Effect of manganese chloride on the neurochemical profile of the rat hypothalamus

Effect of manganese chloride on theneurochemical profile of the rat hypothalamus

Nathalie Just1,2, Cristina Cudalbu1, Hongxia Lei1 and Rolf Gruetter1,2,3

1Laboratory for Functional and Metabolic Imaging, Ecole Polytechnique Federale de Lausanne, Lausanne,Switzerland; 2Department of Radiology, University of Lausanne, Lausanne, Switzerland; 3Department ofRadiology, University of Geneva, Geneva, Switzerland

Manganese (Mn2 +)-enhanced magnetic resonance imaging studies of the neuronal pathways of thehypothalamus showed that information about the regulation of food intake and energy balancecirculate through specific hypothalamic nuclei. The dehydration-induced anorexia (DIA) modeldemonstrated to be appropriate for studying the hypothalamus with Mn2 +-enhanced magneticresonance imaging. Manganese is involved in the normal functioning of a variety of physiologicalprocesses and is associated with enzymes contributing to neurotransmitter synthesis andmetabolism. It also induces psychiatric and motor disturbances. The molecular mechanisms bywhich Mn2 + produces alterations of the hypothalamic physiological processes are not wellunderstood. 1H-magnetic resonance spectroscopy measurements of the rodent hypothalamus arechallenging due to the distant location of the hypothalamus resulting in limited measurementsensitivity. The present study proposed to investigate the effects of Mn2 + on the neurochemicalprofile of the hypothalamus in normal, DIA, and overnight fasted female rats at 14.1 T. Resultsprovide evidence that c-aminobutyric acid has an essential role in the maintenance of energyhomeostasis in the hypothalamus but is not condition specific. On the contrary, glutamine,glutamate, and taurine appear to respond more accurately to Mn2 + exposure. An increase inglutamine levels could also be a characteristic response of the hypothalamus to DIA.Journal of Cerebral Blood Flow & Metabolism (2011) 31, 2324–2333; doi:10.1038/jcbfm.2011.92; published online29 June 2011

Keywords: animal models; astrocytes; calcium channels; GABA; MR spectroscopy

Introduction

Manganese (Mn2 +) is an essential metal necessary forthe normal functioning of a variety of physiologicalprocesses: it has a role in immune system function-ing, regulation of cellular energy, and is cofactor for avariety of enzymes involved in neurotransmittersynthesis and metabolism (Aschner et al, 2007;Benedetto et al, 2009). However, it is also aneurotoxicant and induces psychiatric and motordisturbances as well as cognitive deficits such asmemory and information processing impairments,reduced learning capacity, cognitive slowing, anddecreased mental flexibility (Aschner et al, 2007;

Benedetto et al, 2009). The molecular mechanisms bywhich Mn2 + produces alterations of the hypothala-mic physiological processes are not well understood.

Recently, Mn2 +-enhanced magnetic resonanceimaging techniques demonstrated a growing poten-tial for structural imaging (Silva et al, 2008) and theidentification of neuronal pathways through tracttracing studies (Pautler, 2004). Manganese has thepotential to enter excitable cells through voltage-gated calcium channels and acts as a marker of brainactivation (Aoki et al, 2004). Previous Mn2 +-enhanced magnetic resonance imaging studies estab-lished that several hypothalamic nuclei (the para-ventricular nuclei of the hypothalamus (PVN), thelateral hypothalamus (LH), the ventromedial nucleusof the hypothalamus, and the arcuate nucleus) of ratsand mice show specific enhancement patternsfollowing systemic administration of MnCl2 asso-ciated with food-related challenges (Parkinson et al,2009; Chaudhri et al, 2006). These enhancementswere related to neuronal activity in the aforemen-tioned hypothalamic nuclei.

In relation to these novel and promising techni-ques, dosage and administration routes of Mn2 + havebeen investigated (Silva and Bock, 2008), since acute

Received 9 February 2011; revised 19 April 2011; accepted 3 June2011; published online 29 June 2011

Correspondence: Dr N Just, Laboratory for Functional andmetabolic Imaging (LIFMET), Centre d’Imagerie Biomedicale(CIBM), CH-1015 Lausanne, Switzerland.E-mail: [email protected]

This work was supported by the Centre d’Imagerie BioMedicale

(CIBM) of The University of Lausanne (UNIL), the Ecole Poly-

technique Federale de Lausanne (EPFL), of the Hopital Universi-

taire de Geneve (HUG), and the Centre Hospitalier Universitaire

Vaudois (CHUV) and Leenards and Jeantet foundations.

Journal of Cerebral Blood Flow & Metabolism (2011) 31, 2324–2333& 2011 ISCBFM All rights reserved 0271-678X/11 $32.00

www.jcbfm.com

http://dx.doi.org/10.1038/jcbfm.2011.92

mailto:[email protected]

http://www.jcbfm.com

exposure to high doses of Mn2 + can lead to hepaticfailure and cardiac toxicity (Silva and Bock, 2008;Silva et al, 2004). Only a few NMR (nuclear magneticresonance) spectroscopic studies were conducted toinvestigate the influence of Mn2 + on brain metabo-lism (Zwingmann et al, 2003, 2004, 2007). In ratmodels, Mn2 + affected both astrocytic and neuronalintegrity (Zwingmann et al, 2003, 2004, 2007).Moreover, Mn2 + neurotoxicity has been associatedwith changes in brain energy metabolism andstimulation of oxidative glucose metabolism as wellas protective astrocytic response (Zwingmann et al,2004). Manganese in brain is preferentially accumu-lated in astrocytes and is known to stimulate theproduction of reactive oxygen and nitrogen speciesresulting in oxidative/nitrosative stress (Zwingmannet al, 2004, 2007). Additionally, Mn2 + was shown toinduce cell swelling (Rama Rao et al, 2010). On thecontrary, a model of dehydration-induced anorexia(DIA) was shown to induce cell shrinkage due tohyperosmolar challenge in rats (Jaimes-Hoy et al,2008; Salter-Venzon and Watts, 2008, 2009).

Another benefit of using experimental models ofMn2 + exposure could therefore be the possibilityto investigate neuronal and astrocytic mechanismsof regulation of the metabolism.

In the present work, our aim was to investigatethe effects of Mn2 + on the proton neurochemicalprofile of the hypothalamus under disrupted energybalance conditions and following homeostaticchanges. The following questions were addressed:

� How does Mn2 + affect hypothalamic metabolism?What are the energetic effects of acute Mn2 +

exposure?� Can Mn2 + modify the metabolic response to a

specific challenge? Can this modification becharacterized?

The neurochemical profiles of the hypothalamusof healthy, dehydrated-induced anorexic (Jaimes-Hoy et al, 2008; Watts, 1999), and overnight foodsuppressed (OFS) rats were measured and examinedwith and without the influence of Mn2 + .

Materials and methods

Animal Preparation

All studies were performed following the approval ofService de la consommation et des affaires veterinaires ducanton de Vaud (Switzerland). Thirty female Wistar rats(225 to 250 g body weight on day 0 (D0)) were divided intothree groups: the control group (CTL, n = 10), the OFSgroup (n = 10), and the DIA group (n = 10). Magneticresonance (MR) experiments were conducted on one ratper group per session. Animals were acclimatized in ananimal facility under humidity and temperature-controlledconditions on a 12-hour light/dark cycle during 1 weekbefore the beginning of the experiments. Animals wereallowed ad libitum access to standard chow and drinkingwater during this period.

On day 0 (D0), animals were placed in individual cages.Animals from the CTL group and the OFS group receivedwater and food ad libitum; the dehydrated group (DIA)received 2.5% NaCl (Sigma, Zofingen, Switzerland) asdrinking solution (Jaimes-Hoy et al, 2008; Watts, 1999). OnD3 at 1800 hours, the OFS rats had their chow removeduntil the MR scan took place, 24 hours later. The amountsof food and drink were measured at 1800 hours on D0, D1,D3, D7, D10, and D14. The rats’ body weight was alsomeasured. On D3, rats from each group were anesthetizedwith 2% isoflurane in a mixture of O2 and N2O via afacemask. The tail vein was catheterized for intravenousinfusion of MnCl2. Each rat received an intravenousinfusion of 100 mg/kg at 100 mmol/L MnCl2 at a rate of0.5 mL to 1 mL/h via the tail vein to avoid cardiac arrest.Animals were all the time gently manipulated to avoidstress and suffering. During infusion and until animalswere completely recovered from anesthesia, their tempera-ture was carefully monitored and kept at 371C±11C.

Five female Wistar rats per group did not undergo MnCl2

infusions and were scanned on D4 after the start ofanorexic protocol (D0) for the DIA group of rats.

During the MR experiments, rats were anesthetized with2% isoflurane in a mixture of O2 and N2O. The rats werepositioned in a dedicated holder with head fixation. Theirbody temperature was maintained at 37.51C±0.51C using atemperature-controlled water circulation and a rectalprobe. Their respiration rate was monitored throughoutthe experiment. Magnetic resonance imaging and magneticresonance spectroscopy (MRS) were performed on D424 hours after MnCl2 infusion.

Proton Magnetic Resonance Spectroscopy in theHypothalamus and Hippocampus

Localized proton spectroscopy was performed using Spin-Echo full intensity Acquired Localized (SPECIAL) (Mly-narik et al, 2006), using echo time/repetition time (TE/TR) = 2.8/4,000 milliseconds in combination with outervolume suppression and vapor water suppression (Tkacet al, 1999). A voxel of interest of 50 mL was localized in thehippocampus and the PVN–LH region for all the rats usingthe previously acquired 3D-gradient echo (GRE) images.All first- and second-order shims were readjusted usingFast Automatic Shimming by Mapping along Projections(FASTMAP) (Gruetter and Tkac, 2000) in the voxel ofinterest, which resulted in water linewidth of 23±5 Hz. Toobtain adequate signal-to-noise ratio, 640 scans wereacquired. For absolute quantification, water signal wasacquired using identical parameters, except without watersuppression and averaging eight scans.

Quantification and Statistics

The in vivo 1H-MR spectra were processed using linearcombination (LC) model (Provencher, 1993). In this study,18 metabolite concentrations were quantified using data-bases of simulated spectra of metabolites (Mlynarik et al,2006). Only metabolites with Cramer–Rao lower boundsunder 20% were kept for further analysis.

Manganese toxicity in the hypothalamusN Just et al

2325

Journal of Cerebral Blood Flow & Metabolism (2011) 31, 2324–2333

All data are presented as mean±standard error of themean (s.e.m.) unless otherwise stated. Two-tailed unpairedStudent’s t-tests were applied to compare data acquired inthe three groups of rats. A P value < 0.05 is consideredsignificant.

Results

3D-gradient echo T1-weighted images obtained at14.1 T were used to position carefully the voxels ofinterest in the hippocampus and the hypothalamusof female rats. 3D images provided an increasedspatial resolution in the rat brain and voxels ofinterest were positioned with reference to thePaxinos atlas (Paxinos and Watson, 1998). Despitechallenges due to the depth of the hypothalamus inthe rat, good quality spectra were obtained using asurface radiofrequency (RF) coil at 14.1 T and relativeto the hippocampus (see Supplementary file). Eleven

metabolites were quantified reliably in the hypotha-lamus with Cramer–Rao lower bounds under 20%.The comparison of 1H-MRS spectra of the hypotha-lamus with spectra of the hippocampus in normalfemale rats demonstrated important differences suchas lower Glu, Tau, N-acetyl-aspartate (NAA), PCr, PCr+ Cr (tCr) concentrations in agreement with theprevious findings in mice (Lei et al, 2010). A lowerg-aminobutyric acid (GABA) peak was also found infemale rats. A comparison of 1H-MRS spectra of thehypothalamus with spectra of the hippocampus infemale rats subjected to dehydration by drinking ahypertonic saline solution demonstrated higherconcentrations of most metabolites in both thehippocampus and the hypothalamus (relative differ-ences in hypothalamus: DTau = + 106%, DtCr = +40%, Dmyo-Ins = + 92%, DGABA = + 160%; in hip-pocampus: DTau = + 30%, DtCr = + 50%, Dmyo-Ins =+ 72%). Interestingly, GABA was lower in thehippocampus of DIA rats (DGABA =�50%).

Figure 1 1H-nuclear magnetic resonance (NMR) spectra obtained at 14.1 T of the rat hypothalamus in control (CTL), dehydrated-induced anorexic (DIA), and overnight food-restricted (OFS) rats with metabolite assignments. Peak assignments: Cr, creatine; GABA,g-aminobutyric acid: Gln, glutamine; Glu, glutamate; Lac, lactate; myo-Ins, myo-Inositol; NAA, N-acetyl-aspartate; Tau, taurine. Thevoxels of interest are overlaid on 3D-T1-weighted gradient echo images (TR/TE = 20/5 milliseconds, flip angle = 701, field ofview = 25�25�25 mm3, matrix size = 256�256�128, coronal slices, BW = 25 KHz, five averages). Spectra were alsoobtained 24 hours after an intravenous injection of 100 mg/kg MnCl2 in CTL (MnCl2), DIA (MnCl2), and OFS (MnCl2) rats. Coloredarrows indicate increases or decreases of the metabolites of interest relative to the CTL spectrum: red for GABA and lactate; green formyo-Ins; and violet for Tau. The color reproduction of this figure is available at the Journal of Cerebral Blood Flow and Metabolismjournal online. BW, bandwidth; TE, echo time; TR, repetition time.


2326


1H-MRS spectra are presented in Figure 1 for eachcondition CTL, DIA, and OFS with and withoutMnCl2. Twenty-four hours after infusion of MnCl2

in the rat tail vein, CTL rats demonstrated asignificant increase in GABA and Lac concentrations(t-test, P = 0.01) compared with CTL rats with noMnCl2 infusion (Figure 2A). MnCl2-infused DIA ratsdemonstrated no significant difference with non-MnCl2-infused DIA rats except for Tau (t-test,P = 0.01) (Figure 2B). Under DIA condition, allmetabolite concentrations increased relative to theCTL condition. However, no significant differenceswere detected between CTL and DIA rats in thepresence of MnCl2. In particular, the relative differ-ence for GABA between CTL and DIA was only 3.3%.To verify that changes seen between MnCl2 andnon-MnCl2-infused DIA rats were solely related toosmotic changes due to hypertonic saline in therat hypothalamus, a third group of rats was intro-duced. These rats had their food suppressedovernight (OFS). Five of them received an infusionof MnCl2. Their hypothalamic neurochemical pro-

files 24 hours post-MnCl2 infusion showed lowermetabolite concentrations compared with OFS ratswithout MnCl2 (DGlu =�28%, t-test, P = 0.03;DTau =�55%, t-test, P = 0.04; DNAA, DCrB�40%but P > 0.05) (Figure 2C). Glc was undetectable andGABA and myo-Ins levels increased, although notsignificantly, relative to OFS rats without MnCl2.In MnCl2-infused OFS rats, metabolite concentra-tions had a tendency to decrease relative to theirCTL (DGABA =�6%, DTau =�17%, DGlu =�18%,DLac =�40%), whereas most metabolite concentra-tions increased significantly in OFS rats withoutMnCl2 relative to their CTL. Figure 3 comparesmetabolite concentrations for CTL, DIA, and OFSrats (Figure 3A) and for CTL ( + Mn2 +), DIA ( + Mn2 +),and OFS ( + Mn2 +) (Figure 3B). While significantdifferences were found between the neurochemicalprofile of CTL and the neurochemical profiles ofDIA and OFS, these differences appear to be maskedin the presence of Mn2 + .

The relative concentration changes (percentageof CTL) are presented in Figure 4. Significant

Figure 2 Effect of manganese (Mn2 +) on the neurochemical profile of control (CTL), dehydration-induced anorexia (DIA), andovernight food suppressed (OFS) rats. (A) Comparison between absolute concentrations in CTL and CTL ( + MnCl2). (B) Comparisonbetween absolute concentrations in DIA and DIA ( + MnCl2). DIA rats were examined by 1H-nuclear magnetic resonance (NMR)spectroscopy 4 days after they started drinking a 2.5% NaCl solution (spontaneous drinking) and 24 hours postintravenous infusionof 100 mg/kg MnCl2. (C) Comparison between absolute concentrations in OFS and OFS ( + MnCl2). Rats were fasted 24 hours before1H-NMR spectroscopy. *P < 0.05.


2327


differences were found between DIA and OFS(two-tailed t-test; Gln, Glu, GABA: P < 0.0001; Lac:P = 0.0048; NAA: P = 0.015; Tau: P < 0.05). Significantdifferences also exist between CTL ( + Mn2 +), OFS(Mn2 +), and DIA ( + Mn2 +) relative to the CTL group(two-tailed t-test between CTL ( + Mn2 +) and DIA( + Mn2 +): PCr, Gln: P < 0.0001; between OFS (Mn2 +)and DIA ( + Mn2 +): Gln, Glu, Lac: P < 0.0001).Statistically significant differences between groupsof rats exposed to Mn2 + and rats nonexposed are alsodepicted in Figure 4. Although averaging groups ofrats challenged by different conditions appear arbi-trary, the ‘gross’ effect of Mn2 + on the neurochemicalprofile of the hypothalamus is clearly shown for Glnin DIA, Glu, and NAA in OFS and Tau for allconditions.

Discussion

Manganese is an essential element for normal rodentand human development. It is required for theappropriate functioning of a variety of physiologicalprocesses (Aschner et al, 2007; Benedetto et al,2009). Recently, magnetic resonance imaging techni-ques have taken advantage of the paramagneticproperties of this compound identified as a potentialindirect noninvasive neuronal marker (Duong et al,

2000) and allowing the identification of variousneuronal pathways within the rodent brain in tracttracing studies (Pautler et al, 1998, 2004). However,excessive levels of Mn2 + are toxic for the centralnervous system (Stanwood et al, 2009). Silva et al(2004) have reported behavioral disorders such assomnolence, convulsions, tremors, and respiratoryeffects in both rats and mice at high Mn2 + doses( > 90 mg/kg) independent of the route of administra-tion. Symptoms similar to Parkinson’s disease havealso been reported (Benedetto et al, 2009; Olanow,2004; Pal et al, 1999). Moreover, Mn2 + was alsoreported to disrupt the astrocytes glutamine trans-porter expression and function in astrocytic cellcultures (Sidoryk-Wegrzynowicz et al, 2009) and toalter dopaminergic and GABAergic neurons withinthe basal ganglia (Stanwood et al, 2009). Althoughthe mechanisms underlying Mn2 + neurotoxicityremain unclear and chronic exposure to Mn2 + is atopic of interest since neuropathologic changes inhumans lead to neuronal loss, gliosis and Alzheimerdisease, the present study examined the effects ofacute exposure to Mn2 + in relation to previous Mn2 +-enhanced magnetic resonance imaging studies in thehypothalamus.

The hypothalamus is responsible for metabolicprocesses of the autonomic nervous system. Thisextremely complex brain region coordinates many

Figure 3 Comparison between control (CTL), dehydration-induced anorexia (DIA), and overnight food suppressed (OFS)rats. (A) Without manganese. z Represents significance level(P < 0.05) between CTL and DIA groups. w Representssignificance level (P < 0.05) between CTL and OFS groups.(B) In all, 24 hours postintravenous infusion of manganese. OnlyGlutamine (Gln) demonstrated a significant difference comparedwith CTL and OFS (zP < 0.01). While significant differencesexist between the neurochemical profile of CTL and theneurochemical profiles of DIA and OFS, these differences appearto be masked in the presence of manganese.

Figure 4 Effect of manganese (Mn2 +) on metabolites of thehypothalamus under normal (control, CTL), dehydration-inducedanorexia (DIA), and overnight fasted (OFS) conditions. Thecalculated mean percentages of the CTL without manganese(±s.d.) are represented. Significant differences were foundbetween DIA and OFS (*two-tailed t-test; Gln, Glu, g-amino-butyric acid (GABA): P < 0.0001; Lac: P = 0.0048; NAA:P = 0.015; Tau: P < 0.05). Significant differences also existbetween CTL ( + Mn2 +), OFS (Mn2 +), and DIA ( + Mn2 +)relative to the CTL group (two-tailed t-test zbetween CTL( + Mn2 +) and DIA ( + Mn2 +): PCr, Gln: P < 0.0001; $;between OFS (Mn2 +) and DIA ( + Mn2 +): Gln, Glu, Lac:P < 0.0001). Statistically significant differences between groupsof rats exposed to Mn2 + and rats nonexposed are also depicted(++P < 0.05).


2328


behavioral and physiological mechanisms as well ascomplex homeostatic processes. It must respond tomany different signals internally and externally gener-ated. Moreover, the hypothalamus is richly connectedwith the central nervous system. For an appropriateunderstanding of the function of hypothalamus, wechallenged it with dehydration and fasting andinvestigated its metabolic response to Mn2 + exposure.

Comparison of the Neurochemical Profiles of theHypothalamus and the Hippocampus of the Rat

The neurochemical profile of the normal rathypothalamus was characterized by low PCr, Glu,Tau, and NAA concentrations relative to those of thehippocampus in agreement with a previous study inmice at 14.1 T (Lei et al, 2010). However, GABA andmyo-Ins concentrations were also lower relative tothe rat hippocampus, whereas these metabolites hadsignificantly higher levels in the mouse hypothala-mus relative to the hippocampus (Lei et al, 2010).g-Aminobutyric acid is known to have an importantinhibitory role in the regulation of the hypothala-mic–pituitary–adrenocortical axis (Cullinan et al,2008) and especially at the level of the paraventri-cular nucleus (Cullinan et al, 2008). No GABAergicneurons are located within the PVN itself (Cullinanet al, 2008), which may explain the lower GABAlevels measured compared with the hippocampus inhealthy rats.

In response to changes in brain homeostasis,metabolite concentrations in the hippocampus andthe hypothalamus increased. A significant increaseof GABA in DIA rats relative to CTL rats but alsorelative to the hippocampus was noticed.

In response to changes in osmolality, a cell shrinksor swells and finally exhibits a regulatory volumeincrease or decrease. On hypernatremic challenge, thecell loses water until intracellular and extracellularosmolality are equal. The recovery of water is mediatedthrough accumulation of inorganic (Na+ , K+, Cl�) andorganic (glutamate, glutamine, taurine, myo-Inositol)osmolytes. Gln, Glu, Tau, GABA, myo-Ins rise in thebrain of chronic hypernatremic rats (Sterns and Silver,2007) and return to their baseline levels as soon asnormal osmolar conditions are recovered.

A reduction in brain water due to dehydrationwould be expected to decrease brain metaboliteconcentrations, 4 days after initiation of DIA.Dehydration-induced anorexia rats lost B15% oftheir initial weight (Just and Gruetter, 2011). If it isassumed that the entire weight loss can be attributedto dehydration then a 15% total body water loss canbe assumed and metabolite concentrations can bedownscaled to 15%, which might be overestimatedin the brain, since cellular dehydration is compen-sated by the continuous ingestion of water loadedwith sodium and volume regulation occurs.

Strikingly, levels of NAA also increased in DIA(P > 0.05) and OFS (P < 0.05) rats. N-acetyl-aspartate

is believed to be a marker of neuronal density,integrity, and function. N-acetyl-aspartate reductionshave been interpreted as reflecting neurotoxicity,neurodegeneration, or impaired mitochondrialenergy metabolism. Consistent with this hypothesis,levels of NAA are reduced in a range of pathologicstates with dramatic reductions in patients withtumor or stroke (Zimmerman and Wang, 1997).However, NAA has been identified as an osmolyte(Baslow, 2003). N-acetyl-aspartate is liberated intoextracellular space in response to changing osmoticconditions but NAA was also hypothesized to be amolecular pump in which translocation creates anosmotic environment that passively removes waterfrom the intracellular milieu of neurons (Baslow,2003; Steen and Ogg, 2005).

Studies of psychiatric disorders have also shownthat NAA levels can increase, suggesting a restora-tion or repair of neuronal viability (Moore et al, 2000;Mickael et al, 2003). The administration of lithiumchloride to humans increased whole brain NAAlevels, suggesting an increase of neuronal viability.However, lithium chloride also induces anorecticeffects (Parkinson et al, 2009; Ishii et al, 2004). Therewere no significant difference in Mn2 + uptakes inhypothalamic regions involved in the control offood intake (arcuate nucleus, PVN, and supraopticnucleus) between overnight fasted mice and micereceiving lithium chloride (Parkinson et al, 2009).The upregulation of NAA in OFS rats could beattributed to a transient neuroprotective effect ofNAA in neurons in response to a short fasting period.This hypothesis is consistent with studies, demon-strating that fasting induces neuroprotective effects(Davis et al, 2008; Chiba et al, 2007).

c-Aminobutyric Acid Responses to Hypertonicity andManganese

Without MnCl2, the hypothalamic inhibitory neuro-transmitter GABA increased significantly in bothDIA and OFS rats relative to CTL rats. g-Aminobu-tyric acid is known to suppress feeding in rats andimpact on body weight loss (Turenius et al, 2009).The majority of stress-responsive corticotropin-releasing hormone (CRH) neuronal inputs to thePVN contain the inhibitory neurotransmitter GABA,highlighting the importance of GABAergic neuro-transmission in the regulation of the neuroendocrinestress response (Cullinan et al, 2008). Neurons ofthe PVN synthesize thyrotropin-releasing hormoneand CRH hormones having an important role inenergy homeostasis (Jaimes-Hoy et al, 2008). Underrestricted food regime, rats show decreased thyro-tropin-releasing hormone expression in the PVNwhile anorexic animals show increased thyrotro-pin-releasing hormone expression in the PVN. How-ever, a reduction of CRH expression compared withCTL animals (Jaimes-Hoy et al, 2008) in the PVN wasfound in both food-restricted and DIA animals but to


2329


a greater extent in DIA rats. This finding parallels theincrease of GABA seen in DIA and OFS rats but to agreater extent in DIA rats. Yet, CRH-R2, one of thereceptors of CRH signal involved in anorexigeniceffects (de Gortari et al, 2009) was downregulated inthe PVN of DIA rats but not in food-restrictedanimals (de Gortari et al, 2009). GABAergic projec-tions from regions surrounding the PVN (anteriorhypothalamic area, dorsomedial hypothalamicnucleus, LH, medial preoptic area) convey inhibitorysignals to CRH neurons in response to a stressor suchas DIA and food deficiency and limit hypothalamicactivity by an important braking mechanism thatincludes increase of GABA release at the level of thePVN (Cullinan et al, 2008). It was suggested (Watts,1999) that CRH neurons of the LH are specificallyactivated by the dehydration-induced stimulus andproject to neurons in the PVN so that food inhibitiondue to low energy intake is overridden by dehydra-tion mechanisms.

It is known that infertility can result when foodintake fails to compensate for increased energydemands such as in anorexia nervosa. MnCl2

stimulates the release of luteinizing hormone releas-ing hormone from the hypothalamus but simulta-neously increases GABA release that in turn inhibitsluteinizing hormone releasing hormone release(Prestifilippo et al, 2008). These processes alsoinvolve the activity of nitric oxide synthase (Presti-filippo et al, 2008) and may ultimately contribute tosuppressed reproductive function in both male andfemale rats (Prestifilippo et al, 2008). As a conse-quence, further increases of GABA levels would havebeen expected in rats under both anorexic and Mn2 +

condition. Nonetheless, GABA levels were similar inCTL ( + Mn2 +), DIA ( + Mn2 +), and OFS ( + Mn2 +) rats.Moreover, GABA concentrations were not signifi-cantly increased in DIA ( + Mn2 +) and OFS ( + Mn2 +)rats relative to their corresponding DIA and OFS rats.This suggests that the effects of dehydration and foodrestriction and MnCl2 are not summed. MnCl2

appears to blur the effects of anorexia and foodrestriction on GABA. Dehydration-induced anorexiaanimals demonstrated similar metabolite levels withand without MnCl2 except for Tau. As pointed outearlier, both challenges should induce GABA releasein the PVN but GABA levels were similar in DIA andDIA ( + Mn2 +) rats, suggesting that the effects of onecondition may override the effects of the other. Saltloading and Mn2 + induce opposing effects at thecellular level. When exposed to hypertonicity, a cellshrinks until intracellular and extracellular osmolal-ities are equal. In contrast, Mn2 + induces cellswelling (Rama Rao et al, 2010). Cell swelling andcell shrinkage have opposing effects on Ca2 + con-centration and hormonal secretion (Ben-TabouDe-Leon et al, 2003, 2006). Due to dehydration,morphological plasticity of astrocytes in the PVNoccurred. Thus, synaptic and extrasynaptic transmis-sion in magnocellular neurons of the PVN weremodified with increased synaptic release of gluta-

mate and GABA but reduced ability of Tau to bind toglycinergic receptors and modulate neuronal excit-ability (Oliet and Bonfardin, 2010).

Previous investigations also suggest that signalingof astrocytes to neurons is performed through Ca2 +-dependent release of glutamate. We suggest thatunder hypertonic conditions, the Mn2 + enhance-ment seen in the PVN (and corresponding toneuronal activation) is mediated through astrocyticsignaling. This hypothesis appears consistent within vitro findings, showing that astrocytes increasedtheir oxidative and anaerobic glucose metabolism toprotect against energy failure and oxidative stressinduced by Mn2 + . Simultaneously, Mn2 + inducedneuronal energy failure (Zwingmann et al, 2003).For the maintenance of neuronal energy and neuro-transmitter metabolism, the hypothesis of an astro-cyte–neuron ‘lactate shuttle’ (Pellerin et al, 2007;Zwingmann et al, 2003) might hold.

g-Aminobutyric acid levels increased significantlyfor all the conditions explored in the presentstudy including with Mn2 + . Hence, GABA appearsto be a marker of homeostatic changes within the rathypothalamus but it cannot be considered as aspecific marker.

Spectroscopic Analysis of the Hypothalamus EnergyMetabolism in Manganese Neurotoxicity

The characteristic and significant GABA increases infasted animals (Violante et al, 2009) and MnCl2-infused animals (Zwingmann et al, 2004, 2007)relative to normal animals were already describedfollowing the measurements of brain extracts usingCarbon-13 (13C)-NMR spectroscopy. These metabolicchanges in the glutamine–glutamate–GABA cyclewere attributed to impaired cerebral energy meta-bolism. This hypothesis was further confirmed sinceboth fasted and MnCl2-infused animals demon-strated significantly increased lactate levelsmeasured ex vivo relative to normal animals. Lactatesynthesis is known to counteract decreased mito-chondrial energy production (Zwingmann et al,2004, 2007). Lactate also increased in our in vivostudy for CTL ( + Mn2 +) rats but not for OFS ( + Mn2 +)rats. Violante et al (2009) attributed the increasedGABA levels of fasted mice to a reduction in theGABA degradation pathway since no significantincreases in Glu and Gln were detected. WithoutMnCl2, Glu levels increased significantly in our OFSrats relative to CTL. This would point towardincreased synthesis of GABA through Glu and Glnprecursors while in OFS ( + Mn2 +) rats, no significantdifference was detected relative to CTL, suggesting adecreased degradation of GABA (Violante et al,2009). However, as pointed out earlier, increasinglevels of GABA did not appear to be condition specificincluding in the presence of MnCl2 (Figure 4). Onthe contrary, Mn2 + clearly influenced Tau, NAA, Cr,and PCr as well Glc levels relative to the CTL groupof rats. Moreover, Gln and Glu levels appeared more


2330


specifically affected by Mn2 + for DIA ( + Mn2 +) andOFS ( + Mn2 +) rats. Zwingmann et al (2004, 2007)showed that Mn2 + disturbs glucose metabolism,which is supported by the results of the presentstudy. Additionally, increases in glutamine levelsfollowing exposure of cells to Mn2 + were attributedto an impaired efflux of Gln from astrocytes or animpaired uptake of Gln from neurons (Zwingmann etal, 2003). These results were consistent with aincreased glutamine synthetase activity in Mn2 +-intoxicated rats (Zheng et al, 1999) and match as wellwith findings that simultaneous increases in Glnlevels with decreases in Tau levels are consistentwith alterations in astrocyte morphology (Oliet andBonfardin, 2010). In DIA ( + Mn2 +), Gln increaseswere further exacerbated compared with all the othergroups of rats and relative to CTL, but we also notedthat a significant difference between Gln levels inDIA and OFS rats remains with and without Mn2 + . Infuture studies, the role of glutamine in hypothalamicastrocytes and neurons of anorexic rats should befurther investigated.

Significantly lower Tau levels were noticed forboth OFS ( + Mn2 +) and DIA ( + Mn2 +) rats relative tonon-MnCl2-infused rats. This change was not noticedin CTL rats. Relative to CTL rats, Tau levels did notchange significantly for the OFS ( + Mn2 +) and DIA( + Mn2 +) conditions, whereas higher Tau levels weremeasured in OFS and DIA rats without MnCl2.

Taurine is an amino acid that has neuroprotectiveproperties and is involved as an osmoregulator (Songand Hatton, 2003). Taurine uptake was increased inastrocyte cultures exposed to a hyperosmotic condi-tion (Aschner et al, 2001), while Tau variations havealready been observed in short-term food-deprivedanimals (Benuck et al, 1995). As in the present study,Tau concentrations in the globus pallidus of ratsmeasured in vivo by NMR spectroscopy (Zwingmannet al, 2007) after MnCl2 infusion demonstratedsignificant decreases compared with control rats.Taurine is also an important neurotransmitter knownto modulate GABA neurochemistry (Namima et al,1982, 1983) and share GABA transporters in thebrain. Here, due to Mn2 + exposure taurine respondedinversely to GABA. Previous data suggested thatMn2 + increases GABA by inhibition of GABAtransporter function. In turn, the osmoregulatoryand neuromodulatory functions of taurine are alteredand perturb neural function (Fordahl et al, 2010).

Myo-Inositol concentrations were also increasedwith MnCl2 in the globus pallidus compared withcontrols without MnCl2 whereas NAA levels weredecreased (Zwingmann et al, 2007). Moreover, thesechanges appear to be regional specific (Zwingmannet al, 2007).

Conclusion

In the present work, the metabolism of the hypotha-lamus in rats was characterized both under normal

conditions and following changes in brain energyhomeostasis using a model of DIA and overnightfasting in rats. The data presented in this study showthat it is necessary to perform region-specific1H-MRS for an accurate characterization of theneurochemical profile of the hypothalamus. Inaddition, compared with previous studies (Lei et al,2010), it shows that an improved characterization isperformed when challenges such as dehydration,fasting, and Mn2 + infusion are applied: the data ofthe present study provide evidence that GABA hasan essential role in the maintenance of energyhomeostasis in the hypothalamus. However, resultsalso reveal that GABA is a nonspecific marker ofMn2 + intoxication in the hypothalamus. On thecontrary, glutamine, glutamate, and taurine appearto respond more accurately to Mn2 + exposure. Anincrease in glutamine levels could be a characteristicresponse of the hypothalamus to DIA.

Disclosure/conflict of interest

The authors declare no conflict of interest.

References

Aoki I, Naruse S, Tanaka C (2004) Manganese-enhancedmagnetic resonance imaging (MEMRI) of brain activityand applications to early detection of brain ischemia.NMR Biomed 17:569–80

Aschner M, Guilarte TR, Schneider JS, Zheng W (2007)Manganese: recent advances in understanding its trans-port and neurotoxicity. Toxicol Appl Pharmacol 221:131–47

Aschner M, Mutkus L, Allen JW (2001) Aspartate andglutamate transport in acutely and chronically ethanolexposed neonatal rat primary astrocyte cultures. Neuro-toxicology 22:601–5

Baslow MH (2003) N-acetylaspartate in the vertebratebrain: metabolism and function. Neurochem Res 28:941–53

Benedetto A, Au C, Aschner M (2009) Manganese-induceddopaminergic neurodegeneration: insights into mechan-isms and genetics shared with Parkinson’s disease.Chem Rev 109:4862–84

Benuck M, Banay-Schwartz M, DeGuzman T, Lajtha A(1995) Effect of food deprivation on glutathione andamino acid levels in brain and liver of young and agedrats. Brain Res 678:259–64

Ben-Tabou De-Leon S, Ben-Zeev G, Nussinovitch I (2006)Effects of osmotic shrinkage on voltage-gated Ca2+

channel currents in rat anterior pituitary cells. Am JPhysiol Cell Physiol 290:C222–32

Ben-Tabou De-Leon S, Blotnick E, Nussinovitch I (2003)Effects of osmotic swelling on voltage-gated calciumchannel currents in rat anterior pituitary cells. Am JPhysiol Cell Physiol 285:C840–52

Chaudhri OB, Parkinson JR, Kuo YT, Druce MR, HerlihyAH, Bell JD, Dhillo WS, Stanley SA, Ghatei MA, BloomSR (2006) Differential hypothalamic neuronal activationfollowing peripheral injection of GLP-1 and oxynto-modulin in mice detected by manganese-enhanced


2331


magnetic resonance imaging. Biochem Biophys ResCommun 350:298–306

Chiba T, Fujita S, Kubota H, Inoue D, Mizuno A, KomatsuT, Yamaza H, Higami Y, Shimokawa I (2007) Identifica-tion of fasting-induced genes in the rat hypothalamus:relationship with neuroprotection. Ann NY Acad Sci1119:216–26

Cullinan WE, Ziegler DR, Herman JP (2008) Functional roleof local GABAergic influences on the HPA axis. BrainStruct Funct 213:63–72

de Gortari P, Mancera K, Cote-Velez A, Amaya MI, MartınezA, Jaimes-Hoy L, Joseph-Bravo P (2009) Involvement ofCRH-R2 receptor in eating behavior and in the responseof the HPT axis in rats subjected to dehydration-inducedanorexia. Psychoneuroendocrinology 34:259–72

Davis LM, Pauly JR, Readnower RD, Rho JM, Sullivan PG(2008) Fasting is neuroprotective following traumaticbrain injury. J Neurosci Res 86:1812–22

Duong TQ, Silva AC, Lee SP, Kim SG (2000) FunctionalMRI of calcium-dependent synaptic activity: crosscorrelation with CBF and BOLD measurements. MagnReson Med 43:383–92

Fordahl SC, Anderson JG, Cooney PT, Weaver TL,Colyer CL, Erikson KM (2010) Manganese exposureinhibits the clearance of extracellular GABA andinfluences taurine homeostasis in the striatum ofdeveloping rats. Neurotoxicology 31:639–46

Gruetter R, Tkac I (2000) Field mapping without referencescan using asymmetric echo-planar techniques. MagnReson Med 43:319–23

Ishii Y, Blundell JE, Halford JC, Upton N, Porter R, Johns A,Rodgers RJ (2004) Differential effects of the selectiveorexin-1 receptor antagonist SB-334867 and lithiumchloride on the behavioural satiety sequence in rats.Physiol Behav 81:129–40

Jaimes-Hoy L, Joseph-Bravo P, de Gortari P (2008)Differential response of TRHergic neurons of thehypothalamic paraventricular nucleus (PVN) in femaleanimals submitted to food-restriction or dehydration-induced anorexia and cold exposure. Horm Behav53:366–77

Just N, Gruetter R (2011) Detection of neuronal activity andmetabolism in a model of dehydration-induced ano-rexia in rats at 14.1T using manganese-enhancedMRI and proton MR spectroscopy. NMR Biomed;published online, 4 May 2011; doi:10.1002/nbm.1694

Lei H, Poitry-Yamate C, Preitner F, Thorens B, Gruetter R(2010) Neurochemical profile of the mouse hypothala-mus using in vivo (1) H MRS at 14.1T. NMR Biomed23:578–83

Mickael N, Erfurth A, Ohrmann P, Arolt V, Heindel W,Pfeiderer B (2003) Neurotrophic effects of electroconvul-sive therapy: a proton magnetic resonance study of the leftamygdala region in patients with treatment-resistantdepression. Neuropsychopharmacology 28:720–5

Mlynarik V, Gambarota G, Frenkel H, Gruetter R (2006)Localized short-echo-time proton MR spectroscopy withfull signal-intensity acquisition. Magn Reson Med 56:s965–70

Moore GJ, Bebchuk JM, Hasanat K, Chen G, Seraji-Bozorgzad N, Wilds IB, Faulk MW, Koch S, Glitz DA,Jolkovsky L, Manji HK (2000) Lithium increasesN-acetyl-aspartate in the human brain: in vivoevidence in support of bcl-2’s neurotrophic effects?Biol Psychiatry 48:1–8

Namima M, Okamoto K, Sakai Y (1982) Taurine acts onpresynaptic autoreceptors for GABA in the cerebellum:

effects on Ca2+ influx and GABA release. Jpn JPharmacol 32:746–9

Namima M, Okamoto K, Sakai Y (1983) Modulatory actionof taurine on the release of GABA in cerebellar slices ofthe guinea pig. J Neurochem 40:1–9

Olanow CW (2004) Manganese-induced Parkinsonism andParkinson’s disease. Ann NY Acad Sci 1012:209–23

Oliet S, Bonfardin V (2010) Morphological plasticity of therat supraoptic nucleus-cellular consequences. Eur JNeurosci 32:1989–94

Pal PK, Samii A, Calne DB (1999) Manganese neurotoxi-city: a review of clinical features, imaging and pathol-ogy. Neurotoxicology 20:227–38

Parkinson JR, Chaudhri OB, Kuo YT, Field BCT, HerlihyAH, Dhillo WS, Ghatei MA, Bloom SR, Bell JD (2009)Differential patterns of neuronal activation in thebrainstem and hypothalamus following peripheralinjection of GLP-1, oxyntomodulin and lithiumchloride in mice detected by manganese-enhancedmagnetic resonance imaging (MEMRI). Neuroimage44:1022–31

Pautler RG (2004) In vivo, trans-synaptic tract-tracingutilizing manganese-enhanced magnetic resonance ima-ging (MEMRI). NMR Biomed 17:595–601

Pautler RG, Silva AC, Koretsky AP (1998) In vivo neuronaltract tracing using manganese-enhanced magnetic reso-nance imaging. Magn Reson Med 40:740–8

Paxinos G, Watson C (1998) The Rat Brain in StereotaxicCoordinates. San Diego: Academic Press

Pellerin L, Bouzier-Sore AK, Aubert A, Serres S, Merle M,Costalat R, Magistretti PJ (2007) Activity-dependentregulation of energy metabolism by astrocytes: anupdate. Glia 55:1251–62

Prestifilippo JP, Fernandez-Solari J, De Laurentiis A, MohnCE, de la Cal C, Reynoso R, Dees WL, Rettori V (2008)Acute effect of manganese on hypothalamic luteinizinghormone releasing hormone secretion in adult male rats:involvement of specific neurotransmitter systems. Tox-icol Sci 105:295–302

Provencher SW (1993) Estimation of metabolite concentra-tions from localized in-vivo proton MR spectra. MagnReson Med 30:672–9

Rama Rao K, Jayakumar AR, Reddy PVB, Tong X, CurtisKM, Noremberg MD (2010) Aquaporin-4 in manganese-treated cultured astrocytes. Glia 58:1490–9

Salter-venzon D, Watts AG (2008) The role of hypothalamicingestive behavior controllers in generating dehydrationanorexia: a Fos mapping study. Am J Physiol RegulIntegr Comp Physiol 295:R1009–19

Salter-Venzon D, Watts AG (2009) Site-specific attenuationof food intake but not the latency to eat after hypotha-lamic injections of neuropeptide Y in dehydrated-anorexic rats. Am J Physiol Regul Integr Comp Physiol297:R1813–21

Sidoryk-Wegrzynowicz M, Lee E, Albrecht J, Aschner M(2009) Manganese disrupts astrocyte glutamine trans-porter expression and function. J Neurochem 110:822–30

Silva AC, Bock NA (2008) Manganese-enhanced MRI: anexceptional tool in translational neuroimaging. Schi-zophr Bull 34:595–604

Silva AC, Lee JH, Aoki I, Koretsky AP (2004) Manganese-enhanced magnetic resonance imaging (MEMRI): meth-odological and practical considerations. NMR Biomed17:532–43

Silva AC, Lee JH, Wu CW, Tucciarone J, Pelled G, Aoki I,Koretsky AP (2008) Detection of cortical laminar


2332


architecture using manganese-enhanced MRI.J Neurosci Methods 167:246–57

Song Z, Hatton GI (2003) Taurine and the control of basalhormone release from rat neurohypophysis. Exp Neurol183:330–7

Stanwood GD, Leitch DB, Savchenko V, Wu J, FitsanakisVA, Anderson DJ, Stankowski JN, Aschner M, McLaugh-lin B (2009) Manganese exposure is cytotoxic and altersdopaminergic and GABAergic neurons within the basalganglia. J Neurochem 110:378–89

Steen RG, Ogg RJ (2005) Abnormally high levels of brain N-acetylaspartate in children with sickle cell disease.AJNR Am J Neuroradiol 26:463–8

Sterns RH, Silver SM (2007) Physiology and pathophysiol-ogy: hypenatremic states. In Seldin and Giebisch’s TheKidney (Alpern RJ, Hebert SC, eds). Academic Press:London, pp. 1203–1224

Tkac I, Starcuk Z, Choi IY, Gruetter R (1999) In vivo 1HNMR spectroscopy of rat brain at 1 ms echo time.Magn Reson Med 41:649–56

Turenius CI, Htut MM, Prodon DA, Ebersole PL, Ngo PT, LaraRN, Wilczynski JL, Stanley BG (2009) GABA(A) receptorsin the lateral hypothalamus as mediators of satiety andbody weight regulation. Brain Res 1262:16–24

Violante IR, Anastasovska J, Sanchez-Canon GJ, RodriguesTB, Righi V, Nieto-Charques L, Parkinson JR, Bloom SR,

Bell JD, Cerdan S (2009) Cerebral activation by fastinginduces lactate accumulation in the hypothalamus.Magn Reson Med 62:279–83

Watts AG (1999) Dehydration-associated anorexia: develop-ment and rapid reversal. Physiol Behav 65:871–8

Zheng W, Zhao Q, Slavkovich V, Aschner M, Graziano JH(1999) Alterations if ion homeostasis followingchronic exposure to manganese in rats. Brain Res833:125–32

Zimmerman RA, Wang ZJ (1997) The value of proton MRspectroscopy in pediatric metabolic brain disease. AJNRAm J Neuroradiol 18:1872–9

Zwingmann C, Leibfritz D, Hazell AS (2003) Energymetabolism in astrocytes and neurons treated withmanganese: relation among cell-specific energy failure,glucose metabolism, and intercellular trafficking usingmultinuclear NMR-spectroscopic analysis. J CerebBlood Flow Metab 23:756–71

Zwingmann C, Leibfritz D, Hazell AS (2004) Brainenergy metabolism in a sub-acute rat model of manga-nese neurotoxicity: an ex vivo nuclear magnetic reso-nance study using [1-13C] glucose. Neurotoxicology25:573–87

Zwingmann C, Leibfritz D, Hazell AS (2007) NMR spectro-scopic analysis of regional brain energy metabolism inmanganese neurotoxicity. Glia: 55:1610–7

Supplementary Information accompanies the paper on the Journal of Cerebral Blood Flow & Metabolism website(http://www.nature.com/jcbfm)


2333


Documents

Effect of manganese chloride on the neurochemical profile of the rat hypothalamus