Blog

By Jennifer A. Felsted, Cheng-Hao Chien, Dongqing Wang, Micaella Panessiti, Dominique Ameroso, Andrew Greenberg, Guoping Feng, Dong Kong, and Maribel Rios

Excerpt from the article published in Cell Reports, volume 21, issue 10, P2737-2747, December 2017, DOI: https://doi.org/10.1016/j.celrep.2017.11.048

Editor’s Highlights

• Alpha2delta-1 (α2δ-1) plays a chief role in energy and glucose balance control acting downstream of brain-derived neurotrophic factor (BDNF) in the ventromedial hypothalamus (VMH).
• BDNF signals through the tyrosine kinase B (TrkB) receptor to support neuronal survival, differentiation, and synaptic plasticity.
• BDNF is a key central regulator of energy and glucose balance within the central nervous system.
• Astrocytic BDNF signaling within the ventromedial hypothalamus modulates neuronal activity in response to changes in energy status.
• BDNF signals via the TrkB receptor to regulate synaptic transmission and plasticity in many regions of the mature brain.
• α2δ-1 is a requisite factor in the VMH for energy and glucose balance control and whether steroidogenic factor-1 (SF1)-containing neurons is a critical cellular substrate.
• These actions of α2δ-1 involve effects on the excitatory drive of SF1 neurons and not its canonical calcium channel function.
• α2δ-1 serves dual roles as a voltage-gated calcium channel subunit and as a receptor for thrombospondins (TSPs), facilitating excitatory synapse assembly. 
• α2δ-1 might act on SF1 neurons to negatively regulate serotonin production in the periphery, ultimately affecting lipid metabolism in white adipose tissue (WAT). 
• The mechanistic consequences of central α2δ-1 action is an essential step in creating avenues to tackle metabolic disorders, including diabetes and cardiovascular diseases.

Autor’s Highlights

• Glycemic and lipid but not energy balance control requires intact α2δ-1 in VMH
• α2δ-1 elevates SF1 neuronal activity through calcium channel-independent mechanisms
• α2δ-1 in VMH positively regulates sympathetic tone to metabolic tissues
• Elevated circulating levels of serotonin result from α2δ-1 depletion in VMH

Summary

The central mechanisms controlling glucose and lipid homeostasis are inadequately understood. We show that α2δ-1 is an essential regulator of glucose and lipid balance, acting in steroidogenic factor-1 (SF1) neurons of the ventromedial hypothalamus (VMH). These effects are bodyweight independent and involve regulation of SF1⁺ neuronal activity and sympathetic output to metabolic tissues. Accordingly, mice with α2δ-1 deletion in SF1 neurons exhibit glucose intolerance, altered lipolysis, and decreased cholesterol content in adipose tissue despite normal energy balance regulation. Profound reductions in the firing rate of SF1 neurons, decreased sympathetic output, and elevated circulating levels of serotonin are associated with these alterations. Normal calcium currents but reduced excitatory postsynaptic currents in mutant SF1 neurons implicate α2δ-1 in the promotion of excitatory synaptogenesis separate from its canonical role as a calcium channel subunit. Collectively, these findings identify an essential mechanism that regulates VMH neuronal activity and glycemic and lipid control and may be a target for tackling metabolic disease.

Graphical Abstract

Introduction

Disorders of energy and glucose balance control have a staggeringly high prevalence worldwide and are associated with a myriad of negative health consequences, including diabetes and cardiovascular disease. Insight into how the brain regulates these physiological processes in coordination with peripheral tissues is paramount in developing effective treatments for these afflictions. Previous investigations implicate the ventromedial hypothalamus (VMH) in the regulation of energy, glucose, and lipid balance (King, 2006, Sudo et al., 1991, Takahashi and Shimazu, 1982) and suggest that these effects are mediated in part by controlling sympathetic output in the periphery (Saito et al., 1989, Takahashi et al., 1994). However, the underlying mechanisms remain largely undefined because this region has been under-studied relative to other hypothalamic nuclei.

Our previous investigations suggested a role of alpha2delta-1 (α2δ-1) in energy and glucose balance control acting downstream of brain-derived neurotrophic factor (BDNF) in the VMH (Cordeira et al., 2014). BDNF signals through the tyrosine kinase B (TrkB) receptor to support neuronal survival, differentiation, and synaptic plasticity (Reichardt, 2006). Notably, BDNF is a key central regulator of energy and glucose homeostasis, and the VMH is a critical target for these actions. Accordingly, diminished BDNF signaling in humans and rodents has been associated with overeating, severe obesity, hyperglycemia, hyperinsulinemia, and hyperlipidemia (Gray et al., 2006, Gray et al., 2007, Han et al., 2008, Liao et al., 2012, Lyons et al., 1999, Rios et al., 2001, Unger et al., 2007, Wang et al., 2007, Xu et al., 2003) α2δ-1, for its part, facilitates cell surface trafficking of calcium channels, mediating calcium currents and neurotransmitter release (Davies et al., 2007). Additionally, α2δ-1 was identified as a receptor for thrombospondins, promoting excitatory synapse assembly via calcium channel-independent mechanisms (Eroglu et al., 2009). However, little is known regarding the relevance of this non-canonical role of α2δ-1, influencing whole-animal physiology.

Here we examined whether α2δ-1 is a requisite factor in the VMH for energy and glucose balance control and whether steroidogenic factor-1 (SF1)-containing neurons are a critical cellular substrate for these actions. SF1 neurons, which are exclusive to this hypothalamic region, are primarily glutamatergic and have established roles in energy and glucose homeostasis regulation (Dhillon et al., 2006, Kim et al., 2012, Klöckener et al., 2011, Tran et al., 2003, Zhang et al., 2008). Despite evidence of these important roles, the mechanisms controlling the activity of these cells are largely unknown. We show that α2δ-1 plays an essential role in regulating the excitability and firing activity of SF1 neurons. Accordingly, its deletion in this cell population leads to neuronal hypoactivity, reduced sympathetic tone to white adipose tissue (WAT) and skeletal muscle, and deficits in glycemic and lipid control. These findings help explain previous reports of metabolic disorders in humans prescribed gabapentinoid drugs that inhibit α2δ-1 function (Gee et al., 1996) to combat neuropathic pain and seizure disorders (DeToledo et al., 1997, Hoppe et al., 2008). Importantly, these results inform a central mechanism controlling glucose and lipid homeostasis.

Results

 α2δ-1 in SF1 Neurons Is Not Required for the Regulation of Energy Balance

α2δ-1 is highly expressed in the VMH, particularly in the dorsomedial region, which contains a high density of SF1⁺neurons, a cell population that, in the brain, is exclusive to the VMH (Kurrasch et al., 2007, Taylor and Garrido, 2008). Prior reports showed that α2δ-1 promotes excitatory synapse assembly in a calcium-independent manner in other brain regions (Eroglu et al., 2009). Therefore, we hypothesized that α2δ-1 facilitates energy and glucose balance control by increasing the excitatory drive of SF1 neurons. Immunolabeling studies in reporter mice with SF1-cre-dependent expression of GFP revealed extensive co-localization of α2δ-1 and GFP in the VMH, confirming that α2δ-1 is expressed in SF1 neurons (Figure 1A). SF1 neurons lacking α2δ-1 and cells expressing α2δ-1 but not SF1 were also observed in the VMH (Figure 1A; Figure S1). Accordingly, confocal image analysis revealed that, although 67.8% ± 13.6% of SF1 neurons contain α2δ-1, 77.3% ± 3.0% of α2δ-1⁺ cells in the VMH express SF1.

To directly test the role of α2δ-1 in SF1 neurons, we examined the effect of selectively deleting it in this cell population. Toward this aim, we crossed mice containing loxP sites flanking exon 6 of the α2δ-1 gene with mice expressing cre recombinase under the direction of the SF1 promoter (SF1-Cre mice) (Dhillon et al., 2006) to generate mutants (α2δ-1^{2L/2L:SF1-Cre}) and controls (α2δ-1^2L/2L) (Figure 1B). Figure 1C shows that the α2δ-1 protein content in the mutant VMH was significantly decreased and only 53% that of controls. Immunolabeling studies revealed significant α2δ-1 depletion primarily in the dorsomedial VMH, which contains a high density of SF1 neurons (Figure 1D). Co-immunolabeling with anti-α2δ-1 and anti-SF1 showed that cytoplasmic expression of α2δ-1 exhibited by control SF1 neurons was reduced in α2δ-1^{2L/2L:SF1-Cre} mutants (Figure S2). Moreover, α2δ-1 expression was observed in processes in close proximity to SF1 neurons in mutant VMH. Finally, measurements of the VMH area and the density of SF1 neurons in VMH sections stained with DAPI or immunolabeled with anti-SF1 showed no significant differences in α2δ-1^{2L/2L:SF1-Cre} mice compared with controls (Figures S3A and S3B).
Because SF1 is also expressed in the adrenal glands (Luo et al., 1995a, Luo et al., 1995b), cre-mediated recombination and deletion of α2δ-1 is expected in this tissue in α2δ-1^{2L/2L:SF1-Cre} mice. However, we found that α2δ-1 is not expressed in wild-type adrenal glands, and, thus, its depletion in this region is not a confounding factor (Figure S4A). Accordingly, the serum levels of the adrenal hormone epinephrine were normal in α2δ-1^{2L/2L:SF1-Cre} mice compared with controls, further attesting to normal adrenal function (Figure S4B).
Next we examined the effect of α2δ-1 depletion on energy balance regulation. α2δ-1^{2L/2L:SF1-Cre} mutant mice exhibited normal food intake compared with α2δ-1^2L/2L mice (Figure 2A). Additionally, no significant differences in consumption of chow diet (CD) during re-feeding following a prolonged fast were observed between mutants and controls (Figure S5). Consistent with the observed normal feeding behavior, the body weights of α2δ-1^{2L/2L:SF1-Cre} mutant mice were similar to those of controls (Figure 2B). Furthermore, when challenged with a high-fat diet (HFD, 5.56 kcal/g) beginning at 8 weeks of age, mutant mice exhibited similar levels of food intake and body weight as control mice (Figures 2A and 2B).

Body composition measurements in mice fed a CD showed a similar content of lean and fat mass in mutants compared with controls (Figure 2C). Similarly, energy expenditure and locomotor activity were not affected in α2δ-1^{2L/2L:SF1-Cre}mutants (Figure 2D and E). In aggregate, the findings indicate that α2δ-1 expression in SF1 neurons is not required for energy balance regulation.

 α2δ-1 Deletion in SF1 Neurons Elicits Deficits in Glucose Homeostasis

The VMH is a key glucose-sensing region of the brain facilitating glycemic control (Kang et al., 2006, Song et al., 2001). Therefore, we investigated whether α2δ-1 depletion in SF1 neurons affected glucose homeostasis. Baseline fasting levels of circulating glucose or insulin were not significantly different between α2δ-1^{2L/2L:SF1-Cre} and α2δ-1^2L/2L mice on a CD (Figures 3A and 3B ). However, α2δ-1^{2L/2L:SF1-Cre} mutants exhibited significantly impaired responses to a glucose challenge compared with controls (Figure 3C and 3D).

Consistent with the findings in the glucose tolerance test, we found that α2δ-1^{2L/2L:SF1-Cre} mice exhibited insulin resistance, as indicated by a blunted decrease in blood glucose following an insulin challenge relative to α2δ-1^2L/2L mice (Figure 3E). Accordingly, there was a significant interaction of time and genotype (p = 0.003) and a significant increase in glucose levels in mutants compared with controls 15 min following insulin administration (Figure 3E). However, counterregulatory responses to insulin-induced hypoglycemia appeared to be intact in α2δ-1^{2L/2L:SF1-Cre} mutant mice (Figures 3E and 3F). Our findings indicate an essential role of α2δ-1 in SF1 neurons regulating glucose homeostasis.

 α2δ-1^{2L/2L:SF1-Cre} Mutants Exhibit Alterations in Lipid Homeostasis

The VMH is known to influence peripheral lipid metabolism, but the underlying mechanisms are poorly understood (Kumon et al., 1976, Ruffin and Nicolaidis, 1999). We investigated whether α2δ-1 activity in SF1 neurons might play a part. For this, we measured triacylglycerol (TAG) and total cholesterol content in the liver, WAT, and serum of α2δ-1^{2L/2L:SF1-Cre}mutants and α2δ-1^2L/2L controls fed a CD or following a 24-hr fast. Serum, liver, and WAT levels of TAG were not significantly different between genotypes under fed or fasted conditions (Figures 4A–4C). Serum and liver levels of cholesterol were also normal (Figures 4D and 4F). However, there were significant effects of genotype (p < 0.0003) and feeding status (p < 0.01) on total cholesterol levels in WAT, with fed and fasted mutants exhibiting significant decreases compared with α2δ-1^2L/2Lcontrol mice (Figure 4E). Moreover, examination of fatty acid profiles revealed modest but significant alterations in liver but not in WAT tissue in both fed and fasted α2δ-1^{2L/2L:SF1-Cre}mutant mice, suggesting alterations in hepatic lipid homeostasis (Figures 4G and 4H and data not shown).

We investigated potential alterations in lipolysis within WAT. Lipolysis is normally enhanced during conditions of negative energy balance to hydrolyze triglycerides and elevate free fatty acids and glycerol release from adipose tissue, providing the animal with substrates for oxidative metabolism and ATP production in peripheral tissues. No differences in serum non-esterified free fatty acid (NEFA) concentrations were observed under fed or fasted conditions or in serum glycerol levels under fasted conditions (Figures 5A and 5B ). However, there was a significant interaction between feeding status and genotype (p = 0.02) and a 315% elevation in circulating levels of glycerol in fed mutant mice compared with fed controls, indicating that lipolysis suppression in the fed state was impaired in mutants (Figure 5B). Consistent with increased levels of lipolysis, α2δ-1^{2L/2L:SF1-Cre} mutants exhibited a modest but significant reduction in adipocyte size (Figures 5C and 5D). To ascertain mechanisms underlying increased lipolysis in fed α2δ-1^{2L/2L:SF1-Cre} mutants, we measured levels of total, activated (pHSL_S660), and inhibited (pHSL_S565) levels of the lipolytic enzyme hormone sensitive lipase (HSL) (Haemmerle et al., 2002) in control and mutant WAT under fed and fasted conditions. Consistent with elevated levels of serum glycerol in fed but not in fasted mutants, there was a trend (p = 0.06) toward a significant increase in pHSL_S660 in WAT in the fed state and normal levels in the fasted state in mutants compared with controls (Figures 5E and 5F and data not shown). The levels of HSL and pHSL_S565 were normal in both sets of mutants.

To examine potential alterations in adipokine secretion, serum concentrations of leptin and adiponectin were measured. No significant differences were observed between α2δ-1^{2L/2L:SF1-Cre}and α2δ-1^2L/2L mice (Figure S6). In summary, the data indicate that α2δ-1 in SF1 neurons is a critical player in the regulation of lipid balance, influencing activity of lipolytic enzymes in WAT.

 α2δ-1 Is a Critical Regulator of SF1 Neuronal Activity

We measured the effect of depleting α2δ-1 on the activity of SF1 neurons using whole-cell recordings to inform central mechanisms leading to dysregulated glucose and lipid balance in α2δ-1^{2L/2L:SF1-Cre} mice. Control and mutant mice were generated with selective, cre-dependent expression of red fluorescent tdTomato in SF1 neurons to facilitate the identification of these cells in the VMH. We observed a significant 81% reduction in the firing rate of mutant SF1 neurons compared with controls (Figures 6A and 6B ). Moreover, mutants had a significantly hyperpolarized membrane potential compared with control mice (Figure 6B).

We asked whether alterations in calcium currents might be responsible for the diminished firing rate of SF1 neurons in α2δ-1^{2L/2L:SF1-Cre} mice, considering that the canonical role of α2δ-1 is to serve as an auxiliary high-voltage calcium channel subunit (Davies et al., 2007). We found that calcium currents in mutant SF1 neurons were normal (Figure 6C). Average peak amplitude, voltage dependence, and current kinetics were similar in both groups of animals. Another role of α2δ-1 is to mediate excitatory synaptogenesis in a calcium channel-independent manner through interactions with thrombospondins (Eroglu et al., 2009). Thus, we examined whether mutant SF1 neurons might have alterations in their excitatory drive. A significant decrease in frequency but not amplitude of spontaneous excitatory postsynaptic currents (sEPSCs) was observed in mutant cells (p = 0.03) (Figures 6D and 6E). Measurements performed in the presence of tetrodotoxin (TTX) to evaluate action potential-independent excitatory events revealed a significant decrease in frequency and normal amplitude of miniature excitatory post synaptic currents (mEPSCs) in SF1 neurons in α2δ-1^{2L/2L:SF1-Cre} mice (Figure 6F). Interestingly, the frequency data for sEPSCs and mEPSCs suggest that subpopulations of mutant SF1 neurons might be differentially affected. Accordingly, the VMH of α2δ-1^{2L/2L:SF1-Cre} mutants contained a subgroup of SF1 neurons exhibiting a frequency of sEPSCs and mEPSCs below 2 Hz that was absent in control VMH (Figures 6G and 6H). Notably, this putative mutant subpopulation was significantly different from another subgroup of mutant cells with a frequency of sEPSCs and mEPSCs above 2 Hz (Figures 6G and 6H) that exhibited levels of activity similar to those of wild-type cells. Finally, we found that the frequency and amplitude of spontaneous inhibitory post synaptic currents (sIPSCs) and miniature inhibitory post synaptic currents (mIPSCs) in SF1 neurons were similar in control and mutant mice (Figures 6I and 6J), indicating that α2δ-1 is not required for the regulation of inhibitory transmission in SF1 neurons.

In aggregate, these data indicate that α2δ-1 plays a critical role in regulating the activity of SF1 neurons through calcium channel-independent mechanisms that increase the excitatory drive onto these cells.

 α2δ-1 Depletion in SF1 Neurons Reduces the Sympathetic Tone and Increases Circulating Levels of Serotonin

The VMH can affect peripheral glucose and lipid metabolism through the regulation of sympathetic output to metabolic organs (Saito et al., 1989, Satoh et al., 1999, Shimazu et al., 1991). Thus, we hypothesized that hypoactivity of SF1 neurons in α2δ-1^{2L/2L:SF1-Cre} mutants might result in decreased sympathetic output, triggering glucose intolerance and disrupted lipid homeostasis. To test this, we measured norepinephrine (NE) concentrations in the liver, WAT, brown adipose tissue (BAT), skeletal muscle, and serum of α2δ-1^{2L/2L:SF1-Cre} mutant and α2δ-1^2L/2L control mice. Whereas NE levels in WAT were depleted by 45% in mutants compared with controls, serum levels were only 27% of those in controls (Figure 7A). No differences in NE content were observed in the liver, BAT, or skeletal muscle. Additionally, levels of the NE metabolite dihydroxylphenylglycol (DHPG), which is produced following neuronal reuptake of released NE and closely reflects sympathetic activity (Izzo et al., 1985), was significantly decreased in skeletal muscle of α2δ-1^{2L/2L:SF1-Cre} mutant mice relative to controls (Figure 7B).

The elevated levels of lipolysis observed in α2δ-1^{2L/2L:SF1-Cre}mutants (Figures 5B, 5E, and 5F) were counterintuitive considering the significant reduction in sympathetic tone to WAT they exhibit and that NE promotes this process. Previous investigations demonstrated that serotonin (5-HT), another monoamine produced in the periphery by enterochromaffin cells, robustly induces lipolysis via activation of 5-HT_2B receptors in WAT (Sumara et al., 2012). Thus, we asked whether alterations in peripheral 5-HT might contribute to abnormal lipid metabolism and glycemic control in the mutants. Measurements showed a significant 2.3-fold elevation in serum 5-HT and normal levels of the metabolite 5-hydroxyindoleacetic acid (5-HIAA) in α2δ-1^{2L/2L:SF1-Cre} mutants compared with α2δ-1^2L/2L controls (Figure 7C). These results indicate that α2δ-1 depletion in SF1 neurons leads to blunted sympathetic output and severely increased levels of 5-HT in the periphery, suggesting pathological mechanisms leading to observed alterations in glucose homeostasis and lipid metabolism.

Discussion

Here we demonstrate an essential role of α2δ-1 in regulating the activity of SF1 neurons in the VMH and sympathetic output to metabolic tissues. These studies illustrate a weight-independent mechanism acting in the brain to regulate glycemic control and lipid metabolism. Notably, these actions of α2δ-1 involve effects on the excitatory drive of SF1 neurons and not its canonical calcium channel function. Furthermore, these data reveal a candidate pathological mechanism leading to alterations in glycemic control in lean individuals and patients treated with gabapentinoid drugs, which inhibit α2δ-1 (DeToledo et al., 1997, Gee et al., 1996, Hoppe et al., 2008).

A comprehensive analysis of α2δ-1^{2L/2L:SF1-Cre} mutant mice indicated that α2δ-1 is not required for the regulation of energy balance. This observation was surprising considering our previous finding that rescuing an α2δ-1 deficit broadly in the VMH of adult mice with central BDNF depletion mitigated their hyperphagic behavior and obesity (Cordeira et al., 2014). There are several potential explanations for this discrepancy. Notably, the VMH is comprised of a heterogeneous population of neurons (McClellan et al., 2006), and our immunolabeling studies indicate that α2δ-1 is not exclusively localized to SF1 neurons in this region. Therefore, it is possible that cells in the VMH lacking SF1 are targets of metabolic actions of α2δ-1 influencing energy balance control. It is also possible that deleting α2δ-1 in SF1 neurons during early development triggers compensatory mechanisms that are absent in the mature brain and that partially mitigate the ensuing deleterious effects. Alternatively, the prior observed effects of α2δ-1 gene delivery to the VMH of adult BDNF mutants on energy balance may represent a gain of function.

Despite exhibiting normal energy balance regulation, α2δ-1^{2L/2L:SF1-Cre} mutants are glucose-intolerant, indicating an essential and body weight-independent role of α2δ-1 facilitating glycemic control. Reduced peripheral levels of NE and altered responses during the insulin tolerance test suggest that reduced sympathetic tone and sensitivity to insulin in metabolic organs contribute to this deficit. Notably, reduced SF1 neuronal activity in mutant mice is a plausible central mechanism driving these pathologies. In support, activating SF1 neurons in mice via designer receptors exclusively activated by designer drugs (DREADD) technology elevated insulin sensitivity and glucose uptake in the periphery (Coutinho et al., 2017). Furthermore, the VMH regulates sympathetic outflow, which, for its part, facilitates glucose uptake by skeletal muscle, WAT, and BAT, and some of these effects are insulin-independent (Kang et al., 2004, Song et al., 2001, Minokoshi et al., 1999, Shiuchi et al., 2009, Tiniakos et al., 1996, Carreño and Seelaender, 2004, Bruinstroop et al., 2012, Abe et al., 1993, Liu et al., 1996). Importantly, electrical stimulation of the VMH increased glucose uptake in skeletal muscle and brown adipose tissue in rats, and this effect was abolished by sympathetic denervation, implicating changes in sympathetic tone (Shimazu et al., 1991, Sudo et al., 1991). Considering these findings, SF1 neuronal hypoactivity and concomitant decreases in sympathetic tone in α2δ-1^{2L/2L:SF1-Cre} mutants are expected to contribute to abnormal glycemic control.

Reduced SF1 neuronal activity in α2δ-1^{2L/2L:SF1-Cre} mutants did not appear to compromise counterregulatory responses to hypoglycemia. In contrast, a previous study indicated that chemogenic silencing of SF1 neurons impaired restorative responses to glucoprivation mediated by activation of cholecystokinin (CCK)⁺ cells in the parabrachial nucleus (Garfield et al., 2014). These findings suggest functional diversity in SF1 neurons and that α2δ-1⁺ and CCK-responsive SF1⁺ cells in the VMH might represent distinct subpopulations. Further evidence of heterogeneity in SF1⁺ neurons comes from a recent report indicating that the VMH contains both leptin-activated and leptin-inhibited SF1 neurons (Sohn et al., 2016). These findings are reminiscent of the functional diversity reported in proopiomelanocortin (POMC) neurons of the arcuate nucleus (Sohn et al., 2011, Williams et al., 2010) and illustrate the complexity of neural circuits influencing metabolic function. In support of this idea, our electrophysiology data suggest that the VMH of in α2δ-1^{2L/2L:SF1-Cre} mutants contains two putative subpopulations of SF1 neurons with reduced or similar levels of activity compared with control cells.

Dysregulated glycemic control in α2δ-1^{2L/2L:SF1-Cre} mutants was accompanied by decreased cholesterol content in WAT and increased lipolysis in the fed state, indicative of aberrant lipid homeostasis. Previous investigations indicate that alterations in cholesterol levels greatly affect adipocyte metabolic activity, which can ultimately have deleterious effects on insulin sensitivity (Le Lay et al., 2001). The failure of mutants to suppress lipolysis in the fed state was surprising because NE is known to stimulate this process, and mutants exhibited decreased sympathetic tone in WAT. However, these seemly discrepant findings could be explained by the 2.3-fold elevation in circulating 5-HT observed in mutants. In support, mice with selective gut deletion of tryptophan hydroxylase, the rate-limiting enzyme for 5-HT synthesis, exhibited both reduced peripheral 5-HT and impaired lipolysis while having normal body weights, food intake, and circulating triglyceride levels (Sumara et al., 2012). Notably, wild-type mice implanted with 5-HT-releasing pellets to increase the levels of peripheral 5-HT by 2.5-fold exhibited elevated levels of serum glycerol in the fed state and of the lipolytic enzyme pHSL_S660 in WAT, reminiscent of α2δ-1^{2L/2L:SF1-Cre} mutants. These effects of 5-HT are mediated by direct action on 5-HT_2B receptors on adipocytes. Sumara et al., 2012 also found that peripheral serotonin promotes hepatic gluconeogenesis and prevents glucose uptake by hepatocytes. Therefore, elevated circulating levels of serotonin evident in α2δ-1^{2L/2L:SF1-Cre} mutants are expected to also contribute to alterations in glycemic control. It remains unclear whether elevated serum 5-HT in mutants is a primary effect of deleting α2δ-1 in the VMH or a secondary consequence.

The collective data indicate that α2δ-1 is a critical regulator of SF1 neuronal activity. α2δ-1 serves dual roles as a voltage-gated calcium channel subunit and as a receptor for thrombospondins (TSPs), facilitating excitatory synapse assembly (Davies et al., 2007, Eroglu et al., 2009). The calcium currents of SF1⁺ neurons in α2δ-1^{2L/2L:SF1-Cre} mice were indistinguishable from those of controls. In contrast, the frequency of sEPSCs and mEPSCs was significantly reduced in mutant SF1⁺ neurons, indicating that the latter is the mechanism at play. In agreement, cortical neurons in transgenic mice overexpressing α₂δ-1 exhibited increased density of excitatory synapses and elevated frequency of miniature EPSCs compared with wild-type mice (Eroglu et al., 2009). Conversely, the α₂δ-1 inhibitor gabapentin blocked the positive effects of TSP/α₂δ-1 on excitatory synaptogenesis, whereas calcium channel blockers had no effect, further attesting to the calcium-independent actions of α₂δ-1. Finally, Eroglu et al., 2009showed that α2δ-1 does not influence inhibitory neurotransmission, consistent with our observation that sIPSCs and mIPSCs were normal in α2δ-1^{2L/2L:SF1-Cre} mutants. Considering the collective data, it is possible that α2δ-1 mediates contact of excitatory presynaptic terminals originating within or outside the VMH onto SF1 neurons to increase their activity.

In summary, we demonstrate a requisite role of α2δ-1 regulating SF1 neuronal activity and metabolic function. We propose that α2δ-1 action through non-canonical mechanisms increases the excitatory drive and activity of SF1 neurons, mediating increased sympathetic output to metabolic organs to achieve glucose homeostasis. Additionally, α2δ-1 might act on SF1 neurons to negatively regulate serotonin production in the periphery, ultimately affecting lipid metabolism in WAT. The collective findings inform pathological mechanisms underlying metabolic disturbances in individuals administered the anti-epileptic and anti-nociceptive drugs gabapentin and pregabalin, which bind and inhibit α₂δ-1 (DeToledo et al., 1997, Hoppe et al., 2008). They are interesting, considering a recent human study showing that an individual with a de novo chromosomal truncation encompassing CACNA2D1, the gene encoding α₂δ-1, exhibited metabolic alterations, including hyperinsulism (Vergult et al., 2014). Because this truncation also involved deletion of the metabolic gene CD36, the authors speculated that metabolic alterations were elicited by that deficit. However, the possibility of α₂δ-1 involvement cannot be ruled out. Therefore, further defining the mechanistic consequences of central α2δ-1 action is an essential step in creating avenues to tackle metabolic disorders, including diabetes.