Vagal Hyperactivity Due to Ventromedial Hypothalamic Lesions Increases Adiponectin Production and Release

Adiponectin is an adipokine that plays an important role in preventing development of type 2 diabetes and cardiovascular diseases in patients with obesity or metabolic syndrome (1,2). Adiponectin-deficient mice show insulin resistance with glucose intolerance, despite having a body weight gain similar to wild-type mice (3), and circulating adiponectin levels decrease in parallel with reduced insulin sensitivity in obese humans and in animal models of obesity (4–6). Reduced circulating adiponectin is thought to play a central role in the development of metabolic syndrome in humans (2).

Adiponectin is a highly abundant circulating protein produced almost exclusively in adipose tissue and especially in human visceral adipose tissue (2,7). An in vitro study showed that adiponectin production and release in cultured human adipocytes was positively associated with increased adipocyte volume (8). However, adiponectin production and release in obese humans and animals are thought to be downregulated by feedback inhibition by as-yet-unidentified serum humoral components, resulting in decreased serum adiponectin, with insulin or tumor necrosis factor-α playing a possible role in this mechanism (9). In obese humans, serum adiponectin decreases, and adiponectin mRNA expression in adipose tissue and serum adiponectin are inversely associated with BMI, a measure of the degree of obesity (10). In genetic obese animals, such as ob/ob mice and Zucker obese rats, or diet-induced obese animals, serum adiponectin is also decreased by downregulation of adiponectin production and release in adipose tissue (6,11–13).

The cell size of adipocytes or the fat mass is another important factor in adiponectin production and release in obese humans and animals (8–10,14–16). Marked weight loss in obese humans increases adiponectin gene expression and also elevates serum adiponectin concentration (14). Adiponectin is induced by activation of nuclear peroxisome proliferator–activated receptors (PPARs), especially the PPAR-γ (15,16). Differentiation of preadipocytes into adipocytes is required to induce increased expression of adiponectin, and adiponectin production and release occur specifically in maturing adipocytes (9). Thus, adipocyte size or fat mass has a key role in regulation of adiponectin production and release in adipose tissue in obese humans and animals. Low-grade inflammation also has been found to play an important role in adiponectin production through a feedback inhibition process (17), accordingly a body weight reduction of >10% elevates serum adiponectin and decreases other proinflammatory markers (18).

Ventromedial hypothalamic (VMH) obesity is produced by the destruction of bilateral VMH nuclei in the hypothalamus and is used as a representative animal model of hypothalamic obesity. Among several types of hypothalamic obesity (19), VMH lesions uniquely produce derangements of the autonomic nervous system, including decreased sympathetic and markedly increased parasympathetic nervous activities (19,20). Studies have suggested VMH lesions are involved in the regulation of adipose tissue metabolism (21,22), but changes of adiponectin production and release in white adipose tissues of VMH-lesioned animals have not been investigated in detail.

In the current study, we examined adiponectin production and release, and the resultant serum adiponectin level, simultaneously with morphological changes in visceral and subcutaneous adipose tissues in VMH-lesioned animals and investigated the mechanism of increased adiponectin production and release in these animals. We also examined possible involvement of autonomic derangements in modulation of adiponectin production and release in visceral and subcutaneous adipose tissues using pharmacological modulators of the sympathetic and parasympathetic nervous systems.

Female ddY mice, a common outbred strain of mice in Japan (23), were obtained from Japan SLC (Shizuoka, Japan) at age 10 weeks, and female Sprague-Dawley rats were obtained from Charles River Japan (Kanagawa, Japan) at age 13 weeks. The animals were housed with free access to laboratory chow and water under a 12-h light-dark cycle. All procedures were performed according to the Japanese Physiological Society’s guidelines for animal care. The experimental protocol was approved by the Kiryu University Faculty of Health Care Animal Research Committee.

Bilateral VMH lesions in mice and rats were produced by electrical destruction using a Narishige stereotaxic instrument (Tokyo, Japan) (24,25). In mice, an anodal current of 1 mA was passed for 10 s through a stainless steel needle insulated with Epoxylite, except for 1 mm at the top, for formation of bilateral electrical lesions in the VMH. The coordinates from the Paxinos and Franklin Atlas for VMH lesioning were 1.6 mm posterior to the bregma anteriorly, ± 0.5 mm lateral to the midsagittal line, and 0.2 mm above the base of the skull. In rats, an anodal current of 2 mA was passed for 20 s. The coordinates from the Degroot Atlas for VMH lesioning were at the bregma anteriorly, ± 0.75 mm to the midsagittal line, and 1.0 mm above the base of the skull. Frozen coronal sections were prepared through the region of the brain containing the VMH lesions, and the lesions were verified by microscopic examination of serial brain sections. If the whole VMH was not destroyed bilaterally, the animal was excluded from the data analysis. Usually 2 mA and 20 s of current was powerful enough to destroy the whole VMH if the tip of the electrode was appropriately placed, but sometimes destroyed lesions extended to the arcuate nucleus or to the third ventricle. Such cases were also excluded from the data analysis.

Subdiaphragmatic vagotomy was conducted according to the method of Snowdon and Epstein (26). The rats were anesthetized with 2–3% isoflurane in air and placed supine on an operating table heated to ∼34°C. After the abdomen had been clipped and scrubbed with a disinfectant solution (Isodine; Meiji, Tokyo, Japan), a ventral midline incision (∼20 mm) was made from the xiphoid caudally. Sectioning of the subdiaphragmatic vagus was performed under a dissecting microscope. The stomach and esophagus were retracted through a midline abdominal incision, and the anterior and posterior subdiaphragmatic vagi were dissected free from the esophagus and sectioned at the level proximal to the hepatic branch. The abdominal wall and skin were closed separately with surgical silk.

Atropine sulfate monohydrate, a cholinergic inhibitor (5 mg/kg body weight; Wako Pure Chemical Industries, Osaka, Japan) dissolved in saline was subcutaneously administered twice a day after preparation of the sham VMH or VMH lesions. Saline or carbachol (Sigma-Aldrich, St. Louis, MO) was continuously infused into subcutaneous tissue on the back by an ALZET pump at 60 or 180 µg/kg body weight/h for 5 days. 6-Hydroxydopamine (50 mg/kg body weight; Sigma-Aldrich) was injected intraperitoneally on days 1 and 4.

In mice, the percentage of body fat was measured under pentobarbital anesthesia at 1 week and 4 weeks after the operation using dual-energy X-ray absorptiometry (PIXImus2; GE Yokogawa Medical Systems, Tokyo, Japan) (27). In rats, fat pads in the abdominal area were dissected out, and visceral and subcutaneous adipose tissues were weighed at 1 week (28). The Lee index, a marker of body fatness, was also calculated in rats (29).

At 1 week after preparation of the sham VMH- and VMH-lesioned mice and rats, parts of the dissected visceral and subcutaneous fat pads in mice and rats were placed in 10% formalin solution. Each section was stained with hematoxylin and eosin. The major axis of the adipose cell was measured at original magnification ×400 in 200 adipose cells in six to seven fields randomly selected from each section using BZ-Analyzer II image analysis software (Keyence Corp., Osaka, Japan).

Visceral and subcutaneous fat pads were prepared from mice at 1 and 4 weeks and from rats at 5 days and 1 week after the operation. Anti-rat adiponectin antibody was purchased from Cell Signaling Technology (Beverly, MA). Western blot analyses were performed as described previously (30), with some modifications.

Total RNA was isolated from adipose tissue with Isogen (Nippon Gene, Tokyo, Japan), and reverse transcription was performed using a PrimeScript RT Reagent Kit (Takara Bio Inc., Shiga, Japan). Real-time PCR was performed with a Real-Time PCR System (Applied Biosystems Japan, Tokyo, Japan) (31). Two independent experiments were performed in triplicate.

Serum immunoreactive insulin, leptin, and total serum adiponectin concentrations were measured by ELISA using a Morinaga high-sensitive insulin assay kit, a Morinaga leptin assay kit (Morinaga Institute of Biological Science, Inc., Yokohama, Japan), and a mouse/rat adiponectin ELISA kit (Otsuka Pharmaceutical Co., Tokyo, Japan), respectively. Glucose levels were measured by the glucose oxidase method.

All data are expressed as the mean ± SEM. Differences of means were analyzed by ANOVA, followed by Student t test, Dunnett multiple comparison test, or Tukey post hoc test for post hoc multiple comparisons. For the histogram of the adipocyte distribution, the adipocyte diameter was represented as a dummy variable for each 10 µm, and then a Mann-Whitney U test was applied, with these data considered to be discrete. The distribution of the adipocyte diameters was assessed by Shapiro-Wilk test, followed by a Mann-Whitney U test to evaluate the difference between groups. Significance was set at P < 0.05 for all analyses.

Further information on the methods is provided in the Supplementary Data.

VMH-lesioned mice showed a marked increase in body weight at 1 and 4 weeks after the operation compared with sham VMH-lesioned mice, and VMH-lesioned mice at 4 weeks had a higher body weight than at 1 week (Table 1). A transient increase in food intake in the VMH-lesioned mice was observed at 1 week after the operation, as we observed in our previous study (24), but daily food intake at 4 weeks did not differ significantly between the VMH- and sham VMH-lesioned mice (Table 1). Total body fat at 1 and 4 weeks was significantly higher in VMH-lesioned mice compared with sham VMH-lesioned mice, and VMH-lesioned mice had greater total body fat at 4 weeks compared with 1 week (Table 1).

Blood glucose concentrations did not differ between VMH- and sham VMH-lesioned mice at 1 and 4 weeks after the operation (Table 1), but VMH-lesioned mice showed significantly higher serum insulin at 1 and 4 weeks (Table 1). Serum concentrations of total adiponectin and leptin were also significantly higher in VMH-lesioned mice at 1 and 4 weeks, but serum insulin, total adiponectin, and leptin in VMH-lesioned mice at 4 weeks did not differ significantly from the levels in these mice at 1 week.

Morphological changes in adipocytes at 1 week after VMH-lesioning are shown in Fig. 1. The distributions of adipocyte sizes differed significantly for visceral and subcutaneous adipocytes between VMH- and sham VMH-lesioned mice (Fig. 1A), and the number of large adipocytes in VMH-lesioned mice was markedly increased in both types of adipose tissue, with a particularly marked increase in visceral adipose tissue (Fig. 1A and C). At 1 week, the average diameters of adipocytes in visceral and subcutaneous adipose tissues in VMH-lesioned mice were significantly larger by 1.43 and 1.19 times, respectively, compared with those in sham VMH-lesioned mice (Fig. 1B).

Western blot analysis showed that adiponectin production in VMH-lesioned mice at 1 week significantly increased in visceral and subcutaneous adipose tissues by 3.5 and 2.0 times, respectively, compared with those in sham VMH-lesioned mice at 1 week after the operation (Table 1). Adiponectin mRNA expression in visceral and subcutaneous adipose tissues also significantly increased by 4.6 (P < 0.02) and 2.9 times (P = 0.05), respectively, in VMH-lesioned mice compared with sham VMH-lesioned mice. Thus, VMH-lesioned mice showed elevated serum total adiponectin with increased adiponectin production and release, despite adipocyte hypertrophy in visceral and subcutaneous tissues and increased body fat.

VMH-lesioned rats at 1 week also showed markedly increased food intake and body weight gain, with a significantly increased Lee index for body fat and a significant increase of both visceral and subcutaneous fat pad weights (Table 1). VMH-lesioned rats had similar serum glucose levels to those in sham VMH-lesioned rats, but significantly higher serum insulin, total adiponectin, and leptin (Table 1). Western blot analysis showed significantly increased adiponectin production in visceral and subcutaneous adipose tissues in VMH-lesioned rats (Table 1). Moreover, adiponectin production was significantly higher in visceral adipose tissue than in subcutaneous adipose tissue of VMH-lesioned rats (Supplementary Fig. 1). Morphological changes showed hypertrophy of visceral and subcutaneous adipocytes after 1 week in VMH-lesioned rats (Fig. 2A and C). The distribution of adipocyte sizes in visceral and subcutaneous adipose tissues differed significantly between VMH- and sham VMH-lesioned rats (Fig. 2A), and the average diameters of adipocytes in VMH-lesioned rats were significantly larger in both adipose tissues by 1.32 and 1.18 times, respectively, compared with those in sham VMH-lesioned rats (Fig. 2B). Thus, the results for VMH-lesioned rats at 1 week replicated the findings of elevated serum total adiponectin in VMH-lesioned mice at 1 week.

We hypothesized that the increased parasympathetic nerve activity produced by VMH lesions stimulates adiponectin production and release, which results in elevated serum adiponectin. To test this hypothesis, the effects of atropine, an inhibitor of parasympathetic activity, or subdiaphragmatic vagotomy on these parameters were examined. VMH-lesioned rats administered saline showed marked body weight gain and increased daily food intake for 5 days after lesioning (Fig. 3A and B). These rats also had significantly lower blood glucose levels in the fed condition (Fig. 3C), significantly higher serum insulin (Fig. 3D), and significant elevation of serum total adiponectin and leptin (Fig. 3E and F). Daily, subcutaneous administration of 5 mg/kg atropine significantly and completely inhibited the increased body weight gain (Fig. 3A) and the increased daily food intake (Fig. 3B) and also completely reversed the hyperinsulinemia (Fig. 3D) and elevation of circulating total adiponectin (Fig. 3E) in VMH-lesioned rats, although the decreased blood glucose levels persisted (Fig. 3C) and hyperleptinemia was only partially reversed (Fig. 3F). Daily atropine administration also significantly and completely reversed the increased adiponectin production and release in visceral and subcutaneous adipose tissues (Fig. 4A and B), which resulted in reversal to normal levels of serum total adiponectin in VMH-lesioned rats (Fig. 3E). In contrast, atropine did not affect adiponectin production in both adipose tissues (Fig. 4A and B) or serum adiponectin (Fig. 3E) in sham VMH-lesioned rats.

Subdiaphragmatic vagotomy significantly reduced the marked increase in body weight (Fig. 5A), which inhibited further weight gain, and attenuated the increased food intake (Fig. 5B) usually observed in VMH-lesioned rats (25). Subdiaphragmatic vagotomy reversed reduced blood glucose levels to normal levels (Fig. 5C). Subdiaphragmatic vagotomy completely reversed the hyperinsulinemia (Fig. 5D) and the elevated serum total adiponectin to normal levels (Fig. 5E) but did not affect the elevated leptin levels (Fig. 5F) in VMH-lesioned rats. Subdiaphragmatic vagotomy also reversed adiponectin production in visceral (Fig. 6A) and subcutaneous (Fig. 6B) adipose tissues. In contrast, subdiaphragmatic vagotomy did not affect adiponectin production in both adipose tissues (Fig. 6A and B) or serum adiponectin (Fig. 5E) in sham VMH-lesioned rats.

To examine whether autonomic derangements (increased parasympathetic nerve activity and reduced sympathetic activity), which are similarly induced by VMH lesions, are directly involved in the increased adiponectin production and release, and resultant elevation of serum adiponectin independent of obesity, the effects on serum adiponectin of carbachol, a parasympathetic stimulator and 6-hydroxydopamine, a sympathetic blocker, were determined in normal rats. Continuous subcutaneous infusion of high-dose carbachol significantly increased serum total adiponectin concentrations, which mimicked the effects of VMH lesions in serum adiponectin in mice and rats in this study, but low-dose carbachol infusion did not affect serum total adiponectin (Fig. 7A). In contrast, chemical sympathectomy with 6-hydroxydopamine significantly decreased serum total adiponectin and failed to produce a significant increase in circulating total adiponectin (Fig. 7A). After 5 days, rats treated with carbachol or 6-hydroxydopamine did not show increased body weight (body weight changes in 5 days: saline, −0.60 ± 5.9 g; low-dose carbachol, −0.56 ± 4.3 g; high-dose carbachol, −6.64 ± 2.02 g; 6-hydroxydopamine, −15.4 ± 6.0 g; not significant). Western blot analysis showed that adiponectin in rats with high-dose carbachol infusion significantly increased in both adipose tissues (Fig. 7B and C). Moreover, adiponectin production with high-dose carbachol infusion was significantly higher in visceral adipose tissue than in subcutaneous adipose tissue (Supplementary Fig. 2). Body weight and food intake did not change significantly during the experiments (279 ± 5 to 277 ± 3 g vs. 291 ± 2 to 285 ± 3 g for body weight; 15.7 ± 0.9 to 14.8 ± 0.6 g vs. 15.6 ± 0.8 to 14.5 ± 0.8 g for food intake in rats with saline and high-dose carbachol, respectively). Thus, stimulation of parasympathetic nervous activity, but not suppression of sympathetic nervous activity, elevated serum total adiponectin independently of obesity.

It is well known that circulating adiponectin concentrations are reduced in animal models of obesity and in patients with obesity or metabolic syndrome, despite adipocyte hypertrophy or increased body fat (2–6). However, the details of adiponectin production and release have varied among studies, especially in animal models (10,28,29,32,33). This is probably due to the apparently different etiology of the three types of obese animal models: genetic, diet-induced, and hypothalamic obesity (20,34).

VMH lesion-induced hypothalamic obesity in animals is the only obesity model that shows clear derangements of autonomic nervous activities (hyperactivity of the vagus nerve and hypoactivity of sympathetic nerves) compared with genetic (20), diet-induced obesity (34), and other types of hypothalamic obesity (19). Thus, there is a possibility that this animal model has different characteristics of adiponectin production and release, with resultant change of serum adiponectin, compared with those of other obesity models.

In the current study, we first investigated whether VMH-lesioned mice and rats have abnormal adiponectin production and release in visceral and subcutaneous adipose tissues and resultant change in serum adiponectin. Surprisingly, we found that adiponectin production and release significantly increased in visceral and subcutaneous adipose tissues of these animals, resulting in significant elevation in serum adiponectin, despite the enlarged size of adipocytes in both types of adipose tissues and increased body fat. Similar findings of increased adiponectin mRNA and protein production in visceral adipose tissue were reported by Huypens and Quartier (35) in gold-thioglucose–induced VMH lesion-induced obese mice, a similar animal model of hypothalamic obesity produced by chemical destruction of the VMH. Adiponectin production and release in subcutaneous adipose tissue and the level of circulating adiponectin were not examined by Huypens and Quartier (35).

We next investigated the mechanism of increased adiponectin production and release in both types of adipose tissues and resultant elevated serum adiponectin in VMH-lesioned rats. The autonomic nervous system may be involved in regulation of adiponectin production and release in adipose tissues, and adipose tissue is innervated by sympathetic nerves (36); however, results for involvement of the sympathetic nervous system in adiponectin production and release in humans and animals are in conflict. In animals, sympathetic nervous activation by cold exposure has been shown to decrease adiponectin mRNA in adipose tissue and decrease serum adiponectin (37); to cause no change in mRNA adiponectin in adipose tissue, resulting in no change in serum adiponectin in mice (38); and to increase adiponectin mRNA in adipose tissue in mice (39). In humans, cold exposure has been reported to elevate serum adiponectin levels (40), but conversely, blockade of the sympathetic nervous system has been shown to elevate serum adiponectin in patients with hypertension (41). Involvement of parasympathetic nervous activity in adipose tissue metabolism and adiponectin production and release has not been investigated in detail, in part, due to lack of information on parasympathetic innervations of adipose tissue until these innervations were described in visceral and subcutaneous adipose tissues (42). A recent clinical study also showed that the vagus nerve locally regulates the amount of intra-abdominal fat tissue in humans, and selective vagotomy in gastrectomy results in preferential reduction of visceral fat, indicating the possible involvement of the vagus nerve in visceral fat metabolism in humans (43).

With this background, we investigated the effects of atropine, a parasympathetic and cholinergic inhibitor, on adiponectin production and release in rats 5 days after preparation of VMH lesions. Daily atropine administration significantly attenuated the increase in adiponectin production and release in the visceral and subcutaneous of adipose tissues, which eliminated the elevation of serum adiponectin in the VMH-lesioned rats but did not affect serum adiponectin level in sham-operated animals. Furthermore, subdiaphragmatic vagotomy in VMH-lesioned rats replicated the results of atropine effects on the reversal of increased adiponectin production and release and the elimination of resultant elevated serum adiponectin. Atropine administration and subdiaphragmatic vagotomy also reversed hyperphagia and hyperinsulinemia in the VMH-lesioned rats. However, hyperinsulinemia is unlikely to contribute to the observed increase of adiponectin production and release in these animals, because hyperinsulinemia inhibits adiponectin expression (44). Short-term food restriction and refeeding did not affect the serum adiponectin level, but adipose tissue adiponectin gene expression is upregulated by long-term food restriction in rats (45). Therefore, it is conceivable that hyperphagia after VMH lesioning may not upregulate adiponectin production and release. Taken together, these results indicate that the reversal of the increased adiponectin production and release in visceral and subcutaneous of adipose tissues and the elimination of resultant elevated serum adiponectin was produced by the reduction of hyperactivity of the parasympathetic nervous system by atropine or by subdiaphragmatic vagotomy, but not by the reduction of hyperinsulinemia or hyperphagia, which were also induced by atropine or subdiaphragmatic vagotomy.

On the basis of similar results, Huypens and Quartier (35) speculated that the origin of increased adiponectin production in VMH-lesioned animals is due to reduced sympathetic nervous activity produced by VMH lesions (20). However, our findings clearly show that the increased adiponectin production and release, and resultant elevated serum adiponectin, in these animals is due to increased parasympathetic nervous activity though the cholinergic neuron system. A recent study showed that obese mice treated with monosodium glutamate, which destroys the arcuate nucleus in the hypothalamus to produce another animal model of hypothalamic obesity (19), had no changes in adiponectin mRNA expression in epididymal white adipose tissue and normal levels of circulating adiponectin (46), probably due to the lack of marked vagal hyperactivity in this animal model of hypothalamic obesity (19).

Finally, we investigated whether activities of the autonomic nervous system are directly involved in these phenomena in normal rats. We found that infusion of high-dose carbachol, a parasympathetic stimulator, for 5 days elevated serum adiponectin associated with increased adiponectin production in both adipose tissues without body weight and food intake, whereas 6-hydroxydopamine, a sympathetic blocker, reduced serum adiponectin, without changes of body weight. This suggests that activation of parasympathetic nervous activity directly stimulates adiponectin production and release in adipose tissues, which results in elevated serum adiponectin independent of obesity. Reduced sympathetic activity was reported in diet-induced obesity, a relevant human obesity model, in animals (34). This may, in part, contribute to the decreased adiponectin production and release in these animals in addition to downregulation of feedback inhibition (13). This needs further investigation because the effect of the sympathetic nerve is inconsistent, as already described (37–41).

In conclusion, we clearly demonstrated that VMH lesions increase adiponectin production and release in visceral and subcutaneous adipose tissues, which results in elevated serum adiponectin through activating parasympathetic nervous activity, despite adipocyte hypertrophy in both visceral and subcutaneous adipose tissues and increased body fat in rodents. We also show possible involvement of increased cholinergic neuron system activity in the increase of adiponectin production and release. Such activation may provide a method for reversal of the reduced adiponectin production and release in patients with metabolic syndrome and obesity-related type 2 diabetes.

Funding. This work was supported in part by a Grant-in-Aid for Scientific Research in Innovative Areas (“Molecular Basis and Disorders of Control of Appetite and Fat Accumulation”) to H.S. from the Japanese Ministry of Education, Culture, Sports, Science and Technology.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. Y.S. planned and conducted the experiments, produced sham VMH- and VMH-lesioned mice, and measured serum parameters. H.S. planned and conducted the experiments and measured mRNA adiponectin in the adipose tissues. N.I. planned and conducted the experiments, produced sham VMH- and VMH-lesioned rats, measured serum parameters, and produced subdiaphragmatic vagotomized sham VMH- and VMH-lesioned rats. N.K. and T.Ku. measured adipose tissue adiponectin by Western blot analysis. A.S. conducted histological examinations of adipose tissues. H.K. conducted histological examinations of adipose tissues and data analysis. T.O. planned and conducted the experiments. S.H., H.-J.K., and A.M. conducted body fat determinations in mice and rats. S.S., M.M., and T.Ka. contributed to experimental planning and writing the manuscript. S.I. contributed to experimental planning and produced the final version of the manuscript. S.I. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.