original article the exploration of rotenone as a toxin for...

Available online at www.sciencedirect.com

Journal of Pharmacological and Toxicological Methods 57 (2008) 114–130www.elsevier.com/locate/jpharmtox

Original article

The exploration of rotenone as a toxin for inducing Parkinson's disease inrats, for application in BBB transport and PK–PD experiments

Paulien G.M. Ravenstijn a, Mario Merlini a, Marjolijn Hameetman a, Tracey K. Murray b,Mark A. Ward b, Hywel Lewis b, Gareth Ball b, Cathy Mottart c, Christine de Ville de Goyet c,

Thomas Lemarchand c, Kristel van Belle c, Michael J. O'Neill b,Meindert Danhof a, Elizabeth C.M. de Lange a,⁎

a Leiden Amsterdam Center for Drug Research, Leiden University, Division of Pharmacology, PO Box 9502, 2300 RA Leiden, The Netherlandsb Neurodegeneration Drug Hunting Team, Eli Lilly & Co Ltd., Erl Wood Manor, Windlesham, Surrey GU20 6PH, United Kingdom

c Department of Drug Deposition, Lilly Development Centre S.A., Rue Granbonpré 11, 1348 Mont-Saint-Guibert, Belgium

Received 13 July 2007; accepted 30 October 2007

Abstract

Introduction: In search for a suitable rat model to study potentially affected blood-brain barrier (BBB) transport mechanisms in the course ofParkinsons disease (PD) progression, experiments were performed to characterise Parkinsons disease markers following subcutaneous (SC) andintracerebral (IC) infusion of the toxin rotenone in the rat. Methods: Studies were performed using Male Lewis rats. SC infusion of rotenone(3 mg/kg/day) was performed via an osmotic minipump. IC infusion of rotenone occurred directly into the right medial forebrain bundle at threedifferent dosages. At different times following rotenone infusion, behaviour, histopathology (tyrosine hydroxylase and alpha-synucleinimmunocytochemistry), peripheral organ pathology (adrenals, heart, kidney, liver, lung, spleen and stomach) were assessed. In part of the SC andIC rats, BBB transport profiles of the permeability marker sodium fluorescein were determined using microdialysis. Results: SC rotenone failed toproduce dopaminergic lesions and led to extensive peripheral organ toxicity. BBB permeability for fluorescein following SC rotenone waschanged, however due peripheral toxicity. In contrast, IC rotenone produced a progressive lesion of the nigrostrial dopaminergic pathway over28 days with no associated peripheral toxicity. IC rotenone also exhibited a large increase in amphetamine induced rotational behaviour. Inaddition, a few IC rats showed alpha-synuclein immunoreactivity and aggregation. Following IC rotenone, no changes in BBB permeability weredetected after 14 days. Discussion: SC rotenone only produced peripheral toxicity. IC rotenone appeared to create a progressive lesion of the ratnigrostrial pathway, and may therefore be a more appropriate model of Parkinson's disease progression, compared with the most commonly used6-OH-DA rat model.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Blood-brain barrier; Parkinson's disease; Methods; Microdialysis; Neurotoxicity; Rat; Rotational behaviour; Rotenone; Tyrosine hydroxylase

1. Introduction

Parkinson's disease is a chronic, progressive neurodegenerativedisease inwhich nigrostratial dopaminergic neurones are graduallylost, leading to a decrease in dopamine concentration in the stria-tum (Dauer & Przedborski, 2003). It is the second most common

⁎ Corresponding author. Tel.: +31 71 527 6330; fax: +31 70 514 1260.E-mail address: [email protected] (E.C.M. de Lange).

1056-8719/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.vascn.2007.10.003

neurodegenerative disease and its current therapy focuses mainlyon symptomatic treatment by replacing the loss of dopamine in thestriatum (Mercuri & Bernardi, 2005) by L-Dopa, the dopamineprecursor. The most recent discoveries in the fields of diagnosisand treatment of Parkinson's disease have recently been reviewed(Savitt, Dawson, & Dawson, 2006). The main goal for futuretreatment of Parkinson's disease is the discovery and developmentof neuroprotective/neurotrophic drugs to diminish or, better, to haltthe disease progression (Bonuccelli & Del Dotto, 2006; Chen &

mailto:[email protected]

http://dx.doi.org/10.1016/j.vascn.2007.10.003

115P.G.M. Ravenstijn et al. / Journal of Pharmacological and Toxicological Methods 57 (2008) 114–130

Le, 2006). An important consideration for the development ofneuroprotective drugs for the treatment of Parkinson's disease isthe relationship between disease progression and the pharmaco-kinetic and pharmacodynamic (PK–PD) properties of the drug.

In the PK–PD relationship of anti-Parkinsonian drugs anumber of factors are of importance. One factor is transport ofthe drug across the blood-brain barrier (BBB). Drug transport isgoverned by a number of BBB transport mechanisms (de Boer,van, & Gaillard, 2003). Passive transport across the BBB occursby simple diffusion, either paracellularly or transcellularlyacross the cerebral capillary endothelial cells that compose theBBB. For hydrophilic solutes the diffusion via the paracellulartransport route is highly impeded by the presence of the tightjunctions between the brain capillary endothelial cells. Besides,many active influx and efflux transport mechanisms are presentat the BBB. For example, active efflux transport by P-glyco-protein (Pgp) or the multidrug resistance protein (MRP) as haspreviously been shown for numerous compounds (de Langeet al., 2000; Schinkel, Wagenaar, Mol, & van Deemter, 1996;van, et al., 2001; Vautier et al., 2006) for reviews see (Borst,Evers, Kool, & Wijnholds, 2000; Gottesman, Fojo, & Bates,2002; Loscher & Potschka, 2005; Schinkel, 1999).

BBB functionality is dynamically controlled by blood com-ponents and the surrounding brain cells by direct contact or in-directly by their extracellular products. Thus, BBB functionalitymay vary among different physiologic, pathologic, and chronicdrug treatment conditions. It should be realised that BBB transportof a particular drug into and out of the brain is the sum of all actualBBB transport mechanisms applicable to that particular drug, andthat changes in BBB transport mechanisms by the disease statemay therefore affect BBB transport of the drug. PK–PD relation-ships of symptomatic drugs such as levodopa (L-Dopa) in patientswith different stages of Parkinson's disease have been describedpreviously (Chan, Nutt, & Holford, 2005; Harder & Baas, 1998)and indicate that Parkinson's disease progression is able to in-fluence PK–PD parameters of anti-Parkinsonian drugs. Moreover,in a clinical study from 2005, Parkinson's disease was associatedwith a reduction in Pgp (Kortekaas et al., 2005). Carvey et al.(2005) found an increased leakage of the BBB in 6-hydroxydo-pamine (6-OH-DA) lesioned rat, indicated by the leakage of FITC-labeled albumin or horseradish peroxidase from the vasculatureinto the parenchyma, in both the striatum and substantia nigra(SNc) after 6-OH-DA injection. Furthermore, it has been shownthat blood to brain transport of L-Dopa is reduced in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) Parkinsonian mon-keys (Alexander, Schwartzman, Grothusen, & Gordon, 1994a)implicating changes in the BBB, as caused by the diseased state ofthe animal specifically to the large amino acid transporter.

It is important to reveal in which direction and to what extentthe different BBB transport mechanisms (and actual BBBtransport of a drug) are influenced by Parkinson's disease pro-gression. This should be addressed in a systematic manner, inrelation to the physico-chemical properties of the drug. To that endpreclinical studies are needed using an adequate animal model ofParkinson disease in which Parkinson disease progression can beidentified. The most widely used animal models of Parkinson'sdisease are neurotoxin models where MPTP (mice, cats and

primates) or 6-OH-DA (mice, rats, cats and primates) are used(Beal, 2001). These models are generally acute (Betarbet, Sherer,& Greenamyre, 2002; Schober, 2004), but in some cases MPTPsubchronic models in both primates and mice have been used(Schober, 2004). Recently, a mouse Parkinson's disease modelwas introduced in which MPTP was continuously administeredthrough an osmotic minipump (Fornai et al., 2005). This modelsappears to have the advantage over the other MPTP models thatthere is formation of inclusion bodies in the SN. Although thesedata look very promising, it has not yet been described elsewhere.An intrastriatal injection of 6-OH-DA seems to also show a moreprogressive degeneration of nigral dopaminergic cells comparedto intranigral injections or injections into the median forebrainbundle (MFB) (Cicchetti, Brownell, Williams, Chen, Livni, &Isacson, 2002; Sauer & Oertel, 1994). However, no Lewy bodyformation has yet been observed in any of the 6-OH-DA models.Other approaches, which have extensively been reviewed else-where (Betarbet et al., 2002; Uversky, 2004) include the geneticmodels of Parkinson's disease and the use of reserpine, meth-amphetamine, 3-nitrotyrosine paraquat in combination withmaneb and rotenone to induce Parkinson-like symptoms. Thesubcutaneous (SC) infusion of rotenone to induce Parkinson'sdisease in the rat neurons has been reported, in which a gradualincrease in striatal dopaminergic denervation with associatedalpha-synuclein-positive cytoplasmic inclusions in nigral (Betar-bet et al., 2000; Sherer, Kim, Betarbet, & Greenamyre, 2003).Most recently, the intracerebral (IC) infusion of rotenone into theMFB or SNc has been reported to affect the nigrostriatal pathwayand produce deficits which were reversed by L-Dopa (Alam,Mayerhofer, & Schmidt, 2004; Antkiewicz-Michaluk et al., 2004;Saravanan, Sindhu, & Mohanakumar, 2005; Sindhu, Saravanan,& Mohanakumar, 2005).

In this study we have first investigated the influences of SCadministration of rotenone (3 mg/kg/day) via an osmoticminipump in the rat on bodyweight and behaviour (locomotoractivity) as measured frequently within a period of 4 weeks.Within this period, at different time intervals, rats were sacrificedand brains were used to assess nigrostriatal damage based onimmunohistological staining on tyrosine hydroxylase, while inaddition for all rats peripheral organ pathologywas determined. Ina separate experiment, at 14 days of SC administration ofrotenone–BBB permeability was assessed using sodium fluor-escein as a marker. In a last experiment, peripheral organpathology was performed to determine the effect of 3 mg/kg/dayrotenone systemically. In a subsequent series of experiments, theinfluence of IC administration of rotenone in the MFB at threedifferent doses of rotenone (0.5, 2.0 and 5.0 μg) on nigrostriataldamage (immunohistological staining on tyrosine hydroxylase)was determined. To be able to compare these results to the resultsobtained in the experiments using the subcutaneous rotenonemodel, we carefully monitored bodyweight and peripheral organpathology was also performed on these rats. After confirming wewere producing a suitable lesion we used the highest rotenonedose (5.0 μg) in behavioural experiments and in a microdialysisexperiment using sodium fluorescein as a BBB permeabilitymarker. A summary of all the experiments which are presented inthis paper is shown in Table 1.

116 P.G.M. Ravenstijn et al. / Journal of Pharmacological and Toxicological Methods 57 (2008) 114–130

2. Materials and methods

2.1. Statement on use and care of animals

The experiments described in this paper were approved by theEthical Committee onAnimal Experimentation of the Universityof Leiden (DEC numbers 118 and 5069).

For all experiments (Table 1) inbred adult male Lewis rats(250–300 g, Charles River BV, Maastricht, The Netherlands)were used. The rats were housed in standard plastic cages (six percage before surgery and individually after surgery) with a normal12-hour day/night schedule (lights on 7:30AM) and a temperatureof 21 °C. The animals had access to standard laboratory chow(RMH-TM; Hope Farms, Woerden, The Netherlands) andacidified water ad libitum.

2.2. Surgical methods

2.2.1. Placement osmotic minipumpsIn these experiments, the rats were implanted with an Alzet

osmotic minipump (2ML-4, Alzet CR, Maastricht, The Nether-lands) placed subcutaneously on the back of the rat. For the rats inthe microdialysis experiment, the pump was placed 7 days aftermicrodialysis surgery. The osmotic minipump was filled witheither 3 mg/kg/day Rotenone (Pestanal®, Sigma Aldrich BV,

Table 1Setup of the experiments performed for the subcutaneous and intracerebral rotenone

Frequency of measurement

SalineBodyweight (week)daily until day of sacrificeBehaviour (Rotarod) (week)daily until day of sacrificeHistology (striatum and SNc) Once on day of sacrificeBehaviour (amphetamine challenged rotations) a Not measured for salineBBB permeability (fluorescein) a Once on day of sacrificePeripheral organ pathology Once on day of sacrifice

Vehicle (DMSO/PEG)Bodyweight (week)daily until day of sacrificeBehaviour (Rotarod) (week)daily until day of sacrificeHistology (striatum and SNc) Once on day of sacrificeBehaviour (amphetamine challenged rotations) a Once, 7 days prior to sacrifice (daBBB permeability (fluorescein) a Once on day of sacrificePeripheral organ pathology Once on day of sacrifice

RotenoneBodyweight (week)daily until day of sacrificeBehaviour (Rotarod) (week)daily until day of sacrificeHistology (striatum and SNc) Once on day of sacrificeBehaviour (amphetamine challenged rotations) a Once, 7 days prior to sacrifice (daBBB permeability (fluorescein) a Once on day of sacrificePeripheral organ pathology Once on day of sacrifice

In the experiments for the intracerebral rotenone model, three doses were used unleNA: not applicable.a Histology results on the brains of these rats are not presented in this paper.b A total of 4 rats died spontaneous during these experiments and were not incluc Only 5 μg rotenone was used in these experiments.d Total of 24 rats; 8 per treatment group.

Zwijndrecht, the Netherlands) dissolved in a 1:1-mixture ofdimethylsulfoxide (DMSO, Sigma Aldrich BV, Zwijndrecht, theNetherlands) with polyethylene glycol (PEG 200, Sigma AldrichBV, Zwijndrecht, the Netherlands) as the vehicle (DMSO:PEG;1:1), or filled with the vehicle alone or saline (0.9%NaCl). Pumpswere incubated in sterile saline at 37 °C overnight. For theimplantation, rats were deeply anesthetised using isoflurane(Forene®, Abbott B.V., Hoofddorp, The Netherlands). The Alzetosmotic minipumps were implanted under the skin on the back ofthe rat. The rats were weighed on regular basis to monitor generalwell-being. Rats were euthanised either after 7, 14, 21 or 28 daysor after the development of hunched posture suggestive ofabdominal pain and upon inadequate feeding or grooming,suggestive of ongoing toxicity.

2.2.2. Implantation of blood cannulasThe surgery for the microdialysis study was performed

under anesthesia with an intramuscular injection of 0.1 mg/kgmedetomidine hydrochloride (Domitor 1 mg/ml, Pfizer, Capellea/d IJssel, The Netherlands) and 1 mg/kg ketamine base(Ketalar 50 mg/ml, Parke-Davis, Hoofddorp, The Netherlands).Two indwelling cannulas (pyrogen-free, nonsterile polyethy-lene tubing, Portex Limited) were implanted, one in the leftfemoral artery and one in the left femoral vein. The cannula inthe left femoral vein was used for administration of fluorescein,

models

Subcutaneous rotenone model Intracerebral rotenone model

number of rats (N=) number of rats (N=)

Day 4 Day 14 Day 21 Day 28 Day 4 Day 7 Day 14 Day 28

24 18 12 6 96 72 48 2424 18 12 6 NA6 6 6 6 24 d 24 d 24 d 24 d

NA NA3 NA10 a 10 a 24 d 24 d 24 d 24 d

24 18 12 6 96 72 48 2424 18 12 6 NA6 6 6 6 24 d 24 d 24 d 24 d

y 28) NA 83 810 a 10 a 24 d 24 d 24 d 24 d

60 45 30 15 96 72 48 2460 45 30 15 NA15 15 15 15 24 d 24 d 24 d 24 d

y 28) NA 12 c

7 12c10 a, b 10 a, b 24 d 24 d 24 d 24 d

ss otherwise indicated: 0.5 μg, 2.0 μg and 5.0 μg.

ded in the analysis.

Table 2Validation of the determination of fluorescein: intra-assay and inter-assay variability,coefficients of variability and accuracy

Compound Added(ng/ml)

Intra-assay Inter-assay

(n=4) (n=10)

Found(mean±(ng/ml)SEM)

C.V. Accuracy Found(mean±(ng/ml)SEM)

C.V. Accuracy

(%) (%) (%) (%)

Fluorescein 1.0 0.9±0.03

6.0 93.7 1.1±0.08

22.3 107.0

2.0 2.0±0.02

1.6 98.8 2.0±0.03

4.6 102.1

5.0 5.1±0.03

1.1 101.7 4.9±0.07

4.4 97.5

10.0 10.1±0.08

1.7 101.3 10.3±0.22

6.9 102.7


whereas the cannula in the left femoral artery was used for serialcollection of arterial blood samples. The cannulas were tunneledsubcutaneously and fixed at the back of the neck with a rubberring. The skin in the neck was stitched with normal sutures. Theskin in the groin was closed with wound clips. To preventclotting and cannula obstruction, the cannulas were filled with a25% (w/v) polyvinylpyrrolidone solution (PVP; Brocacef,Maarssen, The Netherlands) in pyrogen-free physiological sa-line (B. Braun Melsungen AG, Melsungen, Germany) containing20 IU/ml heparin (Hospital Pharmacy, Leiden UniversityMedicalCenter, Leiden, The Netherlands). After the implantation of thecannulas, the rats were placed in a stereotaxic frame and the skullwas exposed for brain surgery.

2.2.3. Microdialysis surgery for the subcutaneous modelAfter the implantation of the cannulas, a small hole was drilled

to allow the implantation of a microdialysis guide cannula (CMA/12, Aurora Borealis Control B.V. Schoonebeek, TheNetherlands)in the striatum (relative to bregma (Paxinos, Watson, Pennisi, &

Fig. 1. (A) The effect of the subcutaneous infusion of (○) saline, (□) vehicle (DMSOexpressed as mean values±SEM. Statistical analysis (one-way ANOVA) showed noand no significant change between the saline and vehicle groups for any timepoint (PNand the saline-treated group (Pb0.001) or vehicle-treated group (Pb0.001) for all timgrey) 0.5 μg rotenone, (♢) 2.0 μg rotenone or (▲) 5.0 μg rotenone (n=24) on the bodyANOVA) showed no significant difference between all groups.

Topple, 1985): AP:+0.4; L:+3.2; V:−3.5). Two support screwswere placed to hold the guide, which was glued to the skull withdental acrylic cement (Howmedia simplex rapid+methylacrylate,Drijfhout, Amsterdam, The Netherlands). The osmotic minipumpwas placed 7 days after the microdialysis surgery and the expe-riment was carried out 14 days after pump implantation.

2.2.4. Microdialysis surgery for the intracerebral modelAfter the implantation of the cannulas, the rats received a

unilateral infusion into the right MFB (AP:−2.8; L:+2.0; V:−9.0relative to bregma; Paxinos et al., 1985) at a rate of 0.1 μl/min for30 min of either vehicle (DMSO/PEG, 1:1; n=8; referred to assham) or 5.0 μg of rotenone (n=12). After the infusion, the needlewas kept in place for another 5 min to allow diffusion of the fluidwithout leakage along the track of the needle. Subsequently, twosmall holes were drilled into the skull to allow implantation of amicrodialysis guide cannula (CMA/12, Aurora Borealis ControlB.V. Schoonebeek, The Netherlands) in the left and in the rightstriatum relative to bregma (AP:+0.4; L:±3.2; V:−3.5, relative tobregma; Paxinos et al., 1985). One support screwwas placed as anextra anchor for fixation of the guide, which was glued to the skullwith dental acrylic cement (Howmedia simplex rapid+methyla-crylate, Drijfhout, Amsterdam, The Netherlands). The micro-dialysis experiment was carried out 14 days after microdialysissurgery.

2.3. Experimental methods

2.3.1. Subcutaneous model: Behavioural experimentFor this study (in total n=108), the rats received an Alzet

osmotic minipump filled with either rotenone in vehicle (n=60,for a dose of 3 mg/kg/day), vehicle alone (n=24), or saline(n=24). For each treatment, the rats were divided into 4 groupswith different treatment durations of respectively 4, 14, 21 and28 days after which the rats were given an overdose ofNembutal®(Natriumpentobarbital 60 mg/ml, Ceva Sante Animale,

/PEG) and (●) 3.0 mg/kg/day rotenone on body weight over time. The data aresignificant difference in weight between all treatment groups on day 0 (PN0.05)0.05). Significant differences were observed between the rotenone-treated groupepoints. (B) The effect of an intracerebral infusion of (○) artificial ECF, (● lightweight over time. The data are mean values±SEM. Statistical analysis (one-way

Fig. 2. Illustrates the influence of either subcutaneous administration or intracerebral administration on the behaviour in the rat. A: Subcutaneous rotenone model: Thetime that a rat was able to remain walking on the Rotarod at different days following implantation of the osmotic minipump in (○) saline-treated rats, (△) vehicle(DMSO/PEG)-treated rats, (■) 3.0 mg/kg/day rotenone-treated rats. Panels A1–C1 represent the amount of data points in the 0–100 s region. Panels A2–C2 representthe amount of data points in the 100–200 s region. Panels A3–C3 represent the amount of data points in the 200- to 300-second region. B: Intracerebral rotenonemodel: Amphetamine (5 mg/kg i.p.) induced rotations measured on day 21 after infusion of either sham (DMSO/PEG; 1:1) or 5 μg of rotenone into the MFB. Leftpanel depicts the mean total rotation score (±SEM) after amphetamine injection and the right panel shows the mean asymmetry score over time after amphetamineinjection.


Table 3The effect of a subcutaneous infusion of saline (n=6 per timepoint), vehicle(n=6 per timepoint) or 3 mg/kg/day of rotenone (n=15 per timepoint) on theintensity of the striatal TH-immunoreactivity

Striatum

4 daysmean±SEM

14 daysmean±SEM

21 daysmean±SEM

28 daysmean±SEM

Subcutaneous rotenone modelSaline 104±7 103±11 85±20 102±11Vehicle 99±10 98±13 73±11 103±10Rotenone 88±6 96±7 89±7 95±6

Intracerebral rotenone modelaECF 93±3 91±3 96±4 89±30.5 μg Rotenone 60±12 64±8 55±13 64±102.0 μg Rotenone 50±13 34±15 42±11 98±25.0 μg Rotenone 70±13 44±9 27±9 35±12

The effect of different doses of rotenone (n=8 per dose and per timepoint) andartificial ECF (n=8 per timepoint) infused into the MBF on the percentage(lesioned vs. intact brain side) striatal TH immunostaining in time. The data arepresented as mean values±SEM.

Table 4The effect of a subcutaneous infusion of saline (n=6 per timepoint), vehicle(n=6 per timepoint) or 3 mg/kg/day of rotenone (n=15 per timepoint) on thenigral cell count

Substantia nigra

4 daysmean±SEM

14 daysmean±SEM

21 daysmean±SEM

28 daysmean±SEM

Subcutaneous rotenone modelSaline 163±18 155±13 204±18 146±8Vehicle 137±21 136±43 135±38 131±9Rotenone 129±9 157±17 145±11 141±8

Intracerebral rotenone modelaECF 104±9 101±7 84±9 90±50.5 μg Rotenone 68±15 92±9 62±9 62±62.0 μg Rotenone 72±13 45±12 68±2 91±65.0 μg Rotenone 63±13 74±7 58±6 35±8

The effect of different doses of rotenone (n=8 per dose and per timepoint) andartificial ECF(n=8 per timepoint) infused into the MBF on the percentage(lesioned vs. intact brain side) nigral cell count in time. The data are presented asmean values±SEM.


Naaldwijk, The Netherlands) and the thorax was opened and therat was perfused with 30 ml of saline followed by 30 ml of 10%phosphate buffered formalin (pH=7.0) via the left ventricle of theheart. Brains were removed for histopathology. During the entirecourse of the experiment, the bodyweight of the rats wasmonitored daily from Monday to Friday (Table 1). Beforeimplantation of the minipump, the rats were trained daily for2 weeks on the Rotarod (Ugo Basile, Comerio, Italy). Afterimplantation, the rats' ability to stay on the Rotarod was tested ona daily basis during the week, in two sessions. Before the firstsession the rat was weighed, and then placed on the drum of theRotarod with the head facing in the direction opposite to thedirection of rotation. This forced the rat to move forward in orderto stay on the rotating drum. The drum was started at a speed of2 rotations per minute (r.p.m) and accelerated every 30 s to a finalmaximum speed of 20 r.p.m. after 5 min. At this time, the sessionwas ended and the rat was allowed to recover for a 30 min periodbefore the second session was started. As a behavioural read-out,the time that the rat was able to stay on the drumwas recorded. Allrats that were unable to remain on the Rotarod for at least 250 sbefore the start of the treatment were excluded from data analysis.

2.3.2. Intracerebral model: Behavioural experimentFor this study, male Lewis rats received a unilateral infusion

into the MFB of either DMSO/PEG (1:1; n=8; vehicle) or 5.0 μgof rotenone (n=12). Approximately 21 days after rotenone in-fusion the animals were placed in automated rotometers (Med.Associates Ltd.). The apparatus consisted of perspex bowls whereeach rat was linked to a harness carrying an infrared sensor at thetop connected to a computer with ROTORAT software. The ani-mals were tested for rotations in absence (baseline) and in thepresence (stimulation) of amphetamine (5 mg/kg i.p.) to evaluatethe effects of rotenone treatment on stimulant-induced rotations.On the day of sacrifice (day 28), the animals were given anoverdose of anaesthetic and the thorax was opened and perfusedwith 30 ml of saline followed by 30ml of 10% phosphate buffered

formalin (pH 7.0) via the left ventricle of the heart. Brains wereremoved for histopathology.

2.3.3. Subcutaneous and intracerebral model: Microdialysisexperiment

At 18–24 hours prior to the experiment, the microdialysisprobes (CMA12, membrane length of 4.0 mm; Aurora BorealisControl B.V. Schoonebeek, the Netherlands) were gently insertedinto the guide cannulas after removal of the probe dummies.

The microdialysis experiment was started between 7:00 and8:00 a.m. The inlets of both microdialysis probes were connectedby FEP tubing (fluorinated ethylene propylene tubing; internalvolume of 1.2 μl/100 mm length; Aurora Borealis Control B.V.Schoonebeek, the Netherlands) to syringe pumps (Beehive, BasTechnicol, Congleton, United Kingdom). The probes were per-fused with aECF (composition in mM: NaCl 145; KCl 2.7; CaCl21.2; MgCl2 1.0; ascorbic acid 0.2 in a 2 mM phosphate buffer pH7.4; Moghaddam & Bunney, 1989) at a flow rate of 2 μl/min. Theoutlets consisted of fused silica tubing (I.D. 150μm,O.D. 375μm;Composite Metal Services Ltd, Iikley, United Kingdom) and wereconnected to a microsample collector (Univentor 820; Antec,Leiden, TheNetherlands), inwhich the sampleswere collected andcooled (4 °C). After a stabilization period of 60 min, the in vivorecovery of fluorescein was determined by the retrodialysismethod. For this purpose, the probes were first perfused with afluorescein solution (10 or 20 ng/ml in aECF) for 60 min to collect6 fractions. The relative loss of fluorescein was calculated as anaverage of this period. After this period, the syringes wereswitched to blank aECF for thewashout phase of 90min. After thewashout period, the intravenous administration of fluorescein wasstarted. The venous cannula was connected to a syringe containingfluorescein (Sigma Aldrich BV, Zwijndrecht, the Netherlands) insaline (0.9% NaCl). For a dose of 6 mg/kg of fluorescein, a 2-minute intravenous infusion was started at a rate of 50 μl/min. Inthe first 120 min of the experiment, microdialysis fractions werecollected at 10-minute intervals and from 120–180 min and from


180–240 min at 20- and 30-minute intervals, respectively. Bloodsamples (50μl in heparinised Eppendorf vials)were taken at 0, 0.5,2, 4, 6, 8, 10, 15, 20, 25, 30, 45, 60, 90, 120, 180 and 240min afterstart of the fluorescein infusion. The blood samples were cen-trifuged for 10min at 5000 r.p.m. and the plasma was pipetted intoEppendorf vials. All samples were stored at −20 °C.

2.4. Subcutaneous and intracerebral model: Histopathology

TH immunohistochemistry was performed to quantify thedegree of dopaminergic neurodegeneration. In short, the brainswere cut twice into 6-mm segments using a rodent brain matrix,processed, and embedded in paraffin wax. Coronal sections(8 μm) were cut through the striatum (at 1.2 mm caudal tobregma) and the SNc (from −4.5 to −6.2 mm caudal to bregma)(Paxinos et al., 1985) with a sledge microtome (Microm,Walldorf, Germany). The sections were deparaffinised and re-hydrated and endogenous peroxidase was quenched with 0.3%H2O2 for 30 min. The slides were placed in pepsin (0.2 g ofSigma-p-7000 pepsin in 50 ml of 0.01 M HCl) for 30 min,washed, and non-specific binding was blocked with 1.5% nor-mal goat serum (Vectastain rabbit IgG ABC kit; VectorLaboratories, Burlingame, CA). This was followed by the ap-

Fig. 3. A Intracerebral rotenone model: Images of tyrosine hydroxylase immunostrotenone-infused rats (0.5 (b), 2.0 (c) or 5.0 (d) μg) taken 14 days after infusion. Eachof the animal and the right striatum represents the lesioned side of the same animal. Thwhereas the highest rotenone dose (d) had more pronounced loss in staining and sommodel: Images of tyrosine hydoxylase staining at the level of the substantia nigra forminfusion. The left hand images show the intact substantia nigra with higher magnificatbrain with insets showing a higher magnification of the cell bodies. There is a clear doconcentrations of rotenone.

plication of the primary rabbit polyclonal anti-tyrosine hydroxy-lase antibody (AB152 incubated for 18 h at room temperature;Chemicon International, Temecula, CA), the secondary bioti-nylated antibody (Vectastain rabbit IgG ABC kit for 30 min)and the horseradish peroxidase conjugate (Vectastain rabbit IgGABC kit for 30 min). Visualisation was carried out using 3,3′-diaminobenzidine (Vector SK-4100 Vector Laboratories, Bur-lingame, CA) as a chromogen. The slides were coverslippedusing DPX mountant. Adjacent nigral sections were immunos-tained for alpha-synuclein using a anti-alpha-synuclein (1:200,AB15530, Abcam, UK).

After staining, striatal sections were scanned using SprintS-can 35 with PathScan Enabler™ at a resolution of 1012 dpi.Optimas 5.2 software was used to measure the optical density ofmanually-defined areas of each black and white image producedas a mean grey value (MGV). TH-IR was measured for the slide(background), cortex (control tissue staining), corpus callosum(control non-cellular staining), dorsal striatum (caudate puta-men, CPu) and ventral striatum (nucleus accumbens, NAcc).The values were adjusted for non-specific staining by subtract-ing the MGV of the corpus callosum or cortex, an area thatshould have no specific TH staining. Values obtained wereentered into a spreadsheet sorted by treatment group, and the

aining of coronal sections taken at the level of the striatum from aECF (a) ortreatment group is represented by 4 rats, the left striatum represents the intact sidee lowest (b) andmiddle (c) rotenone doses show partial dopaminergic denervatione rats had almost complete dopaminergic denervation. B: Intracerebral rotenoneaECF (a) or rotenone-infused rats (0.5 (b), 2.0 (c) or 5.0 (d) μg) on day 14 after

ion of the cell bodies as insets. The right hand side shows the lesioned side of these-dependent loss in TH positive cells and processes (see arrows) with increasing

Fig. 3(continued ).


mean, standard deviation and standard error (SEM) calculatedusing formulas. These means and errors were subsequentlyplotted and unless otherwise stated, all striatal TH-IR values arecorrected for cortical TH-IR.

A series of sections through the SN (−4.5 mm to 6.2 mm)were evaluated and stained. After careful evaluation under themicroscope the TH positive cells within the SNc were counted

at one selected stereotaxic level (−5.2 to5.4 mm caudal tobregma) in both vehicle-treated and the rotenone-lesioned ani-mals. The cells were counted manually using a digital camera(Hitachi HV-C20A) attached to a light microscope (LEICADMLB, Leica Microsystem Wetzlar GmbH) at ×20 magnifica-tion and the lesioned side is expressed as a percentage of theintact side. Light microscopy was also used to provide a


qualitative assessment of the presence of alpha-synuclein in theSNc in the intracerebral experiment.

2.5. Subcutaneous and intracerebral model: Peripheral organpathology

The pathological evaluation of the brain and peripheral organsof the subcutaneous experiment was performed on 56 rats and ofthe intracerebral experiment on 96 rats (Table 1). After 7, 14, 21 or

Fig. 4. Intracerebral rotenone model: Examples of alpha-synuclein positive staining(G and H) rotenone or aECF (A and B) into the left medial forebrain bundle (MFB,surgery. Both the intact SNc (A, C, E and G) and the lesioned SNc (B, D, F and H) arediaminobenzidine (DAB) with a haematoxylin counter stain.

28 days of treatment (for the subcutaneous experiment only 14 or28 days), the animals were given an overdose of Nembutal®(Natriumpentobarbital 60 mg/ml, Ceva Sante Animale, Naald-wijk, The Netherlands) and the thorax was opened and the rat wasperfusedwith 30ml of saline followedby 30ml of 10%phosphatebuffered formalin (pH=7.0) via the left ventricle of the heart.Adrenals, heart, kidney, liver, lung, spleen and stomach wereremoved and immersed in 10% neutral buffered formalin(pH=7.0), trimmed after adequate fixation according to a

in the SNc of rats that have received 0.5 μg (C and D), 2 μg (E and F) or 5 μgstereotaxic coordinates: AP:−2.8, Lat:+2.0, V:−9.0) and killed off 28 days postshown at 20× magnification. Alpha-synuclein staining was visualised with 3-6-


standard operating procedure inspired from the Registry ofIndustrial Toxicology Animal-data (RITA) standards (Bahne-mann et al., 1995), and appropriate samples routinely processed totissue blocks embedded with paraffin wax. Tissue sections werecut at approximately 3 μm, stained with Harris' haematoxylin andeosin (HE), Oil red O for neutral fat and Von Kossa's stain formineral/calcium deposits and examined by light microscopy forrelevant changes by a pathologist, without knowledge of thetreatment status.

2.6. Subcutaneous and intracerebral model: Analysis of fluorescein

The plasma and microdialysate samples obtained in thesubcutaneous model experiment were analysed for fluoresceinconcentrations by HPLC with fluorescence detection. TheHPLC system comprised a PSS Suprema size exclusion column(8×300 mm, 10 μm particle size), with a mobile phase of0.05 M ammoniumacetate (pH=9.0) and acetonitrile (4:1 v/v).The flow rate was maintained at a flow of 1.5 ml/min. Thedetection was performed using a fluorescence detector (LC240,Perkin Elmer, United Kingdom) with the excitation wavelengthset at 488 nm and the emission wavelength set at 512 nm. Themicrodialysate samples (15 μl) were injected directly onto theHPLC column. For the analysis of the plasma samples, 10 μlof plasma was added to 990 μl of an icecold artificial extracellu-lar fluid (described above) and vortexed for 10 s. A 50 μl aliquotof this mixture was injected directly onto the HPLC system

Fig. 5. Intracerebral rotenone model: Examples of possible alpha-synuclein overexprcells (n) within the SNc of rats with rotenone-induced lesions of the MFB. Imagehaematoxylin counterstain.

equipped with a refill guard column (2 mm I.D.×20 mm)(Upchurch Scientific, Oak Harbor, WA, USA) packed with C18(particle size 20–40 μm) (Alltech, Breda, The Netherlands).

For the analysis of fluorescein in plasma and microdialysatesamples from the intracerebral experiment, we developed anassay using the Fluostar Optima, which was a fast and reliablemethod. The results of the validation of this assay are presentedin Table 2. The detection limit was 0.08 ng/ml. For the analysisof the microdialysate samples, 10 μl microdialysate waspipetted onto a black 96-well plate (FIA-plate flat bottom/medium binding, Greiner, Alphen a/d Rijn, The Netherlands)and 40 μl of a 0.25 M NaOH solution was added. For theanalysis of the plasma samples, an aliquot of 10 μl plasma waspipetted onto a black 96-well plate and 90 μl of a 0.1 M NaOHsolution was added. The fluorescence intensity was measuredon the Fluostar Optima (BMG Labtech, Offenburg, Germany) atthe emission wavelength of 530 nm and the excitation wave-length of 480 nm.

2.7. Statistical analysis

All data were statistically evaluated for significance by one-way analysis of variance (ANOVA) and a Tukey–Kramer mul-tiple comparison test except for the microdialysis fluorescein datafrom the subcutaneous experiment which were evaluated usingStudent's t-test due to the small sample size of the saline andvehicle groups. Values of Pb0.05 were considered significant.

ession (o) and aggregation (a) in dopaminergic cells compared to non-expressings are at 100× magnification, alpha-synuclein was visualised with DAB and a


3. Results

3.1. Subcutaneous model: Bodyweight and locomotor activity

The subcutaneous infusion of rotenone produced a time-related reduction (Pb0.001) in bodyweight of rotenone-treatedrats, compared with vehicle- or saline-treated rats (Fig. 1A).This reduction was most pronounced in the first 10–14 daysfollowing the start of the treatment. Behavioural analysis usingthe Rotarod was performed at various timepoints after infusionof saline, vehicle or rotenone (Fig. 2A). The analysis was basedon the time-period in which the rat was able to remain on theRotarod (latency time). Results were categorised into 3 panels;panel 1 for latency times ranging between 0 and 100 s, for panel2 between 100 and 200 s, and for panel 3 between 200 and300 s. In saline-treated rats and vehicle-treated rats only 2–3%of the animals had a Rotarod latency lower than 100 s whereasin rotenone-treated rats this number had increased to 30%(panels A1, B1, C1). Also, 77% of the saline-treated rats and86% of the vehicle-treated rats were able to stay on the appa-ratus for more than 200 s, but only 45% of the rotenone-treatedrats were able to stay on the Rotarod for that period of time(panels A3, B3, C3). Therefore, rotenone treatment impairedRotarod performance, with a greater percentage of these ani-mals falling off the apparatus at early timepoints and feweranimals staying on for the full testing period (300 s).

3.2. Intracerebral model: Bodyweight and rotational behaviour

The effect of the intracerebral infusion of either aECF, 0.5 μgrotenone, 2.0 μg rotenone or 5.0 μg rotenone into the MFB on thebody weight of the rat (Fig. 1B) indicated that there was a decreasein body weight of about 10–15% in all treatment groups in the first

Table 5Subcutaneous rotenone model: Incidence (%) of rotenone-related histopathological cha

Scheduled sacrifices groups N=56

Day 14

Saline Vehicle Rotenon

n=10 n=10 n=8

StomachMineralisation, minimal to moderate 0 0 2 (25%)Forestomach mucosal thinning, min to slight 0 0 2 (25%)

HeartMyocardialnecrosis, minimal

0 0 2 (25%)

LiverDecreased glycogen content, marked to severe 0 0 7 (88%)Centrilobular single cell necrosis, slight 0 0 2 (25%)Increased centrilobular inflammation, slight 0 0 1 (13%)Oil red O negative microvacuolation, slight 0 0 1 (13%)

AdrenalMineralisation, cortex, minimal, unilateral 0 0 1 (13%)Increased cortical vacuolation, min to slight 0 0 2 (25%)

2 days after surgery. However, from 2 days onwards until the end ofthe experiment (day 28) the body weight increased as normal in alltreatment groups and overall there were no statistically significantdifferences among the various treatment groups. In agreement withthis, the experimenter's observations of the rats with regard togrooming indicated no noticeable differences among control ratsand rotenone-treated rats. To determine the behavioural conse-quences of intracerebral infusion of rotenone, we performed anexperiment in which we measured amphetamine (5 mg/kg i.p.)induced rotations in rats infused with 5.0 μg of rotenone andcompared these results to sham (=DMSO/PEG; 1:1) infused rats(Fig. 2B). The left panel depicts the mean total rotation score andthe right panel illustrates the time course of the rotations for both thesham infused rats and the rotenone-infused rats. The baseline ofboth groups, measured before the injection of amphetamine, showsno preference of any group for a particular side. However, after theamphetamine challenge the rotenone-infused had a large and sig-nificant increase in the number of ipsiversive rotations.

3.3. Histopathology

3.3.1. Subcutaneous modelLooking at the individual subcutaneous rotenone ratmodel data

from the behavioural and microdialysis experiments, only 2% ofthe rotenone-treated rats showed any loss of tyrosine hydroxylasestaining in the striatum. The loss of striatal staining in theseparticular rats showed a similar pattern to that reported by Betarbetet al. (2000). No loss of cell bodies in the SNc of any animals in thepresent studies was observed. The mean data for the striatum andSNc are shown in Tables 3 and 4, respectively. No significantdifference was seen in the mean striatal TH immunostaining ofcontrol, vehicle and rotenone-treated rats or in themean number ofTH positive cells in the SNc among these groups.

nges in Lewis rats sacrificed on schedule

Day 28 All day combined

e Saline Vehicle Rotenone Saline Vehicle Rotenone

n=10 n=10 n=8 n=20 n=20 n=16

0 0 5 (63%) 0 0 7 (44%)0 0 1 (13%) 0 0 3 (19%)

0 0 0 0 0 2 (13%)

0 0 5 (63%) 0 0 12 (75%)0 0 0 0 0 2 (13%)0 0 0 0 0 1 (6%)0 0 0 0 0 1 (6%)

0 0 0 0 0 1 (6%)0 0 0 0 0 2 (13%)

Fig. 6. Subcutaneous rotenone model: A: Limiting ridge of the stomach, Lewis rat given 3.0 μg/day of rotenone for 28 days. The muscularis and submucosa haveextensive mineralisation (arrow). haematoxylin and eosin (HE) stain. Original magnification 100×. B: Serial section of the change observed in Fig. 6A stainedaccording to von Kossa's method for calcium precipitates (arrow). Original magnification 100×. C: Right ventricular free wall of the myocardium, Lewis rat given3.0 μg/day of rotenone for 14 days. Minimal focus of myocardial necrosis (arrowheads) and adjacent perivascular mononuclear cell infiltration. HE. Originalmagnification 400×. D: Median lobe of the liver of a Lewis rat given 3.0 μg/day of rotenone for 14 days. Slightly increased single cell necrosis (inset) and slightinflammation (aggregates of mixed cells — arrow) are observed in numerous centrilobular areas adjacent to the central veins (CV), along with slight diffusecytoplasmic hepatocellular vacuolation (arrowheads). Occasional focal widening of the perivenous centrilobular spaces suggests loss of centrilobular hepatocytes. HE.Original magnification 100×. E: Median liver lobe from a Lewis rat given saline for 14 days. Clear cells with empty spaces (interpreted as washed-off glycogenstorage) and a few vacuolation (interpreted as lipid droplets) typical of normal liver of a non-fasted rat. CV: central vein. PV: portal vein. HE. Original magnification100×. F: Median lobe of the liver of a Lewis rat given 3.0 μg/day of rotenone for 28 days. Simple severe thinning of hepatic cords due to hepatocellular atrophy, withloss of clear cytoplasm and lipid vacuoles, interpreted as loss of glycogen content, typical of rats with anorexia. HE. Original magnification 100×.


3.3.2. Intracerebral modelThe histological evaluation of the intracerebral infusion of

rotenone indicated that there was no effect of the aECF infusionon the percentage of striatal TH immunostaining (Table 3 andFig. 3A), while there was a dose-dependent decrease in THstaining in rats with rotenone infusion. The dose-dependentnature of the toxin effect was most pronounced on day 14 afterinfusion. However, while the damage appeared to be progres-sive, there were no significant differences among any of the

rotenone doses on day 28 compared to day 14 and a time-dependent decrease can only be observed for the 5.0 μg rote-none dose. Also, there appears to be an increase in TH stainingin both the striatum and SNc for the 2.0 μg dose between day 14and 28. The TH immunostaining of the SNc (Table 4 andFig. 3B) showed that aECF infusion did not alter the number ofTH positive nigral cells, but infusion of rotenone produceddose-dependent decrease in cell number, although less pro-nounced than that observed in the striatum.

Fig. 7. (A) Plasma fluorescein concentration (μg/ml) vs. time (min) on day 14 after the subcutaneous infusion of (○) saline, (□) vehicle (DMSO/PEG) or (●) 3.0 mg/kg/day rotenone. The data are expressed as mean values±SEM. (B) Plasma fluorescein concentration (μg/ml) vs. time (min) on day 14 after an intracerebral infusion of(○) artificial ECF or (●) 5.0 μg rotenone. The data are mean values±SEM. Statistical analysis (one-way ANOVA) showed no significant difference between all groupsfor both rotenone models.


The results of alpha-synuclein immunostaining of the SNcindicated that there were alpha-synuclein positive cells in a fewof the rotenone-treated rats (Figs. 4 and 5).

The alpha-synuclein staining was variable among rats whilethere were more positive cells in some of the rotenone-treated rats,without an increase alpha-synuclein staining in several others.

3.4. Peripheral organ pathology

3.4.1. Subcutaneous modelFour rats (20%) died spontaneously after 3–5 days of rotenone

administration (3.0 mg/kg/day). No peripheral organ pathologywas performed on these rats. These mortalities were attributed totreatment. In the surviving rats, treatment-related and significantchanges were observed in the stomach and minor changes wereobserved in the liver and heart which were considered to be relatedto rotenone administration and are summarised in Table 5. Someminor changeswere also seen in the liver and adrenals, whichwere

Fig. 8. (A) Microdialysare fluorescein concentration (μg/ml) vs. time (min) on day 143.0 mg/kg/day rotenone. A significant difference was seen (timeN60 min) between rvehicle-saline-treated rats (Pb0.01; Student's unpaired t-test). (B) Microdialysate flinfusion of (○,□) artificial ECF or (●,■) 5.0 μg rotenone. Circles represent the contare mean values±SEM. Statistical analysis (one-way ANOVA) showed no significa

most probably the consequence of decreased food consumption orcomplete anorexia and stress, respectively.

Time-related increase in incidence and severity of gastricmineralisation, confirmed by a Von Kossa stain, was observed inthe smooth muscles fibers, the elastic connective tissue and occa-sionally in the glandular epithelial cells of the muscular layer, thesubmucosa, the muscularis mucosae and glandular crypts of thestomach (Fig. 6A and B). In the heart, minimal myofiber necrosischaracterised by significant coagulative necrosis of myofibers andan inflammatory response was observed in two rotenone-treatedrats sacrificed on day 14 (Fig. 6C). In the liver, slight increase incentrilobular single cell necrosis and/or inflammationwas noted in2 rats sacrificed on day 14, above what is usually seen in overnightfasted albino rat controls (Fig. 6D). A decrease in glycogen contentwas found in a majority of rats (Table 5), which was characterisedby a marked decrease or an absence of clear cytoplasmic spaceswithout sharp outlines, a smaller size of hepatocellular trabeculaeand was accompanied by centrilobular hepatocellular atrophy in

after the subcutaneous infusion of (○) saline, (□) vehicle (DMSO/PEG) or (●)otenone- and saline-treated rats (Pb0.05; Student's unpaired t-test) and betweenuorescein concentration (μg/ml) vs. time (min) on day 14 after an intracerebralrol-side of the brain and squares represent the infused side of the brain. The datant difference between all groups.

Table 6Mean area under the curves (AUC in μg/ml⁎min) of the plasma and estimatedextracellular fluid (ECF) curves for both the subcutaneous rotenone model aswell as the intracerebral rotenone model as shown in Figs. 7 and 8

AUCplasma

(μg/ml⁎min)mean±SEM

AUCECF

(μg/ml⁎min)mean±SEM

AUC ratio(μg/ml⁎min)mean±SEM

Subcutaneous rotenone modelSaline 1007±200 3.0±0.4 0.30±0.20Vehicle 1029±117 7.1±1.6 0.69±1.40Rotenone 1502±306 18±5 1.22±1.64

Intracerebral rotenone modelaECF (control) 642±42 1.6±0.2 0.25±0.03aECF (infused) 2.3±0.7 0.36±0.10Rotenone (control) 642±42 1.6±0.2 0.25±0.03Rotenone (infused) 2.6±1.2 0.40±0.23

To determine transport of fluorescein across the BBB, the AUCECF/AUCplasma

ratio was calculated. The data are presented as mean values±SEM.


the most severe cases (Fig. 6E–F). Even though it is only of minorimportance since it is anorexia-related, the change was clearly themost dramatic change in terms of incidence and severity. In theadrenal, minimalmultifocalmineralisation (calcification)was seenin the cortex of a single rat sacrificed on day 28.

3.4.2. Intracerebral modelFor the intracerebral rotenone model, no dose or time-related

changes were observed in the lungs, heart, liver, stomach, spleen,kidney or adrenals of rats given up to 5.0 μg of rotenone up to28 days, suggesting no peripheral toxicity.

3.5. Microdialysis experiment

3.5.1. Subcutaneous modelFor the subcutaneous rotenone model, all microdialysis con-

centrations were corrected for in vitro recovery (10.6±0.7%) inorder to estimate brain extracellular (brain ECF) concentrations.No statistically significant difference was found in concentra-tion–time profiles or in area under the curves (AUC) in plasmaamong any of the groups indicating that the treatment withrotenone or the vehicle had no effect on the overall dispositionof fluorescein (Fig. 7A and Table 6). However a significantdifference in the estimated brain ECF profiles was seen betweenrotenone-treated rats and saline-treated rats (Pb0.05) andbetween vehicle-treated rats and saline-treated rats (Pb0.01)from 80–150 min (Fig. 8A). The BBB transport of fluoresceinwas calculated as the ratio of AUCECF and AUCplasma multi-plied by 100%. The results are shown in Table 6. No statisticallysignificant difference was seen among all treatment groups.

3.5.2. Intracerebral modelFor the intracerebral rotenone model, the concentration–time

profiles of fluorescein in plasma are shown in Fig. 7B. Allmicrodialysate concentrations were corrected for the average invivo recovery as determined during the retrodialysis period (25±1%) in order to estimate brain extracellular fluid (brain ECF)concentrations. The in vivo recovery was equal for the two con-centrations of fluorescein used. The concentration–time profiles of

fluorescein in brain ECF are shown in Fig. 8B. No significantdifferenceswere found in area under the curves (AUC) in plasma orin brain ECF among any of the groups indicating that the treatmentwith rotenone or DMSO/PEG had no effect on the overall dispo-sition of fluorescein (Table 6). No statistical difference was ob-served in overall BBB transport of fluorescein among any of thegroups as indicated by the AUC ratios (Table 6).

4. Discussion

In this study we compared the subcutaneous and intracerebraladministration of rotenone to develop a rat model for Parkinson'sdisease. We examined brain histology, peripheral toxicity, behav-iour and BBB permeability.

4.1. Subcutaneous model

The subcutaneous infusion of rotenone resulted in minordevelopment of dopaminergic lesions whereas profound periphe-ral organ toxicity was found. Although the subcutaneous infusionseemed to affect locomotor behaviour and seemed to induce anincrease in BBB permeability compared to the control infusion(saline and vehicle), these results could not be correlated to nigro-striatal injury, but were most probably a consequence of peripheralorgan toxicity. A time-related increase in incidence and severity ofmineralisation (calcification) in the stomach was observed in ro-tenone-treated rats and is clearly related to the systemic rotenoneadministration. Since there were no significant kidney lesions, themineralisation is probably dystrophic secondary to cellular injury/necrosis, instead of metastatic associated with hypercalcemia orsystemic Ca/P imbalance related to renal failure (uremia) (Cotran,Kumar, & Collins, 1999; Palmer, 1993). However, this changewould deserve further investigations to definitively rule out ametastatic mechanism. The mineralisation (calcification) in thestomach of rotenone-treated rats could be the consequence ofmitochondrial injury due to inhibition of complex I and/or oxida-tive stress and secondary calcium influx (Sousa, Maciel, Vercesi,& Castilho, 2003; Sousa & Castilho, 2005). The reason as to whythe damage is localised to the stomach remains unsolved. Areduced blood-flow to the stomach as a result of chronic excessivestomach distension could be an explanation, although mucosalthinning which is indicative of excessive distension was not con-sistently observed. This change has not been, to our knowledge,previously reported in rats that received rotenone systemically.Stomach distension is usually a non-specific change related tophysiological demise, nausea (pica), severe stress or exaggeratedpharmacological effect and there were no morphological evi-dences of atrophy of parietal cells, which could support the pre-viously published hypothesis of achloridria with decreased wallmass (Lapointe et al., 2004). The observed cardiac changes in 2rotenone-treated rats were attributed to treatment, since none wereobserved in any of the saline- or vehicle-treated rats and thischange is not common in albino rats of that age; however, abackground change cannot strictly be ruled out, since this changemay occasionally be found in control Fischer 344 or SpragueDawley rats (Boorman, Eustis, Elwell, Montgomery, & MacK-enzie, 1990).(Boorman et al., 1990). Liver changes were minor


and related to decreased bodyweight with time and suspecteddecrease in food consumption. All of these changes could besecondary to the gastric lesions, especially the observed decreasein glycogen content. In addition, a few animals had atrophy ofhepatocellular cords along with increased single cell necrosis,inflammation and hepatocellular microvacuolation. Althoughthese changes could be associated with starvation, a rotenone-related effect cannot be ruled out. In contrast to a previous reportwith similar experimental conditions, hepatocellular coagulativenecrosis was not reported in this study (Lapointe et al., 2004).

The peripheral toxicity, high mortality and variability in thedevelopment of dopaminergic lesions in the brain of systemicallyadministered rotenone as we have seen in our experiments areconfirmed by literature (Betarbet et al., 2000; Fleming et al., 2004;Hoglinger et al., 2003; Lapointe et al., 2004; Sherer et al., 2003;Zhu et al., 2004) although it has not been so extensivelyinvestigated as in the experiments described in this paper. Inconclusion, these results indicate that the subcutaneous infusionof rotenone in the rat causes significant systemic toxicity, notablyimportant and adverse changes in the stomach and heart, while notable to produce a sufficient amount of nigrostriatal damage asreferred to as golden standard for Parkinson's disease.

4.2. Intracerebral model

The intracerebral infusion caused a progressive, dose-depen-dent decrease in dopaminergic neurons to about 30% in 28 days,with in a few cases alpha-synuclein immunoreactivity and aggre-gation but without any peripheral toxicity or any effect on body-weight. Although the alpha-synuclein inclusions were not directlyrelated to the severity of the dopamine lesions, they have not beenobserved in previous experiments using an intracerebral infusionof rotenone (Alam et al., 2004; Antkiewicz-Michaluk et al., 2004;Saravanan et al., 2005; Sindhu et al., 2005) and this result istherefore one step closer to mimicking the human pathology ofParkinson's disease. Saravan et al. observed a dose-dependent (2, 6and 12μg) decrease was in striatal biogenic levels already on day 5after the infusion (Saravanan et al., 2005). In our studies, we used alower concentration range which may account for the fact that wedo not see a clear dose-dependent decrease in the first 4 to 7 days.Furthermore, the decrease TH staining in time is most evident at5.0 μg, specifically in the SNc. In the striatum the decrease seemsto reach a minimal plateau after 28 days. Moreover, there appearsto be an increase in TH staining in both the striatum andSNc for the2.0 μg dose between day 14 and 28. This could perhaps beexplained by the relatively low concentration infused, some ratsbeing less susceptible to rotenone and general variability with thistoxin. Variability has also been observed in response to sys-temically treated rotenone rats (Betarbet et al., 2000; Sherer et al.,2003),. It would be of interest to look at higher doses and for alonger period of time and perhaps compare different strains of ratsspecifically for the infusion of rotenone into the MFB. The am-phetamine induced rotational behaviour of the intracerebral in-fused rotenone rats (5.0 μg) showed a significant preference ofthese rats for rotating in the ipsiversive direction on day 21 whichwas consistent with a unilateral lesion in the brain dopaminergicsystem. These results were somewhat different from a study by

Sindhu et al. (2005) where they did not see any significant ste-reotypic rotations inMFB-lesioned rats (12μg) until day 28.Againthese differences may be due to variation among rat strains (Lewisrats compared to Sprague Dawley rats).

In contrast to what would be expected from literature(Alexander, Schwartzman, Grothusen, & Gordon, 1994a; Carveyet al., 2005; Kortekaas et al., 2005), the microdialysis data showedno significant difference between the rotenone and sham infusedrats, indicating no difference in BBB permeability in the striatum inthe Parkinson's diseased brain. However, as themicrodialysis studywas performed at one particular timepoint at a specific location inthe brain, it cannot be concluded that no changes in BBBpermeability would occur in other stages of the disease, or at otherplaces within the brain. While microdialysis is a good technique tostudy BBB transport characteristics as it monitors concentrations inthe ECF in time (de Lange, Danhof, de Boer, & Breimer, 1997;Hammarlund-Udenaes, Paalzow, & de Lange, 1997), it does notprovide spatial information from a specific location in the brain. Inour view further investigations on BBB transport mechanisms areneeded, to be performed in a strict mechanistic and quantitativemanner, as a function of the disease state.

The intracerebral administration of rotenone creates theopportunity to develop an animal model for Parkinson's disease,in which the toxin is delivered in its target organ-reducing side-effects in other parts of the body that may interfere in the inves-tigations — with features getting closer to those in Parkinson'sdisease patients. However, upon local delivery of the toxin, thequestion remains to be answered on how toxins which are capableof inducing Parkinson's disease in human are able to reach theirtarget in the brain. Further experiments to follow the progression atlonger timepoints would be required to evaluate if indeed there isfurther disease progression and further alpha-synuclein develop-ment beyond 28 days. Since age seems to increase the sensitivity ofdopaminergic neurons to rotenone toxicity (Phinney et al., 2006), itwould be worthwhile to investigate this for the IC infused rotenonerats and the influence that age might have on the development ofalpha-synuclein inclusions.

Comparing these results to the more commonly used 6-OH-DA model, in our hands, 6-OH-DA infusion into the nigral cellbodies (O'Neill et al., 2004) or into theMFB (Visanji et al., 2006)produces a more rapid (5–10 days) loss in terminals. Also, wheninfused into the SNc, 6-OH-DA causes an anterograde degenera-tion of the whole nigrostriatal dopaminergic system (O'Neillet al., 2004; Sachs& Jonsson, 1975), resulting in amore rapid lossof tyrosine hydroxylase positive cells. Thus, the seemingly slowereffects of MFB infused rotenone obtained thus far favours a morechronic model than 6-OHDA and this might be due to differencesin the solubility, stability, diffusion of the toxin or the mechanismby which the toxin produces cell death. 6-OH-DA can be infusedinto the striatum and while a rapid loss in terminals is observedthis model also produces slow progressive retrograde damage tothe nigral cells (Murray et al., 2003; Rosenblad et al., 1999), butno alpha-synuclein inclusions have been observed so far in thismodel. Taken as a whole the data presented here indicate thatrotenone infused intracerebrally is able to create a progressive ratmodel for Parkinson's disease, which could then be used in PK–PD and other types of experiments.


Acknowledgement

The authors would like to acknowledge Dr Daniel G.Rudmann for critically reviewing the pathology section of themanuscript and the outstanding technical assistance of Claudinede Boelpaep in preparing the tissues for pathological assess-ment. The described work was financially supported by LillyDevelopment Centre, Mont-Saint-Guibert, Belgium.

References

Alam,M., Mayerhofer, A., & Schmidt,W. J. (2004). The neurobehavioral changesinduced by bilateral rotenone lesion in medial forebrain bundle of rats arereversed by L-Dopa. Behavioural Brain Research, 151(1–2), 117−124.

Alexander, G. M., Schwartzman, R. J., Grothusen, J. R., & Gordon, S. W. (1994a).Effect of plasma levels of large neutral amino acids and degree of parkinsonismon the blood-to-brain transport of levodopa in naive and MPTP Parkinsonianmonkeys. Neurology, 44(8), 1491−1499.

Alexander, G. M., Schwartzman, R. J., Grothusen, J. R., & Gordon, S. W.(1994b). Effect of plasma levels of large neutral amino acids and degree ofparkinsonism on the blood-to-brain transport of levodopa in naive andMPTP Parkinsonian monkeys. Neurology, 44(8), 1491−1499.

Antkiewicz-Michaluk, L., Wardas, J., Michaluk, J., Romaska, I., Bojarski, A., &Vetulani, J. (2004). Protective effect of 1-methyl-1,2,3,4-tetrahydroisoquino-line against dopaminergic neurodegeneration in the extrapyramidal structuresproduced by intracerebral injection of rotenone. International Journal ofNeuropsychopharmacology,, 7(2), 155−163.

Bahnemann, R., Jacobs,M., Karbe, E., Kaufmann,W.,Morawietz, G., Nolte, T., et al.(1995). RITA-registry of industrial toxicology animal-data-guides for organsampling and trimming procedures in rats. Experimental and ToxicologicPathology, 47(4), 247−266.

Beal, M. F. (2001). Experimental models of Parkinson's disease. Nature Reviews.Neuroscience, 2(5), 325−334.

Betarbet, R., Sherer, T. B., & Greenamyre, J. T. (2002). Animal models ofParkinson's disease. Bioessays, 24(4), 308−318.

Betarbet, R., Sherer, T. B., MacKenzie, G., Garcia-Osuna, M., Panov, A. V., &Greenamyre, J. T. (2000). Chronic systemic pesticide exposure reproducesfeatures of Parkinson's disease. Nature Neuroscience, 3(12), 1301−1306.

Bonuccelli, U., & Del Dotto, P. (2006). New pharmacologic horizons in thetreatment of Parkinson disease. Neurology, 67(7 Suppl 2), S30−S38.

Boorman, G. A., Eustis, S. L., Elwell, M. R., Montgomery, C. A., &MacKenzie,W. F. (1990). Pathology of the fisher rat—reference and atlas.

Borst, P., Evers, R., Kool, M., & Wijnholds, J. (2000). A family of drugtransporters: The multidrug resistance-associated proteins. Journal of theNational Cancer Institute, 92(16), 1295−1302.

Carvey, P. M., Zhao, C. H., Hendey, B., Lum, H., Trachtenberg, J., Desai, B. S.,et al. (2005). 6-Hydroxydopamine-induced alterations in blood-brain barrierpermeability. European Journal of Neuroscience, 22(5), 1158−1168.

Chan, P. L., Nutt, J. G., & Holford, N. H. (2005). Importance of within subjectvariation in levodopa pharmacokinetics: a 4 year cohort study in Parkinson'sdisease. Journal of Pharmacokinetics and Pharmacodynamics, 32(3–4),307−331.

Chen, S., & Le, W. (2006). Neuroprotective therapy in Parkinson disease.American Journal of Therapeutics, 13(5), 445−457.

Cicchetti, F., Brownell, A. L., Williams, K., Chen, Y. I., Livni, E., & Isacson, O.(2002). Neuroinflammation of the nigrostriatal pathway during progressive6-OHDA dopamine degeneration in rats monitored by immunohistochem-istry and PET imaging. European Journal of Neuroscience, 15(6), 991−998.

Cotran, R. S., Kumar, V., & Collins, T. (1999). Cellular pathology II:Adaptations, intracellular accumulations, and cell aging. Robbins' patholo-gic basis of disease (pp. 31−49). WB Sauners company.

Dauer, W., & Przedborski, S. (2003). Parkinson's disease: Mechanisms andmodels. Neuron, 39(6), 889−909.

de Boer, A. G., van der Sandt, I. C., & Gaillard, P. J. (2003). The role ofdrug transporters at the blood-brain barrier. Annual Review of Pharma-cology and Toxicology, 43, 629−656.

de Lange, E. C., Danhof, M., de Boer, A. G., & Breimer, D. D. (1997).Methodological considerations of intracerebral microdialysis in pharmacokineticstudies on drug transport across the blood-brain barrier. Brain Research BrainResearch Reviews, 25(1), 27−49.

de Lange, E. C., Marchand, S., van den Berg, D. J., van der Sandt, I. C., de Boer,A. G., Delon, A., et al. (2000). In vitro and in vivo investigations onfluoroquinolones; effects of the P-glycoprotein efflux transporter on braindistribution of sparfloxacin. European Journal of Pharmaceutical Sciences,12(2), 85−93.

Fleming, S. M., Zhu, C., Fernagut, P. O., Mehta, A., DiCarlo, C. D., Seaman,R. L., et al. (2004). Behavioral and immunohistochemical effects of chronicintravenous and subcutaneous infusions of varying doses of rotenone.Experimental Neurology, 187(2), 418−429.

Fornai, F., Schluter, O.M., Lenzi, P., Gesi, M., Ruffoli, R., Ferrucci, M., et al. (2005).Parkinson-like syndrome induced by continuous MPTP infusion: convergentroles of the ubiquitin-proteasome system and alpha-synuclein.Proceedings of theNational Academy of Sciences, United States of America, 102(9), 3413−3418.

Gottesman, M. M., Fojo, T., & Bates, S. E. (2002). Multidrug resistance incancer: Role of ATP-dependent transporters. Nature Reviews.Cancer, 2(1),48−58.

Hammarlund-Udenaes, M., Paalzow, L. K., & de Lange, E. C. (1997). Drugequilibration across the blood-brain barrier-pharmacokinetic considerations basedon the microdialysis method. Pharmaceutical Research, 14(2), 128−134.

Harder, S., & Baas, H. (1998). Concentration–response relationship of levodopain patients at different stages of Parkinson's disease. Clinical Pharmacologyand Therapeutics, 64(2), 183−191.

Hoglinger, G. U., Feger, J., Prigent, A., Michel, P. P., Parain, K., Champy, P., et al.(2003). Chronic systemic complex I inhibition induces a hypokineticmultisystem degeneration in rats. Journal of Neurochemistry, 84(3), 491−502.

Kortekaas, R., Leenders, K. L., van Oostrom, J. C. H., Vaalburg, W., Bart, J.,Willemsen, A. T. M., et al. (2005). Blood-brain barrier dysfunction inParkinsonian midbrain in vivo. Annals of Neurology, 57, 176−179.

Lapointe, N., St Hilaire, M., Martinoli, M. G., Blanchet, J., Gould, P., Rouillard,C., et al. (2004). Rotenone induces non-specific central nervous system andsystemic toxicity. FASEB Journal, 18(6), 717−719.

Loscher, W., & Potschka, H. (2005). Blood-brain barrier active efflux transporters:ATP-binding cassette gene family. NeuroRx., 2(1), 86−98.

Mercuri, N. B., & Bernardi, G. (2005). The ‘magic’ of L-Dopa: Why is it the goldstandard Parkinson's disease therapy? Trends in Pharmacological Sciences,26(7), 341−344.

Moghaddam, B., & Bunney, B. S. (1989). Ionic composition of microdialysisperfusing solution alters the pharmacological responsiveness and basaloutflow of striatal dopamine. Journal of Neurochemistry, 53(2), 652−654.

Murray, T. K., Whalley, K., Robinson, C. S., Ward, M. A., Hicks, C. A., Lodge,D., et al. (2003). LY503430, a novel {alpha}-amino-3-hydroxy-5-methyli-soxazole-4-propionic acid receptor potentiator with functional, neuroprotec-tive and neurotrophic effects in rodent models of Parkinson's disease. TheJournal of Pharmacology and Experimental Therapeutics, 306(2), 752−762.

O'Neill, M. J.,Murray, T. K.,Whalley, K.,Ward,M. A., Hicks, C. A.,Woodhouse,S., et al. (2004). Neurotrophic actions of the novel AMPA receptor potentiator,LY404187, in rodent models of Parkinson's disease. European Jouranl ofPharmacology, 486(2), 163−174.

Palmer, N. (1993). Bones and Joints. Pathology of domestic animals: (pp. 1−181).Academic Press Inc.

Paxinos, G., Watson, C., Pennisi, M., & Topple, A. (1985). Bregma, lambda andthe interaural midpoint in stereotaxic surgery with rats of different sex, strainand weight. Journal of Neuroscience Methods, 13(2), 139−143.

Phinney, A. L., Andringa, G., Bol, J. G., Wolters, E. C., vanMuiswinkel, F. L., vanDam, A. M., et al. (2006). Enhanced sensitivity of dopaminergic neurons torotenone-induced toxicity with aging. Parkinsonism Related Disorder, 12(4),228−238.

Rosenblad, C., Kirik, D., Devaux, B., Moffat, B., Phillips, H. S., & Bjorklund,A. (1999). Protection and regeneration of nigral dopaminergic neurons byneurturin or GDNF in a partial lesion model of Parkinson's disease afteradministration into the striatum or the lateral ventricle. European Journal ofNeuroscience, 11(5), 1554−1566.

Sachs, C., & Jonsson, G. (1975). Mechanisms of action of 6-hydroxydopamine.Biochemical Pharmacology, 24(1), 1−8.


Saravanan, K. S., Sindhu, K. M., &Mohanakumar, K. P. (2005). Acute intranigralinfusion of rotenone in rats causes progressive biochemical lesions in thestriatum similar to Parkinson's disease. Brain Research, 1049(2), 147−155.

Sauer, H., & Oertel, W. H. (1994). Progressive degeneration of nigrostriataldopamine neurons following intrastriatal terminal lesions with 6-hydroxydo-pamine: A combined retrograde tracing and immunocytochemical study in therat. Neuroscience, 59(2), 401−415.

Savitt, J.M., Dawson, V. L., &Dawson, T.M. (2006). Diagnosis and treatment ofParkinson disease: Molecules to medicine. Journal of Clinical Investigation,116(7), 1744−1754.

Schinkel, A. H. (1999). P-Glycoprotein, a gatekeeper in the blood–brain barrier.Advanced Drug Delivery Reviews, 36(2–3), 179−194.

Schinkel, A. H., Wagenaar, E., Mol, C. A., & van Deemter, L. (1996). P-glycoprotein in the blood-brain barrier of mice influences the brainpenetration and pharmacological activity of many drugs. Journal of ClinicalInvestigation, 97(11), 2517−2524.

Schober, A. (2004). Classic toxin-induced animal models of Parkinson's disease:6-OHDA and MPTP. Cell and Tissue Research, 318, 215−224.

Sherer, T. B., Kim, J. H., Betarbet, R., & Greenamyre, J. T. (2003). Subcutaneousrotenone exposure causes highly selective dopaminergic degeneration andalpha-synuclein aggregation. Experimental Neurology, 179(1), 9−16.

Sindhu, K. M., Saravanan, K. S., & Mohanakumar, K. P. (2005). Behavioraldifferences in a rotenone-induced hemiparkinsonian rat model developedfollowing intranigral or median forebrain bundle infusion. Brain Research,1051(1–2), 25−34.

Sousa, S. C., & Castilho, R. F. (2005). Protective effect of melatonin on rotenoneplus Ca2+-induced mitochondrial oxidative stress and PC12 cell death.Antioxidants and Redox Signalling, 7(9–10), 1110−1116.

Sousa, S. C., Maciel, E. N., Vercesi, A. E., & Castilho, R. F. (2003). Ca2+-induced oxidative stress in brain mitochondria treated with the respiratorychain inhibitor rotenone. FEBS Letters, 543(1–3), 179−183.

Uversky, V. N. (2004). Neurotoxicant-induced animal models of Parkinson's disease:Understanding the role of rotenone, maneb and paraquat in neurodegeneration.Cell and Tissue Research, 318, 225−241.

van der Sandt, I. C., Smolders, R., Nabulsi, L., Zuideveld, K. P., de Boer, A. G.,& Breimer, D. D. (2001). Active efflux of the 5-HT(1A) receptor agonistflesinoxan via P-glycoprotein at the blood-brain barrier. European Journalof Pharmaceutical Sciences, 14(1), 81−86.

Vautier, S., Lacomblez, L., Chacun, H., Picard, V., Gimenez, F., Farinotti, R., et al.(2006). Interactions between the dopamine agonist, bromocriptine and theefflux protein, P-glycoprotein at the blood-brain barrier in the mouse. Euro-pean Journal of Pharmaceutical Sciences, 27(2–3), 167−174.

Visanji, N. P., O'Neill, M. J., & Duty, S. (2006). Nicotine, but neither thealpha4beta2 ligand RJR2403 nor an alpha7 nAChR subtype selectiveagonist, protects against a partial 6-hydroxydopamine lesion of the ratmedian forebrain bundle. Neuropharmacology, 51(3), 506−516.

Zhu, C., Vourc'h, P., Fernagut, P. O., Fleming, S.M., Lacan, S., DiCarlo, C. D., et al.(2004). Variable effects of chronic subcutaneous administration of rotenone onstriatal histology. Journal of Comparative Neurology, 478(4), 418−426.

original article the exploration of rotenone as a toxin for...

Documents