Central α-Klotho Suppresses NPY/AgRP Neuron Activity and Regulates Metabolism in Mice

α-Klotho, a well-documented antiaging protein primarily produced in the kidney and choroid plexus (1,2), has recently been observed to have therapeutic potential in rodent models of metabolic disease (3–6). Studies show α-klotho promotes lipid oxidation, protects pancreatic β-cells from oxidative damage, increases energy expenditure, and facilitates insulin release (3–7). Furthermore, circulating α-klotho concentrations are decreased in patients with obesity and diabetes (8), suggesting a possible direct role in the pathophysiology of metabolic disorders. Notably, studies have primarily investigated peripheral α-klotho, which neglects the central function of α-klotho due to its impermeability to the blood-brain barrier (9). The few studies investigating centrally circulating α-klotho demonstrate that α-klotho has antioxidative and anti-inflammatory properties (10), is involved in myelination (11), and can be therapeutic in models of hypertension (12). However, the role of central α-klotho in the regulation of metabolism remains unexplored.

Neuropeptide Y/agouti-related peptide (NPY/AgRP)-expressing neurons are located within the arcuate nucleus (ARC) of the hypothalamus and are critical to homeostatic regulation of metabolism. NPY/AgRP neurons sense nutritional changes in the cerebrospinal fluid (CSF) to regulate feeding behavior (13), energy expenditure (14), and glucose metabolism (15–17). However, disordered overactivity of these neurons results in phenotypes resembling diabetes and obesity (13,15). Some circulating factors, such as leptin and insulin, also modulate NPY/AgRP neurons (18,19), but in metabolic disease states, signaling of these hormones is disrupted. Therefore, identification of novel regulators of this neuron population could facilitate the development of therapeutic tools for the prevention and treatment of metabolic disease.

Recent studies have identified several fibroblast growth factor (FGF) hormones that activate FGF receptor (FGFR)–phosphatidylinositol 3 kinase (PI3K) signaling to elicit antidiabetic effects and regulate NPY/AgRP neurons (20–25). Interestingly, α-klotho serves as a critical scaffolding protein to the FGF23-FGFR complex to promote FGFR activity (26,27). The current study investigates the novel role of central α-klotho in the regulation of NPY/AgRP neurons and whole-body metabolism via an FGFR/PI3K mechanism.

Cell culture experiments were performed on immortal hypothalamic GT1-7 cells cultured in high-glucose (4.5 mg/dL) DMEM, 10% FBS, and 1% penicillin-streptomycin. Cells were treated with 3.65 mmol/L α-klotho (R&D Systems) (10,11,28), treated with 100 ng/mL FGF23 (R&D Systems) (28), and/or pretreated with 10 nmol/L FGFR1 antagonist PD173074 (Fisher Scientific) (29) or 50 nmol/L PI3K inhibitor wortmannin (Fisher Scientific) (25). All experiments used vehicle-treated cells as controls.

C57BL/6 and B6.Tg(NPY-hrGFP)1Lowl/J (NPY-GFP reporter) mice were cared for in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and experimental protocols were approved by Institutional Animal Care and Use Committees of East Carolina University. Mice were housed at 20–22°C with a 12-h light-dark cycle.

The 6-week-old male C57BL/6 mice were given ad libitum access to a high-fat diet with a kilocalorie composition of 58%, 25%, and 17% of fat, carbohydrate, and protein, respectively, for 10 weeks (D12331; Research Diets, New Brunswick, NJ) before undergoing intracerebroventricular (ICV) cannulation.

Prior to the procedure, mice were given oral analgesic meloxicam and anesthetized with intraperitoneal (i.p.) injection of ketamine and xylazine. Mice were placed on a stereotaxic device, and a midline incision was made on the head. A hole was drilled (1.0 mm lateral, −0.5 mm posterior, 2.5 mm deep to the bregma), and a cannula was placed into the lateral ventricle (Supplementary Fig. 1). Another hole was drilled, and a screw was placed approximately at the ipsilateral lambdoid structure to aid in supporting the cannula in the skull with 3M carboxylate dental cement. Mice recovered for 14 days before immunohistochemical experiments and 7 days before all other experiments. All ICV treatments were administered via Hamilton syringe as 2.0 μL between 6:30 and 7:30 p.m.

In high-fat diet–induced obesity (DIO) mice, central administration of either 2.0 μg recombinant α-klotho (R&D Systems) alone, 25.0 μg PD173074 alone, 25.0 μg PD173074 10 min before 2.0 μg α-klotho (20), or vehicle was performed daily. On day 7, glucose tolerance tests (GTTs) or insulin tolerance tests (ITTs) were performed, and on day 12, a body composition analysis was performed using EchoMRI (Echo Medical Systems, Houston, TX). Mice were then euthanized, and tissues were collected. Food intake and body weight data were analyzed from the first 7 days to prevent confounding effects from additional assays.

In DIO mice, ICV administration of either 2.0 μg recombinant α-klotho (R&D Systems), 10 ng wortmannin, 10 ng wortmannin 1 h before 2.0 μg α-klotho (30), or vehicle was performed the night before a GTT or ITT. Food intake was measured for 4 days after the injection.

The 8- to 9-week-old chow-fed male mice underwent ICV cannulation before receiving 3 days i.p. injection of 100 mg/kg streptozotocin (STZ). The dose was determined by pilot studies yielding consistently elevated fed glucose levels between 250 and 600 mg/dL (Supplementary Fig. 2A). At 7 days after STZ injection, mice received 7 days ICV treatment with either 2.0 μg recombinant α-klotho (n = 10) or vehicle (n = 9). Food intake, body weight, and fed glucose levels were monitored daily. To identify the effects of α-klotho treatment on glucose levels independent from food intake, a pair-fed experiment was performed using the same protocols, and on day 7, fasting glucose levels were measured (n = 8–9/group).

The 9-week-old male, chow-fed mice were ICV treated with 1.0 μg anti–α-klotho (ab-α-klotho) (R&D Systems) or vehicle (n = 8/group). Treatments were performed on the evening of day 1 and morning of day 2, while subsequent treatments were performed between 6:30 and 7:30 p.m. for 7 days. GTTs were performed on day 3, and on day 7, mice were euthanized, followed by tissue collection for further assays.

Food intake was measured daily by weighing food (8–9 g) and subtracting from the total food. Bedding was inspected thoroughly for residual bits of food, which were included in measurements. On day 4 in one cohort of DIO mice, food was removed from cages during the light phase and replenished at the beginning of the dark phase (7:30 p.m.). Food intake was measured at 0.5, 1, 1.5, 2, 3, 4, 8, 14, and 24 h after food reintroduction.

For GTTs, 20% glucose solution (1.0 g/kg body weight) was i.p. injected after an overnight fast, and for ITTs, 0.6 units/kg insulin were i.p. injected after a 4-h fast. Tail blood samples were collected 15, 30, 60, 90, and 120 min after injections for analysis using a glucose meter (ReliOn Prime Blood Glucose Monitoring System; ARKRAY Inc., Kyoto, Japan). Serum was isolated from clotted blood spun at 4°C and 2,000g for 30 min. Insulin levels were quantified using an insulin ELISA kit (Crystal Chem).

On day 12 of ICV treatment, a weight-matched cohort of DIO mice was i.p. injected with 10 units/kg insulin or saline. At 7 min after the injection, hypothalamus, epididymal adipose tissue (eWAT), liver, and hindlimb skeletal muscle were flash frozen for future Western blot analysis.

For immunofluorescent analysis, mice were intracardially perfused with PBS followed by 10% formalin before immunohistochemistry was performed as described previously (31). Briefly, brains were sliced into 20-μm coronal sections using a freezing microtome (VT1000 S; Leica, Wetzlar, Germany) and incubated overnight in antibody to phosphorylated ERK (1:500; Cell Signaling Technology, Danvers, MA) or cFOS (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), followed by incubation with Alexa-fluorophore secondary antibody for 2 h. Stains were photographed using an optical microscope (DM6000; Leica), followed by blind analysis using ImageJ. At least three anatomically matched images per mouse were quantified.

Western blot was performed as described previously (31). Briefly, equal protein samples were loaded into a 4–20% HCL gel, transferred to a nitrocellulose membrane, and incubated overnight in 1:500–1:1,000 antibody dilutions in 5% milk with Tris-buffered saline with Tween for phosphorylated (p)AT^ser473, total AKT, pFOXO1^ser256, total FOXO1, pERK^{thr202/thr204} (Cell Signaling Technology), pIR^tyr972 (Invitrogen), and total ERK (Santa Cruz). Image J software was used to quantify mean intensity of equal-area sections representing each sample.

Cell and tissue RNA was extracted by Trizol (Thermo Fisher Scientific, Waltham, MA). The expressions of specific mRNA were analyzed using quantitative real-time PCR (RT-qPCR) (Power SYBR Green PCR Master Mix; Applied Biosystems, Foster City, CA). Reactions were performed in triplicate for each sample, while GAPDH was used as a reference gene for normalization.

We conducted cell-attached voltage clamp recordings of NPY/AgRP neurons as described previously (32,33). Briefly, mice were deeply anesthetized by isoflurane followed by intracardial perfusion with chilled N-methyl-d-glucamine solution and sliced into 200- to 300-μm sections. Slices recovered for an hour in HEPES recovery solution, and recordings were conducted in a normal artificial cerebrospinal fluid (aCSF) bath. For whole cell recordings, gigaohm seals were obtained, and the cells were broken into using negative pressure. Data were sampled at 10 kHz. Current clamp recordings were stabilized for repeated firing under the baseline condition. Equivalent length periods (0.5–5 min) were set within each recording during perfusion of aCSF or α-klotho (3.65 mmol/L). Firing rate (Hz) was calculated by dividing the number of events by the number of seconds. Bath application of terodotoxin (TTX) (1.0 μmol/L) prior to α-klotho treatment was used to determine action potential–independent effects on membrane potential. Voltage clamp whole cell recordings were conducted at a −70 mV holding potential and a high KCl intracellular solution (130 mmol/L KCl, 5 mmol/L CaCl₂, 10 mmol/L EGTA, 10 mmol/L HEPES, 2 mmol/L magnesium adenosine triphosphate, 0.5 mmol/L sodium guanosine triphosphate, and 5 mmol/L phosphocreatine) was used. For voltage clamp recordings of miniature inhibitory postsynaptic currents (mIPSCs), glutamatergic blockade was induced using NMDA receptor blocker AP5 (50 μmol/L) and AMPA receptor blocker cyanquixaline (10 nmol/L), followed by α-klotho administration, and then currents were abolished with picrotoxin (100 μmol/L) (34).

Unpaired t tests were used for in vivo mouse experiments to compare differences between groups in food intake, body weight, and body composition. To compare pre-post within-group changes over the course of the experiment, paired t tests were performed. To compare differences at GTT or ITT time points, two-way ANOVA with repeated measures for time and Bonferroni corrections for multiple comparisons were used. Unpaired t tests or one-way ANOVA with Tukey correction for multiple comparisons were used in cell culture experiments when appropriate. Paired t tests or repeated-measures ANOVA with Tukey correction for multiple comparisons was used in patch clamp electrophysiology experiments when appropriate. All analyses were performed using GraphPad Prism statistics software, and a P value <0.05 was considered statistically significant.

The data sets generated during the current study are available from the corresponding author on reasonable request.

A 7-day ICV α-klotho treatment in DIO mice significantly reduced body weight (4.9%) compared with vehicle-treated controls (Fig. 1A and B). These changes were, at least in part, due to decreased food intake both daily (14.8%) and after a daytime food restriction (11.4%) (Fig. 1E–H). ICV α-klotho treatment also improved glucose clearance and insulin release during a GTT, as well as insulin sensitivity during an ITT (Fig. 1I–N).

To determine if the effects of central α-klotho on glucose metabolism were independent from changes in body weight, a single ICV α-klotho injection was performed in DIO mice the night before a GTT or ITT. Acute ICV α-klotho treatment improved glucose clearance during a GTT (Fig. 2A–C) and even decreased food intake the following day (Fig. 2D). Interestingly, acute ICV α-klotho treatment had no effects on insulin sensitivity (Fig. 2E–H). These data suggest central α-klotho regulates glucose metabolism independent from changes in body weight and insulin sensitivity. Supporting this hypothesis, 12 days of ICV α-klotho had no effects on insulin-stimulated signaling in hypothalamus, skeletal muscle, eWAT, or liver in weight-matched DIO mice (Supplementary Fig. 3).

To begin to investigate alternative peripheral mechanisms through which central α-klotho improves glucose regulation, basal hepatic gene expression was analyzed. Hepatic PEPCK mRNA was significantly reduced (0.75-fold reduction) in DIO mice treated with α-klotho for 12 days, suggesting attenuated hepatic gluconeogenesis, despite no changes in pyruvate kinase, glucose-6-phsophatase, or glucokinase. α-Klotho–treated mice also had reduced hepatic lipid accumulation and upregulated ACC1 and ACC2 mRNA (Supplementary Fig. 4).

The therapeutic potential of α-klotho was also investigated in a model of type 1 diabetes induced by STZ treatment. Similar to DIO mice, ICV α-klotho decreased body weight (5.3%), suppressed food intake (27.8%), and attenuated hyperglycemia (20.2% reduction in fed glucose levels) in STZ-treated mice compared with vehicle-treated controls (Fig. 3A–G). Even in pair-fed STZ-treated mice, ICV α-klotho attenuated hyperglycemia and trended to improve fasting glucose levels (Fig. 3K–M). These data further demonstrate glucoregulatory and anorexic action of central α-klotho in both type 1 and type 2 diabetes models.

To determine the effects of central α-klotho inhibition on energy and glucose homeostasis, we performed central administration of ab-α-klotho antibody. A 2-day ab-α-klotho treatment significantly impaired glucose tolerance compared with vehicle-treated controls despite similar body weights (Fig. 4A–C). There were no differences in liver PEPCK or glucose-6-phsophatase. However, gene expression of glucokinase and pyruvate kinase trended to be lower in ab-α-klotho mice (P = 0.14 and 0.15, respectively) (Supplementary Fig. 5D). Surprisingly, 7 days of ab-α-klotho significantly decreased body weight with no changes in food intake (Fig. 4D–G). Taken together with ICV α-klotho experiments, these data suggest a distinct glucoregulatory role of central α-klotho independent of body weight and food intake.

We next aimed to investigate the effects of α-klotho on NPY/AgRP neurons considering their critical role in energy homeostasis. A single ICV α-klotho injection in NPY-GFP reporter mice before an overnight fast significantly reduced cFOS colocalization with NPY/AgRP neurons by 49.0% (Fig. 5A–C). Furthermore, electrophysiological recordings revealed α-klotho treatment decreases NPY/AgRP neuron firing rate and membrane potential (0.79 vs. 0.22 Hz and −52.7 vs. −57.8 mV, respectively) (Fig. 5D–F).

To determine if α-klotho’s suppressive effects on NPY/AgRP neurons are due to pre- or postsynaptic events, brain slices were pretreated with TTX (1.0 μmol/L) to block action potentials. In the presence of TTX, α-klotho still decreased membrane potential (−53.4 vs. −58.4 mV), suggesting postsynaptic action of α-klotho on NPY/AgRP neurons (Fig. 5G and H). We also observed α-klotho to increase the magnitude, but not the frequency, of mIPSCs in NPY/AgRP neurons under glutamatergic blockade (25.9 vs. 34.4 pA) (Fig. 5I–M), indicating α-klotho is directly antagonizing NPY/AgRP neurons by modulating receptor availability or intracellular signals (35). Overall, these experiments illustrate that α-klotho directly suppresses NPY/AgRP neuron activity by, at least in part, increasing receptor-mediated inhibitory signals.

To investigate the potential of α-klotho to alter cell signaling and gene expression in the hypothalamus, we first used the GT1-7 immortal hypothalamic cell line (36). A 30-min α-klotho treatment increased phosphorylation of ERK^{thr202/tyr204}, AKT^ser473, and FOXO1^ser256 (Supplementary Fig. 6A). Additionally, α-klotho treatment during both overnight and 2 h of serum starvation significantly reduced AgRP mRNA (by 28.5% and 30.3%, respectively), suggesting hormonal action of α-klotho in GT1-7 cells (Supplementary Fig. 6C). To investigate if α-klotho has hormonal action in the hypothalamus in vivo, we performed acute ICV α-klotho administration in healthy, fed mice and observed elevated phosphorylated ERK after 90 min in the ARC compared with vehicle-treated controls (Supplementary Fig. 6B).

Previous studies demonstrate the importance of α-klotho as a scaffolding protein increasing the affinity of FGF23 to FGFR1 (26,27). In hypothalamic GT1-7 cells, 30-min FGF23 (100 ng/mL) treatment had no effects on phosphorylated ERK or AKT (Fig. 6A–C), while cotreatment with FGF23 and α-klotho had no synergistic effect compared with α-klotho alone. This suggests, at least in hypothalamic GT1-7 cells, α-klotho is independent of exogenous FGF23-mediated signaling. When cells were pretreated with FGFR1 inhibitor PD173074 (10 nmol/L), α-klotho–mediated cell signaling and suppression of AgRP mRNA were abolished (Fig. 6A–D), indicating hypothalamic α-klotho action is dependent on FGFR1 activity. Moreover, immunofluorescent staining of cFOS revealed that ICV pretreatment with PD173074 also inhibited the ability of α-klotho to decrease NPY/AgRP neuron activity in vivo (Fig. 6E–G).

PI3K is a downstream mediator of FGFR1 and is also an important negative regulator of NPY/AgRP neurons (25,37). PI3K inhibition using wortmannin (50 nmol/L) also eliminated α-klotho’s ability to suppress AgRP gene expression (Fig. 6D). Taken together, these data demonstrate the importance of FGFR1/PI3K signaling in hypothalamic α-klotho function.

To determine if the therapeutic effects of α-klotho in DIO mice were dependent on FGFRs, we centrally injected PD173074 to inhibit endogenous FGFR function. Mice receiving α-klotho treatment alone experienced significantly decreased food intake and body weight compared with all groups (Fig. 7A–D), while PD173074 treatment alone induced weight gain (Fig. 7A and B). PD173074 treatment blunted α-klotho–mediated reductions in food intake and body weight, suggesting the effects of central α-klotho on energy balance are mediated by FGFR signaling. Surprisingly, both α-klotho with an FGFR inhibitor and α-klotho alone groups experienced improved glucose clearance compared with vehicle-treated controls (Fig. 7E and F), suggesting FGFRs may not be involved in central α-klotho–mediated glucose regulation.

Similar to FGFR antagonism, central inhibition of PI3K abolished α-klotho’s ability to suppress food intake and improve glucose clearance (Fig. 8). These data indicate that PI3K is critical to α-klotho–mediated regulation of food intake and glucose metabolism. Overall, coupled with in vitro cell signaling experiments, these data demonstrate a novel α-klotho/FGFR/PI3K mechanism in the central regulation of metabolism.

To our knowledge, this is the first study to provide evidence that α-klotho functions as a hypothalamic hormonal agent. ICV α-klotho administration improved glucose regulation, suppressed food intake, and reduced body weight in mouse models of type 1 and 2 diabetes, illustrating the therapeutic potential of central α-klotho in metabolic disease states. Although deeper investigation is required to identify the direct connection between central α-klotho activity and peripheral glucose metabolism, the current study determined that the glucose-lowering effects of α-klotho are independent from insulin sensitivity but, rather, are mediated by augmented insulin secretion during a glucose challenge. However, STZ experiments revealed that some of the glucoregulatory actions of α-klotho are independent of insulin altogether. Basal hepatic PEPCK mRNA was decreased in α-klotho-treated DIO mice, suggesting decreased hepatic glucose output may be an alternative mechanism.

The central and peripheral pools of α-klotho have distinct, independent functions due to α-klotho’s inability to cross the blood-brain barrier (9). While recent publications have demonstrated metabolic roles of α-klotho in the blood, including long-term α-klotho injection improving adiposity in DIO mice (6) and ameliorating diabetic cardiomyopathies in STZ-treated mice (38), the current study identifies distinct differences between central and peripheral α-klotho–mediated metabolic regulation. For example, α-klotho’s effects on food intake and glucose metabolism seem to be mainly via central mechanisms, while peripherally circulating α-klotho regulates gene expression to promote lipid oxidation and energy expenditure (6,38). Notably, whole-body α-klotho knockout and knockdown models have been previously utilized to investigate α-klotho’s functions (1,7), but these approaches do not distinguish between peripheral and central α-klotho function.

ICV ab-α-klotho was used in this study as a novel approach specifically impairing central α-klotho signaling, and as expected, ab-α-klotho treatment impaired glucose clearance. Although central α-klotho concentrations have yet to be quantified in patients with diabetes, past studies show blood α-klotho concentrations to be decreased in some populations with diabetes (8,39). Thus, our data connecting central α-klotho impairment and disordered glucose regulation may provide new insight into the pathophysiology of metabolic disorders.

Contrary to our hypotheses, central α-klotho inhibition resulted in decreased body weight with no differences in food intake. α-Klotho knockout mice also experience weight loss, primarily due to atrophy of metabolically active organs, resulting in premature death (1,7). These findings highlight the complicated and diverse metabolic functions of α-klotho. For example, while evidence from the current study and past literature describes α-klotho as an antidiabetic agent (3–6,38), overexpression of α-klotho has been shown to elicit insulin resistance (1). Notably, α-klotho–overexpressing mice do not experience hyperglycemia, adiposity, or hyperphagia associated with clinical insulin resistance (1). Moreover, α-klotho is an important negative modulator of insulin and IGF-I signaling to regulate apoptosis and ROS buffering (10,11). The many complex physiological roles of α-klotho may explain the unexpected results in response to central α-klotho inhibition.

The current study identified central α-klotho as a novel antagonist of NPY/AgRP neurons. Considering NPY/AgRP neuron overactivity is associated with disordered feeding, body weight, and glucose regulation (13,40), our data provides encouraging evidence of α-klotho as a potential therapeutic target in metabolic disease prevention. At the present, it is unclear if NPY/AgRP neurons are the primary mediators of central α-klotho’s regulation of metabolism, underscoring the importance of further investigation into the specific neuronal effectors and cell signaling involved. However, the observed ICV α-klotho phenotype has many similarities to the previously described effects of NPY/AgRP neuron inhibition, including suppressed food intake, reduced body weight, improved glucose clearance and insulin release, and decreased hepatic gluconeogenic gene expression (13,15–17,41,42).

Similar to findings in studies using hippocampal and oligodendrocyte progenitor cells (10,11,28), α-klotho induced phosphorylation of ERK^{thr202/tyr204}, AKT^ser473, and FOXO1^ser256 in hypothalamic GT1-7 cells—all of which are established signaling molecules involved in downregulating NPY/AgRP gene transcription and activity (18,43). Furthermore, the observed ICV α-klotho phenotype resembles FGFR activation, which also results in suppressed food intake, improved glucose regulation, attenuated NPY/AgRP neuron activity, and decreased liver gluconeogenic gene expression (20–24). α-Klotho serves as a nonenzymatic scaffold to increase FGF23 affinity to FGFR1 (27). Thus, we investigated the potential importance of a hypothalamic α-klotho–FGFR1 signaling mechanism. Similar to previous studies in hippocampal cells, our results show that hypothalamic α-klotho–mediated signaling and AgRP mRNA regulation in GT1-7 cells were abolished with pretreatment with FGFR1 antagonist PD173074 (28). Additional experiments determined that PI3K signaling, a downstream mediator of FGFR1 (37) and potent regulator of NPY/AgRP neurons (25), was also required for α-klotho–mediated AgRP mRNA suppression. Future studies should further investigate the possible involvement of a novel α-klotho–FGFR1–PI3K axis in the homeostatic modulation of NPY/AgRP neurons.

We further investigated the involvement of FGFR/PI3K signaling to central α-klotho–mediated regulation of metabolism. Central FGFR or PI3K inhibition blunted ICV α-klotho’s effects on food intake and body weight, while only PI3K inhibition affected α-klotho–mediated glucose regulation. Overall, these data support the hypothesis that central FGFR-PI3K signaling is critical to α-klotho–mediated regulation of metabolism. However, studies investigating the function of central FGFRs in metabolism yield mixed results depending on animal model and experimental approach. ICV PD173074 (FGFR inhibitor) impairs glucose clearance in healthy rats, but it is described as stress related (23,44). ICV PD173074 in DIO mice elicits no phenotype (21,24). Furthermore, antibody-mediated inhibition of FGFR1 in rodents and monkeys increases energy expenditure, decreases food intake, and reduces body weight, while genetic deletion of FGFR1 in NPY/AgRP neurons also results in no metabolic phenotype (45–47). Additionally, the specificity for PD173074 in vivo is unclear; thus, it likely has nonspecific antagonism of other FGFRs. Future studies should investigate the specific roles of FGFRs, their isoforms, and their neuronal effectors in central regulation of metabolism by performing selective deletion of FGFRs in specific neurons of mature mice using the inducible Cre-LoxP system.

In addition to FGFR-PI3K signaling, there are likely unknown concurrent mechanisms underlying central α-klotho–mediated metabolic regulation. Other neuron populations, such as proopiomelanocortin (POMC) neurons, which are closely associated with NPY/AgRP neurons, may be involved. Our cell culture and immunohistochemistry data also may suggest ERK as an additional cell signaling mechanism of hypothalamic α-klotho action. ERK signaling is downstream of α-klotho, negatively regulates NPY/AgRP neurons, possibly via Kruppel-like factor 4, and is involved in hypothalamic FGF1- and FGF19-mediated glucose lowering (21,43,48).

To summarize, this study identifies α-klotho as a novel antagonist of NPY/AgRP neurons and demonstrates α-klotho’s importance to central regulation of metabolism via an α-klotho–FGFR1–PI3K signaling axis. Our data revealed central administration of α-klotho to yield various therapeutic effects in models of type 1 and 2 diabetes, including improved glucose regulation, suppressed food intake, and reduced body weight. To our knowledge, this study provides the first evidence of α-klotho as a novel hypothalamic regulator of energy balance and glucose metabolism, thus providing new insight into the pathophysiology of metabolic disease.

Funding. The funding for this project was provided by East Carolina University for start-up funds and the National Institute of Diabetes and Digestive and Kidney Disease (DK121215) to H.H.

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

Author Contributions. T.L. wrote the manuscript and performed the majority of the experiments. B.T.L. performed the electrophysiology experiments and reviewed and edited the manuscript. P.L. helped with cell culture experiments and mouse phenotyping. W.B. helped with the immunohistochemistry experiments. Z.R. helped with quantitative PCR. A.P. performed blind quantification of immunofluorescent stains. J.S. performed blind quantification of immunofluorescent stains. H.H. oversaw the project and wrote the manuscript. H.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for its integrity and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the 79th Scientific Sessions of the American Diabetes Association, San Francisco, CA, 7–11 June 2019.