iences 260 (2007) 87–94www.elsevier.com/locate/jns

Journal of the Neurological Sc

Effect of the branched-chain α-keto acids accumulating in maple syrupurine disease on S100B release from glial cells

Cláudia Funchal ⁎, Francine Tramontina, André Quincozes dos Santos, Daniela Fraga de Souza,Carlos Alberto Gonçalves, Regina Pessoa-Pureur, Moacir Wajner

Universidade Federal do Rio Grande do Sul, Instituto de Ciências Básicas da Saúde, Departamento de Bioquímica,Rua Ramiro Barcelos 2600 anexo, 90035-003 Porto Alegre, RS, Brazil

Received 15 January 2007; received in revised form 6 March 2007; accepted 9 April 2007Available online 17 May 2007

Abstract

Accumulation of the branched-chain α-keto acids (BCKA), α-ketoisocaproic acid (KIC), α-keto-β-methylvaleric acid (KMV) and α-ketoisovaleric acid (KIV) and their respective branched-chain α-amino acids (BCAA) occurs in tissues and biological fluids of patientsaffected by the neurometabolic disorder maple syrup urine disease (MSUD). The objective of this study was to verify the effect of the BCKAon S100B release from C6 glioma cells. The cells were exposed to 1, 5 or 10 mM BCKA for different periods and the S100B release wasmeasured afterwards. The results indicated that KIC and KIV, but not KMV, significantly enhanced S100B liberation after 6 h of exposure.Furthermore, the stimulatory effect of the BCKA on S100B release was prevented by coincubation with the energetic substrate creatine andwith the N-nitro-L-arginine methyl ester (L-NAME), a nitric oxide synthase inhibitor, indicating that energy deficit and nitric oxide (NO)were probably involved in this effect. Furthermore, the increase of S100B release was prevented by preincubation with the protein kinaseinhibitors KN-93 and H-89, indicating that KIC and KIV altered Ca2+/calmodulin (PKCaMII)- and cAMP (PKA)-dependent protein kinasesactivities, respectively. In contrast, other antioxidants such as glutathione (GSH) and trolox (soluble vitamin E) were not able to prevent KIC-and KIV-induced increase of S100B liberation, suggesting that the alteration of S100B release caused by the BCKA is not mediated byoxidation of sulfydryl or other essential groups of the enzyme as well as by lipid peroxyl radicals. Considering the importance of S100B forbrain regulation, it is conceivable that enhanced liberation of this protein by increased levels of BCKA may contribute to theneurodegeneration characteristic of MSUD patients.© 2007 Elsevier B.V. All rights reserved.

Keywords: Branched-chain α-keto acids; Maple syrup urine disease; S100B; Antioxidants; Creatine; Kinase inhibitors

1. Introduction

Mammalian mitochondrial branched-chain α-keto aciddehydrogenase complex (BCKD) catalyzes the oxidativedecarboxylation of the branched-chainα-keto acids (BCKA),α-ketoisocaproic acid (KIC), α-keto-β-methylvaleric acid(KMV), and α-ketoisovaleric acid (KIV), which originatetheir corresponding branched-chain amino acids (BCAA)leucine, isoleucine and valine, respectively. In patients with

⁎ Corresponding author. Tel.: +55 51 3316 5535; fax: +55 51 3316 5564.E-mail address: csfunchal@yahoo.com.br (C. Funchal).

0022-510X/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.jns.2007.04.011

maple syrup urine disease (MSUD), or branched-chain ketoaciduria, the activity of the BCKD is severely deficient. Themetabolic blockage at this step results in the accumulation ofBCKA and BCAA in tissues of the affected individuals [1,2].

The symptomatology of MSUD includes ketoacidosis,failure to thrive, poor feeding, apnea, ataxia, seizures, coma,psychomotor delay and mental retardation [3,1]. Neuro-pathologic findings characteristic of the disease are cerebraledema, atrophy of the cerebral hemispheres, spongydegeneration of the white matter and delayed myelination[4,1]. MSUD usually presents as a heterogeneous clinicalphenotype, ranging from the severe classical form to mildvariants, possibly due to distinct residual enzyme activity

88 C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

[5]. Although the mechanisms of brain damage in MSUDare still unclear, it appears that leucine and KIC areconsidered to be the main neurotoxic metabolites sinceincreased plasma concentrations of these metabolites areassociated with the appearance of neurological symptoms[6,1]. In this context, it has been demonstrated that theBCKA may affect energy metabolism in rat brain [7–15].Furthermore, it has been postulated that competition of KIV,KIC, and their hydroxyderivatives with L-glutamate fordecarboxylation and the consequent reduction of γ-amino-butyric acid (GABA) concentration [16], impairment ofmyelin development [17–19,4], low plasma and brain levelsof essential amino acids [20,21], oxidative stress [22–25]and reduced brain uptake of essential amino acids leading todecreased neurotransmitter synthesis also contribute to braininjury [26].

Glial cells play a vital role in the homeostatic regulationof the central nervous system (CNS). Astrocytes, the mostabundant glial cell type in the brain, are involved inneurotransmitter uptake, neuronal metabolic support, pHregulation, and protection against toxic episodes [27,28]. Inaddition, glial cells preserve neuronal survival throughinactivation of reactive oxygen species (ROS) [29]. Becauseof its similarities with primary glial cells in culture, the C6 ratglioma cell line has provided a useful model to study glialcell properties, glial growth factors and sensitivity of glialcells to various substances and conditions [30,31].

S100B is a calcium-binding protein of 21 kDa expressedprimarily in astrocytes and secreted by these cells [32]. Manystudies have suggested an intracellular role for this proteinparticularly in the regulation of cytoskeleton and cell cyclesuch as modulation of CapZ (an actin-binding protein) [33],microtubules [34–36], and glial fibrillary acidic protein(GFAP) [37,35]. Moreover, extracellular S100B stimulatesglial proliferation, survival of neurons, and extension ofneurites in cell culture [32]. On the other hand, abnormalexpression of S100B has supported its role in severalneurodegenerative disorders, including Alzheimer's disease[38].

Although several studies demonstrate a relationshipbetween synaptic plasticity and S100B, there are few reportson S100B and glutamatergic transmission. It has been shownthat toxic levels of glutamate reduce S100B secretion inhippocampal astrocytes and brain slices [39,40]. Morerecently, we have shown that S100B secretion is increasedwhen glutamate uptake is reduced [41]. Interestingly, adecrease of brain glutamate has been observed in MSUDpatients during a period of metabolic decompensation [42]. Inaddition, we found that the BCKA impairs glutamate uptake[43], favoring brain damage in excitotoxic conditions.

Therefore, considering that excitotoxicity, energy deple-tion and oxidative stress are involved in the neuropatholo-gical features observed in patients affected by MSUD in thisstudy we investigated the effect of BCKA on S100B releasefrom C6 glioma cells. We also tested the influence of theenergetic substrate creatine, various antioxidants [glu-

tathione (GSH), α-tocopherol (trolox) and Nω-nitro-L-arginine methyl ester (L-NAME)] and kinase inhibitors (H-89 and KN-93) on the effects elicited by the BCKA.

2. Materials and methods

2.1. Materials

α-Ketoisocaproic acid, α-keto-β-methylvaleric acid, α-ketoisovaleric acid, N-[2-hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid] (HEPES), creatine, glutathione (GSH),trolox, Nω-nitro-L-arginine methyl ester (L-NAME), anti-body anti-S100B (SH-B1) and material for cell culture werepurchased from Sigma (St. Louis MO, USA). Dulbecco'smodified Eagle's medium (DMEM) was purchased fromGibco BRL (Carlsbad, CA, USA) and fetal bovine serum(FBS) from Cultilab (Campinas, SP, Brazil). KN-93 and H-89 were purchased from Calbiochem (La Jolla, CA, USA).Peroxidase-conjugated anti-S100B were obtained fromDAKO (Sao Paulo, Brazil).

2.2. Cell culture

C6 glioma cells were purchased from American TypeCulture Collection (Rockville, Maryland, USA). Cells wereseeded in flasks and cultured in DMEM (pH 7.4)supplemented with 5% fetal bovine serum (FBS), 2.5 mg/ml Fungizone® and 100 U/l gentamicin. Cells were kept at atemperature of 37 °C, a minimum relative humidity of 95%,and an atmosphere of 5% CO2 in air. Exponentially growingcells were detached from the culture flasks using 0.25%trypsin/ethylene-diaminetetraacetic acid (EDTA) and seededin 24-well plates (5×103 cells/well).

2.3. Incubation

After cells reached confluence, the culture medium wasremoved by suction and the cells were incubated for 6, 12 or24 h at 37 °C in an atmosphere of 5% CO2/95% air inDMEM (pH 7.4) containing 0% FBS in the absence(controls) or presence of KIC, KMV or KIV (1, 5 or10 mM). In some experiments the cells were coincubatedwith BCKA plus 5 mM creatine, 1 mM GSH, 10 μM Troloxor 0.5 mM L-NAME or preincubated for 20 min with 10 μMKN-93 or 10 μM H-89. C6 glioma cells were also incubatedin the presence of the BCAA leucine, valine and isoleucine at10 mM concentration.

2.4. S100B release

S100B concentrations were determined in the mediumafter the exposure of the cells to the metabolites and drugs asdescribed above. The measurement was carried out byELISA, as previously described [44]. Briefly, 50 μl ofsample plus 50 μl of barbital buffer were incubated for 3 h ona microtiter plate previously coated with monoclonal anti-

Fig. 1. Effect of α-ketoisocaproic acid (KIC) (A), α-keto-β-methylvalericacid (KMV) (B) and α-ketoisovaleric acid (KIV) (C) on S100B release fromcultivated C6 glioma cells. Results represent mean±SEM of 4–5independent experiments and are expressed in percentage of controls(0.1601 ng S100B/ml). Statistically significant difference determined byone-way ANOVA followed by the Tukey test is indicated: ⁎Pb0.001, ascompared to controls.

Fig. 2. Effect creatine (5 mM) on the increase of S100B release from C6glioma cells provoked by 10 mM α-ketoisocaproic acid (KIC) or α-ketoisovaleric acid (KIV). Results represent mean±SEM of 4–6 indepen-dent experiments and are expressed in percentage of controls (0.1589 ngS100B/ml). Statistically significant difference determined by one-wayANOVA followed by the Tukey test is indicated: ⁎Pb0.001, as comparedto controls.

89C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

S100B (SH-B1). Peroxidase-conjugated anti-S100 was thenadded and incubated for 1 h. The color reaction with o-phenylenediamine was measured at 492 nm. The standardS100B curve ranged from 0.1 to 4 ng/ml.

2.5. Protein determination

Protein concentrations were determined by the method ofLowry et al. (1951) [45], using bovine serum albumin asstandard.

2.6. Statistical analysis

Data from the experiments are presented as mean±SEMand were analyzed statistically by one-way analysis ofvariance (ANOVA) followed by the Tukey test when the Fvalue was significant. Values of Pb0.05 were considered tobe significant. All analyses were carried out in an IBMcompatible PC using the Statistical Package for SocialSciences (SPSS) software.

3. Results

We initially verified that S100B concentrations (ngS100B/ml) in the cultured medium increased when C6glioma cells were exposed for 6 h (0.1594±SEM 0.0092),12 h (0.2770±SEM 0.0084) and 24 h (0.4167±SEM0.0025). We then tested the influence of 1, 5 and 10 mMKIC, KMV or KIV on S100B release from C6 glioma cellsduring various incubation periods. We verified that 10 mMof KIC or KIV significantly increased (190%) S100Bconcentrations in the incubation medium when C6 gliomacells were exposed to these metabolites for 6 h (Fig. 1A andC). In contrast, KMV was not able to alter S100Bconcentrations in the incubation medium (Fig. 1B). Thefigure only shows the data of 6 h incubation but similarresults were obtained with 12 and 24 h incubation (resultsnot shown). Interestingly, the presence of BCAA for up to24 h at concentrations as high as 10 mM did not induce anyalteration of S100B release by the cultured cells (results notshown).

We further observed that coincubation of 5 mM creatinewith 10 mM KIC or 10 mM KIV for 6 h prevented theBCKA-induced increase of S100B release, indicating thatlack of creatine/or phosphocreatine, reflecting energydeprivation, is probably involved in this effect (Fig. 2).Identical results were obtained with 12 and 24 h incubation(results not shown).

We also tested the influence of various antioxidants,namely GSH, trolox and L-NAME, on the increased S100Brelease induced by KIC and KIV. We observed thatcoincubation of C6 cells with 1 mM GSH (a naturallyoccurring thiol-reducing agent) or 10 μM trolox (solublevitamin E) and 10 mM of each BCKA (KIC or KIV) did notprevent the enhanced S100B release caused by these ketoacids (Fig. 3A and B). In contrast, coincubation of C6 cellswith 0.5 mM L-NAME, a nitric oxide (NO) synthaseinhibitor, and 10 mM KIC or KIV fully prevented theincrease of S100B release caused by these metabolites at 6 hincubation, indicating that NO is partially involved in such

Fig. 4. Effect of the kinase inhibitors KN-93 (A) or H-89 (B) on theenhanced S100B release from C6 glioma cells provoked by 10 mM α-ketoisocaproic acid (KIC) or α-ketoisovaleric acid (KIV). Results representmean±SEM of 4–6 independent experiments and are expressed inpercentage of controls (6 h=0.1577 ng S100B/ml; 12 h=0.2801 ngS100B/ml; 24 h=0.4127 ng S100B/ml). Statistically significant differencedetermined by one-way ANOVA followed by the Tukey test is indicated:∗Pb0.001, as compared to controls.

Fig. 3. Effect the antioxidants glutathione (GSH, 1 mM) (A), trolox (10 μM)(B) and L-NAME (0.5 mM) (C) on the stimulatory effect of 10 mM α-ketoisocaproic acid (KIC) or α-ketoisovaleric acid (KIV) on S100B releasefrom cultivated C6 glioma cells. Results represent mean±SEM of 4–6independent experiments and are expressed in percentage of controls(6 h=0.1627 ng S100B/ml; 12 h=0.2759 ng S100B/ml; 24 h=0.4271 ngS100B/ml). Statistically significant difference determined by one-wayANOVA followed by the Tukey test is indicated: ∗Pb0.001, as comparedto controls.

90 C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

effects (Fig. 3C). However, when the cells were exposed forlonger periods (12 or 24 h) to L-NAME and the BCKA, thisantioxidant was not capable to prevent the increased S100Brelease caused by KIC or KIV (Fig. 3C).

Next, we investigated whether protein kinase A (PKA) orprotein kinase dependent of calcium and calmodulin II(PKCaMII) were involved in the enhanced secretion ofS100B provoked by KIC and KIV in C6 cells. We pre-incubated these cells with the specific protein kinaseinhibitors H-89 (10 μM) or KN-93 (10 μM), inhibitors ofPKA and PKCaMII, respectively, followed by incubation ofthe cells with 10 mM KIC or KIV. Results showed that bothinhibitors prevented the increased S100B release caused bythe BCKA when incubation proceeded for 6 and 12 h (Fig.4A and B). However, KN-93 was not able to prevent sucheffect after 24 h of incubation (Fig. 4A). These resultssuggest that the increase of S100B release caused by theseorganic acids is mediated by PKA and PKCaMII.

4. Discussion

In the present study we investigated the effect of the ketoacids accumulating in MSUD on S100B release fromcultivated C6 rat glioma cells, which have similar propertiesto primary glial cells [30,46–49]. It should be noted that glialcell viability is reduced in some neurodegenerative diseases[50,51] and S100B expression, particularly extracellularS100B, is used as a parameter of glial death in pathologicalsituations with brain injury. In fact, elevation of peripheral(cerebrospinal fluid or serum) S100B concentrations hasbeen observed in several common neurological andpsychiatric disorders [52], as well as in phenylketonuria[53] and in hyperammonemia [54]. In this context, our grouphas shown that elevated extracellular ammonia inducesS100B secretion in astrocyte cultures [55]. The pathophy-siological significance of S100B elevation is not wellunderstood since it could represent a glial response toprotect neurons from damage (S100B is a neurotrophiccytokine), or, conversely, reflect glial injury or death oncehigh levels of this protein are per se apoptotic to neural cells[38]. In summary, S100B increase may reflect a direct orindirect neurological damage.

Although some secretagogos of S100B have beenidentified, including forskolin, lysophosphatidic acid, ser-otonin and glutamate, the mechanisms of S100B secretionstill remain poorly established [41]. In this study we show,for the first time, that elevated extracellular levels of KIC and

91C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

KIV, metabolites accumulating in the neurometabolicdisorder MSUD, alter the basal S100B secretion in C6cultivated glioma cells. In contrast, KMV does not alter thisparameter, indicating that the effects elicited by KIC andKIV were selective, rather a non-specific effect of BCKA.

The conversion of the BCKA to the BCAA may occur inthe brain tissue because of the relatively high level of theamino acid transaminases in this tissue [56,57]. However, itseems unlikely that our present results should be attributed tothe respective BCAA, namely leucine and valine. This isbased on the fact that we could not detect increase of S100Bdue to the presence of 10 mM concentration of these aminoacids in the cell cultures for periods up to 24 h. Furthermore,we cannot exclude the possibility that a reduction ofintracellular glutamate levels could be at least in partresponsible for the increased S100B released by thecultivated cells, since it has been well demonstrated thatthe BCKA reduce glutamate concentrations in the brain [11].

Although the mechanisms responsible for the neurologi-cal dysfunction in MSUD are not well defined, variousreports indicate that energy depletion and oxidative stressare probably involved in the brain injury observed in theaffected patients [1,6–9,13,14,22–25,58–60]. Therefore, wetested whether these mechanisms could be involved in theBCKA-induced increase of S100B secretion by cultivatedC6 glioma cells. We verified that creatine totally preventedthe increase of S100B secretion induced by KIC and KIV,suggesting that the effect of these keto acids was probablymediated by a reduction of cellular energy charge. In thiscontext, increasing evidence has demonstrated that creatineaffords neuroprotection [61–64]. The mechanisms impli-cated in creatine neuroprotection are controversial butappear to at least in part depend on an increase in cellularphosphocreatine stores [65]. Phosphocreatine can to acertain extent prevent the dangerous fall of ATP levels bydonating its phosphate group to ADP to resynthesize ATPeven in the absence of oxygen and glucose [66,67]Furthermore, brain phosphocreatine concentrations can besubstantially increased by treatment with creatine. In fact,when excess creatine is available, the creatine kinasereaction is shifted toward the synthesis of phosphocreatine[68]. Moreover, an antioxidant action by creatine has alsobeen reported [69], and this, too, could participate inneuroprotection by creatine. On the other hand, it has beenrecently described that creatine may behave as a neuromo-dulator and more precisely as a cotransmitter in the brainthat modulates the functioning of postsynaptic GABAA

receptors, exerting therefore a neuroprotection [70]. Inter-estingly, KIC blood increased concentrations are associatedto encephalopatic crises with seizures in MSUD patients [6]and convulsions were induced in rats by intrastriataladministration of KIV [71]. Therefore, our present resultsreinforce the concept of a neuroprotective role of creatinecaused by KIC and KIV.

Our group has recently described that the BCKA thataccumulate in MSUD were able to induce morphological

alterations and oxidative stress in C6 cells that wereprevented by the use of antioxidants [72]. Therefore, weinvestigated the effect of various antioxidants on theenhanced effect of KIC and KIV on S100B release in orderto test whether this effect could be also mediated byoxidative stress. We verified that GSH, that acts as anaturally occurring thiol-reducing agent [73], and trolox(soluble vitamin E), which is an excellent trapping agent forlipid peroxyl radicals [74], were not able to prevent theBCKA effect. These data suggest that the alteration ofS100B release caused by the BCKA is not mediated byoxidation of sulfydryl or other essential groups of theenzyme as well as by lipid peroxyl radicals. Otherwise, thenitric oxide synthase inhibitor L-NAME fully prevented thestimulatory effect of the BCKA on S100B release from C6glioma cells after 6 h incubation, indicating that NO and/orits toxic derivative peroxynitrite are probably involved inthese effects. Interestingly, these nitrogen reactive speciesare known inhibitors of respiratory chain complexes [75,76],suggesting a possible relationship between NO formationand the protective actions observed with the energeticsubstrate and antioxidant agent creatine. On the other hand,L-NAME was unable to prevent KIV and KIC-stimulatoryaction on S100B secretion when the glial cells wereincubated for longer periods (12 and 24 h). It could betherefore presumed that the amount of L-NAME in theculture medium was insufficient to overcome NO and/orperoxynitrite synthesis by C6 cells.

Both calcium mobilization and cAMP accumulation havebeen also proposed as mediators of S100B release [39,77],implying that these mechanisms could be involved in thestimulation of S100B release induced by KIC and KIV. Aprevious study from our group demonstrated that KICincreased the phosphorylation of cytoskeletal proteins incerebral cortex of rats via Ca2+/calmodulin (PKCaMII)- andcAMP (PKA)-dependent protein kinases activities [78].Therefore, we also studied the effect of the kinase inhibitorsKN-93 and H-89 on the increased S100B release caused byKIC and KIV. The PKA inhibitor H-89 fully prevented theincrement of extracellular S100B observed at 6, 12 and 24 hincubation, indicating the involvement of cAMP and PKAin the increased secretion of this protein caused by theBCKA. Moreover, KN-93, the inhibitor of PKCaMII, alsoprevented this increment, but only in the first 12 h.Considering that the signal pathways mediated by cAMPand calcium are interrelated [79], our present results indicatethat Ca2+/calmodulin (PKCaMII)- and cAMP (PKA)-dependent protein kinases activities are also involved inthe increased S100B release caused by the BCKA. More-over, it has been recently demonstrated that cAMP andcalcium are able to activate nNOS via protein kinases [80]and recent evidence suggests that cAMP induce eNOSactivation through PKA [81].

As regards the pathophysiological significance of ourpresent data showing an increase of S100B release causedby 10 mM KIC and KIV, it is known that the levels of these

92 C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

keto acids may reach milimollar concentrations in theblood and CSF [1]. Considering that these concentrationsmay be higher in neural cells, as occurs with other or-ganic acids accumulating in various organic acidemias inwhich these compounds can be trapped by neural cells[82,83], it is possible that our results may be of relevanceto explain at least in part the neurological damage ofMSUD patients.

In summary, our results show for the first time thatthe BCKA accumulating in MSUD enhance S100B releasein C6 glioma cells. We also observe that these effects areprobably mediated by energetic deficit, production ofreactive nitrogen species and alterations on the phosphoryla-tion system since they are prevented by creatine, L-NAMEand kinase inhibitors. Since KIC and KIV are considered tobe potent neurotoxins related to the neurological symptomsin MSUD [1,23–25,56,57,69], it seems desirable to measureS100B peripheral levels in CSF and serum of affectedpatients to test whether this protein may serve as a marker ofbrain toxicity in this disease, as occurs in pathologicalconditions with brain injury [53,54]. Finally, considering thatthe current diet therapy is apparently insufficient to maintainnormal brain development and functioning in MSUDpatients, it may be interesting to test whether creatine andantioxidant supplementation to the diet would benefit thesepatients.

Acknowledgements

This work was supported by Conselho Nacional deDesenvolvimento Científico e Tecnológico (CNPq), Funda-ção de Amparo à Pesquisa do Estado do Rio Grande do Sul(FAPERGS) and Pró Reitoria de Pesquisa da UniversidadeFederal do Rio Grande do Sul (PROPESq-UFRGS).

References

[1] Chuang DT, Shih VE. Maple syrup urine disease (branched-chainketoaciduria). In: Scriver CR, Beaudet AL, Sly WL, Valle D, editors.The metabolic and molecular bases of inherited disease. 8th ed. NewYork: McGraw-Hill; 2001. p. 1971–2005.

[2] Ogier de Baulny H, Saudubray JM. Branched-chain organic acidurias.Semin Neonatol 2002;7:65–74.

[3] Nyhan WL. Maple syrup urine disease: branched-chain ketoaciduria.In: Nyhan WL, editor. Abnormalities in amino acid metabolism inclinical medicine. Norwalk: Appleton-Century-Crofts; 1984. p. 21–35.

[4] Treacy E, Clow CL, Reade TR, Chitayat D, Mamer OA, Scriver CR.Maple syrup urine disease: interrelationship between branched chainamino-, oxo-, and hydroxyacids; implications for treatment; associa-tion with CNS dysmelination. J Inherit Metab Dis 1992;15:121–35.

[5] Schadewaldt P, Wendel U. Metabolism of branched-chain amino acidsin maple syrup urine disease. Eur J Pediatr 1997;156:62–6.

[6] Snyderman SE, Norton PM, Roitman E. Maple syrup urine diseasewith particular reference to diet therapy. Pediatrics 1964;34:454–72.

[7] Howell RK, Lee M. Influence of alpha-ketoacids on the respiration ofbrain in vitro. Proc Soc Exp Biol Med 1963;113:660–3.

[8] Halestrap A, Brand MD, Denton RM. Inhibition of mitochondrialpyruvate transport by phenylpyruvate and α-ketoisocaproate. BiochimBiophys Acta 1974;367:102–8.

[9] Land JM, Mowbray J, Clark JB. Control of pyruvate and β-hydroxy-butyrate utilization in rat brain mitochondria and its relevance tophenylketonuria and maple syrup urine disease. J Neurochem1976;26:823–30.

[10] Danner DJ, Elsas JL. Disorders of branched chain amino acid and ketoacid metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D,editors. The metabolic basis of inherited disease. New York: Mc GrawHill; 1989. p. 671–92.

[11] Yudkoff M, Daikhin Y, Lin ZP, Nissin I, Stern J, Pleasure D, et al.Interrelationships of leucine and glutamate metabolism in culturedastrocytes. J Neurochem 1994;62:1192–11202.

[12] Zielke HR, Zielke CL, Baab PJ, Collins RM. Large neutral amino acidsauto exchange when infused by microdialysis into the rat brain:implications for maple syrup urine disease and phenylketonuria.Neurochem Int 2002;40:347–54.

[13] Pilla C, Cardozo RF, Dutra-Filho CS, Wyse AT, Wajner M,Wannmacher CM. Creatine kinase activity from rat brain is inhibitedby branched-chain amino acids in vitro. Neurochem Res 2003;28:675–9.

[14] Sgaravati AM, Rosa RB, Schuck PF, Ribeiro CAJ, Wannmacher CMD,Wyse ATS, et al. Inhibition of brain energy metabolism by the α-ketoacids accumulating in maple syrup urine disease. Biochim BiophysActa 2003;1639:232–8.

[15] Funchal C, Schuck PF, Santos AQ, Jacques-Silva MC, Gottfried C,Pessoa-Pureur R, et al. Creatine and antioxidant treatment prevent theinhibition of creatine kinase activity and the morphological alterationsof C6 glioma cells induced by the branched-chain alfa-keto acidsaccumulating in maple syrup urine disease. Cell Mol Neurobiol 2006;26:67–79.

[16] Tashian R. Inhibition of brain glutamic acid descarboxylase byphenylalanine, leucine and valine derivates: a suggestion concerningthe neurological defect in phenylketonuria and branched-chain ketoaciduria. Metabolism 1961;10:393–400.

[17] Appel SH. Inhibition of brain protein synthesis: an approach to abiochemical basis of neurological dysfunction in the amino-acidurias.Trans N YAcad Sci 1966;29:63–70.

[18] Taketomi T, Kunishita T, Hara A, Mizushima S. Abnormal protein andlipidic composition of the cerebral myelin of a patient with maple syrupurine disease. Jpn J Exp Med 1983;53:109–16.

[19] Tribble D, Shapira R. Myelin proteins: degradation in rat brain initiatedby metabolites causative of maple syrup urine disease. BiochemBiophys Res Commun 1983;114:440–6.

[20] Wajner M, Vargas CR. Reduction of plasma concentrations of largeneutral amino acids in patients with maple urine disease during crises.Arch Dis Child 1999;80:579.

[21] Wajner M, Coelho DM, Barschak AG, Araújo PR, Pires RF, LulhierFLG, et al. Reduction of large neutral amino acid concentration inplasma and CSF of patients with maple syrup urine disease duringcrises. J Inherit Metab Dis 2000;23:505–12.

[22] Fontella FU, Gassen E, Pulrolnik V, Wannmacher CMD, Klein AB,Wajner M, et al. Stimulation of lipid peroxidation in vitro in rat brainby the metabolites accumulating in maple syrup urine disease. MetabBrain Dis 2002;17:47–54.

[23] Bridi R, Araldi J, Sgarbi MB, Testa CG, Durigon K, Wajner M, et al.Induction of oxidative stress in rat brain by the metabolites accu-mulating in maple syrup urine disease. Int J Dev Neurosci 2003;21:327–32.

[24] Bridi R, Braun CA, Zorzi GK, Wannmacher CM, Wajner M, Lissi EG,et al. Alpha-keto acids accumulating in maple syrup urine diseasestimulate lipid peroxidation and reduce antioxidant defences incerebral cortex from young rats. Metab Brain Dis 2005;20:155–67.

[25] Bridi R, Latini A, Braum CA, Zorzi GK, Wajner M, Lissi E, et al.Evaluation of the mechanisms involved in leucine-induced oxidativedamage in cerebral cortex of young rats. Free Radic Res 2005;39:71–9.

[26] Araújo P, Wassermann GF, Tallini K, Furlanetto V, Vargas CR,Wannmacher CM, et al. Reduction of large neutral amino acid level in

93C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

plasma and brain of hyperleucinemic rats. Neurochem Int 2001;38:529–37.

[27] Blanc EM, Bruce-Keller AJ, Mattson MP. Astrocytic gap junctionalcommunication decreases neuronal vulnerability to oxidative stress-induced disruption of Ca2+ homeostasis and cell death. J Neurochem1998;70:958–70.

[28] Takuma K, Baba A, Matsuda T. Astrocyte apoptosis: implications forneuroprotection. Prog Neurobiol 2004;72:111–27.

[29] Kobayashi K, Hayashi M, Nakano H, Fukutani Y, Sasaki K, ShimazakiM, et al. Apoptosis of astrocytes with enhanced lysosomal activity andoligodendrocytes in white matter lesions in Alzheimer's disease.Neuropathol Appl Neurobiol 2002;28:238–51.

[30] Vernadakis A, Kentroti S, Brodie C, Mangoura D, Sakellaridis N. C6glioma cells of early passage have progenitor properties in culture. AdvExp Med Biol 1991;296:181–95.

[31] Vernadakis A, Lee K, Kentroti S, Brodie C. Role of astrocytes in aging:late passage primary mouse brain astrocytes and C6 glial cells asmodels. Prog Brain Res 1992;94:391–409.

[32] Donato R. S100: a multigenic family of calcium-modulated proteins ofthe EF-hand type with intracellular and extracellular functional roles.Int J Biochem Cell Biol 2001;33:637–68.

[33] Ivanekov VV, Jamieson GA, Dimlich RV, Gruenstein RV. Character-ization of S-100B binding epitopes. Identification of a novel target, theactin capping protein, CapZ. J Biol Chem 1995;270:14651–8.

[34] Donato R. Calcium-sensitivity of brain microtubule proteins in thepresence of S-100 proteins. Cell Calcium 1985;6:343–61.

[35] Sorci G, Agneletti AL, Bianchi R, Donato R. Association of S100Bwith intermediate filaments and microtubules in glial cells. BiochimBiophys Acta 1998;1448:277–89.

[36] Sorci G, Agneletti AL, Donato R. Effects of S100A1 and S100B onmicrotubule stability. An in vitro study using triton-cytoskeletons fromastrocyte and myoblast cell lines. Neuroscience 2000;99:773–83.

[37] Bianchi R, Giambanco R, Donato R. S100 protein, but not calmodulin,binds to the glial fibrillary acidic protein and inhibits its polymerizationin a Ca2+-dependent manner. J Biol Chem 1993;268:12669–74.

[38] Van Eldik LJ, Wainwright MS. The Janus face of glial-derived S100B:beneficial and detrimental functions in the brain. Restor NeurolNeurosci 2003;21:97–108.

[39] Gonçalves D, Karl J, Leite M, Rotta L, Salbego C, Rocha E, et al. Highglutamate decreases S100B secretion stimulated by serum deprivationin astrocytes. NeuroReport 2002;13:1533–5.

[40] Büyuküysal RL. Protein S100B release from rat brain slices during andafter ischemia: comparison with lactate dehydrogenase leakage.Neurochem Int 2005;47:580–8.

[41] Tramontina F, Leite MC, Goncalves D, Tramontina AC, Souza DF,Frizzo JK, et al. High glutamate decreases S100B secretion by amechanism dependent on the glutamate transporter. Neurochem Res2006;31:815–20.

[42] Yudkoff M, Daikhin Y, Nissim I, Horyn O, Luhovyy B, Nissim I. Brainamino acid requirements and toxicity: the example of leucine. J Nutr2005;135:158S–1531S.

[43] Funchal C, Rosa AM, Wajner M, Wofchuk S, Pessoa-Pureur R.Reduction of glutamate uptake into cerebral cortex of developing ratsby the branched-chain α-keto acids accumulating in maple syrup urinedisease. Neurochem Res 2004;29:747–53.

[44] Tramontina F, Karl J, Gottfried C, Mendez A, Gonçalves D, Portela LV,et al. Digitonin-permeabilization of astrocytes in culture monitored bytrypan blue exclusion and loss of S100B by ELISA. Brain Res Protoc2000;6:86–90.

[45] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measure-ment with the Folin phenol reagent. J Biol Chem 1951;193:265–7.

[46] Mangoura D, Sakellaridis N, Jones J, Vernadakis A. Early and latepassage C6 glial cell growth: similarities with primary glial cells inculture. Neurochem Res 1989;14:941–7.

[47] Goya L, Feng PT, Aliabadi S, Timiras PS. Effect of growth factors onthe in vitro growth and differentiation of early and late passages of C6glioma cells. Int J Dev Neurosci 1996;14:409–17.

[48] Haghighat N, McCandless DW. Effect of ammonium chloride onenergy metabolism of astrocytes and C6-glioma cells. Metab Brain Dis1997;12:287–98.

[49] Bonini P, Cicconi S, Cardinale A, Vitale C, Serafino AL, Ciotti MT,et al. Oxidative stress induces p53-mediated apoptosis in glia: p53transcription-independent way to die. J Neurosci Res 2004;75: 83–95.

[50] Smale G, Nichols NR, Brady DR, Finch CE, Horton WEJ. Evidencefor apoptotic cell death in Alzheimer's disease. Exp Neurol1995;133:225–30.

[51] Sonnewald U, Qu H, Aschner M. Pharmacology and toxicology ofastrocyte–neuron glutamate transport and cycling. J Pharmacol ExpTher 2002;301:1–6.

[52] Donato R. Intracellular and extracellular roles of S100 proteins.Microsc Res Tech 2003;60:540–51.

[53] Schulpis KH, Kariyannis C, Papassotiriou I. Serum levels of neuralprotein S-100B in phenylketonuria. Clin Biochem 2004;37:76–9.

[54] Wiltfang J, Nolte W, Otto M, Wildberg J, Bahn E, Figulla HR, et al.Elevated serum levels of astroglial S100 beta in patients with livercirrhosis indicate early and subclinical portal-systemic encephalopathy.Metab Brain Dis 1999;14:239–51.

[55] Leite MC, Brolese G, Almeida LMV, Pinero CC, Gottfried C,Gonçalves CA. Ammonia-induced alteration in S100B secretion inastrocytes is not reversed by creatine addition. Brain Res Bull 2006;70:179–85.

[56] Hutson SM, Berkich D, Drown P, Xu B, Aschner M, LaNoue KF. Roleof branched-chain aminotransferase isoenzymes and gabapentin inneurotransmitter metabolism. J Neurochem 1998:863–74.

[57] Bixel M, Shimomura Y, Hutson S, Hamprecht B. Distribution of keyenzymes of branched-chain metabolism in glial and neuronal cells inculture. J Histochem Cytochem 2001;49:407–18.

[58] Jouvet P, Rustin P, Taylor DL, Pocock JM, Felderhoff-Mueser U,Mazarakis ND, et al. Branched-chain amino acids induce apoptosisin neural cells without mitochondrial membrane depolarizationor cytochrome c release: implications for neurological impairmentassociated with maple syrup urine disease. Mol Biol Cell 2000;11:1919–32.

[59] Funchal C, de Lima Pelaez P, Oliveira Loureiro S, Vivian L, Dall BelloPessutto F, Almeida LMV, et al. α-Ketoisocaproic acid regulatesphosphorylation of intermediate filaments in postnatal rat corticalslices through ionotropic glutamatergic receptors. Dev Brain Res2002;139:267–76.

[60] Funchal C, Dall Bello Pessutto F, Almeida LMV, de Lima Pelaez P,Oliveira Loureiro S, Vivian L, et al. α-Keto-β-methylvaleric acidincreases the in vitro phosphorylation of intermediate filaments incerebral cortex of young rats though the GABAergic system. J NeurolSci 2004;217:17–24.

[61] Wilken B, Ramirez JM, Probst I, Rchter DW, Henefeld F. Creatineprotects the central respiratory network of mammals under anoxicconditions. Pediatr Res 1998;43:8–14.

[62] Sullivan PG, Geiser JD, Mattson MP, Scheff SW. Dietary supplementcreatine protects against traumatic bran injury. Ann Neurol 2000;48:723–9.

[63] Zhang W, Narayanan M, Friedlander RM. Additive neuroprotectiveeffects of minocycline with creatine in a mouse model of ALS. AnnNeurol 2003;53:267–70.

[64] Scheff SW, Dhillon HS. Creatine-enhanced diet alters levels of lactateand free fatty acids after experimental brain injury. Neurochem Res2004;29:469–79.

[65] Brustovetsky N, Brustovetsky T, Dubiski JM. On the mechanisms ofneuroprotection by creatine and phosphocreatine. J Neurochem2001;76:425–34.

[66] Clarke DD, Sokoloff L. Circulation and energy metabolism of thebrain. In: Siegel GJ, Agranoff BW, Wayne Albers R, Fisher SK, UhlerMD, editors. Basic neurochemistry. Philadelphia (PA): LippincottWiliams and Wilkins; 1999. p. 638–69.

[67] Wyss M, Kaddurah-Daouk R. Creatine and creatinine metabolism.Physiol Rev 2000;80:1107–212.

94 C. Funchal et al. / Journal of the Neurological Sciences 260 (2007) 87–94

[68] Lesman M, Korzhevskii DE, Mourovets V, Kostkin VB, Izvarina N,Perasso L, et al. Intracerebroventricular administration of creatineprotects against damage by global cerebral ischemia in rat. Brain Res2006;1114:187–94.

[69] Lawler JM, Barnes WS, Wu GY, Song W, Demaree S. Direct anti-oxidant properties of creatine. Biochem Biophys Res Commun 2002;290:47–52.

[70] Almeida LS, Salomons GS, Hogenboom FC, Jakobs C, SchoffelmeerANM. Exocytotic release of creatine in rat brain. Synapse 2006;60:118–23.

[71] Coitinho AS, Mello CF, Lima TT, Bastiani J, Figuera MR, Wajner M.Pharmacological evidence that alpha-ketoisovaleric acid inducesconvulsions through GABAergic and glutamatergic mechanisms inrats. Brain Res 2001;894:68–73.

[72] Funchal C, Latini A, Jacques-Silva MC, Santos AQ, Buzin L, GottfriedC, et al. Branched-chain alpha-ketoacids accumulating in maple syrupurine disease induce morphological alterations and oxidative stress inglial cells. Neurochem Int 2006;49:640–50.

[73] Meister A, Anderson ME. Glutathione. Annu Ver Biochem 1983;52:711–60.

[74] Arstall MA, Bailey C, Gross WL, Bak M, Balligand JL, Kelly RA.Reversible S-nitrosation of creatine kinase by nitric oxide in adult ratventricular myocytes. J Mol Cell Cardiol 1998;30:979–88.

[75] Radi R, Cassina A, Hodara R. Nitric oxide and peroxynitriteinteractions with mitochondria. Biol Chem 2002;383:401–9.

[76] Han Z, Chen YR, Jones CI, Meenakshiundaram G, Zweier JL,Alevriadou BR. Shear-induced reactive nitrogen species inhibit

mitochondrial respiratory complex activities in cultures vascularendothelial cells. Am J Physiol Cell Physiol 2007;292:C1103–12.

[77] Davey GE, Murmann P, Heizmann CW. Intracellular Ca2+ and Zn2+

levels regulate the alternative cell density-dependent secretion ofS100B in human glioblastoma cells. J Biol Chem 2001;276:30819–26.

[78] Funchal C, Zamoner A, Santos AQ, Loureiro SO, Wajner M, Pessoa-Pureur R. Alpha-ketoisocaproic acid increases the phosphorylation ofintermediate filament proteins from rat cerebral cortex by mechanismsinvolving Ca2+ and cAMP. Neurochem Res 2005;30:1139–46.

[79] Balazs R, Miller S, Chun Y, Cotman CW. Receptor-coupledphospholipase C and adenylyl cyclase function with different calciumpools in astrocytes. NeuroReport 1998;11:1397–401.

[80] Guix FX, Uribesalgo I, Coma M, Munoz FJ. The physiology andpathophysiology of nitric oxide in the brain. Prog Neurobiol 2005;76:126–52.

[81] Hashimoto A, Miyakoda G, Hirose Y, Mori T. Activation ofendothelial nitric oxide synthase by cilostazol via a cAMP/proteinkinase A- and phosphatidylinositol 3-kinase/Akt-dependent mechan-ism. Atherosclerosis 2006;189:350–7.

[82] Hoffmann GF, Meier-Augestein W, Srockler S, Surtees R, Rating D,Nyan WL. Physiology and pathophysiology of organic acids incerebrospinal fluid. J Inherit Metab Dis 1993;16:646–69.

[83] Sauer SW, Okun JG, Fricker G, Mahringer A, Muller I, Crnic LR, et al.Intracerebral accumulation of glutaric and 3-hydroxyglutaric acidssecondary to limited flux across the blood–brain barrier constitute abiochemical risk factor for neurodegeneration in glutaryl-CoAdehydrogenase deficiency. J Neurochem 2006;97:899–910.