The pallidopyramidal syndromes: nosology, aetiology and pathogenesis

Purpose of review The aims of this review is to suggest a new nomenclature and classification system for the diseases currently categorized as neurodegeneration with brain iron accumulation (NBIA) or dystonia-parkinsonism, and to discuss the mechanisms implicated in the pathogenesis of these diseases. Recent findings NBIA is a disease category encompassing syndromes with iron accumulation and prominent dystonia–parkinsonism. However, as there are many diseases with similar clinical presentations but without iron accumulation and/or known genetic cause, the current classification system and nomenclature remain confusing. The pathogenetic mechanisms of these diseases and the causes of gross iron accumulation and significant burden of neuroaxonal spheroids are also elusive. Recent genetic and functional studies have identified surprising links between NBIA, Parkinson's disease and lysosomal storage disorders (LSD) with the common theme being a combined lysosomal–mitochondrial dysfunction. We hypothesize that mitochondria and lysosomes form a functional continuum with a predominance of mitochondrial and lysosomal pathways in NBIA and LSD, respectively, and with Parkinson's disease representing an intermediate form of disease. Summary During the past 18 months, important advances have been made towards understanding the genetic and pathological underpinnings of the pallidopyramidal syndromes with important implications for clinical practice and future treatment developments.


INTRODUCTION
The term pallidopyramidal degeneration (PPD) was first introduced by Davison in 1954 [1] who described a series of five patients presenting with the triad of progressive parkinsonism, spasticity and dystonia combined with pyramidal and pallidal lesions and blue discoloration of the globus pallidus following the initial report by Hunt in 1917 [2] of a single case with juvenile parkinsonism and eosinophilic spheroidal structures. Subsequently, Hallervorden and Spatz in 1992 reported a family with five affected sisters with brown discoloration of the globus pallidus [3], a syndrome that was named Hallervorden-Spatz syndrome.
During the past decade, the advent of genetic technologies has allowed a more systematic delineation of the clinical presentations and genetic underpinnings of PPD starting with the identification of the first mutations in PANK2 [4], a finding that led to the renaming of this disease class to neurodegeneration with brain iron accumulation (NBIA) [4,5]. Despite the fact that a molecular diagnosis and modern neuropathological analysis is not possible for the initial cases described by Davison due to the lack of preserved tissue and blood, it is likely that the brown-blue discoloration of the globus pallidus represents gross iron accumulation and the eosinophilic formations neuroaxonal spheroids, and that all belong to the modern disease entity of NBIA (Fig. 1a).
In this review, we argue that the use of the term NBIA is not ideal and suggest that the more general term pallidopyramidal syndromes (PPS) conceived by Davison would perhaps be more appropriate [6]. In this context, we also suggest a modified classification system better reflecting the clinical and pathological phenotypes associated to PPS. Finally, we outline possible disease mechanisms providing a mechanistic basis for some of the features unique to PPS that were first highlighted by Davison, and suggest a model tying the pathogenesis of lysosomal storage diseases (LSD), Parkinson's disease, and PPS.

CLASSIFICATION OF PALLIDOPYRAMIDAL SYNDROMES
At present, according to the OMIM classification system, a disease is classified as NBIA based on the clinical features including Davison's PPD triad (Fig. 1b), and gross iron accumulation on T2 Ã MRI. Further classification in four subtypes depends on the pattern of iron accumulation on MRI, and on the underlying mutated genetic locus ( Table 1).

Caveats of the current classification system
The current classification system is far from ideal for two reasons: (1) The use of iron accumulation as a classification criterion is debatable. Iron accumulation is not a consistent finding among diseases with otherwise indistinguishable clinical presentations [7][8][9][10] leading to the use of the oxymoron term 'NBIA without brain iron' [10]. Also, the importance of iron for disease pathogenesis and progression remains elusive [10], especially as this has been reported in various seemingly unrelated disorders and even in healthy individuals [10][11][12]. (2) The use of a classification system based on mutated genetic loci has two weaknesses. First, as the complete genetic landscape of NBIA is unknown, there are several 'idiopathic' syndromes that are not included in the current classification system (Fig. 1c). Second, patients with mutations in the same gene often present with substantially divergent clinical features [8,13].
Taking into account these weaknesses, we suggest that at present a clinical and pathological classification system of NBIA would be more suitable for clinical practice.

Pallidopyramidal syndromes: Proposed clinical classification system
In our suggested classification system, a disease must be characterized by Davison's triad with or without iron accumulation on MRI to be classified as PPS. Further sub-classification is based on the age at onset of symptoms as this feature can serve as a starting point to prioritise genetic testing (Table 2A).

KEY POINTS
We suggest the use of the more general term 'pallidopyramidal syndromes' instead of NBIA when referring to syndromes with prominent dystoniaparkinsonism, and an alternative clinicopathological classification system for these diseases better suited to clinical practice.
Often, a differential diagnosis has to be made from phenocopies (atypical, usually milder presentations of syndromes caused by mutations in different genetic loci, reviewed in [9]). However, we do not include these in the suggested classification system as they probably do not fit in the nosological entity originally described by Davison and      ]. Here, we discuss the potential pathogenic processes underpinning these lesions and their implications for our suggested pathological classification system.

a-Synuclein and pallidopyramidal syndromes
Although a-synuclein deposition consistently occurs in various neurodegenerative diseases it is still unclear whether this is the primary event driving disease pathogenesis or is just an epipheno- Here, drawing mainly from studies on Parkinson's disease, we argue that most recent evidence implicates lysosomal dysfunction and/or lipid abnormalities in the aggregation and spreading of a-synuclein and then discuss the implications of this observation for the pathogenesis of PPS.

Lewy body formation is probably caused by lysosomal dysfunction
Lewy bodies have been reported in four disease categories: Parkinson's disease, PPS, LSD and dementia with Lewy bodies. Interestingly, the common denominator in most of these situations with a-synuclein accumulation appears to be lysosomal dysfunction: (  , a protein whose precise function is currently unknown. Finally, Lewy bodies are frequent in 'sporadic' Parkinson's disease in similar distribution and severity to GBA-associated disease [39]. The significance of these observations and the relation of a-synuclein accumulation to mitochondrial dysfunction are discussed later.
Interactions between a-synuclein and lipids both within the lysosomal context [51] and the cytoplasm [92 && ,93] also seem to underlie a-synuclein homeostasis. Such extensive a-synuclein-lipid interactions are in keeping with the highlighted frequent involvement of ceramide metabolism pathways in parkinsonian disorders with Lewy Bodies in neuropathology [94]. Given that GlcCer interactions have been shown to stabilize a-synuclein oligomers, we can hypothesize that similar a-synuclein-lipid interactions in the cytoplasm could have similar consequences.
Why are Lewy bodies absent from some pallidopyramidal syndromes but present in others? & ]: the defect initiates from the globus pallidus but cannot spread to other brain regions due to the absence of a-synuclein involvement. As recent evidence implicates lysosomal dysfunction and/or lipid abnormalities in the aggregation and spreading of a-synuclein, this observation has two possible implications for the pathogenic mechanisms of PKAN: (1) Probably, lysosomal dysfunction is not a primary event in the pathogenesis of PKAN. , and traumatic brain injury [31]; however, these are also observed in healthy, aged individuals [123]. Even though neuroaxonal spheroids have not been ultrastructurally studied in genetically confirmed cases and systematically compared between various diseases, limited electron microscopy studies on nongenetically confirmed HDLS and on mouse models of PLA2G6 have indicated that these structures likely contain mitochondria [124][125][126] in addition to other molecules [31,32 &

, 35
& ,119]. Given the similarities in the staining patterns of the spheroids between diseases, it is likely that they represent identical or highly homologous structures, a remarkable finding given the diversity in clinical presentations of associated diseases.
As neuroaxonal spheroids are present in such a variety of serious neurodegenerative diseases, the mechanisms underlying their formation are intriguing. Here, we hypothesize that spheroids could result from impaired mitochondrial trafficking as a reaction to severe neuronal damage drawing from evidence provided from the study of PKAN.
(1) Specifically in the case of PPS, perhaps their formation stems from a primary mitochondrial dysfunction, a relationship that would seem more clear and convincing in PKAN. As mitochondria heavily rely on CoA provision for energy generation, it is expected that PANK2 mutations would have a devastating effect on mitochondrial integrity [127 && ]. The increased number of large degenerate mitochondria [127 && ] could result in the overload of the macroautophagy pathway with the formation of large, indigestible autophagosomes that cannot be uptaken by the lysosomes [128]; thus, neuroaxonal spheroids could represent these indigestible autophagocytic vesicles. Alternatively, damaged mitochondria could impinge on lysosomal function indirectly through impaired microtubule trafficking [129 && ,130 & ].
(2) Mitochondrial trafficking impairment could occur secondarily to mitochondrial defects. It has been recently shown that Miro, a mitochondrial trafficking protein [131,132], is selectively targeted by PINK1 and parkin in mitochondrial damage in order to halt mitochondrial      or absence of a-synuclein accumulation and/or Lewy bodies (Table 2B). In addition, as neuroaxonal spheroids are a common feature of several neurodegenerative diseases, we suggest the establishment of a separate disease category of 'spheroidopathies' (Table 2B, Fig. 2, Table 3).

DISEASE MODEL HYPOTHESIS: THE 'PARKINSONIAN MITOCHONDRIAL-LYSOSOMAL TRIANGLE'
There appears to be a clear relationship between PPS, Parkinson's disease and LSD clinically [9,138,139] ]. Interestingly, for some of the mutated molecules involved in the dysfunction of these two organelles, there is functional and neuropathological evidence mapping them clearly in one of the two pathways. However, for the rest there seems to be an overlap: even though for each mutated gene there is strong functional evidence that only one of the two organelles should be affected, there are circumstantial pathological features indicating that perhaps the second organelle is affected too (Table 4).
Thus, in general, Lewy bodies consistently occur in cases with mutations in lysosomal enzymes whereas these are found only occasionally in relation to mutations in mitochondrial proteins. This observation would support the hypothesis that mitochondrial dysfunction does not directly cause a-synuclein accumulation; indeed, to date, there is not strong enough functional evidence that ]. Such an overlap in pathologies would suggest that there is a functional link between lysosomes and mitochondria and that unknown events (or perhaps even stochasticity) could shift the balance between the two pathways and some well determined genetic forms of Parkinson's disease develop inconsistent pathological features. We term this functional continuum 'Parkinsonian mitochondrial-lysosomal triangle' and suggest that PPS and LSD lie in the extreme ends of this triangle with Parkinson's disease as an intermediate form of disease (Fig. 3a).
If this theory holds true, we can make two interesting hypotheses: ( (2) How can lysosomes and mitochondria be functionally connected? Though the exact nature of this link is unknown, it is thought to take the form of mitophagy [185] and to be bidirectional (Fig. 3b). Indeed, there is evidence suggesting that the dysfunctional lysosomes 'attack' mitochondria in ATP13A2 patient fibroblasts [160 && ] and that lysosomal dysfunction could result in an accumulation of dysfunctional mitochondria in mouse models of LSD [186]. Conversely, damaged mitochondria can impact on autophagy through impaired microtubulemediated vesicular trafficking resulting in a more generalized lysosomal dysfunction (including inhibition of a-synuclein degradation)   ]. It has also intriguingly been hypothesised that MAPT variants could impact on the type of pathology exhibited with LRRK2 mutations shifting the balance between tau and Lewy bodies [195,196 &

CONCLUSION
We propose a simplified classification of PPS that allows incorporation of the increasing genetic findings. Although the precise pathogenic underpinnings of PPS are far from clear, numerous reports suggest interesting links on multiple levels between PPS, Parkinson's disease and LSD with a central role for combined mitochondrial and lysosomal dysfunction, a relation which will be further dissected as identification of novel disease-causing genes adds the missing pieces to the puzzle [203 && ]. rology, UCL for useful discussions. The authors would also like to thank the Medical Research Council (MRC), the National Organisation for Rare Disorders (NORD), the Dystonia Medical Research Foundation (DMRF), the dystonia coalition and the Parkinson's disease foundation (PDF) for funding their research.   76. Gitler AD, Chesi A, Geddie ML, et al. Alpha-synuclein is part of a diverse and highly conserved interaction network that includes PARK9 and manganese toxicity. Nat Genet 2009; 41:308-315. 77.