Dyskinesias and Plasticity
Director: Kathy Steece-Collier, Ph.D.
Dopamine (DA) replacement therapy, in particular levodopa (i.e.: Sinemet®), is currently the gold standard and considered current, best medical therapy for individuals with Parkinson’s disease (PD). However, as long-term levodopa therapy continues, the risk increases for developing the side-effect known as levodopa-induced dyskinesia (LID). LIDs are drug-induced involuntary movements that have a significant, negative impact on quality of life for patients. While numerous alterations in specific brain regions are known to be associated LID, the lack of a mechanism explaining why PD patients develop LIDs has hindered the progress of therapeutics development.
In our laboratory, we developed a dyskinesia rating scale for evaluating LID behaviors in a rat model of parkinsonism (depicted in Figure 1) in conjunction with movement disorder specialist, Roger Kurlan, M.D., at the University of Rochester, NY. Laboratories world-wide have established that LIDs in animal models of PD have strong behavioral, pharmacological (response to drugs) and anatomical similarities to that in PD patients. In our Udall center studies aimed at understanding mechanisms of LID, we use this rat model to examine specific attributes of “striatal pathology” in LID. So what do we mean by “striatal pathology” and what specifically are we investigating?
|Freeze-frame images of LID behavior in a parkinsonian rat showing approximately 3 seconds of real time video-footage following levodopa administration. Note that the right forepaw is hyperkinetic, showing up & down movements (arrows show relative right forepaw position). There is intermittent downward hyperextension of the forelimb across the front of the body with the neck pulling in the opposite direction. Twisting of axial musculature is maintained throughout the array of other excess, abnormal movements.|
In PD there is death of brain cells, or neurons, in a brain region called the substantia nigra (SN). These SN make and use the chemical transmitter dopamine (DA) to communicate with other brain regions involved with voluntary motor function. The primary brain region with which SN DA neurons communicate is the STRIATUM. Thus when SN DA neurons die in PD there is a secondary decrease in DA in the striatum resulting in “striatal pathology”. The neurons of the parkinsonian striatum undergo in numerous changes, some of which are adaptive and beneficial, while other are maladaptive, and are hypothesized to be involved with development of LID behaviors. Our research in the MSU Udall Center brings together a unique combination of researchers with expertise in the areas of PD, applied gene therapy, and maladaptive plasticity to create a team highly qualified to identify and test the therapeutic potential of specific striatal targets that could prevent and/or reverse the devastating side-effect of LID in PD.
|Schematic depiction of “Striatal Pathology”. The above drawings are oversimplifications of particular aspects of the interconnections, or circuitry, of neurons in the striatum of the normal, healthy striatum (a) and in the parkinsonian striatum where some maladaptive changes are depicted (b). There are additional maladaptive changes in the parkinsonian striatum when a graft of new nigral dopamine neurons is transplanted into this brain region. Studies on which these diagrams are based are detailed in Steece-Collier et al., 2012; PMID: 22712056.|
Relevant Publication: Anatomy of Graft-induced Dyskinesias: Circuit Remodeling in the Parkinsonian Striatum. Steece-Collier K, Rademacher DJ, Soderstrom K. Basal Ganglia. 2012;2(1):15-30.
Project Title: “Profiles of Maladaptive Plasticity: Impact on Graft and Levodopa Efficacy”
The primary target of afferent input from nigral DA and cortical glutamate neurons is the medium spiny neuron (MSN) within the striatum. The numerous dendritic spines found on normal MSNs are critical sites for synaptic integration of DA and glutamate signaling, which is essential for normal motor behavior. In advanced PD there is a marked atrophy of dendrites and spines on these neurons. Similar pathology is observed in mice and rats with severe DA depletion. The impact of altered dendritic morphology of MSNs in the parkinsonian striatum on therapeutics such as DA terminal replacement or traditional pharmacotherapy remains equivocal.
|Schematic illustration of the loss of ascending dopamine fibers in the parkinsonian striatum and the accompanying loss of dendritic spines. Images modified from Steece-Collier et al., 2012; PMID: 22712056.|
The mechanism underlying spine loss involves dysregulation of intraspine CaV1.3 calcium channels and pharmacological blockade of the CaV1.3 channels with dyhydropyridines such as nimodipine or isradipine can prevent spine loss despite severe DA depletion (Day et al., 2006; PMID: 16415865). In this Udall Dyskinesia Project, we proposed to undertake series of studies aimed at addressing two primary issues related to spine loss and synaptic plasticity in the parkinsonian brain. The first set of studies proposed to examine whether degenerative changes in spine density of MSN impacted DA terminal replacement strategies using grafted DA neurons as a model to interrogate the ability of new DA terminals to integrate and establish functional connections within the parkinsonian striatum. A second set of studies proposed to examine whether altered spine morphology played a role in the development of levodopa-induced and/or DA graft-induced dyskinetic behavior. In addition, the MSU Center brought all project and core leaders together to examine, in parallel, the gene expression changes of relevance to each project’s thematic research. For the dyskinesia project, we performed a comprehensive gene expression analysis to find new targets to exploit for therapeutic translation.
The hypotheses and findings from these principal studies are presented below.
Hypothesis 1: Loss of dendritic spines on striatal MSNs interferes with the capacity of grafted DA neurons to provide full therapeutic benefit in parkinsonian subjects.
- Parkinsonian rats received a suboptimal number of DA neurons grafted into the striatum; prior to grafting half the subjects received slow-release nimodipine pellets to prevent dendritic spine loss.
- Vibrissae sensorimotor forelimb use and levodopa-induced dyskinesias behaviors both showed significant improvement in rats with preserved striatal dendritic spine density.
- Preventing spine loss also appeared to allow for increased graft reinnervation of striatum.
- Nimodipine treatment in this model did not impact grafted DA neuron survival nor did it have acute pharmacological interaction with these behaviors.
- These results demonstrate that even with grafting suboptimal numbers of cells, maintaining normal spine density on target MSNs results in overall superior behavioral outcome when new DA terminals intervene to remodel the parkinsonian striatum.
|Preserving spine density with nimodipine provided enhanced behavioral benefit of grafted dopamine (DA) neurons, evidenced by the significant reduction in the occurrence of levodopa-induced dyskinesias (LIDs) in allograft + nimodipine group compared to allograft only rats that experience dendritic spine loss at the 18-20 wk post-graft time point (*P = 0.02). Image modified from Soderstrom et al., 2010, PMID: 20105237.|
Publication: Impact of dendritic spine preservation in medium spiny neurons on dopamine graft efficacy and the expression of dyskinesias in parkinsonian rats. Soderstrom KE, O'Malley JA, Levine ND, Sortwell CE, Collier TJ, Steece-Collier K. Eur J Neurosci. 2010;31(3):478-90. PMID: 20105237
Hypothesis 2: Loss of dendritic spines on striatal MSNs is a key factor in the emergence of pathological plasticity in the basal ganglia that is responsible for levodopa-induced dyskinesias (LIDs) in parkinsonian subjects. To test this general hypothesis, we designed two separate experiments:
Experiment 1. We tested the primary hypothesis that loss of dendritic spines is a key step in the emergence of pathological activity that results in the development of LIDs.
- In a separate cohort of animals, run in parallel with the DA grafted rats for “Hypothesis 1”, the impact of dendritic spine preservation on LIDs in non-grafted parkinsonian rats was assessed.
- We found that preventing the loss of striatal dendritic spines allowed for significant buffering against dyskinesia development in severely parkinsonian rats.
|Maintaining dendritic spine density on medium spiny neurons in severely parkinsonian rats with chronic nimodipine pellets significantly delayed the onset of levodopa-induced dyskinesias (LIDs) as evidenced by a significant reduction in the severity of LIDs at early (**P = 0.005) and middle (+P = 0.013) time-points. Adjacent image modified from Soderstrom et al., 2010, PMID: 20105237|
- While spine preservation delayed the onset of LIDs in our model, this benefit was lost overtime with repeated daily, high-dose (12 mg/kg) levodopa.
- Acute pharmacological studies demonstrated no inhibitory or enhancing interaction of acute calcium channel blockade with nimodipine on LIDs.
- These data suggest that the CaV1.3 channel is a potential antidyskinetic target, however, there is reason to suggest that the limitation in scope and loss of protection over time are related to pharmacological limitations of the antagonist drugs tested.
- This Udall funded research has spawned additional funding to continue our investigation into this promising target, i.e.: the striatal CaV1.3 channels. We have now developed a short-hairpin RNA to silence striatal CaV1.3 channels. The shRNA has been cloned into a rAAV genome resulting in a CaV1.3 recombinant adeno-associated viral (rAAV)-shRNA vector that will allow us to test whether continuous, high potency and target-selective, gene-level silencing of striatal CaV1.3 channel can produce lasting antidyskinetic efficacy, devoid of the many pharmacological limitations inherent to all currently available dihydropyridines including nimodipine and isradipine.
|Unilateral injection of CaV1.3 rAAV-shRNA into the rat striatum. Our vectors (CaV1.3 and the scrambled control) are tagged with green fluorescent protein (GFP) thus native GFP fluorescence (green, left hemisphere) can be seen in these unprocessed brains sections.|
Publication: Impact of dendritic spine preservation in medium spiny neurons on dopamine graft efficacy and the expression of dyskinesias in parkinsonian rats. Soderstrom KE, O'Malley JA, Levine ND, Sortwell CE, Collier TJ, Steece-Collier K. Eur J Neurosci. 2010;31(3):478-90. PMID: 20105237.
Experiment 2. We hypothesized that regardless of whether spine loss influences LIDs, dyskinesia indices would be correlated with abnormal spine morphology and/or aberrant synapse formation on MSNs in animals that display these behaviors.
- Using detailed immunoelectron microscopy and light microscopic analyses of Golgi-impregnation neuron reconstruction we showed for the first time how striatal MSNs adapt in subjects with or without LIDs.
- Our data demonstrate dramatic rewiring that involves specifically corticostriatal, but not thalamostriatal contacts.
- Specifically, initial corticostriatal synaptic loss that occurs with following a nigrostriatal DA lesion is followed by a restoration of these excitatory synapses to control levels in dyskinetic rats, however, this restoration involves re-establishment of atypical input patterns onto spines and excessive inputs onto dendrites.
|Schematic diagram of synaptic rewiring in the dyskinetic striatum. Corticostriatal synapses onto a MSN in a (A) control (sham/saline), (B) parkinsonian, L-dopa-naïve (6-OHDA lesion) and (C) parkinsonian, dyskinetic+ (D+) rat. Asymmetric (excitatory) corticostriatal inputs (VGLUT1+) contact spines (boutons with white centers) or dendrites (terminals with grey dots). Some synapses are MSBs (black center), i.e. contacting two spines, each on a different neuron. (B) Illustration of the loss of spines and their synapses following the 6-OHDA lesion (dotted lines). (C) A diagram of the changes the corticostriatal inputs undergo in the D+ animals, which includes a full restoration of synapses onto spines, an aberrant increase in multisynaptic (mushroom) spines (M) and an increase in contacts onto dendrites. Image and legend modified from Zhang et al., 2013, PMID: 23843533|
- We also found evidence that MSNs adapt structurally to re-establishment of synaptic inputs, increasing distal, but not proximal spines; and increasing the number of mushroom spines that accommodate more than one excitatory synapse (which is to our knowledge the first evidence of multi-synaptic input coming onto a single dendritic spine in the striatum).
- Cumulatively these data suggest that the enduring motor dysfunctions and aberrant LTP seen in rats with LIDs are possibly associated with newly sprouted long-range connections that form multiple asymmetric connections with distal spines (presumably mushroom spines), rather than remodeling of existing contacts as MSBs.
- Publication: Aberrant Restoration of Spines and their Synapses in L-DOPA-Induced Dyskinesia: Involvement of Corticostriatal but Not Thalamostriatal Synapses. Zhang Y, Meredith GE, Mendoza-Elias N, Rademacher DJ, Tseng KY, Steece-Collier K. Journal of Neurosci. 2013 Jul 10;33(28):11655-67. PMID: 23843533
- In order to perform studies with Golgi and electron microscopy in the same animals, we also spent extensive effort in designing new methodologies for successful combination of Golgi, EM, and light level immunohistochemistry in adjacent section. Publication: Advances in thin tissue Golgi-Cox impregnation: fast, reliable methods for multi-assay analyses in rodent and non-human primate brain. Levine ND, Rademacher DJ, Collier TJ, O'Malley JA, Kells AP, San Sebastian W, Bankiewicz KS, Steece-Collier K. Journal of Neurosci Methods. 2013; 213(2):214-27; PMID: 23313849.
|The above image shows schematic and micrographs demonstrating the approach and feasibility of performing multiple postmortem analyses in the same rat brain. Rats were perfused with 3.75% acrolein in 2% PF to allow adjacent sections to be processed for thin-tissue Golgi impregnation and electron microscopic visualization central nervous system ultrastructure. Adjacent image modified from Levine et al., 2013, PMID: 23313849.|
Hypothesis 3: The loss of spines on MSNs in the parkinsonian striatum promotes aberrant graft-host connections that underlie the development of graft-induced dyskinesias (GID) noted in many PD graft recipients and rodents with embryonic DA neuron grafts.
- In the same cohort of parkinsonian rats used in “Hypothesis 1” (above), we found that rats with preserved dendritic spine density showed a delay in the induction of GID-like behaviors suggesting a role for dendritic spine loss in the development of GID.
|Spine density preservation resulted in a transient improvement in the expression of the GID, forelimb tapping dyskinesia. Continuous nimodipine treatment resulted in a significant decrease in the severity score of TPD in dopamine-grafted rats (DA graft + nimodipine) when compared with rats receiving dopamine grafts alone (DA graft, *P = 0.04) at a middle time-point post-grafting; however, this effect was lost at later post-graft time-points (P = 0.191). DA= dopamine.|
- This was in keeping with our previous findings (Soderstrom et al., 2008; PMID: 18672063) that demonstrated as the number of typical axo-spinous synapses between grafted and host cells decreased, and the number of atypical axo-dendritic increased, the expression of aberrant GID behavior increased.
- Unfortunately, there was a gradual escalation of GID, albeit more protracted in parkinsonian rats with preserved spine density.
- While the mechanism(s) responsible for the gradual re-emergence of GID in this study is unknown, our previous work (Soderstrom et al., 2008; PMID: 18672063) suggests that additional synaptic changes independent of the state of MSN spine integrity may be playing a role.
Publication: Impact of dendritic spine preservation in medium spiny neurons on dopamine graft efficacy and the expression of dyskinesias in parkinsonian rats. Soderstrom KE, O'Malley JA, Levine ND, Sortwell CE, Collier TJ, Steece-Collier K. Eur J Neurosci. 2010 Feb;31(3):478-90. Epub 2010 Jan 25. PMID: 20105237
Hypothesis 4: Normalizing intraspine calcium with nimodipine in severely DA-depleted animals after the initial spine loss has occurred will allow for recovery of spine density in young subjects where plasticity is high, but not in those of advanced age, and this accounts for suboptimal benefit of DA terminal replacement therapy in aged parkinsonian subjects.
- In this Aim, we asked the question of whether the parkinsonian striatum could be “remodeled” or normalized with regard to MSN spine density with calcium channel antagonism, and if so, whether this would be refractory in the aged brain.
- Counter to our hypothesis, we found that dendritic spine loss in severely parkinsonian rats is reversible irrespective of host age; young = 3 mo and aged = 22 month old at time of lesion.
|Dendritic spine loss and recover in young and aged rats. Rats were rendered parkinsonian and left untreated for 3 weeks, a time sufficient for dendritic spine loss. A subset of rats was sacrificed to confirm spine loss, and the remainder of the rats received either slow release nimodipine or vehicle pellets for 21 days.|
- We also found that the agednormal rats contain significantly fewer spines that their young counter part, suggesting that the aged parkinsonian brain has significantly more morphological pathology than the young parkinsonian brain.
- Further, these studies suggest that age-related striatal spine loss does not appear to be responsive to CaV1.3 calcium channel modulation.
- In a follow-up study, we found that there appears to additional dysfunction of remaining spines in the aged brain because even when 5x as many cells are grafted into the aged striatum, half the number of synaptic contacts were apparent compared to young brain; and this anatomical finding correlated with inferior behavioral recovery in aged subjects compared to young (Collier et al., in preparation).
|To help understand the capacity of the aged, parkinsonian striatum to be remodeled with new DA terminals, we used a grafting model and examined whether increasing the number of grafted DA neurons would translate to enhanced behavioral recovery. Young (3 mo), middle-aged (15 mo), and aged (22 mo) parkinsonian rats were grafted with proportionately increasing numbers of embryonic ventral mesencephalic (VM) cells to evaluate whether the limitations of the graft environment in subjects of advancing age can be offset by increased numbers of transplanted neurons. Despite robust survival of grafted neurons in aged rats, reinnervation of the striatum remained inferior and behavioral recovery delayed or absent. This study demonstrates that: 1) counter to previous evidence, under certain conditions the aged striatum can support robust survival of grafted DA neurons; and 2) factors inherent in the aging host that inhibit integration of graft and host continue to present obstacles to full therapeutic efficacy of DA cell-based therapy.|
Publication: Enhancing the number of dopamine neurons grafted into the aged parkinsonian host does not allow for enhanced graft efficacy. Timothy J. Collier, Jennifer O’Malley, Caryl E. Sortwell, Katrina Paumier, Kibrom Gebre-Egziabher, Kathy Steece-Collier (in preparation)
Gene array: A secondary goal of the Udall Dyskinesia Project was to use these studies to identify novel targets that could be manipulated to minimize or prevent LIDs.
In this Udall Dyskinesia Project we employed comprehensive gene microarray analyses, followed by qPCR confirmation, to examine differential transcript expression between striata of 6-OHDA lesioned, parkinsonian rats that developed dyskinesias (LID+) and those that did not (LID-), both having undergone chronic daily exposure to high dose (12mg/kg) levodopa to find potential causes for LIDs. This study capitalized on the observation that levodopa administration in parkinsonian humans and rodents causes LIDs in only a subset of subjects. Over 80 genes were differentially expressed in LID+ vs LID- striatum. Three novel genes were found to be highly upregulated (>30-fold) in the LID+ striatum compared to the LID- striatum. We have developed overexpression and silencing vectors to begin to explore the utility of two of these genes as antidyskinetic targets for therapeutic develop in patients.
Udall Projects & Services
Director: Kathy Steece-Collier, Ph.D.
Director: Caryl E. Sortwell, Ph.D.
Director: Jack W. Lipton, Ph.D.
Director: Timothy J. Collier, Ph.D
Director: Fredric P. Manfredsson, Ph.D.
Director: Timothy J. Collier, Ph.D