPGL), in the hypothalamus of birds and rodents. NPGL is localized in the lateral posterior part of the arcuate nucleus (ArcLP), and it enhances feeding behavior and fat accumulation in mice. Various neurotransmitters, such as catecholamine, glutamate, and γ-aminobutyric acid (GABA), produced in the hypothalamus are also involved in energy metabolism. The colocalization of neurotransmitters and NPGL in neurons of the ArcLP leads to the elucidation of the regulatory mechanism of NPGL neurons. In this study, we performed double immunofluorescence staining to elucidate the relationship between NPGL and neurotransmitters in mice. The present study revealed that NPGL neurons did not co-express tyrosine hydroxylase as a marker of catecholaminergic neurons and vesicular glutamate transporter-2 as a marker of glutamatergic neurons. In contrast, NPGL neurons co-produced glutamate decarboxylase 67, a marker for GABAergic neurons. In addition, approximately 50% of NPGL neurons were identical to GABAergic neurons. These results suggest that some functions of NPGL neurons may be related to those of GABA. This study provides insights into the neural network of NPGL neurons that regulate energy homeostasis, including feeding behavior and fat accumulation.

Neurotransmitters are also involved in the regulation of feeding behavior [6,21]. Serotonin suppresses appetite [2]. Dopamine attenuates appetite in the Arc and the LHA, whereas it increases appetite in the ventromedial hypothalamus (VMH) [22][23][24]. Glutamate participates in the regulation of feeding behavior [21]. In addition, γ-aminobutyric acid (GABA), an inhibitory neurotransmitter produced by NPY/AgRP neurons, enhances feeding behavior by inhibiting anorexigenic neurons, such as POMC neurons [25,26]. Thus, feeding behavior is intricately modulated by various neurotransmitters and neuropeptides. However, this complex system remains to be fully elucidated. To further understand the mechanisms of feeding behavior and energy homeostasis, we sought to identify previously unknown hypothalamic neuropeptides in vertebrates.
Recently, we discovered a novel gene related to feeding behavior and energy homeostasis in the chicken hypothalamus [27]. The small secretory protein derived from this novel gene was named neurosecretory protein GL (NPGL) because the C-terminal amino acid sequence is Gly-Leu-NH 2 [27]. Genome database analysis suggested that the homologous gene of NPGL is conserved in vertebrates, including chickens, mice, rats, and humans [27][28][29].
Functional analysis of NPGL has been conducted in mammalian models, such as mice and rats. Npgl mRNA expression is increased by fasting in mice, whereas it is reduced by a high-fat diet [29]. In addition, the administration of NPGL and Npgl overexpression in the hypothalamus increases food intake and fat accumulation in rats and mice [28,30,31]. These data suggest that NPGL is involved in the energy metabolism of rodents.
In a previous study, we found that NPGL-producing cell bodies are localized in the lateral posterior part of the Arc (ArcLP), which is involved in feeding regulation, and fibers of NPGL neurons project to several hypothalamic regions, including POMC neurons in mice [29]. Furthermore, some NPGL neurons co-produce the orexigenic neuropeptide galanin in mice [31]. However, the production of neurotransmitters by NPGL-producing neurons remains unknown.
The information will provide vital knowledge to clarify the regulatory mechanisms of NPGL neurons on energy metabolism. The aim of this study was to elucidate the neural characteristics of NPGL neurons. In this study, we performed double immunofluorescence staining to investigate the relationship between NPGL and neurotransmitters, that is, catecholamine, glutamate, and GABA, in the mouse hypothalamus using antibodies against marker proteins of neurotransmitters.

Animals
C57BL/6J mice (n = 4-6 in each double immunofluorescence staining) were purchased from Nihon SLC (Shizuoka, Japan) and singly housed under standard conditions (25 ± 2 • C under a 12-h light/12-h dark cycle) with ad libitum access to water and normal chow (CE-2; CLEA Japan, Tokyo, Japan). Animal surgery was performed under isoflurane anesthesia. All animal experiments were performed in accordance with 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 number: C19-18).

Immunohistochemistry
In stereotaxic surgery, male mice (8 weeks old) were injected with colchicine (30 µg/ 2.5 µL) into the lateral ventricle under isoflurane anesthesia. The colchicine treatment enabled the detection of NPGL-immunoreactive cell bodies [29]. After 2 days of colchicine treatment, the brains were cut into 20-or 40-µm sections with a cryostat at -20 • C. The procedure using immunofluorescence staining of the floating sections was conducted as previously described [29,31].
The co-expression of NPGL-immunoreactive cell bodies with neurotransmitters was surveyed using double-label immunofluorescence as follows: a guinea pig antibody against NPGL (1:500 dilution) and a rabbit antibody against tyrosine hydroxylase (TH), as a marker of catecholaminergic neurons (1:2000 dilution, AB152; Merck Millipore, Burlington, MA, USA), were used for the detection of NPGL neurons. For catecholaminergic neurons, a rabbit antibody against NPGL (1:250 dilution), and a guinea pig antibody against vesicular glutamate transporter 2 (vGLUT2), as a marker of glutamatergic neurons (1:1000 dilution, RRID_2571621; Frontier Institute Co., Ltd, Hokkaido, Japan), were used for the detection of NPGL neurons and glutamatergic neurons.

NPGL-Immunoreactive Cells Co-Expressing with GAD67 Counting
For quantitative analysis of NPGL-immunoreactive cells co-expressing GAD67, images of all localized NPGL-immunoreactive cells were photographed. Based on the observed images, the total number of NPGL-immunoreactive cells was counted at 20 µm intervals, that is, skipping one slice. Next, the number of NPGL-immunoreactive cells co-expressing GAD67 was counted, and the ratio of the number of NPGL-immunoreactive cells coexpressing GAD67 to the total number of NPGL-immunoreactive cells was calculated. All results are presented as the mean ± standard error of the mean (SEM, n = 6).

Double Immunofluorescence Staining of NPGL and TH as a Marker of Catecholaminergic Neurons
To examine the co-expression of NPGL and catecholamine in the hypothalamus, double immunofluorescence staining was performed using antibodies against NPGL and TH. NPGL-immunoreactive cells were localized in the ArcLP ( Figure 1B,E). TH-immunoreactive cells were also observed in the neighborhood of NPGL-producing cells in the ArcLP ( Figure 1C,F), in addition to other regions of the ArcLP ( Figure 1C). The merged image showed that NPGL-immunoreactive cells were different from TH-positive neurons, that is, catecholaminergic neurons ( Figure 1D,G). Similar results were obtained in repeated experiments with four mice.

Double Immunofluorescence Staining of NPGL and vGLUT2 as a Marker of Glutamatergic Neurons
Next, we investigated the colocalization of NPGL and glutamatergic neurons using an antibody against vGLUT2. NPGL-immunoreactive cells were localized in the ArcLP ( Figure 2B,E,H). vGLUT2-immunoreactivities were observed in the granular form in the hypothalamus ( Figure 2C,F,I). The merged images showed that NPGL-immunoreactive cells did not contain vGLUT2-immunoreactive granules ( Figure 2D,G,J). Similar results were obtained in repeated experiments with four mice.

Double Immunofluorescence Staining of NPGL and vGLUT2 as a Marker of Glutamatergic Neurons
Next, we investigated the colocalization of NPGL and glutamatergic neurons using an antibody against vGLUT2. NPGL-immunoreactive cells were localized in the ArcLP ( Figure 2B,E,H). vGLUT2-immunoreactivities were observed in the granular form in the hypothalamus ( Figure 2C,F,I). The merged images showed that NPGL-immunoreactive cells did not contain vGLUT2-immunoreactive granules ( Figure 2D,G,J). Similar results were obtained in repeated experiments with four mice.

Double Immunofluorescence Staining of NPGL and GAD67 as a Marker of GABAergic Neurons
In the final experiment, we analyzed the colocalization of NPGL and GABAergic neurons using an antibody against GAD67. NPGL-immunoreactive cells were localized in the ArcLP ( Figure 3B,E). GAD67-immunoreactive cells were also localized in the ArcLP, in addition to other hypothalamic regions ( Figure 3C,F). The merged images indicated that NPGL was co-expressed with GAD67 in some NPGL-immunoreactive cells ( Figure 3D,G). Similar results were obtained in repeated experiments with six mice. The number of NPGL-immunoreactive cells expressing GAD67 was counted. Based on the estimated number of cells, approximately 50% of NPGL cells were co-produced with GAD67, showing GABAergic neurons (Table 1).   addition to other hypothalamic regions ( Figure 3C,F). The merged images indicated that NPGL was co-expressed with GAD67 in some NPGL-immunoreactive cells ( Figure 3D,G). Similar results were obtained in repeated experiments with six mice. The number of NPGL-immunoreactive cells expressing GAD67 was counted. Based on the estimated number of cells, approximately 50% of NPGL cells were co-produced with GAD67, showing GABAergic neurons (Table 1).

Discussion
Our previous studies have indicated that NPGL participates in energy homeostasis, including feeding behavior and fat accumulation in rodents [28][29][30][31]. In addition, we found that NPGL cell bodies are localized in the ArcLP, and NPGL-containing fibers project to several anterior hypothalamic regions in mice [29]. However, the colocalization of neurotransmitters in NPGL-producing neurons remains unknown. In this study, we used an immunohistochemical analysis to investigate whether NPGL neurons produce other neurotransmitters in the mouse hypothalamus. The data showed that a subset of NPGLproducing neurons were GABAergic. These results suggest that GABA is also involved in the neural network and/or function of NPGL neurons.
Catecholamines, including adrenaline, noradrenaline, and dopamine, are excitatory neurotransmitters. In rodents, the injection of noradrenaline or dopamine into the hypothalamus stimulates and inhibits feeding behavior, respectively [2,32]. Glutamate is an excitatory neurotransmitter [21]. In the hypothalamus, some glutamatergic neurons participate in the regulation of feeding behavior [21,33]. However, the present study revealed that NPGL-producing cells did not co-produce TH and vGLUT2. Hence, NPGL neurons are not identical to the catecholaminergic and glutamatergic neurons.
GABA is an inhibitory neurotransmitter in adult mammals [21,26]. In the hypothalamus, GABA is involved in feeding regulation [26]. In the Arc, GABA produced in NPY/AgRP neurons is orexigenic [21,26,34]. In contrast, approximately half of anorexigenic POMC neurons are GABAergic neurons that project to the dorsomedial hypothalamus (DMH) [35]. MCH-expressing neurons containing GABA in the LHA project to several parts of the hypothalamus, and MCH enhances feeding behavior via GABA [26]. In the DMH, GABAergic neurons project to the paraventricular nucleus of the hypothalamus (PVH) and stimulate feeding behavior [36].
Taken together, these results show that GABA is an important factor in regulating feeding behaviors in both stimulatory and inhibitory manners in the different hypothalamic regions. In the present study, approximately 50% of NPGL neurons were identical to GABAergic neurons. Previous studies have revealed that fibers of NPGL neurons in the ArcLP innervate POMC neurons in the Arc [29]. Some orexigenic NPY/AgRP neurons exert hyperphagic effects by inhibiting POMC neurons via GABA [25].
Therefore, similar to NPY/AgRP neurons, NPGL may exert its hyperphagic effects by inhibiting POMC neurons via GABA. However, it is unclear whether NPGL neurons containing GABA project to the POMC neurons. In the future, it will be necessary to elucidate whether NPGL neurons innervate POMC neurons with or without GABA using retrograde tracers. This type of study will reveal the connection between NPGL/GABA, NPGL, and POMC neurons.
In addition to the cell bodies of NPGL in the ArcLP, fibers of NPGL neurons spread to several areas of the hypothalamus, including the Arc, PVH, DMH, VMH, and LHA [29]. There are some anorexigenic factors, such as corticotropin-releasing hormone (CRH) in the PVH and pituitary adenylate cyclase-activating polypeptide (PACAP), brain-derived neurotrophic factor (BDNF), and steroidogenic factor 1 (SF-1) in the VMH [37][38][39][40]. Furthermore, it has been reported that endogenous bioactive peptides, such as kisspeptin and hemopressin as well as exogenous peptides derived from food proteins, participate in appetite control through the orexigenic and anorexigenic factors as well as the receptors for neurotransmitters or cannabinoids [41][42][43].
Therefore, NPGL neurons may influence these various factors or receptors to elicit some important biological functions, including energy homeostasis, feeding behavior, and instinctive behavior. Information on the receptor for NPGL is essential to elucidate the target sites of NPGL. However, a receptor for NPGL has not yet been identified. On the other hand, much attention has been focused on the relationship between eating disorder and dysfunction of neuropeptides [44].
Several synthetic bioactive peptides related to urocortin 3, growth hormone-releasing hormone, neuromedin U, and kisspeptin, have antidepressant-like or anxiety-like effects via the receptors for neurotransmitters or the release of neurotransmitters [45][46][47][48]. The dysregulations of neurotransmitters, including GABA and glutamate cause mood disorders and Alzheimer's disease [49,50]. Taken together, further studies of NPGL may uncover the mechanisms of eating disorders, including anorexia nervosa and brain dysfunctions.
In the present study, approximately 50% of NPGL-producing cells did not contain GABA. About half of the POMC neurons are also GABAergic; however, some POMC neurons express vGLUT2, a marker of glutamatergic neurons [35]. These data suggest that different subpopulations of POMC neurons play diverse physiological roles [35]. Therefore, it is likely that NPGL-producing neurons, like POMC neurons, have multiple neuronal networks and physiological functions.
Previous studies have revealed that NPGL is co-expressed with galanin, which exerts feeding behavior [51], in approximately 30% of NPGL-producing cells in mice [31]. In addition, some galanin-producing neurons are GABAergic [52,53]. Therefore, it is necessary to analyze whether galanin neurons containing GABA are co-expressed with NPGL using triple-immunostaining for NPGL, GABA, and galanin in future studies. Hence, NPGLproducing cells may have heterogeneities and/or subclasses based on neurotransmitter and neuropeptide phenomena.
In the future, to clarify the more complex characteristics of NPGL-producing cells containing GABA and other neuropeptides, single-cell transcriptional analysis of NPGL