A. Bjorklund and M. A. Cenci (Eds.) Progress in Brain Research, Vol. 184 ISSN: 0079-6123 Copyright ! 2010 Elsevier B.V. All rights reserved.

CHAPTER 15

Neural grafting in Parkinson’s disease: unraveling the mechanisms underlying graft-induced dyskinesia

Emma L. Lane!,†, Anders Björklund‡, Stephen B. Dunnett§, and Christian Winkler{

†Welsh School of Pharmacy, Cardiff University, South Wales, UK ‡Wallenberg Neuroscience Center, Department of Experimental Medical Science, Lund University, Lund, Sweden

§Brain Repair Group, School of Biosciences, Cardiff University, South Wales, UK {Department of Neurology, University Hospital Freiburg, Freiburg, Germany

Abstract: The development of neural transplantation as a treatment for Parkinson’s disease has been compromised by a lack of functional efficacy and the appearance of transplant-induced motor side-effects in some patients. Since the first reports of these graft-induced dyskinesias (GID), and the realization of their impact on the progress of the field, a great deal of experimental work has been performed to determine the underlying cause(s) of this problematic side-effect. In this review we describe the clinical phenomenon of GID, explore the different representations of GID in rodent models, and examine the various hypotheses that have been postulated to be the cause. Based on the available clinical and preclinical data we outline strategies to avoid GID in future clinical trials using fetal cell transplants or cell preparations derived from stem cells.

Keywords: Parkinson’s disease; Transplantation; Dopamine; Dyskinesia; Graft-induced dyskinesia; L-dopa-induced dyskinesia; Amphetamine-induced dyskinesia

Introduction of dopamine (DA) levels in the caudate–putamen. The cardinal motor symptoms of this disease—

Parkinson’s disease (PD) is characterized by a slowness of movement (bradykinesia), muscle progressive degeneration of dopaminergic neurons rigidity, and resting tremor—can be effectively in the substantia nigra with concomitant reduction alleviated by pharmacotherapy using DA agonists

or the DA-precursor L-dihydroxyphenylalanine (L-dopa). However, the efficacy of these drugs ! Corresponding author.

Tel:. "44-292-0874112; Fax: "44-292-0876749; wanes as the disease progresses and long-term E-mail: LaneEL@cardiff.ac.uk treatment with L-dopa commonly leads to the

DOI: 10.1016/S0079-6123(10)84015-4 295

296

Table 1. Definitions of abnormal movements

Dyskinesia Disturbance of movement according to the original definition; very often used to describe an increase of movements (hyperkinesia) following pharmacotherapy. Dyskinesia induced by chronic L-dopa medication is of two types: (1) peak-dose dyskinesia, seen during the on-phase of each L-dopa dose; and (2) diphasic dyskinesia, seen at the beginning and end of the L-dopa dose when dopamine levels are rising or decreasing

Chorea Derived from the Greek word for dance; movements are involuntary, usually short-lasting and quick, non-rhythmic and non-repetitive, primarily involving distal body parts

Dystonia Disturbance of muscle tone according to the original definition; usually associated with increased muscle tone due to involuntary, longer-lasting muscle contractions that put the head, body, or limbs into abnormal positions

Stereotypy Coordinated repetitive movements; usually suppressible by will LID L-dopa-induced dyskinesia—abnormal movements following administration of L-dopa; changes of dyskinesia after

transplantation usually mean changes in LID GID Graft-induced dyskinesia—abnormal movements following intracerebral transplantation and unrelated to drug stimulation AID Amphetamine-induced dyskinesia—abnormal movements in rodents with intracerebral dopaminergic transplants

induced by administration of amphetamine; used as an animal model of GID

development of motor fluctuations and abnormal involuntary movements (AIMs) known as dyski-nesia (see Table 1). These involuntary move-ments, typically choreic or dystonic in nature, can in themselves be severely debilitating and affect the quality of life. With limited treatment options available in the

later stages of PD, the need for alternative approaches was recognized several decades ago. After convincing results in animal models of PD, one of these approaches, neural transplantation (Bjorklund, 1992; Herman and Abrous, 1994; Winkler et al., 2000), moved into the clinic. To date, several hundred PD patients have received intracerebral transplants of fetal DAergic tissue obtained following elective surgical terminations of pregnancy. Scientific evaluation of grafted patients in several open-label trials demonstrated that fetal DAergic grafts survive and reinnervate the host striatum, release DA, and induce pro-nounced and continuous improvement of motor deficits despite an ongoing disease process in the brain (Hagell and Brundin, 2001; Kordower et al., 2008a, b Li et al., 2008; Piccini et al., 1999; Winkler et al., 2005). To validate the findings of the open-label trials and to guard against misinterpretation through placebo effects (de la Fuente-Fernandez and Stoessl, 2002; Goetz et al., 2008), two sham-surgery-controlled double-blind trials were funded by the NIH. However, in these trials improvements

of motor deficits were limited as compared to the open-label trials and improvements to motor func-tion were mild and only observed in sub-populations of transplanted patients (Freed et al., 2001; Ma et al., 2010; Olanow et al., 2003). An additional development was the reports from both NIH trials of unanticipated side-effects, the appearance of a new type of dyskinesia unrelated to ongoing medi-cation, now termed graft-induced dyskinesia (GID). During a temporary and self-imposed moratorium on transplantation for PD, there has been active discussion, alongside considerable new experi-mentation, aimed at resolving the issues identified as key in maximizing the functional benefit derived from the graft and minimizing side-effects. In this review we focus specifically on the devel-oping understanding of these side-effects.

The clinical problem of graft-induced dyskinesia (GID)

The first reports of unanticipated side-effects of transplants came in the Denver–Columbia trial (Freed et al., 2001; Greene et al., 1999; Ma et al., 2002), in which severe involuntary, predominantly dystonic, movements were observed in 4 out of 33 grafted patients, 6–12 months after grafting (for details, see summary in Tables 2A and 2B). Unusually, these dyskinesias were unrelated to the

297

intake of L-dopa, persisting for several days after the drug had been withdrawn (Cho et al., 2005; Graff-Radford et al., 2006; Olanow et al., 2003). A video-based review of the patients transplanted in open-label studies in the Lund–London–Marburg cohort then revealed off-medication dyskinesia in 7 out of 14 grafted patients (Hagell and Cenci, 2005; Hagell et al., 2002). Severity peaked 24–48 months after grafting, reaching a significant clinical amplitude in two patients. In 2003, Olanow and colleagues reported off-medication dyskinesia

Table 2A. Explicit reports of graft-induced dyskinesia in grafted patients

GID first described by… Freed et al. (2001) Hagell et al. (2002) Olanow et al. (2003)

Double-blind Open-label Double-blind No. patients 33 14 23 No. of patients with GID 5 (15%) 6 (43%) 13 (56%) Age of patients <60 Time course of GID—onset– 6–12 months 6–12 months 6–12 months peak NR 24–48 months 12 months Phenomenology Severe cranial Concurrent hyperkinesias and Repetitive stereotypic

dystonia or persistent dystonias. Repetitive stereotypic alternating contractions of dyskinesia in arm or ballistic movements flexor and extensor muscles generalized dyskinesia at low frequency

Pre-graft LID Yes Yes Yes Effect of transplant on PD Improvement Improvement Improvement symptoms (UPDRS)a

Effect of transplant on LID NR Improvement Improvement Effect of dopaminergic drugs Worsen Worsen Worsen on GIDb

Response to amantadine Two responded well, NR NR one transiently

DBS, Type and no. of patients GPi DBS (3)c/d STN (1) or GPi (1) DBS GPi (1) STN (3) DBS Response to DBS NR GPi DBS reduced GID, STN DBS GPi DBS 1 did well initially,

did note then developed generalized GID, STN DBS improved allf

Relationship to…..? Pre-graft FDOPA PET NR Yesg Noh

Post-graft FDOPA PET— Yesci No Noc

graft area Post-graft FDOPA PET— NR Change in ventral striatum associated NR ventral striatum (non-graft with poorer response, including GIDj

areas) Pre-graft UPDRS NR NR Noh

Post-graft UPDRS “on” NR Improvement Some improvement with GID. None without GID

Post-graft UPDRS “off” NR No Noh

Pre-graft L-dopa response No, but predictive of NR Noh

positive graft effect

(Continued)

in 56% of their patients in a second NIH placebo-controlled trial (the Tampa–Mount Sinai trial, Olanow et al., 2003). In all studies, off-medication dyskinesia was clearly distinguishable from L-dopa-induced dyskinesia, in being either more dystonic (Denver–Columbia trial), more stereotypic and rhythmic (Tampa–Mount Sinai trial), or both (Lund–London–Marburg cohort). In a recent video-based re-assessment of the patients in the Tampa–Mount Sinai trial the GIDs were described as repetitive, stereotypic movements in the lower

298

Table 2A. Explicit reports of graft-induced dyskinesia in grafted patients (Continued )

GID first described by… Freed et al. (2001) Hagell et al. (2002) Olanow et al. (2003)

Post-graft LID Immunosuppression

Tissue storage

NR None used

Yesl

Yesk

Increased or development of GID on withdrawalj

Yesm

Noh

Used for first 6 months, onset coincided with withdrawal No

Summary of clinical trials reporting graft-induced dyskinesia, GID, following intracerebral transplantation of fetal ventral mesencephalon in patients with Parkinson’s disease. Details are provided with reference to the article that first reported the phenomena but additional details have also been obtained from subsequent analyses and are referenced separately.

Abbreviations: DBS—deep brain stimulation; FDOPA PET—[18F] fluorodopa positron emission tomography; GID—graft-induced dyskinesia; GPi—internal segment of globus pallidus; LID—L-dopa-induced dyskinesia; NR—not reported; PD—Parkinson’s disease; STN—subthalamic nucleus; UPDRS—Unified Parkinson’s Disease Rating Scale a Data only available for first year (Freed et al., 2001), GID associated with residual PD symptoms (Olanow et al., 2003). b Some improvement in GID with dose reduction in L-dopa reported (Freed et al., 2001; Olanow et al., 2003). c Ma et al., 2002. d Freed et al., 2004. e Herzog et al., 2008. f Cho et al., 2009. g Reported as a trend, rs = -0.529, p = "0.064 (Hagell et al., 2002) h Olanow et al 2009. i High FDOPA signal in “hotspot” areas of the graft was only present in patients with GID (Ma et al., 2002). j Piccini et al 2005. k p = 0.015 (Hagell et al., 2002). l All tissue was cultured for 4 weeks and transplanted as “noodles”, no relationship between this and GID has been reported. m Two of the patients most severely affected by GID had tissue stored in hibernation media at 4"C for 1–8 days (Hagell et al., 2002).

Table 2B. Implicit reports of graft-induced dyskinesia in grafted patients

GID first described by… Defer et al. (1996) Jacques et al. (1999) Hauser et al. (1999)

Open-label Open-label Open-label No. patients in study 5 60 6 No. with GID-like 3 (60%) 1 (1.5%) 1 (17%) behaviors Age of patients 48–64 55 50 Phenomenology Worsened bilateral peak-dose “on” Increased Developed at 8 year post-

dyskinesia 2–6 months post-op, later dyskinesia following transplant with severe “off” development of asymmetric transplantation (no dyskinesia “groping contralateral peak-dose dyskinesia other details) movement” of right handa

Pre-op LID Yes NR Yes Effect of transplant on Improvement Improvement Improvement PD symptoms (UPDRS) Effect of antiparkinsonian Appearance despite stable Not alleviated by NR medication on dyskinesia medication, variable regimes used change in medical

regime Further treatment and NR Pallidotomy Improvement with GPi DBSa

response resolved

Early reports by Defer et al. (1996) and Jaques et al. (1999) and the recent publication by Hauser et al. (1999) as well as subsequent publications on the same patient cohorts mention changes in dyskinetic behaviors post-transplantation. There is not sufficient detail to conclusively determine whether these are true GIDs; therefore, they are included but with the limited details available as Table 1B. a Graff-Radford et al., 2006.

299

extremities with residual parkinsonism in other body regions, representing a prolonged form of diphasic dyskinesias (Olanow et al., 2009). Careful reading of transplantation studies published prior to the two controlled trials actually suggests that the alteration in the profile of dyskinesia after grafting is not a new phenomenon. Several early studies did report unusual dyskinetic behaviors both “on” and “off” L-dopa following transplanta-tion, although details are limited (Defer et al., 1996; Hauser et al., 1999; Herzog et al., 2008; Jacques et al., 1999; Kopyov et al., 1997) (see Table 2B). While, for the most part, GIDs have been mild,

some patients in each of the three studies described in Table 2A developed GIDs of a mag-nitude and severity that has necessitated surgical intervention by deep brain stimulation (DBS) (Freed et al., 2001). In several of these cases the expression of GID has been effectively controlled by DBS of the subthalamic nucleus or the internal segment of the globus pallidus (Cho et al., 2005; Freed et al., 2004; Graff-Radford et al., 2006; Herzog et al., 2008). Clearly, this is not an accep-table long-term solution to the problem of GID. A number of successful, individual transplantation cases have clearly demonstrated the potential ben-efits afforded by cell transplantation in PD and unraveling the mechanism(s) underlying the slow and protracted development of GID is thus an urgent priority. This goes beyond the continuation of fetal tissue transplant trials, with implications for the clinical development of alternative sources of cells, such as cells derived from embryonic or adult stem cells, or from induced pluripotent cells.

Finding the cause of GID

In an attempt to understand the root cause of GID, each clinical center has analyzed their patients as the basis for generating hypotheses for mechanisms (Hagell et al., 2002; Ma et al., 2002; Olanow et al., 2003). However, comparing PD transplantation studies, including those without any report of GID (e.g., Cochen et al., 2003; Mendez et al., 2000a, 2008),

has highlighted fundamental differences in the pro-cedures used (for details, see Winkler et al., 2005). This has made it impossible to define a common cause of GID. Nevertheless, a number of technical and clinical parameters have been considered for their potential role in the development of GID. We categorize these as either related to the patients (e.g., the selection of patients, their striatal and extra-striatal DA loss before and after surgery, use of immunosuppression, etc.) or to the transplantation procedure itself (including the preparation of the tissue, storage, and surgical technique). An outline of the procedure and the leading hypotheses at each stage of the process are summarized in Fig. 1. However, if we are to clarify the mechanisms under-lying these phenomena in patients, we need a valid animal model to allow more controlled experimental analysis.

Modeling GID in animals

When GID was first described in patients, transplant-induced AIMs had been observed in rodent models of PD but only in response to L-dopa, i.e., “L-dopa-induced dyskinesia” (LID). To quantify LIDs, clin-ical rating scales were adapted to score AIMs of the head, limbs, and body that developed in rodents. First described by Cenci and colleagues (Cenci et al., 1998; Lee et al., 2000), this model has now been extensively characterized enabling exam-ination of the molecular and behavioral impact of different PD therapies. Similarities to LID seen in PD patients, or in the more extensively used MPTP-treated primate model, have validated this approach (Dekundy et al., 2007; Lundblad et al., 2002; Maries et al., 2006; Winkler et al., 2002). Truly spontaneous GID, however, has only

been described in two animal studies (Lane et al., 2006; Vinuela et al., 2008). 6-hydroxydopamine (6-OHDA)-lesioned rats with intrastriatal dopami-nergic grafts displayed short bursts of mild AIMs at early post-grafting intervals but these were incon-sistent and unreliable (Lane et al., 2006; Vinuela et al., 2008). Even non-pharmacological stimuli,

Dissect Patient selection

DA A9 vs: A10 Embryonic age AgeDA vs: 5-HT Severity of disease

Collect Surgical procedure Preexisting LID Extrastriatal degeneration

Culture

Fresh

No. of cells transplanted Hibernate Location of cell deposits

Tissue distribution, hotspots Duration of culture

Culture process

Duration of hibernation Postoperative events Use of GDNF

Tissue preparation Immunosupression Ongoing extrastriatal degeneration

300

Fig. 1. Schematic representation of factors hypothesized to contribute to GID.Many parameters vary between the different grafting studies, most of which have been or continue to be considered as factors that may contribute to graft-induced dyskinesia (GID), beginning with tissue-related components from as early as the dissection of the embryo. The dissected tissue may or may not include serotonergic (5-HT) neurons or different proportions of DAergic (DA) A9 vs. A10 neuronal subtypes, depending on the dissection landmarks and/or the age of the donor fetus. Tissue has variously been used fresh, stored in so-called hibernation medium or cultured for some time, and either dissociated into a crude cell suspension or cut into small tissue fragments. There has been some exploration of the use of glial cell line-derived neurotrophic factor (GDNF) in the hibernation procedure—its possible role in GID is unknown. Patient-related factors include pre-transplantation characteristics such as the severity of the disease and extent of DA denervation; the presence or absence of L-dopa-induced dyskinesia, LID; and age at the time of surgery. Surgical and post-surgical conditions that may play a role in the development of GID include surgical technique, the number of transplanted cells, and their location and distribution through the caudate putamen, the use of immunosuppression, and ongoing extrastriatal degeneration.

such as exposure to a novel environment or tail to baseline). This aberrant contralateral graft-pinches, have been unable to consistently provoke mediated behavior reflects an overcompensation spontaneous dyskinesia (Carlsson et al., 2006; Lane induced by DA released from the transplanted et al., 2006; Vinuela et al., 2008). cells, does not correlate with the number of surviv-With hindsight, observations in the 6-OHDA rat ing DA neurons in a mature graft, and follows a

model of PD had already provided indications for different temporal profile to the turning behavior aberrant graft-induced behavior, associated with induced by DA released in the contralateral intact adverse biochemical or cellular activity in the striatum seen in animals with lesions alone grafted striatum. In unilateral 6-OHDA-lesioned (Herman et al., 1993; Lane et al., 2006; Torres rats, the DA-releasing agent amphetamine evokes et al., 2007, 2008) (see Fig. 2). a profound rotational behavior contralateral to the In an attempt to reproduce the clinical phe-lesioned and transplanted striatum (instead of nomenon of GID, grafted rats were observed restoring lesion-induced ipsilateral rotations back following injection of amphetamine and, using

A L-dopa-induced B Amphetamine-induced C Amphetamine-induced dyskinesia rotations dyskinesia

Sev

erity

of A

IMs

Mag

nitu

de o

f rot

atio

nsIp

sila

tera

l C

ontra

late

ral

Sev

erity

of A

IMs

Time (h) Time (h) Time (h)

6-OHDA-lesioned rats 6-OHDA-lesioned rats with mesencephalic grafts

301

Fig. 2. The most common rodent model of PD is the unilateral 6-OHDA-lesioned rat.In response to chronic L-dopa administration, supersensitization of the DA receptors on the lesioned side causes the animal to perform circling movements, rotations, in a direction contralateral to the lesion for 3 h and develop severe abnormal involuntary movements (AIMs) known as L-dopa-induced dyskinesia, LID (Fig. 2a, grey line). Following transplantation of DA-rich ventral mesencephalic VM tissue the severity of these behaviors is significantly reduced (Fig. 2a, black line). In response to an injection of amphetamine, the DA imbalance causes a rat with a 6-OHDA lesion to turn in the direction ipsilateral to the lesion (Fig. 2b, grey line). With the addition of a DAergic graft into the lesioned striatum, the direction of the rotational response reversed to a contralateral bias with a biphasic time course (rapid onset and vigorous rotation at 15–30 min after injection, return to lower rotation rates during the next 2–3 h, and a further increase of rotations during the last 1–2 h (Herman et al., 1993; Lane et al., 2006), total duration 6 h, Fig. 2b, black line). Very low levels of motor and orolingual stereotypy are observed in lesion-only animals following amphetamine, which may contribute to the occasional low-level AID scores (Fig. 2c, grey line). Some grafted animals may also show this very mild stereotypic response. However, importantly, many of the grafted rats will develop amphetamine-induced dyskinesia (AID), a more severe behavioral phenotype reminiscent of the LID behaviors, over the 12–16 post-transplantation (Fig. 2c, black line). This behavior has a different time course to the rotational response, gradually increasing over the first 30–45 min and then remaining stable over the next 4–5 h before slowly returning to baseline (Cenci et al., 1998; Lane et al., 2006). Behaviorally they tend to be less severe than LIDs and have a greater hyperkinetic component with extensive orolingual and limb movements (Carlsson et al., 2006). The profile of behavior induced by amphetamine may relate to the location of the graft and the distribution of its DAergic innervation within the striatum. Different striatal regions are known to contribute to different behavioral patterns with amphetamine-induced orolingual stereotypies originating in the ventro-lateral striatum (Kelley et al., 1988), overlapping with the caudal-lateral striatal areas known to contribute to LIDs (Cenci et al., 1998) (see Table 3 for other parameters that influence the expression of AIDs).

the rat LID scale, AIMs were observed in a pro- absence of any anti-PD medication and develops portion of transplanted animals, behaviors that progressively following transplantation in a pro-are not present following amphetamine adminis- portion of animals. Importantly, these involuntary tration to lesion-only animals (Carlsson et al., movements can be clearly distinguished from 2006; Lane et al., 2006). These amphetamine- amphetamine-induced rotation (see Fig. 2), and induced dyskinesias (AID) have since been despite showing some similarities to LID, AID observed in several further studies (Tables 2A is comprised to a larger extent by dystonic axial and 2B) (Carlsson et al., 2007; Lane et al., 2008, movements and hyperkinetic limb and orofacial 2009a, b). As spontaneous GID in rodent PD dyskinesia (Carlsson et al., 2006), which are char-models is too inconsistent and unpredictable, acteristic features of clinical GID. AID has come to serve as an experimental There are alternative reported approaches to model for the study of GID. Similar to GID studying GID in rodent models, the main feature seen in grafted patients, AID is observed in the of which is the characterization of AIMs in grafted

302

Table 3. Animal models of post-grafting dyskinesia in PD (adapted from Lane et al. 2009a)

Post-transplantation Spontaneous transient Amphetamine- L-dopa-induced AIMs L-dopa-induced forelimb-behaviors forepaw tapping and body induced AIMs facial stereotypies

twisting

Term GID AID LID LID

Observation period 2 ! 1 min (sessions at repeated intervals)

Dopamine neuron grafts Presentc,d

Graft size/No. of Expressed in small and dopamine cells large graft groupsc,d

Location of dopamine Not observedb

cells in striatum

A9 vs A10 dopamine Not observedf

neurons Role of dopamine Some observed in both transporter (DAT) DAT KO and WT groupsi

5-HT neurons Not observedb,c

Inflammation Not observedg

18 ! 1 min (every 20 mins for 6h)c,d,e,f,g

Presentc,d,e,f

Not observedf

No within-group correlationc

Worsened with more caudal graft vs rostralb

Not observedf

Not observed in WT or DAT KOi

No effect c,e,j,k

No effectg

9 ! 1 min (every 20 min for 3h)a–l

Improvea–l

Improve with more cellsa,c,i

Improved with more caudal graft vs rostralb

A9 correlates with LID improvementf

Greater LID improvement with DAT KO graftsi

Worsen e,j

No effectg Increased tappingh

1 ! 2 mins (30 mins after L-dopa)a,b

Presentd,i,l

No effecti Less with larger graftsd

Greater with “hotspot” graft of same cell number as diffuse graftd,l

–

–

– No effecth

All these models were produced in rats with unilateral 6-OHDA lesions of the nigrostriatal pathway, receiving transplants of embryonic VM tissue in the DA-denervated striatum. Effects are reported on post-grafting dyskinesia, defined here as either spontaneous, non-drug-induced behaviors (GID), dyskinetic behaviors following amphetamine administration (AID) or L-dopa (LID). Full paper citations are given in the reference list. - = issue not reported/studied in this model; “Not observed” = behavior reported as not being present in these studies; “No effect” = reported as no difference between groups. WT = transplanted cells from wildtype mice; DAT KO = transplanted cells from DA transporter knockout mice; AIMs = Abnormal involuntary movements; 5-HT = serotonergic. a Lee et al. (2000) b Carlsson et al. (2006) c Lane et al. (2006) d Maries et al. (2006) e Carlsson et al. (2007) f Kuan et al. (2007) g Lane et al. (2008) h Soderstrom et al. (2008) i Vinuela et al. (2008) j Carlsson et al. (2009) k Lane et al. (2009) l Steece-Collier et al. (2009).

animals shortly after administration of L-dopa one characteristic feature of GID in patients is (Maries et al., 2006; Soderstrom et al., 2008; that it is unrelated to L-dopa intake and persistent, Steece-Collier et al., 2009; Vinuela et al., 2008) in some cases after prolonged withdrawal from (Table 3). In these studies, grafted animals display L-dopa. This suggests that mechanisms underlying repetitive and stereotypic grabbing and gnawing, GID and LID may be related but still distinct. In or a tapping behavior. The increase in stereotypy this review we term behaviors produced solely by observed in these experimental studies provides the graft GID, while AID is used as an experi-an interesting feature for further study (Cenci and mental model to study a type of GID that shares Hagell, 2006; Olanow et al., 2003). Nonetheless, some characteristic features, particularly its

303

dystonic and stereotypic nature, with GID seen in patients. In contrast, the use of L-dopa to stimulate dyskinesia is a parallel to the so-called peak-dose dyskinesia seen in the clinical “ON” state. There-fore, changes to LID as a result of transplantation are referred to as post-transplantation LID.

Selection of patients to prevent GID

In studies of grafted patients, correlation analyses have variously suggested that potential functional benefits derived from the graft are limited, if— prior to transplantation—PD patients are either too old, too severely affected by the disease, show a bad preoperative L-dopa response, or when degeneration of the DA system is not restricted to the putamen but extends to more ventral regions (Freed et al., 2001; Olanow et al., 2003; Piccini et al., 2005; Winkler et al., 2005). The Lund open-label trial demonstrated a trend toward a negative correlation between severity of GID and preoperative putaminal 18F-DOPA uptake, suggesting that more severely affected PD patients may carry a higher risk for the develop-ment of GID (Hagell et al., 2002; Piccini et al., 2005). In this same study, there was a further correlation between the severities of post-operative peak on-phase dyskinesia and off-phase dyskinesia (Hagell and Cenci, 2005). In the Olanow et al. (2009) study patients with or without GID dis-played similar levels of LID, and also similar striatal 18F-DOPA uptake on PET, both pre-and post-operatively. Nevertheless, the mechanisms of LID and GID may still be related in the sense that patients who have developed severe LID prior to surgery run an increased risk of developing GID after transplantation. Indeed, in one rodent study we have reproduced the positive correlation between pre-operative LID and post-operative AID (Lane et al., 2006). In a further study of the role of pre-transplantation L-dopa in GID we found that 6-OHDA-lesioned rats had a greater tendency to develop AID when the animals had been exposed to chronic L-DOPA prior to

transplantation (Lane et al., 2009b) irrespective of graft size. Most transplantation patients will have received long-term L-dopa treatment, and a good pre-operative L-dopa response has been part of the inclusion criteria for many transplantation stu-dies. Indeed, a good L-dopa response was found to be predictive of a positive outcome in the Denver–Columbia trial (Freed et al., 2001). Thus, further studies have examined the relation-ship of AID to the severity of pre-transplantation LID. Animals with moderate-to-severe LID showed a greater risk of developing AID, while animals with no or only minor LID, despite chronic L-dopa treatment, developed no or only very subtle GID (Lane et al., unpublished obser-vations; Winkler et al., 2009). The implication of these clinical and preclinical data is therefore that patients with severe LID may be less appropriate candidates for neural transplantation. In conclu-sion, PD patients selected for neural transplanta-tion should not be too severely affected by the disease and with DAergic degeneration restricted to the putamen in order to obtain maximum functional benefit. Prior to grafting the patients should have shown a good therapeutic L-dopa response, but little or no LID, in order to reduce the risk for the development of GID.

Is immunosuppression required?

It is now well established that the immune privi-lege of the brain is not complete. Thus, while intracerebral DAergic grafts may survive for long periods after transplantation, they may be infiltrated by immune cells as suggested by post-mortem analyses from the two US controlled trials (Freed et al., 2001; Olanow et al., 2003). Beha-vioral recovery after grafting may therefore have been affected by an ongoing immune response induced by either lack of immunosuppression, as in the Denver–Columbia trial (Freed et al., 2001), or by withdrawal of low-dose cyclosporine by 6 months after grafting and concomitant waning of the behavioral benefit, as was the case in the

304

Tampa–Mount Sinai trial (Freed et al., 2001; Olanow et al., 2003). In contrast, in the Lund– London–Marburg patients, triple immunosuppres-sion was continued for at least 12 months after grafting and behavioral improvement continued after gradual reduction and eventual withdrawal of immunosuppressive drugs (Piccini et al., 2000; Wenning et al., 1997). GID has so far not been convincingly associated with an ongoing non-rejecting inflammatory process within and around the graft. Indeed, in a recent animals study we did not observe any AID following either rejection, or a non-rejecting inflammation, of an intracerebral DAergic graft (Lane et al., 2008). Nonetheless, while GID developed and peaked 6–12 months after grafting in the two NIH-sponsored trials (Freed et al., 2001; Olanow et al., 2003), the peak of GID was only achieved at 24–48 months after grafting in patients that had received long-term triple immunosuppression in the Lund cohort (Piccini et al., 2005). Interestingly, the severity of GID increased in these patients after withdrawal of immunosuppression, suggestive of a delayed immune reaction (Piccini et al., 2005). Post-mortem studies have so far not been performed on any patients with GID, and while it is clear that patients receiving cell suspension grafts with very little infiltration of immune cells seen in post-mortem analysis did not show any GID (Mendez et al., 2005, 2008), these behaviors have not been described in patients that do show some immune reaction (Kordower et al., 1997). Until more con-clusive data is available, we consider it advisable that grafted patients should continue to receive immunosuppression for at least 1 year after trans-plantation, in order to maximize potential benefit from the graft and reduce the risk of GID.

Tissue preparation, graft composition, and surgical protocols

Collection, preparation, and storage of tissue used for transplantation has been diverse between centers (Winkler et al., 2005). Patients in the

Tampa–Mount Sinai trial received grafts of tissue that had been stored in so-called hibernation med-ium for 2 days, and patients in the Denver–Colum-bia trial received grafts after culture of the tissue for up to 4 weeks. Interestingly, the two patients with the highest GID scores in the Lund trials had received tissue that had been hibernated for 1–8 days, whereas the other patients with transplants of fresh tissue displayed lower GID scores (Hagell et al., 2002). Although none of the parameters for collection, preparation, and storage of tissue have so far been correlated with GID development, further standardization of these procedures will be required before transplantation may be safely used in larger numbers of patients. One of the first hypotheses of the cause of GID

was the possibility of a DA “overload”, too many DAergic cells producing an excess of DA into the striatum (Freed et al., 2001). Subsequent 18F-DOPA PET scans did not find excessive DA synthesis and storage in patients with GID (Hagell et al., 2002; Olanow et al., 2009). In one study (Ma et al., 2002) a detailed analysis of the PET scans revealed “hotspots” of 18F-DOPA uptake within the grafted area suggesting a non-homogeneous cell distribution may underlie the emergence of GID. However, examination of the PET scans in the Lund series of patients did not support these findings, with no evidence of excessive or focal DA release (Hagell et al., 2002; Piccini et al., 2005) (Tables 2A and 2B). Animal experiments have also suggested that the number of DA neurons in the graft, whilst important in the reduction of LID, is not directly related to the development of AID or GID (Table 3). Various studies have reported similar levels of AID, despite a range in the order of 2000–17,000 DAergic cells (Carlsson et al., 2006; Lane et al., 2006). Nevertheless, a grafting cannula and microsyringe has been developed which pro-duces a more even distribution of cells throughout the graft area (Mendez et al., 2000b). This tool should ensure that cell distribution in future clinical studies is as homogeneous as possible. The inclusion of different DA cell types in the

transplant composition, i.e., neurons of the A9 and

305

A10 lineage forming the developing substantia nigra and ventral tegmental area, respectively, has been under debate. These two populations have different distributions in the mature graft as observed in rodent transplants, with A9 neurons being located around the periphery with axonal projections into the striatum, whereas the smaller A10 neurons lie more centrally with limited axonal outgrowth beyond the perimeter of the graft (Kuan et al., 2007; Thompson et al., 2005). The number of grafted A9 neurons has been correlated with the degree of both functional recovery and reduction of LID in rodents (Kuan et al., 2007), but negative effects of either the A9 or the A10 cell population within a graft have not yet been observed. Selec-tion of either cell population for grafting is unlikely to be a viable approach for the continuation of fetal cell transplantation. However, these observation do emphasize the importance of ensuring precise pre-differentiation of stem cell populations des-tined for transplantation, not simply into generic DAergic neurons but into the precise regional phe-notype required for appropriate patterns of rein-nervation of their distinct forebrain targets (e.g., Girk2", calbindin, and CCK neurons characteris-tic of A9 nigral neurons).

Can serotonin neurons play a role?

One aspect of graft composition that has been studied in more detail with respect to the develop-ment of GID is the inclusion of serotonin neurons into the transplant. Serotonin neurons develop in close proximity to the DA neurons of the ventral mesencephalon and some contamination of the graft tissue is unavoidable, the degree of which depends on the precise dissection parameters. Since grafts of serotonin neurons have been shown to worsen LID in a rat model of PD (Carlsson et al., 2007, 2008), and as serotonin neurons have been demonstrated in large num-bers in post-mortem analysis of transplants in PD patients (Mendez et al., 2008), the role of seroto-nin in GID has been examined by transplantation

of brainstem raphé neurons and by pharma-cotherapy (Tables 2A and 2B). In one rodent transplant study, grafts contained large numbers of both DA and serotonin neurons, increasing striatal serotonin innervation by 300%, but this was not correlated with the extent of AID observed in these animals (Lane et al., 2006). Another study has shown that even when the number of serotonin neurons exceeds the DA neurons in the graft tenfold, there is no impact on the severity of AIDs development (Garcia et al., 2009). AIDs develop in rats grafted with DA neurons regardless of the presence of sero-tonin neurons (Carlsson et al., 2007; Lane et al., 2009a), suggesting that the development of GID is dependent on the DA component, and not the serotonin component, of the graft. Nevertheless, there is an apparent modulation of AID through serotonergic mechanisms (worsened by 5-HT reuptake inhibition, improved by 5HT1A inhibi-tion) (Lane et al., 2006, 2009a). Lesion of the serotonin-containing host raphé nucleus did not affect the severity of GID in the grafted animals, suggesting that this pharmacological effect is on the grafted tissue. Although the grafted serotonin neurons are

unlikely to be involved in the development of AID in the rodent model, there are some recent observations that point to a possible role in the expression of GID in grafted patients. Politis and coworkers report two patients that had received VM transplants in putamen or in both putamen and caudate nucleus 13 and 16 years earlier under-went PET scanning using a ligand for the seroto-nin transporter (SERT) by 11C-DASB PET (Politis et al., 2010). In these patients UPDRS motor scores in “off” were gradually reduced by about 70%, and from the fourth and eighth year after surgery they no longer needed any dopami-nergic medication and 18F-dopa and 11C-raclo-pride PET showed dopaminergic restoration and DA release within normal levels in the grafted putamen. Both of them, however, developed mod-erately disabling GID over time. 11C-DASB PET revealed excessive binding, 2–3-fold above

306

normal, in the grafted parts of the striatum (puta-men in one case, and putamen and caudate nucleus in the other). Interestingly, the GIDs were almost completely abolished by systemic administration of a 5-HT1A partial agonist, buspir-one, which dampens transmitter release from ser-otonergic neurons, indicating that they were caused by the serotonergic hyperinnervation. These data, consistent with findings in the rat model of AID (Lane et al., 2006), suggest a role of excess serotonin neuron activity in the induc-tion of GID, either by a direct amphetamine-like action of serotonin on the dopaminergic nerve terminals, or by dysregulated release of DA as a “false transmitter” from the serotonergic term-inals. It is known that extracellular DA under certain conditions can be eliminated by re-uptake into serotonergic terminals where it competes with 5-HT for vesicular storage and release (Kannari et al., 2006; Saldana and Barker, 2004; Schmidt and Lovenberg, 1985). Due to lack of normal auto-regulatory feedback the release of DA from sero-tonergic terminals will be dysregulated, with abnormal swings, resulting in GID. Additionally, local injection of serotonin into the striatum has been shown to induce dose-dependent stereotypic behaviors that are mediated by release of endogen-ous DA, probably via reversal of the DA transpor-ter (Jacocks and Cox, 1992; Yeghiayan and Kelley, 1995; Yeghiayan et al., 1997). In the Politis et al. study, therefore, it is proposed that, in areas of serotonin hyper-innervation, these two mechanisms may co-operate to induce excess, dysregulated DA release causing GID, and that the effective block-ade of GID by buspirone is explained by its ability to dampen serotonin neuron activity by activation of the inhibitory 5-HT1A autoreceptors.

Conclusions

The search for the underlying cause of GID is complicated by the fact that we may not be dealing with a single phenomenon. Although differences in design of the clinical trials make inter-trial

comparisons difficult, it seems clear that the GIDs observed in the Denver–Columbia and Tampa–Mount Sinai trials differ with respect to their expression and clinical phenotype, although the time course is approximately similar. Further-more, whilst patients from each trial have come to post-mortem analysis, none of those patients have had any significant level of GID, leaving open questions about graft distribution, cellular compo-sition, and immune/inflammatory response in this subgroup. Experimental studies of GID are ham-pered by the fact that spontaneous, graft-induced AIMs analogous to GIDs in patients are not observed in any of the animal PD models studied so far. Nevertheless, the study of AIDs in 6-OHDA-lesioned rats has provided a useful tool for the study of mechanisms underlying a type of GID that resembles, in both its protracted devel-opment and motoric expression, the dystonic– stereotypic type of GID seen in patients. Despite these shortcomings, we are making significant pro-gress in determining the combination of factors that interact in the development of GID in patients. Today there is growing optimism that avoidance of these abnormal behaviors, and indeed improved functional efficacy, can be achieved by better standardization of the trans-plantation protocol across centers, more wide-spread, homogenous distribution of the graft cell suspension, and careful selection of the most sui-table candidates for DA cell replacement, charac-terized by a good therapeutic response to L-dopa in the absence of any significant or troublesome L-dopa-induced dyskinesia and DAergic denerva-tion which is restricted to the caudate–putamen. A new joint European initiative chaired by Dr Roger Barker (Cambridge University) and funded by the Seventh Framework Program of the European Union, is now underway. This new multi-center clinical trial is designed to re-estab-lish fetal cell transplantation as a safe and effective treatment for PD and thereby opening the way for future clinical trials of alternative stem-cell based therapies that will avoid the practical and ethical problems associated with the use of fetal tissue.

307

Acknowledgments

We would like to thank Bengt Mattsson for his assistance in generating the figures for this manu-script. Our own studies have been funded by the UK Medical Research Council, Parkinson’s Disease Society of Great Britain, Michael J Fox Foundation, Parkinsonfonden, Swedish Research Council, and German Parkinson’s disease Foundation. The authors declare no financial conflicts of interest.

References

Bjorklund, A. (1992). Dopaminergic transplants in experimen-tal parkinsonism: Cellular mechanisms of graft-induced functional recovery. Current Opinion in Neurobiology, 2, 683–689.

Carlsson, T., Carta, M., Munoz, A., Mattsson, B., Winkler, C., Kirik, D., et al. (2008). Impact of grafted serotonin and dopamine neurons on development of L-DOPA-induced dyskinesias in parkinsonian rats is determined by the extent of dopamine neuron degeneration. Brain, 132, 319–335.

Carlsson, T., Carta, M., Winkler, C., Bjorklund, A., & Kirik, D. (2007). Serotonin neuron transplants exacerbate L-DOPA-induced dyskinesias in a rat model of Parkinson’s disease. Journal of Neuroscience, 27, 8011–8022.

Carlsson, T., Winkler, C., Lundblad, M., Cenci, M. A., Bjorklund, A., & Kirik, D. (2006). Graft placement and uneven pattern of reinnervation in the striatum is important for development of graft-induced dyskinesia. Neurobiology of Disease, 21, 657–668.

Cenci, M. A., & Hagell, P. (2006). Dyskinesia and neural graft-ing in Parkinson’s disease. In P. Brundin & W. Olanow (Eds.), Restorative Therapies in Parkinson’s Disease, 184–224.

Cenci, M. A., Lee, C. S., & Bjorklund, A. (1998). L-DOPA-induced dyskinesia in the rat is associated with striatal overexpression of prodynorphin- and glutamic acid decar-boxylase mRNA. European Journal of Neuroscience, 10, 2694–2706.

Cho, C., Alterman, R., Miravite, J., Shils, J., & Taglati, M. (2005). Subthalamic DBS for the treatment of “runaway” dyskinesias after embryonic or fetal tissue transplant. Move-ment Disorder, 20, 1237.

Cochen, V., Ribeiro, M. J., Nguyen, J. P., Gurruchaga, J. M., Villafane, G., Loc’h, C., et al. (2003). Transplantation in Parkinson’s disease: PET changes correlate with the amount of grafted tissue. Movement Disorder, 18, 928–932.

de la Fuente-Fernandez, R., & Stoessl, A. J. (2002). The pla-cebo effect in Parkinson’s disease. Trends in Neurosciences, 25, 302–306.

Defer, G. L., Geny, C., Ricolfi, F., Fenelon, G., Monfort, J. C., Remy, P., et al. (1996). Long-term outcome of unilaterally transplanted parkinsonian patients. I. Clinical approach. Brain, 119(Pt 1), 41–50.

Dekundy, A., Lundblad, M., Danysz, W., & Cenci, M. A. (2007). Modulation of L-DOPA-induced abnormal involun-tary movements by clinically tested compounds: Further validation of the rat dyskinesia model. Behavioural Brain Research, 179, 76–89.

Freed, C., Breeze, R., Fahn, S., & Eidelberg, D. (2004). Pre-operative response to levodopa is the best predictor of trans-plant outcome. Annals of Neurology, 55, 896–379.

Freed, C. R., Greene, P. E., Breeze, R. E., Tsai, W. Y., DuMouchel, W., Kao, R., et al. (2001). Transplantation of embryonic dopamine neurons for severe Parkinson’s disease. New England Journal of Medicine, 344, 710–719.

Garcia, J. A., Carlsson, T., Nikkhah, G., & Winkler, C. (2009). Role of dopamine and serotonin grafts for the expression of L-dopa- and graft-induced dyskinesia. Society for Neu-roscience Abstracts, 430. 10/I14.

Goetz, C. G., Wuu, J., McDermott, M. P., Adler, C. H., Fahn, S., Freed, C. R., et al. (2008). Placebo response in Parkinson’s disease: Comparisons among 11 trials covering medical and surgical interventions. Movement Disorder, 23, 690–699.

Graff-Radford, J., Foote, K. D., Rodriguez, R. L., Fernandez, H. H., Hauser, R. A., Sudhyadhom, A., et al. (2006). Deep brain stimulation of the internal segment of the globus palli-dus in delayed runaway dyskinesia. Archives of Neurology, 63, 1181–1184.

Greene, P., Fahn, S., Tsai, W. Y., Breeze, R., Eidelberg, D., Winfield, H., et al. (1999). Severe spontaneous dyskinesias: A disabling complication of embryonic dopaminergic tissue implants in a subset of transplanted patients with advanced Parkinson’s disease. Movement Disorder, 14, 904.

Hagell, P., & Brundin, P. (2001). Cell survival and clinical outcome following intrastriatal transplantation in Parkinson disease. Journal of Neuropathology and Experimental Neu-rology, 60, 741–752.

Hagell, P., & Cenci, M. A. (2005). Dyskinesias and dopamine cell replacement in Parkinson’s disease: A clinical perspec-tive. Brain Research Bulletin, 68, 4–15.

Hagell, P., Piccini, P., Bjorklund, A., Brundin, P., Rehncrona, S., Widner, H., et al. (2002). Dyskinesias following neural transplantation in Parkinson’s disease. Nature Neuroscience, 5, 627–628.

Hauser, R. A., Freeman, T. B., Snow, B. J., Nauert, M., Gau-ger, L., Kordower, J. H., et al. (1999). Long-term evaluation of bilateral fetal nigral transplantation in Parkinson disease. Archives of Neurology, 56, 179–187.

Herman, J. P., & Abrous, N. D. (1994). Dopaminergic neural grafts after fifteen years: Results and perspectives. Progress in Neurobiology, 44, 1–35.

308

Herman, J. P., Rouge-Pont, F., Le Moal, M., & Abrous, D. N. (1993). Mechanisms of amphetamine-induced rotation in rats with unilateral intrastriatal grafts of embryonic dopaminergic neurons: A pharmacological and biochemical analysis. Neuroscience, 53, 1083–1095.

Herzog, J., Pogarell, O., Pinsker, M. O., Kupsch, A., Oertel, W. H., Lindvall, O., et al. (2008). Deep brain stimulation in Parkinson’s disease following fetal nigral transplantation. Movement Disorder, 23, 1293–1296.

Jacocks, H. M., III&Cox, B. M. (1992). Serotonin-stimulated release of [3H]dopamine via reversal of the dopamine trans-porter in rat striatum and nucleus accumbens: A comparison with release elicited by potassium, N-methyl-D-aspartic acid, glutamic acid and D-amphetamine. Journal of Pharmacology and Experimental Therapeutics, 262, 356–364.

Jacques, D. B., Kopyov, O. V., Eagle, K. S., Carter, T., & Lieberman, A. (1999). Outcomes and complications of fetal tissue transplantation in Parkinson’s disease. Stereotactic and Functional Neurosurgery, 72, 219–224.

Kannari, K., Shen, H., Arai, A., Tomiyama, M., & Baba, M. (2006). Reuptake of L-DOPA-derived extracellular dopa-mine in the striatum with dopaminergic denervation via serotonin transporters. Neuroscience Letters, 402, 62–65.

Kelley, A. E., Lang, C. G., & Gauthier, A. M. (1988). Induction of oral stereotypy following amphetamine microinjection into a discrete subregion of the striatum. Psychopharmacol-ogy, 95, 556–559.

Kopyov, O. V., Jacques, D. S., Lieberman, A., Duma, C. M., &Rogers, R. L. (1997). Outcome following intrastriatal fetal mesencephalic grafts for Parkinson’s patients is directly related to the volume of grafted tissue. Experimental Neurol-ogy, 146, 536–545.

Kordower, J. H., Chu, Y., Hauser, R. A., Freeman, T. B., & Olanow, C. W. (2008a). Lewy body-like pathology in long-term embryonic nigral transplants in Parkinson’s disease. Nature Medicine, 14, 504–506.

Kordower, J. H., Chu, Y., Hauser, R. A., Olanow, C. W., &Freeman, T. B. (2008b). Transplanted dopaminergic neu-rons develop PD pathologic changes: A second case report. Movement Disorder, 23, 2303–2306.

Kordower, J. H., Styren, S., Clarke, M., DeKosky, S. T., Ola-now, C. W., & Freeman, T. B. (1997). Fetal grafting for Parkinson’s disease: Expression of immune markers in two patients with functional fetal nigral implants. Cell Transplant, 6, 213–219.

Kuan, W. L., Lin, R., Tyers, P., & Barker, R. A. (2007). The importance of A9 dopaminergic neurons in mediating the functional benefits of fetal ventral mesencephalon trans-plants and levodopa-induced dyskinesias. Neurobiology of Disease, 25, 594–608.

Lane, E. L., Brundin, P., & Cenci, M. A. (2009a). Ampheta-mine-induced abnormal movements occur independently of both transplant- and host-derived serotonin innervation

following neural grafting in a rat model of Parkinson’s dis-ease. Neurobiology of Disease, 35, 42–51.

Lane, E. L., Soulet, D., Vercammen, L., Cenci, M. A., & Brundin, P. (2008). Neuroinflammation in the generation of post-transplantation dyskinesia in Parkinson’s disease. Neu-robiology of Disease, 32, 220–228.

Lane, E. L., Vercammen, L., Cenci, M. A., & Brundin, P. (2009b). Priming for L-DOPA-induced abnormal involun-tary movements increases the severity of amphetamine-induced dyskinesia in grafted rats. Experimental Neurology, 219, 255–258.

Lane, E. L., Winkler, C., Brundin, P., & Cenci, M. A. (2006). The impact of graft size on the development of dyskinesia following intrastriatal grafting of embryonic dopamine neu-rons in the rat. Neurobiology of Disease, 22, 334–345.

Lee, C. S., Cenci, M. A., Schulzer, M., & Bjorklund, A. (2000). Embryonic ventral mesencephalic grafts improve levodopa-induced dyskinesia in a rat model of Parkinson’s disease. Brain, 123(Pt 7), 1365–1379.

Li, J. Y., Englund, E., Holton, J. L., Soulet, D., Hagell, P., Lees, A. J., et al., (2008). Lewy bodies in grafted neurons in sub-jects with Parkinson’s disease suggest host-to-graft disease propagation. Nature Medicine, 14, 501–503.

Lundblad, M., Andersson, M., Winkler, C., Kirik, D., Wierup, N., & Cenci, M. A. (2002). Pharmacological validation of behavioural measures of akinesia and dyskinesia in a rat model of Parkinson’s disease. The European Journal of Neu-roscience, 15, 120–132.

Ma, Y., Feigin, A., Dhawan, V., Fukuda, M., Shi, Q., Greene, P., et al. (2002). Dyskinesia after fetal cell transplantation for parkinsonism: A PET study. Annals of Neurology, 52, 628–634.

Ma, Y., Tang, C., Chaly, T., Greene, P., Breeze, R., Fahn, S., et al. (2010). Dopamine cell implantation in Parkinson’s disease: Long-term clinical and (18)F-FDOPA PET out-comes. Journal of Nuclear Medicine, 51, 7–15.

Maries, E., Kordower, J. H., Chu, Y., Collier, T. J., Sortwell, C. E., Olaru, E., et al. (2006). Focal not widespread grafts induce novel dyskinetic behavior in parkinsonian rats. Neu-robiology of Disease, 21, 165–180.

Mendez, I., Dagher, A., Hong, M., Hebb, A., Gaudet, P., Law, A., et al. (2000a). Enhancement of survival of stored dopa-minergic cells and promotion of graft survival by exposure of human fetal nigral tissue to glial cell line-derived neuro-trophic factor in patients with Parkinson’s disease. Report of two cases and technical considerations. Journal of Neuro-surgery, 92, 863–869.

Mendez, I., Hong, M., Smith, S., Dagher, A., & Desrosiers, J. (2000b). Neural transplantation cannula and microinjector system: Experimental and clinical experience. Technical note. Journal of Neurosurgery, 92, 493–499.

Mendez, I., Sanchez-Pernaute, R., Cooper, O., Vinuela, A., Ferrari, D., Bjorklund, L., et al. (2005). Cell type analysis of functional fetal dopamine cell suspension transplants in

309

the striatum and substantia nigra of patients with Parkinson’s disease. Brain, 128, 1498–1510.

Mendez, I., Vinuela, A., Astradsson, A., Mukhida, K., Hallett, P., Robertson, H., et al. (2008). Dopamine neurons implanted into people with Parkinson’s disease survive without pathol-ogy for 14 years. Nature Medicine, 14, 507–509.

Olanow, C. W., Goetz, C. G., Kordower, J. H., Stoessl, A. J., Sossi, V., Brin, M. F., et al. (2003). A double-blind controlled trial of bilateral fetal nigral transplantation in Parkinson’s disease. Annals of Neurology, 54, 403–414.

Olanow, C. W., Gracies, J. M., Goetz, C. G., Stoessl, A. J., Freeman, T., Kordower, J. H., et al. (2009). Clinical pattern and risk factors for dyskinesias following fetal nigral trans-plantation in Parkinson’s disease: A double blind video-based analysis. Movement Disorder, 24, 336–343.

Piccini, P., Brooks, D. J., Bjorklund, A., Gunn, R. N., Grasby, P. M., Rimoldi, O., et al. (1999). Dopamine release from nigral transplants visualized in vivo in a Parkinson’s patient. Nature Neuroscience, 2, 1137–1140.

Piccini, P., Lindvall, O., Bjorklund, A., Brundin, P., Hagell, P., Ceravolo, R., et al. (2000). Delayed recovery of movement-related cortical function in Parkinson’s dis-ease after striatal dopaminergic grafts. Annals of Neurol-ogy, 48, 689–695.

Piccini, P., Pavese, N., Hagell, P., Reimer, J., Bjorklund, A., Oertel, W. H., et al. (2005). Factors affecting the clinical outcome after neural transplantation in Parkinson’s disease. Brain, 128, 2977–2986.

Politis, M., Wu, K., Loane, C., Quinn, N., Brooks, D. J., Rehncrona, S., et al. (2010). Serotonergic neurons cause “off”-dyskinesias in Parkinson’s patients with neural trans-plants. Science Translational Medicine, 2, 38–46.

Saldana, S. N., & Barker, E. L. (2004). Temperature and 3,4-methylenedioxymethamphetamine alter human seroto-nin transporter-mediated dopamine uptake. Neuroscience Letters, 354, 209–212.

Schmidt, C. J., & Lovenberg, W. (1985). In vitro demonstration of dopamine uptake by neostriatal serotonergic neurons of the rat. Neuroscience Letters, 59, 9–14.

Soderstrom, K. E., Meredith, G., Freeman, T. B., McGuire, S. O., Collier, T. J., Sortwell, C. E., et al. (2008). The synaptic impact of the host immune response in a parkinsonian allograft rat model: Influence on graft-derived aberrant behaviors. Neurobiology of Disease, 32, 229–242.

Steece-Collier, K., Soderstrom, K. E., Collier, T. J., Sortwell, C. E., & Maries-Lad, E. (2009). Effect of levodopa priming on dopamine neuron transplant efficacy and induction of abnor-mal involuntary movements in parkinsonian rats. The Jour-nal of Comparative Neurology, 515, 15–30.

Thompson, L., Barraud, P., Andersson, E., Kirik, D., & Bjork-lund, A. (2005). Identification of dopaminergic neurons of nigral and ventral tegmental area subtypes in grafts of fetal ventral mesencephalon based on cell morphology, protein expression, and efferent projections. Journal of Neu-roscience, 25, 6467–6477.

Torres, E. M., Dowd, E., & Dunnett, S. B. (2008). Recovery of functional deficits following early donor age ventral mesencephalic grafts in a rat model of Parkinson’s disease. Neuroscience, 154, 631–640.

Torres, E. M., Monville, C., Gates, M. A., Bagga, V., & Dun-nett, S. B. (2007). Improved survival of young donor age dopamine grafts in a rat model of Parkinson’s disease. Neu-roscience, 146, 1606–1617.

Vinuela, A., Hallett, P. J., Reske-Nielsen, C., Patterson, M., Sotnikova, T. D., Caron, M. G., et al. (2008). Implanted reuptake-deficient or wild-type dopaminergic neurons improve ON L-dopa dyskinesias without OFF-dyskinesias in a rat model of Parkinson’s disease. Brain, 131, 3361–3379.

Wenning, G. K., Odin, P., Morrish, P., Rehncrona, S., Widner, H., Brundin, P., et al. (1997). Short- and long-term survival and function of unilateral intrastriatal dopaminergic grafts in Parkinson’s disease. Annals of Neurology, 42, 95–107.

Winkler, C., Garcia, J. A., Carlsson, T., & Nikkhah, G. (2009). Predictive value of severity of preoperative L-dopa-induced dyskinesia for the development of graft-induced dyskinesia in the rat Parkinson model. Society for Neuroscience Abstracts 430.19/I15

Winkler, C., Kirik, D., & Bjorklund, A. (2005). Cell transplan-tation in Parkinson’s disease: How can we make it work? Trends in Neurosciences, 28, 86–92.

Winkler, C., Kirik, D., Bjorklund, A., & Cenci, M. A. (2002). L-DOPA-induced dyskinesia in the intrastriatal 6-hydroxy-dopamine model of Parkinson’s disease: Relation to motor and cellular parameters of nigrostriatal function. Neurobiol-ogy of Disease, 10, 165–186.

Winkler, C., Kirik, D., Bjorklund, A., & Dunnett, S. B. (2000). Transplantation in the rat model of Parkinson’s disease: Ectopic versus homotopic graft placement. Progress in Brain Research, 127, 233–265.

Yeghiayan, S. K., & Kelley, A. E. (1995). Serotonergic stimulation of the ventrolateral striatum induces orofacial stereotypy. Pharmacology Biochemistry and Behavior, 52, 493–501.

Yeghiayan, S. K., Kelley, A. E., Kula, N. S., Campbell, A., & Baldessarini, R. J. (1997). Role of dopamine in behavioral effects of serotonin microinjected into rat striatum. Pharma-cology Biochemistry and Behavior, 56, 251–259.