Deletion of the Neurotrophic Factor neudesin Prevents Diet-induced Obesity by Increased Sympathetic Activity

Some neurotrophic factors, which are potent regulators of neuronal development and function, have recently been implicated in the control of energy balance by increasing energy expenditure. We previously identified neudesin as a novel neurotrophic factor with potential roles in the central nervous system. Although neudesin is also expressed in various peripheral tissues including adipose tissue, its physiological roles have not yet been elucidated. We found that neudesin knockout (KO) mice were resistant to high-fat diet-induced obesity and obesity-related metabolic dysfunctions. neudesin KO mice exhibited increased energy expenditure due to increased sympathetic activity, which resulted in increased heat production and fatty acid oxidation in brown adipose tissue and enhanced lipolysis in white adipose tissue. Thus, neudesin, which may be a negative regulator of sympathetic activity, could represent a novel regulator of the development of obesity and obesity-related metabolic dysfunctions.

Thus, neurotrophic factors generally induce anti-obesity effects by suppressing food intake or increasing energy expenditure.
We previously identified the neurotrophic factor, neudesin, which was well conserved in various vertebrates, including humans, mice, and zebrafish [18][19][20] . Previous studies reported that neudesin was expressed in the brain and spinal cord of embryonic mice and appears to be essential for neuronal function [18][19][20][21][22] . The administration of recombinant neudesin to the hypothalamus was reported to repress appetite 24 . Novais et al. generated neudesin KO mice and showed that the neudesin deletion resulted in shorter dendrites in the dentate gyrus of the hippocampus and increased anxious-like behaviors 25 . This is the first and only study that described the physiological role of neudesin in the central nervous system (CNS); however, its systemic roles are not yet fully understood. neudesin was also expressed in various peripheral tissues in adult mice, including adipose tissue, and several studies suggested that it may play roles in peripheral tissues 19,20,23 . Thus, determining the system effects of neudesin in various tissues is essential for elucidating its physiological effects. Therefore, we also generated neudesin KO mice that were resistant to high-fat diet (HFD)-induced obesity and obesity-related metabolic dysfunctions. The present findings represent the first report that a neurotrophic factor suppresses energy expenditure. neudesin, which may be a negative regulator of sympathetic activity, could represent a novel regulator of the development of obesity and obesity-related metabolic dysfunctions.

Results
neudesin KO mice were resistant to diet-induced obesity. neudesin was previously reported to be expressed in the CNS and various peripheral tissues of adult mice [18][19][20]23 (Fig. S1A). We generated neudesin KO mice to elucidate its physiological roles. neudesin KO mice showed the expected Mendelian ratios ( Fig. S1B-G) and were apparently normal and fertile. We then examined the body weights and tissue weights of WT and neudesin KO mice. Body weights at 8 weeks of age and E18.5 were similar between wild type (WT) and neudesin KO mice. The weights of principal tissues including the liver, heart and kidney but not WAT at 8 weeks of age were also similar between WT and neudesin KO mice (Fig. 1A, Fig. S2 and Table. S1). Tibia lengths, which reflected growth 26 , were also similar between WT and neudesin KO mice (Fig. S1H). These results suggested that the growth of neudesin KO mice was essentially normal. However, increases in body weight were suppressed in neudesin KO mice as they grew, and at 16 weeks of age, their body weights were lower than those of WT mice. Increases in body weights due to HFD in WT mice were strongly suppressed in neudesin KO mice (Fig. 1A).
We measured the weights of various tissues at 16 weeks of age and the obtained results were summarized in Table 1. The WAT weight was significantly lower in neudesin KO mice than in WT mice when they were fed normal chow (NC). When WT mice were fed HFD, WAT weight was greater than when fed NC, whereas this increase was strongly suppressed in neudesin KO mice (Table 1). We next examined the morphology of white adipocytes (Fig. 1B). The sizes of white adipocytes in neudesin KO mice fed NC were not significantly different from those in WT mice fed NC. HFD caused adipocyte hypertrophy in WT mice, which was significantly suppressed in neudesin KO mice (Fig. 1C).
The weights of the liver and BAT were similar between WT and neudesin KO mice when they were fed NC (Table 1). Increases were observed in the weights of the liver and BAT of WT mice fed HFD, whereas these increases were significantly less in neudesin KO mice fed HFD. We then histologically examined the liver and BAT of WT and neudesin KO mice (Fig. 1D,E). The liver was apparently normal in both WT and neudesin KO mice fed NC, whereas the marked accumulation of lipids was observed in WT mice fed HFD. However, HFD-induced lipid accumulation in the liver was significantly decreased in neudesin KO mice (Fig. 1D). BAT was also apparently normal in both WT and neudesin KO mice fed NC, whereas the formation of large lipid droplets was observed in WT mice fed HFD. The formation of lipid droplets was suppressed in neudesin KO mice (Fig. 1E). The weights of other tissues examined were similar between WT and neudesin KO mice. These results suggested that neudesin KO mice were resistant to diet-induced obesity (DIO). neudesin KO mice were protected from obesity-induced metabolic dysfunctions. We measured various metabolic parameters and the results were summarized in Table 2. The levels of glucose, total cholesterol, and leptin were significantly increased in WT mice fed HFD, and this increase in metabolic parameters was generally suppressed in neudesin KO mice fed HFD. These results suggested that neudesin KO mice were protected from obesity-induced metabolic dysfunctions.
Obesity is known to cause insulin resistance. Blood glucose and plasma insulin levels were similar between WT and neudesin KO mice when fed NC (Table 2). WT mice showed increased glucose and insulin levels when fed HFD, whereas this HFD-induced increase in glucose and insulin levels was suppressed in neudesin KO mice ( Table 2). These results indicated that neudesin KO mice were protected from insulin resistance. We then performed glucose tolerance and insulin tolerance tests. Glucose tolerance in WT mice was impaired by HFD and this effect was significantly improved in neudesin KO mice ( Fig. 2A,B). In parallel, insulin sensitivity was aggravated in WT mice fed HFD, but was also significantly improved in neudesin KO mice (Fig. 2C,D). These results suggested that neudesin KO mice were protected from HFD-induced insulin resistance.
Obesity can also induce the infiltration of macrophages to adipose tissue, which causes inflammation. We examined the expression levels of F4/80, a macrophage marker 27  WT mice fed HFD were significantly higher in mesenteric WAT and tended to be higher in subcutaneous WAT compared with those in WT mice fed NC. However, its expression levels in neudesin KO mice fed HFD were not significantly but tended to be lower than those in WT mice fed HFD (Fig. 2E). Thus, neudesin KO mice might be protected from obesity-induced inflammation. Energy expenditure was increased in neudesin KO mice. We found that neudesin KO mice were resistant to DIO and obesity-induced metabolic dysfunctions including diabetes. We attempted to elucidate why neudesin KO mice were resistant to DIO. Food intake at 16 weeks of age was slightly but significantly lower in neudesin KO mice than in WT mice fed NC, whereas food intake at 16 weeks of age was similar between WT and neudesin KO mice fed HFD (Fig. S3A). Thus, resistance to DIO was independent of food intake. We estimated fat absorption by determining fecal weights and fecal TG contents and found that more than 99% TG in HFD was absorbed in both WT and neudesin KO mice (Fig. S3B). Thus, the deletion of neudesin had little effect on fat absorption.
The enhancement observed in energy expenditure could have induced resistance to DIO. The rectal temperature of neudesin KO mice fed HFD was significantly higher than that of WT mice fed HFD, although voluntary activity was similar between WT and neudesin KO mice (Fig. 3A,B). Respiratory gas analysis showed that oxygen consumption was significantly increased in neudesin KO mice fed HFD (Fig. 3C). The respiratory quotient was significantly lower in neudesin KO mice fed HFD than in WT mice (Fig. 3D). Fatty acid oxidation was significantly increased in neudesin KO mice fed HFD (Fig. 3E). Carbohydrate oxidation was similar between WT mice and neudesin KO mice (Fig. 3F). These results indicated that energy expenditure in neudesin KO mice fed HFD was increased, which contributed to the observed resistance to DIO.
Sympathetic activity was increased in neudesin KO mice. We attempted to elucidate why energy expenditure in neudesin KO mice was increased. As increased sympathetic activity often results in increased energy expenditure 6,28 , we examined the effects of the neudesin deletion on sympathetic activity. Heart rate, which is a common indicator of sympathetic activity 29 , was significantly increased in neudesin KO mice (Fig. 4A).
We next measured catecholamine levels in BAT and WAT, which also reflected sympathetic activity 30 . Norepinephrine levels were essentially similar in BAT and WAT of neudesin KO mice fed NC, whereas norepinephrine levels were significantly higher in BAT and tended to be higher in WAT of neudesin KO mice fed HFD (Fig. 4B,C). Norepinephrine levels in plasma were also significantly increased in neudesin KO mice fed NC and HFD (Fig. 4D). These results suggested that sympathetic activity was increased in neudesin KO mice fed HFD. In addition, we found that recombinant neudesin suppressed the expression  Table 2. Metabolic parameters of WT and neudesin KO mice at 16 weeks of age. Data are means ± standard error of means (SEMs) (n = 4-9; *, p < 0.05, **, p < 0.01, ***, p < 0.001, vs WT mice within the same condition. #, p < 0.05, ###, p < 0.001 vs. feeding within the same genotype) levels of tyrosine hydroxylase (Th), which encodes a rate-limiting enzyme in the synthesis of norepinephrine, in differentiated PC12 cells, which show sympathetic neuron-like properties 31 (Fig. 4E). Thus, neudesin may suppress sympathetic activity by reducing Th expression.
Enhanced sympathetic activity has been shown to stimulate β3-adrenergic receptors (β3AR) in adipose tissue 7 , increase heat production and fatty acid oxidation in BAT 9 , and promote lipolysis in WAT 8 . We then examined the physiological effect of this enhanced sympathetic activity on BAT and WAT. We firstly examined the expression levels of Ucp1, Cpt1, Pparα, and Pgc-1α in BAT. Ucp1 is expressed exclusively in BAT and is indispensible for thermogenesis. CPT1 catalyzes the entry of fatty acids into mitochondria and is crucial for fatty acid oxidation. PPARα and PGC-1α are the transcriptional regulators of thermogenesis and fatty acid oxidation. The expression levels of Ucp1, Cpt1, Pparα, and Pgc-1α in BAT of neudesin KO mice at 4 and 16 weeks of age fed NC were similar to those in WT a mice ( Fig. 5A and Fig. S4). However, these expression levels in BAT of neudesin KO mice at 16 weeks fed HFD were significantly higher than those in WT mice (Fig. 5A). We examined the expression levels of β3AR in BAT. Its levels in WT mice fed HFD were significantly lower than those in WT mice fed NC. However, its levels in neudesin KO mice fed HFD were similar to those in neudesin KO mice fed NC but significantly higher than those in WT mice fed HFD (Fig. 5B). The lipolytic activity of subcutaneous WAT in neudesin KO mice increased slightly when they were fed NC and increased markedly when they were fed HFD (Fig. 5C). These results indicated that sympathetic activity was increased in adipose tissue of neudesin KO mice. Thus, the energy expenditure was increased by the increased sympathetic activity in WAT and BAT of neudesin KO mice, resulting in the resistance to DIO. The hypothalamus greatly contributes to the regulation of sympathetic activity, and several neuropeptides including POMC, NPY, AGRP, BDNF, and CRF in the hypothalamus are the important regulators of sympathetic activity and affect the development of obesity [32][33][34] . To examine the roles of the neuropeptides in the hypothalamus in increased sympathetic activity in neudesin KO mice fed HFD, we examined the expression levels of Pomc, Npy, Agrp, Bdnf, and Crf in the hypothalamus. Their expression levels in neudesin KO mice fed HFD were   essentially comparable to those in WT mice (Fig. S5). The results suggest that these neuropeptides did not contribute to the increased sympathetic activity observed in neudesin KO mice fed HFD.

Discussion
In the present study, we generated neudesin KO mice to elucidate its physiological roles. The results obtained showed that neudesin KO mice were resistant to DIO and obesity-related metabolic dysfunctions. Energy expenditure in neudesin KO mice was increased by increased sympathetic activity. Thus, neudesin appeared to be a suppressor of energy metabolism in vivo.
neudesin was originally identified as a neurotrophic factor. Neurotrophic factors promote the survival, differentiation, and maintenance of neurons in the nervous system. Previous studies suggested that neudesin was expressed in the CNS and required for CNS functions and development [18][19][20][21][22]24,25 . For example, Novais et al. reported that neudesin KO mice exhibited increased anxious-like behaviors caused by shorter dendrites in the dentate gyrus of the hippocampus 25 . Byerly et al. also reported that neudesin was expressed in the paraventricular nuclei (PVN) and arcuate nucleus (ARC), which are important areas of the hypothalamus for regulating appetite 35,36 , and that the administration of recombinant neudesin into the hypothalamus repressed appetite 24 . However, we showed that food intake by neudesin KO mice fed NC was decreased, suggesting that neudesin increases food intake. This discrepancy may be explained by differences in the physiological analysis using neudesin KO mice and pharmacological analysis by the administration of recombinant neudesin. In contrast, as food intake was similar between WT and neudesin KO mice when fed HFD, resistance to DIO in neudesin KO mice was independent of food intake. Resistance to DIO in neudesin KO mice may be mediated by increased sympathetic activity. PVN and ARC also regulate energy expenditure by affecting sympathetic activity 37,38 . However, our results showed that the expression levels of hypothalamic neuropeptides, which are the important regulators of sympathetic activity and affect the development of obesity 32-34 , in neudesin KO mice fed HFD were similar to those in WT mice. Thus, increased sympathetic activity observed in neudesin KO mice fed HFD may not be mediated by hypothalamic regulation in the CNS. We found that recombinant neudesin might suppress Th expression in differentiated PC12 cells with sympathetic neuron-like properties 31 . Additionally, neudesin KO mice exhibited increased levels of norepinephrine in adipose tissue, which are good indicators of sympathetic activity 30 . Thus, peripheral neudesin may suppress sympathetic activity by suppressing Th expression in peripheral neurons. However, further studies are required to elucidate the roles of central and peripheral neudesin in regulating sympathetic activity.
neudesin was shown to be expressed in various peripheral tissues, including WAT and BAT. We could not detect neudesin in plasma (unpublished observation), indicating that it is a local signaling molecule in peripheral tissues. neudesin is expected to act on targeted cells through a specific neudesin receptor, which has not yet been identified [18][19][20][21][22][23] . neudesin in adipose tissue may suppress sympathetic activity by activating the neudesin receptor expressed in peripheral neurons in or near adipose tissue. The expression levels of neudesin in WAT were significantly increased in WT mice fed HFD, whereas its expression in BAT was essentially unchanged (Fig. S6). However, the precise mechanism underlying the increase in sympathetic activity in neudesin KO mice remains unknown. Further studies are required to elucidate the roles of neudesin in regulating sympathetic activity. Additionally, previous studies suggested that neudesin could act on non-neuronal cells besides neurons 23,39 . Thus, some of the functions of neudesin could not be explained by its neurotrophic factor properties. We need to examine the role of neudesin other than as a neurotrophic factor in the future to obtain a better understanding of its physiological roles.
Although neurotrophic factors generally increase energy expenditure, neudesin suppresses it. To the best of our knowledge, this is the first study to show that a neurotrophic factor suppresses energy expenditure. Thus, the present study may shed new light on the physiological roles of neurotrophic factors in energy metabolism. In conclusion, we showed that neudesin was a possible regulator of energy expenditure and may contribute to the development of obesity. neudesin KO mice were resistant to DIO and obesity-related metabolic dysfunctions, and exhibited increased energy expenditure. Increased energy expenditure was attributed to a systemic increase in sympathetic activity. Our results suggested the physiological importance of neudesin in regulating energy expenditure and its role as a suppressor of sympathetic activity. We expect that neudesin could represent a novel regulator of the development of obesity and obesity-related metabolic dysfunctions.

Methods
Gene Targeting. We produced knockout mice in which neudesin was absent in the whole body.
Animal Experiments. All mice were maintained in a light-controlled room (lights on from 0800 to 2000 h). WT and neudesin KO mice were allowed free access to a normal diet (MF; 3.6 kcal/g, 12% kcal fat, source: soybean; Oriental Yeast, Tokyo, Japan) or high-fat diet (HFD; 5.24 kcal/g, 60% kcal fat, Research Diet, Inc., New Jersey, USA). The HFD-feeding experiments began at 4 weeks of age. All experiments were performed using male mice. Mice were weighed weekly and sacrificed to obtain tissues and plasma samples at 16 weeks of age. All animal studies were conducted in accordance with the International Guiding Principles for Biomedical Research Involving Animals, and approved by the Animal Research Committee of Kyoto University Graduate School of Pharmaceutical Sciences.
The unfixed livers of WT and neudesin KO mice at 16 weeks of age were frozen in optimum cutting temperature compound (Sakura Finetek, Tokyo, Japan). Frozen sections (16μm) were stained with hematoxylin and eosin.
Biochemical Analysis. Blood glucose levels were measured with a Glutest R kit (Sanwa Kagaku Kenkyusho, Nagoya, Japan). Plasma insulin levels were measured using a mouse insulin ELISA kit (Morinaga, Tokyo, Japan). Plasma TG, non-esterified free fatty acid (NEFA), and total cholesterol levels were measured using the TG E-test, NEFA C-test, and cholesterol E-test (Wako, Osaka, Japan), respectively. Plasma leptin levels were measured using a mouse leptin ELISA kit (Morinaga). Norepinephrine, epinephrine, and dopamine levels in plasma, WAT, and BAT were measured by HPLC, as described previously 29,42 . Glucose Tolerance and Insulin Tolerance Tests. The glucose tolerance and insulin tolerance tests were performed at 16 weeks of age by an intraperitoneal injection of glucose (1.5 mg/g) or human regular insulin (1 units/kg) (MP Biomedicals, Ohio, USA) after 6 h of fasting. Blood samples were taken at different time points from a tail vein. Blood glucose and plasma insulin levels were measured, as described above.
Measurement of Food Intake. Daily food intakes by WT and neudesin KO mice were measured at 16 weeks of age. The measurement of food intake was performed for 3 days and the average of 3 days was used in the analysis.

Measurement of fecal TG levels.
Fecal lipids were extracted from dried feces (50 mg) of WT and neudesin KO mice at 8 weeks of age using the Folch method and fecal triglyceride (TG) levels were measured using TG E-test.

Respiratory Gas Analysis.
Oxygen consumption by WT and neudesin KO mice was measured at 12 weeks of age with an indirect calorimetric system. Briefly, room air was pumped through an acrylic metabolic chamber, and subjected to gas analysis (model ARCO-2000; Arco System, Tokyo, Japan). The respiratory quotient, fatty acid oxidation, and carbohydrate oxidation were calculated based on O 2 consumption and CO 2 production 43 .
Measurement of Heart Rates. Heart rates were measured in conscious mice at 16 weeks of age using a tail-cuff system (Softron, Tokyo, Japan) 29 . Heart rates were measured daily for at least 3 days and the average of the 3 days was used in the analysis.

Production of Recombinant Mouse neudesin Protein.
A recombinant neudesin protein was obtained as described previously 18 . Briefly, High Five cells infected with a recombinant baculovirus containing neudesin cDNA with His 6 tag at the 3' terminus of the coding region were cultured at 27 °C for 96 hr in Grace's Insect Medium (Gibco, Rockville, MD, USA) with 10% fetal bovine serum (FBS). A recombinant neudesin protein was purified from the culture medium by affinity chromatography using Ni-NTA agarose (QIAGEN).
Cultures of PC12 Cells. PC12 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Gibco) supplemented with 10% horse serum (Gibco), 5% FBS, and 1% penicillin-streptomycin (Nacalai Tesque, Kyoto, Japan) at 37 °C. PC12 cells were seeded at 4.0 × 10 4 cells/well on 24-well plates. Medium was replaced on the next day at the beginning of the culture, and NGF (50 ng/ml; alomone labs, Jerusalem, Israel) in DMEM supplemented with 1% FBS and 1% penicillin-streptomycin was added. NGF was removed 24 hrs later and recombinant neudesin in DMEM supplemented with 1% FBS and 1% penicillin-storeptomycin was added. Cells were collected 6hrs later and total RNA was extracted. The expression levels of Th were examined by qRT-PCR as described above.
Measurement of Lipolytic Activity. The subcutaneous WAT of WT and neudesin KO mice at 16 weeks of age was homogenized in homogenizing buffer (50 mM Tris-HCl, pH8.0, 0.25 M sucrose, 1 mM EDTA, and 1% Protease Inhibitor Cocktail (Sigma-Aldrich, St. Louis, USA). Cell debris was removed by centrifugation at 1,000 g for 15 min to obtain cell extracts. The protein concentration in tissue extracts was determined using the Bio-Rad protein assay kit (Bio-Rad Laboratories, California, USA) with bovine serum albumin as the standard. The lipolytic activity of tissue extracts was determined using p-nitrophenyl laurate as the substrate 44,45 .