Neuroglial Metabolic Compartmentation Underlying Leptin Deficiency in the Obese ob/ob Mice as Detected by Magnetic Resonance Imaging and Spectroscopy Methods

Abstract

Manganese-Enhanced Magnetic Resonance Imaging (MEMRI), ¹H and ¹³C High-Resolution-Magic Angle Spinning (HR-MAS) Spectroscopy, and genomic approaches were used to compare cerebral activation and neuronal and glial oxidative metabolism in ad libitum fed C57BL6/J leptin-deficient, genetically obese ob/ob mice. T₁-weighted Magnetic Resonance Images across the hypothalamic Arcuate and the Ventromedial nuclei were acquired kinetically after manganese infusion. Neuroglial compartmentation was investigated in hypothalamic biopsies after intraperitoneal injections of [1-¹³C]glucose or [2-¹³C]acetate. Total RNA was extracted to determine the effects of leptin deficiency in the expression of representative genes coding for regulatory enzymes of hypothalamic energy pathways and glutamatergic neurotransmission. Manganese-Enhanced Magnetic Resonance Imaging revealed enhanced cerebral activation in the hypothalamic Arcuate and Ventromedial nuclei of the ob/ob mice. ¹³C HR-MAS analysis showed increased ¹³C accumulation in the hypothalamic glutamate and glutamine carbons of ob/ob mice after the administration of [1-¹³C]glucose, a primarily neuronal substrate. Hypothalamic expression of the genes coding for glucokinase, phosphofructokinase, pyruvate dehydrogenase, and glutamine synthase was not significantly altered while pyruvate kinase expression was slightly upregulated. In conclusion, leptin deficiency associated with obesity led to increased cerebral activation in the hypothalamic Arcuate and Ventromedial nuclei, concomitant with significant increases in neuronal oxidative metabolism and glutamatergic neurotransmission.

Keywords

brain imaging diet energy metabolism magnetic resonance MRI neuronal–glial interactions

Introduction

Obesity is a pandemic syndrome associated with the most prevalent and morbid pathologies in developed countries including heart disease, atherosclerosis, diabetes, and cancer (Das, 2010). The syndrome is thought to arise from disturbances in the highly elaborated mechanisms controlling food intake and energy expenditure involving both systemic and cerebral signalling systems. Body adiposity is currently believed to be regulated systemically through an endocrine ‘adiposity’-negative feedback loop, mainly through the hormones insulin and leptin (Morton et al, 2006). After a meal, increased circulating leptin levels induce an anorexigenic response consisting of a reduction in food intake and an increase in energy expenditure, whereas in fasting periods, decreased plasma levels of leptin promote increased food intake and decreased energy consumption (Coll et al, 2007). Disruptions in the leptin signalling systems are often associated with obesity. For example, obese humans normally present ‘leptin resistance’, a state of reduced responsiveness to plasma leptin, hampering the inhibition of food intake (Lee et al, 2001). Likewise, the leptin-null ob/ob mice exhibit decreased energy expenditure, hyperphagia, and obesity. The administration of exogenous leptin is known to correct the above-mentioned defects in the ob/ob mouse (Pelleymounter et al, 1995).

Abundant evidence indicates that the Central Nervous System is the site for leptin's antiobesity activity, the hypothalamus being the primary cerebral structure supporting leptin performance (Luquet and Magman, 2009). Leptin receptors are expressed at high levels in several hypothalamic regions, mainly in the Arcuate and Ventromedial nuclei both in neurons and in astrocytes (Hakansson et al, 1998; Hsuchou et al, 2009). Indeed by rescuing leptin receptors in the hypothalamus of mice otherwise lacking them, blood glucose and insulin levels are normalized (Coppari et al, 2005). In addition, mice with hypothalamic lesions are resistant to leptin treatment (Koyama et al, 1998). After a meal, leptin released from adipose tissue binds to leptin receptors in the hypothalamus. On leptin binding, the orexigenic neurons releasing neuropeptide Y (NPY) and agouti-related peptide (AgRP) are inhibited and the activity of the anorexigenic neurons expressing pro-opiomelanocortin and the cocaine–amphetamine-regulated transcript is increased (Elmquist et al, 2005). Both orexigenic and anorexigenic neurons project to the Paraventricular nuclei and other hypothalamic sites, to modulate the activity of the melanocortin pathway controlling food intake and energy expenditure (Ghamari-Langroudi et al, 2011).

In addition to the production of orexigenic or anorexigenic neuropeptides, hypothalamic neurons release excitatory and inhibitory amino-acid neurotransmitters as glutamate and gamma-aminobutyric acid (GABA) from glutamate–glutamine and GABA transcellular cycles. However, the neuroglial interactions underlying hypothalamic leptin-dependent activation remain poorly understood. Recently, it was shown that the anorexigenic pro-opiomelanocortin and the orexigenic NPY/AgRP neurons in the Arcuate nucleus of hypothalamus display GABAergic activity (Hentges et al, 2004) whereas other studies indicated that key neurons involved in the regulation of energy balance are glutamatergic and/or densely innervated by glutamatergic nerve terminals (Collin et al, 2003).

A variety of Magnetic Resonance Imaging (MRI) and Magnetic Resonance Spectroscopy methods have been shown recently to provide comprehensive information on cerebral activation and the underlying metabolic coupling mechanisms operating between neurons and glial cells. In particular, Manganese-Enhanced MRI (MEMRI) approaches have revealed the activation of individual hypothalamic nuclei through the localized increased intensity of T₁-weighted MRI (Kuo et al, 2007). Similarly, ¹³C Nuclear Magnetic Resonance investigations of the cerebral metabolism of [1-¹³C]glucose or [2-¹³C]acetate contributed comprehensive information on the operation of the neuronal and glial tricarboxylic acid (TCA) cycles and the transcellular exchanges of glutamate–glutamine or GABA between neurons and glia of the whole brain (Gruetter et al, 2002; Garcia-Espinosa et al, 2004). More recently, we showed that it is possible to improve the spatial resolution and investigate directly hypothalamic metabolism using ¹³C High-Resolution-Magic Angle Spinning (HR-MAS), a novel technique yielding high quality spectra from very small hypothalamic biopsies (Violante et al, 2009).

On these grounds, we investigate here the substrate dependence and neuroglial compartmentation underlying hypothalamic function in control and leptin-deficient, genetically obese ob/ob mice, combining the use of MEMRI and ¹³C HR-MAS. Our results show that leptin deficiency is associated with an increased cerebral activation in the hypothalamic Arcuate and Ventromedial nuclei, as well as with enhanced hypothalamic neuronal oxidative capacity and glutamatergic neurotransmission. These changes do not appear to involve pleiotropic effects of leptin deficiency on gene expression of regulatory enzymes from energy metabolism or glutamatergic neurotransmission, reflecting mainly a relatively enhanced activity of NPY/AgRP hypothalamic pathway in leptin-deficient animals.

Materials and methods

Animals and Research Design

Experimental design and protocols were approved by the Institutional Animal Care and Use Committee and met the guidelines of the appropriate government agency. Male control C57BL6/J and leptin-deficient B6.V-Lep^ob/J ob/ob mice (8- to 12-weeks old) were obtained from the Jackson Laboratories (Bar Harbor, ME, USA). Animals were housed under constant temperature (22°C to 25°C) and maintained on a 12-hour light/dark cycle with ad libitum access to standard chow food and water.

In the morning of the study, blood glucose levels were assessed using a standard glucometer. Plasma insulin or leptin levels were determined using commercially available ELISA kits (Millipore Corporation, Billerica, MA, USA or LINCO Research, St Charles, MO, USA). Brain volumes were determined in each animal from quantitative volumetry studies of T₁-weighted MRI sections as described below.

The effects of leptin deficiency on hypothalamic function were assessed noninvasively, after the hypothalamic uptake of infused manganese (II) chloride (MnCl₂) (Sigma-Aldrich, St Louis, MO, USA) by T₁-weighted MRI in control (n=6) and ob/ob (n=5) mice fed ad libitum. Anesthesia was induced with a mixture of oxygen/isofluorane (96/4%) and maintained by administering oxygen/isofluorane (99/1%). Anesthetized animals received a 30-minute constant (0.2 mL/h) infusion of 0.033 mmol/kg body weight of MnCl₂ in the tail vein, a nontoxic dose (Silva et al, 2004). Manganese (Mn²⁺) infusion was started after two initial baseline acquisitions, after manganese accumulation in the hypothalamus by successive T₁-weighted MRI acquisitions for a total experimental time ∼2 hours.

¹³C HR-MAS studies were performed in a parallel cohort of animals. In these experiments, control (n=10) and ob/ob (n=9) mice, fed ad libitum, received either an intraperitoneal saline injection of [1-¹³C]glucose (20 μmol/g body weight, 99.9% ¹³C, Cambridge Isotope Laboratories, Andover, MA, USA) or [2-¹³C]acetate (60 μmol/g body weight, 99% ¹³C, Cambridge Isotope Laboratories). Fifteen minutes after, cerebral metabolism in animals under terminal anesthetic was arrested within 0.8 seconds using a high-power (5 kW) microwave fixation system (TMW-6402C, Muromachi Kikai Co. Ltd., Tokyo, Japan). The hypothalamic region was dissected from the fixed brains and kept frozen at −80°C until ¹H and ¹³C HR-MAS spectroscopy analysis.

Finally, to investigate the genetic basis of the observed effects we prepared biopsies from the hypothalamus of ad libitum fed control (n=4) and ob/ob (n=5) mice and determined the expression of genes involved in the regulation of the glycolysis or TCA and glutamate–glutamine cycles.

Magnetic Resonance Imaging

We used T₁-weighted MRI to monitor kinetically Mn²⁺ accumulation under high-spatial resolution conditions and T₂-weighted MRI to determine total cerebral volume in vivo. Magnetic Resonance Imaging was performed using a 7.0-T horizontal-bore (160 mm) superconducting magnet (PharmaScan, Bruker BioSpin Gmbh, Ettlingen, Germany) equipped with a ¹H selective birdcage resonator of 38 mm. Spin-echo multi-slice T₁-weighted imaging of sagittal sections across the mouse brain (field of view 22 mm, 256 × 256 matrix size, 1 mm slice thickness, 5 slices, 85 μ² in plane resolution) was acquired successively using repetition time=500 ms, echo time=11.8 ms and three averages, resulting in a total acquisition time of 6 minutes 24 seconds for each scan. In each scan, regions of interest corresponding to the hypothalamic Arcuate and Ventromedial nuclei, and the fourth ventricle were selected and analyzed, after alignment and regionalization with a standard mouse brain atlas (Paxinos and Franklin, 2001). Signal intensity changes, were assessed using a standard image-processing software (NIH ImageJ 1.37v, http://rsbweb.nih.gov/ij/index.html), normalizing the values of each animal to those obtained before Mn²⁺ infusion. T₂-weighted MRI axial images were acquired under identical resolution conditions with repetition time=3,000 ms, echo time=60 ms and three averages. T₂ sections swept the entire mouse brain and cerebral areas in each slice were quantified by using the freely accessible Image-processing software (NIH ImageJ 1.37v). The brain volume of each slice was obtained by multiplying the MRI determined brain area by the slice thickness, the sum of all slices providing the brain volume.

¹H and ¹³C High-Resolution-Magic Angle Spinning Magnetic Resonance Spectroscopy

¹H and ¹³C HR-MAS spectra were acquired at 11.7 T with a Bruker AVANCE wide bore spectrometer equipped with a magic angle spinning accessory (Bruker Biospin, Rheinstetten, Germany). Biopsies from the hypothalamus were placed in 50 μL HR-MAS Zirconium Oxide rotors, transferred to the HR-MAS probe and spun at 4 kHz (277 K). ¹H-decoupled ¹³C HR-MAS spectra were acquired using a pulse-acquire sequence with WALTZ-16 ¹H decoupling applied during the acquisition and relaxation delay periods. Conditions were π/4 pulse width, 64 k data points acquired in 8 k scans with 5 seconds recycle delay with a total acquisition time ∼16 hours. The chemical shifts of the ¹³C resonances were referred to those of the lactate C3 peak (21.9 parts per million (p.p.m.)) in the glucose infusion experiments or to the acetate C2 peak (24.5 p.p.m.) in the acetate infusion experiments. Spectral deconvolution was performed using a PC-based Nuclear Magnetic Resonance analysis program, NUTS (Acorn, Freemont, CA, USA). ¹H HR-MAS spectra were acquired at the beginning and after the overnight ¹³C acquisition for quantification of total metabolite concentrations and verification of sample stability. Acquisition parameters for the ¹H HR-MAS sequence were 5 seconds water presaturation, echo time=144 ms, τ=1 ms, n=100, 10 kHz spectral width, 32 k data points, and 128 scans. Total metabolite concentrations were quantified by using a modified version of the LCModel processing software programme (Linear Combination of Model Spectra, http://s-provencher.com/pages/schtm; Provencher, 1993) with a customized set of spectra from cerebral metabolites (Righi et al, 2009).

Gene Expression Assays

Total RNA was prepared from hypothalamic biopsies using the RNAspin Mini RNA Isolation kit (Amersham Biosciences, GE Healthcare, Piscataway, NJ, USA) and retrotranscribed to cDNA using the high capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA). The expression of the genes for the main enzymes regulating glycolysis, TCA cycle, and glutamate–glutamine cycling was assayed by Real-Time PCR (RT-PCR) assays using the individual probes for each gene commercialized as TaqMan1 (Applied Biosystems). The following genes (gene symbol, TaqMan1 probe) were investigated: GCK, Glucokinase (mCG21380, Mn00439129_m1); PFK, Phosphofructokinase (mCG8414, Mm00445461_m1); PDH, Pyruvate Dehydrogenase (mCG14551, Mm00468675_m1); PK, Pyruvate Kinase (mCG17567, Mm00443090_m1); and GS, Glutamate-ammonia Ligase or Glutamine Synthetase (mCG20133, Mm00725701_s1).

Statistical Analysis

Results are shown as mean±s.e.m. and unpaired Student's t-test or two-way analysis of variance was used to compare control and ob/ob mice. Comparisons with P<0.05 were considered statistically significant.

Results

Physiological and Endocrine Variables

Ob/ob mice are obese when compared with control animals (53±1 versus 26±1 g, P<0.01). As expected, in the fed state, ob/ob mice glycemia was slightly lower than in controls (9.7±0.6 versus 12.5±0.8 mmol/L, P<0.05) (probably reflecting the hyperinsulinemia observed, 7.7±1.6 versus 0.6±0.2 ng/mL, P<0.01, in agreement with previous results; Hinoi et al, 2008). Elevated blood glucose levels detected in either ob/ob or control mice reflect the fact that animals were under heavy anesthesia. After overnight feeding, leptin secretion was virtually undetectable in ob/ob mice relative to control animals (0.1±0.1 versus 2.5±1.0 ng/mL, P<0.05). Analysis of T₂-weighted MRI scans (data not shown) revealed that the cerebral volume was significantly reduced in the ob/ob mice relative to their counterpart controls (428±18 versus 475±7 mm³, P<0.05), in agreement with a recent report where obesity was associated with detectable brain volume deficits in cognitively normal elderly subjects (Raji et al, 2009).

Manganese-Enhanced Magnetic Resonance Imaging of Leptin Deficiency

Figure 1 shows a representative T₁-weighted image of the mouse brain after Mn²⁺ uptake (upper panel) and the time course of normalized signal intensity changes in the hypothalamic Arcuate and Ventromedial nuclei as well as in the fourth ventricle of control and ob/ob mice during Mn²⁺ infusion (lower panel), localized as described in Materials and methods. The enhancement of signal intensity in the forth ventricle (V4) was similar in both groups of animals. However, normalized signal intensities in the hypothalamic Arcuate (Arc) and Ventromedial (VMH) nuclei during Mn²⁺ infusion were consistently higher in the ob/ob mice than in the controls. The average five-fold increase in MEMRI signal observed in the nuclei of leptin-deficient animals is consistent with significantly enhanced hypothalamic activation.

Figure 1

Representative T₁-weighted image of the mouse brain (sagittal section) after intravenous infusion of Manganese (Mn²⁺) in a control mouse (upper panel). A coronal section of the area of interest is also represented in the upper left insert. The change in normalized signal intensities with time is shown in the lower panels for the hypothalamic Arcuate nucleus (Arc), hypothalamic Ventromedial nucleus (VMH) and the forth ventricle (V4) in ob/ob mice, n=5 (black squares) and in control animals, n=6 (gray squares). Data are shown as average±s.e.m., ^∗P<0.05 relative to controls.

Glutamatergic and GABAergic Neurotransmission Under Leptin Deficiency

¹H HR-MAS spectra from representative hypothalamic tissue biopsies after [1-¹³C]glucose or [2-¹³C]acetate administration are illustrated in Figures 2A and 2B, respectively. Comparisons between the ¹H spectra acquired before and after each ¹³C spectrum (not shown) showed that samples did not suffer significant degradation during the ∼16-hour acquisition period necessary to obtain the ¹³C HR-MAS spectra (Violante et al, 2009). Hypothalamic metabolite concentrations for control and ob/ob mice (relative to N-acetyl aspartate), after the administration either of [1-¹³C]glucose or of [2-¹³C]acetate are presented in Figure 3. After the administration of [1-¹³C]glucose, ob//ob mice showed reduced hypothalamic glucose concentrations relative to control animals whereas the relative concentrations of GABA, glutamate, and glutamine were similar in both groups. After the intraperitoneal injection of [2-¹³C]acetate, no significant differences were observed between hypothalamic metabolite concentrations of control and ob/ob mice. However, in control animals the hypothalamic myo-inositol concentrations were lower than those determined in the animals receiving [1-¹³C]glucose infusions and did not change appreciably after the [2-¹³C]acetate infusion.

Figure 2

¹H High-Resolution-Magic Angle Spinning (HR-MAS) spectra (500 MHz, 277 K, 4 KHz) from representative hypothalamic tissue biopsies of a control mouse after (A) [1-¹³C]glucose or (B) [2-¹³C]acetate administrations. Assignments: Ac, acetate; GABA, gamma-aminobutyric acid; Glu, glutamate; Gln, glutamine; Glx, glutamate and/or glutamine; Lac, lactate; NAA, N-acetyl aspartate; p.p.m., parts per million.

Figure 3

Relative concentrations of acetate, gamma-aminobutyric acid (GABA), glucose, glutamine, glutamate, lactate, and myo-inositol with respect to N-acetyl aspartate (NAA) in the hypothalamus of control and ob/ob mice after the administration of either [1-¹³C]glucose or [2-¹³C]acetate. Data are represented as average±s.e.m., ^∗P<0.05 is indicated. White columns correspond to controls infused with [2-¹³C]acetate (n=5), light gray columns to controls infused with [1-¹³C]glucose (n=5), gray columns to ob/ob mice infused with [2-¹³C]acetate (n=4), and finally dark gray columns to ob/ob mice infused with [1-¹³C]glucose (n=5). Acetate/NAA ratios are shown multiplied by 10.

Representative ¹³C HR-MAS spectra from hypothalamic biopsies after [1-¹³C]glucose or [2-¹³C]acetate administration are depicted in Figures 4A and 4B, respectively. In the experiments with [1-¹³C] glucose, the areas of ¹³C HR-MAS resonances were normalized to the area of the myo-inositol (mI) C1, C3 (73.2 p.p.m.) resonance (Cerdán et al, 1985). The incorporation of ¹³C from the labelled precursor in the different ¹³C HR-MAS detectable carbons revealed that [1-¹³C]glucose and [2-¹³C]acetate were efficiently used as energy substrates in the hypothalamus of both types of animals. Moreover, the different labelling patterns observed in the hypothalamus parallel those previously described in the whole brain, with higher ¹³C labelling obtained after the [1-¹³C]glucose infusions (Cerdán et al, 1990; Chapa et al, 2000). The lower incorporation detected after [2-¹³C]acetate administration reflects the smaller incorporation of ¹³C label from labelled acetate into hypothalamic metabolites, suggesting a relatively lower contribution of oxidative glial metabolism to hypothalamic labelling.

Figure 4

¹³C High-Resolution-Magic Angle Spinning (HR-MAS) spectra (125 MHz, 277 K, 4 KHz rotation) from representative hypothalamic biopsies of a control mouse after (A) [1-¹³C]glucose or (B) [2-¹³C]acetate administration. Ac, acetate; GABA, gamma-aminobutyric acid; Glu, glutamate; Gln, glutamine; Lac, lactate; p.p.m., parts per million.

Figure 5 compares ¹³C incorporation from [1-¹³C]glucose or [2-¹³C]acetate in individual carbons of metabolites from the hypothalamus of control and ob/ob mice. In Figure 5A, the administration of [1-¹³C]glucose resulted in significantly increased ¹³C incorporation (ca. three-fold) in glutamate and glutamine carbons of ob/ob mice without a concomitant increase in the total concentration of these metabolites (see Figure 3). This suggests that the leptin-null ob/ob mice depict net increases in neuronal oxidative capacity and glutamatergic neurotransmission together with elevated glutamate–glutamine cycling. Since the cerebral glucose concentration decreases slightly in ob/ob animals (Figure 3) and its ¹³C content increases (Figure 5A), the large increases in neuronal oxidative capacity and glutamine cycling reported here, may include a small contribution derived from increased ¹³C enrichment in the cerebral glucose precursor. With labelled acetate, the smaller concentrations of myo-inositol in control and ob/ob mice (Figure 3), made the natural abundance myo-inositol C1,C3 ¹³C resonances not detectable in these animals, precluding their absolute quantification. As described for the whole brain, the glutamate/glutamine labelling ratio was higher with [1-¹³C]glucose as substrate and smaller with [2-¹³C]acetate as substrate (Figure 5B), a similar metabolic compartmentation effect to that described in the whole brain (Cerdán et al, 1990; Chapa et al, 2000). No significant difference in the glutamate C4/glutamine C4 ratio of control and ob/ob mice infused with [2-¹³C]acetate was measured while a significant increase in the glutamate C4/glutamine C4 ratio was detected, as expected, in the [1-¹³C]glucose-infused animals. These findings suggest that glial oxidative metabolism and the glutamine cycle are not significantly modified during leptin-dependent hypothalamic appetite regulation in the [2-¹³C]acetate-infused animals. Taken together, our results disclose a significantly increased neuronal oxidative capacity and glutamate–glutamine cycling in the hypothalamus of ob/ob mice.

Figure 5

(A) ¹³C concentration of relevant hypothalamic metabolite carbons relative to the C1,C3 myo-inositol resonance (73.2 p.p.m.) after administration of [1-¹³C]glucose for control (n=5) and ob/ob mice (n=5). (B) Hypothalamic GluC4/GlnC4 ratios obtained after the administration of [1-¹³C]glucose or [2-¹³C]acetate. White columns correspond to controls infused with [2-¹³C]acetate (n=5), light gray columns to controls infused with [1-¹³C]glucose (n=5), gray columns to ob/ob mice infused with [2-¹³C]acetate (n=4), and finally dark gray columns to ob/ob mice infused with [1-¹³C]glucose (n=5). Assignments: GABA, gamma-aminobutyric acid; Glc, glucose and its respective anomers α and β; Glu, glutamate; Gln, glutamine; Lac, lactate. Data are represented as average±s.e.m., ^∗P<0.05 and ^∗∗P<0.01 relative to controls are indicated.

Gene Expression Studies in Control and ob/ob Mice Hypothalamus

To investigate if the metabolic and glutamate–glutamine cycling differences observed between control and ob/ob mice were due to genetic determinants, we investigated the expression of the genes coding for the main regulatory enzymes of energy metabolism including glycolysis, the TCA cycle, and glutamateric neurotransmission in ob/ob mice relative to controls. In particular, we assayed the expression of GCK, PFK, PDH, PK, and GS (Figure 6).

Figure 6

Relative expression of the genes coding for the main regulatory enzymes of glycolysis, tricarboxylic acid (TCA) cycle metabolism, and glutamate–glutamine cycling, in ob/ob mice (n=5) (dark gray columns) relative to controls (n=4). Data are represented as average±s.e.m. and ^∗P<0.05 is indicated. GCK, glucokinase; PFK, phosphofructokinase; PDH, pyruvate dehydrogenase; PK, pyruvate kinase; and GS, glutamate-ammonia ligase (glutamine synthetase).

No significant differences were observed in the bulk expression of the main regulatory enzymes of glycolysis (GCK and PFK) and glutamate–glutamine cycle (GS). However, gene expression of PK was significantly upregulated in the ob/ob mice hypothalamus.

Discussion

In the fed state, leptin is secreted from adipose tissue and binds to the hypothalamic leptin receptor inhibiting appetite. In the leptin-deficient ob/ob mice, hypothalamic leptin signalling is drastically reduced and hyperphagia develops leading to obesity. In this study, MEMRI was used to compare hypothalamic activation in ad libitum fed control and in the obese leptin-deficient ob/ob mice. It is a valuable tool for assessing noninvasively regions of cerebral activity since it combines two important properties of manganese: (1) as an indicator of calcium (Ca²⁺) influx entering the neurons during activation and (2) as a paramagnetic label, its intracellular accumulation yielding positive MRI contrast in activated areas (Silva et al, 2004). Calcium is an essential second messenger necessary for brain function as neuronal depolarization leads to a rise in intracellular Ca²⁺ levels derived from increased transport across the plasma membrane through ligand-gated or voltage-gated Ca²⁺ channels. Intracellular Ca²⁺ rise is essential for neurotransmitter release involved in neuronal signal transduction (Nicholls et al, 1992). Since Mn²⁺ competes effectively with Ca²⁺ transport, the MEMRI signal detected is thought to reflect increased Ca²⁺ uptake during cerebral activation and can be used to interrogate whole-brain neuronal tracts (Chuang and Koretsky, 2006). Indeed, MEMRI was previously used successfully to investigate alterations of cerebral activity in key appetite-regulatory regions of the hypothalamus after peripheral injection of orexigenic or anorexigenic peptides in healthy control mice (Kuo et al, 2007). Our results show increased Mn²⁺ accumulation concomitant with enhanced cerebral activity in the hypothalamic Arcuate and Ventromedial nuclei of the leptin-null obese ob/ob mice as compared with controls. It is well known that homeostatic levels of leptin inhibit the orexigenic NPY/AgRP neurons and stimulate the anorexigenic neurons expressing pro-opiomelanocortin and cocaine–amphetamine-regulated transcript in the hypothalamic Arcuate nucleus (Elmquist et al, 2005). On fasting, reduced circulating leptin levels release the inhibition of the NPY/AgRP neurons, resulting in increased orexigenic activity (Takahashi and Cone, 2005). In the leptin-deficient ob/ob mice, the inhibition of NPY/AgRP neurons is almost completely eliminated, driving the orexigenic response even higher. This is supported by previous results where a higher excitatory tone onto NPY neurons and a marked net increase in inhibitory tone onto the pro-opiomelanocortin neurons was observed in hypothalamic Arcuate nucleus slices of the ob/ob mice (Pinto et al, 2004). In contrast with the hypothalamic Arcuate nucleus, the hypothalamic Ventromedial nucleus does not produce large amounts of orexigenic or anorexigenic neuropeptides. However, a vesicular glutamate transporter protein, a marker for glutamatergic neurons, is expressed in the vast majority of these neurons suggesting that glutamate, most probably through glutamate/glutamine neuroglial cycling, may have an important role integrating the metabolic information in the hypothalamic Ventromedial nucleus (Collin et al, 2003). Early studies on the regulation of appetite and obesity emphasized the role of peptide hormones, neuropeptides, and monoamines acting on membrane receptors to trigger intracellular signalling cascades, with relatively little attention dedicated to the role of amino-acid neurotransmission and associated neuroglial interactions (Morton et al, 2006). The view that amino-acid neurotransmitters are present in key hypothalamic neurons that regulate appetite and energy status modified this perception (Meister, 2007). In vivo and in vitro¹H and ¹³C Nuclear Magnetic Resonance approaches have been successfully used in the study of neurotransmitter metabolism and the associated neuroglial interactions during sensory or motor activation (Sappery-Marinier et al, 1992; Hyder et al, 1996). However, the relatively large voxel size used in the acquisition of in vivo¹³C Magnetic Resonance spectra precludes its use for studying the relatively small appetite-controlling hypothalamic area of small rodents. Likewise, the relatively large amounts of cerebral tissue needed to prepare brain extracts for high-resolution ¹³C Nuclear Magnetic Resonance constituted an important limitation. Recently, a new methodology combining ¹³C labelled substrates with ¹³C HR-MAS Spectroscopy was proposed to provide further insights into hypothalamic neurotransmitter metabolism (Violante et al, 2009). ¹³C HR-MAS Spectroscopy offers the advantage of providing information on cerebral metabolic pathways by allowing for the use of high-resolution spectroscopy in smaller tissue biopsies.

¹³C HR-MAS Spectroscopy analysis indicates that the leptin-dependent hypothalamic activation involves mainly increases in neuronal oxidation and glutamatergic neurotransmission. Indeed, it has been previously reported that the injection of glutamate in the hypothalamus rapidly elicits an intense feeding response in satiated healthy rats (Stanley et al, 1993). In contrast with other cerebral activation paradigms (e.g., motor and sensorial) where increased glycolytic responses are observed as lactate increases (Ueki et al, 1988), the neuronal oxidative response appears to be predominant in leptin-dependent appetite regulation. Increased neuronal oxidative metabolism in ob/ob mice involves necessarily increased neuronal pyruvate availability, consistent with the bulk PK gene upregulation as measured by RT-PCR. Hypothalamic PK expression appears to be associated with the prandial state as in fasting (low leptin levels). PK is overexpressed relative to the fed state (high leptin) (Poplawski et al, 2010). In this sense, leptin deficiency may further account for the overexpression of the hypothalamic PK gene in the ob/ob mice as detected in the present study. This finding appears to contrast with the modelling comments by Jolivet et al (2010) where neurons are proposed to be unable to upregulate glycolytic flux. However, this discrepancy may only be apparent. Our gene expression assays measure the net balance in hypothalamic gene expression, making the results compatible with opposite changes in the neuronal or glial compartments. In this sense, it is possible that the net increase in hypothalamic PK expression reflects the balance between a more marked increase in astrocytic PK expression and a decreased neuronal PK expression. In any case, the increased PK expression found did not result in increased lactate levels, suggesting that the ultimate metabolic impact of hypothalamic PK upregulation may be limited by posttranslational events. It should be mentioned in this respect here, however, that independent work reported also that neurons are able to increase their oxidative metabolism in parallel with an increase in pyruvate availability as generated by neuronal glycolysis during sustained checkerboard stimulation of human visual cortex in vivo (Gjedde and Marrett, 2001).

The elevated incorporation of ¹³C from [1-¹³C]glucose into hypothalamic glutamine in ob/ob mice reveals increased glutamine and glutamate cycling as associated with increased glutamatergic activity (Sibson et al, 1997), providing also additional information on the origin of increased neuronal pyruvate. Our results reveal that the increased [4-¹³C]glutamine labelling is produced in ob/ob mice under conditions where glial oxidative metabolism, as monitored by the oxidation of [2-¹³C]acetate, is not significantly enhanced. This suggests that the ATP required for augmented glutamine synthesis, may be mainly derived from astrocytic anaerobic glycolysis, supporting a glial origin for the additional pyruvate required to fuel the increased neuronal oxidation detected. Notably, we found no significant change in the lactate ¹H or ¹³C HR-MAS resonances in ob/ob mice, implying that the increased lactate produced to support the augmented glutamine synthesis, is eventually consumed, most probably as pyruvate to fuel the neuronal cycle. Furthermore, since leptin deficiency stimulates NPY/AgRP neurons and inhibits POMPCT/cocaine–amphetamine-regulated transcript neurons, it is most probable that the increased oxidative activity and glutamatergic neurotransmission detected by ¹³C HR-MAS reflect these circumstances, as visualized by the corresponding increases in MEMRI signal intensities in the hypothalamic Arcuate and Ventromedial nuclei. It should be mentioned, however, that these results do not exclude the operation of additional signalling molecules and mechanisms regulating appetite and obesity, including ghrelin, peptide YY, fatty acids, insulin, and glucose itself (Morton et al, 2006). To the best of our knowledge, this is the first time that a positive correlation is established between increased cerebral oxidative activity and glutamatergic neurotransmission and the elevated MEMRI signal observed during cerebral activation. These concepts are summarized in Figures 7A and 7B in the frame of previous neuroendocrine concepts (Morton et al, 2006; Coll et al, 2007).

Figure 7

Summary of the effects of leptin in the hypothalamus of control (A) and ob/ob mice (B), as revealed by Manganese-Enhanced Magnetic Resonance Imaging (MEMRI) and ¹³C High-Resolution-Magic Angle Spinning (HR-MAS). In the Arcuate nucleus of control animals, leptin (gray triangles, red arrow up) binds to the leptin receptors (inverted Y) inhibiting (negative red circles) the orexigenic Agouti-related peptide/Neuropeptide Y (AgRP/NPY) neurons and stimulating (positive red circles) the anorexigenic pro-opiomelanocortin/cocaine- and amphetamine-regulated transcript (POMC/CART) neurons. In the genetically obese, leptin-null ob/ob mice (red arrow down), the physiological inhibition of the AgRP/NPY neurons in the Arcuate and downstream Paraventricular (PVN) connections is almost fully released (positive red circles), resulting in augmented neuronal oxidative capacity (darker TCA_n), enhanced glutamatergic neurotransmission and neuroglial glutamate/glutamine cycling (darker transcellular arrows) without appreciable effects in the astrocytic oxidative capacity (TCA_g). Together, these changes lead to relative increases in MEMRI signal intensities in the Arcuate (MEMRI red circles +) and ¹³C labelling from [1-¹³C]glucose in glutamate/glutamine in AgRP/NPY neurons and downstream paraventicular (PVN) connections and melanocortin (MC) pathway, resulting in an uncompensated orexigenic response and obesity. A, Astrocyte; ARC, Arcuate; GLU, glutamate; MC4, MC3, melanocortin receptors; N, neuron; TCA, tricarboxylic acid cycle; PVN, Paraventricular.

In conclusion, the data presented in this work show that leptin deficiency associated with obesity leads to increased cerebral activation in the hypothalamic Arcuate and Ventromedial nuclei. These changes involve enhanced neuronal oxidative capacity and glutamatergic neurotransmission whereas pleiotropic effects of leptin deficiency in gene expression appear to be much less pronounced under these conditions.

The possibility to visualize noninvasively hypothalamic leptin activity by MEMRI and evaluate the underlying neuroglial interactions by ¹³C HR-MAS opens a new avenue in the molecular imaging of signalling cascades and may become instrumental for the development of novel therapeutic strategies for obesity and other feeding disorders.

Footnotes

Acknowledgements

The authors are indebted to Mr Javier Pérez CSIC for careful drafting of the illustrations. The authors would also like to acknowledge Santiago Canals CSIC for helpful discussion regarding MEMRI experimental protocols and Jesús Pacheco-Torres CSIC for insights of the LC-Model software program.

The authors declare no conflict of interest.

References

Cerdán

Kunnecke

Seelig

(1990) Cerebral metabolism of [1,2-¹³C₂]acetate as detected by in vivo and in vitro ¹³C NMR. J Biol Chem 265:12916–26

Cerdán

Parrilla

Santoro

Rico

(1985) ¹H NMR detection of cerebral myo-inositol. FEBS Lett 187:167–72

Chapa

Cruz

Garcia-Martin

Garcia-Espinosa

Cerdán

(2000) Metabolism of (1-¹³C)glucose and (2-¹³C, 2-²H3)acetate in the neuronal and glial compartments of the adult rat brain as detected by ¹³C, ²H NMR Spectroscopy. Neurochem Int 37:217–28

Chuang

Koretsky

(2006) Improved neuronal tract tracing using manganese enhanced magnetic resonance imaging with fast T(1) mapping. Magn Reson Med 55:604–11

Coll

Farooqi

O'Rahilly

(2007) The hormonal control of food intake. Cell 129:251–62

Collin

Backberg

Ovesjo

Fisone

Edwards

Fujiyama

Meister

(2003) Plasma membrane and vesicular glutamate transporter mRNAs/proteins in hypothalamic neurons that regulate body weight. Eur J Neurosci 18:1265–78

Coppari

Ichinose

Lee

Pullen

Kenny

McGovern

Tang

Liu

Ludwig

Chua

Jr Lowell

Elmquist

(2005) The hypothalamic arcuate nucleus: a key site for mediating leptin's effects on glucose homeostasis and locomotor activity. Cell Metab 1:63–72

Das

(2010) Obesity: genes, brain, gut, and environment. Nutrition 26:459–73

Elmquist

Coppari

Balthasar

Ichinose

Lowell

(2005) Identifying hypothalamic pathways controlling food intake, body weight, and glucose homeostasis. J Comp Neurol 493:63–71

10.

Garcia-Espinosa

Rodrigues

Sierra

Benito

Fonseca

Gray

Bartnik

Garcia-Martin

Ballesteros

Cerdán

(2004) Cerebral glucose metabolism and the glutamine cycle as detected by in vivo and in vitro ¹³C NMR spectroscopy. Neurochem Int 45:297–303

11.

Ghamari-Langroudi

Srisai

Cone

(2011) Multinodal regulation of the arcuate/paraventricular nucleus circuit by leptin. Proc Natl Acad Sci USA 108:355–60

12.

Gjedde

Marrett

(2001) Glycolysis in neurons, not astrocytes, delays oxidative metabolism of human visual cortex during sustained checkerboard stimulation in vivo. J Cereb Blood Flow Metab 21:1384–92

13.

Gruetter

Seaquist

Ugurbil

(2002) A mathematical model of compartmentalized neurotransmitter metabolism in the human brain. Am J Physiol Endocrinol Metab 281:E100–12

14.

Hakansson

Brown

Ghilardi

Skoda

Meister

(1998) Leptin receptor immunoreactivity in chemically defined target neurons of the hypothalamus. J Neurosci 8:559–72

15.

Hentges

Nishiyama

Overstreet

Stenzel-Poore

Williams

Low

(2004) GABA release from proopiomelanocortin neurons. J Neurosci 24:1578–83

16.

Hinoi

Gao

Jung

Yadav

Yoshizawa

Myers

Chua

Jr Kim

Kaestner

Karsenty

(2008) The sympathetic tone mediates leptin's inhibition of insulin secretion by modulating osteocalcin bioactivity. J Cell Biol 183:1235–42

17.

Hsuchou

Pan

Barnes

Kastin

(2009) Leptin receptor mRNA in rat brain astrocytes. Peptides 30: 2275–80

18.

Hyder

Chase

Behar

Mason

Siddee

Rothman

Shulman

(1996) Increased tricarboxylic acid flux in rat brain during forepaw stimulation detected with ¹H, ¹³C NMR. Proc Natl Acad Sci USA 93:7612–7

19.

Jolivet

Allaman

Pellerin

Magistretti

Weber

(2009) Comment on recent modelling studies of astrocyte-neuron metabolic interactions. J Cereb Blood Flow Metab 30:1982–6

20.

Koyama

Shimabukuro

Chen

Wang

Lee

Kalra

Dube

Kalra

Newgard

Unger

(1998) Resistance to adenovirally induced hyperleptinemia in rats. Comparison of ventromedial hypothalamic lesions and mutated leptin receptors. J Clin Invest 102:728–33

21.

Kuo

Parkinson

Chaudhri

Herlihy

Dhillo

Small

Bloom

Bell

(2007) The temporal sequence of gut peptide CNS interactions tracked in vivo by magnetic resonance imaging. J Neurosci 27: 12341–8

22.

Lee

Reed

Price

(2001) Leptin resistance is associated with extreme obesity and aggregates in families. Int J Obes Relat Metab Disord 25:1471–3

23.

Luquet

Magman

(2009) The central nervous system at the core of the regulation of energy homeostasis. Front Biosci 1:448–65

24.

Meister

(2007) Neurotransmitters in key neurons of the hypothalamus that regulate feeding behaviour and body weight. Physiol Behav 92:263–71

25.

Morton

Cummings

Baskin

Barsh

Schwartz

(2006) Central nervous system control of food intake and body weight. Nature 443:289–95

26.

Nicholls

Martin

Wallace

Fuchs

, (eds) (1992) From Neuron to Brain, 4th edn., Sinnauer Associates: Sunderland, MA, USA

27.

Paxinos

Franklin

, (eds) (2001) The Mouse Brain in Stereotaxic Coordinates. Academic Press, Inc.: New York

28.

Pelleymounter

Cullen

Baker

Hecht

Winters

Boone

Collins

(1995) Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269:540–3

29.

Pinto

Roseberry

Liu

Diano

Shanabrough

Cai

Friedman

Horvath

(2004) Rapid rewiring of arcuate nucleus feeding circuits by leptin. Science 304:110–5

30.

Poplawski

Mastaitis

Yang

Mobbs

(2010) Hypothalamic responses to fasting indicate metabolic reprogramming away from glycolysis toward lipid oxidation. Endocrinology 151:5206–17

31.

Provencher

(1993) Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med 30:672–9

32.

Raji

Parikshak

Becker

Lopez

Kuller

Hua

Leow

Toga

Thompson

(2009) Brain structure and obesity. Hum Brain Mapp 73:1899–905

33.

Righi

Roda

Paz

Mucci

Tugnoli

Rodriguez-Tarduch

Barrios

Schenetti

Cerdan

García-Martín

(2009) ¹H HR-MAS and genomic analysis of human tumor biopsies discriminate between high and low grade astrocytomas. NMR Biomed 22:629–37

34.

Sappery-Marinier

Calabrese

Fein

Hugg

Biggins

Weiner

(1992) Effects of photic stimulation on human visual cortex lactate and phosphates using ¹H and ³¹P Magnetic Resonance Spectroscopy. J Cereb Blood Flow Metab 12:618–23

35.

Sibson

Dhankhar

Mason

Behar

Rothman

Shulman

(1997) In vivo ¹³C NMR measurements of cerebral glutamine synthesis as evidence for glutamate-glutamine cycling. Proc Natl Acad Sci USA 94:2699–704

36.

Silva

Lee

Aoki

Koretsky

(2004) Manganese-enhanced magnetic resonance imaging (MEMRI): methodological and practical considerations. NMR Biomed 17:532–43

37.

Stanley

Spears

Dee

(1993) Lateral hypothalamic injections of glutamate, kainic acid, D,L-alpha-amino-3-hydroxy-5-methyl-isoxazole propionic acid or N-methyl-D-aspartic acid rapidly elicit intense transient eating in rats. Brain Res 613:88–95

38.

Takahashi

Cone

(2005) Fasting induces a large, leptin-dependent increase in the intrinsic action potential frequency of orexigenic arcuate nucleus neuropeptide Y/agouti-related protein neurons. Endocrinology 146:1043–7

39.

Ueki

Linn

Hossmann

(1988) Functional activation of cerebral blood flow and metabolism before and after global ischemia of rat brain. J Cereb Blood Flow Metab 8:486–94

40.

Violante

Anastasovska

Sanchez-Canon

Rodrigues

Righi

Nieto-Charques

Parkinson

Bloom

Bell

Cerdán

(2009) Cerebral activation by fasting induces lactate accumulation in the hypothalamus. Magn Reson Med 62:279–83

Neuroglial Metabolic Compartmentation Underlying Leptin Deficiency in the Obese ob/ob Mice as Detected by Magnetic Resonance Imaging and Spectroscopy Methods

Abstract

Keywords

Introduction

Materials and methods

Animals and Research Design

Magnetic Resonance Imaging

1H and 13C High-Resolution-Magic Angle Spinning Magnetic Resonance Spectroscopy

Gene Expression Assays

Statistical Analysis

Results

Physiological and Endocrine Variables

Manganese-Enhanced Magnetic Resonance Imaging of Leptin Deficiency

Glutamatergic and GABAergic Neurotransmission Under Leptin Deficiency

Gene Expression Studies in Control and ob/ob Mice Hypothalamus

Discussion

Footnotes

Acknowledgements

References

¹H and ¹³C High-Resolution-Magic Angle Spinning Magnetic Resonance Spectroscopy