Sprouting of dopamine terminals and altered dopamine release and uptake in Parkinsonian dyskinaesia

Sprouting of dopamine terminals and altereddopamine release and uptake in ParkinsoniandyskinaesiaJoohyung Lee,1Wen-Mei Zhu,1Davor Stanic,1David I. Finkelstein,2 Marjorie H. Horne,3

Jasmine Henderson,4 Andrew J. Lawrence,1,5 Louise O’Connor,1DorisTomas,1 John Drago1,5 andMalcolm K. Horne1,5

1Brain Injury and Repair, Howard Florey Institute,University of Melbourne, Parkville,VIC, 3010, 2The Mental Health ResearchInstitute of Victoria,155 Oak Street, Locked Bag11, Parkville,VIC, 3052, 3Faculty of Education, Australian Catholic University,115 Victoria Pde, Fitzroy,VIC and 4Department of Pharmacology, Bosch Institute and School of Medical Sciences, BoschBuilding, University of Sydney, NSW, 2006, Australia and 5Centre for Neuroscience, University of Melbourne, Parkville, VIC,3010, Australia

Correspondence to: Malcolm K. Horne, Howard Florey Institute, Level 2 Alan Gilbert Building, 161 Barry St, Carlton Sth,3053, AustraliaE-mail: [email protected]

Failed storage capacity, leading to pulsatile delivery of dopamine (DA) in the striatum, is used to explain theemergence of ‘wearing off’ and dyskinaesia in Parkinson’s disease. In this study, we show that surviving DA neu-rons in 6 -OHDA lesioned rats sprout to re-innervate the striatum, andmaintain terminal density until »60% ofneurons are lost.We demonstrate that DA terminal density correlates with baseline striatal DA concentration([DA]). Electrochemical and synaptosome studies in 6 -OHDA lesioned rats and primates suggest that impairedstriatal DA re-uptake and increased DA release from medial forebrain bundle fibres contribute to main-taining striatal DA levels. In lesioned rats where terminal density fell by 60% or more, L-DOPA administrationincreased striatal DA levelsmarkedly.The striatal [DA] produced by L-DOPAdirectly correlated with the extentof dyskinaesia, suggesting that dyskinaesia was related to high striatal [DA]. While sprouting and decreaseddopamine uptake transporter function would be expected to contribute to the marked increase in L-DOPAinduced [DA], the increased [DA] was most marked when DAergic fibres were `60% denervated, suggestingthat other release sites, such as serotonergic fibres might be contributing. In conclusion, the extent of dyskinae-sia was directly proportional to the extent of DA terminal denervation and levels of extra-synaptic striatal DA.We propose that sprouting of DA terminals and decreased dopamine uptake transporter function prevent theappearance of Parkinsonian symptoms until about 60% loss of nigral neurons, but also contribute to dysregu-lated striatal DA release that is responsible for the emergence of dyskinaesia and ‘wearing off’.

Keywords: dopamine uptake transporter; levodopa; Parkinson’s disease; striatum; voltammetry

Abbreviations: AIM=abnormal involuntary movement; AUC=area under curve; DA=dopamine; DAT=dopaminereuptake transporter; HPLC=high-performance liquid chromatography; IAS= index of arbour size; MFB=medial forebrainbundle; SERT=serotonin re-uptake transporter; TH=tyrosine hydroxylase

Received January 8, 2008. Revised March 5, 2008. Accepted April 14, 2008. Advance Access publication May16, 2008

IntroductionL-DOPA ameliorates symptoms of early Parkinson’s disease,implying that increasing trans-synaptic delivery of dopa-mine (DA) by surviving terminals compensates for loss ofrelease sites. In the absence of L-DOPA, motor function inParkinson’s disease is reduced, reflecting impaired endo-genous DA neurotransmission, whereas L-DOPA therapy

improves motor function (Nutt et al., 2002) and byimplication, DA neurotransmission. L-DOPA treatmentimproves motor function at the onset of disease, butwithin 5 years of treatment, about 50% of patients findthat this beneficial effect of L-DOPA is marred by thedevelopment of involuntary, often debilitating, movementsknown as dyskinesias (Bezard et al., 2001) as well as motor

doi:10.1093/brain/awn085 Brain (2008), 131, 1574^1587

� The Author (2008). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]

by guest on May 22, 2016

http://brain.oxfordjournals.org/D

ownloaded from

http://brain.oxfordjournals.org/

fluctuations, characterized by shortened duration ofefficacy, manifest as ‘wearing off’ (Poewe et al., 1986;Hely et al., 1994; Nutt et al., 1995, 2002; Reardon et al.,1999; McColl et al., 2002). The efficacy of L-DOPA isshortened rather than reduced and duration roughlycorrelates with plasma levels of L-DOPA (Cotzias et al.,1967; Nutt et al., 1995), suggesting that the capacity to storeendogenously synthesized L-DOPA is compromised.

Dyskinaesia almost invariably accompanies the short-ening benefit from L-DOPA and the threshold and timecourse for dyskinaesia and the anti-Parkinsonian effect ofL-DOPA are similar (Mouradian et al., 1988; Nutt et al.,1992, 2002; Metman et al., 1997; Nutt, 2001). Whiledyskinaesias, motor fluctuations and the short-duration res-ponse are not necessarily the same thing, it seems likely thatthey are interrelated inextricably through some common,causal mechanism. Any theory explaining these phenomenamust explain how, in the face of failing storage and endo-genous DA neurotransmission, L-DOPA can deliver ade-quate DA to produce these increased unwanted movements.The failing endogenous capacity presumably occurs becauseof diminishing synthetic capacity, as a consequence ofterminal loss. Yet, when these terminals are provided withthe DA precursor L-DOPA, they can provide sufficient DAto not only improve motor response but also produceincreased and superfluous movements.

Failing storage of synthesized DA, or the ‘storagehypothesis’ is the most commonly proffered explanationfor ‘wearing off’. The normal nigrostriatal neuron istonically active (Grace and Bunney, 1984a, b), firing atabout 6 Hz and continuously releasing low levels of DA, butDA release is increased with bursts of activity associatedwith salient events (Mirenowicz and Schultz, 1994).Synthesis (Onali et al., 1988; Lindgren et al., 2001), release(Ungerstedt et al., 1982; Bowyer and Weiner, 1987), andre-uptake of DA (Hersch et al., 1997; Robinson, 2002) areunder neuronal control to maintain its continuous delivery.According to the storage hypothesis, as the number ofstriatal DA terminals fall, synthetic capacity is increased sothat DA neurotransmission can be maintained while thesubstrate is provided by L-DOPA. Eventually there is noendogenous DA storage because all DA synthesized fromL-DOPA is immediately released. Thus, as compensatorymechanisms are overwhelmed and storage capacity fails,delivery becomes pulsatile and dysregulated. However thereis surprisingly little direct evidence to support this theoryor to explain how it may work. It seems unsatisfactory thatrelease is maintained yet storage falls as these mechanismsare closely coupled. Recently, it has been suggested thatserotonin innervation of the striatal complex, whichsurvives the effects of Parkinson’s disease, may be impor-tant in converting exogenous L-DOPA to DA (Carlssonet al., 2007; Carta et al., 2007). While serotonergic neuronscan store and release DA in an activity-dependent manner,they have no functional DA re-uptake and storage andrelease would be poorly regulated. The implication is that

the role of serotonergic neurons becomes increasinglyimportant as ‘wearing off’ and dyskinaesias becomeprominent (Carta et al., 2007).

In this study, we propose that impaired DA transporter(DAT) function and dysregulated DA release, appearingas a consequence of compensatory sprouting of survivingneurons, may also contribute to impaired storage andabnormal DA release. DA neurons that survive injurysprout to maintain DA terminal density as a compensatoryresponse (Stanic et al., 2003a). Because of this regulatoryprocess, in rodents, normal DA terminal density is main-tained in the striatum until 470% of nigral cells aredestroyed (Finkelstein et al., 2000; Stanic et al., 2003a).Although sprouting maintains terminal density, newlyformed DA terminals have altered structure and function(Stanic et al., 2003b) and have a diminished capacity forDA re-uptake. The net effect is that re-innervating ter-minals would be capable of delivering a bolus of DA inresponse to L-DOPA but impaired re-uptake will result inincreased DA diffusion, affecting more sites than would beanticipated from the density of innervation alone, an effectcompounded by the increased terminal arbour. New ter-minals formed by sprouting have both a greater number ofvesicles and larger-sized vesicles, which would intuitivelysuggest that these terminals are capable of delivering largeramounts of DA into the synaptic cleft (Stanic et al., 2003b).Although larger vesicle numbers and size suggest increasedcapacity for DA release, it may also reflect increaseddemand for synthesis in lieu of the impaired transportthrough the DAT. As a consequence of large terminalarbours, impaired synaptic clearance of DA, and alteredrelease capacity, there will be abnormal DA delivery tounusually large regions of the striatum. This would providea mechanism for impaired storage yet augmented releaseimplied by the storage hypothesis.

This study measured striatal DA release and re-uptakein dyskinetic animals and correlated this with the extent ofterminal loss and sprouting in rodents and primates. Thefindings show that despite falling terminal density, extra-synaptic DA release increases. These data provide directsupport for the storage hypothesis as an explanation fordyskinaesia and also provide support for the view that asstriatal terminal density falls, DA is released almost as soonas it synthesized, and that re-uptake is greatly reduced toaugment extra-synaptic DA.

Materials and MethodsAll methods conformed to the Australian National Health and

Medical Research Council published code of practice for the useof animals in research and were approved by the MonashUniversity and Howard Florey Institute Animal Ethics

Committee. One hundred male Wistar rats weighing 250–350 gwere used. Twenty-two male Common Marmosets (Callithrixjacchus), aged between 2 and 3 years at the commencement of the

study and weighing �350–450 g were used in this study.

Sprouting of dopamine terminals in dyskinaesia Brain (2008), 131, 1574^1587 1575



ownloaded from


Lesion of SNpc in rats and marmosetsIn rats, varying lesions of the SNpc were made by injecting

1.5–3ml of a 2.5 mg/ml solution of 6-OHDA (Sigma-RBI, St. Louis,

MO, USA) into the right SNpc at 3.5 mm anterior, 1.7 mm lateral

to lambda and 7.1 mm below dura and 3.5 mm anterior; 2.1 mm

lateral and 6.8 mm below dura (Paxinos and Watson, 1998).

Anaesthesia was induced with sodium pentobarbitone; 60 mg/kg

i.p with atropine (0.24 mg/kg i.p.) and maintained with a mixture

of xylazine (2.4 mg/kg i.m) and ketamine (28 mg/kg i.m).In Marmosets, partial lesions of the SNpc were made by

injecting 2.5 mg/ml of 6-OHDA into four sites in the SNpc (site 1:

AP = 6.0 mm posterior to the inter-aural line, ML = 1.5 mm lateral

to the midline, DV = 6.5 mm below the dura; site 2: AP = 6.5 mm,

ML = 2.5 mm, DV = 7.0 mm, site 3: AP = 5.0 mm, ML = 1.5 mm,

DV =6.0 mm, site 4: AP = 5.0 mm, ML = 2.75 mm, DV = 6.5 mm).

Anaesthesia was induced with saffan (0.5 ml of 12 mg/ml solution,

i.m.) and maintained with injections of 0.1 ml of a 7:3 mixture of

ketamine and xylazine (28 and 2.4 mg/kg, i.m.).

Behavioural studies

Assessment of abnormal involuntarymovements (AIMs)in ratsRats with varying lesion sizes received L-DOPA and benserazide

(20 and 6.5 mg/kg i.p., twice daily) for 2 weeks. The development

and manifestation of AIMs, a correlate of dyskinaesia in rodents,

were monitored on treatment Day 1, 5 and 10, as previously

described (Cenci et al., 1998; Winkler et al., 2002). To briefly

summarize, rats were observed for 1 min every 20 min after the

injection of L-DOPA (20 mg/kg i.p.) or vehicle for 3 h. Four

subtypes of AIMs were scored: (i) axial AIMs: dystonic posturing

or choreiform twisting of the neck and upper body towards the

side contralateral to the lesion; (ii) limb AIMs: abnormal,

purposeless movements of the forelimb and digits contralateral

to the lesion; (iii) orolingual AIMs: empty jaw movements and

contralateral tongue protrusion and (iv) locomotive AIMs:

increased locomotion with contralateral side bias. Each subtype

was scored as 0: absent: 1, present during less than half of the

observation time: 2, present during more than half of the

observation time: 3, present all the time but suppressible by a

startling stimulus; 4, present all the time and not suppressible. The

total AIMs from one observation were 16 and as there were 10

observations (200 min), allowing a maximum score of 160.

Assessment of dyskinesia in primatesDyskinaesia in Marmoset monkeys was measured for 60 min,

commencing 30 min after oral administration of a single dose of

20 mg/kg L-DOPA and 3.75 mg/kg Benserazide (Madopar� tablets

in honey and administered orally, Roche, Australia). Dyskinetic

movements were assessed every 10 min and were scored as:

0 = absent, 1 = mild: fleeting and rare (55 in 10 min) dyskinetic

postures and movements, 2 = moderate: more prominent and

abnormal movements (5–20 in 10 min), but not interfering with

normal behaviour, 3 = marked: frequent (21–50 in 10 min) and at

times continuous dyskinaesias intruding on the normal repertoire

of activity and 4 = severe: virtually continuous dyskinetic activity,

disabling to the animal and replacing normal behaviour as

previously described (Pearce et al., 1995).

BradykinaesiaThis was the time taken by marmosets to reach through a slot in aPerspex barrier and grasp a small piece of sweetened bread(�0.5 cm cube). This test was performed five times on each testweek (Henderson et al., 1998).

DA release and uptake in vivo and in vitro

Measurement of striatal DA release using in vivovoltammetry in ratsThirty micrometer nafion-coated (5% solution, Sigma–Aldrich)carbon fibre microelectrodes (Textron Systems, Lowell, MA, USA)were glued to a fused glass capillary (i.d. = 40 mm, SGE, VIC,Australia) with separation between carbon-fibre electrode tips andcapillary delivery tube of �200mm with an Ag/AgCl referenceelectrode. Voltammograms were recorded at 10 Hz (potential+550 mV, square-wave pulses), electrode linearity and sensitivitywere determined against standard solutions of dopamine (Sigma–Aldrich) and average responses (nA/ms) were translated into DAconcentration. DA concentrations in the dorsal striatum wererecorded by inserting carbon fibre electrodes and 325� 70 nlof 200 mM DA was injected through the fused silica capillary.KCl (70 mM, 200� 50 nl) was applied through the capillaryto measure DA release from the dorsal striatum. Clearancewas expressed in terms of the time (in seconds) for the DAelectrochemical signal to decrease to 50% of peak amplitude (T50,shown in Fig. 3).

Measurement of evoked DA release with medial fore-brain bundle (MFB) stimulation in rats and marmosetsMFB stimulation was delivered through a twisted, bipolar stimu-lation electrode (Plastics One, Roanoke, VA, USA) positionedin the right MFB (coordinates for rat: AP �4.4 to �4.6 mmfrom bregma, ML 1.4 mm and DV 8.0 to 9.0 mm below dura;marmoset: AP +6.5 mm anterior to the inter-aural line; ML1.4 mm lateral to the midline; DV 5.0 to 5.5 mm below dura. Theposition of the stimulating electrode was adjusted to record amaximal response in the recording electrode. A constant currentstimulator delivered 50 Hz trains of �500mA square wave pulsesfor 2 ms repeated every 90 s. Each train (50 Hz at 2 ms) evoked DArelease with a ‘bell-shaped’ curve appearance (Fig. 3E). Stimulustrains were delivered in a block of six and the DA curves resultingfrom each train were averaged together to make a single DArelease curve for that time point. DA release and uptake overlonger periods were recorded by repeating this protocol every20 min for 140 min. The DA signal was conditioned (Clampfit,Version 8.2, Axon Instrument, USA) by stimulus-locked averagingfor 10 s before and 60 s after the stimulus, artefact rejection,application of high (0.00125 Hz cut-off) and low (1 Hz) pass filter(Clampfit, Butterworth IR filter) and a Boxcar filter with 9 pointsrunning average. This smoothed DA waveform was used forsubsequent measurement and statistical analysis. Electrodes wereelectrochemically treated by passing 20 Hz triangular pulses for50 s followed by a 5-s 2.2 V square pulses and then coated withtwo layers of nafion. Electrodes were calibrated before and afterimplantation and were found to have selectivity for DA overascorbic acid of4200 : 1. Chloride-plated silver wire was used as areference electrode. Twenty minutes was allowed for equilibrationafter implantation and prior to recording.

1576 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


Measurement of extra-synaptic DA release usingstriatal in vivomicrodialysis in ratsRats were anaesthetized with urethane (0.8 g/kg i.p.) and amicrodialysis probe (CMA/12, membrane length 3 mm, OD0.5 mm: Stockholm, Sweden) was placed in the right striatumat the coordinates with respect to bregma; AP +0.50 mm; ML+2.5 mm; DV 6.5 mm below dura. A modified Ringer solution(147 mM NaCl, 4 mM KCl, 1.2 mM CaCl2, 1.1 mM MgCl2) wasperfused through the probes at 1 ml/min. Following 120 min ofequilibration, samples were collected on ice every 20 min (20 ml)into a vial containing 10 ml of 0.1 M perchloric acid. After threebasal samples collections, L-DOPA and benserazide (20 and6.5 mg/kg i.p.) was administered and another seven samples werecollected. The in vitro recovery of the probes for DA was 15%.

Detection of DA levels with high-performance liquidchromatography (HPLC)The DA levels in microdialysate samples were measured by HPLC.Detection was via a BAS LC4B glass carbon-working electrodeoperating at +650 mV versus an Ag/AgCl reference. The mobilephase comprised the following (mM): KH2PO4, 70; EDTAdi-sodium salt, 0.5; octane sulphonic acid, sodium salt, 8; pH3.0 with 15% HPLC grade methanol (Lawrence et al., 2005). Flowrate was 1 ml/min. The detection limit of this system, defined asmasses of standards producing peak heights of double the baselinenoise, was 1.5 pg for DA (all standards from Sigma-Aldrich, USA).

DA uptake and clearance in striatal synaptosomesThe rate of DA clearance and release was measured electro-chemically in striatal synaptosome preparations, as previouslydescribed (Stanic et al., 2003b). To briefly summarize, rats weredecapitated, brains removed and cut in a coronal plane (4.3 mmposterior to bregma) to separate the striatum from the SNpc. Thestriatum was then hemisected and the dorsal striatum placedimmediately in KRH buffer (mM: NaCl, 125; K2HPO4, 1.5;MgSO4, 1.5; CaCl2, 1.25; D-glucose, 10; HEPES, 25; ascorbic acid,0.1; pargyline, 1; and EDTA, 0.1, pH 7.4) at 4�C and oxygenated.The brain areas were homogenized in 25 ml of cold sucrose,centrifuged at 2000 g for 10 min at 4�C. The supernatant wascentrifuged at 16 000 g for 15 min at 4�C and the resulting pelletremained on ice until it was re-suspended in KRH buffer so as toobtain a concentration of 1000 w/v for the transport assay orin vitro electrochemistry. One millilitre of synaptosome suspensionwas added to each well and pre-incubated at 37�C for 3 min.Recordings were made as described above for in vivo measure-ments. In preparations from normal and lesioned animals, 12 ml of0.25 mM DA were injected by micropipette into the synaptosomesuspension. At this concentration, clearance of DA into synapto-somes was too rapid in samples from normal animals to allow ameaningful comparison with clearance from lesioned animals.Therefore, for measurement of DA transport into synaptosomesfrom normal animals, the concentration of DA added to thepreparation was subsequently increased to 0.5 mM. DA releasefrom synaptosomes was examined by adding 50 ml 1 M KCl to thepreparation once DA concentration in the well returned to abaseline level (i.e. when no further DA was being taken up intosynaptosomes). Clearance was expressed as time (in seconds) forthe DA electrochemical signal to decrease to 50% of peakamplitude (T50).

Histology and immunohistochemistryDAT, serotonin re-uptake transporter (SERT), Fos B andtyrosine hydroxylase (TH) immunohistochemistryTwenty micro meter-thick coronal sections were cut seriallythrough the striatum and immunoreacted against dopaminereuptake transporter (DAT), serotonin reuptake transporter(SERT) and Fos B antibodies. DAT immunohistochemistry wasperformed (Stanic et al., 2003b) by incubating sections in rat anti-DAT primary antibody (Chemicon, Temecula, CA, 1:3000)followed by a biotinylated secondary antibody (rabbit anti-ratIgG, 1:500, Vector, Burlingame, CA, USA) and reacted with cobaltand nickel-intensified diaminobenzidine (DAB, Sigma-Aldrich PtyLtd). SERT immunohistochemistry was performed by incubatingsections with mouse anti-SERT primary antibody (Chemicon,Temecula, CA, 1:1000) followed by a biotinylated secondary anti-body (sheep anti-mouse IgG 1:500; Chemicon, Temecula, CA) andreacted with cobalt and nickel-intensified DAB. Fos B immuno-histochemistry was performed by incubating sections in rabbitanti-Fos B primary antibody (Santa Cruz, CA, 1:800) followed bya biotinylated secondary antibody (anti-rabbit IgG, 1:800, Vector,Burlingame, CA) and reacted with cobalt and nickel-intensifieddiaminobenzidine (DAB, Sigma-Aldrich). Fifty micro meter-thickcoronal sections were cut serially through the SNpc and every fifthsection was stained with neutral red and a parallel series reactedwith antibodies against tyrosine hydroxylase (TH), as previouslydescribed (Stanic et al., 2003b). The following antibodies anddilutions were used: Anti-TH primary antibody (BoehringerMannheim, Castle Hill, Australia, 1:3000) and a biotinylated secon-dary antibody (Sheep, anti-Mouse IgG, 1:500, Silenus, Australia).

Fractionator sampling schemeThe fractionator design for estimating the number of SNpcneurons and the number of DAT-ir axonal varicosities werepublished in detail previously with the following modifications(Stanic et al., 2003a). Counts of SNpc neurons, stained for neutralred or TH-ir, were made at regular predetermined intervals(x= 125mm, y= 200 mm) derived by means of a grid program(Stereo Investigator, MicroBrightField, VT, USA) with an unbiasedcounting frame (40 mm� 27 mm = 1080mm2). Therefore, the areasampling fraction is 1080/(125� 200) = 0.043. In all animals,50 mm thick sections through the SNpc, each 180 mm apart, wereanalysed, the fraction of sections sampled being 0.25. Lesion sizewas the number of SNpc neurons estimated in lesioned animals,expressed as a percent of the average number determined in thenormal SNpc.

DAT-ir and TH-ir varicosities in the dorsal 1.5 mm of thestriatum ipsilateral to the lesioned SNpc were counted from 20 mmthick serial sections, each 500mm apart (the section samplingfraction being 1/25). Counts were made at regular intervals(x= 800mm, y= 800 mm) with the area of unbiased counting framebeing 40mm2 (8mm� 5 mm). Thus, the area sampling fraction is40/(800� 800) = 0.00 006 25. The entire z-dimension of eachsection was sampled, the section thickness sampling fractionbeing 1. Varicosities were identified as a dilated element (usuallyround or oval shaped) of immunoreactive axons (Finkelstein et al.,2000). The total number of SNpc neurons or axonal varicositieswas estimated by multiplying the number of neurons orvaricosities counted within the sampled regions with thereciprocals of the fraction of the sections sampled, fraction ofthe sectional area sampled and the fraction of the section thickness




ownloaded from


sampled (West et al., 1991). The density of axonal varicosities wascalculated by dividing the estimated number of varicosities by thevolume of dorsal striatum estimated by the sampling scheme.Coefficients of error and coefficients of variance were calculated asestimates of precision.

StatisticsFor comparisons of T50, peak [DA] and [DA] from baseline,a Student unpaired t-test was applied to estimate overallsignificance between control and lesion groups followed by posthoc t-tests. In DA release or AIMS studies, the comparisons of DAlevels or AIM scores that had been recorded repeatedly,were performed using a repeated-measures ANOVA. Bonferronipost hoc tests were used to estimate overall significanceswhere appropriate. Relationships between variables werestudied using either linear or sigmoidal regression whereappropriate. A probability level of 5% (P50.05) was consideredsignificant for all statistical tests. Data are expressed asmeans� SEM.

Results6-OHDA induced lesions of varying sizes were made in therat SNpc. Sixteen weeks was allowed for stable sprouting ofDA terminals and re-innervation of the striatum to occur(Stanic et al., 2003a). At the end of the 16 weeks, animalswere treated with L-DOPA or vehicle for 2 weeks and ass-essed for dyskinaesia and extra-synaptic DA levels anduptake in the striatum. The animals were then killed and

the density of DAT-ir terminals in the striatum and TH-ircells in the SNpc were estimated using formal stereologicaltechniques.

The relationship between lesion size, striatalDAT terminal density and sprouting of DAterminals following SNpc lesions in ratsThe density of DAT-ir terminals in the striatum was plottedagainst the number of surviving TH-ir neurones inthe SNpc (Fig. 1A). In keeping with previous studies(Finkelstein et al., 2000; Stanic et al., 2003a), the density ofDAT-ir terminals 16 weeks after a lesion was normal inanimals with lesion sizes less than 60% (Fig. 1A). As lesionsize exceeds 60%, terminal density fell progressively to zero.This implies that when lesion size is 560%, new terminalsform or sprout, from the axons of the surviving neurons asa compensatory response. As discussed in detail previously(Parish et al., 2001, 2002; Finkelstein et al., 2004), an indexof this sprouting can be obtained by dividing the density ofDAT-ir terminals by the number of TH-ir neurons in theSNpc, to give an ‘index of arbour size’ (IAS). Expressing theIAS of the lesioned striatum as a percentage of the IASof the intact striatum provides percentage of sprouting(Fig. 1B). Figure 1B demonstrates that the percentage of DAsprouting is proportional to lesion size, confirming thatpre-synaptic DA terminals compensate for neuronal lossuntil the lesion size become 460%.

Fig. 1 SNpc lesions of varying sizes were made in 28 rats (with eight controls). (A) Estimates of DA terminal density (DAT-ir terminals)for each lesioned animal were plotted against lesion size (number of TH-ir neurons in the SNpc as a percentage of normal) and themean (�SEM) terminal density of control animals is shown as a circle. (B) Extent of sprouting was estimated by expressing ‘index ofarbour size’ (see Results section) of each animal as a percentage of the normal ‘index of arbour size’ and plotting this against lesionsize (number of TH-ir neurones in the SNpc as a percentage of normal). (C^F) Photomicrographs of DAT-ir axonal varicosities andfibres in the dorsal striatum of (C) Normal, (D) normal lesioned (E) partial lesioned and (F) extensive lesioned animals, 16 weeksfollowing SNpc lesion. (Scale bar=20 mm.)

1578 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


An important point is that the extent of denervation ofthe striatum is likely to be more relevant to DA deliveryinto the striatum than the extent of neuronal loss.To reflect this we will express terminal loss as a per-centage of denervation compared with the normal stria-tum: % denervation = (1�normal density)� 100. Normal,normal lesioned, partial and extensive denervations aredefined as 0, 1–30, 30–60 and 60–100% denervation,respectively (Fig. 1C–F).

The relationship between striataldenervation, striatal DA concentrationand dyskinaesia in ratsTo determine whether there was a relationship between theextent of denervation and the development of dyskinaesias,we assessed AIMs in rats following a single dose of L-DOPAon days 1, 5 and 10. On Day 10 of testing, dyskineticmovements, measured by the total AIM score, appearedwithin 20 min of L-DOPA administration and peakedbetween 80 and 120 min (Fig. 2A). The severity of thedyskinaesia was the greatest in the extensively denervatedanimals, whilst the normal animals did not exhibit anydyskinaesia following L-DOPA administration (Fig. 2A).The total AIM scores (the sum of the AIM scores from eachtime point from 0 to 180 min) between Day 1, 5 and 10were similar (P40.05, one-way ANOVA, Fig. 2B), indicat-ing that severity of dyskinaesia did not depend on previousexposure to L-DOPA. Dyskinaesia only became apparentwhen there was loss 460% of DAT-ir terminals in thestriatum and increased markedly in severity with furtherloss of terminal density (Fig. 2C).

The concentration of extra-synaptic DA in the striatumthat followed the administration of a single dose ofL-DOPA (20 mg/kg, i.p.) was measured using in vivomicrodialysis and HPLC (Fig. 2D and E). L-DOPAtreatment did not affect the baseline DA concentration inunlesioned animals, as there was no significant differencebetween 14 days of L-DOPA treatment (95.0� 5.2 fmol/sample) and 14 days of vehicle treatment (102.3� 3.4 fmol/sample). However, basal striatal DA levels fell progressivelyand linearly (r= 0.8) from �100 fmol/20 min of dialysates toabout �50–70 fmol/sample of dialysates as the denervationfell from normal to extensive loss of terminals (Fig. 2F).

A single dose of L-DOPA (20 mg/kg, i.p.) produced aninsignificant rise in DA concentration in unlesioned animalsand animals with lesions but normal density of DAT-irterminals. However, when DAT-ir terminals were reduced,there was a robust rise in DA concentration: 210% (partial)and 510% (extensive) increase, respectively (Fig. 2D and E).The DA concentrations peaked (156� 17 and 354� 32fmol/sample, respectively) between 40 and 60 min afterL-DOPA administration (Fig. 2D and E). In rats withmarkedly reduced DAT-ir terminal density, striatal DAlevels remained significantly elevated until recording ceased,140 min after L-DOPA injection (two-way ANOVA, Fig. 2D

and E). The rise in striatal [DA] was similar in vehicle andL-DOPA-treated rats.

Total striatal DA concentration was plotted against thedensity of DAT terminals in the striatum (Fig. 2G),showing that, following a dose of L-DOPA, striatal [DA]did not rise above baseline until there was about 60% lossof terminals. The relationship between density of DATterminals and total [DA] (Fig. 2G) was similar to thatbetween density of DAT terminals and total AIM score(Fig. 2C). This impression was confirmed by plotting AIMSscore against DA concentration (Fig. 2H) and showed thatseverity of dyskinaesia was linearly related to striatal DAconcentration (r= 0.9).

Striatal DA re-uptake is impaired in newlysprouted DA axon terminals in ratsWe examined DAT function in lesioned animals to establishwhether altered re-uptake may contribute to the very highlevels of striatal [DA] following a dose of L-DOPA. This wasdone using striatal synaptosome preparations and measur-ing clearance of DA from the striatum with carbon fibreelectrodes located in the dorsal striatum of lesioned andnormal rats. Rats with SNpc lesions 465% were usedbecause it was not possible to estimate percentage ofdenervation, especially with the synaptosomal preparation.

Figure 3A provides an example of the measurement ofDA concentration in the dorsal striatum in vivo, madebefore and after local application of DA (200mM). Innormal animals, DA concentration rose rapidly to a peakand was also cleared promptly. In a lesioned animal, thetime to peak DA concentration was significantly longer andclearance was greatly prolonged. Clearance of DA, measuredas the time taken for DA concentration to fall from peaklevel to 50% of peak level (T50), was approximately twice aslong in lesioned animals (P50.001, in vivo, Fig. 3B). DAclearance was consistently prolonged at each of three sitesin the striatum (data not shown). Peak amplitudes of DAconcentration were similar in lesioned and control animals(Fig. 3C, in vivo DA) as was peak DA concentration evokedby KCl (P= 0.074, in vivo KCl, Fig.3C). Rate of DAclearance (T50) was also measured from in vitro synapto-some preparations extracted from the dorsal striatum ofnormal (n= 6) and 6-OHDA lesioned rats (n= 12). Averagelesion size was 59� 29%. Clearance of DA into synapto-somes was measured electrochemically by directly injectingknown amounts of DA into the synaptosome preparationresulting in the same concentration of DA. Clearance of DAby synaptosomes from lesioned animals was markedlyprolonged (Fig. 3B, in vitro). Clearance of 0.25 nM DA/mgby synaptosomes from lesioned animals was measured, butbecause the clearance by normal synaptosomes was somuch more rapid, a greater concentration of DA was usedfor normal animals (0.5 nM DA). Peak amplitudes of DAconcentration evoked by KCl were the same in both normaland lesioned rats (Fig. 3C, in vitro).




ownloaded from


The binding properties of DAT in the rat dorsal striatumwere measured using [3H]Mazindol and the data arerepresented in Table 1. As previously shown (Stanic et al.,2003b), the Kd value of the high-affinity site (Kd1) in lesioned

animals was similar to that observed in normal animals,although the density (Bmax1) was reduced by almost 40%,P50.05). The affinity of the second binding site (Kd2) wasreduced by lesioning, concurrent with a 5-fold increase in the

Fig. 2 (A) Time course of total AIM scores/observation in rats following a single dose of L-DOPA (20mg/kg i.p.) on Day 10 of behaviouraltesting.There was little evidence of dyskinaesia in animals without lesions (open triangle) or with lesions but normal density (filled square).Dyskinaesia was marked in animals with extensive (filled triangle) and to a lesser extent, partial denervation (filled circle). (B) The totalAIM scores on Days 1, 5 and 10 of daily L-DOPA administration to rats with extensive striatal denervation. The total AIM score is the sumof scores from administration of a single dose of L-DOPA (20mg/kg i.p.) to 180min later. This plot shows that the dyskinaesia is presenton the first dose of L-DOPA and duration of L-DOPA administration has no effect on severity of dyskinaesia. (C) Total AIM score plottedagainst % denervation [(1^ density/normal density) � 100] and fitted with a sigmoidal curve. (r=0.9). (D and E) Time course ofextra-synaptic DA release in the striatum, before (^40 to 0min) and after (20 to 140 min) administration of L-DOPA (20mg/kg) in animalswith extensive (filled triangle), partial (filled circle), normal density lesioned (filled square) and normal density unlesioned (unfilled triangle).(D) Animals which had received 14 days of L-DOPA, and (E) animals which had received L-DOPA for the first time.(F) Baselineconcentration of extra-synaptic striatal DA plotted against percentage of denervation (r=0.8).(G) Total [DA], following administrationof a single dose of L-DOPA, plotted against percentage of denervation with a sigmoidal curve. (r=0.9).(H) Total [DA], followingadministration of a single dose of L-DOPA, plotted against total AIMs.This relationship was confirmed by plotting total AIMs againsttotal [DA] to yield a linear line with an r-value of 0.9.

1580 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


density (Bmax2) of the lower-affinity site (P50.05). [3H]DAtransport measured in synaptosomes (Stanic et al., 2003b)from the dorsal striatum of rats with large lesions (460%) andsmall lesions (560%) was markedly reduced (Fig. 3D).

In large lesions, DA transport falls to levels found withmazindol treatment. Taken together, these studies suggestthat striatal DA re-uptake is impaired in the newly sproutedterminals following SNpc lesions.

Fig. 3 (A) An example of the measurements of DA concentration in the dorsal striatum, made before and after local application of325� 70 nl of 200mMDA in the vicinity of carbon-fibre recording electrodes. In normal animals (black), DA concentration rose rapidly toa peak and was also cleared promptly. Following a lesion (white), the time to peak DA concentration was significantly longer and clearancewas greatly prolonged. (B) Left-hand bars: in vivo recordings of T50 following injection of DA in the vicinity of the striatal recordingelectrode in normal (white) and lesioned (black) striatum. Right-hand bars: T50 after application of 0.5mMDA to a synaptosomepreparations from the dorsal striatum of normal (white) and lesioned (black) rats, 16 weeks following lesion. �P50.05 significantly differentfrom normal. (C) Left-hand bars: peak DA (� SEM) concentration following injection of DA into the dorsal striatum in the vicinity of therecording electrode. Middle bars: peak DA (� SE) concentration following local application of KCl (70mM, 200� 50 nl) to the dorsalstriatum. Forty-four measurements were made from 24 animals (14 normal, 10 lesioned). Right-hand bars: mean DA (� SEM) release fromstriatal synaptosomes of normal and lesioned animals evoked by 1M KCl. (D) [3H]DA transport into synaptosomes prepared from thedorsal striatum from normal rats (filled circle), rats with lesions of560% (filled square), rats with lesions of460% (filled triangle) and ratstreated with mazindol (open square). The rate of DA transport over the first 5min (R0^5) and the saturation concentration (S) wascalculated. Both R0^5 and S were reduced to almost half of normal animals when lesions were560%, and reduced to levels similar tomazindol treatment when lesions were460%. Reproduced with modifications and with permission fromWiley-Blackwell Publishing Ltd(Stanic et al., 2003b). (E) Parameters of DA release evoked in dorsal striatum by a train of electrical stimuli delivered to the MFB. DAP ispeak DA and T50 is the time taken for [DA] to fall to 50% of the peak. (F) T50 measured from the evoked response following MFBstimulation. Average value for normal animals is shown as (ominus), and from individual lesioned animals as (filled diamond). The line is alogarithmic regression [T50 = 4.28 + 4.29 ln(LS)] fitted to the data with an r= 0.6.




ownloaded from


The capacity for remaining nigral fibres to release DA wasassessed by measuring the DA released into the striatum of therat following MFB stimulation (Fig. 3E and F). Striatal DArelease was evoked by a train of electrical stimuli delivered tothe MFB and DA was detected in the dorsal striatum (Fig. 3E).There were 17 animals with lesions ranging from 59 to 82%and percentage of denervation ranging from 5 to 82%, as wellas 10 normal animals. T50 rose from a mean of 13.6� 3.4 s innormal animals to as long as 29.5 s in rats with large lesions.These data were fitted with a logarithmic regression (r= 0.6,Fig. 3F), showing that T50 rose abruptly initially, and thenplateaued. Examination of Fig. 3B shows T50 measuredin vivo, is approximately twice as long in lesioned animalsas in normal animals, whereas T50 measured in vitro is�30-fold greater in normal animals. In the case of in vitropreparations, re-uptake is the primary mechanism operatingwhereas in vivo, factors such as diffusion and metabolismoperate and we speculate that these dominate when re-uptakebecomes very slow and may explain why the curve plateaus.

Relationship between DA denervation,striatal DA release and dyskinaesia inmarmosetsPartial lesions ranging from 20 to 90%, were made in 14marmosets and compared with eight unlesioned animals.Histological estimates of lesion size and density of DAT-irterminals in the striatum were made 6 months after the lesion.Presumably reflecting the greater heterogeneity of geneticbackground of marmosets compared with in-bred rats, thenumber of TH-ir cells in the SNpc and of striatal DAT-irterminal density varies considerably. To overcome thisproblem, the lesioned side was expressed as a percentage ofthe unlesioned side (percentage of lesion size and percentageof DAT density) to allow reference parameters of DA releaseand re-uptake to be compared with striatal denervation inindividual animals. Accordingly, lesions ranged from 20%to 90% (Fig. 4A). Unlike rats, percentage of DAT density didnot return to normal but plateaued near 60–80% of normal,

even with lesions560% (Fig. 4A). With larger lesions, termi-nals were few and sparse and approached 100% loss. Six ofthese lesioned and six unlesioned marmosets were used in thestudies described below.

Immediately after lesioning, the right arm of all lesionedmonkeys was bradykinetic as measured by their perfor-mance of the staircase task (40.9� 24.7 s: mean� SEcompared with normal, 17.5� 4.6 s) but had returned tonormal by 20 weeks after lesioning (20.9� 6.7 s, P= 0.13,t-test), although individual animals with large lesionsremained significantly impaired. Prior to lesioning noneof the marmosets rotated spontaneously. One week post-lesion, most animals spontaneously turned towards thelesioned side but 20 weeks post-lesion this plateaued at1.8� 2.7 turns per hour even in animals with greatlydiminished percentage of DAT density. This suggests thatremaining terminals were capable of maintaining similarlevels of DA to the contralateral side. Rotation in responseto acute administration of L-DOPA was assessed at 1, 2, 6and 10 months post-lesion and produced marked rotationin the opposite direction to the lesion (data not shown):this was greatest in animals with the fewest terminals(largest lesions). This implies that DA terminals in thepartially innervated striatum can, in the presence ofL-DOPA, produce nigrostriatal dopaminergic signallingthat exceeds the normal side.

Relationship between DA denervation and striatal DArelease evoked by MFB stimulation in marmosetsThe effect of MFB stimulation on evoked DA release wascompared in control and lesioned marmosets and con-firmed that evoked DA was substantially greater in lesionedanimals than in normal marmosets (Fig. 4B–E). T50, peakDA (DAP) and area under curve (AUC) were measured,which were all increased in lesioned animals (Fig. 4B–E).

To analyse the effect of a single dose of L-DOPA, MFBstimulation was performed before administration and thenevery 30 min for 120 min (Fig. 4F–I). Individual parameters(�DAP, �AUC, T50) were analysed at each time point andfor each monkey. The data in Fig. 4F and H werenormalized with respect to pre L-DOPA treatment, whilstFig. 4G and I show absolute values. Following a single doseof L-DOPA, �DAp increased by 2-fold in normal animalsand remained elevated at 120 min following L-DOPAadministration (Fig. 4F). In lesioned animals, these valueswere already significantly elevated, and L-DOPA failed tocause further elevation in �DAp (Fig. 4F). Similarly, �AUCincreased following L-DOPA administration in normalanimals and remained elevated for another 90 min(Fig. 4H). In lesioned animals, �AUC were alreadysignificantly elevated and thus, L-DOPA failed to causefurther elevation (Fig. 4H). In the case of AUC and DAP,the concentration of extra-synaptic DA increased in normalanimals following a single dose of L-DOPA, whereas therewas little change in lesioned animals (Fig. 4G and I).

Table 1 Scatchard analysis of [3H]Mazindol bindingto DA transporter in the dorsal striatum of normaland SNpc lesioned animals 16 weeks after injury(mean� SD)

Dorsalstriatum

Kd1 (nM) Bmax1(fmol/mgprotein)

Kd2 (nM) Bmax2(fmol/mgprotein)

Normal 7.68� 2.2 1441�483 306�182 4284�1631(n=5) (n=5) (n=5) (n=4)

Contralateral 9.87� 2.7 1517�439 126� 62 4688� 2308(n=7) (n=8) (n=7) (n=7)

Lesioned 9.6�3.1 954� 281 8349� 6938 22739�11465(n=7) (n=7) (n=7) (n=6)

Reproduced with permission fromWiley-Blackwell Publishing Ltd(Stanic et al., 2003b).

1582 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


T50 was unaltered by L-DOPA administration in eitherlesioned or normal animals (data not shown).

Relationship between DA denervation and dyskinaesiain marmosetsDyskinetic movements were scored 30 min after adminis-tration of 20 mg/kg of L-DOPA and 3.75 mg/kg benserazidep.o. and were assessed after 1, 2, 6, 10, 11 and 12 months oflong-term L-DOPA treatment. Choreioform and dystonicmovements, affecting upper limbs, face, trunk, neck andorolingual regions were observed in the four lesionedmarmosets treated with L-DOPA (Fig. 4J) and these wereprominent in animals with large lesions. Although therewere only four animals, the extent of dyskinaesia appearedrelated to the extent of striatal denervation. The percentageof denervation of the two markedly dyskinetic animals were75 and 83% of normal (SNpc lesions of 77 and 80%,respectively), whereas percentage of denervation in thetwo animals with the less severe dyskinesia were 20 and24% (SNpc lesion size of 41 and 48%). As with rats,rotation rates induced by L-DOPA were increased inanimals with the greatest denervation and in the oppositedirection to the lesioned side.

Modelling of DA release in the striatumevoked by MFB stimulationThe methods and results illustrated in Figs 2–4 provideinformation about T50 and AUC but only indirect infor-mation about DA release. We therefore used the informa-tion from Figs 2F and 3F and a simple linear differentialequation _C ¼ �� kC to model the release of DA evokedby MFB stimulation. We have assumed that DA is releasedat a constant rate, represented by the constant �, over thefirst 10 s after stimulus. The re-uptake of DA was assumedto be directly proportional to the concentration and that re-uptake, diffusion and metabolic breakdown can berepresented by a constant k. T50 values from the logarithmicregression and the polynomial used to fit the data in Figs2F and 3F were used to produce a family of 20 curves(Fig. 5A shows an example of one of the curves)corresponding to percentage of denervation from 1% to100%. These were used to obtain the corresponding DAP

Fig. 4 (A) Percentage SNpc lesion size (expressed as a percentageof SNpc neurons in comparison to the unlesioned side) plottedagainst the density of DA terminals in the striatum (expressedas a percentage of the unlesioned striatum in lesioned marmosets).(open square) lesioned animals that received long-term L-DOPAtreatment; (open diamond) lesioned untreated animals; (filledtriangle) untreated lesioned marmosets not used subsequentlyin the experiment.(B) The mean DA overflow following stimula-tion of the MFB (n=6, black line=normal; dotted line= lesionedmarmosets).(C^E) Mean DAP (C), T50 (D) and AUC (E) wereplotted from normal (white) and lesioned (black) animals(�P5.05 significantly different from normal, t-test). (F^I)Serial measurements of DAP (F) and AUC (H) were madein normal (filled diamond) and lesioned (open cirlce) and animalsfollowing a bolus injection of L-DOPA. MFB stimulation wasperformed at each of the time points and the relevantparameter was plotted. For each monkey (n=6), the valuesobtained after L-DOPAwere normalized with respect to thepre-L-DOPA value. Post-L-DOPA values remained relativelyconstant in each animal, so these post-L-DOPA values werepooled and compared with the pre-L-DOPA values (G) and (H),(N=normal animals; Les= lesioned animals; mean� SEM).

Note that both DAP and AUC increase in unlesioned animalfollowing L-DOPAwhereas in lesioned animals, DAP increasesonly modestly and not significantly (�P50.05 significantly differentfrom normal, t-test). (J) Dyskinaesia (open circle) was assessedin the four lesioned marmosets treated with long-term L-DOPAand this was plotted against striatal denervation. As with rats,dyskinaesia is greatest in animals with least innervation. Alsoshown are net rotations (R-L) for the same four animals. Rotation30min before (white bars) and 30^60min after (black bars) asingle dose of L-DOPA is shown. Note that the animals with thegreatest denervation rotated the most after a dose of L-DOPAconsistent with an increased capacity to deliver DA into thestriatum.




ownloaded from


shown in Fig. 5B. The conclusion from this modelling isthat there must be a sharp and dramatic increase in DAreleased from the remaining release sites to account for theincreased DA release. While changes in T50 may beimportant they are not sufficient alone to produce themarked increase when denervation is significant. Weestimate that the amount of DA release by each remaining

terminal when denervation is 95% is about 40-fold greaterthan normal.

Fos B expression and SERT immunoreactiveterminals in the striatum following L-DOPAtreatmentWhen lesions of the SNpc are made, sprouting of DAterminals is confined to the dorsal striatum. If sproutedterminals are responsible for abnormal DA release andactivation of post-synaptic medium spiny neurons (MSNs)to produce dyskinaesia, then MSN activation should beconstrained to the dorsal striatum. To address this, animalswere killed 60 min after the administration of L-DOPA andprocessed for Fos B immunohistochemistry. As previouslyreported (Cenci et al., 1999), there was a markedupregulation of Fos B immunoreactive neurons in thedorsolateral striatum of lesioned animals treated withL-DOPA (Fig. 6B), that was not apparent in sham animalstreated with L-DOPA (Fig. 6A).

One explanation for the marked increase in DA with largelesions is unregulated activity of serotoninergic axons thathave sprouted new terminals in the MFB. However there wasno change in the density of SERT-immunoreactive terminalsin animals with large lesions compared to sham animals(Fig. 6C and D), making this an unlikely explanation.

DiscussionDyskinaesias are produced by both pre and post-synapticdisturbances (Cenci and Lundblad, 2006). Excess striatalDA following a dose of L-DOPA appears to be an important

Fig. 5 (A) The linear differential equation ð _C ¼ ð�� kCÞÞ wasused to model the release of DA evoked by MFB stimulation.We have assumed that dopamine is released at a constant rate,represented by the constant �, over the first 10 s after stimulus.k represented the re-uptake, diffusion and metabolic breakdownof DA and was assumed to be directly proportional to theconcentration of [DA]. T50 values from the logarithmic regressionand the polynomial used to fit the data in Figs 2F and 3F wereused to produce a family of 20 curves, one of which is shown in5A, corresponding to percentage of denervation from1^100%.These were used to obtain the corresponding DAP shown inFig. 5B. (B) The corresponding DAP obtained from curves in (A)were plotted against percentage of dernervation. This predicts asharp and dramatic increase in DA released from the remainingrelease sites when terminal density is560%, to account for theincreased DA release.

Fig. 6 (A and B) Photomicrographs of FosB immunoreactive nuclei in the dorsolateral striatum in (A) sham and (B) 6 -OHDA-lesionedrats treated with L-DOPA (scale bar=250mm). (C and D) Photomicrographs of SERT immunoreactive nuclei in the striatum in (C) shamand (D) 6 -OHDA-lesioned rats treated with L-DOPA (scale bar=20mm).

1584 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


component of the pre-synaptic mechanism responsible forgenesis of dyskinaesia (Meissner et al., 2006). The aims ofthis study were to examine in detail the relationshipbetween striatal DA denervation, striatal DA concentrationsand dyskinaesia.

An important consideration in this study is theexamination of DAT-ir terminal density and the use of6-OHDA lesioned animals that have been allowed toestablish stable re-innervation of the striatum. In keepingwith earlier reports (Finkelstein et al., 2000; Stanic et al.,2003a), sprouting of surviving neurons prevents striataldenervation occurring until about 60% of neurons are lost.The process of compensation through sprouting is likely tobe present in a slowly progressive condition such asParkinson’s disease. When sprouting can no longermaintain terminal density, striatal [DA] begins to fall andis likely to be analogous to when humans with Parkinson’sdisease become symptomatic.

While L-DOPA replacement does not significantly raisestriatal [DA] levels in rats with striatal denervation of560%, striatal [DA] levels increased by 3-fold when thedenervation was extensive (i.e. 460%). The point on thegraph (Fig.2G) where striatal [DA] begins to markedlyincrease corresponds to the point where dyskinaesia scoresincrease (Fig. 2C). Indeed, the extent of dyskinaesia waslinearly related to the striatal [DA] following L-DOPAadministration, suggesting that expression of dyskinaesiawas associated with high striatal [DA]. This reflects thefindings of PET studies, demonstrating that pre-synapticDA release in Parkinson’s disease patients correlates withhigher dyskineasia scores (Pavese et al., 2006).

We previously showed that DA terminals newly formedby sprouting have altered structure and function, includingimpaired DAT function, increased terminal bouton size,increased number of vesicles, mitochondria and contactsonto more proximal targets (Stanic et al., 2003b). The neteffect is that re-innervating terminals could deliver a largeractivity-dependent bolus of DA, which in the presence ofimpaired re-uptake could diffuse further, affecting DAreceptor sites over a wider area of the striatum. As aconsequence of large terminal arbours, impaired synapticclearance of DA and altered release capacity, there will beabnormal DA delivery to unusually large regions of thestriatum. The increased capacity for DA release andimpaired DAT function will seriously compromise thecapacity to store DA and contribute to the shorteningduration of response to L-DOPA observed in Parkinson’sdisease. The benefit of L-DOPA would progressively shortenas DAT function fails, reflected as ‘wearing off’ due toimpaired storage and dyskinaesia by higher DA levels. Thisis keeping with observations that pre-synaptic mechanismsare important in both short-duration and long-durationresponses encountered in Parkinson’s disease and dyski-naesias (de la Fuente-Fernandez et al., 2004). These findingsalso support the clinical observations that greater DATlevels are directly associated with lower DA turnover and

lower changes in synaptic DA concentration in Parkinson’sdisease patients (Sossi et al., 2007). Thus, decreased levelsand/or impairment of DAT, although potentially serving asa compensatory mechanism in early disease, may ultimatelyresult in increased DA turnover and higher oscillations insynaptic DA concentration.

Stimulation of the MFB pathway could be considered toreflect activity-dependent release of DA. In the monkey,and most likely in the rat, AUC and DAP were alreadyincreased in lesioned animals and L-DOPA treatment didnot increase it further. In unlesioned animals, bothparameters approximately doubled and approached the(presumably maximum) output of lesioned animals. Thisimplies that unlesioned animals were capable of storing theDA synthesized from L-DOPA treatment whereas lesionedanimals delivered it directly into the striatum.

This study has not fully resolved the question of theorigin of the very high levels of DA in animals with fewremaining terminals. While we have suggested that impair-ed transporter function and altered storage may beimportant factors, other contributions may also be impor-tant. Indeed the presence of DA in animals with effectivelyno DA terminals, and unstimulated with L-DOPA, arguesfor a persisting site of DA synthesis and release. Convincingevidence of the role of serotonergic fibres capacity toprovide DA as a false transmitter has been provided (Cartaet al., 2006, 2007). While this mechanism would convertL-DOPA to DA, the persisting levels of DA suggest thatanother source is active. The lack of increase in the densityof SERT-positive fibres in lesioned animals suggest that anyincreased capacity of serotonergic fibres to convert L-DOPAto DA is by upregulating DA synthesizing enzymes, ratherthan increasing release sites.

Another explanation for increased striatal [DA] anddyskinaesia is the possible synthesis of DA from exogenousL-DOPA in non-neuronal aromatic l-amino acid decarbox-ylase (AADC) containing cells (Melamed et al., 1980;Betarbet et al., 1997; Brown et al., 1999; Porritt et al.,2000a) found in rats (Betarbet et al., 1997), monkeys(Ikemoto et al., 1997) and humans (Porritt et al., 2000a).It is argued that striatal AADC immunoreactive neurons areinduced by DA denervation (Meredith et al., 1999), butAADC positive neurons are present in the normal brain anddenervation does not change their numbers (Lopez-Realet al., 2003). Their number is increased in Parkinson’sdisease (Porritt et al., 2000b), but curiously their distribu-tion is in areas where there is continuing nigral innervation(Porritt et al., 2000a; Lopez-Real et al., 2003). While Porrittand colleagues (2000a) argue that their numbers aresufficient to contribute to L-DOPA conversion in the DAdenervated striatum, the TH and AADC-positive cells areprobably different populations (Ikemoto et al., 1997;Meredith et al., 1999) and the level of enzyme activitydoes not seem to be sufficient to convert L-DOPA into DA(Nakamura et al., 2000). Perhaps most importantly, theevidence from this study shows that release of DA was from




ownloaded from


terminals whose axons are activated by MFB stimulation.Clearly, further work is required to clarify the nature andsource of DA in the severely denervated striatum. Whileserotonergic fibres may well be the important source whendenervation is complete, it may be possible that severalmechanisms may be contributing in the partially denervatedanimal, which more truly reflects the state in theparkinsonian patient.

It has been argued that L-DOPA treatment is necessary toinduce dyskinaesia (Meissner et al., 2006) and certainlydyskinaesia does not occur without L-DOPA administration.However, we found the severity of dyskinaesia to be similar inanimals previously exposed to L-DOPA to those exposed to L-DOPA de novo. This argues that the changes in the terminalsper se rather than some predisposing action of L-DOPA is thecentral factor for the genesis of dyskinaesia.

The actual cellular events underlying dyskinaesia are notclear although there has been speculation that they representabnormal persistence of changed weighting of cortico-striatalglutamatergic synapses (Picconi et al., 2002, 2003; Cenci andLundblad, 2006; Cenci, 2007). Plasticity at the corticalprojection onto spiny neurons was altered by selective DAreceptor blockade and following dopamine denervation butrestored by L-DOPA therapy (Picconi et al., 2002, 2003). It isconceivable that increased concentration of DA at post-synaptic receptors could readily enhance (Pickel et al., 1992)corticostriatal glutamatergic transmission (Reynolds et al.,2001). The timing of coincident synaptic influence at aHebbian synapse is critical for the persistence of LTP (Danand Poo, 2006) and it is possible that one of the consequencesof the altered re-uptake is that DA effects have a much longerduration and their co-incidence with glutamatergic synapsesmay be altered both by this and their more proximaltermination on dendrites (Letzkus et al., 2006).

In summary, we propose that in Parkinson’s disease,surviving terminals either sprout or adopt many of thecharacteristics of sprouted terminals, in particular, impairedDA transport and impaired DA storage. This, in turn,contributes to the emergence of dyskinaesia and ‘wearing off’.

AcknowledgementsThese studies have been supported by funds from theNational Health and Medical Research Foundation(NHMRC), NHMRC C.J. Martin fellowship ID 300083 toD.S., and the Bethlehem Griffiths Research Foundation.

ReferencesBetarbet R, Turner R, Chockkan V, DeLong MR, Allers KA, Walters J,

et al. Dopaminergic neurons intrinsic to the primate striatum.

J Neurosci 1997; 17: 6761–8.

Bezard E, Brotchie JM, Gross CE. Pathophysiology of levodopa-induced

dyskinesia: potential for new therapies. Nat Rev Neurosci 2001; 2:

577–88.

Bowyer JF, Weiner N. Modulation of the Ca++-evoked release of

[3H]dopamine from striatal synaptosomes by dopamine (D2) agonists

and antagonists. J Pharmacol Exp Ther 1987; 241: 27–33.

Brown WD, Taylor MD, Roberts AD, Oakes TR, Schueller MJ, Holden JE,

et al. FluoroDOPA PET shows the nondopaminergic as well as

dopaminergic destinations of levodopa. Neurology 1999; 53: 1212–8.

Carlsson T, Carta M, Winkler C, Bjorklund A, Kirik D. Serotonin neuron

transplants exacerbate L-DOPA-induced dyskinesias in a rat model of

Parkinson’s disease. J Neurosci 2007; 27: 8011–22.

Carta M, Carlsson T, Kirik D, Bjorklund A. Dopamine released from 5-HT

terminals is the cause of L-DOPA-induced dyskinesia in parkinsonian

rats. Brain 2007; 130: 1819–33.

Carta M, Lindgren HS, Lundblad M, Stancampiano R, Fadda F, Cenci MA.

Role of striatal L-DOPA in the production of dyskinesia in

6-hydroxydopamine lesioned rats. J Neurochem 2006; 96: 1718–27.

Cenci MA. Dopamine dysregulation of movement control in L-DOPA-

induced dyskinesia. Trends Neurosci 2007; 30: 236–43.

Cenci MA, Lee CS, Bjorklund A. L-DOPA-induced dyskinesia in the rat is

associated with striatal overexpression of prodynorphin- and glutamic

acid decarboxylase mRNA. Eur J Neurosci 1998; 10: 2694–706.

Cenci MA, Lundblad M. Post- versus presynaptic plasticity in L-DOPA-

induced dyskinesia. J Neurochem 2006; 99: 381–92.

Cenci MA, Tranberg A, Andersson M, Hilbertson A. Changes in the

regional and compartmental distribution of FosB- and JunB-like

immunoreactivity induced in the dopamine-denervated rat striatum by

acute or chronic L-dopa treatment. Neuroscience 1999; 94: 515–27.

Cotzias GC, Van Woert MH, Schiffer LM. Aromatic amino acids and

modification of parkinsonism. N Engl J Med 1967; 276: 374–9.

Dan Y, Poo MM. Spike timing-dependent plasticity: from synapse to

perception. Physiol Rev 2006; 86: 1033–48.

de la Fuente-Fernandez R, Schulzer M, Mak E, Calne DB, Stoessl AJ.

Presynaptic mechanisms of motor fluctuations in Parkinson’s disease:

a probabilistic model. Brain 2004; 127: 888–99.

Finkelstein DI, Stanic D, Parish CL, Drago J, Horne MK. Stereological pro-

cedures for determination of neuronal arbour size. In: Gerfen C, Wray S,

editors. Current protocols in neuroscience. New York: Wiley; 2004

p. 1.13.1–1.13.16.

Finkelstein DI, Stanic D, Parish CL, Tomas D, Dickson K, Horne MK.

Axonal sprouting following lesions of the rat substantia nigra.

Neuroscience 2000; 97: 99–112.

Grace AA, Bunney BS. The control of firing pattern in nigral dopamine

neurons: burst firing. J Neurosci 1984a; 4: 2877–90.

Grace AA, Bunney BS. The control of firing pattern in nigral dopamine

neurons: single spike firing. J Neurosci 1984b; 4: 2866–76.

Hely MA, Morris JG, Reid WG, O’Sullivan DJ, Williamson PM, Rail D, et al.

The Sydney Multicentre Study of Parkinson’s disease: a randomised,

prospective five year study comparing low dose bromocriptine with

low dose levodopa-carbidopa. J Neurol Neurosurg Psychiatry 1994; 57:

903–10.

Henderson JM, Annett LE, Torres EM, Dunnett SB. Behavioural effects of

subthalamic nucleus lesions in the hemiparkinsonian marmoset

(Callithrix jacchus). Eur J Neurosci 1998; 10: 689–98.

Hersch SM, Yi H, Heilman CJ, Edwards RH, Levey AI. Subcellular

localization and molecular topology of the dopamine transporter in the

striatum and substantia nigra. J Comp Neurol 1997; 388: 211–27.

Ikemoto K, Kitahama K, Jouvet A, Arai R, Nishimura A, Nishi K, et al.

Demonstration of L-dopa decarboxylating neurons specific to human

striatum. Neurosci Lett 1997; 232: 111–4.

Lawrence AJ, Parish CL, Chen F, Lodge DJ, Krstew EV, Card K, et al.

Chronic corticotropin-releasing factor type 1 receptor antagonism with

antalarmin regulates the dopaminergic system of Fawn-Hooded rats.

J Neurochem 2005; 94: 1523–34.

Letzkus JJ, Kampa BM, Stuart GJ. Learning rules for spike timing-

dependent plasticity depend on dendritic synapse location. J Neurosci

2006; 26: 10420–9.

Lindgren N, Xu ZQ, Herrera-Marschitz M, Haycock J, Hokfelt T, Fisone G.

Dopamine D(2) receptors regulate tyrosine hydroxylase activity

and phosphorylation at Ser40 in rat striatum. Eur J Neurosci 2001; 13:

773–80.

1586 Brain (2008), 131, 1574^1587 J. Lee et al.



ownloaded from


Lopez-Real A, Rodriguez-Pallares J, Guerra MJ, Labandeira-Garcia JL.

Localization and functional significance of striatal neurons immunor-

eactive to aromatic L-amino acid decarboxylase or tyrosine hydroxylase

in rat Parkinsonian models. Brain Res 2003; 969: 135–46.

McColl CD, Reardon KA, Shiff M, Kempster PA. Motor response to

levodopa and the evolution of motor fluctuations in the first decade of

treatment of Parkinson’s disease. Mov Disord 2002; 17: 1227–34.

Meissner W, Ravenscroft P, Reese R, Harnack D, Morgenstern R,

Kupsch A, et al. Increased slow oscillatory activity in substantia nigra

pars reticulata triggers abnormal involuntary movements in the

6-OHDA-lesioned rat in the presence of excessive extracellular striatal

dopamine. Neurobiol Dis 2006; 22: 586–98.

Melamed E, Hefti F, Wurtman RJ. Nonaminergic striatal neurons convert

exogenous L-dopa to dopamine in parkinsonism. Ann Neurol 1980; 8:

558–63.

Meredith GE, Farrell T, Kellaghan P, Tan Y, Zahm DS, Totterdell S.

Immunocytochemical characterization of catecholaminergic neurons in

the rat striatum following dopamine-depleting lesions. Eur J Neurosci

1999; 11: 3585–96.

Metman LV, van den Munckhof P, Klaassen AA, Blanchet P, Mouradian MM,

Chase TN. Effects of supra-threshold levodopa doses on dyskinesias in

advanced Parkinson’s disease. Neurology 1997; 49: 711–3.

Mirenowicz J, Schultz W. Importance of unpredictability for reward

responses in primate dopamine neurons. J Neurophysiol 1994; 72:

1024–7.

Mouradian MM, Juncos JL, Fabbrini G, Schlegel J, Bartko JJ, Chase TN.

Motor fluctuations in Parkinson’s disease: central pathophysiological

mechanisms, Part II. Ann Neurol 1988; 24: 372–8.

Nakamura K, Ahmed M, Barr E, Leiden JM, Kang UJ. The localization and

functional contribution of striatal aromatic L-amino acid decarboxylase

to L-3,4-dihydroxyphenylalanine decarboxylation in rodent parkinso-

nian models. Cell Transplant 2000; 9: 567–76.

Nutt JG. Motor fluctuations and dyskinesia in Parkinson’s disease.

Parkinsonism Relat Disord 2001; 8: 101–8.

Nutt JG, Carter JH, Lea ES, Sexton GJ. Evolution of the response to levodopa

during the first 4 years of therapy. Ann Neurol 2002; 51: 686–93.

Nutt JG, Carter JH, Woodward WR. Long-duration response to levodopa.

Neurology 1995; 45: 1613–6.

Nutt JG, Woodward WR, Carter JH, Gancher ST. Effect of long-term

therapy on the pharmacodynamics of levodopa. Relation to on-off

phenomenon. Arch Neurol 1992; 49: 1123–30.

Onali P, Olianas MC, Bunse B. Evidence that adenosine A2 and dopamine

autoreceptors antagonistically regulate tyrosine hydroxylase activity in

rat striatal synaptosomes. Brain Res 1988; 456: 302–9.

Parish CL, Finkelstein DI, Drago J, Borrelli E, Horne MK. The role of

dopamine receptors in regulating the size of axonal arbors. J Neurosci

2001; 21: 5147–57.

Parish CL, Stanic D, Drago J, Borrelli E, Finkelstein DI, Horne MK. Effects

of long-term treatment with dopamine receptor agonists and antagonists

on terminal arbor size. Eur J Neurosci 2002; 16: 787–94.

Pavese N, Evans AH, Tai YF, Hotton G, Brooks DJ, Lees AJ, et al. Clinical

correlates of levodopa-induced dopamine release in Parkinson disease: a

PET study. Neurology 2006; 67: 1612–7.

Paxinos G, Watson C. The rat brain in stereotaxic co-ordinates. Sydney:

Academic Press; 1998. p. C10–33; C37–43.

Pearce RK, Jackson M, Smith L, Jenner P, Marsden CD. Chronic L-DOPA

administration induces dyskinesias in the 1-methyl-4-phenyl-1,2,3,6-

tetrahydropyridine-treated common marmoset (Callithrix Jacchus).

Mov Disord 1995; 10: 731–40.

Picconi B, Centonze D, Hakansson K, Bernardi G, Greengard P, Fisone G,

et al. Loss of bidirectional striatal synaptic plasticity in L-DOPA-induced

dyskinesia. Nat Neurosci 2003; 6: 501–6.

Picconi B, Pisani A, Centonze D, Battaglia G, Storto M, Nicoletti F, et al.

Striatal metabotropic glutamate receptor function following experi-

mental parkinsonism and chronic levodopa treatment. Brain 2002; 125:

2635–45.

Pickel VM, Johnson E, Carson M, Chan J. Ultrastructure of spared

dopamine terminals in caudate-putamen nuclei of adult rats neonatally

treated with intranigral 6-hydroxydopamine. Brain Res Dev Brain Res

1992; 70: 75–86.

Poewe WH, Lees AJ, Stern GM. Low-dose L-dopa therapy in Parkinson’s

disease: a 6-year follow-up study. Neurology 1986; 36: 1528–30.

Porritt MJ, Batchelor PE, Hughes AJ, Kalnins R, Donnan GA,

Howells DW. New dopaminergic neurons in Parkinson’s disease

striatum. Lancet 2000a; 356: 44–5.

Porritt MJ, Batchelor PE, Hughes AJ, Kalnins R, Donnan GA,

Howells DW. New dopaminergic neurons in Parkinson’s disease

striatum. Lancet 2000b; 356: 44–5.

Reardon KA, Shiff M, Kempster PA. Evolution of motor fluctuations in

Parkinson’s disease: a longitudinal study over 6 years. Mov Disord 1999;

14: 605–11.

Reynolds JN, Hyland BI, Wickens JR. A cellular mechanism of reward-

related learning. Nature 2001; 413: 67–70.

Robinson MB. Regulated trafficking of neurotransmitter transporters:

common notes but different melodies. J Neurochem 2002; 80: 1–11.

Sossi V, de la Fuente-Fernandez R, Schulzer M, Troiano AR, Ruth TJ,

Stoessl AJ. Dopamine transporter relation to dopamine turnover in

Parkinson’s disease: a positron emission tomography study. Ann Neurol

2007; 62: 468–74.

Stanic D, Finkelstein DI, Bourke DW, Drago J, Horne MK. Timecourse of

striatal re-innervation following lesions of dopaminergic SNpc neurons

of the rat. Eur J Neurosci 2003a; 18: 1175–88.

Stanic D, Parish CL, Zhu WM, Krstew EV, Lawrence AJ, Drago J, et al.

Changes in function and ultrastructure of striatal dopaminergic

terminals that regenerate following partial lesions of the SNpc.

J Neurochem 2003b; 86: 329–43.

Ungerstedt U, Herrera-Marschitz M, Stahle L, Zetterstrom T. Models for

studying the synaptic mechanisms: correlative measurements of

transmitter release and drug altered behaviour. In: Spiegelstein MY,

Levy A, editors. Behavioural models and the analysis of drug action.

Amsterdam: Elsevier; 1982. p. 57–70.

West MJ, Slomianka L, Gundersen HJ. Unbiased stereological estimation of

the total number of neurons in the subdivisions of the rat hippocampus

using the optical fractionator. In: Spielgestein MY, Levy A, editors. Anat

Rec 1991; 231: 482–97.

Winkler C, Kirik D, Bjorklund A, Cenci MA. L-DOPA-induced dyskinesia

in the intrastriatal 6-hydroxydopamine model of parkinson’s disease:

relation to motor and cellular parameters of nigrostriatal function.

Neurobiol Dis 2002; 10: 165–86.




ownloaded from


Sprouting of dopamine terminals and altered dopamine release and uptake in Parkinsonian dyskinaesia

Documents

Transcript of Sprouting of dopamine terminals and altered dopamine release and uptake in Parkinsonian dyskinaesia