Journal of Neurology

, Volume 258, Supplement 2, pp 299–306

Basic science in Parkinson’s disease: its impact on clinical practice

Authors

    • Department of Neurology, University HospitalRWTH Aachen University
    • JARA Brain
  • Manfred Gerlach
    • Department for Child and Adolescent Psychiatry, Psychosomatics and PsychotherapyUniversity of Würzburg
  • Gabriele Gille
    • Department of NeurologyTU Dresden
  • Wilfried Kuhn
    • Department of NeurologyLeopoldina-Krankenhaus
  • Martina Müngersdorf
    • Neurologie am Hackeschen Markt
  • Peter Riederer
    • Clinic and Policlinic of Psychiatry, Psychosomatics and PsychotherapyUniversity of Würzburg
  • Martin Südmeyer
    • Department of NeurologyUniversity of Düsseldorf
    • Department of NeurolgyUniversity of Ulm
Article

DOI: 10.1007/s00415-011-6040-y

Cite this article as:
Schulz, J.B., Gerlach, M., Gille, G. et al. J Neurol (2011) 258: 299. doi:10.1007/s00415-011-6040-y

Abstract

Failures in clinical studies that were aimed to prove disease-modifying effects of treatments in Parkinson’s disease (PD) raise the question as to whether basic sciences have had an impact in clinical practice. This question implies that despite well-publicized results obtained by intensive genetic and pathogenetic research, e.g. the identification of mutations and cellular biochemical pathways that underlie Parkinson-specific neurodegeneration, no relevant disease-modifying treatment options have been developed. This view neglects the fact that today there are plenty of dopaminergic and non-dopaminergic and surgical treatment options, and that PD was not treatable 50 years ago. This progress was made possible only by basic science. In this review, we underline the success of previous basic science for daily practice in PD and its impact for the understanding and development of an early diagnosis. Early, even pre-symptomatic diagnosis might be key to successfully establish disease-modifying treatments.

Keywords

Parkinson’s diseaseBasic scienceHistoryTreatmentGenetics

Introduction

“Until we are better informed respecting the nature of the disease, the employment of internal medicine is scarcely warrantable.”—J. Parkinson [53].

When first describing what was later termed Parkinson's disease (PD), James Parkinson called for a better understanding of the pathophysiology for the development of therapies. The question now arises whether basic science has fulfilled the promises to contribute to the understanding and treatment of PD.

Initially, only empiric treatments were available. In the 1880s Charcot used belladonna alkaloids for treatment. Treatment with belladonna alkaloids followed the clinical observation of Ordenstein, in 1876, that treatment of hypersalivation with Atropa belladonna in Parkinson's disease patients also had positive effects on their tremor [52]. Trihexyphenidyl was introduced in 1950 for the treatment of PD and is still used today if the tremor does not respond to dopamimetic treatment. In 1968 Schwab recognized an improvement of Parkinsonian symptoms in a patient with PD who received the anti-influenza prophylaxis amantadine. In 1969, he published his treatment results of 153 patients; two-third of his patients showed benefit from amantadine treatment [63].

In contrast to these empirical treatment advances, the introduction of levodopa was well supported through a basic science rationale. Levodopa was introduced after Carlsson and colleagues had shown that the striatum contained high concentrations of dopamine [14]. Later, Ehringer and Hornykiewicz observed in 1960 that striatal dopamine depletion is the major biochemical alteration in the brain of PD patients [25]. Independently, I. Sano demonstrated a striatal deficiency of dopamine in a single case of Parkinson's disease and tried DL-DOPA as early as 1960 (although he did not believe in the success of the DOPA-therapy in later years—in contrast to the Viennese group).

The natural course of disease progression before the introduction of levodopa therapy showed a broad variation; however, in primary Parkinsonism, time to death was 9.4 years [33]. The mortality risk was increased to 2.6 in men and 3.8 in women. After the introduction of levodopa, dopamimetics, MAO-B- and COMT-inhibition and deep brain stimulation, the natural course has changed. The mortality of PD patients has decreased and now ranges from 1.0 to 3.4 [36]. The still increased mortality is mainly attributed to non-dopaminergic symptoms including dementia, postural instability and dysphagia [19, 32].

Neurochemistry and neuropharmacology of PD

Identification of the underlying biochemical alterations was essential for the development of a potent symptomatic treatment of PD. Subsequent to the isolation of 3,4-dihydroxyphenylalanine (l-DOPA) from pods and seeds of Vicia faba, broad beans, in 1911 by Torquato Torquati, the successful synthesis of dl-DOPA by Funk [28] in the same year and the exploration of this aromatic amino acid’s metabolism via its decarboxylation to dopamine (DA) [34] and the monoamines oxidative deamination (MAO; discovered and named “amine oxidase” by Blaschko [8] but already investigated in the 1920th by M. Hare-Bernheim), translation to their physiological role became evident in studies related to actions of l-DOPA and DA on cardiovascular functions. In the 1950s experimental studies with reserpine demonstrated its monoamines depleting potential by blockade of respective vesicular stores. Reserpine was shown to be useful in reducing sympathetic function, thus controlling hypertension. It was “sedative,” especially in psychiatric patients (schizophrenia), and led to parkinsonian-like impairment of motor performance. Carlsson et al. [12] in 1957 described the dl-DOPA induced antagonism of reserpine-induced sedation in mice and rabbits. Subsequent publications demonstrated DA to be a neurotransmitter localized in specific pathways in the brains. The discovery of a significant reduction of DA in the striatum of patients with PD [25, 58] and the description of short-lasting beneficial effects of dl-DOPA [58] and l-DOPA [5] led Birkmayer and Hornykiewicz to suggest combining l-DOPA with drugs that would reduce the metabolism of l-DOPA/DA. Carlsson et al. [12] had already shown that iproniazid, a MAO-inhibitor, potentiated the effect of DOPA. Sano 1960 used various MAO-inhibitors in PD. Birkmayer and Hornykiewicz [4] described the clinical effect of various MAO-inhibitors in PD. They also mentioned adverse reactions.

Liver toxicity and hypertensive crisis were the main arguments against using MAO-inhibitors in neuropsychiatric patients. Only the discovery of MAO-isoforms MAO-A and MAO-B in 1968 by Johnston [37] and the synthesis and pharmacological distribution of a selective irreversible MAO-B inhibitor, selegiline, by Hungarian scientists [40], led Peter Riederer to suggest treating patients with PD [6].

As DA is metabolized not only by MAO but also via catechol-O-methyltransferase (COMT), [13] suggested inhibiting this enzyme in PD in order to potentiate the effects of l-DOPA. This resulted only in the 1990s in the marketing of entacapone and tolcapone as adjuvant therapies of levodopa.

Once discovered, dopamine receptor agonists were been recognized as a treatment strategy for PD in extension. However, bulbocapnine (1920s) and apomorphine [62] were used much earlier and without knowledge of their DA receptor activity. It was then Corrodi et al. [17] discovered that bromocriptine exerts substantial effects on central catecholamines. Calne et al. [11] were the first to demonstrate an anti-PD activity of bromocriptine. Many more DA-receptor agonists with different D1/D2 receptor subtype profiling followed, showing (1) significant locomotor inducing action in PD, and (2) being anti-dyskinetic in Levodopa induced dyskinetic patients.

Based on neurochemical [43] and electrophysiological [9] findings, aminoadamantanes, like amantadine and memantine, have been identified as glutamate receptor channel antagonists. Since then such compounds, as well as direct antagonists of glutamate receptor subtypes, have been developed in order to antagonize excitotoxicity in PD. However, it is noteworthy that amantadine, with its antiviral potency, was shown to have antiakinetic effects in PD long before the above mentioned discovery [63].

The MPTP model and its relevance on the clinical practise

During the period 1979–1982, observations were made on a number of young drug-dependent Californians who had injected a new “synthetic heroin” and had developed a serious and irreversible Parkinson syndrome [18]. These patients exhibited all the symptoms typical of idiopathic PD and responded well to treatment with l-DOPA and dopamine-receptor agonists. Analysis of the “synthetic heroin” showed it contained not only about 25% of the actual active agent 1-methyl-4-phenyl-4-propionoxypiperidine, but also up to 2.9% 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [44].

MPTP was shown to cause a selective destruction of nigro-striatal dopaminergic neurons in both human and non-human primates as well as in various other mammalian species. This stimulated a great deal of work on the mechanisms involved in its neurotoxic effects and their possible relationship to PD [29]. The mechanisms underlying the selective neurotoxicity of MPTP involve the conversion by the action of the B form of monoamine oxidase (MAO-B) to 1-methyl-4-phenylpyridinium ion (MPP+), which is the effective neurotoxin. This compound is taken up into mitochondria of dopaminergic neurons, and then disrupts oxidative phosphorylation by inhibiting complex I of the electron transport chain. It also leads to increased production of reactive oxygen species, release of calcium and iron, and programmed cell death.

The mechanisms involved in the neurotoxic effects of MPTP were the rationale for the development of disease-modifying (previously named neuroprotective) treatment strategies of PD. These strategies include the use of MAO-B inhibitors (selegiline, rasagiline), radical scavengers (α-tocopheroal, coenzyme Q10) iron chelators (desferal), anti-apoptotic compounds (minocyclin, selegiline, rasagiline), and bioenergetic drugs (creatin).

A biochemical link between MPTP toxicity and PD was established when several groups reported reduced complex I activity in the brain, platelets and skeletal muscle of patients with PD [60]. The ability of MPTP to give rise to a parkinsonian condition has also led to suggestions that there may be a naturally occurring or environmental toxin that causes PD. Potential candidates for such endo- and exogenous neurotoxins include 4-phenylpyridine, 1,2,3,4-tetrahydroisoquinoline and its derivatives, β-carbolines, paraquat, and heavy metals. Indeed, epidemiological studies have demonstrated an increased risk for PD in rural areas of industrialized societies associated with the use of pesticides, herbicides and heavy metals (iron, manganese), although these findings have not been confirmed by others and, as yet, no specific toxin has been identified [31].

Subthalamic nucleus deep brain stimulation for Parkinson's disease

Deep brain stimulation (DBS) for treatment-refractory tremor or fluctuating motor symptoms in PD is a routine neurosurgical intervention. Benabid et al. [1] were the first to describe the use of a chronically implantable DBS-system for long-term electrical stimulation. Since then, more than 70,000 DBS interventions have been performed worldwide and various large-scale, multicenter, randomized controlled studies have revealed that DBS in PD significantly increases quality of life [21]. However, until this accurate, precise and safe stereotactic DBS procedure was commercially available for clinical use, decades of neuroscientific research was necessary, beginning with the ablation therapy in patients with parkinsonism in the late 1940s up to the currently used chronic DBS. As an example for translational research in DBS we herein briefly outline how the STN was targeted in PD.

In the 1980s a better understanding of the basal ganglia anatomy and function, and the availability of 1-methyl-4-phenyl-1,2,3,6 tetrahydropyridine (MPTP) animal model revealed the crucial role of the STN in the pathophysiology of PD [3, 20]. It was recognized that the dopaminergic loss within the nigrostriatal pathway was associated with shifts in the balance between the neuronal activity of the basal ganglia ‘motor’ circuit, which accounts for the development of cardinal parkinsonian motor symptoms [20]. In this respect, increased STN activity was established as a relevant feature for the internal pallidal hyperactivity that causes pronounced inhibition of thalamocortical projection neurons. Based on this novel concept of basal ganglia organization, Bergman et al. demonstrated that in MPTP-treated monkeys lesioning of the STN with ibotonic acid dramatically improved all of the parkinsonian motor symptoms. They postulated that surgical STN inactivation might be a potential clinical application for the treatment of PD [3]. This theoretical and practical reasoning lead Benazzouz et al. to apply chronic DBS in MPTP-treated primates to reversibly inactivate the STN in the early 1990s. In their experiments they clearly demonstrated that alleviation of parkinsonian motor symptoms was possible by high-frequency stimulation at 100–130 Hz. Furthermore, they observed, contrary to the traditional view that STN lesioning causes severe hemiballism, that STN-DBS was well tolerated [2]. Finally, the group of Benabid in Grenoble translated these outstanding basic research findings in primate models of PD to human trials. In 1995, a small case series of three advanced PD patients treated with bilateral STN DBS was published [47]. Chronic STN DBS was found to significantly improve PD motor symptoms by about 42–84% and further lead to increased activity of daily living. The maximum effects were seen by a high-frequency stimulation over 50 Hz, which remained stable up to 2,000 Hz and allowed postoperative levodopa dosage reduction of up to 50%. In addition, they demonstrated that the DBS adaptability enabled controlling of side-effects, making this technique preferable to lesioning techniques.

Today, chronic DBS is mainly applied to the STN with a prevalent frequency of 130 Hz, due to its effect on all cardinal motor symptoms and since it allows for a reduction of the dopaminergic medication. Since the first publications of STN stimulation in PD a variety of methodological refinements (e.g. optimized preoperative imaging techniques, rechargeable batteries, smaller impulse generators) have reduced operation times and lowered morbidity. Nevertheless, many questions remain to be answered regarding the mechanisms and effects of DBS. For example, there is a question concerning to what extent DBS might exert a neuroprotective effect is still unresolved, which would imply a recommendation for its application early in the course of the disease. Further technological advances and non-toxic animal models might help to increase the understanding of DBS and to define new brain targets in PD and other movement disorders.

Genetics

Although PD was considered a sporadic disorder until the late 1980s, twin studies suggested that there might also be hereditary factors underlying its pathogenesis. Today it has become clear that mutations in several genes underlie autosomal-dominant or autosomal-recessive inheritance in families with PD. Some of them are rare, e.g. mutations in α-synuclein (<20 families reported world-wide), some more frequent, e.g. mutations in LRRK2 (about 5% in the US and Europe but 40% in North African Arabs and Ashkenazi Jews). In total, at least 16 loci have been identified to cause familial forms in AD. In addition, genome-wide association studies (GWAS) suggest common genetic variants in the aetiology of typical PD but also indicate population-specific genetic heterogeneity in this disease [64]. GWAS in a European population predicted strong association of PD at the α-synuclein (SNAC) and tau (MPTA) locus, and a locus neighbouring LRRK2. The results were replicated in the Japanese population for SNAC and LRRK2, but not for MAPT. Furthermore, a new genetic locus, PARK16, was identified in both populations. These new data show that genes identified in rare forms of PD may have implications in all patients with typical idiopathic PD.

Major breakthroughs have been achieved by studying the gene products identified in genetic studies. The best example is the identification of α-synuclein. It quickly became clear that α-synuclein is not only mutated in rare forms of PD but that aggregated, non-soluble α-Synuclein is the major constitutive of all Lewy bodies, which are the pathological hallmark for the diagnosis of all typical cases of idiopathic PD. This facilitated a new research area in PD and led to the Braak staging, which then initiated new research activities in our understanding of pre-symptomatic stages of PD. This includes early symptoms such as impaired smell, depression, idiopathic rapid-eye-movement sleep behaviour disorder (RBD), and late symptoms such as dementia. It took some time to understand that neither the aggregated form of α-Synuclein nor the Lewy body itself provide the toxicity that leads to the degeneration of dopaminergic neurons, but its pre-fibrillar forms, the early and soluble oligomers [38]. Although there is no disease-modifying treatment available yet that is based on inhibiting α-Synuclein aggregation or removal of misfolded α-Synuclein, there is no doubt that future initiatives will include such strategies. Furthermore, screening for smell impairment or RBD will lead to the identification of a population with increased risk to develop PD, a cohort that may be worthwhile to follow in clinical trials testing disease-modifying interventions.

Identified genes and their interactions have implicated specific biochemical pathways as underlying PD. This includes the ubiquitin-proteasomal (mutations in α-synuclein, Parkin, and UCH-L1) or lysosomal (mutations in ATP13A2, glucocerebrosidase) degradation pathways as well as mitochondrial dysfunction (mutations in DJ-1, Pink1) [59]. It is currently an open question as to whether these pathways interfere or lead to the same downstream molecular mechanisms. Taking into account that (1) there is impairment of complex I in the substantia nigra of PD patients, (2) MPTP toxicity replicates many biochemical, pathological and clinical features of PD and (3) pre-fibrillar α-Synuclein may interfere with mitochondrial function, impairment of mitochondrial function may be the common downstream converging point resulting in neuronal dysfunction and neuronal death. Whether mitochondrial inhibition per se may lead to the formation of Lewy bodies remains less clear. However, it may not be sufficient to inhibit PD at this common downstream converging point. If so, it would be necessary to understand the underlying pathological pathway of each individual patient to identify the best possible therapy at an upstream pathogenetic point.

Neuropathology and clinical association

The Lewy body, first described in 1912, has long been considered the pathologic hallmark of PD. However, the concepts of the neuropathology of PD have changed in the last 20 years [49]. The common wisdom that Lewy bodies should be considered mandatory for the pathologic diagnosis of PD has been challenged. It is not known whether Lewy bodies are harmful and interfere with normal cell function or are indicative of a failed cellular self-preservation mechanism [30]. Meanwhile, α-Synuclein-aggregation in Lewy bodies of the substantia nigra has been recognised as a pathologic substrate of PD. However, a more widespread distribution of α-Synuclein pathology, including both noradrenergic, serotonergic and cholinergic sites, and a broad range of other central and peripheral nervous system areas has received increased attention in recent years. Furthermore, Heiko Braak [10] has proposed that PD starts not in the substantia nigra but in the lower brainstem and olfactory bulb and tract. According to this model, degeneration of dopaminergic areas starts after involvement of a variety of other central and peripheral areas in nervous system (Braak stage 3). It has to be concluded from these observations that PD is a neurodegenerative multisystem process. The pathology of PD can now be associated with a wide spectrum of non-motor and motor symptoms [49]. Many of these non-motor features including e.g. olfactory, gastrointestinal and sleep dysfunction, predate the classic motor symptoms and now can be used as early markers of the disease. The clinical recognition of premotor problems can be helpful to improve both symptomatic and putative neuroprotective therapy. However, the progressive nature of PD remains unsolved to date. Further research is necessary to identify the cause of PD and to characterize the subsequent multisystem processes in correlation to clinical symptoms and the development of new therapeutic possibilities.

Cell replacement strategies

When the first successful reports of functional nigral grafts in animal models of PD were published in the late 70s [7, 54] and early 80s, e.g. [23, 24] hopes for a similarly successful application in the clinical practice were flying high. Although selected cases of PD patients benefited for several years from transplantation of human fetal ventral mesencephalic allografts, e.g. [48, 56, 68] or dissociated human embryonic dopaminergic neurons, e.g. [26, 27, 68], the overall outcome of fetal mesencephalic transplantation studies was disappointing. While these open label studies showed clinical improvements and less requirement of anti-parkinsonian medications double-blind placebo controlled trials were unable to repeat the observed positive outcomes [26, 50]. Graft-induced dyskinesia (off-medication dyskinesia), especially, developed as a serious side-effect [51]. Skepticism about the ultimate success of fetal nigral transplantation culminated with the recent finding of Lewy body (LB) pathology in these grafts [16, 41, 42, 45, 46]. Thus, the pathogenic process underlying PD obviously propagated to grafted neurons.

Although animal studies paved the way for transplantation therapy in PD patients, researchers could not foresee long-term drawbacks such as LB pathology and severe dyskinesia. It seems that the ideal animal model is almost impossible to achieve. However, nobody will seriously dispute that animal studies remain an imperative prerequisite in approaching any translation of promising therapy concepts into the clinic. Present expectations are directed towards stem cells as a potential replacement therapy for lost dopaminergic neurons. Stem cells can be easily propagated and are amenable to manipulation in the culture dish. But a lot of pre-clinical research in cell culture and animal models still needs to be done before the transplantation of stem cells can be regarded as a feasible therapeutic strategy. Principally, stem cells might serve as a source for allografts [embryonic stem cells (ESCs), fetal neural precursor cells] or autografts [endogenous adult neural precursor cells, carotid body stem cells, mesenchymal stem cells, induced pluripotent stem (iPS) cells] [70]. Human ESCs extracted from the blastocyst were best differentiated into dopaminergic neurons in vitro using a combination of feeder cells and growth factors [55, 57, 61], but propagation of selected neural precursor cells from cultured aggregates of ESCs proved to be even more successful [15]. Although transplantation studies with ESC-derived dopaminergic neurons have demonstrated behavioural and neurological recovery in rat [39] and primate PD models [65], transplantation studies with human ESCs were much less efficient in animal models [71]. In contrast to the ethical concerns and potential immune rejection connected with the use of differentiated ESCs, iPS cells from adult somatic cells are not dependent on human embryos and offer the advantage of an autograft. Recently, iPS cells with properties of ESCs were successfully generated from human dermal fibroblasts using a sophisticated culturing protocol [66]. Moreover, mouse fibroblast derived iPS cells exhibiting a dopaminergic phenotype have been shown to improve behaviour in a rat model of PD [69]. Both, ESCs and iPS cells share the problem of possible teratoma formation after transplantation, although this might be decreased by preselecting more differentiated cells and removing pluripotent cells. On the other hand, this process might reduce the survival of grafted cells to a significant extent, contributing to the second problem of the so-far poorly efficient generation of iPS cells or true dopaminergic neurons from ESCs. Finally, the question remains as to whether the quality and duration of induction into a true dopaminergic neuron persist after the transplantation process. These hindrances, however, have not restrained some companies from offering doubtful and potentially harmful stem cell therapies to PD patients whose bone marrow derived mesenchymal stem cells were injected into their brains, although these cell types give rise to dopaminergic neurons only when differentiated with defined protocols in vitro [22, 67]. The International Society for Stem Cell Research (ISSCR) has stated a distinct position on unproven commercial stem cell interventions (see their “Guidelines for the Clinical Translation of Stem Cells”). The existence of “black sheep” in the field of stem cell application once more underscores the necessity and urgent need of profound basic research before cell replacement therapy for PD can be translated into the clinic.

Challenges of the future

It can be easily concluded that during the previous 50 years, significant progress has been made in the clinical care for patients suffering from Parkinson's disease. This is obviously mainly due to impressive progress made in the fields of neuropharmacology and neurochemistry and reflected by the Nobel prize given to Arvid Carlsson. This progress has made Parkinson's disease—not only in public opinion—the primary treatable neurodegenerative disease. In addition, we could include deep brain stimulation in our therapeutic armamentarium. However, treatment remains mainly symptomatic and progress must be made in preventive, disease modification and regenerative approaches.

This attempt is currently hampered by the lack of animal models which reflect one key feature of the disease—its progressive nature— which in the ideal world should at least partially mirror the staging as suggested by Heiko Braak [10]. This progression is not present in the neurotoxic models, which served the community well with regard to the investigation of the neurochemistry and neuropharmacology of Parkinson's disease. It also generated hypotheses, such as the presence of mitochondrial damage, which proved to be useful in the study of human Parkinson's disease. Also, cellular–neuronal and non-neuronal models are urgently needed for the understanding of disease pathogenesis and—in a second step—high-throughput drug screening. It could be that the routine use of human iPS cells will approach this goal.

Although the Braak stages have given us significant insights into the clinical and preclinical stages of Parkinson's disease, we do not know yet whether the key to the pathogenesis should be seen in neuronal and axonal degeneration alone. Alternatively, cell death is non-cell autonomous, in other words, not only neurons but also microglial, astrocytic and the dysfunction of other cells are part of the disease process [35]. This aspect of Parkinson's disease is not yet explored. However, improved knowledge could be important since answers would potentially open novel therapeutic avenues. In particular, an attempt could be made on this basis to identify drugs which are promising to modify the natural history of the disease, and regenerative approaches could be developed, or both.

In addition to the presence of effective symptomatic therapeutic strategies, the field of Parkinson's disease has another important advantage—it is obvious that the disease has preclinical stages and is systemic, meaning that it affects tissues outside the central nervous system. Based on this opportunity it seems to be possible that early preclinical and systemic biomarkers (blood or other accessible tissues) can be developed. In an ideal world these biomarkers should be non-invasive and cheap, making mass screening of a given population possible. This again, together with appropriate disease-modifying drugs or regenerative strategies potentially supported by adjustments of lifestyle, opens the exciting outlook of prevention of the disease James Parkinson described close to 200 years ago.

Acknowledgments

Manfred Gerlach, Gabriele Gille, Wilfried Kuhn, Martina Müngersdorf, Peter Riederer, Martin Südmeyer, Albert Ludolph declare no conflict of interest. Jörg B. Schulz serves on scientific advisory boards for Santhera Pharmaceuticals and Takeda Pharmaceutical Company Limited; has received funding for travel and/or speaker honoraria from GlaxoSmithKline, Merz Pharmaceuticals, LLC, Santhera Pharmaceuticals, Lundbeck Inc, Pfizer Inc, and Takeda Pharmaceutical Company Limited; serves as Editor-in-Chief for the Journal of Neurochemistry, and as Associate Editor for the Journal of Neuroscience; has received research support from the BMBF, and the DFG.

Conflict of interest

None.

Copyright information

© Springer-Verlag 2011