PD has been increasingly associated with the lysosome after the discovery of several autosomal dominant and recessive genes associated with PD correlated with other genes important for the functioning of this organelle. The aggregation of α-syn has been reported in the brain of patients and animal models of different LSDs, including GD, Fabry disease, Niemann-Pick disease, Sandhoff disease, Tay-Sachs disease, MLD, β-galactosialidosis and GM2 gangliosidosis, Krabbe’s disease, and Neuronal Ceroid Lipofuscinosis [9].
Elevated ChT activity has already been described in other neurological disorders including stroke, multiple sclerosis and cerebral adrenoleukodystrophy (C-ALD) [11]. ChT levels even correlated with the degree of severity of white matter lesions which corroborates to the importance of macrophages and astrocytes in the natural immune response of these demyelinative lesions [11, 12]. The more severe involvement of the brain in MLD might be the responsible for the higher levels of ChT found in this group in comparison to the GD patients. Higher levels of Cht in CSF were also demonstrated in other LSDs with more severe neurological involvement than GD1as GM1-gangliosidosis and GM2-gangliosidosis, for example [13]. Another possible explanation for this finding is that the GD group is treated with ERT which is already extensively associated with the lowering of ChT levels [14]. However, as this treatment doesn’t cross the brain blood barrier and ChT is not clearly demonstrated to breach it, the ERT role on the lower levels of ChT in the GD group still needs to be further elucidated [15].
Lower HEX activity was found in the GD group when compared to MLD patients in our cohort. In a study that analyzed α-syn, phosphorylated tau protein and lysosomal enzymes in CSF of 71 PD patients compared with 45 neurological controls, CSF GCase activity was significantly decreased, and HEX activity was moderately increased in the PD group [16]. HEX is associated with the Sandhoff disease (SD) caused by the accumulation of ganglioside GM2 and with shared features with PD including the α-syn inclusions frequently found in the SD brain of both patients and mouse models [17, 18]. Interestingly, a new study that biochemically characterize α-syn changes in the transgenic HEX knockout mice model showed that HEX expression decreases α-syn association to lipid compartments and prevented nigrostriatal accumulation of AAV6-overexpressed α-syn [18]. To the best of our knowledge this is the first time that low HEX is described in the CSF of GD patients. We hypothesize that a lower HEX activity in GD might also be contributing to the increased risk of PD found in these patients; however, a bigger sample of patients is necessary to establish a better association.
We hypothesize that the usefulness of CSF endolysosomal enzyme activity levels as biomarkers of PD, and their role in the prodromal period of PD, are likely to be of greater value when used in combination with measures of other substrates, such as glycosphingolipids and complex gangliosides. We could not find any explanation in the literature for the increase of Beta-glucuronidase seen in the GD group, as this enzyme was mostly associated with bacterial meningitis [19]. Currently, there is no treatment available that can stop the progression of PD. However, new studies suggest that re-establishing normal levels of GCase, which has been reported to be reduced in PD patients, may be a treatment option for patients carrying GBA variants, as demonstrated by some animal models [18, 20]. One limitation of our study was the availability of few lysosomal enzymatic assays and the restricted number of GD CSF samples, which did not allow for a proper assessment of the lysosomal environment in GD patients as intended.
The NMS assessment allows for a more complete follow-up of GD patients, as it was recently reported that most GD patients would like to be informed about their PD risk in the clinical setting by the physician in charge at the time of GD diagnosis [21]. After a three-year follow-up, it was interesting to note that the patient with the highest number of NMS and the lowest SST score was the one who developed PD after two years (P17). In a study that performed a six-year longitudinal evaluation of 31 GD1 patients, the GD subjects had significantly worse scores in UPSIT (Pennsylvania Smell Identification Test), UMSARS (Unified Multiple System Atrophy Rating Scale), MoCA, and MDS-UPDRS III compared to controls, with hyposmia being a predictive factor for deterioration in this population. Due to the importance of NMS in our cohort, we advocate for the routine assessment of smell, even by directed anamnesis, as multiple tools can be used in clinical practice to assess hyposmia [23].
In another two-year follow-up study, it was shown that RBD and UPDRS III scores were significantly worse in GD patients and PD-GBA variants compared to controls at baseline [24]. On the other hand, olfactory and cognitive assessments were lower in the GD group and did not differ significantly from baseline [24]. Although we followed up our patients for almost the same period, no consistent pattern was observed between both cohorts. One hypothesis for the lack of difference in our NMS follow-up compared to Avenali's and Beavan's cohorts might have been the older mean age of these cohorts (mean ages = 52 and 61 years, respectively) compared to ours [22, 24].
It was intriguing to observe that ESS significantly improved, and the BDI score was not worsened during the follow-up, which happened during the SARS-CoV-2 pandemic, as these were described as common during this period [25, 26]. We hypothesize that the changes in habits during the pandemic, mainly working habits (including less time spent in commute due to working from home), might be responsible for these findings.
The term "variant severity" has been used in the literature to define the effect of the variant on the phenotype of patients with GD, classifying them as "severe" (e.g., 84 GG, V394L, D409H, L444P) or "mild" (e.g., N370S, R496H) [27, 28]. However, this terminology has been criticized because the pathogenic variant N370S, which is considered "mild," has great phenotypic heterogeneity. Homozygotes can remain asymptomatic for decades, while others can present severe clinical manifestations in childhood, including hepatosplenomegaly and pancytopenia [29]. It has been reported that motor, cognitive, olfactory, and psychiatric symptoms are more severe in PD carriers of "severe" GBA variants compared to those carrying mild or no GBA variants [27]. We could not find any difference in the NMS assessment between patients presenting at least one L444P allele and the rest of the cohort. Our single patient presenting motor symptoms (P17) only had "mild" variants, which leads to the hypothesis that other factors influence the neurodegeneration of PD beyond the patient's genotype.