1. Protein Clearance System Impairment in PD-Ubiquitin-Proteasome System (UPS)
For misfolded proteins to be cleared, ubiquitin ligases are required to tag them with ubiquitin molecules to signal proteolysis. These ubiquitylated proteins are then handed over to the 26S proteasome—the most used proteasome in mammals with a 20S catalytic core and 19S regulatory particles at both ends that break misfolded proteins into small peptides. Non-proteasomal endopeptidases and aminopeptidases then digest these small peptides into monomeric amino acids for reused in new protein synthesis
[1].
Failures in the ubiquitin-proteasome system (UPS) in Parkinson’s disease (PD) are reported in both familial and idiopathic cases (iPD) cases. In familial cases, an E3 ubiquitin ligase, parkin, is associated with early-stage PD (<30 years)
[2]. Interestingly, another member of E3 ubiquitin ligase, thyroid hormone receptor-interacting protein 12 (TRIP12), which accumulates in sporadic PD brains, degrades glucosylceramidase (GCase) enzymes and controls their turnover
[3]. Postmortem analysis of nigral tissues has shown a 30% deficiency in proteasome enzymatic function in iPD cases compared to healthy controls
[4]. Though the mechanism of this proteasome enzymatic dysfunction is unknown, the selective loss of 20S core α-subunits restricted to dopaminergic neurons of the SN in sporadic PD could be a possible explanation
[5]. Taken together, UPS dysfunction in iPD has the potential to explain the pathologic mechanism underlying protein accumulation and Lewy bodies formation in the PD brain, however, more questions are yet to be addressed.
2. Protein Clearance System Impairment in PD-Autophagy-Lysosome Pathway Fails in Clearing Sphingolipids
Autophagy is a conserved lysosomal degradation process for selective removal of damaged subcellular organelles (mitochondria, peroxisomes, endoplasmic reticulum), midbody ring structures, ribosomes and aggresomes
[6]. This process requires lysosomes to remove unwanted material in a so-called autophagy-lysosome pathway and has a critical role in PD pathology
[7]. Indeed, some PD-related genes (SNCA, VPS35, Parkin, PINK1, FBXO7, ATP13A2 and GBA) are involved in autophagy processes and their abnormal levels are implicated in the PD brain
[8]. Lysosomes are hosting multiple hydrolyses as the ultimate hub for the degradation of various types of macromolecules including lipids and, as such, the autophagy-lysosome pathway is involved in orchestrating lipid metabolism
[9]. Groups of lipids such as eicosanoids, endocannabinoids, oxysterols, fatty acids and sphingolipids with cell-regulatory characteristics are classified as bioactive lipids and deemed as key players in PD pathophysiology
[10]. Sphingolipids are a member of bioactive lipids with a reciprocal effect on the autophagy process in PD. Accordingly, defective sphingolipid metabolism/catabolism pathways are reported through a few clinical studies on PD subclasses.
Sphingolipid pathway impairment is not restricted to only PD neurons or PD CSF but seems to be a systemic feature of PD evidenced by reduced GCase activity in blood and fibroblasts
[11], which are invaluable for non-invasive biomarker purposes. The GCase activity is pathologically informative in familial PD (PD-GBA) and research is underway to elucidate its diagnostic potential in iPD.
2.1. Is GCase Activity a Potential Biomarker for iPD?
The GCase enzyme is a 536 amino acid length lysosomal-membrane-anchored protein that catalyzes glucosylceramide into glucose and ceramide
[12], and glucosylsphingosine into glucose and sphingosine
[13]. Mutations in the GBA gene are implicated in different types of Gaucher disease (GD) (type 1, 2 and 3) and PD. In GD, homozygous mutations in the GBA gene cause insufficient activity of GCase, which is the gold standard marker for GD diagnosis
[14][15]. Heterozygous mutations in GBA that cause haploinsufficiency in GCase activity are the greatest risk factor for PD
[16][17]. Across different studies, 5–25% of iPD are carriers of these mutations, which are associated with a 5–20-fold increase in developing PD compared to non-carriers
[16][18]. There are some differences in GBA mutations for GD and PD. For instance, some GBA alterations that are considered “mild mutations” and do not cause GD (E326K (p.E365K) and T369 M (p.T408M), still predispose carriers to parkinsonism
[19][20]. GBA heterozygous mutations are clinically associated with early-onset PD with severe cognitive symptoms
[16][17]. PD age of onset correlates with GBA mutation status and PD cases with homozygous mutations manifest the disease approximately 6–11 years earlier than in heterozygotes; however, the overall PD incidence is comparable
[21].
In definitive PD cases, which are diagnosed by postmortem confirmation of αSyn accumulation into Lewy bodies, reduced GCase activity is reported in both GBA mutation carriers and iPD
[22][23]. In these independently conducted studies on the SN tissues of wild type PD-GBA (WT) cases, the GCase enzyme catalytic activity is reduced as well as its mRNA and protein levels
[22][23][24]. There is strong and emerging evidence that mutations in the GBA gene can cause PD and, additionally, that GCase activity is altered in idiopathic disease without mutation. The mutation frequency in healthy cases is still lower than in PD. In a study with 79 PD and 61 healthy controls, 4.9% of controls and 12.7% of PD are GBA mutation carriers
[25]. In another study conducted in Australia, 18.75% of 48 PD and 6.8% of 44 healthy controls are GBA mutation carriers
[26]. These data suggest that GCase reduced activity is not merely due to mutations and other mechanisms regulating GCase trafficking, and activation should be explored. Due to the prognostic value of reduced GCase activity and its implication in definitive iPD, several studies asked the question of whether GCase could be a potential biomarker for iPD diagnosis by looking at different biospecimens; however, the answer remains inconclusive. Analysis of CSF specimens, that interchange with interstitial fluid in the brain, has indicated GCase activity as non-informative
[27][28][29], while others associate reduced GCase activity with PD irrespective of GBA mutation status
[29]. Blood interchanges with CSF through the blood–CSF barrier and is expected to be a partial proxy for detecting changes in the brain and CSF. Data coming from blood samples (serum, plasma and dried blood spots) are also inconsistent. Though GCase activity is reduced in the serum of PD-GBA and PD wild-type
[27], in dried blood samples, for example, reduced GCase activity is not necessarily associated with PD risk
[30]. This discrepancy could be due to the subjective nature of the clinical diagnosis of selected PD cases for analysis, biospecimen sourced monocytes versus dried blood samples or different strategies in evaluating the GCase activity
[31]. Altogether, though PD diagnosis by GCase activity in isolation remains inconclusive, its prognostic value holds the hope for increasing diagnosis accuracy by incorporating GCase with upstream or downstream regulators such as saposin C (SapC) and post-translational modifications. This would provide stronger evidence for an altered pathway, from GCase transcription to activity, and be less reliant on a single marker.
3.2. GCase from Birth to Death—What Can Go Wrong?
GCase enzyme activity depends on several regulatory mechanisms including structure-functional changes, pH and the availability of its activator SapC; disruptions in any of these steps could potentially affect downstream GCase activity. The GBA gene located at 1q21 encodes the lysosome-membrane-anchored GCase enzyme and its mutations are implicated in iPD
[15]. After the GCase protein is synthesized in the endoplasmic reticulum (ER) it becomes glycosylated (particularly at Asn19), which is deemed crucial for its catalytic activity
[32]. One possible mechanism for the reduced downstream activity is the disruptions in the structure-functional changes of GCase impacting intercellular trafficking, interaction with SapC and substrate stability, all of which could underlie GCase loss-of-function in wild type carriers
[33]. Currently, 20–30% of carrier diagnoses in GD by GCase activity are false positives and false negatives and introducing an activity-dictating structural pattern for GCase could also be useful for GD carrier diagnosis
[14]. There is no currently published evidence exploring the impacts of GCase structural changes on catalytic activity in iPD and further studies are warranted.
3.3. Saposin C and GCase
The prosaposin gene (PSAP) encodes a precursor protein that forms into sphingolipid activator proteins (SAPs)—saposin A-D—that facilitate lysosomal hydrolysis of sphingolipids
[34]. In lysosomes, SapC interaction with GCase alters GCase catalytic activity by inducing conformational changes
[35]. Furthermore, this interaction stabilizes the GCase enzyme against protein clearance systems
[36]. Accordingly, among many mutations reported in GBA, two common and clinically relevant mutations in GD, N270S and L444P, are suggested to destabilize the SapC and GCase complex
[37]. Full sequence analysis of PSAP in patients suffering rapid eye movement sleep behavior disorder—a sign of the prodromal stage of synucleinopathies such as PD
[38]—report an association of three different and rare mutations with idiopathic rapid eye movement sleep behavior disorder
[38]; however, full sequencing of PSAP was not found to fully support a role for PSAP or SapC in PD patients
[39]. Regardless of the current argument on the contribution of PSAP mutations to PD, its impaired function could be indirectly inferred from GCase function, ideally from GCase substrates levels as its activity analysis in isolation could be impacted by external factors.
3.4. Do GCase Substrates Differentiate iPD?
The inconsistency of data on GCase activity is also represented at the substrate level. Here, the conflict starts from brain data, with some reports of no accumulation of GCase substrates (glucosylceramide and glucosylsphingosine) in the PD-GBA brain
[40], yet the accumulation of glucosylsphingosine has been seen in specific age groups of iPD
[41]. However, in a recent study on CSF, a high level of glucosylceramides is reported
[27]. Though these limited data on various brain and CSF specimens do not support a direct association between glucosylceramide level and PD
[42], further studies on a broader scale are warranted.
The GCase enzyme interacts with sphingolipid pathways and changes in specific substrates could have variable indications in PD diagnosis; therefore, there is growing interest in a comprehensive analysis of sphingolipid metabolism for biomarker discovery.
3. Sphingolipids in PD Diagnosis
Post-mitotic neuron survival is partly governed by sphingolipids through modulating the autophagy process
[43]. Autophagy eliminates aggregated proteins and its impairment is reported through a few clinical studies on PD subjects
[44][45][46]. Sphingolipids are becoming linked to the core pathological feature of PD, αSyn aggregation. This is specifically supported by a recently proposed bidirectional link between GCase deficiency and Lewy body formation
[47][48]. Though the exact molecular mechanism is still unidentified, the data is in line with the role sphingolipids play in neuronal survival, exemplified by the monosialoganglioside GM1. GM1 is a member of glycosphingolipids that binds and assists in αSyn structural stability and its deficiency has been associated with αSyn aggregation and neurodegeneration
[49][50]. GM1 and αSyn interactions are supported through several clinical studies. GM1 level deficiency in the PD brain is not only restricted to substantia nigra pars compacta (SN), but it is also reduced in the less affected part of the PD brain (occipital cortex), and systemically in CSF and blood across different clinical studies. Interestingly, the level of GM1′s metabolic precursor, GD1a
[51], shows the same changing trend in PD tissues. The lower GM1 and its precursor levels could predispose dopaminergic neurons to degeneration since it is essential for proper glial cell-derived neurotrophic factor (GDNF) signalling that protects dopaminergic neuron survival
[52].
Ceramides are the precursor of complex sphingolipids, which are implicated in PD pathology
[53] and cognitive decline
[54], nevertheless, data on the diagnostic potential of ceramides is inconsistent across different studies on PD brain, CSF and blood or plasma. This discrepancy could be explained by different GBA mutation status of patients and controls, sample processing (blood versus plasma), involvement of the tissue of study in PD pathogenesis (less involvement of occipital cortex compared to anterior cingulate) or the time of sampling (antemortem vs. postmortem CSF sampling). Therefore, further studies are required to standardize the diagnostic potential of ceramides.
The data on other components of the sphingolipid pathway is also limited; however, reduced activity of the enzymes involved in this pathway, and in some cases, the accumulation of corresponding substrates, introduce a systemic disruption in sphingolipid metabolism in PD subjects
[23][27][55][56][57][58]. This pathway dysregulation in definitive PD is partially represented in peripheral biospecimens such as CSF and blood. Here, GM1, GD1a levels, galactosidase activity, and to some extent, GCase activity and its substrate glucosylceramide, could successfully differentiate PD cases even in CSF and blood. These data suggest that monitoring sphingolipids in CSF and blood, which are both available antemortem and have lower degrees of invasiveness, have the potential to increase iPD diagnosis.
Sphingolipids’ pathway dysregulation is becoming indicated in prodromal PD. The GBA gene mutations are listed as an intermediate-strength genetic factor for prodromal PD, where the total PD risk in mutation carriers at age 65 is 9 times higher (18% in PD versus 2% in age-matched controls)
[59]. Interestingly, serum levels of the sphingolipids are also indicated in RBD cases, which are considered a sign of prodromal PD. A meta-analysis of 1280 RBD patients with an average follow-up of 4.6 years reports that 14.56% of RBD patients develop PD
[60]. Though the GlcCer is not changed, a significant reduction in the level of lactosylceramide, Gb3, Gb4, GM2, GM3, GM1a and GD1a is reported compared to healthy and PD controls
[57].
Altogether, data on using sphingolipid metabolism as a diagnostic tool are promising but limited, and future studies with large patient populations and stratified controls could help confirm the specificity of different parts of the sphingolipid catabolism pathway and their respective diagnostic value for iPD.
4. Would Integrating Biomarkers Make iPD Diagnosis Possible?
Integrating metabolites representing iPD pathophysiology with lower specificity and sensitivity could enhance the accuracy of iPD differentiation from other parkinsonism. There is no published evidence using this approach, yet preliminary data on biomarker integration support enhanced diagnostic accuracy. For instance, a combination of β-glucocerebrosidase, cathepsin D and β-hexosaminidase improves diagnostic accuracy—from an area under the curve value of 0.72; with a sensitivity of 0.67; and specificity of 0.77 for only GCase activity, to an area under the curve value of 0.77; sensitivity, 0.71; and specificity, 0.85 for the combination in differentiating PD from healthy controls
[29]. This improved diagnostic accuracy could be further increased by incorporating other markers that have similar behavior in differentiating PD from controls such as other lysosomal enzymes (GCase, acid sphingomyelinase; aSMase, acid-alpha galactosidase; GLA, acid alpha-glucosidase; GAA, galactocerebrosidase; GALC) and their relative activities, with and without association with each other
[56].
Integrated biomarker set accuracy can also be further improved by incorporating markers from other generalist and specific pathways that represent the same trend of changes in CSF and/or serum as those seen in the PD brain such as 3-hydroxykynurenine
[61][62], apolipoprotein D
[63][64] and potentially neurofilament light chain
[65]. With the same hypothesis, the diagnostic accuracy of multiple biochemical markers could be increased by combining them with imaging data. The data measured through different assays and research conditions could be combined by converting them to the percentage of changes. Expressing metabolic changes as a percentage of changes would also allow different types of assays to be compared to each other. Finally, an iPD specific biomarker signature could be used for both differential diagnosis of iPD between its subtypes, and inform prognosis, considering the different weightings of each marker.