The in vivo recognition of the underlying pathology in neurodegenerative disorders is the initial step towards the development of disease-modifying, protein-targeted treatments. To this end, CSF Aβ₄₂, τ_Τ and τ_P-181 are established AD biomarkers, which can predict an underlying Alzheimer’s disease pathology with high diagnostic accuracy. Biomarkers of other proteinopathies (e.g., synucleinopathies, TDP-43 proteinopathies) are presently lacking.

CSF α-syn species have been studied to this end, including t-α-syn, and, more importantly, o-α-syn and pS129-α-syn. Most studies in the field have focused on differences of α-syn levels and species among synucleinopathies, or between synucleinopathies and AD or PSP, reporting lower t-α-syn and higher pS129-α-syn and o-α-syn values in synucleinopathies [7,8,10,15,16,28,33]. α -Syn oligomers were increased in plasma of PD and brain extracts of DLB patients [34,35]. A comprehensive evaluation of different α-syn species in a variety of movement and cognitive disorders is presently lacking.

The present study examined CSF α-syn species (t-, o- and pS129-α-syn) in a variety of neurodegenerative disorders of presumed underlying proteinopathy. These included synucleinopathies (PD, MSA), tauopathies (PSP, CBD) and mixed tauopathy with amyloidosis (AD). Moreover, we further included patients with FTD (either 3R-tauopathy or TDP-43-proteinopathy) and VD, as paradigms of diseases with neuronal death due to neurodegeneration and ischemia, respectively.

In our cohort, CSF t-α-syn was numerically lower and the pS129-α-syn/α-syn ratio higher in patients with PD and MSA compared to other study groups. These differences reached statistical significance when patients were grouped based on their presumed underlying proteinopathy (i.e., synucleinopathy vs. tauopathy), with suboptimal specificity. Importantly, when patients with CSF blood contamination were excluded, these differences in α-syn and pS129-α-syn/α-syn ratio were more pronounced, providing improved sensitivity and specificity for a synucleinopathy diagnosis. Still, it has to be noted that o-α-syn and pS129-α-syn were not different between synucleinopathies and tauopathies, and that the main determining factor of differentiation between the two groups was t-α-syn. This index may prove useful, in conjunction with other markers, in the differentiation of PD and MSA on the one hand, and PSP and CBD on the other, a distinction that is not always easy on clinical grounds.

Our findings are in agreement with the literature, with most relevant studies reporting a decrease in α-syn and an increase in pS129-α-syn and o-α-syn in PD [7,8,9]. Despite these statistically significant differences, there is considerable inter-subject overlap in α-syn species values among synucleinopathies and other proteinopathies. Thus, at present, CSF α-syn species cannot be utilized as clinically relevant biomarkers for the differentiation of synucleinopathies from other disorders, due to their suboptimal diagnostic accuracy. This is partly due to the great variability of α-syn species among patients with PD, as noted elsewhere [15].

Importantly, there were no significant t-α-syn alterations in non-synucleinonopathic disorders in our cohort (i.e., AD, FTD, VD, PSP and CBD). This is in general agreement with the literature [7,10,16]. An increase in t-α-syn levels in AD vs. control subjects has been previously reported [19]. Moreover, there are neuropathological data that support the colocalization of α-syn and tau oligomers [36], whereas pS129-α-syn correlated with Braak stage and Ab levels in another study [37]. A correlation between tau species and α-syn has also been reported in AD [19]. In the present study, there were no correlations between α-syn species and classical CSF biomarkers in the AD cohort. However, in the absence of a control group, CSF α-syn species alterations in non-synucleinopathies could not be further studied in this cohort.

The significant improvement of diagnostic accuracy of α-syn and the pS129-α-syn/α-syn ratio after excluding patients with CSF blood contamination (i.e., >50 RBC) in our cohort, re-emphasizes the need for standardization of pre-analytical factors in CSF biomarker studies [29]. CSF blood contamination in particular is a major confounding factor in α-syn level determination, as has been reported before [38]. The disparity among studies regarding CSF α-syn species could in part be attributed to this diversity of peri-analytical factors. These factors may include time of lumbar puncture, type of tubes used for CSF collection, blood contamination of CSF, CSF samples storage and shipment protocol, sample thawing procedure etc. [39]. More importantly, there are differences in analytical factors, such as the type of capture and detection antibodies applied [40].

There have been reports regarding correlations of biomarkers with clinical features. More specifically, it has been postulated that pS129-α-syn correlates with clinical severity in PD, as measured by the UPDRS [8]. An increase in pS129-α-syn with PD disease progression has also been supported [41]. Moreover, o-α-syn correlated with PD motor severity in another study [19]. Phosphorylated and oligomeric α-syn species may vary, depending on disease stage, in PD (i.e., preclinical vs. clinical vs. PD with dementia), raising the question of whether these α-syn species are in fact state or stage markers [18]. This correlation with disease stage was also evident in asymptomatic LRRK2 mutation carriers, as well as patients with idiopathic REM sleep behavior disorder (iRBD) [42,43,44]. In the present study, CSF α-syn species did not correlate with any clinical feature (i.e., disease duration, UPDRS). This is in agreement with a study by Foulds et al., where α-syn exhibited stability over PD disease progression [15]. However, only longitudinal studies, with multiple α-syn species determinations over different PD disease stages, could provide robust data regarding possible correlations between α-syn species and PD severity.

The main focus of this study was the significance of α-syn species as possible biomarkers in neurodegenerative disorders. However, in the cohort of patients with movement disorders, CBD patients had significantly higher τ_T values compared to PSP and τ_Τ/Aβ₄₂ ratios compared to PD and PSP. A relative increase in τ_T in CBD has been reported previously. The relative increase in τ_T in CBD is difficult to interpret, particularly when considering that no significant CSF τ_T alteration is evident in PSP, another 4R-tauopathy [30,45].

This study has certain limitations. Initially, the number of patients per study group is suboptimal. This is expected in a single-center cohort, due to the rarity of some of these disorders (e.g., CBD) and due to methodological issues (i.e., CSF blood contamination). Studies of larger cohorts could provide more robust information. Secondly, the absence of a control group precludes the direct comparison of study groups to healthy controls. The inclusion of a control group could further assist in disentangling the complex interplay between CSF biomarkers, clinical phenotype and underlying pathology. Lastly, all diagnoses were based on clinical criteria. To optimize the diagnostic accuracy, all patients fulfilled probable diagnostic criteria and were followed up to exclude patients with atypical clinical features. Moreover, all patients with a non-amnestic phenotype and typical CSF AD profile were excluded, thus minimizing the possibility that the included subjects represented atypical species of AD.

Despite these shortcomings, the present study provides data on CSF α-syn species in a variety of neurodegenerative disorders of diverse underlying pathologies and investigates any possible correlations of them with classical CSF AD biomarkers and clinical features. Larger studies with neuropathological data are needed to establish the relationship of CSF biomarkers, clinical phenotype and the underlying pathology.