Molecular and Cellular Endocrinology 375 (2013) 14–26

Contents lists available at SciVerse ScienceDirect

Molecular and Cellular Endocrinology

journal homepage: www.elsevier .com/locate /mce

Congenital hypothyroidism alters the oxidative status, enzyme activitiesand morphological parameters in the hippocampus of developing rats

0303-7207/$ - see front matter � 2013 Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.mce.2013.05.001

⇑ Corresponding author. Address: Departamento de Bioquímica, Centro deCiências Biológicas, UFSC, Campus Universitário, Caixa Postal 5069, Bairro Trindade,CEP 88040-970 – Florianópolis, Santa Catarina, Brazil. Tel.: +55 48 3721 4747; fax:+55 48 3721 9672.

E-mail addresses: arianezamoner@yahoo.com, ariane.zamoner@ufsc.br (A. Za-moner).

Daiane Cattani a, Paola Bez Goulart a, Vera Lúcia de Liz Oliveira Cavalli a, Elisa Winkelmann-Duarte b,André Quincozes dos Santos c, Paula Pierozan c, Daniela Fraga de Souza c, Viviane Mara Woehl b,Marilda C. Fernandes d, Fátima Regina Mena Barreto Silva a, Carlos Alberto Gonçalves c,Regina Pessoa-Pureur c, Ariane Zamoner a,⇑a Departamento de Bioquímica, Universidade Federal de Santa Catarina, Florianópolis, SC, Brazilb Departamento de Ciências Morfológicas, Universidade Federal de Santa Catarina, Florianópolis, Brazilc Departamento de Bioquímica, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazild Laboratório de Pesquisa em Patologia, Universidade Federal de Ciências da Saúde de Porto Alegre, Porto Alegre, Brazil

a r t i c l e i n f o

Article history:Received 27 August 2012Received in revised form 17 April 2013Accepted 1 May 2013Available online 18 May 2013

Keywords:HypothyroidismGlutamateCytoskeletonGlutamine synthetaseCalciumOxidative stress

a b s t r a c t

Congenital hypothyroidism is associated with delay in cell migration and proliferation in brain tissue,impairment of synapse formation, misregulation of neurotransmitters, hypomyelination and mentalretardation. However, the mechanisms underlying the neuropsychological deficits observed in congenitalhypothyroidism are not completely understood. In the present study we proposed a mechanism by whichhypothyroidism leads to hippocampal neurotoxicity. Congenital hypothyroidism induces c-Jun-N-termi-nal kinase (JNK) pathway activation leading to hyperphosphorylation of the glial fibrillary acidic protein(GFAP), vimentin and neurofilament subunits from hippocampal astrocytes and neurons, respectively.Moreover, hyperphosphorylation of the cytoskeletal proteins was not reversed by T3 and poorly reversedby T4. In addition, congenital hypothyroidism is associated with downregulation of astrocyte glutamatetransporters (GLAST and GLT-1) leading to decreased glutamate uptake and subsequent influx of Ca2+

through N-methyl-D-aspartate (NMDA) receptors. The Na+-coupled 14C-a-methyl-amino-isobutyric acid(14C-MeAIB) accumulation into hippocampal cells also might cause an increase in the intracellular Ca2+

concentration by opening voltage-dependent calcium channels (VDCC). The excessive influx of Ca2+

through NMDA receptors and VDCCs might lead to an overload of Ca2+ within the cells, which set off glu-tamate excitotoxicity and oxidative stress. The inhibited acetylcholinesterase (AChE) activity might alsoinduce Ca2+ influx. The inhibited glucose-6-phosphate dehydrogenase (G6PD) and gamma-glutamyltransferase (GGT) activities, associated with altered glutamate and neutral amino acids uptake couldsomehow affect the GSH turnover, the antioxidant defense system, as well as the glutamate–glutaminecycle. Reduced levels of S100B and glial fibrillary acidic protein (GFAP) take part of the hypothyroid con-dition, suggesting a compromised astroglial/neuronal neurometabolic coupling which is probably relatedto the neurotoxic damage in hypothyroid brain.

� 2013 Elsevier Ireland Ltd. All rights reserved.

Introduction

Hypothyroidism during brain development produces hypomye-lination, delay in cell migration and proliferation, impairment ofsynapse formation, and dysregulation of neurotransmitters(Zoeller and Rovet, 2004). However, the mechanisms whereby

congenital hypothyroidism leads to several neuropsychologicaldeficits are not completely understood.

Astrocytes are the key regulators of the uptake and metabolismof the excitatory neurotransmitter glutamate controlling the gluta-mate–glutamine cycle in the brain by the uptake of glutamatethrough glutamate transporters and metabolizing it into glutaminevia glutamine synthetase (GS) (Gras et al., 2006). The removal ofglutamate from the extracellular compartment mainly occursthrough the activity of two sodium-dependent astroglial glutamatetransporters termed GLAST (EAAT1) and GLT-1 (EAAT2) (Gegelash-vili and Schousboe, 1997; Danbolt, 2001). Otherwise, the system‘‘A’’ is the major pathway for the uptake of glutamine by neurons,while the release of this amino acid from glia is mainly mediated

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 15

by a system N transporter (Kanamori and Ross, 2004). Reduction inneuronal uptake of glutamine is accompanied by concomitantreduction in intracellular glutamate concentrations, indicating thatsystem A transporter-mediated glutamine uptake can be a prere-quisite for the formation of glutamate inside neurons (Jenstadet al., 2009).

Analysis of the levels of specific astrocyte markers has beenused to evaluate the implications of astrocytes in brain damage.In this context, the Ca2+-binding protein S100B and the cytoskeletalprotein glial fibrillary acidic protein (GFAP) are among the impor-tant astrocyte markers. S100B has many putative intracellular tar-gets, including cytoskeletal proteins in glial and neuronal cells. Theeffects of secreted S100B vary from neurotrophic to apoptotic ac-tions depending upon its concentration and signaling pathwayactivation (Donato, 2001, 2003). Otherwise, our group has previ-ously found that congenital hypothyroidism lead to down-regula-tion of GFAP in astrocytes of cerebral cortex of young rats(Zamoner et al., 2008a).

Perturbation of glutamatergic neurotransmission can be re-garded as a major cause for a number of neurological disordersand to be crucial for distinct neurodegenerative diseases (Rothsteinet al., 1992, 1995; Gillessen et al., 2002; Zamoner and Pessoa-Pur-eur, 2011). Moreover, the neurodegenerative diseases are fre-quently associated with oxidative stress (Jellinger, 1999; Cooksonand Shaw, 1999; Markesbery and Carney, 1999. The lowered GSHcontent appears to be the first indicator for oxidative stress duringthe progression of neurodegenerative diseases (Nakamura et al.,1997). The synthesis of GSH appears to be dependent on the regu-lation of synaptic glutamate concentration by astrocytes (Dringenand Hirrlinger, 2003). Thus, a disturbance in glutamate homeosta-sis might compromise the GSH content and function in CNS. In linewith this, we have previously found that excitotoxic mechanismsdecreased glutamate uptake and overactivated the N-methyl-D-aspartate (NMDA) receptors, with subsequent free radical produc-tion, leading to the misregulation of the phosphorylating systemassociated with the intermediate filament proteins from neuralcells (Zamoner et al., 2008a; Pierozan et al., 2012). These findingssuggest that alteration in the homeostasis of the cytoskeletoncould be among the responses to different neurotoxic conditions.

The hippocampus is involved in various memory processes thatare essential for creating new memories, and lesions to hippocam-pus result in impaired learning and memory (Zhang et al., 2011). Inrecent years many studies have determined the great vulnerabilityof the different areas of the hippocampus to developmentally in-duced hypothyroidism. Dong et al. (2005) found impaired induc-tion of long-term synaptic plasticity (LTP), reflecting CA1 damagein adult hypothyroid rats. However, LTP was not modified by con-genital hypothyroidism during development (PN10–PN18) (Varaet al., 2003), while the intrinsic excitability of CA1 neurons was(Sanchez-Alonso et al., 2012). Gong et al. (2010) described histo-logical alterations in hippocampal neurons of offspring in theCA1, CA3 and DG regions with nerve fiber malfunction and thusimpairments in hippocampal development of hypothyroid youngrats.

Therefore, in the present study we used a rat model of hypothy-roidism generated by neonatal exposure to propylthiouracil (PTU)to examine neurochemical parameters implicated in neuronaldevelopment, synaptic function, learning and memory. In spite ofthe relevance of glutamate metabolism for normal neuronal andastroglial functions, the relationship between hypothyroidismand glutamatergic excitotoxicity remains unclear. Given that astro-cyte and glutamate dysfunctions are described in several neurode-generative conditions, in the present report we investigated someaspects of the neurometabolic coupling of astrocytes and neurons.We focused on glutamate metabolism, oxidative defenses,astrocyte markers GFAP and S100B, the phosphorylating system

associated with cytoskeleton, as well as histological parametersof neural cells in hippocampus of hypothyroid rats in development.The present report provides an important contribution to the exist-ing literature regarding alterations in hippocampal developmentinduced by neonatal hypothyroidism.

Experimental procedures

Radiochemical and compounds

Anti-GLAST and anti-GLT-1 antibodies were obtained fromTocris Bioscience (Ellisville, MO, USA). [32P]Na2HPO4 waspurchased from CNEN, São Paulo, Brazil. L-[2,3-3H] glutamic acid([3H] glutamate) (specific activity 49 Ci/mmol) and the horseradishperoxidase conjugated anti rabbit IgG were purchased from Amer-sham (Oakville, Ontario, Canada). Paraplastic resin (Histosec) andEntellan were from Merk Brasil. Monoclonal anti-S100B antibody(SH-B1), rabbit polyclonal GFAP antiserum, diaminobenzidine, rab-bit peroxidase-anti-peroxidase (PAP), Triton X-100, polyclonalanti-GFAP, anti-GFAP (clone G-A-5), mouse anti-b-Actin, thyroxine(T4), 3,5,30triiodo-L-thyronine (T3), benzamidine, leupeptin, anti-pain, pepstatin, chymostatin, acrylamide; bis-acrylamide andanti-mouse IgG were from Sigma Chemical (St. Louis, MO, USA).Polyclonal anti-S100 was from DAKO. 45CaCl2 (specific activity of321 kBq/mg of Ca2+) and Optiphase Hisafe III biodegradable liquidscintillation were purchased from PerkinElmer (Waltham, MA).Anti-SAPK/JNK and anti-phospho SAPK/JNK antibodies were fromCell Signaling Technology, Inc. (Danvers, MA, USA) and the anti-body Ms. X NF-M/NF-H (anti KSP repeats) was from Chemicon(Temecula, CA, USA). a-[14C] methylaminoisobutyric acid ([14C]MeAIB) (sp.act. 1.85 GBq/mmol) was purchased from Du Pont,NEN Products, Mass, USA. The Immobilon™ Western chemilumi-nescent HRP substrate was obtained from Millipore. The mouseanti-NeuN (NEUronal Nuclei; clone A60) antibody was obtainedfrom Chemicon International (São Paulo, Brazil). All other chemi-cals were of analytical grade.

Animals

Wistar rats were obtained from our breeding stock. Rats weremaintained on a 12-h light/12-h dark cycle in a constant tempera-ture (22 �C) colony room. On the day of birth the litter size wasculled to eight pups. Dams were provided ad libitum with waterand the Nuvilab CR-1 irradiated chow obtained from Nuvital(Curitiba, PR, Brazil). All procedures were conducted in accordancewith ethical recommendations of the Brazilian Veterinary Medi-cine Council and the Brazilian College of Animal Experimentation(Protocol CEUA/PP00318). Results were obtained at least fromthree independent experiments using three or four animals pergroup in each experiment.

Induction of hypothyroidism

Congenital hypothyroidism was induced by adding 0.05% 6-pro-pyl-2-thiouracil (PTU) in the drinking water from gestation day 9and continually up to lactation day 15 (Zamoner et al., 2008a).Euthyroid rats, receiving only water during the same period, wereused as controls.

Preparation of hippocampus slices

Rats were killed by decapitation, the blood was collected, andthe hippocampus was dissected onto Petri dishes placed on iceand the parietal region was cut into 300 lm thick slices with aMcIlwain chopper.

16 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

32P in vitro incorporation

Preincubation and 32P labelingTissue slices were initially preincubated at 30 �C for 20 min in a

Krebs-Hepes medium containing 124 mM NaCl, 4 mM KCl, 1.2 mMMgSO4, 25 mM Na-HEPES (pH 7.4), 12 mM glucose, 1 mM CaCl2,and the following protease inhibitors: 1 mM benzamidine,0.1 lM leupeptin, 0.7 lM antipain, 0.7 lM pepstatin and 0.7 lMchymostatin. After preincubation, the medium was changed andincubation was carried out at 30 �C with 100 lL of the basic med-ium containing 80 lCi of [32P] orthophosphate, with or withoutaddition of 1 lM T3 or 0.1 lM T4, when indicated. The labelingreaction was normally allowed to proceed for 30 min at 30 �Cand stopped with 1 mL of cold stop buffer (150 mM NaF, 5 mM,EDTA, 5 mM EGTA, 50 mM Tris–HCl, pH 6.5, and the proteaseinhibitors described above. Slices were then washed twice withstop buffer to remove excess radioactivity.

Preparation of the high salt-Triton insoluble cytoskeletal fractionBriefly, after the labeling reaction, slices were homogenized in

400 lL of ice-cold high salt buffer containing 5 mM KH2PO4, (pH7.1), 600 mM KCl, 10 mM MgCl2, 2 mM EGTA, 1 mM EDTA, 1% Tri-ton X-100 and the protease inhibitors. The homogenate was centri-fuged at 15,800g for 10 min at 4 �C in an Eppendorf centrifuge. Thesupernatant was discarded and the pellet was homogenized withthe same volume of the high salt medium. The resuspendedhomogenate was centrifuged as described and the supernatantwas discarded again. The Triton-insoluble IF-enriched pellet, con-taining neurofilament (NF) subunits of low, medium and highmolecular weight (NF-L, NF-M and NF-H respectively), vimentinand GFAP, was dissolved in 1% SDS and protein concentrationwas determined.

Polyacrylamide gel electrophoresis (SDS–PAGE)For electrophoresis analysis, samples were dissolved in 25%

(v/v) of a solution containing 40% glycerol, 5% mercaptoethanol,50 mM Tris–HCl, pH 6.8 and boiled for 3 min. Equal protein con-centrations were loaded onto 10% polyacrylamide gels, analyzedby SDS–PAGE. The gels containing the IF-enriched cytoskeletalfraction were exposed to X-ray films (T-mat G/RA) at �70 �C withintensifying screens, and the autoradiograph was obtained.

Western blot analysis

In some experiments total tissue extracts were processed toWestern blot analysis. The nitrocellulose membranes were washedfor 10 min in Tris-buffered saline (TBS; 0.5 M NaCl, 20 mM Trizma,pH 7.5), followed by 2 h incubation in blocking solution (TBS plus5% defatted dried milk). After incubation, the blot was washedtwice for 5 min with TBS plus 0.05% Tween-20 (T-TBS), and thenincubated overnight at 4 �C in blocking solution containing0.25 lg/mL rabbit anti-GLT-1 antibody, anti-GFAP (clone G-A-5)diluted 1:500 or mouse anti-b-actin diluted 1:2000. The blot wasthen washed twice for 5 min with T-TBS and incubated for 2 h inTBS containing horseradish peroxidase conjugated secondary anti-body anti-rabbit IgG diluted 1:2000 or anti-mouse IgG diluted1:1000. The blot was washed twice again for 5 min with T-TBSand twice for 5 min with TBS. The blot was then developed usingthe Immobilon™ Western chemiluminescent HRP substrate).

Quantitation of Western blot and 32P incorporation into IF proteins

The radioactivity incorporated into cytoskeletal proteins as wellas the immunoblots were quantified by scanning the films with aHewlett–Packard Scanjet 6100C scanner and determining optical

densities with an OptiQuant version 02.00 software (PackardInstrument Company).

Glutamate uptake assay

Tissue slices were incubated with Hank’s balanced salt solution(HBSS) containing 137 mM NaCl; 0.63 mM Na2HPO4; 4.17 mMNaHCO3; 5.36 mM KCl; 0.44 mM KH2PO4; 1.26 mM CaCl2;0.41 mM MgSO4; 0.49 mM MgCl2 and 5.55 mM glucose, in pH7.4. The assay was started by the addition of 0.1 mM L-glutamateand 0.66 lCi/mL L-[2,3-3H] glutamate. Incubation was stoppedafter 5 min by removal of the medium and rinsing the slices twicewith ice-cold HBSS. Slices were then lysed in a solution containing0.5 M NaOH. Sodium-independent uptake was determined usingN-methyl-D-glucamine instead of sodium chloride. Sodium-depen-dent glutamate uptake was obtained by subtracting the non-spe-cific uptake from the specific uptake. Radioactivity was measuredwith a scintillation counter.

Glutamine synthetase activity

Briefly, homogenate (0.1 mL) was added to 0.1 mL of reactionmixture containing 10 mM MgCl2; 50 mM L-glutamate; 100 mMimidazole–HCl buffer (pH 7.4); 10 2-mercaptoethanol; 50 hydrox-ilamine-HCl; 10 ATP and incubated for 15 min at 37 �C. The reac-tion was stopped by the addition of 0.4 mL of a solutioncontaining (in mM): 370 ferric chloride; 670 HCl; 200 trichloroace-tic acid. After centrifugation, the supernatant was measured at530 nm and compared to the absorbance generated by standardquantities of c-glutamylhydroxamate treated with ferric chloridereagent.

Glutathione content

GSH content was determined as described by Browne andArmstrong (1998). Briefly, slices were homogenized in sodiumphosphate buffer (0.1 M, pH 8.0) containing 0.005 M EDTA andprotein was precipitated with 1.7% meta-phosphoric acid. Super-natant was assayed with o-phthaldialdeyde (1 mg/mL) at roomtemperature for 15 min. Fluorescence was measured using excita-tion and emission wavelengths of 350 and 420 nm, respectively.GSH concentrations were calculated as nmol/mg protein.

Lipoperoxidation (TBARS levels) evaluation

The lipid peroxidation extent was spectrophotometricallydetermined at 535 nm in hippocampal slices by the thiobarbituricacid method described by Bird and Draper (1984). Values were ex-pressed in nmol TBARS mL�1.

S100B measurement

The S100B immunocontent was determined by ELISA in sam-ples of hippocampus and serum of control and hypothyroid youngrats. ELISA for S100B was carried out as previously described (Leiteet al., 2008).

45Ca2+ uptake

Slices of hippocampus from 15-day-old male rats were preincu-bated in Krebs Ringer-bicarbonate (KRb) buffer (122 mM NaCl;3 mM KCl; 1.2 mM MgSO4; 1.3 mM CaCl2; 0.4 mM KH2PO4;25 mM NaHCO3) for 15 min in a Dubnoff metabolic incubator at32 �C, pH 7.4 and gassed with O2:CO2 (95:5; v/v). After that, themedium was changed by fresh KRb with 0.1 lCi/mL 45Ca2+ during

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 17

5, 15 or 30 min. Extracellular 45Ca2+ was thoroughly washed off in awashing solution containing 127.5 mM NaCl, 4.6 mM KCl, 1.2 mMMgSO4, 10 mM HEPES, 11 mM glucose, 10 mM LaCl3, pH 7.3(30 min in washing solution). The presence of La3+ during thewashing stage was found to be essential to prevent release of theintracellular 45Ca2+ (Zamoner et al., 2007). After washing, tissuewas lysed in 0.5 M NaOH, the protein concentration was deter-mined (Lowry et al., 1951). Radioactivity was measured with ascintillation counter.

[14C] MeAIB accumulation

For amino acid accumulation experiments, hippocampal sliceswere pre-incubated in KRb buffer for 30 min in a Dubnoff metabolicincubator at 37 �C, pH 7.4 and gassed with O2:CO2 (95:5; v/v).The slices were then incubated in fresh KRb buffer for 60 min.[14C] MeAIB (3.7 kBq/mL) was added to each sample during theincubation period (Silva et al., 2001). After incubation the sliceswere lysed in 0.5 M NaOH, the protein concentration was deter-mined (Lowry et al., 1951), and radioactivity was measured witha scintillation counter. The results were expressed as the tissue/medium (T/M) ratio: cpm/mL tissue fluid per cpm/mL incubationmedium.

Enzymatic activity of GGT, acetylcholinesterase and G6PD

Slices of hippocampus were homogenized in cold 0.1 M Trisbuffer, pH 8.5 (10% homogenate w/v) to determine GGT activityor in 0.2 M Tris buffer, pH 7.4 to quantify acetylcholinesteraseand G6PD activities. Sample aliquots were saved for total proteindeterminations (Lowry et al., 1951).

GGT assayGamma-glutamyl transferase (GGT) activity was measured with

the use of the modified technique described previously by Orlow-sky and Meister (1963) using L-c-glutamyl p-nitroanilide as sub-strate and glycylglycine as the acceptor molecule. The resultswere expressed as IU/L/lg protein.

G6PD assayFor measuring the glucose-6-phosphate dehydrogenase (G6PD)

activity, hippocampus slices were incubated in the presence ofNADP+ leading to the oxidation of glucose-6-phosphate to 6-phos-phoglunate. The NADPH produced was measured in a kineticmode.

Cholinesterase assayCholinesterase enzymatic assay was performed by using acetyl-

thiocholine as a substrate. The rate of hydrolysis of acetylthiocho-line was measured at 412 nm for 2 min, through the release of thethiol compound that reacts with DTNB (5,50-dithio-bis(2-nitroben-zoic acid) and produces the color-forming compound TNB. Theactivity was expressed in U/lg protein.

Histological procedures for cell counting

Control and hypothyroid animals were anaesthetized with xyla-zine (0.1 mL/100 g body weight, i.m.) and ketamine (0.1 mL/100 gbody weight, i.m.) and perfused with phosphate buffered saline(PBS) containing heparin (50 mL in the 15-day-old rats) followedby 4% paraformaldehyde diluted in 0.1 M phosphate buffer (pH7.4) at 4 �C of the same flow rate and total amount. The perfusionrate was approximately 100 mL. After perfusion, the brain was ex-tracted from the skull, weighed, and placed in the same fixing solu-tion for 72 h. After fixation, the brain was washed for 1 h inrunning water, dehydrated in increasing concentrations of ethanol

(70%, 80%, 90%, 95% and absolute ethanol) and cleared with xylol.Brains were then embedded in paraplastic resin (Histosec). Coronalserial sections (6 lm thickness) were obtained with a microtomeand serially collected on glued slides. The tissue was stained withcresyl violet, and then the sections were dehydrated through analcohol series, cleared with xylene, and coverslipped with Entellan.The cell body layer of the CA1, CA2, CA3, CA4 and dentate gyrofields of the hippocampus were identified according to a rat brainatlas (Paxinos and Watson, 2008).

Estimation of the numerical density of cells

Images were captured with a light microscope (Olympus, BX-41) equipped with a digital camera (3.3 Mpixel QCOLOR3C, Qimag-ing™) and image acquisition software (Qcapture Pro 5.1, Qimag-ing™). Sections were viewed with a 100� planachromaticimmersion oil objective. For each frame, the thickness of the sec-tion was measured by focusing on the upper and lower section sur-faces, and verified objectively by manually focusing to control thez-axis (Mandarim-de-Lacerda et al., 2010). The neurons with awell-focused nucleus when focusing upward, but not downward,were counted. The nucleus of the cell should be inside the test area,but it could not touch two borders of the previously chosen testareas (forbidden lines) (Gundersen, 1977; Howard and Reed,2005).

For the sake of efficiency, we analyzed the nuclei of the neurons.The number of cell nuclei was counted inside a frame of 2958 lm3

in the lookup plane. The center of the nucleus, as opposed to thetop of nucleus, was measured. The thickness of the disector (t)was 5 lm, representing 1/4–1/3 of the height of the nuclei, whichin practice is the more important constraint on section thickness(Gundersen and Osterby, 1981). Beginning with the third posteriorsection, we counted each 80 section rostrally. We analyzed 10 sec-tions throughout the hippocampus, between figures 27 (bregma:2.30) and 42 (bregma: 6.04) of the atlas (Paxinos and Watson,2008).

Immunohistochemistry for GFAP and NeuN

After fixation, the brain was cryoprotected with a 15% sucrose/PBS solution until the sample was saturated, followed by immer-sion in a 30% sucrose/PBS solution. The sample was then frozenin cooled isopentane, and coronal serial sections (50 lm thickness)of the hippocampus were obtained using a cryostat and rat brainatlas (Paxinos and Watson, 2008) for anatomical references.

The samples were collected in PBS solution every 100 lm. Thetissue was processed for GFAP immunohistochemistry by usingthe antibody peroxidase–antiperoxidase (PAP) procedure. Accord-ing to this procedure, free-floating sections were treated in 10%methanol and 3% H2O2 for 30 min and washed carefully. Then,the sections were pre-incubated in 3% normal goat serum (NGS)in PBS containing 0.3% Triton X-100 (PBS-Tx) for 30 min and incu-bated with rabbit polyclonal anti-GFAP diluted 1:150 or mouseanti-NeuN antibody, diluted 1:500 in 3% NGS in PBS-Tx for 48 hat 4 �C. The NeuN antibody specifically recognizes the DNA-bind-ing, neuron-specific protein NeuN, which is present in most neuro-nal cell types of all vertebrates tested. NeuN protein distributionsare apparently restricted to neuronal nuclei, perikarya and someproximal neuronal processes (Mullen et al., 1992), while GFAP isan IF protein, used as astrocyte biomarker. After washing severaltimes with PBS-Tx, sections were incubated in an anti-rabbit IgGsolution diluted 1:50 in PBS-Tx at room temperature for 2 h. Sec-tions were washed again in PBS and incubated in anti-rabbit (forGFAP) or anti-mouse (for NeuN) peroxidase-anti-peroxidase(PAP) diluted 1:500 in PBS for 2 h at room temperature. Theimmunohistochemical reaction was performed by incubating the

18 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

sections in a histochemical medium containing 0.06% 303-diam-inobenzidine (DAB) dissolved in PBS for 10 min and then in a sim-ilar solution containing 1 lM of 3% H2O2 per mL of DAB mediumfor 10 min. Afterwards, the sections were rinsed in PBS, dehydratedin ethanol, cleared with xylene and covered with Entellan and cov-erslips. The tissue of all animals in this experiment was processed(perfusion and immunodetection) in the same day, using the samesolutions to avoid changes in background and differences in thechromogen reaction.

Statistical analysis

Variations in astroglial cell density and neuron density for eachpyramidal cell layer of the hippocampus induced by hypothyroid-ism were analyzed by a one-way ANOVA. Post-hoc analyses werecarried out using the Tukey HSD test when appropriate. Biochem-ical data were analyzed statistically by Student t test or one-wayANOVA followed by Bonferroni multiple comparison test. Differ-ences were considered to be significant when p < 0.05.

Results

In the present study, we induced congenital hypothyroidism byadministration of 0.05% PTU to the dams in the drinking water(Zamoner et al., 2008a, 2008b). The effectiveness of PTU is evi-denced by decreased T3 and T4, associated with high TSH levels(Table 1). Therefore, neurochemical and histochemical parameterswere analyzed in the hippocampus of 15-day-old hypothyroidmale pups.

Hypothyroidism disrupted the homeostasis of the cytoskeleton andthis effect was not reversed by thyroid hormones

We initially investigated the in vitro effect of congenital hypo-thyroidism on the activity of the phosphorylating system associ-ated with the glial fibrillary acidic protein (GFAP), vimentin andneurofilament (NF) subunits of low (NF-L), medium (NF-M) andhigh (NF-H) molecular weight, of hippocampal astrocytes and neu-rons respectively. Results showed that hypothyroid rats presentedhyperphosphorylation of all IF subunits, as compared with euthy-roid animals (Fig. 1 A). Moreover, when hypothyroid tissue sliceswere incubated with 1 lM T3 or 0.1 lM T4, IF hyperphosphoryla-tion was not reversed, except for GFAP hyperphosphorylation,which was partially reversed by T4 (Fig. 1A).

The lysine–serine–proline (KSP) repeats located in the carboxyl-terminal tail domains of NF-M and NF-H were the main phosphor-ylating sites targeted in the hippocampal neurons of hypothyroidanimals, as depicted in Fig. 1B and this effect could have been med-iated, at least in part, by the c-Jun-N-terminal kinase (JNK), whichis a mitogen-activated protein kinase (MAPK) family member

Table 1Effect of congenital hypothyroidism on serum levels of thyroid hormones (T3 and T4)and thyroid-stimulating hormone (TSH) in immature rats.

Control Hypothyroid

FT3 (ng/mL) 3.14 ± 0.018 0.38 ± 0.0021*

FT4 (ng/dL) 2.47 ± 0.015 0.25 ± 0.0017*

TSH (ng/mL)-rTSH 0.87 ± 0.0028 4.47 ± 0.0035*

T3 (ng/mL) 8.75 ± 0.01 0.69 ± 0.002*

T4 (lg/dL) 6.11 ± 0.001 0.80 ± 0.001*

Serum levels of free and total T3 (FT3 and T3, respectively), free and total T4 (FT4 andT4, respectively) and TSH (rTSH = rat TSH). Data are reported as means ± S.E.M;n = 10 in both groups. Statistically significant differences from controls, as deter-mined by Student’s t test are indicated.* P < 0.0001.

which appeared activated/phosphorylated in the hypothyroid hip-pocampus (Fig. 1C). KSP repeats hyperphosphorylation was not re-versed by supplementation with T3, but it was partially reversed byT4, suggesting that this hormone is able to partially restore thehomeostasis of JNK signaling pathway preventing the disruptionof the neural cytoskeleton. These results shed light into the effectsof congenital hypothyroidism on the signaling mechanisms associ-ated with the cytoskeleton of neural cell in the hippocampus ofrats. Also, the total inability of T3 and the partial ability of T4 inrestoring this equilibrium, suggest the complexity of these actions.We, therefore, assessed other neurochemical parameters to betterevaluate the consequences of congenital hypothyroidism in thehippocampus during development.

Hypothyroidism decreases the levels of the astrocyte proteins S100B,GFAP, GLAST and GLT-1 in hippocampus

Astrocytes are active participants in brain metabolism, synapticplasticity and information processing, thus to further assess theconsequences of hypothyroidism on astrocyte function, we ana-lyzed the expression of two important astrocyte biomarkers,named S100B and GFAP, as well as the levels of the glutamatetransporters GLAST and GLT1. S100B levels significantly decreasedin hippocampal slices from hypothyroid animals. However, theserum levels of this protein were increased (Fig. 2A). Moreover,down-regulation of GLAST and GLT-1 immunocontent observedin hippocampus (Fig. 2B) was consistent with the low glutamateuptake (Fig. 3A). Furthermore, GFAP levels were significantlydecreased in this brain structure, suggesting down-regulatedexpression, increased turnover or reduced cell number (Fig. 2C).

Congenital hypothyroidism alters [3H]-glutamate, [14C]-MeAIB and45Ca2+ uptake

Congenital hypothyroidism significantly decreased the astro-cyte [3H]-glutamate uptake (Fig. 3A). Accordingly, tissue slices ex-posed to 45Ca2+ during 5, 15 and 30 min showed increased 45Ca2+

accumulation into hippocampal cells after 30 min exposure tothe radiolabel (Fig. 3B), compatible with glutamate excitotoxicityin the hippocampus of hypothyroid rats. However, the increased[14C]-MeAIB accumulation in hippocampus (Fig. 3C) suggestedhigher glutamine uptake by neurons in the hypothyroid hippocam-pus, as compared with euthyroid animals.

Hypothyroid condition leads to oxidative stress and modifications onenzymatic activities

We aimed to investigate the status of GSH–GSSG in the hippo-campus of hypothyroid rats since GSH plays a fundamental role inprotecting cells by destroying hydrogen peroxide and hydroxyl freeradicals. Congenital hypothyroidism lead to reduced GSH content(Fig. 4A), associated with higher levels of TBARS (Fig. 4 B), stronglysuggesting oxidative stress generation. The decreased GSH levelscould be associated with decreased [3H]-glutamate uptake(Fig. 3A) in hypothyroid pups, since glutamate is partially directedto GSH production in astrocytes. Consistent with low glutamatelevels, the astrocytic enzyme glutamine synthetase (GS) is downregulated, since this enzyme uses glutamate to generate glutamineinto the astrocytes (Fig. 4C). Thus, it is feasible that decreased GSactivity could be related with glutamate excitotoxicity in congeni-tal hypothyroidism. Also, regeneration of GSH from its oxidizedform (GSSG) requires NADPH produced in the glucose-phosphatedehydrogenase (G6PD) reaction. Thus, the lower levels of GSH,could be related with the significant reduction on G6PD activityin hippocampus from hypothyroid animals (Fig. 4D). In addition,gamma-glutamyltransferase (GGT) levels appeared lowered in

Fig. 1. Effect of congenital hypothyroidism on the phosphorylating system associate with intermediate filament subunits in the cytoskeletal fraction from slices of rathippocampus. (A) Hippocampal slices were incubated in the presence of 32P orthophosphate for 30 min. The high-salt Triton insoluble cytoskeletal fraction was extracted andthe radioactivity incorporated into the intermediate filament subunits was measured. Representative loading control: equal amounts of the cytoskeletal fraction fromcontrols (C), hypothyroid (H), T3-treated hypothyroid (H + T3) and T4-treated hypothyroid (H + T4) samples stained by Coomassie blue. (B) Effect of congenital hypothyroidismon the immunocontent of phosphoNF-H KSP repeats in rat hippocampus. Representative immunoblots and a representative loading control (beta-actin) are presented. (C)Immunocontent of total- and phospho-JNK1/2 in hypothyroid hippocampus of 15-day-old rats. Total tissue homogenate from control and hypothyroid hippocampus wereobtained and the phosphorylation/activation of JNK1/2 was determined by Western blot. Representative immunoblots are presented. Scans from 10 different animals fromeach group were quantified. Data are reported as means ± S.E.M of 8 animals. Statistically significant differences from controls, as determined by one-way ANOVA followed byTukey–Kramer multiple comparison test were indicated: �P < 0.001; ��P < 0.0001.

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 19

hypothyroid animals (Fig. 4 E). The primary role of cellular gammaglutamyltransferase (GGT) is to metabolize extracellular reducedglutathione (GSH), allowing for precursor amino acids to be assim-ilated and reutilized for intracellular GSH synthesis (Lee et al.,2004). Therefore, it is feasible that low GGT activity could be re-lated with low cytoplasmic GSH levels in the hippocampus ofhypothyroid animals.

Hypothyroidism significantly reduces the acetylcholinesteraseactivity in the spinal cord of rat pups (Koohestani et al., 2012)moreover, it has been described that free radicals are associatedwith cholinergic dysfunction (Ahmad et al., 2012). In line with this,we found decreased acetylcholinesterase activity in hippocampalslices of hypothyroid rats, although no modifications were ob-served in serum samples from these pups (Fig. 5), suggesting thataltered cholinergic transmission could be associated with the

oxidative imbalance we are evidencing in the congenitalhypothyroidism.

Hypothyroidism reduces neuron and glial cell number in hippocampus

In order to tentatively establish a link between the neurochem-ical parameters analyzed and histochemical alterations in thehypothyroid hippocampus, we carried out histochemical analysis.Fig. 6 shows the difference in cresyl violet stained pyramidal neu-ron density in the hippocampus from control and hypothyroidpups (CA1, CA2, CA3, CA4 and dentate gyros – DG). Because theneuron density did not differ between the two hemispheres, weeliminated this factor from the analysis by grouping data fromthe two hemispheres. The number of neurons was decreased inhypothyroid hippocampus at all the analyzed regions. On the other

Fig. 2. Effect of congenital hypothyroidism on the levels of the astrocyte proteinsS100B, GFAP and glutamate transporters GLAST and GLT-1. S100B immunocontentwas determined by ELISA in serum in hippocampus of rats. GFAP, GLAST and GLT-1immunocontents were determined by Western blot in homogenate of rat hippo-campus. Representative immunoblots were showed: C = Control; H = hypothyroid.b-actin immunoblots were showed as protein loading control. Data are reported asmeans ± S.E.M. of 10 animals and expressed as% of controls. Statistically significantdifferences from controls, as determined by Student’s t-test were indicated:�P < 0.001; ��P < 0.0001.

Fig. 3. Effect of congenital hypothyroidism on [3H]-glutamate uptake (A), aminoacid accumulation (B) and 45Ca2+ uptake (C) in rat hippocampus. Data are reportedas means ± S.E.M. of 8 animals in each group. Statistically significant differencesfrom controls, as determined by Student’s t-test were indicated: ��P < 0.0001.

20 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

hand, the number of glial cells in CA4 and dentate gyros fromhypothyroid group was lower than in the control group while nodifferences were observed between groups in the other hippocam-pal regions. Otherwise, it is important to note that despite the

immunohistochemical analysis depicted in Fig. 7 shows that NeuNimmunoreactive neurons were detected in all hippocampal subre-gions, both in control and hypothyroid groups, decreased NeuN po-sitive immunoreactivity confirmed the decreased neuronalpopulation in all areas of hippocampus in both hemispheres. Inthe immunohistochemistry depicted in Fig. 8 we can observe theastroglial (GFAP positive immunoreactivity) cell counting amongthe pyramidal cells of the hippocampal areas in both hemispheresof 15-day-old male pups. Interestingly, the effect of hypothyroid-ism is different in the right hemisphere from that on the left one.Results demonstrated lower GFAP positive cells in CA3 and CA4 re-gions in the right hemisphere of the hippocampus, while in the lefthemisphere the GFAP positive cells were decreased in CA4 anddentate gyros. Surprisingly, in CA1 of the left hemisphere GFAPimmunoreactivity was increased, according to cell counting.

Discussion

The present study demonstrated that congenital hypothyroid-ism is associated with JNK activation, cytoskeleton misregulation,as well as downregulation of astrocyte glutamate transporters

Fig. 4. Congenital hypothyroidism leads to oxidative stress and modifies the activity of enzymes involved in GSH metabolism in rat hippocampus. GSH content (A),thiobarbituric acid-reactive substances (TBARS) measurement of lipid peroxidation (B), glutamine synthetase activity (C), c-glutamyl trasferase activity (D) and G6PD activity(E). Data are reported as means ± S.E.M. of 8 animals. Statistically significant differences from controls, as determined by Student’s t-test were indicated: �P < 0.001;��P < 0.00001; ���P < 0.000001.

Fig. 5. Effect of congenital hypothyroidism on acetylcholinesterase activity inserum and hippocampus of immature rats. Data are reported as means ± S.E.M. of10 animals. Statistically significant differences from controls, as determined byStudent’s t-test were indicated: ��P < 0.0001.

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 21

leading to decreased glutamate uptake with subsequent influx ofCa2+ through NMDA receptors in rat hippocampus. The Na+-cou-pled 14C-MeAIB accumulation into hippocampal cells also mightcause an increase in the intracellular Ca2+ concentration by open-ing VDCC. The excessive influx of Ca2+ through NMDA receptorsand VDCCs leads to an overload of Ca2+ within the cells, which

set off glutamate excitotoxicity and oxidative stress. These eventssuggest a compromised neural defense system that is probably re-lated to the neurotoxic damage in hypothyroid brain.

Interestingly, hypothyroidism altered the phosphorylating sys-tem associated with cytoskeleton of both neuron and astrocyteIFs, but supplementation with T3 did not reverse either neuronaland astrocyte IF hyperphosphorylation. On the other hand, thephosphorylation of GFAP and of KSP repeats on NF-H were partiallyreversed by T4, suggesting that this hormone might restore, at leastin part, the homeostasis of the phosphorylating system associatedwith the cytoskeleton. These data are consistent with previousevidence from Zamoner et al. (2008a) showing that hypothyroid-induced hyperphosphorylation of IF was not reversed byshort-term exposure to T3 in cerebral cortex of young rats.

NF-M and NF-H KSP repeats are among the preferential targetsfor the MAPKs, including JNK, and misregulated phosphorylation ofKSP sites, in particular those of NF-H, regulates NF axonal transport(Jung et al., 2000). In this context, hyperphosphorylation of thesesites plays a role in several neurological disorders (Sihag et al.,2007). Activation of JNK is regarded as a molecular switch in stresssignal transduction. Several stimuli, including ROS (Seki et al.,2012), are able to activate JNK pathway, playing critical roles inregulating cell fate, including gene expression, cell proliferationand programmed cell death. Also, Sherrin et al. (2011) describeda role for hippocampal JNK in memory and synaptic plasticity.Therefore, it is feasible that ROS generation in the hypothyroid

Fig. 6. Cell counting from cresyl violet staining showing the effects of hypothyroidism on neuronal (A) and astroglial (B) cell density in the hippocampal pyramidal cell layerof 15-day-old male rats. Panel C: hippocampal overview and selected regions of CA1, CA2, CA3, CA4 and dentate gyrus – DG. Pyramidal cells were recognized as having adistinct nucleolus within an ovoid nucleus and had elongated perikarya. The arrow indicates a neuron, which is recognized as having a larger size than other cells and havinga distinct nucleolus within the nucleus. Bars represent the mean ± S.E.M of 5 animals in each group (control and hypothyroid). Results were compared between groups byusing a one-way ANOVA followed by Tukey–Kramer multiple comparison test. Statistically significant differences from controls were indicated: �P < 0.01; ��P < 0.001;���P < 0.0001.

Fig. 7. Immunohystochemistry for NeuN of the hippocampus (CA1, CA2, CA3, CA4 and dentate gyrus – DG) of 15-day-old male rats. The arrow indicates a NeuNimmunoreactive neuron. Bars represent the mean ± S.E.M of 5 animals in each group (control and hypothyroid). Results were compared between groups (control andhypothyroid). Statistically significant differences from controls, as determined by one-way ANOVA followed by Tukey–Kramer multiple comparison test were indicated:�P < 0.01; ��P < 0.001.

22 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

hippocampus of 15-day-old rats elicited the JNK signaling path-way, misregulating the homeostasis of the cytoskeleton and con-tributing to delayed hippocampal development.

In the CNS, astrocytes are the main protectors of neurons fromexcitotoxicity and this protection is conferred mainly by clearanceof extracellular glutamate (Danbolt, 2001; Gras et al., 2006). Taking

Fig. 8. Immunohystochemistry for GFAP in the right (A) and left (B) hemispheres of the hippocampus (CA1, CA2, CA3, CA4 and dentate gyrus – DG) of 15-day-old male rats.The arrows indicate GFAP positive immunoreactive astrocytes. Bars represent the mean ± S.E.M of 5 animals in each group (control and hypothyroid). Results were comparedbetween groups (control and hypothyroid). Statistically significant differences from controls, as determined by one-way ANOVA followed by Tukey–Kramer multiplecomparison test were indicated: �P < 0.01; ��P < 0.001.

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 23

into account these findings, our results demonstrating decreasedglutamate uptake, increased 45Ca2+ uptake, decreased levels of glu-tamate transporters (GLAST and GLT-1) and diminished GS activityin hippocampal slices from hypothyroid rats support a possibleglutamate-induced excitotoxicity in this brain structure. In thiscontext, in astrocytes, the enzyme GS rapidly converts glutamateinto glutamine (Danbolt, 2001; Dringen and Hirrlinger, 2003),which serves as an important precursor for the synthesis of the pri-mary excitatory neurotransmitter by neurons through the neuro-nal enzyme glutaminase. The hypothyroid condition clearlyshowed a disruption of glutamate–glutamine cycle, by interferingon the excitatory neurotransmitter metabolism (by inhibiting GS)and transport (by decreasing GLAST and GLT-1 levels). Theseevents may induce neuronal excitotoxicity due to an excessiveextracellular glutamate concentration (Coulter and Eid, 2012).

Moreover, considering that GS expression and activity areup-regulated by T3 the reduced GS activity we are evidencing isconsistent with the hypothyroid condition. Because purines,

pyrimidines, and certain amino acids necessary for the synthesisof myelin components are, in part, provided by the GS pathway,T3 effect on myelination development and maturation could bemediated in part, through the GS gene regulation (Baas et al.,l998). Thus, myelination delay and reduced myelin amount inhypothyroid rat brain (Ibarrola and Rodriguez-Pena, 1997; Tosicet al., 1992) could be related with the decreased GS activity/expression we are evidencing.

In CNS, S100B is found predominantly in the cytoplasm and nu-cleus of astrocytes, where it regulates cytoskeleton and cell prolif-eration, along with other members of the S100 family. Part of thisprotein content (less than 1%) is secreted in a regulated manner(Gonçalves et al., 2008). As a consequence, S100B can be detectedin the cerebrospinal fluid and in serum. S100B protein levels (intraand extracellular) have been used as a parameter of astrocyte acti-vation and/or death in several situations of brain injury (Kleind-ienst and Ross Bullock, 2006; Berger et al., 2007; Tanaka et al.,2008; Murabayashi et al., 2008). Our present results show that

24 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

hypothyroidism leads to elevated S100B serum levels associatedwith decreased content of this protein in the hippocampus ofdeveloping rats. These data would support a role for serumS100B as peripheral marker of brain damage in the hypothyroid-ism. Our results are consistent with previous data describing ele-vated S100 serum levels in perinatal acidosis (Maschmann et al.,2000) and higher S100B in the serum of newborns with hypoxic–ischemic encephalopathy (Murabayashi et al., 2008). However,many extracerebral sources contribute to the serum S100B con-tent, including adipocytes, chondrocytes, lymphocytes, bone mar-row cells and melanoma cells (Donato, 2001).

On the other hand, considering that hypothyroidism in develop-ing rats results in abnormal and delayed brain development(Bernal, 2007), we could ascribe the high S100B serum levels as aconsequence of delayed glial maturation and or decreased glial cellnumber in these rats. In fact, a developmental postnatal decreasehas been described in serum S100B of children (Portela et al.,2002) and in CSF S100B of rats (Tramontina et al., 2002). Brain tis-sue decrease in S100B content, glutamate transporters and gluta-mine synthetase, as well as in GFAP content are all associatedwith astrocyte maturity, and reinforces the possibility of delayedastroglial development.

It is also described that elevated extracellular glutamate levelsdecrease S100B release in hippocampal astrocytes (Tramontinaet al., 2006). Taking into account that we have demonstrateddecreased levels of glutamate transporters in cerebral cortex(Zamoner et al., 2008a) and hippocampus of hypothyroid animals,supporting high extracellular concentrations of glutamate, it is fea-sible that the reduced intracellular S100B levels might be related toglutamate excitotoxicity. Also, the reduced glutamate uptake leadsto decreased intracellular glutamate that therefore becomesunavailable as a GSH precursor (Dringen and Hirrlinger, 2003).Glutamate and GSH metabolisms intricately interact as a resultof complex mechanisms and these interactions might be alteredin hypothyroidism.

GSH is a critical component of astrocyte neuroprotective func-tions. In this context, reduced GSH content associated with in-creased TBARS levels observed in hippocampus might induceneurotoxic damage in hypothyroid rat brain. Moreover, the highlevel of O2 consumption confers a particular vulnerability of brainto oxidative stress. Thus, down-regulation of GSH, the main anti-oxidant of brain, might be responsible, at least in part, by the oxi-dative stress in hypothyroid brain (Dasgupta et al., 2007), sincecompromised astroglial GSH system may contribute to a lowerdefense capacity (Drukarch et al., 1997). GSH-dependent redoxsystems are reliant upon a continuous supply of NADPH as themajor electron donor. The cellular supply of NADPH is mainly be-lieved to be provided by G6PD activity (Halliwell and Gutteridge,2007). In this context, the inhibition of G6PD observed hereinmight lead to decreased supply of NADPH and subsequent oxida-tive stress generation in hippocampus of hypothyroid animals andmight be also associated with decreased levels of GSH. Glutathi-one synthesis is orchestrated by the enzymes of the gamma-glut-amyl cycle and deficiencies in the enzymes involved in the GSHmetabolism could be associated with neurological disorders.GGT is an astroglial ecto-enzyme which transfers the gamma-glutamyl moiety of GSH to an acceptor amino acid, and thus itsactivity can generate putative precursors of neuronal GSH(Dasgupta et al., 2007; Koga et al., 2011). GSH levels, as well asGGT and G6PD activities were determined trying to connect ourresults to deficiencies in the antioxidant defense system inhypothyroid rat brain. Thus, the inhibition in GGT activityobserved by us might diminish the reservoir of glutamatenecessary to GSH synthesis leading to a decrease in the levels ofthis important antioxidant in brain. However, this phenomenonrequires further investigation.

Cholinesterase inhibition observed in hippocampal slices fromhypothyroid animals could be related to modifications in braincognition. According to Wu et al. (2010) the increased acetylcho-line levels could be implicated in the activation of nicotinic recep-tors (nAChR) allowing the influx of both extracellular Na+ and Ca2+.Therefore, taking into account our findings, the increased 45Ca2+

uptake induced by hypothyroid condition could probably be asso-ciated with decreased acethylcholinesterase activity observedherein. Also, the Na+-coupled 14C-MeAIB accumulation into hippo-campal cells induced by hypothyroidism might cause a rapid in-crease in the intracellular Ca2+ concentration. In this context,Young et al., 2010 demonstrated the stimulation of Ca2+ signalingin response to MeAIB in STC-1 cells. It is also described that aminoacid transport through system A transporter is a major mechanismby which amino acids induce Ca2+ signaling. Taking into accountour findings, alterations in system A transporter, in glutamate up-take, in Ca2+ uptake in glutamate–glutamine cycle and in oxidativestress were probably interconnected and must be together one ofthe key factors involved in cognitive alterations induced by con-genital hypothyroidism.

Hypothyroidism during the fetal and/or the neonatal period re-sults in several histological changes in the neonatal rat brain, suchas reduced synaptic connectivity, delayed myelination, disturbedneuronal migration and modifications in neurotransmission lead-ing to deleterious effects for central nervous system development(Bernal, 2007).

The hippocampus integrates the encoding, storage and recall ofmemories, binding the spatio-temporal and sensory informationthat constitutes experience and keeping episodes in their correctcontext. The rapid and accurate processing of data relies on inter-connected networks corresponding to the anatomical subfields ofDG, CA3 and CA1 (Jones and McHugh, 2011). Hasegawa et al.,2010 showed reductions of the hippocampal and brain volume ofthe hypothyroid rats when compared to control animals. The mostsignificant morphological findings of the present study concern thedecreased density of neurons in all pyramidal cell layer of the hip-pocampus (CA1, CA2, CA3, CA4 and DG) in both hemispheres of 15-day-old hypothyroid rat hippocampus and the differential altera-tions in density of astroglial cells in each hemisphere of the brain.The decreased density of neurons in developmental hypothyroid-ism is in line with Gong et al. (2010), who found decreased numberof surviving cells in the hippocampus of young hypothyroid rats.These findings could support the cognitive impairment associatedwith hypothyroid condition in adults.

Astrocytes are the main glial cells involved in the developmentof the nervous system. These cells are involved in neuron migra-tion and maturation, myelination, ionic regulation, metabolism ofneurotransmitters and synaptic integration (Bezzi et al., 2001;Kettenmann and Ransom, 1995). In the present study we foundthat the GFAP positive cells were increased in CA1 from the lefthemisphere of hypothyroid hippocampus, however, the right onewas unaffected by the hypothyroidism. These data suggest a later-ality effect. Our results are in line with Raftogianni et al. (2012)who described that protein expression is lateralized in the hippo-campus of 13-day-old rat pups and most of the proteins, includingGFAP were up-regulated in the left hippocampus in response to anearly experience involving either receipt or denial of the expectedreward of maternal contact. The authors propose this effect couldreflect developmental changes in both hemispheres.

Since no differences were observed in the laterality of hippo-campal neuronal density, it is feasible that changes in synapticplasticity could be more directly related to the increased densityof astrocytes in the left hemisphere. Altogether, these data clearlydemonstrate the differential vulnerability of astrocyte from the dif-ferent hippocampal subregions to thyroid hormone deficit, result-ing in a compromised astroglial defense system that is probably

D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26 25

related to a misregulation in the glutamate–glutamine cycle,oxidative stress, and neurotoxic damage in hypothyroid brain.These deficiencies in hippocampal astrocyte functions appear toplay a role in the physiopathology of the neurological dysfunctionsof hypothyroidism.

Taking together our present results, we propose a hypothyroid-ism-induced glutamatergic excitotoxicity, considering the altera-tions in calcium and glutamate uptake and transporters,inasmuch as glutamate homeostasis could be somehow involvedin almost all aspects of normal and abnormal brain activities.

Conflict of interest

Authors declare that there is no conflict of interest that could beperceived as prejudicing the impartiality of the research reported.

Funding

This work was supported by grants from Conselho Nacional deDesenvolvimento Científico e Tecnológico (CNPq-Brazil)/EditalUniversal 14/2011/research Grant #479483/2011-6, Fundação deApoio à Pesquisa Científica e Tecnológica do Estado de SantaCatarina (FAPESC) research grant ‘‘Programa de Pesquisa para oSUS – PPSUS: Gestão compartilhada em Saúde – FAPESC/MS-CNPq/SES-SC 04/2009’’ #15972/2009-7, PGFAR, PPG-BQA, FUN-PESQUISA-Universidade Federal de Santa Catarina, Coordenaçãode Aperfeiçoamento de Pessoal de Nível Superior (CAPES-Brazil)and CAPES-REUNI.

Acknowledgements

We thank INTERCIENTÍFICA (São José dos Campos, SP, Brazil) forthe generous gift of G6PD kit. Special thanks to the Studies in Biol-ogy Multi-user’s Laboratories (Laboratórios Multiusuários de Estu-dos em Biologia – LAMEBs I and II) for their facilities and to all Labtechnicians for their kindly assistance. Daiane Cattani was regis-tered on Pharmacy Postgraduate Program/UFSC-Brazil (PGFAR),Vera Lúcia de Liz Oliveira Cavalli was registered on BiochemistryPostgraduate Program/UFSC-Brazil (PPGBQA).

References

Ahmad, A., Khan, M.M., Javed, H., Raza, S.S., Ishrat, T., Khan, M.B., Safhi, M.M., Islam,F., 2012. Edaravone ameliorates oxidative stress associated cholinergicdysfunction and limits apoptotic response following focal cerebral ischemiain rat. Mol. Cell. Biochem. 367, 215–225.

Baas, D., Fressinaud, C., Vitkovic, L., Sarlieve, L.L., 1998. Glutamine synthetaseexpression and activity are regulated by 3,5,3’-triodo-L-thyronine andhydrocortisone in rat oligodendrocyte cultures. Int. J. Dev. Neurosci. 16, 333–340.

Berger, R.P., Beers, S.R., Richichi, R., Wiesman, D., Adelson, P.D., 2007. Serumbiomarker concentrations and outcome after pediatric traumatic brain injury. J.Neurotrauma 24, 1793–1801.

Bernal, J., 2007. Thyroid hormone receptors in brain development and function.Nature Clin. Pract. Endocrinol. Metab 3, 249–259.

Bezzi, P., Domercq, M., Vesce, S., Volterra, A., 2001. Neuron-astrocyte cross-talkduring synaptic transmission: physiological and neuropathologicalimplications. Prog. Brain Res. 132, 255–265.

Bird, R.P., Draper, A.H., 1984. Comparative studies on different methods ofmalondyhaldehyde determination. Methods Enzymol. 90, 105–110.

Browne, R.W., Armstrong, D., 1998. Reduced glutathione and glutathione disulfide.Methods Mol. Biol. 108, 347–352.

Cookson, M.R., Shaw, P.J., 1999. Oxidative stress and motor neuron disease. BrainPathol. 9, 165–186.

Coulter, D.A., Eid, T., 2012. Astrocytic regulation of glutamate homeostasis inepilepsy. Glia 60, 1215–1226.

Danbolt, N.C., 2001. Glutamate uptake. Prog. Neurobiol. 65, 1–105.Dasgupta, A., Das, S., Sarkar, P.K., 2007. Thyroid hormone promotes glutathione

synthesis in astrocytes by up regulation of glutamate cysteine ligase throughdifferential stimulation of its catalytic and modulator subunit mRNAs. Free Rad.Biol. Med. 42, 617–626.

Donato, R., 2001. S100: a multigenic family of calcium-modulated proteins of theEF-hand type with intracellular and extracellular functional roles. Int. J.Biochem. Cell Biol. 33, 637–668.

Donato, R., 2003. Intracellular and extracellular roles of S100 proteins. Microsc. Res.Tech. 60, 540–551.

Dong, J., Yin, H., Liu, W., Wang, P., Jiang, Y., Chen, J., 2005. Congenital iodinedeficiency and hypothyroidism impair LTP and decrease C-fos and C-junexpression in rat hippocampus. Neurotoxicology 26, 417–426.

Dringen, R., Hirrlinger, J., 2003. Glutathione pathways in the brain. Biol. Chem. 384,505–516.

Drukarch, B., Schepens, E., Jongenelen, C.A.M., Stoof, J.C., Langeveld, C.H., 1997.Astrocyte-mediated enhancement of neuronal survival is abolished byglutathione deficiency. Brain Res. 770, 123–130.

Gegelashvili, G., Schousboe, A., 1997. High affinity glutamate transporters:regulation of expression and activity. Mol. Pharmacol. 52, 6–15.

Gillessen, T., Budd, S.L., Lipton, S.A., 2002. Excitatory amino acid neurotoxicity. Adv.Exp. Med. Biol. 513, 3–40.

Gonçalves, C.A., Leite, M.C., Nardin, P., 2008. Biological and methodological featuresof the measurement of S100B, a putative marker of brain injury. Clin. Biochem.41, 755–763.

Gong, J., Dong, J., Wang, Y., Xu, H., Wei, W., Zhong, J., Liu, W., Xi, Q., Chen, J., 2010.Developmental iodine deficiency and hypothyroidism impair neuraldevelopment, up-regulate caveolin-1 and down-regulate synaptophysin in rathippocampus. J. Neuroendocrinol. 22 (129–139), 2010.

Gras, G., Porcheray, F., Samah, B., Leone, C., 2006. The glutamate-glutamine cycle asan inducible, protective face of macrophage activation. J. Leukoc. Biol. 80, 1067–1075.

Gundersen, H.J., 1977. Notes on the estimation on the numerical density of arbitraryprofiles: the edge effect. J. Microsc. 111, 219–223.

Gundersen, H.J., Osterby, R., 1981. Optimizing sampling efficiency of stereologicalstudies in biology: or ‘‘do more less well’’. J. Microsc. 121, 65–73.

Halliwell, B., Gutteridge, J.M.C., 2007. Free Radicals in Biology and Medicine, fourthed. Oxford University Press, New York.

Hasegawa, M., Kida, I., Wada, H., 2010. A volumetric analysis of the brain andhippocampus of rats rendered perinatal hypothyroid. Neurosci. Lett. 479, 240–244.

Howard, C.V., Reed, M.G., 2005. Unbiased Stereology: Three-DimensionalMeasurement in Microscopy, second ed. New York BIOS - ScientificPublication, New York.

Ibarrola, N., Rodriguez-Pena, A., 1997. Hypothyroidism coordinately and transientlyaffects myelin protein gene expression in most rat brain regions duringpostnatal development. Brain Res. 752, 285–293.

Jellinger, K.A., 1999. The role of iron in neurodegeneration – prospects forpharmacotherapie of Parkinson’s disease. Drugs Aging 14, 115–140.

Jenstad, M., Quazi, A.Z., Zilberter, M., Haglerød, C., Berghuis, P., Saddique, N., Goiny,M., Buntup, D., Davanger, S., S Haug, F.M., Barnes, C.A., McNaughton, B.L.,Ottersen, O.P., Storm-Mathisen, J., Harkany, T., Chaudhry, F.A., 2009. System Atransporter SAT2 mediates replenishment of dendritic glutamate poolscontrolling retrograde signaling by glutamate. Cereb. Cortex 19, 1092–1106.

Jones, M.W., McHugh, T.J., 2011. Updating hippocampal representations: CA2 joinsthe circuit. Trends Neurosci. 34, 526–535.

Jung, C., Yabe, J.T., Shea, T.B., 2000. C-terminal phosphorylation of the heavymolecular weight neurofilament subunit is inversely correlated withneurofilament axonal transport velocity. Brain Res. 856, 12–19.

Kanamori, K., Ross, B.D., 2004. Quantitative determination of extracellularglutamine concentration in rat brain, and its elevation in vivo by system Atransport inhibitor, alpha-(methylamino)isobutyrate. J. Neurochem. 90, 203–210.

Kettenmann, H., Ransom, B.R., 1995. Neuroglia. Oxford University Press, New York.Kleindienst, A., Ross Bullock, M., 2006. A critical analysis of the role of the

neurotrophic protein S100B in acute brain injury. J. Neurotrauma 23, 1185–1200.

Koga, M., Serritella, A.V., Messmer, M.M., Hayashi-Takagi, A., Hester, L.D., Snyder,S.H., Sawa, A., Sedlak, T.W., 2011. Glutathione is a physiologic reservoir ofneuronal glutamate. Biochem. Biophys. Res. Commun. 409, 596–602.

Koohestani, F., Brown, C.M., Meisami, E., 2012. Differential effects of developmentalhypo- and hyperthyroidism on acetylcholinesterase and butyrylcholinesteraseactivity in the spinal cord of developing postnatal rat pups. Int. J. Dev. Neurosci.30, 570–577.

Lee, D.H., Blomhoff, R., Jacobs Jr., D.R., 2004. Is serum gamma glutamyltransferase amarker of oxidative stress? Free Radic. Res. 38, 535–539.

Leite, M.C., Galland, F., Brolese, G., Guerra, M.C., Bortolotto, J.W., Freitas, R., Almeida,L.M., Gottfried, C., Goncalves, C.A., 2008. A simple, sensitive and widelyapplicable ELISA for S100B: methodological features of the measurement ofthis glial protein. J. Neurosci. Methods. 169, 93–99.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurementwith the Folin phenol reagent. J. Biol. Chem. 193, 265–267.

Orlowsky, M., Meister, A., 1963. Gamma-glutamyl-p-nitroanilide: a new convenientsubstrate for determination and study of l- and D-gamma-glutamyltranspeptidase activities. Biochim. Biophys. Acta 73, 679–681.

Mandarim-de-Lacerda, C.A., Fernandes-Santos, C., Aguila, M.B., 2010. Image analysisand quantitative morphology. Methods Mol. Biol. 611, 211–225.

Markesbery, W.R., Carney, J.M., 1999. Oxidative alterations in Alzheimer’s disease.Brain Pathol. 9, 133–146.

Maschmann, J., Erb, M.A., Heinemann, M.K., Ziemer, G., Speer, C.P., 2000. Evaluationof protein S-100 serum concentrations in healthy newborns and sevennewborns with perinatal acidosis. Acta Paediatr. 89, 553–555.

Mullen, R.J., Buck, C.R., Smith, A.M., 1992. NeuN, a neuronal specific nuclear proteinin vertebrates. Development 116, 201–211.

26 D. Cattani et al. / Molecular and Cellular Endocrinology 375 (2013) 14–26

Murabayashi, M., Minato, M., Okuhata, Y., Makimoto, M., Hosono, S., Masaoka, N.,Okada, T., Yamamoto, T., Mugishima, H., Takahashi, S., Harada, K., 2008. Kineticsof serum S100B in newborns with intracranial lesions. Pediatr. Int. 50, 17–22.

Nakamura, K., Wang, W., Kang, U.J., 1997. The role of glutathione in dopaminergicneuronal survival. J. Neurochem. 69, 1850–1858.

Paxinos, G., Watson, C., 2008. The Rat Brain in Stereotaxic Coordinates, sixth ed.Academic Press, San Diego.

Pierozan, P., Zamoner, A., Soska, Â.K., de Lima, B.O., Reis, K.P., Zamboni, F., Wajner,M., Pessoa-Pureur, R., 2012. Signaling mechanisms downstream of quinolinicacid targeting the cytoskeleton of rat striatal neurons and astrocytes. Exp.Neurol. 233, 391–399.

Portela, L.V., Tort, A.B., Schaf, D.V., Ribeiro, L., Nora, D.B., Walz, R., Rotta, L.N., Silva,C.T., Busnello, J.V., Kapczinski, F., Gonçalves, C.A., Souza, D.O., 2002. The serumS100B concentration is age dependent. Clin. Chem. 48, 950–952.

Raftogianni, A., Stamatakis, A., Papadopoulou, A., Vougas, K., Anagnostopoulos, A.K.,Stylianopoulou, F., Tsangaris Th, G., 2012. Effects of an early experience ofreward through maternal contact or its denial on laterality of proteinexpression in the developing rat hippocampus. PlosOne 7, 1–16.

Rothstein, J.D., Martin, L.J., Kuncl, R.W., 1992. Decreased glutamate transport by thebrain and spinal cord in amyotrophic lateral sclerosis. N. Engl. J. Med. 326,1464–1468.

Rothstein, J.D., Van Kammen, M., Levey, A.I., Martin, L.J., Kuncl, R.W., 1995. Selectiveloss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann.Neurol. 38, 73–84.

Sánchez-Alonso, J.L., Sánchez-Aguilera, A., Vicente-Torres, M.A., Colino, A., 2012.Intrinsic excitability is altered by hypothyroidism in the developinghippocampal CA1 pyramidal cells. Neuroscience 207, 37–51.

Seki, E., Brenner, D.A., Karin, M., 2012. A liver full of JNK: signaling in regulation ofcell function and disease pathogenesis, and clinical approaches.Gastroenterology 143, 307–320.

Sherrin, T., Blank, T., Todorovic, C., 2011. C-Jun N-terminal kinases in memory andsynaptic plasticity. Rev Neurosci. 22, 403–410.

Sihag, R.K., Inakaki, M., Yamagushi, T., Shea, T.B., Pant, H.C., 2007. Role ofphosphorylation on the structural dynamics and function of types III and IVintermediate filaments. Exp. Cell Res. 313, 2098–2109.

Silva, F.R., Leite, L.D., Barreto, K.P., D’Agostini, C., Zamoner, A., 2001. Effect of 3,5,3’-triiodo-L-thyronine on amino acid accumulation and membrane potential inSertoli cells of the rat testis. Life Sci. 69, 977–986.

Tanaka, Y., Marumo, T., Omura, T., Yoshida, S., 2008. Early increases in serum S100Bare associated with cerebral hemorrhage in a rat model of focal cerebralischemia. Brain Res. 1227, 248–254.

Tosic, M., Torch, S., Comte, V., Dolivo, M., Honegger, P., Matthieu, J.M., 1992.Triiodothyronine has diverse and multiple stimulating effects on expression ofthe major myelin protein genes. J. Neurochem. 59, 1770–1777.

Tramontina, F., Conte, S., Gonçalves, D., Gottfried, C., Portela, L.V., Vinade, L., Salbego,C., Gonçalves, C.A., 2002. Developmental changes in S100B content in braintissue, cerebrospinal fluid, and astrocyte cultures of rats. Cell. Mol. Neurobiol.22, 373–378.

Tramontina, F., Leite, M.C., Gonçalves, D., Tramontina, A.C., Souza, D.F., Frizzo, J.K.,Nardin, P., Gottfried, C., Wofchuk, S.T., Gonçalves, C.A., 2006. High glutamatedecreases S100B secretion by a mechanism dependent on the glutamatetransporter. Neurochem. Res. 31, 815–820.

Vara, H., Muñoz-Cuevas, J., Colino, A., 2003. Age-dependent alterations of long-termsynaptic plasticity in thyroid-deficient rats. Hippocampus 13, 816–825.

Wu, P.C., Fann, M.J., Kao, L.S., 2010. Characterization of Ca2+ signaling pathways inmouse adrenal medullary chromaffin cells. J. Neurochem. 112, 1210–1222.

Young, S.H., Rey, O., Sternini, C., Rozengurt, E., 2010. Amino acid sensing byenteroendocrine STC-1 cells: role of the Na+-coupled neutral amino acidtransporter 2. Am. J. Physiol. Cell Physiol. 298, C1401–13.

Zamoner, A., Royer, C., Barreto, K.P., Pessoa-Pureur, R., Silva, F.R., 2007. Ionicinvolvement and kinase activity on the mechanism of nongenomic action ofthyroid hormones on 45Ca2+ uptake in cerebral cortex from young rats.Neurosci. Res. 57, 98–103.

Zamoner, A., Heimfarth, L., Pessoa-Pureur, R., 2008a. Congenital hypothyroidism isassociated with intermediate filament misregulation, GLAST glutamatetransporters down-regulation and MAPK activation in developing rat brain.Neurotoxicology 29, 1092–1099.

Zamoner, A., Barreto, K.P., Filho, D.W., Sell, F., Woehl, V.M., Guma, F.C., Pessoa-Pureur, R., Silva, F.R., 2008b. Propylthiouracil-induced congenitalhypothyroidism upregulates vimentin phosphorylation and depletesantioxidant defenses in immature rat testis. J. Mol. Endocrinol. 40, 125–135.

Zamoner, A., Pessoa-Pureur, R., 2011. Nongenomic actions of thyroid hormones:every why has a wherefore. Immunol. Endocr. Metab. Agents Med. Chem. 11,165–178.

Zhang, H.M., Lin, N., Dong, Y., Su, Q., Luo, M., 2011. Effect of perinatal thyroidhormone deficiency on expression of rat hippocampal conventional proteinkinase C isozymes. Mol. Cell Biochem. 353, 65–71.

Zoeller, R.T., Rovet, J., 2004. Timing of thyroid hormone action in the developingbrain: clinical observations and experimental findings. J. Neuroendocrinol. 16,809–818.