R

Ter

P

a

b

c

d

h

���

a

ARR1AA

KPAARM

1

c

U

0h

Behavioural Brain Research 243 (2013) 6– 15

Contents lists available at SciVerse ScienceDirect

Behavioural Brain Research

j ourna l ho me pa ge: www.elsev ier .com/ locate /bbr

esearch report

he behavioural and neuropathological impact of intranigral AAV-�-synuclein isxacerbated by systemic infusion of the Parkinson’s disease-associated pesticide,otenone, in rats

ádraig Mulcahya,b, Aideen O’Dohertyb, Alexia Paucarda, Timothy O’Brienc, Deniz Kirikd, Eilís Dowda,b,∗

Department of Pharmacology & Therapeutics, National University of Ireland, Galway, IrelandNational Centre for Biomedical Engineering Science, National University of Ireland, Galway, IrelandRegenerative Medicine Institute, National University of Ireland, Galway, IrelandDepartment of Experimental Medical Science, Lund University, Sweden

i g h l i g h t s

We model Parkinson’s disease using genetic and environmental risk factors in rats.Exposure to alpha-synuclein and rotenone models the main features of the disease.The model may shed light on genes by environment interactions in Parkinson’s disease.

r t i c l e i n f o

rticle history:eceived 18 September 2012eceived in revised form3 December 2012ccepted 27 December 2012vailable online 4 January 2013

eywords:arkinson’s diseasenimal modelsAV-�-synucleinotenoneotor function

a b s t r a c t

Despite the widely held belief that Parkinson’s disease is caused by both underlying genetics and exposureto environmental risk factors, it is still widely modelled in preclinical models using a single geneticor neurotoxic insult. This single-insult approach has resulted in a variety of models that are limitedwith respect to their aetiological, construct, face and/or predictive validity. Thus, the aim of the currentstudy was to investigate the interplay between genes and the environment as an alternative approachto modelling Parkinson’s disease. To do so, rats underwent stereotaxic surgery for unilateral deliveryof the Parkinson’s disease-associated gene, ˛-synuclein, into the substantia nigra (using AAV vectors).This was followed 13 weeks later by subcutaneous implantation of an osmotic minipump deliveringthe Parkinson’s disease-associated pesticide, rotenone (2.5 mg kg−1 day−1 for 4 weeks). The effect of thegenetic and environmental insults alone or in combination on lateralised motor performance (Corridor,Stepping and Whisker Tests), nigrostriatal integrity (tyrosine hydroxylase immunohistochemistry) and�-synucleinopathy (�-synuclein immunohistochemistry) was assessed. We found that exposing AAV-�-synuclein-treated rats to rotenone led to a model in which the classical Parkinson’s disease triad of

progressive motor dysfunction, nigrostriatal neurodegeneration and �-synucleinopathy was evident.However, delivering rotenone systemically was also associated with bilateral motor dysfunction and lossof body weight. Thus, although we have shown that Parkinson’s disease can be modelled in experimentalanimals by combined exposure to both genetic and environmental risk factors, this approach is limitedby systemic toxicity of the pesticide rotenone. Direct intracerebral delivery of rotenone may be moreuseful in longer-term studies as we have previously shown that it overcomes this limitation.

. Introduction

Parkinson’s disease is a chronic neurodegenerative disorderharacterised neuropathologically by intracytoplasmic formation

∗ Corresponding author at: Department of Pharmacology & Therapeutics, Nationalniversity of Ireland, Galway, Ireland. Tel.: +353 91 492776; fax: +353 91 525700.

E-mail address: Eilis.Dowd@nuigalway.ie (E. Dowd).

166-4328/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.bbr.2012.12.051

© 2013 Elsevier B.V. All rights reserved.

of Lewy bodies and progressive loss of nigrostriatal dopaminergicneurons, and symptomatically by a bradykinetic motor syndrome[1,2]. Although the aetiology of this condition is poorly understood,it is thought to arise as a result of complex interactions betweenunderlying genetics and environmental factors [3,4]. Indeed, it isprobable that Parkinson’s disease is the result of the net interac-

tions of multiple risk factors encountered over the lifetime, whichin addition to those promoting risk, such as exposure to pesticidesand other environmental toxins [5], geographical location [6,7],

ral Brain Research 243 (2013) 6– 15 7

hsa

amrhvuatacd[luceefotptPma

aeaaebsrtaa

2

2

CoRi

2

2ogfWpmTtra

2

h

Table 1Final groups used in this study. Rats were given a unilateral intranigral infusionof either AAV-GFP or AAV-�-synuclein. Thirteen weeks later, this was followed bysubcutaneous implantation of an osmotic minipump for delivery of either vehicle(DMSO:PEG (1:1)) or rotenone (2.5 mg kg−1 day−1) to yield 4 final groups as shown.

Group Virus injected Minipump implanted n

Control AAV-GFP Vehicle 10�-Synuclein AAV-�-synuclein Vehicle 7

P. Mulcahy et al. / Behaviou

ead trauma [8], age, race and gender [9] and genetic predispo-ition [10], also includes factors that have shown to be protectivegainst the disease, such as caffeine and cigarette smoking [11,12].

Despite the understanding that Parkinson’s disease arises fromn interaction between genes and the environment, it is widelyodelled in experimental animals using a single genetic or neu-

otoxic insult [13,14]. Unfortunately, this single-insult approachas limited these models on a number of important measures ofalidity. For example, because the most widely used neurotoxinssed to model Parkinson’s disease (that is, 6-hydroxydopaminend MPTP) are rarely associated with the human condition, usinghese neurotoxins to model the condition in animals limits theiretiological validity [15]. Most models also fail to recapitulate thelassic neuropathological and symptomatic features of the humanisease, and thus are limited in both their construct and face validity16]. Moreover, the poor predictive validity of the well estab-ished models is one of the reasons that has been suggested tonderlie the high failure rate of experimental anti-Parkinsonianompounds in clinical trials [17–21]. Therefore, the limitations ofxisting models have led to a drive to develop more relevant mod-ls of Parkinson’s disease with improved aetiological, construct,ace and predictive validity. In this regard, we have recently devel-ped a novel dual-hit approach to modelling the disease in whichhe disease-associated ˛-synuclein gene and the disease-associatedesticide, rotenone, are sequentially administered into the rat subs-antia nigra [22]. This approach has generated a relevant and robustarkinson’s disease model in which the classic triad of progressiveotor decline, �-synucleinopathy and nigrostriatal degeneration

re expressed.In this study, we further investigate the interplay between genes

nd the environment as an approach to modelling Parkinson’s dis-ase. Specifically, we investigated the impact of exposing rats with

high nigrostriatal load of �-synuclein to the Parkinson’s disease-ssociated pesticide, rotenone. In our previous studies (Mulcahyt al., 2011; [22]) the pesticide was administered directly into therain, whereas in the present study, it was delivered by chronicubcutaneous infusion via an osmotic minipump to more accu-ately mimic systemic rotenone exposure in humans. The aim ofhe study was to determine if combining the established geneticnd environmental risk factors would exacerbate the behaviouralnd neuropathological features induced by either factor alone.

. Materials and methods

.1. Animals

All procedures were carried out in accordance with the European Communitiesouncil Directive (86/609/EEC), and were approved by the Animal Ethics Committeef the National University of Ireland, Galway. Male Sprague–Dawley rats (Charlesivers, UK) were used in all experiments. All behavioural testing and quantitative

mmunohistochemistry was conducted blind to the treatment of the rats.

.2. Experimental design

Thirty-four male Sprague–Dawley rats were used, weighing 279 ± 10 g (range:64–302 g) on arrival. Baseline performance in the Corridor Test was establishedver 1 week pre-operatively. Rats were then divided into 2 performance matchedroups for intranigral infusion of AAV-GFP or AAV-�-synuclein. Post-viral motorunction was then assessed at weeks 4, 8 and 12 in the Corridor, Stepping and

hisker Tests. Thirteen weeks after viral infusion, rats were further sub-divided intoerformance-matched groups and were subcutaneously implanted with osmoticinipumps to deliver rotenone (2.5 mg kg−1 day−1) or vehicle (DMSO:PEG (1:1)).

his yielded 4 final groups which are detailed in Table 1. Post-rotenone behaviouralesting was carried out over a further 4 week period after which the rats were sac-ificed by transcardial fixation and processed for quantitative tyrosine hydroxylasend �-synuclein immunohistochemistry.

.3. AAV preparation

AAV2 plasmids carrying normal human �-synuclein or GFP, and theelper plasmid, pDG-5, were purified using calcium chloride density gradient

Rotenone AAV-GFP Rotenone 9Combined AAV-�-synuclein Rotenone 8

ultracentrifugation. The AAV-�-synuclein and AAV-GFP viral vectors were producedby co-transfecting HEK-293 T cells with the relevant AAV2 plasmid and pDG-5 asdescribed previously [23]. Vectors were then purified by a single step column purifi-cation protocol using Q-sepharose columns, and titres were established using realtime PCR. The final viral titres were 2.93 × 1010 drp �l−1 for AAV-�-synuclein and1.07 × 1010 drp �l−1 for AAV-GFP, and both of the transgenes were driven by thehuman synapsin promoter.

2.4. Surgery

Unilateral intranigral AAV infusion was conducted under isofluorane anaesthe-sia (5% in O2 for induction and 2% in O2 for maintenance) in a stereotaxic framewith the nose bar set at −2.3 mm. The substantia nigra was infused (1 �l min−1 for2 min with 2 min for diffusion) with either AAV-�-synuclein or AAV-GFP at stereo-taxic coordinates AP −5.3, ML +2.0; (from bregma) and DV −7.2 below dura. Theosmotic minipumps (Alzet® , model 2ML4) for delivery of rotenone or vehicle werealso subcutaneously implanted (through a small incision on the back of the rat)under isofluorane anaesthesia (5% in O2 for induction and 2% in O2 for maintenance)13 weeks after viral infusion.

2.5. Behavioural testing

The Corridor Test of contralateral neglect has been described in more detailelsewhere [24,25]. In brief, hungry rats (deprived of food overnight) were placedinto a long, narrow alleyway (150 cm) and were free to retrieve food reward (crispedrice cereal (CocoPops®)) from pots on the left and right sides. Trials were completedwhen rats made a total of 20 retrievals or after a maximum trial time of 5 min hadelapsed. The Stepping Test of forelimb akinesia has been described in more detailelsewhere [26]. In brief, both hindlimbs and one forelimb were gently restrained bythe experimenter, and the number of forehand and backhand adjusting steps madeby the unrestrained forelimb was counted when the rat was moved sideways alonga table surface for 90 cm over 5 s. The Whisker Test of sensorimotor integrationhas been described in more detail elsewhere [27]. In brief, both hindlimbs and oneforelimb were gently restrained by the experimenter, and the number of vibrissae-evoked forelimb-placings made by the unrestrained forelimb was counted when therats’ whiskers were brushed against the side of a table. This was repeated 10 timesfor each rat on both sides.

2.6. Immunohistochemistry

Rats were sacrificed by terminal anaesthesia (50 mg kg−1 pentobarbital i.p.)and transcardially perfused with 100 ml heparinised saline followed by 150 ml 4%paraformaldehyde. Brains were rapidly removed, post-fixed in 4% paraformalde-hyde (in 0.2 M phosphate buffer) overnight and stored in 25% sucrose plus 0.1%sodium azide solution (in 0.2 M phosphate buffer). Serial coronal sections (40 �m)were cut using a freezing stage sledge microtome and a 1:6 series of sectionswas used for all quantitative immunohistochemistry. For immunohistochemicalanalyses, following quenching of endogenous peroxidise activity (using a solu-tion of 3% hydrogen peroxide/10% methanol in distilled water) and blockingof non-specific secondary antibody binding (using 3% normal horse serum inTris-buffered saline (TBS) with 0.2% Triton X-100 at room temperature for 1 h),sections were incubated overnight at room temperature with the appropriateprimary antibody (mouse anti-tyrosine hydroxylase (MAB318, 1:1000, Millipore)and human anti-�-synuclein (AB36615, 1:1000, Abcam)) diluted in TBS with0.2% Triton X-100. Sections were then incubated in a biotinylated secondaryantibody (horse anti-mouse (BA2001, 1:200, Vector)) for 3 h followed by 2 hincubation in streptavidin–biotin–horseradish peroxidase solution (Vector). Sec-tions were developed in 0.5% solution of diaminobenzidine tetrahydrochloride(Sigma) in Tris–buffer containing 0.3 �l ml−1 of hydrogen peroxide. Sections weremounted on gelatine coated microscope slides, dehydrated in ascending con-centrations of alcohols, cleared in xylene and coverslipped using DPX mountant(BDH chemicals).

2.7. Image analysis

Photomicrographs of immunostained sections were captured under bright-field illumination (using a Nikon DXM1200 C digital camera mounted on a Nikon

8 ral Bra

dwdsifcca(otft(id

2

brs

3

3o

aArlrPlrmtiSwuaro

FcHla

P. Mulcahy et al. / Behaviou

issecting microscope) and all image analysis was completed using Image J soft-are. For quantification of AAV-�-synuclein and/or rotenone-induced nigrostriatalamage, the number of tyrosine hydroxylase immunopositive cell bodies in theubstantia nigra, and the optical density of tyrosine hydroxylase immunoreactiv-ty in striatum, was quantified in 3 coronal sections from each region (AP distancesrom bregma (mm): striatum: +0.7, −0.3, −0.7; nigra: −6.0, −6.4, −6.8). For quantifi-ation of AAV-�-synuclein expression, the number of �-synuclein immunopositiveell bodies in the substantia nigra, and the optical density of �-synuclein immunore-ctivity in the striatum were quantified in a single coronal section from each regionAP distances from bregma (mm): striatum: −0.3; nigra: −6.4). For determinationf the optical density of immunostaining, images were converted to greyscale andhe mean grey values of the selected regions of the striatum (delimitated using theree-hand tool in Image J) were recorded. Mean grey values were then convertedo optical density measurements according to the algorithm specified in Image JOptical density = log10(255/mean grey value)). The optical density of backgroundmmunostaining (measured in the unstained cortex) was subtracted from all opticalensity values obtained.

.8. Statistical analyses

Data were analysed using 1-way or 2-way repeated measures ANOVA followedy post hoc Newman Keuls where appropriate. All tests used are indicated at theelevant point in the text. Differences between groups were considered statisticallyignificant when P < 0.05. All data are expressed as mean ± s.e.m.

. Results

.1. Effect of intracerebral ˛-synuclein and/or systemic rotenonen the rats’ general health

After virus surgery, all groups of rats continued to gain weightt the same rate as the Control group indicating that neitherAV-�-synuclein nor AAV-GFP had any detrimental impact on theats’ general health (Fig. 1A). However, from the second week fol-owing minipump implantation, both groups of rotenone-infusedats began to lose weight steadily (Group × Time, F(21,238) = 11.54,

< 0.001, post hoc Newman Keuls confirmed significant weightoss in the Rotenone and Combined groups). By the time of sac-ifice, both groups of rats implanted with a rotenone-infusinginipump (Rotenone and Combined) had lost almost 10% of

heir pre-implantation body weight, whereas both groups of ratsmplanted with a vehicle-infusing minipump (Control and �-ynuclein) had gained almost 5% over their pre-implantation bodyeight (Fig. 1B). Post-mortem assessment revealed a build up of

ndigested food in the stomachs of 11 of the 17 rotenone-infusednimals suggesting the weight loss may have been caused by aotenone-induced decrease in gastrointestinal motility as previ-usly reported [28].

ig. 1. Effect of intracerebral �-synuclein and/or systemic rotenone on the rats’ body wourse of the entire study, and (B) change in body weight after minipump implantation.

owever, after implantation of the minipumps, both groups of rats that were implantedose weight at a steady rate. Data are shown as mean ± s.e.m. ***P < 0.001 vs. Control by 2nd Combined groups from Week 15.5).

in Research 243 (2013) 6– 15

3.2. Systemic rotenone infusion exacerbates the behaviouralimpact of intranigral AAV-˛-synuclein

One of the key features of any model of Parkinson’s disease isthe progressive emergence of motor impairments in tests of spon-taneous motor function. Thus, to assess the impact of the single orcombined Parkinsonian insults on motor function, we subjected therats to a battery of behavioural tests that assess different aspectsof the motor disorder. Because the AAV-�-synuclein was adminis-tered unilaterally, we selected various tasks that assess lateralisedmotor function, namely the Corridor, Stepping and Whisker Tests.However, as rotenone was administered systemically, motor per-formance on both the ipsilateral (the side of AAV infusion) andcontralateral (the side opposite AAV infusion) sides was assessed(Figs. 2 and 3 respectively).

In its own right, unilateral intranigral infusion of AAV-�-synuclein caused contralateral (but not ipsilateral) motorimpairments in the Corridor Test of contralateral neglect, the Step-ping Test of forelimb akinesia and the Whisker Test of sensorimotorintegration (Figs. 2 and 3). Similarly, in its own right, continuoussystemic infusion of rotenone (2.5 mg kg−1 day−1 for 4 weeks) alsocaused motor dysfunction in all tests on both the ipsilateral andcontralateral sides (Figs. 2 and 3). However the most importantfinding of the present study was the additive effect of the geneticand environmental insults on the contralateral side of the rats’ bod-ies only.

In the Corridor Test, the impairments in contralateral retrievalinduced by either AAV-�-synuclein or rotenone were exacer-bated by combining the two insults (Fig. 3A; Group, F(3,30) = 40.76,P < 0.0001; post hoc Newman Keuls confirmed significant motordysfunction in the Combined group relative to the �-Synucleinand Rotenone groups on this side). Importantly, this additive effectwas not seen on the ipsilateral side (Fig. 2A; Group, F(3,30) = 155.40,P < 0.0001; post hoc Newman Keuls confirmed that the Combinedgroup did not differ from the Rotenone group). Similarly, in theStepping Test, the impairments in contralateral backhand steppinginduced by either AAV-�-synuclein or rotenone were exacerbatedby combining the two risk factors (Fig. 3B; Group, F(3,30) = 32.01,P < 0.0001; post hoc Newman Keuls confirmed significant motordysfunction in the Combined group relative to the �-Synuclein and

Rotenone groups on this side). As with the Corridor Test, this addi-tive effect was not seen on the ipsilateral side in this test (Fig. 2B;Group, F(3,30) = 17.11, P < 0.0001; post hoc Newman Keuls confirmedthat the Combined group did not differ from the Rotenone group). In

eight. Body weight data are shown as (A) actual body weight measured over theAfter intranigral virus injection, all rats continued to gain weight at the same rate.

with minipumps releasing rotenone (2.5 mg kg−1 day−1 s.c. for 4 weeks) began to-way repeated measures ANOVA with post hoc Newman Keuls (for both Rotenone

P. Mulcahy et al. / Behavioural Brain Research 243 (2013) 6– 15 9

Fig. 2. Effect of intracerebral �-synuclein and/or systemic rotenone on the rats’ ipsilateral motor function. Systemic rotenone infusion (2.5 mg kg−1 day−1 s.c.), but notunilateral intranigral AAV-�-synuclein, caused significant motor impairments in the (A) Corridor, (B) Stepping and (C) Whisker Tests. The increase in ipsilateral retrievals inthe Corridor Test in the �-Synuclein group represents a bias in retrieving food from the ipsilateral side in rats with a unilateral (contralateral) impairment. The XY plots depictt d datav ewma

caeFGs�

eett

he data collected over the course of the study, while bar charts represent collapses. Control; ++P < 0.01, +++P < 0.001 vs. �-Synuclein by 1-way ANOVA with post hoc N

ontrast, in the Whisker Test, there was no statistically significantdditive effect of AAV-�-synuclein and rotenone on either ipsilat-ral or contralateral vibrissae-elicited forelimb placings (Ipsilateral,ig. 2C: Group, F(3,30) = 31.13, P < 0.0001; Contralateral, Fig. 3C:roup, F(3,30) = 30.18, P < 0.0001; post hoc Newman Keuls confirmedignificant motor dysfunction in the Combined group relative to the-Synuclein but not the Rotenone group on both sides).

To confirm that the additive effect observed on the contralat-

ral side in the Corridor and Stepping Tests was genuine, we alsoxpressed these data as percentage change in performance rela-ive to pre-implantation of the minipump. Expressing the data inhis manner avoided any possible confounding effect of the lower

from the post-operative period only. Data are shown as mean ± s.e.m. ***P < 0.001n Keuls. CTL, Control; SYN, �-Synuclein; ROT, Rotenone; COM, Combined.

behavioural “baseline” in the Combined group after virus surgery(see Fig. 3A and B). When the data was expressed in this man-ner (Fig. 4), the additive effect of the environmental and geneticinsults was still clearly evident in the Corridor (Fig. 4A; Group,F(3,30) = 63.48, P < 0.0001; post hoc Newman Keuls confirmed sig-nificant motor dysfunction in the Combined group relative to the�-Synuclein and Rotenone groups) and Stepping Tests (Fig. 4B;Group, F(3,30) = 47.23, P < 0.0001; post hoc Newman Keuls confirmed

significant motor dysfunction in the Combined group relative tothe �-Synuclein and Rotenone groups) thereby excluding any con-founding effect of the lower behavioural “baseline” in the Combinedgroup after virus surgery.

10 P. Mulcahy et al. / Behavioural Brain Research 243 (2013) 6– 15

Fig. 3. Effect of intracerebral �-synuclein and/or systemic rotenone on the rats’ contralateral motor function. In their own right, both unilateral intranigral AAV-�-synucleinor systemic rotenone (2.5 mg kg−1 day−1 s.c.) caused significant motor impairments in the (A) Corridor, (B) Stepping and (C) Whisker Tests. However, the impact of thecombined genetic and environmental insults on contralateral motor function was significantly more pronounced than either insult alone in the Corridor and Stepping Tests.T chartsm +P < 0S

3n

irtts((ifi

he XY plots depict the data collected over the course of the study, while the bar

ean ± s.e.m. *P < 0.05, ***P < 0.001 vs. Control; #P < 0.05 vs. Rotenone; ++P < 0.01, ++

YN, �-Synuclein; ROT, Rotenone; COM, Combined.

.3. Systemic rotenone infusion exacerbates theeurodegenerative impact of intranigral AAV-˛-synuclein

Once behavioural testing was complete (17 weeks after intran-gral virus injection and 4 weeks after minipump implantation),ats were sacrificed for immunohistochemical analyses. Quantita-ive tyrosine hydroxylase immunohistochemical staining revealedhat, in its own right, unilateral intranigral infusion of AAV-�-ynuclein caused nigrostriatal degeneration on the injected side

Fig. 5). In contrast, continuous systemic infusion of rotenone2.5 mg kg−1 day−1) did not have any impact on nigrostriatalntegrity when injected alone. However, the most importantnding was that nigrostriatal degeneration at the level of the

represent collapsed data from the post-operative period only. Data are shown as.001 vs. �-Synuclein by 1-way ANOVA with post hoc Newman Keuls. CTL, Control;

nigrostriatal terminals was worsened in rats with combined treat-ment. Thus, in the striatum, the loss of nigrostriatal terminalsinduced by AAV-�-synuclein was exacerbated by subsequentchronic exposure to rotenone (Group, F(3,30) = 21.23, P < 0.0001;post hoc Newman Keuls confirmed significant terminal loss inthe Combined group relative to the �-Synuclein and Rotenonegroups on this side). This combined effect was not seen in thesubstantia nigra where combining the genetic and environmen-tal insults did not exacerbate the AAV-�-synuclein-induced cell

body loss (Group, F(3,30) = 19.64, P < 0.0001; post hoc NewmanKeuls confirmed that the Combined group did not differ from the�-Synuclein group). This indicates that combining intranigral AAV-�-synuclein and systemic rotenone infusion results in a model in

P. Mulcahy et al. / Behavioural Brain Research 243 (2013) 6– 15 11

Fig. 4. Effect of intracerebral �-synuclein and/or systemic rotenone on the rats’ contralateral motor function relative to pre-implantation performance. Contralateral motorperformance in the (A) Corridor, (B) Stepping and (C) Whisker Tests was also expressed as a percentage of “baseline” values obtained prior to implantation of the rotenone-or vehicle-infusing minipumps. Expressing the data in this manner confirmed that combining the genetic and environmental insults had an additive effect on performancein the Corridor and Stepping Tests as the impact of AAV-�-synuclein and rotenone was significantly more pronounced in the Combined group. The XY plots depict the datac final t# post h

wn

3˛

s

ollected over the course of the study, while the bar charts represent data from theP < 0.05 vs. Rotenone; ++P < 0.01, +++P < 0.001 vs. �-Synuclein by 1-way ANOVA with

hich the nigrostriatal neurons degenerate in a “dying-back” man-er.

.4. Systemic rotenone infusion does not exacerbate the

-synucleinopathy induced by intranigral AAV-˛-synuclein

To determine if unilateral intranigral AAV-�-synuclein orystemic rotenone infusion could induce any changes in the

ime-point only. Data are shown as mean ± s.e.m. **P < 0.01, ***P < 0.001 vs. Control;oc Newman Keuls. CTL, Control; SYN, �-Synuclein; ROT, Rotenone; COM, Combined.

expression of human �-synuclein in the rat brain, quantitativeimmunohistochemical staining for �-synuclein was completed(Fig. 6). As expected, unilateral intranigral infusion of AAV-�-synuclein caused a significant increase in the density of human

�-synuclein in the substantia nigra and the striatum but thiswas not enhanced when combined with rotenone infusion (Stri-atum: Group, F(3,35) = 14.63, P < 0.0001; Substantia nigra: Group,F(3,35) = 18.74, P < 0.0001; post hoc Newman Keuls confirmed greater

12 P. Mulcahy et al. / Behavioural Brain Research 243 (2013) 6– 15

Fig. 5. Effect of intracerebral �-synuclein and/or systemic rotenone on the rats’ ipsilateral nigrostriatal pathway. Unilateral intranigral AAV-�-synuclein, but not systemicrotenone infusion (2.5 mg kg−1 day−1 s.c.), caused significant degeneration of tyrosine hydroxylase immunopositive nigrostriatal (A) terminals and (B) cell bodies. However,the impact of the combined genetic and environmental insults on striatal dopaminergic deafferentation was significantly more pronounced than either insult alone. Dataa ; +P <

h

�r

4

imacuwirtewawcat

iio

re shown as mean ± s.e.m. *P < 0.05, ***P < 0.001 vs. Control; #P < 0.005 vs. Rotenoneydroxylase immunoreactivity.

-synuclein expression in the �-Synuclein and Combined groupselative to the Control and Rotenone groups).

. Discussion

This overall aim of this experiment was to investigate thenterplay between genes and the environment as an approach to

odelling Parkinson’s disease by exposing rats to both geneticnd environmental risk factors that are relevant to the humanondition. To this end, we administered rotenone to rats by contin-ous subcutaneous infusion (2.5 mg kg−1 day−1 for 4 weeks) fromeeks 13–17 after unilateral intranigral AAV-�-synuclein admin-

stration. We found that exposing AAV-�-synuclein-treated rats tootenone led to a model in which the classical Parkinson’s diseaseriad of progressive motor dysfunction, nigrostriatal neurodegen-ration and �-synucleinopathy was evident. However, this methodas not without its limitations as systemic rotenone infusion was

lso associated with bilateral motor dysfunction and loss of bodyeight. Thus, although we have shown that Parkinson’s disease

an be modelled in animals by combined exposure to both geneticnd environmental risk factors, this model is limited by systemicoxicity of the pesticide rotenone.

In line with previous reports, in its own right, unilateralntranigral infusion of AAV-�-synuclein led to contralateral motormpairments which were underpinned by ipsilateral expressionf the �-synuclein protein throughout the entire nigrostriatal

0.05 vs. �-Synuclein by 1-way ANOVA with post hoc Newman Keuls. TH-ir, tyrosine

pathway, as well as loss of ipsilateral nigrostriatal cell bodies andterminals [22,29–34]. Interestingly, the motor dysfunction causedby AAV-�-synuclein progressively emerged over the 17 weeks ofthe study and was particularly pronounced in the Corridor Testin line with our previous study [22]. This again highlights theAAV-�-synuclein model to be an excellent progressive model inits own right [33,35]. In contrast, systemic infusion of rotenonecaused bilateral motor dysfunction that was neither associatedwith �-synucleinopathy nor significant nigrostriatal neurode-generation. One possible explanation for why rotenone causedmotor dysfunction without a significant degenerative impacton the nigrostriatal pathway could be that rotenone is not aselective dopaminergic neurotoxin. Thus a toxic effect on otherneurotransmitter systems involved in motor control could haveaccounted for the motor dysfunction observed. However, analternative explanation for the behavioural results observed is thatthe systemic toxic effects of rotenone could have affected the rats’ability to perform the motor tasks. In this study, rotenone causedsignificant and progressive loss of body weight from the secondweek after minipump implantation which is in line with previousreports of rotenone-induced systemic organ toxicity [36–39,28].

Despite the limitations of bilateral motor dysfunction and body

weight loss induced by systemic rotenone, this study did revealsome important effects of combining the genetic and environmen-tal risk factors. The most important finding was the additive effectof systemic rotenone and unilateral intranigral AAV-�-synuclein

P. Mulcahy et al. / Behavioural Brain Research 243 (2013) 6– 15 13

Fig. 6. Effect of intracerebral �-synuclein and/or systemic rotenone on �-synuclein expression. Unilateral administration of AAV-�-synuclein into the rat substantia nigracaused significant expression of �-synuclein in the nigrostriatal pathway at the level of (A) the terminals in the striatum and (B) the cell bodes in the substantia nigra. However,t ed thC imm

otogo[dtecosaatutcpamwfasr

here was no evidence that subsequent administration of systemic rotenone affectontrol by 1-way ANOVA with post hoc Newman Keuls. �-synuclein-ir, �-synuclein

n ipsilateral nigrostriatal integrity and contralateral motor func-ion. Thus, combining the two risk factors resulted in exacerbationf nigrostriatal terminal loss on the ipsilateral side, and pro-ressive motor dysfunction in the Corridor and Stepping Testsn the contralateral side only in line with our previous report22]. This indicates that �-synuclein-overexpressing nigrostriatalopaminergic neurons were more sensitive to rotenone leadingo more pronounced degeneration ipsilateral but not contralat-ral to the side of AAV-�-synuclein infusion. Interestingly, thisombined neurodegenerative effect was only evident at the levelf the nigrostriatal terminals and not cell bodies which is repre-entative of the human disease where the nigrostriatal neuronspparently exhibit a “dying-back” degeneration from the stri-tum to the substantia nigra [40–44]. Although the mechanismshat underlie �-synuclein-induced neurodegeneration are not fullynderstood, mitochondrial dysfunction and oxidative stress arehought to contribute (for review see [45]). Since rotenone is alassical inhibitor of complex I of the mitochondrial electron trans-ort chain [46], mitochondrial dysfunction and oxidative stresslso underlie rotenone-induced neuronal death. Thus, because theotor and neuropathological features were significantly enhancedhen rats were exposed to both the genetic and environmental risk

actors, this indicates that the processes underlying �-synucleinnd rotenone-induced neuronal death are additive in this modelystem. The findings of the present study are in line with aecent study showing that �-synuclein transgenic rats are more

e level of �-synuclein expression. Data are shown as mean ± s.e.m. ***P < 0.001 vs.unoreactivity.

sensitive to the degenerative effects of systemic rotenone even inthe absence of overt nigrostriatal neurodegeneration [47]. This wasattributed to �-synuclein-induced oxidative stress and changes instriatal dopamine metabolism in the transgenic rats which mayalso contribute to the heightened sensitivity to rotenone seen inthe present study.

The data generated by this study support those reported inour previous study investigating the interplay between genesand the environment as an approach to modelling Parkinson’sdisease [22]. In both studies, exposing rats with a high nigrostri-atal load of �-synuclein (induced 13 weeks earlier by intranigralinjection of AAV-�-synuclein) to rotenone modelled the triadof classic features associated with human Parkinson’s disease,namely, �-synucleinopathy, nigrostriatal neurodegeneration andmotor dysfunction. However, these two studies differed in theirroute and chronicity of exposure to the pesticide. Thus, in our previ-ous study [22], rotenone was administered directly into the brain asa single intranigral injection (0.9 �g), whereas in this study, it wasadministered systemically by a subcutaneously implanted osmoticminipump (2.5 mg kg−1 day−1 for 4 weeks). Although the systemicroute may more accurately mimic human exposure to agriculturalpesticides, this approach is severely limited by body weight loss

and renders the model unsuitable for long-term preclinical drugdiscovery studies. In contrast, we have shown that direct intracere-bral administration of rotenone completely overcomes the toxicityassociated with systemic delivery (Mulcahy et al., 2011; [22]), and

1 ral Bra

tl

irtterafaitmsob

C

A

(

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

4 P. Mulcahy et al. / Behaviou

herefore the dual intranigral model may be more useful for suchong-term studies.

In summary, we have shown that unilateral intranigral admin-stration of AAV-�-synuclein and systemic administration ofotenone to rats leads to a model of Parkinson’s disease in whichhere is progressive motor dysfunction with underlying nigros-riatal �-synucleinopathy and neurodegeneration. The sequentialxposure to these two known risk factors for Parkinson’s diseaseesulted in a model with more exaggerated motor impairmentsnd neuropathology than either the genetic or environmental riskactor alone. Thus this model has improved aetiological, constructnd face validity over more widely used single insult models andt may be useful for examining the interplay between genes andhe environment in Parkinson’s disease. However, the use of this

odel in long-term drug-discovery programmes is limited by theystemic toxic effects of the pesticide. In such longer term studies,ur recently developed sequential intranigral dual-hit model maye invaluable [22].

onflict of interest

All authors declare that there are no conflicts of interest.

cknowledgment

This project was funded by Science Foundation Ireland08/RFP/NSC1344).

eferences

[1] Braak H, Del Tredici K, Rub U, de Vos RA, Jansen Steur EN, Braak E. Staging ofbrain pathology related to sporadic Parkinson’s disease. Neurobiology of Aging2003;24:197–211.

[2] Jankovic J. Parkinson’s disease: clinical features and diagnosis. Journal of Neu-rology, Neurosurgery and Psychiatry 2008;79:368–76.

[3] Vance JM, Ali S, Bradley WG, Singer C, Di Monte DA. Gene-environment interac-tions in Parkinson’s disease and other forms of parkinsonism. Neurotoxicology2010;31:598–602.

[4] Gao HM, Hong JS. Gene-environment interactions: key to unraveling the mys-tery of Parkinson’s disease. Progress in Neurobiology 2011;94:1–19.

[5] Tanner CM, Kamel F, Ross GW, Hoppin JA, Goldman SM, Korell M, et al.Rotenone, paraquat, and Parkinson’s disease. Environmental Health Perspec-tives 2011;119:866–72.

[6] Svenson LW, Platt GH, Woodhead SE. Geographic variations in the prevalencerates of Parkinson’s disease in Alberta. Canadian Journal of Neurological Sci-ences 1993;20:307–11.

[7] Imaizumi Y. Geographical variations in mortality from Parkinson’s disease inJapan, 1977-1985. Acta Neurologica Scandinavica 1995;91:311–6.

[8] Goldman SM, Tanner CM, Oakes D, Bhudhikanok GS, Gupta A, angston JW.Head injury and Parkinson’s disease risk in twins. Annals of Neurology2006;60:65–72.

[9] Van Den Eeden SK, Tanner CM, Bernstein AL, Fross RD, Leimpeter A, BlochDA, et al. Incidence of Parkinson’s disease: variation by age, gender, andrace/ethnicity. American Journal of Epidemiology 2003;157:1015–22.

10] Sato S, Hattori N. Genetic mutations and mitochondrial toxins shed newlight on the pathogenesis of Parkinson’s disease. Parkinson’s Disease2011;2011:979231.

11] Gorell JM, Rybicki BA, Johnson CC, Peterson EL. Smoking and Parkinson’s dis-ease: a dose-response relationship. Neurology 1999;52:115–9.

12] Ross GW, Petrovitch H. Current evidence for neuroprotective effects of nico-tine and caffeine against Parkinson’s disease. Drugs and Aging 2001;18:797–806.

13] Bove J, Prou D, Perier C, Przedborski S. Toxin-induced models of Parkinson’sdisease. NeuroRx 2005;2:484–94.

14] Blesa J, Phani S, Jackson-Lewis V, Przedborski S. Classic and new animal modelsof Parkinson’s disease. Journal of Biomedicine and Biotechnology 2012:845618.

15] Grunblatt E, Mandel S, Youdim MB. MPTP and 6-hydroxydopamine-inducedneurodegeneration as models for Parkinson’s disease: neuroprotective strate-gies. Journal of Neurology 2000;247(Suppl. 2):II95–102.

16] Lane E, Dunnett S. Animal models of Parkinson’s disease and L-dopa

induced dyskinesia: how close are we to the clinic? Psychopharmacology2008;199:303–12.

17] Linazasoro G. Recent failures of new potential symptomatic treatments forParkinson’s disease: causes and solutions. Movement Disorders 2004;19:743–54.

[

in Research 243 (2013) 6– 15

18] Willis GL, Kennedy GA. The implementation of acute versus chronic animalmodels for treatment discovery in Parkinson’s disease. Reviews in the Neuro-sciences 2004;15:75–87.

19] Biglan KM, Ravina B. Neuroprotection in Parkinson’s disease: an elusive goal.Seminars in Neurology 2007;27(106):112.

20] Kieburtz K, Ravina B. Why hasn’t neuroprotection worked in Parkinson’s dis-ease? Nature Clinical Practice Neurology 2007;3:240–1.

21] Meredith G, Sonsalla P, Chesselet M-F. Animal models of Parkin-son’s disease progression. Acta Neuropathologica 2008;115:385–98.

22] Mulcahy P, Walsh S, Paucard A, Rea K, Dowd E. Characterisation of a novelmodel of Parkinson’s disease by intra-striatal infusion of the pesticide rotenone.Neuroscience 2011;181:234–42.

23] Grimm D, Kern A, Rittner K, Kleinschmidt Jr A. Novel tools for production andpurification of recombinant adenoassociated virus vectors. Human Gene Ther-apy 1998;9:2745–60.

24] Dowd E, Monville C, Torres EM, Dunnett SB. The Corridor Task: a simple testof lateralised response selection sensitive to unilateral dopamine deafferenta-tion and graft-derived dopamine replacement in the striatum. Brain ResearchBulletin 2005;68:24–30.

25] Fitzsimmons DF, Moloney TC, Dowd E. Further validation of the corridor taskfor assessing deficit and recovery in the hemi-Parkinsonian rat: restorationof bilateral food retrieval by dopamine receptor agonism. Behavioural BrainResearch 2006;169:352–5.

26] Olsson M, Nikkhah G, Bentlage C, Bjorklund A. Forelimb akinesia in therat Parkinson model: differential effects of dopamine agonists and nigraltransplants as assessed by a new stepping test. Journal of Neuroscience1995;15:3863–75.

27] Schallert T, Fleming SM, Leasure JL, Tillerson JL, Bland ST. CNS plasticityand assessment of forelimb sensorimotor outcome in unilateral rat modelsof stroke, cortical ablation, parkinsonism and spinal cord injury. Neuro-pharmacology 2000;39:777–87.

28] Greene JG, Noorian AR, Srinivasan S. Delayed gastric emptying and entericnervous system dysfunction in the rotenone model of Parkinson’s disease.Experimental Neurology 2009;218:154–61.

29] Kirik D, Rosenblad C, Burger C, Lundberg C, Johansen TE, Muzyczka N,et al. Parkinson-like neurodegeneration induced by targeted overexpress-ion of alpha-synuclein in the nigrostriatal system. Journal of Neuroscience2002;22:2780–91.

30] Yamada M, Iwatsubo T, Mizuno Y, Mochizuki H. Overexpression of alpha-synuclein in rat substantia nigra results in loss of dopaminergic neurons,phosphorylation of alpha-synuclein and activation of caspase-9: resemblanceto pathogenetic changes in Parkinson’s disease. Journal of Neurochemistry2004;91:451–61.

31] Gorbatyuk OS, Li S, Sullivan LF, Chen W, Kondrikova G, Manfredsson FP, et al.The phosphorylation state of Ser-129 in human alpha-synuclein determinesneurodegeneration in a rat model of Parkinson disease. Proceedings of theNational Academy of Sciences of the United States of America 2008;105:763–8.

32] Azeredo da Silveira S, Schneider BL, Cifuentes-Diaz C, Sage D, Abbas-Terki T,Iwatsubo T, et al. Phosphorylation does not prompt, nor prevent, the formationof alpha-synuclein toxic species in a rat model of Parkinson’s disease. HumanMolecular Genetics 2009;18:872–87.

33] Ulusoy A, Decressac M, Kirik D, Bjorklund A. Viral vector-mediated overexpress-ion of alpha-synuclein as a progressive model of Parkinson’s disease. Progressin Brain Research 2010;184:89–111.

34] Koprich JB, Johnston TH, Huot P, Reyes MG, Espinosa M, Brotchie JM. Progres-sive neurodegeneration or endogenous compensation in an animal model ofParkinson’s disease produced by decreasing doses of alpha-synuclein. PLoS One2011;6:e17698.

35] Lindgren HS, Lelos MJ, Dunnett SB. Do alpha-synuclein vector injections providea better model of Parkinson’s disease than the classic 6-hydroxydopaminemodel? Experimental Neurology 2012;237:36–42.

36] Ferrante RJ, Schulz JB, Kowall NW, Beal MF. Systemic administration of rotenoneproduces selective damage in the striatum and globus pallidus, but not in thesubstantia nigra. Brain Research 1997;753:157–62.

37] Antkiewicz-Michaluk L, Karolewicz B, Romanska I, Michaluk J, Bojarski AJ,Vetulani J. 1-Methyl-1,2,3,4-tetrahydroisoquinoline protects against rotenone-induced mortality and biochemical changes in rat brain. European Journal ofPharmacology 2003;466:263–9.

38] Alam M, Schmidt WJ. L-DOPA reverses the hypokinetic behaviour andrigidity in rotenone-treated rats. Behavioural Brain Research 2004;153:439–46.

39] Lapointe N, St-Hilaire M, Martinoli MG, Blanchet J, Gould P, Rouillard C, et al.Rotenone induces non-specific central nervous system and systemic toxicity.FASEB Journal 2004;18:717–9.

40] Cabral GA, Harmon KN, Carlisle SJ. Cannabinoid-mediated inhibition ofinducible nitric oxide production by rat microglial cells: evidence for CB1receptor participation. Advances in Experimental Medicine and Biology2001;493:207–14.

41] Iseki E, Kato M, Marui W, Ueda K, Kosaka K. A neuropathological study

of the disturbance of the nigro-amygdaloid connections in brains frompatients with dementia with Lewy bodies. Journal of the Neurological Sciences2001;185:129–34.

42] Raff MC, Whitmore AV, Finn JT. Axonal self-destruction and neurodegeneration.Science 2002;296:868–71.

ral Bra

[

[

[

[

[47] Cannon JR, Geghman KD, Tapias V, Sew T, Dail MK, Li C, et al. Expres-

P. Mulcahy et al. / Behaviou

43] Morfini G, Pigino G, Opalach K, Serulle Y, Moreira JE, Sugimori M, et al. 1-Methyl-4-phenylpyridinium affects fast axonal transport by activation of caspase andprotein kinase C. Proceedings of the National Academy of Sciences of the UnitedStates of America 2007;104:2442–7.

44] Chu Y, Morfini GA, Langhamer LB, He Y, Brady ST, Kordower JH. Alterationsin axonal transport motor proteins in sporadic and experimental Parkinson’sdisease. Brain 2012;135:2058–73.

45] Cookson MR, van der Brug M. Cell systems and the toxic mechanism(s) of alpha-synuclein. Experimental Neurology 2008.

in Research 243 (2013) 6– 15 15

46] Degli Esposti M. Inhibitors of NADH-ubiquinone reduc-tase: an overview. Biochimica et Biophysica Acta 1998;1364:222–35.

sion of human E46K-mutated alpha-synuclein in BAC-transgenic ratsreplicates early-stage Parkinson’s disease features and enhances vulner-ability to mitochondrial impairment. Experimental Neurology 2012;240C:44–56.