MPTP Parkinsonism and Implications for Understanding Parkinson's Disease

Langston JW, Ballard P, Tetrud JW, Irwin I. Chronic parkinsonism in humans due to a product of meperidine‐analog synthesis. Science 1983;219:979–980.

In 1983, Langston et al first reported chronic parkinsonism in individuals following self‐injection of intravenous heroin contaminated with 1‐methyl‐4‐phenyl‐1,2,3,6‐tetrahydropyridine (MPTP). ¹ This astute clinical observation and subsequent work led to the discovery of MPTP as a dopaminergic neurotoxin that resulted in the development of the most validated and successful preclinical model of any neurological disease. The MPTP primate and subsequent MPTP rodent models have enabled a better understanding of basal ganglia circuitry and revolutionized translational drug discoveries and surgical treatments for Parkinson's disease (PD). In addition, knowledge of the conversion of MPTP to 1‐methyl‐4‐phenylpyridinium (MPP+) via monoamine‐oxidase (MAO) B (MAO‐B) inhibition opened up the notion of environmental toxins as potential causes of PD as well as the role of mitochondria in PD pathophysiology.

MPTP‐Induced Parkinsonism: Clinical Clues?

The original 4 cases were individuals from northern California who had recently injected a new synthetic form of heroin. Within a week, all individuals experienced jerking of the limbs, muscle stiffness, and visual hallucinations followed by progressive difficulty with mobility and generalized slowness over a few weeks. The clinical features were noted to be strikingly similar to a severe presentation of PD, including reduced eye blinking, sialorrhea, facial seborrhea, akinesia and freezing of gait, flexed posture, and cogwheel rigidity. ¹ There was a significant improvement with levodopa (l‐dopa), and 2 patients had an additional benefit from dopamine agonists and continued to be dependent on these medications for symptom control with no remission. Early onset of fluctuations with on–off responses to l‐dopa and dyskinesia were also reported. ² Subsequent analysis of the heroin showed contamination with MPTP. ¹ Thus, the subacute onset of an akinetic‐rigid syndrome, following exposure to MPTP, led to the seminal description of MPTP‐induced parkinsonism.

The presence of bilateral clinical features and the subacute onset suggested that the parkinsonism was likely secondary to a toxin, which is in keeping with other drugs that have been described to cause acute parkinsonism, such as manganese exposure, or dopamine antagonists. Moreover, “secondary forms of parkinsonism” are often associated not only with pure akinesia but also include additional clinical features. For example, in postencephalitic parkinsonism, oculogyric crisis and early psychosis were often reported. For a contemporary context, the ongoing COVID‐19 pandemic has led to concerns about parkinsonism as a consequence of exposure to severe acute respiratory syndrome coronavirus 2. To date, the handful of cases described with this infection are of bilateral parkinsonism. ³

The benefit of l‐dopa on MPTP parkinsonism and the presence of early motor fluctuations and dyskinesia also offered clinical clues to understanding basal ganglia pathophysiology. This effect indicated functioning postsynaptic striatal dopamine receptors, unlike some toxic insults that cause parkinsonism with striatal lesions when l‐dopa is less or not effective. The rapid and severe degree of dopaminergic denervation at the time of l‐dopa initiation also accounted for the relatively faster onset of fluctuations when compared with idiopathic PD. This is similar to the classical description of postencephalitic parkinsonism in Sacks’ Awakenings. ⁴ The lesson is that dyskinesia can occur on early exposure to l‐dopa if the degree of nigrostriatal dopamine denervation is severe enough to trigger striatal changes in dopamine receptor activation. This is commonly seen in the MPTP primate models when dyskinesia can be induced on first exposure to l‐dopa due to the >90% nigral dopamine cell loss. ⁵ These clinical observations thus show that the pathophysiology of l‐dopa‐induced dyskinesia may not just be attributed to chronic l‐dopa exposure (so‐called “priming”) but also to the degree of nigrostriatal presynaptic dopamine cell loss.

The other interesting aspect is that despite the similarity of clinical features to idiopathic PD, pathological study of 3 subjects with MPTP‐induced parkinsonism revealed no Lewy bodies but, rather, evidence of inflammation. ⁶ To date, there have been no further cases reported in the literature with pathological examination.

MPTP Selective Targeting of Substantia Nigra Dopaminergic Neurons: Understanding Mechanisms of Neurotoxicity

MPTP neurotoxicity is mediated by the metabolite MPP+ produced from MPTP after crossing the blood–brain barrier. MPTP is then converted to MPP+ via MAO‐B in astrocytes and serotonergic neurons. Subsequently, MPP+ is selectively taken up into dopaminergic neurons of the substantia nigra pars compacta (SNpc) via the dopamine transporter (DAT). ⁷ Studies of mice models lacking DAT showed complete protection of dopaminergic neurons against MPTP toxicity. ⁸ In keeping with the pattern of dopamine cell loss in PD, the susceptibility of dopaminergic neurons to MPTP is predominantly SNpc dopaminergic neurons. Also, similar to PD, MPTP induces a more severe loss of dopaminergic innervation in the dorsal striatum that is targeted mainly by ventral SNpc neurons than in the ventral striatum, which is targeted by ventral tegmental area neurons. Thus, the pattern of dopamine loss induced by MPTP has enabled a better understanding of the very selective mechanism whereby dopamine signaling leads to the generation of clinical signs and the relative sensitivity of these basal ganglia circuits to neurotoxicity.

Other catecholamines are implicated in the pathology of PD, including noradrenaline (norepinephrine) and serotonin (5HT). MPP+ can affect noradrenergic neurons, and potentially serotonergic neurons, via binding and uptake to the norepinephrine transporter and serotonin transporter, respectively, but to a lesser degree than DAT. ⁹ A double hit model using the serotonergic toxin 3,4‐methylenedioxymethamphetamine combined with MPTP has been used to model 5HT loss combined with dopamine loss. ¹⁰ Therefore, the MPTP models have also allowed insights into nondopaminergic mechanisms underlying PD. ¹¹

Mitochondria and MPTP: Another Clue to the Pathogenesis of PD

Identification of the key role that mitochondria play in the pathophysiology of PD was also a discovery from the mechanism of action of MPP+. Once in the neuron, MPP+ is concentrated in the mitochondria, leading to inhibition of complex I of the respiratory chain, with impairment of adenosine triphosphate production. ⁷ The resulting oxidative stress, along with α‐synuclein aggregation, causes early and progressive loss of vesicular monoamine transporter‐2, ¹² which leads to the degeneration of dopaminergic neurons. ⁷ Subsequent data have demonstrated coenzyme Q10 deficiency in PD; however, the exact relationship between this and complex I deficiency in causing nigral dopamine cell loss remains unclear. ¹³ Recent studies have also confirmed mitochondrial dysfunction in PD attributed to genetic causes such as PTEN Induced Kinase 1 (PINK1), Protein deglycase DJ‐1 (DJ‐1, α‐synuclein, Leucine Rich Repeat Kinase 2 (LRRK2), and parkin mutations. ¹⁴ These findings have led to the investigation of therapies potentially associated with the reduction of mitochondrial injury, such as coenzyme Q10 and creatine replacement. Nevertheless, such agents have not been found to provide any significant symptomatic benefit in large clinical trials. ¹⁵ , ¹⁶

Environmental Toxins and PD

The discovery of MPTP also shifted the understanding of PD pathophysiology to reconsider the role of environmental toxins. Thus, the pesticides paraquat and rotenone have a similar structure to MPTP. Paraquat impairs the protective mechanisms against oxidative stress by undergoing redox cycling and producing superoxides, ¹⁷ whereas rotenone inhibits mitochondrial complex I. ¹⁸ Recently, large epidemiological studies have reported parkinsonism after exposure to pesticides. ¹⁹ This possible association between the development of PD and pesticide exposure has been widely reported in the media, partly related to the recent reapproval of several pesticides, including paraquat, by the Environmental Protection Agency, which has been banned or is being phased out in other large agricultural nations. ²⁰ As a result, the discovery of MPTP led to advances in the understanding of mechanisms involved in dopaminergic impairment. Specifically, it has provided strong evidence to the leading hypothesis that PD results from a cascade of processes triggered by a complex interaction between genetics and environmental factors.

The MPTP Model of Parkinsonism Is the Gold Standard Preclinical Model

An important outcome of discovering MPTP was the generation of an animal model of parkinsonism. The MPTP model has been key to understanding the pathophysiology of PD, as well as an essential tool for drug development in PD. The MPTP model is now also used to study different stages of PD, from presymptomatic to advanced motor complications as well as nonmotor symptoms. ¹¹

The first reports of an MPTP primate model of PD were published in 1983. ²¹ , ²² Administration of MPTP to macaques, squirrel monkeys, or common marmosets resulted in clinical motor features of parkinsonism that developed over a few weeks and then stabilized. In primates, the phenotype of bradykinesia is usually seen as reduced spontaneous body movements, slowness of movements, and diminished exploratory behavior plus rigidity (if able to assess), a flexed posture, and poor balance. The parkinsonian features respond to l‐dopa, and animals develop l‐dopa‐induced dyskinesia, similar to that seen in humans. ²³ , ²⁴ Nonmotor features of PD, such as drooling, sleep disturbances, bladder and gastrointestinal dysfunction, cognitive impairment, and psychosis‐like behaviors to medications, have been also been described. ¹¹ , ²⁵ The clinical features of the MPTP primate have excellent face validity as a model of PD, and it remains a powerful tool in preclinical studies of symptomatic therapeutics for PD.

One criticism is that MPTP models an advanced stage of PD, with more than 90% dopaminergic cell loss. The classic model is generated by MPTP administered for 5 days with a recovery period, then gradual development of parkinsonism over 2 to 3 months followed by stabilization. ²⁶ The observed parkinsonism lasts several months, but some animals may recover. Moreover, there is no progressive neurodegenerative process. To develop an earlier stage of parkinsonism, possibly with progression, lower doses and more chronic dosing schedules for 2 to 3 weeks have been attempted. ²⁷ To date there are no conclusive therapeutics that have been demonstrated to slow progression using this modified MPTP primate model. Other challenges with the bilateral MPTP model have been caring for the animals immediately post MPTP. Thus, a hemiparkinsonism model was developed by a single‐dose infusion of MPTP into the carotid artery ²⁸ ; this is less symptomatic and easier to maintain, and the unaffected side can act as a control. However, there has not been widespread adoption of this model.

Logistic and ethical challenges with primates have led to the development of MPTP rodent models. Rats are relatively resistant to MPTP because of the presence of MAO in the blood–brain barrier, which oxidizes MPTP to MPP+ outside the brain. ²⁹ Mice appear to be more sensitive to MPTP, and the first MPTP mouse model was initially described in 1984. ³⁰ Some mouse model strains also appear to be resistant to MPTP administration, which points to the possibility that such responses could be related to differences in MAO‐B levels and that underlying genetic factors may play a large role in the development of symptoms. ³¹ Other models have used MPTP combined with transgenics, which are better defined as “model fusion” to allow exploration of genetic and environmental interaction. ³²

In addition, the understanding of neural mechanisms underlying parkinsonism and l‐dopa‐induced dyskinesia came from the MPTP model. Thus, the classic basal‐ganglia‐thalamocortical circuits described by Alexander et al ³³ were developed from studies using the MPTP primate. The seminal work by Crossman et al ³⁴ and Delong et al ³⁵ demonstrated the interconnections of the striatopallidal GABAergic “direct” and “indirect” pathways and outputs from the basal ganglia. As a consequence, the report of subthalamic nucleus (STN) overactivation led to the development of STN deep brain stimulation (DBS) as a therapeutic target for PD. ³⁶ Moreover, other advances have been possible, such as the understanding of dopamine receptor signaling with enhanced D1/D3 receptor changes and the roles of many nondopaminergic circuits involved in the pathophysiology of dyskinesia. ³⁷

The MPTP model has also been essential for the evaluation of all currently used PD therapeutics. Thus, amantadine, dopamine agonists, and l‐dopa formulations were all assessed in the MPTP model. ³⁸ Despite these successes, challenges still exist. ³⁹ As discussed, the model replicates an advanced stage of the disease, which is often not the main target for therapeutic initiation. Logistics usually preclude chronic treatment in the model, so interventions are usually evaluated as single‐dose acute studies. Recent attempts have been made to improve the translational benefit from primates. One strategy included the comparison of animal rating scales to humans to determine a “clinically relevant difference.” The results suggested that drugs administered as monotherapy with a ≥50% reduction of global parkinsonism in the MPTP primate may be necessary to predict the possibility of clinical efficacy in patients with PD. In the specific case of dyskinesia, the predictability threshold should be set at ≥25% reduction of peak‐dose dyskinesia. ⁴⁰ Tolerability also clearly differs in human disease and has been a frequent cause of failure at subsequent phase 2 level clinical trials.

MPTP Model: Lack of Pathology for Neuroprotection Strategies?

For evaluating neuroprotective targets, the MPTP model appears to be less useful. The model results in a rapid and acute degeneration that does not accurately reflect the slow temporal progression of symptoms in PD. Even with chronic administration of MPTP over weeks, the death of neurons occurs relatively quickly and is unlikely to mimic the natural course of disease in humans. In addition, common methods of administration of MPTP lack α‐synuclein aggregation or Lewy bodies formation. ⁴¹ As discussed previously, this is in keeping with the pathological evidence from the few cases reported in humans. Some aged MPTP‐treated primates have been reported to develop intraneuronal inclusions in the SNpc, locus coeruleus, nucleus basalis of Meynert, dorsal motor nucleus of the vagus, and raphe nucleus, but the relevance to human disease is unclear. ⁴¹ Recently, α‐synuclein aggregation in midbrain dopaminergic neurons and dystrophic nigrostriatal axons have been described in chronically MPTP‐treated young adult squirrels and rhesus monkeys. ⁴² Nevertheless, these models have not yet been widely used for the evaluation of potential neuroprotection.

Another obstacle associated with the MPTP model is that male animals are predominantly used. This relates to the logistics of using mixed colonies or potentially pregnant female animals. Previous data suggested that estrogen might be protective against MPTP and thus cause variability in the extent of dopamine cell loss. ⁴³ This also means that the model might only be reflective of male PD. Recent findings have also reported significant sex‐related differences in the disease progression and potentially in medication response variability. ⁴⁴

Given the limitations for modeling disease progression over time, the MPTP model is unlikely to remain a useful tool in future neuroprotective studies. Indeed, the early widespread application of the MPTP model might partially explain the current lack of effective therapeutic agents. As noted previously, despite effective preclinical studies, clinical trials using the MAO‐B inhibitors selegiline and rasagiline were “inconclusive” as neuroprotective agents, as determined by evidence‐based medicine criteria, and to date no potential neuroprotective therapy has evidence for benefit. ⁴⁵ Other failures included glial cell line–derived neurotrophic factor infusion. ⁴⁶ More recent developments in animal models are thus now turning to transgenic models (α‐synuclein) for neurodegeneration studies. ⁴⁷ The key question in choosing the preclinical model to use is what is the outcome of the intervention? For symptomatic efficacy, the MPTP model is superior; for neuroprotection, alternative models are likely preferable. ⁴⁸

Preventing Dopamine Cell Death via MAO‐B Inhibitors: Lessons in Clinical Trial Design?

The discovery of MAO‐B as the enzyme facilitating the oxidation of MPTP initiated a series of preclinical and clinical studies exploring the efficacy of MAO‐B inhibitors as neuroprotective agents. Although proven in MPTP primate models, ⁴⁹ there was a failure to show neuroprotection (as clinically measured) in human studies. ⁵⁰ , ⁵¹ This was partly caused by the additional action of MAO‐B inhibitors in enhancing dopamine levels with a symptomatic effect that was difficult to separate from a purported neuroprotective effect in clinical trials. However, the lessons learned in novel trials facilitated the differentiation of these 2 endpoints, including prolonged wash‐out and delayed‐start designs. Such strategies changed the field of clinical trials in PD, allowing the possibility of more successful trial outcomes in the future.

Conclusions

The seminal paper and clinical observations by Langston and colleagues in 1983 drastically impacted the PD field. Besides leading to innumerous advances in the understanding of the dopaminergic system and PD pathophysiology, the publication has also provided the basis for the development of widely used animal models in research. To date, such insights resulted in the investigation of MAO‐B inhibitors for PD as well as a better understanding of the basal ganglia circuitry, which has subsequently led to the development of DBS. Moreover, the discovery that MPTP directly damaged the nigrostriatal pathway played a key role in igniting the search for environmental factors that could lead to the development of PD. Although the movement disorders field continues to move forward by questioning PD as a single entity as well as focusing on genetic models and clinical trials development, the MPTP model remains an important and complementary tool to provide pivotal preclinical data.

Author Roles

(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Manuscript Preparation: A. Writing of the First Draft, B. Review and Critique.

S.A.: 1A, 1B, 1C, 2A, 2B

D.G.D.L.: 1A, 1B, 1C, 2A, 2B

S.H.F.: 1A, 1B, 1C, 2B

Disclosures

Ethical Compliance Statement

We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines. This paper did not require the approval of an institutional review board, and informed patient consent was not necessary.

Funding Sources and Conflicts of Interest

There were no funding sources for this paper and no conflicts of interest.

Financial Disclosures for the Previous 12 Months

Susan H. Fox received clinic support from the Edmond J. Safra Foundation for Parkinson Research, National Parkinson Foundation, and the Toronto Western and General Foundation; salary from UHN Department of Medicine Practice Plan; research funding from The Michael J. Fox Foundation for Parkinson Research, National Institutes of Health (Dystonia Coalition), CIHR, and Parkinson Canada; and honoraria from the International Parkinson and Movement Disorder Society and American Academy of Neurology. She is the site principal investigator for clinical trials for Biotie, Cynapsus, Eisai, and Revance and has received consultancy/speaker fees from Acadia, Atuka, Sunovion, Teva, and Paladin and royalties from Oxford University Press. Sohaila AlShimemeri and Daniel G. Di Luca have no financial disclosures to declare.