B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ loca te /b ra in res

Research Report

Alterations in striatal dopamine catabolism precede loss ofsubstantia nigra neurons in a mouse model of juvenileneuronal ceroid lipofuscinosis

Jill M. Weimera, Jared W. Benedicta, Yasser M. Elshatorya, Douglas W. Shorta,c,Denia Ramirez-Montealegrea, Deborah A. Ryana, Noreen A. Alexandere,f,Howard J. Federoff a,d, Jonathan D. Coopere,f, David A. Pearcea,b,c,d,⁎aCenter for Aging and Developmental Biology, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USAbDepartment of Biochemistry and Biophysics, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USAcDepartment of Environmental Medicine, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USAdDepartment of Neurology, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USAePediatric Storage Disorders Laboratory, Centre for the Cellular Basis of Behaviour, King’s College London, Institute of Psychiatry,De Crespigny Park, London, SE5 9NU, UKfDepartment of Neuroscience, Centre for the Cellular Basis of Behaviour, King’s College London, Institute of Psychiatry,De Crespigny Park, London, SE5 9NU, UK

A R T I C L E I N F O

⁎ Corresponding author. University of RochesRochester, NY 14642, USA. Fax: +1 585 276 19

E-mail address: david_pearce@urmc.roch

0006-8993/$ – see front matter © 2007 Elsevidoi:10.1016/j.brainres.2007.05.018

A B S T R A C T

Article history:Accepted 14 May 2007Available online 21 May 2007

Batten disease, or juvenile neuronal ceroid lipofuscinosis (JNCL), results from mutations inthe CLN3 gene. This disorder presents clinically around the age of 5 years with visual deficitsprogressing to include seizures, cognitive impairment, motor deterioration, hallucinations,and premature death by the third to fourth decade of life. The motor deficits includecoordination and gait abnormalities, myoclonic jerks, inability to initiate movements, andspasticity. Previous work from our laboratory has identified an early reduction in catechol-O-methyltransferase (COMT), an enzyme responsible for the efficient degradation ofdopamine. Alterations in the kinetics of dopamine metabolism could cause theaccumulation of undegraded or unsequestered dopamine leading to the formation oftoxic dopamine intermediates. We report an imbalance in the catabolism of dopamine in3 month Cln3−/− mice persisting through 9 months of age that may be causal to oxidativedamage within the striatum at 9 months of age. Combined with the previously reportedinflammatory changes and loss of post-synaptic D1α receptors, this could facilitate cell lossin striatal projection regions and underlie a general locomotion deficit that becomesapparent at 12 months of age in Cln3−/−mice. This study provides evidence for early changesin the kinetics of COMT in the Cln3−/− mouse striatum, affecting the turnover of dopamine,likely leading to neuron loss and motor deficits. These data provide novel insights into thebasis of motor deficits in JNCL and how alterations in dopamine catabolism may result inoxidative damage and localized neuronal loss in this disorder.

© 2007 Elsevier B.V. All rights reserved.

Keywords:Batten diseaseCatechol-O-methyltransferaseCLN3DopamineOxidative stress

ter School of Medicine and Dentistry, Center for Aging and Developmental Biology, Box 645,72.ester.edu (D.A. Pearce).

er B.V. All rights reserved.

99B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

1. Introduction

Juvenile neuronal ceroid lipofuscinosis (JNCL) is an autosomalrecessive neurodegenerative disorder with an average age ofonset around 5 years with patients typically not surviving pasttheir third decade of life. This disease, resulting from amutation in the CLN3 gene, manifests with visual deficits butprogresses to include seizures, motor problems, cognitiveimpairment, and, in some cases, hallucinations and schizo-phrenic-like behavior. The extrapyramidal motor deficitsassociated with the disease include stereotypic cog-wheel

Fig. 1 – Deficits genes involved in catecholamine metabolism isEvaluation of several published microarray data sets performedthe expression of Comt (3.56 fold decrease, 10-week whole braincatecholamine catabolism (Dopamine, A and Norepinephrine, B)catecholamine is achieved through conversion of tyrosine to DOdopamine, norepinephrine, and epinephrine. Of the enzymes intyrosine hydroxylase, DOPA decarboxylase, dopamine β-hydroxywere altered in the Cln3−/− mice aside from a slight, yet significanenzyme involved in conversion of norepinephrine to epinephrin10-week whole brain microarray; Brooks et al., 2003] (C).

rigidity, balance impairment, hypokinesia, flexed posture, andshuffling gait that typically render patients immobile by theirmid-teens (Hofman et al., 1999).

Imaging studies in JNCL patients have shown dysfunctionin the striatum as measured by a reduction in dopaminetransporter (DAT) levels and [18F] fluorodopa uptake, an analogof dopamine (Ruottinen et al., 1997; Aberg et al., 2000). Oncereleased into the synaptic cleft, dopamine activates either theD1 or D2 class of receptors. PET studies of JNCL patientsshowed a reduction in the striatal D1, but not D2 receptors(Rinne et al., 2002). In this study, we have investigated these

restricted to a decreased in COMT in the Cln3−/− mice.on the Cln3−/− mice showed a selective down regulation inmicroarray) with a sparing of all other enzymes involved in(Brooks et al., 2003; Elshatory et al., 2003). Synthesis ofPA, followed by the subsequent selective conversion tovolved in the production of catecholamines, specificallylase, and norephinephrine N-methyltransferase (Pnmt) nonet, decrease in the level of Pnmt, suggesting disruption in thee [less than 1.5 fold change (p≤0.05, Student's t-test),

100 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

basal ganglia deficits in Cln3 knockout mice (Cln3−/−), a modelof Batten disease. Although NCL mouse models have shownmany pathological characteristics similar to diseased patients(Mitchison et al., 1999; Seigel et al., 2002; Sappington et al.,2003; Pontikis et al., 2004; Pontikis et al., 2005; Kielar et al.,2007), neuroanatomical analysis of the Cln3−/− mice has failedto reveal any regional volumetric changes of the striatum,although detailed analysis of target cell loss in this area hasnot been performed (Pontikis et al., 2004). Interestingly, in theCln3−/− mice, as well as other NCL models, there is a markedincrease in the levels of both reactive astrocytes andmicrogliawithin this compartment, indicative of cellular damage(Pontikis et al., 2004; Pontikis et al., 2005; Kielar et al., 2007).Interestingly, gene expression studies of the Cln3−/− mousebrain have uncovered a decrease in the expression of catechol-O-methyl transferase (COMT) (Elshatory et al., 2003). This is akey enzyme involved in the degradation of catecholamineswithin the striatum, as well as other brain regions, calling intoquestion the integrity of the metabolism and function of thisneurotransmitter family in the Cln3−/− mouse. Althoughlimited behavioral analysis of these mice has unmaskedcerebellar related motor deficits (Kovacs et al., 2006), noanalysis of basal ganglia specific motor changes has yet beenperformed.

In this study,wedemonstratea reduction inboth theproteinand activity levels of COMT in Cln3−/− mice by three months ofage which results in an alteration in the degradation ofcatecholamines. Specifically, we show a decrease in the levelsand catabolism of dopamine.We reasoned that a diminution ofdopamine catabolism would lead to elevation in reactiveoxidative forms of dopamine, particularly within the non-vesicular compartment, that could increase oxidative damagewithin the striatum. Recent attempts to elucidate the eventsleading up to cell loss and microglial activation in the Cln3−/−

mice have demonstrated elevated oxidative burden in differentanatomical compartments, often with activated microgliallocated immediately adjacent to neurons reactive for potentantioxidant markers (Benedict et al., in press). Elevatedoxidative stresswithin the striatum could result from or triggerthe previously reported inflammatory response (Pontikis et al.,2004), subsequently leading to restricted cellular changes,including the postsynaptic receptor down regulation reportedin JNCL (Rinne et al., 2002). Indeed,we saw amarked increase inoxidized proteins and the down regulation of D1α receptors in

Fig. 2 – Decreased COMT protein and activity levels in 3 and 9 mcatechol-O-methyl transferase (COMT) revealed changes in bothin brain samples fromCln3−/−mice compared towild-typemice. S3- and 9-month-old male mice. Top panels show the representatiwhile the lower band shows β-actin immunoblotting. HistogramS-COMT in wild-type and Cln3−/− mice. Densitometry of MB-COMwith background subtraction being performed for both COMT andactivity of COMT was assayed by enriching for the S- or MB-COMWestern blot), mixing samples with the co-factor S-adenosylmeacid. O-methylation of dihydroxybenzoic acid (DHBAc) to vanillicHPLC (A, B, lower row). Samples were prepared from cortex (A) a(right column) -old male Cln3−/− and 129/SvJ wild-type mice. Decrin cortex and striatum (A, B, lower row). Significant decreases instriatum (B). Activity is represented as mean pg vanillic acid/mg(two-way ANOVA).

the Cln3−/− striatum by nine months of age, paralleling theincrease in glial activation. In exploringwhether this alterationwithin the basal ganglia culminates in cell loss in afferent orefferent striatal targets,wedemonstrateda loss of parvalbuminpositive substantia nigra pars reticulata neurons at ninemonths.Therefore, this suggests that an early decrease in COMTactivity in the Cln3−/− mice leads to dramatic changes incatabolism of dopamine, ultimately affecting striatal circuitry,including oxidative damage and progressive loss of nigralneurons. This cell loss provides a plausible explanation for themotor deficits we report in Cln3−/− mice and provides a likelymechanism for the extrapyramidal dysfunction seen in JNCLpatients.

2. Results

2.1. Disruption in key enzyme involved in catecholaminecatabolism in Cln3−/− mice

The Cln3−/− mouse model for Batten disease has a substantialdecrease in the levels of catechol-O-methyl transferase (Comt)mRNA in the brain (Elshatory et al., 2003). This enzyme, alongwith monoamine oxidase, MAO, facilitates the degradation ofdopamine both within the striatum and other portions of thenervous system (Fig. 1A). Degradation of norepinephrinerequires two additional enzymes, aldehyde reductase andaldehyde dehydrogenase (Fig. 1B). Mining of several pub-lished microarray data sets performed on brains from Cln3−/−

mice showed a selective downregulation in the expression ofComt but none of the other enzymes involved in catechola-mine catabolism were altered (Figs. 1A and B) (Brooks et al.,2003; Elshatory et al., 2003). Normal synthesis of catechola-mines is achieved through conversion of tyrosine to DOPA,dopamine, norepinephrine, and epinephrine, sequentially.Disruption in the initial synthesis of catecholamines couldimpact downstream regulation of these proteins, as well asenzymes involved in their turn-over. Enzymes involved in theproduction of catecholamines, specifically tyrosine hydro-xylase, DOPA decarboxylase, dopamine β-hydroxylase, andnorephinephrine N-methyltransferase (Pnmt) were alsoincluded in the previously conducted microarray studies. Ofthese enzymes, none were altered in the Cln3−/− mice, excepta slight, yet significant, decrease in the level of Pnmt,

onth Cln3−/− mice cortex and striatum. Immunoblotting forthe membrane bound (MB) and soluble (S) form of the proteinampleswere prepared from cortex (A) and striatum (B) of bothve level of COMT from the indicated samples (13% SDS-PAGE)s in the lower panel depict relative levels of MB-COMT andT and S-COMT bands were standardized to β-actin densities,actin bands (mean relative intensity±S.E.M.). O-methylatingT fractions by centrifugation at 100,000×g (confirmed bythionine (SAM), followed by reaction with dihydroxybenzoicacid was assessed after 60 min incubation times bynd striatum (B) of both 3- (left column) and 9-montheases in the activity of S-COMTwere seen at both time pointsactivity of the MB-COMT were seen only at 9 months in theprotein/min±S.E.M.; *P≤0.05, **P≤0.01, ***P≤0.001

101B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

suggesting disruption in the enzyme involved in conversionof norepinephrine to epinephrine (less than 1.5 fold change,p=0.049, 10-week whole brain microarray; Brooks et al.,2003) (Fig. 1C). Therefore, we chose to further explore COMTas a putative enzymes involved in the pathogenesis of Battendisease.

To establish that this decrease in Comt mRNA levels trans-lated to alterations in protein levels, we performed immuno-blotting analysis for COMT on proteins from microdissectedcortex and striatumsamples. COMT is found in the brain in twoforms, as a predominant, high affinity membrane bound (MB-

COMT) isoform located primarily in the rough-ER, and in a lowaffinity, soluble isoform (S-COMT) found throughout thecytosol and nucleus (Salminen et al., 1990; Lundstrom et al.,1991; Malherbe et al., 1992; Winqvist et al., 1992). COMT hasbeen shown to be most prevalent within the prefrontal cortex,however, biochemical and immunohistological studies havedemonstrated this enzyme to be present in nearly allmammalian tissues (Guldberg and Marsden, 1975). Compar-ison of COMT protein levels in wild-type and Cln3−/− mice invarious brain regions revealed isoform-specific, brain region,and age-dependent differences.

102 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

In frontal cortex, while COMT protein levels were notdifferent between wild-type and Cln3−/− mice at 3 months ofage, by 9 months of age, Cln3−/− cortex showed lower levels ofMB-COMT (Fig. 2A). Themost apparent changes in COMTwerefound in the striatum, where there were decreased COMTprotein levels at both 3 and 9 months of age, with the S-COMTbeing significantly lower at 3 months and the MB-COMT by9 months (Fig. 2B). Thus, MB-COMT protein levels weredecreased in the cortex and striatum of 9 month old Cln3−/−

mice, while S-COMT protein levels decreased in the striatumonly at three months of age.

To establish whether alterations in the ratio of soluble tomembrane COMT levels affect the activity of this enzyme, weassayed the function of both forms of this protein by the O-methylation of dihydroxybenzoic acid in protein extracts.Within the cerebral cortex, the activity level of both forms ofCOMT paralleled the decreases in protein levels at 3months ofage, although by 9 months of age there was no significantchange in activity (Fig. 2A). Within the striatum, earlyreductions in the expression of COMT followed closely witha reduction in the activity of the enzyme which wasmaintained at 9 months of age, demonstrating a substantial,prolonged impairment in the levels and function of thisenzyme (Fig. 2B).

In summary, within the striatum, there is a globalreduction in both the low affinity soluble and high affinitymembrane bound form of the enzyme occurring as early as3 months of age which persists at 9 months of age. Deficits inthis enzyme are not restricted to the striatum, with both thefrontal cortex and cerebellum (data not shown) showingimpairment in COMT levels and function. The low affinity,soluble form is predominately expressed within astrocytes of

Fig. 3 – Alterations in level of Dopamine, DOPAC, and HVA in mhomogenates were sonicated and catecholamine metabolites weand DOPAC came off at 3.33, 5.65, and 2.73 min, and showed marespectively. The levels of dopamine, HVA, and DOPACwere assacortex and midbrain and presented as level of metabolite in ng nrevealed global disruption in the levels of these metabolites in Clthere was a significant decrease in the levels of dopamine and Dpresented as mean ng metabolite/μg total protein±S.E.M., *P≤0.

the brain, whereas the MB-COMT is primarily found inneurons (Rivett et al., 1983a,b). Although neuroanatomicalanalysis has failed to reveal any volumetric decrease in thestriatum of Cln3−/− mice, by 5 months of age there is anelevation in both reactive astrocytes and microglia within thiscompartment, indicative of on-going cellular damage andtargeted cellular stress within the striatum (Pontikis et al.,2004).

2.2. Disruption in dopamine catabolism in Cln3−/− mice

To explore whether decreased activity of COMT within theCln3−/− mice influenced turnover of catecholamines, weexamined the levels of dopamine and its metabolites usingHPLC-EC in cortical and striatal samples. At both 3 and9 months, there was a significant decrease in the level ofdopamine in both the cerebral cortex and at 3 months withinthe striatum (Figs. 3A–B). These findings demonstrate anoverall decrease in the levels of dopamine in Cln3−/− mice.Several mechanisms could account for this decrease indopamine levels which is apparent by 3 months of age. Onepossibility is that there is an alteration in the metabolism ofdopamine. Alternatively, decreases in dopamine could be theresult of an increase in dopamine oxidation. Once oxidized,the dopamine may form covalent adducts on proteins, whichin turn would cause a shift from the normal dopamine peakmaking the levels of dopamine appear artificially low.

A decrease in the activity of enzymes that degradecatecholamines, such as COMT, could alter the catabolism ofdopamine. This process occurs by two alterative pathways,either degradation via COMT, leading to the formation ofhomovanillic acid (HVA) via an homovanillyl-aldehyde

idbrain and cortex wild-type and Cln3−/− mice. Deproteinizedre fractionated on an ESA column (70-0636). Dopamine, HVAximum responses at channels set to 25, 250 and 100 mV,yed 3- and 9-month-old Cln3−/− and 129/SvJ wild-typemice inormalized to the total amount of protein in μg. This analysisn3−/− mice. In both the cerebral cortex (A) and the striatum (B)OPAC, although HVA levels remained unaffected. Data are05, **P≤0.01, ***P≤0.001 (two-way ANOVA).

103B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

intermediate, or by monoamine oxidase B, also leading to theformation of HVA, but via a 3,4-dihydroxyphenylacetic acid(DOPAC) intermediate (Kandel et al., 2000). Therefore, bycomparing levels of dopamine, and two of these majormetabolites, DOPAC and HVA, the effectiveness of COMTversus MAO-B catabolism of dopamine may be determined.The levels of HVA and DOPAC in 3- and 9-month-old Cln3−/−

and 129/SvJ wild-type mice in cortex and midbrain weremeasured by HPLC-EC. This analysis revealed global disrup-tion in the levels of the DOPAC metabolite. In the cerebralcortex there was a ∼10 fold decrease in the levels of DOPAC at3 months of age (Fig. 3A), whereas there were significantdecrease in the levels of DOPAC within the striatum, with agreater than 9 fold reduction in the level of this metabolite at3 months and a ∼5 fold reduction at 9 months of age (Fig. 3B).These findings reveal a significant alteration in the catabolismof dopamine in both brain regions. One would speculate that,in order to maintain a comparable level of HVA with adecrease in COMT there would have to be a substantialincrease in the levels of DOPAC. Instead there was a markeddecrease in the levels DOPAC, suggesting that the machineryinvolved in dopamine catabolism is slowing in response to thedecreased levels of dopamine. Additionally, this decrease incatabolism could increase the amount of time the residualdopamine is free within the synapse, therefore profoundlyaffecting the normal functioning of the frontal cortex andstriatum [as well as similar observations noted in thecerebellum (data not shown)] in Cln3−/− mice.

2.3. Alteration in dopamine catabolism within thestriatum leads to an increase in oxidative stress and loss ofneurons in the Cln3−/− mice

Alterations in COMT activity can potentially have confoundingeffects on the striatum. The presence of elevated levels ofundegradeddopaminewithin the synaptic compartment couldlead to theappearanceof toxic, reactive oxidative derivatives ofdopamine. These toxic intermediates have been shown topromote oxidative damage in several neurodegenerative dis-ease, including Parkinson and Alzheimer’s disease (Dalle-Donne et al., 2003a; Glaser et al., 2005; Hsu et al., 2005). Todeterminewhether there is elevatedoxidative stresswithin thestriatum of Cln3−/− mice, we measured the levels of proteincarbonyls within the striatum and cortex. Derivatization ofcarbonyl groups on amino acid side chains with 2,4-dinitro-phenylhydrazine (DNPH) leads to the formation of a stableproduct, 2,4-dinitrophenyl (DNP) hydrozone which can bedetected using an antibody to DNP (Dalle-Donne et al., 2003b).At 3monthsof age, therewasnochangebetweenwild-typeandCln3−/− mice in the level of DNP, but at 9 months of age therewas a significant increase in the level of carbonyls within thestriatum of Cln3−/− mice (Fig. 4A). No changeswere noted in thecerebral cortex, again suggesting that these changes arerestricted to the striatum (Fig. 4A). Appearance of oxidativedamage within the striatum lags behind alterations in COMTactivity, as well as, in the appearance of microglial activation,suggesting that oxidative stress due to carbonylation occursonly in the later stages of disease pathogenesis.

Microglial activation, oxidative damage, and cellular stresswithin the striatum could have multiple effects, not only on

cells within the striatum itself, but also cells which project toand from this region. It has been previously demonstrated inthe Cln3−/− mouse that by as late as 15 months of age there isno overall atrophy of the striatum, suggesting gross pre-servation of this brain region (Pontikis et al., 2004). This samestudy showed an increase in the activation of both microgliaand astrocytes within the striatum by 5 months of age,suggesting that an ongoing cellular insult occurs within thisregion of the brain. These changes in oxidative damage andglial activation could have direct effects upon cell popula-tions which project to and from the striatum. Cells within thesubstantia nigra pars compacta (SNpc) receive modulatoryprojections from the striatum. Additionally, cells in thesubstantia nigra pars reticulata (SNpr) receive efferentconnections from cells of the striatum and project to thethalamus. We first explored whether there were regionalvolume changes within either of these connection regions byperforming Cavalieri estimates of volume at 6 months of age,an intermediate time point between when we first notedchanges in COMT levels and elevation in oxidative damagewithin the striatum. Indeed, we saw a decrease in the volumeof the SNpr of the Cln3−/− mouse at 6 months of age (Fig. 4B).To determine whether regional volume changes were theresult of either cell loss or a decrease in individual cell size,we performed optical fractionator cell counts. We usedparvalbumin (PV) stained sections, to visualize the GABAergicneurons of the SNpr, and tyrosine hydroxylase (TH) stainedsections, to visualize the dopaminergic neurons of the SNpc.Although there was a substantial decrease in the volume ofthe SNpr at 6 months of age, there was not a significant lossof PV positive cells within this region until 9 months of age(Fig 4C). By this time point, there was a near 35% reduction inthe number of cells within SNpr of the Cln3−/− mice. Incontrast, the TH positive cells of the SNpc remainedpreserved even at the later time points (Fig. 4D). The THpositive cells of neighboring ventral tegmental area (VTA)also appeared preserved (data not shown). Nucleator esti-mates of neuronal volume detected no change in the size ofthe remaining neurons within either region at 6 months ofage (data not shown).

2.4. Degeneration in motor skills in the Cln3−/− mouse

To explore whether these cellular and molecular perturbationof the basal ganglia circuitry results in defects in motor skills,we evaluated locomotor behavior at 3, 9, and 12months of age.To determine whether there was a compromise in generalmotor activity, we evaluated total motor output in distanceand time. Although no differences in these parameters wereobserved between genotypes at 3 or 9 months of age, by12 months of age there was an overt impairment in locomotoractivity (Fig. 5). At this later time point, Cln3−/− mice showed a∼3 fold reduction in the total distance traveled and a ∼2 foldreduction in the total amount of time spent ambulating duringthe 30 min test session (Figs. 5E–F). These impairments werespread evenly across the 30min trial periodwith no detectabledifferences within individual 50 ms time intervals. Thisdemonstrates that these differences are not the result ofslowness in initiating movements at the onset of the trial or atiring of the animal.

104 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

2.5. Compensation for decreased COMT activity

Within the striatum, inhibition of COMT activity and altera-tion in dopamine levels could lead to changes within themicroenvironment of the synapse. Because undegradeddopamine can auto-oxidize and lead to the formation of

chemically reactive intermediates which are capable ofpromoting oxidative damage, cells within the striatum mightup-regulate other proteins, shifting toward an alternativepathway for sequestering or degrading this catecholamine.

Although COMT has been associated with postsynapticmedium spiny neurons within the striatum, this enzyme

Fig. 5 – Degenerative changes inmotor activity of Cln3−/−mice. Locomotor testingwas performed on 129/SvJ and Cln3−/−mice atthree different time points in an open field activity chamber. Animals were habituated for 2 days for 30 min followed bytesting trial of 30min. Separate groups ofmicewere tested at 3, 9, and 12months of age. The total distance each animal traveled(A, C, E) and the amount of time spent moving (B, D, F) was calculated. No differences in total distance traveled or time spentambulatingwere detected at 3 or 9months of age. By 12months of age, therewas a significant decrease in distance traveled andamount of time spent ambulating in Cln3−/− mice (E–F). Mean time or distance traveled±S.E.M., twoway ANOVA, F(1,23)=2.421,P=0.004 (time) and (F(1,23)=2.560, P=0.001 (distance).

105B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

predominantly resides within astrocytes juxtaposed to thesynapse. Astrocytes do not express a dopamine specifictransporter; therefore, movement of dopamine into thesecells is facilitated by the norepinephrine transporter (NET).

Fig. 4 – Elevated oxidative striatal stress and cellular atrophy infrom 3- and 9-month-oldmale Cln3−/− andwild-type (129/SvJ) cortonto PVDF and carbonyls were derivatized with DNPH to form arevealed an increase in the level of oxidative stress at 9 monthsHistograms depict densitometry of derivatized carbonyls standardSypro Ruby Red (Mean relative intensity±S.E.M.). Unbiased Cavastained brain sections. Regional volumeswere expressed inμm3

homozygous Cln3−/− mice. There was a significant regional atropwithin the substantia nigra pars reticulata (SNpr) (B, left panel), asubstantia nigra pars compacta (SNpc, B, right panel). This cell loand was progressive as the animals aged. Optical fractionator ceabsolute numbers of cells within each region. To distinguish celparvalbumin (PV) and the cells of the SNpc were labeled with tyrdecrease was seen in the number of PV positive SNpr cells, prog1788±21.3, Cln3−/−: 1214±135.87) although no changewas seen at1916±228.27; 6 month: 129/SvJ: 1785±87.79, Cln3−/−: 1501±72.68)points (3 month: 129/SvJ: 2906±287.23, Cln3−/−: 2023±329.48; 6 m129/SvJ: 3859±74.71, Cln3−/−: 3324±153.05) (D). Data are presente

Although different isoforms of NET are expressed in the brain(Porzgen et al., 1995), there were no changes in the proteinlevels of the predominant 69 kDa transporter at either 3 or9 months of age in the striatum or the cerebral cortex of the

the substantia nigra pars reticulata of Cln3−/− mice. Proteinsex (A, upper row) and striatum (A, lower row)were dot blottedstable product that was detected by immunoblotting. Thisof age in the striatum of Cln3−/− mice (A, lower right panel).ized to total protein, determined by staining of the PVDFwithlieri estimates of regional volume were measured on Nissland themean volume of each region calculated for control andhy was seen in Cln3−/− vs. 129/SvJ (+/+) at 6 months of agelthough no difference was observed in the volume of thess was specific to the parvalbumin positive cells of the SNprll counts of the SNpr and SNpc were performed to determinels types, cells of the SNpr were immuno-labeled withosine hydroxylase at 3, 6, and 9 months of age. A significantressing to near 35% cell loss by 9 months of age (129/SvJ:the early time points (3month: 129/SvJ: 1766±166.89, Cln3−/−:(C). No decrease within the SNpc was noted at any of the timeonth: 129/SvJ: 3100±193.69, Cln3−/−: 3650±109.83; 9 month:d as mean±S.E.M., *p≤0.05, **p≤0.01 (Student's t-test).

106 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

Cln3−/− mice (Figure S1). The dopamine transporter (DAT) is aplasma membrane transporter expressed at presynapticterminals and facilitates the reuptake of dopamine into nigralprojection neurons. Due to the potential for an increase inundegraded dopamine within the striatum, we investigatedwhether these afferent terminals altered the levels of DAT inan effort to sequester dopamine into the presynaptic term-inals. Western blot analysis showed no difference in the levelof the 75 kDa DAT protein in the Cln3−/− mouse (Figure S1).These data suggest that this potential alteration in dopaminecatabolism does not result in the compensatory upregulationof catecholamine specific transporters. The activity of NETand DAT is regulated by the intracellular signaling events andthe subsequent internalization of the transporter complex.

Fig. 6 – Decreased D1α receptor expression at 9 month Cln3−/− mreceptor revealed a decrease in the expression level of the protecompared to wild-type (129/SvJ) mice. Sampleswere prepared fromice. Top panels show level of D1α receptor from the indicated sHistograms in the lower panel depict relative levels of D1α receptobandswere standardized to actin densities, with background subData are presented as mean±S.E.M., *p≤0.05 (Student's t-test).

Therefore, we can not rule out the possibility that NET orDAT’s activity is being down regulated by removal of theprotein from the cell surface (Torres et al., 2003; Holton et al.,2005).

Once sequestered within the presynaptic terminals via theDAT transporter, dopamine can either be repackaged intosynaptic vesicles by the vesicular monoamine transporter(VMAT) for re-release or degraded by the monoamine oxidaseenzyme, MAO-B. Examination of levels of both VMAT andMAO-B revealed that neither was altered in an immunoblotassay, suggesting that there are not compensatory changes inthe level of the proteins involved in the intracellular handlingof dopamine in the presynaptic nerve terminals of the SNpc(data not shown). However, the finding that there is a decrease

ice striatum. Immunoblotting for the dopamine 1α (D1α)in in striatum brain samples from 9-month-old Cln3−/− micem cortex (A) and striatum (B) of both 3- and 9-month-oldmaleamples while the lower band shows β-actin immunoblotting.r inwild-type andCln3−/−mice. Densitometry of D1α receptortraction being performed for both receptor and β-actin bands.

107B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

in the dopaminemetabolite DOPAC (Fig. 3), which is specific tothe MAO pathway of degradation, does indicate compromisein some aspect of this catabolic pathway.

Alteration in the levels of synaptic dopamine levels couldaffect receptor expression on post-synaptic neurons of thestriatum. The D1α and D2 receptors are the post-synapticdopaminergic receptors and establish the direct and indirectpathway between the striatum and output nuclei of the basalganglia (Kandel et al., 2000). Patients with Batten disease havea down-regulation in the expression of D1α receptor of themedium spiny striatal neurons (Rinne et al., 2002). To addresswhether a similar decrease in receptor levels occurs in theCln3−/− mice, Western blot analysis was performed on corticaland striatal samples at 3 and 9 months of age. A significantdecrease in D1α receptor levels was noted by 9 months of agewithin the striatum, though no difference was seen at theearlier time point (Fig. 6). Additionally, there was no change inD1α receptor levels within the cortex of the Cln3−/− mice.These findings confirm that postsynatpic changes that occurwithin the Cln3−/− mouse striatum resemble those seen inJNCL patients. Moreover, these data suggest that, although

Fig. 7 – Summary of dopaminemetabolism and handling in JNCLthe metabolites dopamine and DOPAC. (3) Decrease in the postsyDAT, MAO-B, VMAT, or NET.

alterations in the levels of COMT protein and activity occur inthe Cln3−/− mice as early as 3 months of age, the effect onpostsynaptic proteins is not seen until a number of monthslater.

3. Discussion

Mechanisms by which JNCL patients develop motor deficitsduring the mid- to late-stages of the disease are poorlyunderstood and exploration of these changes in humanpatients has been limited to imaging studies. Connections inthe basal ganglia nuclei are thought to fine tunemotor signalsrelayed via the cortex and thalamus and alterations withinany aspect of this circuitry alters movements. In this study,using a genetic mouse model of JNCL, we demonstratealterations in MB-COMT and S-COMT protein and activitylevels which are most apparent in the striatum of the Cln3−/−

mice. The alterations in dopamine metabolism that we reportare summarized in Fig. 7. COMT catalyzes the transfer of amethyl group from S-adenosyl-ι-methionine to one of the

. (1) Decrease in both COMT levels and activity. (2) Decrease innaptic D1α receptor. (4) No difference in the expression of

108 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

hydroxyls of a catechol, leading to catecholamine inactivation(Axelrod and Tomchick, 1958). In addition, a shift in the ratioof membrane to soluble form of COMT, as seen in this study,alters the kinetics of dopamine metabolism (Guldberg andMarsden, 1975). Over time, this is likely to lead to anaccumulation of undegraded, unsequestered dopaminewhich auto-oxidizes, initiating an increase in oxidativedamage as we demonstrated by an increased carbonyl burden.Eventually, this cellular stress could progress to the downregulation of the D1α receptor that we see which is similar tothat seen in the striatum of JNCL patients (Rinne et al., 2002).

In addition to early changes in COMT levels, these findingsreveal a substantial overall decrease in the levels of dopaminein the Cln3−/− mice. There is no apparent disruption in theenzymes involved in catecholamine synthesis. The decreasein COMT could result in excess dopamine that could auto-oxidize within the synaptic compartment. HPLC-EC methodsused here would be inefficient at detecting these modifiedforms of dopamine, which are unstable and that quicklyadduct with other proteins. Regardless of the precise mechan-ism involved, there is a significant decrease in functionaldopamine within the brain of Cln3−/− mice as well as adecrease in dopamine catabolism, as measured by decreasesin DOPAC levels. Alterations in COMT could precipitateblunted dopamine catabolism, or alternatively, a decrease inthe global levels of dopamine could lead to a decrease in theproduction of enzymes essential for degradation. Althoughthe present data are insufficient to determine which of thesecellular events occurs first and further studies will be requiredto determine the exact time course of these early events, wedo demonstrate that oxidative damage does not occur untillater ages in the striatum of the Cln3−/− mice. Althoughsustained oxidative stress within this cellular compartmentcould exacerbate glial activation and ultimately lead to celldeath within the striatum, close examination of the timing ofthese events is suggestive that oxidative damage is onlyoccurring secondarily to other earlier molecular and patholo-gical eventswithin the striatum.Nevertheless, the changeswereport offer a plausible explanation for precipitating the loss ofneurons in striatal projection regions and, ultimately in motordeficits that are measurable in Cln3−/− mice. The initial motoranalysis presented here provides evidence for a degenerativechange in locomotor activity dependant on a progressive lossof cells within the substantia nigra paralleling the late-stageonset of motor changes seen in JNCL patients. Recently, wehave reported a defect in motor coordination by as early as2 months of age in Cln3−/− mice (Kovacs et al., 2006). Althoughthese deficits appear to be cerebellar specific and distinguish-able from the late onset general locomotor impairmentreported here, we cannot rule out that coordination deficitsmay be confounding overall locomotor activity, warranting asophisticated assessment of these motor impediments.

Prior to this study, the only reported changes within thestriatumof JNCL patients have come from imaging studies andobservations of motor deficits in patients. A year-long studywhere JNCL patients were treated with L-DOPA, the precursorof dopamine, revealed an initial response with a significantimprovement on a Unified Parkinson Disease Rating Scale(UPDRS) although these patients did not respond as well totreatment as idiopathic Parkinson patients (Aberg et al., 2001).

Concomitant administration of COMT inhibitors with L-DOPA,or the genetic predisposition to express COMT with adiminished activity, has been suggested to increase thebenefits of L-DOPA (Tai and Wu, 2002). As such, the decreasedlevels of COMT evident in Cln3−/− mice, and presumably JNCLpatients, might create an environment more conducive to L-DOPA treatment. The improvement observed in JNCL patientstreated with L-DOPA was not permanent though and afterapproximately 12 months UPDRS scores returned to pretreat-ment levels. Interestingly, a case study of late infantile NCL(LINCL) showed a decrease in cerebrospinal fluid levels of HVAand HVA/DOPAC ratios, suggesting a similar deficit indopamine turnover exist in patients with LINCL, as well as,providing a useful model for measurement and detection ofaltered dopamine metabolites in patients (Barisic et al., 2003).Not unlike the JNCL patients, subsequent treatment with L-DOPA proved ineffective in treatment. Therefore, a morecomprehensive assessment of the molecular changes affect-ing dopaminemetabolism and COMT activity in JNCL patientswould be prudent in lieu of continued L-DOPA therapy.

In addition to affecting the central nervous system (CNS),decreases in the level of COMT could have confounding affectson other systems by impeding subsequent degradation ofcatecholamines, including norepinephrine and epinephrine,both which rely on dopamine as a precursor. As thesecatecholamines regulate cardiovascular function, in tandemwith a depression in COMT activity, disruption in theirmetabolism could lead to aberrant stimulation of the heart.Aside from the reported pathology in the CNS, the only othersystem to be affected in JCNL patients is the cardiovascularsystem. Abnormalities become apparent in the later stages ofthe disease and can include bradycardia and poor blood flowand exploration of whether alterations in catecholaminelevels affect the cardiovascular system in JNCL patients isnecessary (Hofman et al., 1999). Themechanism by which lossof CLN3 would result in changes in the level of COMT andcatecholamines remains elusive, as does themanner in whichthis loss targets oxidative damage to the striatum whilesparing the cortex. We demonstrate a decrease in the numberof GABAergic neurons in the SNpr but observed no change innumber of dopaminergic neurons in the SNpc. As the onlydopamine producing cells of the basal ganglia, the observationof a decrease in dopamine within the striatum would suggestthat cells of the SNpc, although not reduced in number, mayhave significant functional impairment. Previous analysis ofneuroanatomical changes within the striatum have focusedon changes in structural volume, which could reflex not onlychanges in cell number but also alterations in the cell volume,invasion of activated glial cells, or compensatory changes inthe associated neuropil. This study provides a framework for amore detailed assessment of neuronal populations within thestriatum, suggesting that cells projecting to the SNpr may beselectively targeted. Based on the previous findings ofdecreased D1α receptor expression in JNCL, we investigatedwhether the expression of this receptor were changed in Cln3−/−

mice. It should be reiterated that, in addition to D1α, mediumspiny neurons, as well as presynaptic terminals within thestriatum, also express the D2 class of receptors. This receptoris known to be autoregulated by the levels of dopamine withinthe synaptic cleft. Therefore, changes in this class of receptors

109B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

can not be excluded and changes could further contribute tothe pathology reported here in Cln3−/− mice.

Disruption in dopamine function could contribute to boththe motor deficits as well as the Schizophrenic like behaviorsassociated with Batten disease. Indeed, polymorphismswithin COMT have been reported to decrease the levels ofenzyme within the brain (Weinshilboum and Raymond, 1977;Boudikova et al., 1990) and have been associated withSchizophrenia, motor impairments, and cognitive decline(Fan et al., 2005; Galderisi et al., 2005; Horowitz et al., 2005).As JNCL progresses, hallucinations and Schizophrenic-likebehavior becomes common. Recently, Disc-1, a gene believedto be involved in cortical development and linked to Schizo-phrenia, has been shown to be naturallymutated in the 129S6/SvEv genetic background (Koike et al., 2006), the same strainof mice as the Cln3−/− mice. This would suggest that whilesome of our findings may be pertinent to Schizophrenic-likesymptoms in Batten disease, that these mice are not anappropriate model for the study of this aspect of Battendisease.

The function of CLN3 protein remains unknown, however,recent studies have suggested a role for CLN3 in cellulartransport (Fossale et al., 2004; Luiro et al., 2004; Weimer et al.,2005). Therefore, loss of CLN3 may affect not only theintracellular transport of various cargos, but the communica-tion between neuronal populations. A recent study reportedthat lymphoblasts from patients with JNCL had decreasedlevels of the amino acid arginine, as well as altered transportof this molecule across the lysosomal membrane (Ramirez-Montealegre and Pearce, 2005). Although a common linkbetween dopamine and arginine metabolism is not known,there are certain reports of catecholamines influencing nitricoxide biosynthesis from arginine through transport into cellsof this amino acid (Lin et al., 2005). As studies of Cln3−/− miceunveil vulnerable cell populations, comparative studiesbetween the relative cellular expression of CLN3 and diseaseprogression could provide further insights into the function ofCLN3, and its influence on the metabolism of molecules suchas arginine and dopamine. In this study we provide novelevidence for oxidative damage within the striatum leading toselective degeneration of cells within the efferent nigralregion, providing potential insight into JNCL progression. Bynarrowing the cause of this neuronal loss to an enzymaticalteration within the striatum, this study has providedvaluable information on the pathogenesis of JNCL and couldprovide new therapeutic targets aimed at alleviating exptra-pyramidal motor deficits associated with the disease andpreventing further nigral degeneration.

4. Experimental procedures

4.1. Locomotor activity and motor function

Automated open field locomotor activity chambers equippedwith infrared photobeams (Med Associates, St. Albans, Ver-mont) were used to quantify locomotor activity. Mice wereinitially habituated to the locomotor activity chambers in two30-min sessions occurring on consecutive days. Animals werehabituated at the same time each day and housed in a room

with a 12-h light–dark cycle between testing. On the third day,locomotor testing was performed identical to the pre-testinghabituation for 30 min. Photobeam breaks were recordedevery 50 ms for 30 min for horizontal, vertical and ambulatorymovements. Separate groups of mice were tested at three,nine, and 12 months of age and results of locomotor activityare reported as the means of total distance each animaltraveled or the amount of time spent moving within the30 min test interval. To further discern differences in distanceand time traveled, we compared between genotype at each50 ms recorded interval for distance or time. Data wereanalyzed by two-way ANOVA and presented as mean±S.E.M.;asterisks indicate significance compared with control atP≤0.05 (Bonferroni’s post-hoc t-test).

4.2. Tissue preparation and immunoblotting

Wild-type (129/SvJ) and homozygous Cln3−/− mice on a 129/SvJbackground were used for this study. All procedures werecarried out in accordance with NIH guidelines and theUniversity of Rochester Animal Care and Use CommitteeGuidelines with all animals housed under identical condi-tions. Cortex and striatum samples were microdissected from3- and 9-month-old Cln3−/− and wild-type (129/SvJ) mice (n=6/test group). Samples were prepared solely from male mice toavoid the complication of sex differential expression of COMT.Tissue was flash-frozen immediately upon collection, andthen homogenized on ice in approximately four volumes ofice-cold 50 mM sodium phosphate buffer (0.5 mMDTT, pH 7.5)using a motor-driven homogenizer. Protease inhibitors(Sigma, St. Louis, MO) and 2×solubilization buffer (50 mMsodium phosphate buffer, pH 7.5, 0.5 mM DTT, 0.2% SDS) wereadded. Protein concentration was determined on aliquotsfrom each of the tissue homogenates using a Lowry proteinassay (BioRad Inc., Hercules, CA). SDS PAGE (10–13%) andimmunoblotting was performed using standard techniques(as outlined in Elshatory, et al., 2003). Antibodies used forimmunoblotting included anti-β-Actin (Sigma, St. Louis, MO),anti-COMT (Transduction Labs, Lexington, KY), anti-DAT,anti-D1α (Chemicon, Temecula, CA), and anti-MAO-B (SantaCruz Biotechnologies, Santa Cruz, CA). Immunoblots wereexposed to film for various lengths of time and scanned fordensitometry using AlphaImager v5.5 (Imgen Technologies,Alexandria, VA) and the mean pixel density of each protein ofinterest was normalized to the mean pixel density of β-actin.

4.3. MB-COMT and S-COMT activity assays

Cortex and striatum samplesweremicrodissected from 3- and9-month-old Cln3−/− and wild-type (129/SvJ) mice (n=3/testgroup). Following homogenization, 500 μl of homogenate wasultracentrifuged in a Beckman Optima TL ultracentrifugeusing a TLA 100.2 rotor for 30 min at 2 °C (48,000 rpm or100,000×g). The supernatant, enriched in S-COMT, was placedin a fresh tube, and the pellet, enriched in MB-COMT, wasresuspended in 500 μl of ice-cold 50 mM sodium phosphatebuffer (0.5 mM DTT, pH 7.5) and placed in a fresh tube. Prior toperforming the COMT assay, an aliquot of the soluble andpellet fractions were used to estimate protein concentrationand for Western blot analysis to confirm proper separation of

110 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

the MB- and S-COMT isoforms. The COMT assay used here is amodified version of those described previously (Reenila et al.,1995; Reenila et al., 1997; Ellingson et al., 1999; Pihlavisto andReenila, 2002). Reaction samples contained 180 μl of tissue(from either the S- or the MB-COMT-enriched fractions) in atotal reaction volume of 500 μl. MgCl2 was added to a finalconcentration of 5.5 mM, and the COMT co-substrate S-adenosylmethionine was added to a final concentration of2 mM. The COMT substrate used, dihydroxybenzoic acid(DHBAc), was added to a final concentration of 0.2 mM tostart the reactions. Samples were incubated at 37 °C for60 min, upon which time reactions were terminated with theaddition of 100 μl of 2 M HClO4. Reaction blanks contained allof the above components with the exception of tissue extract.Another reaction control contained all reaction componentsabove in addition to CaCl2 (2.5 mM), which, at this concentra-tion, inhibits COMT activity by about 50%. The reactionproduct peak for vanillic acid was detected at 4.6 min, withmaximal response detected at 325 mV. HPLC analysis wasperformed as described above, except for the injection volumebeing 20 μl. Vanillic acid standard was dissolved in DMSO anddiluted to the range of levels detected for experimentalsamples (Sigma, St. Louis, MO). Activity was expressed aspicogram of vanillic acid formed per minute per mg protein.Data were analyzed by two-way ANOVA and presented asmean±S.E.M.; asterisks indicate significance compared withcontrol at ⁎=P≤0.05, ⁎⁎=P≤0.01, ⁎⁎⁎=P≤0.001 (Bonferroni’s post-hoc t-test).

4.4. Dopamine, DOPAC and HVA level analysis

Aliquots from each tissue homogenate (800 μl) were deprotei-nized by the addition of 200 μl of 2 M HClO4 (Sigma, St. Louis,MO) and sonicated on ice using a tip sonicator to liberatebound catecholamines and metabolites. Samples were thencentrifuged for 10 min at 13,000 rpm (4 °C) and supernatantswere passed through 0.2 μm syringe-driven filters (VWRInternational, Bridgeport, NJ) and stored at −80 °C untilchromatographic analysis. Pellets from each spin werecollected, resuspended in 0.5 NaOH to neutralize, sonicated,respun, and supernatants collected for a Lowry protein assay.Standard solutions for HVA, Dopamine, and DOPAC wereprepared in the ranges of detected tissue metabolites (Sigma,St. Louis, MO). Metabolites were expressed as picogram of thecompound of interest per mg protein injected into the HPLCsystem. Samples were injected at a volume of 50 μl using anautosampler with a cooled tray system (ESA Model 542autosampler). A two-pump solvent delivery apparatus wasutilized (ESA 582 module), with compound separationachieved using ESA column 70-0636 (ESA, Chelmsford, MA).Fractionated samples were subsequently detected using anin-line electrochemical detection system, utilizing an 8-channel array with potentials set to 400, 325, 250, 175, 100,25, −50, and −125 mV (ESA Coularray 5600A detector). ESAMD-TM mobile phase was delivered at a rate of 0.9 ml/min forthe duration of the sample runs. Data was analyzed using theESA Coularray software package. Dopamine, HVA, and DOPACpeaks were detected at 3.33, 5.65 and 2.73 min, with maximalresponses found at channels set to 25, 250 and 100 mV,respectively. Values for each peak were then normalized to

total protein amount and data presented as ng of catechola-mine per mg protein. Analyses were performed using threeeach of 3- and 9-month-old Cln3−/− and wild-type (129/SvJ)mice, with brains microdissected into cerebellum, cortex, andstriatum. Homogenates were subdivided into five separatesamples prior to processing, each an n=15 for each test group.Data were analyzed by two-way ANOVA and presented asmean±S.E.M.; asterisks indicate significance compared withcontrol at ⁎=P≤0.05, ⁎⁎=P≤0.01, ⁎⁎⁎=P≤0.001 (Bonferroni’s post-hoc t-test).

4.5. Carbonyl dot blot

Protein samples were vacuum dot blotted (Gibco Life Tech-nologies, Rockville, MD) onto PVDF which had been activatedin 100% methanol and transfer buffer (25 mM Tris, 200 mMglycine, and 20% methanol) in the amount of 25 and 50 μg.Samples were collected, rinsed in water and incubated for15 min in Sypro Ruby Red (Molecular Probes, Eugene, OR).Following washing in water, images were captured under UVlight and used for densitometry of total protein amounts.Membranes were then incubated in 2 N HCl with DNPH(0.1 mg/ml) for 5 min, washed in 2 N HCl and then washedseven times in 100% methanol for 5 min each. Following onewash in PBST (10 mM sodium phosphate, pH 7.2, 0.9% NaCl,0.1% Tween 20), membranes were incubated for 1 h in blockingbuffer (PBSTwith 5% nonfat driedmilk), washed three times inPBST, and then carbonyl formation was detected by incuba-tion with anti-DNP diluted in blocking buffer for 2 h (Chemi-con, Temecula, CA). Blots were then washed in PBST andincubated in secondary antibody (Goat anti-rabbit-HRP, Che-micon, Temecula, CA) for 1 h at room temperature. Densito-metry of immunoblots was performed to determine theaverage pixel value over dot minus background and normal-izing to total protein loaded per spot based on Sypro Ruby Redstaining. Data were analyzed and presented as mean±S.E.M.;asterisks indicate significance compared with control atP≤0.05 (Student’s t-test).

4.6. Measurements of substantia nigra volume and cellcounts

Unbiased Cavalieri estimates of the volume of the substantianigra pars compact and pars reticulata were made from6 month male wild-type (129/SvJ) and Cln3−/− mice (n=4/testgroup) as described previously, with no prior knowledge ofgenotype (Bible et al., 2004; Pontikis et al., 2004). Theboundaries of nuclei were defined by reference to landmarksin Paxinos and Franklin (Paxinos and Franklin, 2001).Regional volumes were expressed in μm3 and the meanvolume of each region calculated for wild-type (129/SvJ) andhomozygous Cln3−/− mice. Neuronal survival within the parscompacta and pars reticulata were investigated using Stereo-Investigator software (Microbrightfield Inc., Williston, VT) toobtain unbiased optical fractionator estimates of cell num-bers from either Nissl stained sections at 6 months of age orcell type specific immuno-stained sections at 3, 6, or9 months of age. For immunohistochemical analysis, brainsamples were collected and sectioned as outlined above. Freefloating sections were processed using standard immuno-

111B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

histochemical techniques as described previously (Bible et al.,2004). Antibodies used were tyrosine hydroxylase (Chemicon,Temecula, CA) and parvalbumin (Swant Inc, Bellinzona,Switzerland). Unbiased optical fractionator cell counts wereperformed as described previously (Bible et al., 2004; Pontikiset al., 2004), with a random starting section chosen, followedby every second stained section thereafter. Only neuronswith a clearly identifiable nucleus were sampled and allcounts were carried out using a 100× oil objective (NA 1.4).Data were analyzed and presented as mean±S.E.M.; asterisksindicate significance compared with control at p≤0.05(Student’s t-test). The mean co-efficient of error (CE) for allindividual optical fractionator and nucleator estimates wascalculated according to the method of Gundersen and Jensenand was less than 0.08 in all these analyses (Gundersen andJensen, 1987).

Acknowledgments

The authors wish to thank Charlie Pontikis for assistance inanalysis of stereology data; and Timothy Curran, AndrewSerour, and Sarah Leistman for technical assistance. Thiswork was funded in part by National Institutes of Health (NIH)NS40580 and NS44310 (DAP), NS41930 (JDC), student fellow-ships NIH T32 MH065181 (JMW), NIH T32 ES07026 (DWS),Batten Disease Support and Research Association (JMW, JWB),The Luke and Rachel Foundation (DMR) and NIH R25GMO64113 (YME).

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at doi:10.1016/j.brainres.2007.05.018.

R E F E R E N C E S

Aberg, L., Liewendahl, K., Nikkinen, P., Autti, T., Rinne, J.O.,Santavuori, P., 2000. Decreased striatal dopamine transporterdensity in JNCL patients with parkinsonian symptoms.Neurology 54, 1069–1074.

Aberg, L.E., Rinne, J.O., Rajantie, I., Santavuori, P., 2001. A favorableresponse to antiparkinsonian treatment in juvenile neuronalceroid lipofuscinosis. Neurology 56, 1236–1239.

Axelrod, J., Tomchick, R., 1958. Enzymatic O-methylation ofepinephrine and other catechols. J. Biol. Chem. 233, 702–705.

Barisic, N., Logan, P., Pikija, S., Skarpa, D., Blau, N., 2003. R208Xmutation in CLN2 gene associated with reduced cerebrospinalfluid pterins in a girl with classic late infantile neuronal ceroidlipofuscinosis. Croat. Med. J. 44, 489–493.

Benedict, J.W., Sommers, C.A., Pearce, D.A., in press. Progressiveoxidative damage in the CNS of a murine model of juvenileBatten disease. Journal of Neuroscience Research.

Bible, E., Gupta, P., Hofmann, S.L., Cooper, J.D., 2004. Regional andcellular neuropathology in the palmitoyl protein thioesterase-1null mutant mouse model of infantile neuronal ceroidlipofuscinosis. Neurobiol. Dis. 16, 346–359.

Boudikova, B., Szumlanski, C., Maidak, B., Weinshilboum, R., 1990.Human liver catechol-O-methyltransferase pharmacogenetics.Clin. Pharmacol. Ther. 48, 381–389.

Brooks, A.I., Chattopadhyay, S., Mitchison, H.M., Nussbaum, R.L.,Pearce, D.A., 2003. Functional categorization of geneexpression changes in the cerebellum of a Cln3-knockoutmouse model for Batten disease. Mol. Genet. Metab. 78, 17–30.

Dalle-Donne, I., Giustarini, D., Colombo, R., Rossi, R., Milzani, A.,2003a. Protein carbonylation in human diseases. Trends Mol.Med. 9, 169–176.

Dalle-Donne, I., Rossi, R., Giustarini, D., Milzani, A., Colombo, R.,2003b. Protein carbonyl groups as biomarkers of oxidativestress. Clin. Chim. Acta 329, 23–38.

Ellingson, T., Duddempudi, S., Greenberg, B.D., Hooper, D.,Eisenhofer, G., 1999. Determination of differential activities ofsoluble and membrane-bound catechol-O-methyltransferasein tissues and erythrocytes. J. Chromatogr., B, Biomed. Sci.Appl. 729, 347–353.

Elshatory, Y., Brooks, A.I., Chattopadhyay, S., Curran, T.M., Gupta,P., Ramalingam, V., Hofmann, S.L., Pearce, D.A., 2003. Earlychanges in gene expression in two models of Batten disease.FEBS Lett. 538, 207–212.

Fan, J.B., Zhang, C.S., Gu,N.F., Li, X.W., Sun,W.W.,Wang,H.Y., Feng,G.Y., St Clair, D., He, L., 2005. Catechol-O-methyltransferasegene Val/Met functional polymorphism and risk ofschizophrenia: a large-scale association study plusmeta-analysis. Biol. Psychiatry 57, 139–144.

Fossale, E., Wolf, P., Espinola, J.A., Lubicz-Nawrocka, T., Teed, A.M.,Gao, H., Rigamonti, D., Cattaneo, E., MacDonald, M.E., Cotman,S.L., 2004. Membrane trafficking and mitochondrialabnormalities precede subunit c deposition in a cerebellar cellmodel of juvenile neuronal ceroid lipofuscinosis. BMCNeurosci. 5, 57.

Galderisi, S., Maj, M., Kirkpatrick, B., Piccardi, P., Mucci, A.,Invernizzi, G., Rossi, A., Pini, S., Vita, A., Cassano, P., Stratta, P.,Severino, G., Del Zompo,M., 2005. Catechol-O-MethyltransferaseValMet polymorphism in schizophrenia: associations withcognitive and motor impairment. Neuropsychobiology 52,83–89.

Glaser, C.B., Yamin, G., Uversky, V.N., Fink, A.L., 2005. Methionineoxidation, alpha-synuclein and Parkinson’s disease. Biochim.Biophys. Acta 1703, 157–169.

Guldberg, H.C., Marsden, C.A., 1975. Catechol-O-methyltransferase: pharmacological aspects and physiological role.Pharmacol. Rev. 27, 135–206.

Gundersen, H.J., Jensen, E.B., 1987. The efficiency of systematicstereology and its prediction. J. Microsc. 147, 229–263.

Hofman, I., Kohlschutter, A., Santavuori, P., Gottlob, I., Goebel,H.H., Lake, B.D., Schutgens, R.B.H., Greene, N.D., Leung, K.-Y.,Mitchison, H.M., Munroe, P.B., Taschner, P.E., 1999. TheNeuronal Ceroid Lipofuscinses (Batten Disease). IOS Press,Amsterdam, The Netherlands.

Holton, K.L., Loder, M.K., Melikian, H.E., 2005. Nonclassical,distinct endocytic signals dictate constitutive andPKC-regulated neurotransmitter transporter internalization.Nat. Neurosci. 8, 881–888.

Horowitz, A., Shifman, S., Rivlin, N., Pisante, A., Darvasi, A., 2005. Asurvey of the 22q11 microdeletion in a large cohort ofschizophrenia patients. Schizophr. Res. 73, 263–267.

Hsu, M., Srinivas, B., Kumar, J., Subramanian, R., Andersen, J., 2005.Glutathione depletion resulting in selective mitochondrialcomplex I inhibition in dopaminergic cells is via anNO-mediated pathway not involving peroxynitrite:implications for Parkinson’s disease. J. Neurochem. 92,1091–1103.

Kandel, E.R., Schwartz, J.H., Jessell, T.M., 2000. Principles of NeuralScience, 4th Edition. McGraw-Hill, New York, NY.

Kielar, C., Maddox, L., Bible, E., Pontikis, C.C., Macauley, S.L.,Griffey, M.A., Wong, M., Sands, M.S., Cooper, J.D., 2007.Successive neuron loss in the thalamus and cortex in a mousemodel of infantile neuronal ceroid lipofuscinosis. Neurobiol.Dis. 25, 150–162.

112 B R A I N R E S E A R C H 1 1 6 2 ( 2 0 0 7 ) 9 8 – 1 1 2

Koike, H., Arguello, P.A., Kvajo, M., Karayiorgou, M., Gogos, J.A.,2006. Disc1 is mutated in the 129S6/SvEv strain and modulatesworking memory in mice. Proc. Natl. Acad. Sci. U. S. A. 103,3693–3697.

Kovacs, A.D., Weimer, J.M., Pearce, D.A., 2006. Selectivelyincreased sensitivity of cerebellar granule cells to AMPAreceptor-mediated excitotoxicity in a mouse model of Battendisease. Neurobiol. Dis. 22, 575–585.

Lin, W.C., Tsai, P.S., Huang, C.J., 2005. Catecholamines’enhancement of inducible nitric oxide synthase-induced nitricoxide biosynthesis involves CAT-1 and CAT-2A. Anesth. Analg.101, 226–232 (table of contents).

Luiro, K., Yliannala, K., Ahtiainen, L., Maunu, H., Jarvela, I., Kyttala,A., Jalanko, A., 2004. Interconnections of CLN3, Hook1 and Rabproteins link Batten disease to defects in the endocyticpathway. Hum. Mol. Genet. 13, 3017–3027.

Lundstrom, K., Salminen, M., Jalanko, A., Savolainen, R., Ulmanen,I., 1991. Cloning and characterization of human placentalcatechol-O-methyltransferase cDNA. DNA Cell Biol. 10,181–189.

Malherbe, P., Bertocci, B., Caspers, P., Zurcher, G., Da Prada, M.,1992. Expression of functional membrane-bound and solublecatechol-O-methyltransferase in Escherichia coli and amammalian cell line. J. Neurochem. 58, 1782–1789.

Mitchison, H.M., Bernard, D.J., Greene, N.D., Cooper, J.D., Junaid,M.A., Pullarkat, R.K., de Vos, N., Breuning, M.H., Owens, J.W.,Mobley, W.C., Gardiner, R.M., Lake, B.D., Taschner, P.E.,Nussbaum, R.L., 1999. Targeted disruption of the Cln3 geneprovides a mouse model for Batten disease. The Batten mousemodel consortium. Neurobiol. Dis. 6, 321–334.

Paxinos, G., Franklin, K.B.J., 2001. The Mouse Brain in StereologicalCoordinates, 2nd Edition. Academic Press, San Diego, CA.

Pihlavisto, P., Reenila, I., 2002. Separation methods for catecholO-methyltransferase activity assay: physiological andpathophysiological relevance. J. Chromatogr., B Analyt.Technol. Biomed. Life Sci. 781, 359–372.

Pontikis, C.C., Cella, C.V., Parihar, N., Lim, M.J., Chakrabarti, S.,Mitchison, H.M., Mobley, W.C., Rezaie, P., Pearce, D.A., Cooper,J.D., 2004. Late onset neurodegeneration in the Cln3−/− mousemodel of juvenile neuronal ceroid lipofuscinosis is preceded bylow level glial activation. Brain Res. 1023, 231–242.

Pontikis, C.C., Cotman, S.L., Macdonald, M.E., Cooper, J.D., 2005.Thalamocortical neuron loss and localized astrocytosis in theCln3(Deltaex7/8) knock-in mouse model of Batten disease.Neurobiol. Dis. 20, 823–836.

Porzgen, P., Bonisch, H., Bruss, M., 1995. Molecular cloning andorganization of the coding region of the humannorepinephrine transporter gene. Biochem. Biophys. Res.Commun. 215, 1145–1150.

Ramirez-Montealegre, D., Pearce, D.A., 2005. Defective lysosomalarginine transport in juvenile Batten disease. Hum. Mol. Genet.14, 3759–3773.

Reenila, I., Tuomainen, P., Mannisto, P.T., 1995. Improved assay ofreaction products to quantitate catechol-O-methyltransferase

activity by high-performance liquid chromatography withelectrochemical detection. J. Chromatogr., B, Biomed. Appl.663, 137–142.

Reenila, I., Tuomainen, P., Soinila, S., Mannisto, P.T., 1997. Increaseof catechol-O-methyltransferase activity in rat brain microgliaafter intrastriatal infusion of fluorocitrate, a glial toxin.Neurosci. Lett. 230, 155–158.

Rinne, J.O., Ruottinen, H.M., Nagren, K., Aberg, L.E., Santavuori, P.,2002. Positron emission tomography shows reduced striataldopamine D1 but not D2 receptors in juvenile neuronal ceroidlipofuscinosis. Neuropediatrics 33, 138–141.

Rivett, A.J., Francis, A., Roth, J.A., 1983a. Localization ofmembrane-bound catechol-O-methyltransferase.J. Neurochem. 40, 1494–1496.

Rivett, A.J., Francis, A., Roth, J.A., 1983b. Distinct cellularlocalization of membrane-bound and soluble forms ofcatechol-O-methyltransferase in brain. J. Neurochem. 40,215–219.

Ruottinen, H.M., Rinne, J.O., Haaparanta, M., Solin, O., Bergman, J.,Oikonen, V.J., Jarvela, I., Santavuori, P., 1997. [18F]fluorodopaPET shows striatal dopaminergic dysfunction in juvenileneuronal ceroid lipofuscinosis. J. Neurol. Neurosurg. Psychiatry62, 622–625.

Salminen, M., Lundstrom, K., Tilgmann, C., Savolainen, R.,Kalkkinen, N., Ulmanen, I., 1990. Molecular cloning andcharacterization of rat liver catechol-O-methyltransferase.Gene 93, 241–247.

Sappington, R.M., Pearce, D.A., Calkins, D.J., 2003. Optic nervedegeneration in a murine model of juvenile ceroidlipofuscinosis. Invest. Ophthalmol. Visual Sci. 44,3725–3731.

Seigel, G.M., Lotery, A., Kummer, A., Bernard, D.J., Greene, N.D.,Turmaine, M., Derksen, T., Nussbaum, R.L., Davidson, B.,Wagner, J., Mitchison, H.M., 2002. Retinal pathology andfunction in a Cln3 knockout mouse model of juvenile neuronalceroid lipofuscinosis (Batten disease). Mol. Cell. Neurosci. 19,515–527.

Tai, C.H., Wu, R.M., 2002. Catechol-O-methyltransferase andParkinson’s disease. Acta Med. Okayama 56, 1–6.

Torres, G.E., Gainetdinov, R.R., Caron, M.G., 2003. Plasmamembranemonoamine transporters: structure, regulation andfunction. Nat. Rev., Neurosci. 4, 13–25.

Weimer, J.M., Chattopadhyay, S., Custer, A.W., Pearce, D.A., 2005.Elevation of Hook1 in a disease model of Batten disease doesnot affect a novel interaction between Ankyrin G and Hook1.Biochem. Biophys. Res. Commun. 330, 1176–1181.

Weinshilboum, R., Raymond, F., 1977. Variations incatechol-O-methyltransferase activity in inbred strains of rats.Neuropharmacology 16, 703–706.

Winqvist, R., Lundstrom, K., Salminen, M., Laatikainen, M.,Ulmanen, I., 1992. The human catechol-O-methyltransferase(COMT) gene maps to band q11.2 of chromosome 22 andshows a frequent RFLP with BglI. Cytogenet. Cell Genet. 59,253–257.