Synucleinopathies: Clinical and Pathological Implications

The synucleinopathies are a diverse group of neurodegenerative disorders that share a common pathologic lesion composed of aggregates of insoluble α-synuclein protein in selectively vulnerable populations of neurons and glia. Growing evidence links the formation of abnormal filamentous aggregates to the onset and progression of clinical symptoms and the degeneration of affected brain regions in neurodegenerative disorders. These disorders may share an enigmatic symmetry, ie, missense mutations in the gene encoding for the disease protein (α-synuclein) cause familial variants of Parkinson disease as well as its hallmark brain lesions, but the same brain lesions also form from the corresponding wild-type brain protein in the more common sporadic varieties of Parkinson disease. It is likely that clarification of this enigmatic symmetry in 1 form of synucleinopathy will have a profound impact on understanding the mechanisms underlying all these disorders. Furthermore, these efforts will likely lead to novel diagnostic and therapeutic strategies in regard to the synucleinopathies.

The synucleins are a family of soluble proteins whose function is not well understood. These proteins are characterized by an acidic carboxyl terminus and 5 or 6 imperfect repeat motifs (KTKEGV) distributed throughout the aminoterminus.^1-6 There are currently 4 members of the synuclein family that range between 127 to 140 amino acids in length. The first synuclein was described in 1988, after it was purified from the Torpedo electroplaque and from rat brain,⁷ and it is known as α-synuclein (αS).^1-6,8 Later, however, it was also named the nonamyloid component (NAC) of plaque precursor protein after the NAC peptide was isolated from amyloid-rich senile plaques (SPs) of brains of patients with Alzheimer disease (AD)^1-6,8 and shown to be identical to residues 61 to 95 of αS. The αS gene then was mapped to chromosome 4q21.3-q22.^1,2,9

The second member of the synuclein family, which now is known as β-synuclein (βS), is highly homologous to αS. It was originally isolated from the bovine brain and initially termed phosphoneuroprotein-14.^1-6,10 The gene for βS was mapped to chromosome 5q35.^1,2,6 Because of the similar localization of both proteins, predominantly in the presynaptic terminals of neurons, it has been speculated that αS and βS may be involved in synaptic function.^1,2,6

The third member of the synuclein family, now designated γ-synuclein (γS), was isolated from breast cancer tissue and initially termed breast cancer gene–specific product 1.^1,2,6,11,12 In addition to breast tissue, γS is expressed in the brain and spinal cord but is most abundant in the peripheral nervous system. Unlike αS and βS, γS is predominantly cytosolic in location,^1,11,12 and it has been mapped to chromosome 10q23.^1,2,11,12 The most recently described member of the synuclein family is termed synoretin, but the gene encoding this protein has not been mapped at this time.¹³ Synoretin is also distributed throughout the cytoplasm, but it is mainly expressed in the retina, and there is only low-level expression in the brain.^1,13 While αS has been extensively implicated in mechanisms underlying neurodegenerative disorders,^1-6,14-16 until recently, there has been no evidence to suggest a role for βS and γS in neurologic disease.¹⁴

Parkinson disease (PD) is the most common neurodegenerative movement disorder, and it is well characterized clinically as well as pathologically.^1,14,15,17 It is estimated that there are more than 500 000 patients with PD in the United States alone. The cardinal clinical features of PD include (1) bradykinesia, (2) cogwheel rigidity, (3) resting tremor, and (4) postural instability. Dementia is a variable manifestation of PD, but it occurs in approximately 40% of patients during the protracted course of the disease. Pathologically, PD is characterized by the loss of the dopaminergic neurons from the substantia nigra pars compacta (SNpc), the presence of extracellular melanin released from degenerating neurons, reactive gliosis, and eosinophilic intracytoplasmic inclusions known as Lewy bodies (LBs) in residual SNpc neurons.^1,4,15-18 The LB, first described by Frederick Lewy in 1913 in the basal forebrain,¹⁹ is present in essentially all cases of PD, and the detection of LBs in SNpc neurons of the postmortem brain is required to establish a definite diagnosis of PD.^1,4,15-18 The primary neurochemical deficit in PD is the loss of dopaminergic projections to the striatum as a result of the massive degeneration of dopaminergic nigral neurons. This neurochemical defect can be partially corrected by the administration of levodopa and related therapies that substitute for the loss of SNpc neurons, which thereby ameliorates at least the motor impairments of PD, particularly early in the course of the disease.^1,15,17 However, in addition to the dopaminergic neurons of the SNpc, other populations of neuromelanin-containing and non–neuromelanin-containing neurons in the brainstem and basal forebrain degenerate and accumulate LBs that are similar to those found in the SNpc, possibly accounting in part for many of the secondary clinical features in PD, including autonomic instability, seborrhea, sleep disturbances, and possibly dementia.^1,17

While nigral and other subcortical LBs were well documented in PD, it was not until 1961 that Okazaki et al²⁰ described cortical LBs in a patient with severe dementia and quadriparesis.^1,17 This original description opened investigation into a new category of late-life dementias, collectively referred to as dementia with Lewy bodies (DLB).^1,3,5,14-18 This brain disorder may be the second most common form of dementia after AD. Clinical features that are most characteristic of DLB include (1) progressive dementia with a fluctuating clinical course, (2) extrapyramidal signs (typically bradykinesia and cogwheel rigidity but rarely resting tremor), (3) prominent visual hallucinations, and (4) increased sensitivity to neuroleptic medications.

In the decades that followed the original descriptions of the neuropathologic and clinical features of PD, a large body of experimental evidence developed that correlated the neuropathologic features of PD with its different clinical manifestations. However, recent dramatic insight and clues into the cellular, molecular, and genetic pathobiology of PD have now emerged that challenge the conventional and traditional views about the basic mechanisms underlying brain degeneration in PD as well as the common, although unclear, overlap between PD and AD.^1,5,14-18 Recent evidence highlights the clinical heterogeneity of both PD and AD.^1-6,14-18 Patients with PD may develop an AD-like dementia in the latter stages of the disease, and as many as 25% to 35% of patients with AD develop extrapyramidal signs during their illness.^1,15,17,18

Investigators have demonstrated by immunohistochemical examination that LBs contain many different cytoskeletal and noncytoskeletal proteins, but before 1997 the 2 most consistently described proteins were neurofilament and ubiquitin.^1,17 However, in 1997, a landmark finding by Polymeropoulos and coworkers²¹ detailed 5 Mediterranean families with autosomal dominant PD caused by a missense mutation in the αS gene leading to an A-to-T substitution at amino acid 53. Subsequently, a second mutation was described in 1 family of German origin with an A-to-P substitution at amino acid 30.²² Furthermore, in 1997 and shortly thereafter, it was discovered that antibodies specific for αS detect numerous LBs and dystrophic Lewy neurites (LNs) in sporadic PD and DLB.^{1-6,14-16,18,23}

There is now mounting evidence to support the concept that αS is the primary building block of the fibrillary component of LBs.^1-6,14-16 Antibodies that recognize αS stain LBs more intensely and more consistently than antibodies directed against other protein components, and immunoelectron microscopic studies have demonstrated that LB fibrils are intensely decorated with αS antibodies in situ.^1-6,14-16 There is ample evidence that insoluble αS filaments accumulate in brains of patients with PD and DLB and that recombinant αS protein can assemble in vitro into elongated filaments with ultrastructural features similar to those of LB filaments visualized in situ.^1,24 In addition, αS aggregation within the dystrophic LNs is associated with brain regions rich in perikaryal LBs, particularly the CA2/3 region of the hippocampus. Accordingly, it appears increasingly plausible that this pathologic finding may play an important role in the loss of neuronal function.^1,5,14,16

Increasing interest in the pathobiological features of αS has been driven in part by the expanding spectrum of LB-related disorders (Figure 1).^1-6,14-18 For example, αS-rich cortical LBs similar to those found in the mesencephalic neurons of patients with PD have also been detected in cortical neurons of some demented patients.^1-6,14-18 The presence of numerous cortical and subcortical LBs, SPs, and neurofibrillary tangles defines a subtype of AD referred to as the LB variant of AD.^{1,5,14,15,17,18} Furthermore, the presence of abundant cortical LBs in the brains of patients with a late-life AD-like dementia who present with extrapyramidal and neuropsychiatric symptoms without significant SPs or neurofibrillary tangles is diagnostic of the disorder referred to as DLB.^1-6,14-18 Notably, it appears that the number of cortical LBs correlates with the severity of dementia and that the burden of αS-immunoreactive LBs is the most specific and sensitive marker for dementia in patients with PD and DLB.^15,17,25 In addition, LBs have been described in the amygdala of patients with familial forms of AD with mutations in the genes for the amyloid precursor protein and presenilin-1 and -2^1,2,5,26 as well as in the brains of aged patients with Down syndrome.^1,2,5,27

In recognition that LB dementias may be the second most common form of neurodegenerative dementing illness in the elderly after AD and because of the confusion regarding nomenclature and diagnostic criteria for LB dementias, a consortium on DLB formalized standard procedures for the definitive pathological and clinical diagnosis of this disorder.^1-6,14-18 The recommendations of the consortium included the use of antibodies against ubiquitin to detect LBs. However, ubiquitin immunoreactivity is present in a number of neurodegenerative lesions, some of which may be very difficult to distinguish from LBs (eg, neurofibrillary tangles).^1,17 Thus, the use of antibodies specific for αS to detect LBs and LNs is a much more specific technique that should improve the standardization of the postmortem diagnosis of neurodegenerative diseases characterized by LBs and LNs as well as other diseases with αS-immunoreactive lesions.^1,25

Significantly, αS has been implicated as a major component of the tubulofilamentous inclusions found in oligodendrocytes in multiple system atrophy known as glial cytoplasmic inclusions (GCIs).^2,3,5,6,28 Multiple system atrophy consists of a syndrome complex with parkinsonism and a combination of cerebellar, autonomic, and gait abnormalities as well as variable cognitive decline. These disorders include Shy-Drager syndrome, striatonigral degeneration, and olivopontocerebellar atrophy, all of which are characterized by the presence of αS-immunoreactive GCIs throughout the neocortex, hippocampus, brainstem, spinal cord, and dorsal root ganglia. In addition, neurodegeneration with brain iron accumulation, type I (also referred to as adult neuroaxonal dystrophy or Hallervorden-Spatz syndrome), is a rare neurodegenerative disorder characterized clinically by parkinsonism, cognitive decline, cerebellar signs, and bulbar symptoms. Pathologically, iron deposition is found in the globus pallidus, red nucleus, SNpc, and dentate nucleus of the cerebellum, and axonal swellings known as spheroids are seen. In addition, both GCIs and LB-like neuronal cytoplasmic inclusions are seen in both cortical and subcortical structures. The spheroids, GCIs, and neuronal cytoplasmic inclusions are all easily detected by antibodies against αS.^28,29 The αS component of brain regions with these lesions is largely found in the insoluble fractions, similar to the pattern of distribution seen in LB disorders. Furthermore, axonal lesions after traumatic brain injury also have been demonstrated recently to express αS.²⁹ Besides the above-mentioned disorders, there are several other neurodegenerative diseases in which αS-immunoreactive lesions may contribute to the pathologic features seen in the disorder but may not be the major protein constituent of the lesion. For example, a subset of the Pick bodies within the dentate gyrus of the hippocampus in patients with Pick disease is strongly immunoreactive for αS; however, the major building block of Pick bodies appears to be tau protein.^3,5,16 In addition, glial inclusions in amyotrophic lateral sclerosis also show αS immunoreactivity (J.E.G., V.M.-Y.L., J.Q.T., unpublished data, 1999).

It has been demonstrated recently that αS, βS, and γS are expressed in olfactory epithelium, especially the olfactory receptor neurons. These neurons retain the ability to regenerate continuously throughout life.³⁰ Although olfactory dysfunction is a feature of several neurodegenerative diseases and dystrophic neurites were detected with antibodies to αS and βS in the postmortem examination of olfactory epithelium of patients with synucleinopathies, similar patterns of pathologic and normal synuclein expression were seen in control subjects and patients with other neurodegenerative diseases. This suggests that synuclein may play a role in the regeneration and plasticity of adult human olfactory epithelium.³⁰

Although βS and γS are abundant in the brain, initially they had not been implicated in a neurodegenerative disease.¹⁴ However, it was recently demonstrated that hilar neurons in the hippocampus of patients with PD and DLB were surrounded by accumulations of αS- and βS-immunoreactive vesicles in presynaptic terminals.¹⁴ This pathologic finding was not seen in normal control subjects or in patients with other neurodegenerative disorders, such as AD or Pick disease. The marked abundance and variable size of these profiles suggested that they reflect pathologic aggregation of αS and βS in the axon terminals of dentate gyrus mossy fiber projections to the hilar neurons. The abundance of these vesicles appeared to parallel the presence of αS-immunoreactive LNs in the CA2/3 region of the hippocampus and LBs in the entorhinal cortex of patients with PD and DLB. Moreover, axonal spheroidlike lesions were identified in the molecular layer of the dentate gyrus of PD and DLB, with antibodies against γS but not with antibodies to αS or βS. The synuclein-rich lesions colocalized with other presynaptic proteins (ie, synaptophysin, synapsin, and synatobrevin), suggesting that the pathologically altered axon terminals or their corresponding synapses may be dysfunctional. This finding is significant because it describes brain abnormalities in a neurodegenerative disease that contains pathologic accumulation of not only αS but also βS and γS. Accordingly, one can conclude that βS and γS, in addition to αS, may play mechanistic roles in the onset or progression of several neurodegenerative disorders. Indeed, because αS pathologic features may be seen in other neurodegenerative diseases as well as in a few normal individuals, βS and γS pathologic features may be more specific to LB disorders.¹⁴

The role αS plays in amyloid plaque formation has been called into question, as several investigators have failed to confirm the initial observation of NAC immunoreactivity in AD SPs.^1,31 Monoclonal antibodies to the NAC peptide have been demonstrated to stain LBs and LNs, but they failed to stain any AD SPs.³¹ It is possible that, in the original isolation of the NAC peptide, the brain regions that were demonstrated to contain SPs also contained abundant αS-rich LNs, and the contaminating LNs may have served as the source for the αS in isolated amyloid plaques.

The finding of insoluble lesions in different neuronal and glial cell populations involving highly varied clinical syndromes suggests that many of these neurodegenerative disorders share a commonality.^1-6,14-18,32 Despite the fact that these syndromes express different phenotypic symptoms and pathologic lesions, the mechanisms underlying filament formation may be similar. Notably, the assembly of normally soluble protein subunits into insoluble filaments in these neurodegenerative diseases does not occur de novo in normal brain.^1,24 Thus, one can conceptualize that another way to approach these different disorders is to consider the disease state as one of an abnormality in protein metabolism (ie, synucleinopathy)^1-6,14-18,32 (Table 1). It is plausible that the common feature of synucleinopathies is the mechanisms underlying the posttranslational modification of synuclein that alters the biophysical properties of this synaptic protein, leading to the formation of insoluble fibrillar aggregates. These biochemical changes most likely lead to a cascade of cellular responses that result in neuronal dysfunction and death that present as a recognizable clinical and pathologic phenotype.^1-6,14-18,32 It is reasonable to propose that future investigative efforts that pursue molecular analyses of shared protein abnormalities across several disorders (ie, synuclein protein in PD; DLB; familial AD; multiple system atrophy; neurodegeneration with brain iron accumulation, type I, or tau protein in Pick disease; frontotemporal dementia; prion disease; corticobasal degeneration; and progressive supranuclear palsy) will provide insights into disease mechanisms underlying 1 or more neurodegenerative disorders characterized by abundant filamentous lesions. In this manner, preventive and potentially curative strategies for these disorders may be possible.

No doubt, the development of transgenic animal models that recapitulate the key pathologic aspects of human neurodegenerative disease will help answer questions regarding the mechanisms leading to neurodegenerative disease. At present, there has been only 1 published report suggesting that mice that overexpress αS develop perikaryal and nuclear αS aggregates, although these aggregates are not fibrillar and it is unclear whether these aggregates are associated with neuronal dysfunction or loss.³³ It is likely that many reports of other efforts to generate αS-, as well as βS- and γS-, transgenic mice will soon appear.

Further studies will be necessary to determine to what extent insoluble aggregates of αS as LBs, LNs, GCIs, or spheroids can lead to the demise of neurons; however, current evidence suggests that their presence may have several deleterious effects. First, the meshwork of αS filaments may serve as a molecular "trap" for proteins destined for transport within axons, preventing transit of vital proteins from soma to axon terminals. As a result, distal regions of axons and dendrites may be deprived of proteins essential for survival or function, resulting in a loss of structural and molecular integrity of the axons and dendrites. Second, the presence of vital cytoskeletal proteins entrapped within the αS filaments may deprive axons of important structural integrity and cause a "dying back" process that may "disconnect" the SNpc from the basal ganglia (in the case of PD) and one cortical region from another (in DLB, for example).^1,5,14,17 Third, the presence of αS aggregates in neurites and axons (ie, dystrophic LNs) and γS in perforant pathway projections also may contribute to the dysfunction of the nervous system, possibly by impeding axonal transport.^1,14 Last, recent studies demonstrating the accumulation of synuclein and other presynaptic proteins in the degenerating terminals of the hippocampal perforant pathway projections could interfere with the unidirectional flow of information in this vital circuit important in memory and behavior.¹⁴ Thus, the accumulation of a normally soluble protein into pathologic insoluble aggregates could not only serve as a marker of disease but also compromise the function and viability of neurons.^1,5,14,17 Efforts to further elucidate the pathobiological features of synuclein proteins are likely to lead to improved strategies for antemortem diagnosis and the development of novel therapeutic interventions for a diverse group of neurodegenerative disorders, including PD; DLB; LB variant of AD; multiple system atrophy; neurodegeneration with brain iron accumulation, type I; and other synucleinopathies.

Accepted for publication May 10, 2000.

This work was supported by grants AG10124 and AG09215 from the National Institute on Aging, Bethesda, Md, and a Pioneer award from the Alzheimer's Association, Chicago, Ill.

We thank Benoit I. Giasson, PhD, and John E. Duda, MD, for their critical review of the manuscript.

Corresponding author and reprints: James E. Galvin, MD, MSc, Alzheimer's Disease Research Center, Department of Neurology, Campus Box 8111, Washington University School of Medicine, 660 S Euclid Ave, St Louis, MO 63110 (e-mail: galvinj@neuro.wustl.edu).