Introduction
Parkinson’s disease (PD), PD dementia (PDD), and dementia with Lewy bodies (DLB) represent a spectrum of neurodegenerative diseases [
1‐
5] collectively known as α-synucleinopathies due to the aggregation of α-synuclein into intracellular Lewy bodies (LBs). The presence of LBs in the brainstem is associated with dopaminergic neuron loss and motor dysfunction in those diagnosed with PD [
6]. However, up to 80% of patients develop dementia (PDD) during their disease course [
7], and this progression is associated with the presence of LBs in cortical areas [
8]. Progression to dementia is also associated with increasing tau co-pathology in a pattern that is similar to that seen in Alzheimer’s disease (AD) [
4,
9], but with a greater temporal neocortical distribution [
10]. Together, these studies have suggested that α-synuclein and tau pathologies are associated with each other in PD-PDD and may directly influence each other or be driven by a common factor to augment the progression of neurodegeneration. Identifying factors that could influence the progression of disease could provide a valuable target for modifying disease trajectory.
While most PD is idiopathic, insights into disease pathogenesis have been gleaned from genetic mutations which elevate the risk of PD. One of the most common genes mutated in PD encodes leucine-rich repeat kinase (
LRRK2) [
11]. The most prevalent mutation in
LRRK2, p.G2019S, confers a 25 to 42.5% risk of PD [
12], with a similar age of onset and disease duration as idiopathic PD [
11], suggesting that it may phenocopy idiopathic PD (iPD). Indeed, the most common feature of
LRRK2 PD, as in iPD, is the loss of substantia nigra neurons [
13]. While LBs are another feature of some
LRRK2 PD, approximately 21–54% of reported
LRRK2 mutation carriers have no apparent LBs [
13,
14], suggesting that some other disease factor may be responsible for the observed clinical disease. A top candidate for this disease factor is tau, since 79% of
LRRK2 mutation carriers have been reported to have some degree of tau pathology [
13].
Further support for tau pathology as the neuropathological substrate of the parkinsonism observed in
LRRK2 mutation carriers has come from the identification of progressive supranuclear palsy (PSP)-like tau inclusions observed in several cases [
15‐
18]. However, the PSP-like tau pathology in these cases was mild, and AD-like tau was also a prominent neuropathological feature [
17,
18]. Multiple population studies have found that
LRRK2 mutations are very rare in pathologically-confirmed primary tauopathies PSP or corticobasal degeneration (CBD) [
19,
20], suggesting that
LRRK2 mutations are primarily associated with PD. What genetic studies do not clarify is whether or not
LRRK2 mutations could drive tau pathology in the context of
LRRK2 PD. It is also not clear if the tau observed in
LRRK2 mutation carriers is PSP tau or AD tau, and if this pathology is sufficient to be classified as the neuropathological substrate of dopaminergic neuron loss.
The current study uses quantitative pathology analysis to determine the type and extent of protein pathologies present in 12 cases with
LRRK2 mutations. In addition to pathological α-synuclein, tau and Aβ staining used for typical neuropathological assessments, we also used an AD tau-selective antibody, GT-38 [
21,
22], to investigate the type of tau present in
LRRK2 PD. We find that tau pathology is a prominent feature of
LRRK2 PD, and that this tau pathology is largely AD-type tau. AD tau staging in LRRK2 PD follows a similar distribution to iPD and iPDD and is accompanied by abundant concurrent Aβ pathology in most cases. Further, tau is not an independent disease factor in
LRRK2 PD, but is associated with the degree of α-synuclein pathology and progression to dementia. Together, these results suggest that
LRRK2 PD is similar to iPD in its accumulation of AD type tau. It will be important for future studies to address whether LRRK2 directly influences the development of tau pathology and whether LRRK2 inhibitors affect tau pathology.
Discussion
The presence of a familial disease that clinically mimics idiopathic disease can provide important mechanistic information about the pathogenesis of disease. The mutated protein may be involved in the biological pathway of disease and may be a therapeutic target for both familial and idiopathic forms of disease. In the case of PD, patients with or without
LRRK2 mutations seem to have similar disease onset and symptoms [
18,
27]. However, whether or not
LRRK2 PD and iPD share the same pathological substrate has been unclear. It has been suggested that there may be two neuropathological classes of
LRRK2 PD [
16,
28]. The most prominent of these is typical PD, with nigral degeneration and LBs; the second presents with nigral degeneration in the absence of LBs, but potentially with PSP-like tau pathology. The presence of LBs has been associated with progression to dementia in
LRRK2 mutation carriers, while cases without LBs manifest primarily motor phenotypes [
14]. While AD tau has been described in
LRRK2 mutation carriers [
13,
17,
18], the prevalence and extent of this pathology and its relationship with PSP-like tau has not been well-characterized. Therefore, in the current study, we sought to quantitatively characterize the level and extent of α-synuclein, tau, AD tau and Aβ pathologies in 12
LRRK2 mutation carriers.
Similar to previous reports, nigral LB pathology was present in 64% (7/11) of our LRRK2 mutation carriers. Remarkably, the 4 cases without nigral LBs also had no discernible Lewy pathology throughout all regions analyzed, clearly suggesting that phosphorylated α-synuclein is not a neuropathological substrate of disease in these patients. While the absence of α-synuclein pathology is rare in PD, we identified 9 “atypical PD” patients out of 109 total iPD patients in our database that had a similar nigral degeneration without LB pathology. Further, the tau pathology burden in the midbrain of LB-negative LRRK2 mutation carriers was low, and was recognized by GT-38, suggesting that it is predominantly AD tau. Our data is therefore not consistent with tau acting as the neuropathological substrate of nigral neuron loss.
In contrast, AD-type tau is overall a prominent feature of
LRRK2 PD. This finding is corroborated by a recent report of 6 p.G2019S
LRRK2 mutation carriers which had variable LB pathology, but intermediate to high AD pathology in 83% (5/6) of the brains [
29]. The level tau and Aβ pathology seen in this
LRRK2 PD cohort is more than would be expected for a non-demented cohort of individuals of a similar age [
30,
31], but consistent with the prevalence of AD dementia in the population by this age [
32]. Previous research has indicated that AD tau is a prominent feature of iPD and bridges the neuropathological space between PD, PDD, DLB, and AD, with higher AD pathology associated with a higher likelihood of progression to dementia [
4,
9]. As in iPD-PDD, α-synuclein pathology burden is associated with both tau pathology and progression to dementia [
9]. The AD tau pathology in
LRRK2 PD cases also follows similar Braak stages to those observed in iPD, suggesting that
LRRK2 PD is similar in pathological presentation to iPD.
Included in this study are two rare variants of unknown pathogenicity. The p.R793M variant was first identified in two families with a history of autosomal-dominant PD [
33]. The pathogenicity of this variant was later called into question due to its detection in two healthy individuals in Norway, although the individuals were relatively young (59 and 61 years old) at the time [
26]. The p.L1165P variant was described later in conjunction with the current p.R793M case as having abundant α-synuclein and tau pathology [
24]. While little else is known about the p.L1165P variant pathologically, it has been shown to elevate kinase activity in an in vitro kinase assay [
34], similar to p.G2019S and other variants [
35]. Interestingly, these rare variants are unique among the cohort examined due to the high degree of tau pathology in the absence of any Aβ pathology. This makes these cases more likely to be primary tauopathies. However, the tau in these cases is also recognized by GT-38 and is present in areas indicative of AD pathology, so the tau in these cases may represent a continuum of AD that lacks Aβ pathology [
36] or primary age-related tauopathy (PART) [
37].
One limitation of the current study is the cohort size. While 11 PD-PDD patients with LRRK2 mutations is a small group, it appears from our clinical and neuropathological assessment that these cases have captured the heterogeneity present in the larger LRRK2 PD population, allowing us to understand general patterns of pathology. A second limitation of this study is the possible ascertainment bias from recruiting individuals from neurological disease centers. While this may lead to under sampling of neurologically normal cohorts, LRRK2 genotyping was done in 94.9% of all individuals in the CNDR database where DNA was available. It will be interesting to see in the future if non-symptomatic LRRK2 mutation carriers harbor any precursors of the neuropathology observed in our cases. One p.G2019S LRRK2 mutation carrier without PD was included in our assessment, and had no apparent neuropathology. A final limitation is that the high prevalence of AD tau pathology could mask other, milder forms of tau pathology. Though we looked closely for non-AD tau in the substantia nigra, it is possible that non-AD tau is more prominent in other regions, including those that were not examined as part of the current study.
Beyond a neuropathological phenomenon, the prevalence of AD-type tau in
LRRK2 PD and iPD has important implications for modeling, diagnosis, and therapeutic treatment of PD. While much animal research has focused on investigating α-synuclein pathology in LRRK2 mutant animals, hyper-phosphorylated tau has been a consistent phenomenon observed in several LRRK2 mutant mouse lines [
38‐
40]. Further, recent work has suggested that mutant LRRK2 may enhance spread of virally-expressed tau protein [
41]. Our description of AD-type tau as a prominent pathology in
LRRK2 mutation carriers suggests that tau hyperphosphorylation may be an important component of
LRRK2-related pathogenesis, and further investigation of the relationship between LRRK2 kinase activity and tau pathology in animal models is warranted.
Diagnosing PD and assessing therapeutic efficacy have been limited by the lack of reliable biomarkers for PD pathology. However, biomarkers are available for AD tau, including positron emission tomography (PET) imaging ligands [
42] and cerebrospinal fluid tau and Aβ levels [
43]. The presence of AD tau in iPD and
LRRK2 PD could serve as diagnostic and prognostic tools to segregate cases that will or will not progress to dementia, although further work is needed to validate tau as a biomarker in PD [
44,
45]. Further, if tau pathology is modulated by the same biological mechanism as α-synuclein pathology, which may be the case for
LRRK2 PD, tau pathology could be used as a clinical trial endpoint measure of interest for therapy trials. Finally, if elevated LRRK2 kinase activity is directly modulating tau pathology, then LRRK2 inhibitors, which are already in clinical trials for PD, may modulate tau pathology in other primary tauopathies. However, whether LRRK2 kinase activity directly modulates tau pathology is not known, and future studies are warranted to investigate the role that LRRK2 plays in tau pathogenesis and whether LRRK2 inhibitors could serve as a potential therapeutic for tau pathology.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.