Rotenone Blocks the Glucocerebrosidase Enzyme and Induces the Accumulation of Lysosomes and Autophagolysosomes Independently of LRRK2 Kinase in HEK-293 Cells

Parkinson’s disease (PD) is a neurodegenerative disorder caused by the progressive loss of dopaminergic (DAergic) neurons in the substantia nigra and the intraneuronal presence of Lewy bodies (LBs), composed of aggregates of phosphorylated alpha-synuclein at residue Ser129 (p-Ser129α-Syn). Unfortunately, no curative treatment is available yet. To aggravate matters further, the etiopathogenesis of the disorder is still unresolved. However, the neurotoxin rotenone (ROT) has been implicated in PD. Therefore, it has been widely used to understand the molecular mechanism of neuronal cell death. In the present investigation, we show that ROT induces two convergent pathways in HEK-293 cells. First, ROT generates H2O2, which, in turn, either oxidizes the stress sensor protein DJ-Cys106-SH into DJ-1Cys106SO3 or induces the phosphorylation of the protein LRRK2 kinase at residue Ser395 (p-Ser395 LRRK2). Once active, the kinase phosphorylates α-Syn (at Ser129), induces the loss of mitochondrial membrane potential (ΔΨm), and triggers the production of cleaved caspase 3 (CC3), resulting in signs of apoptotic cell death. ROT also reduces glucocerebrosidase (GCase) activity concomitant with the accumulation of lysosomes and autophagolysosomes reflected by the increase in LC3-II (microtubule-associated protein 1A/1B-light chain 3-phosphatidylethanolamine conjugate II) markers in HEK-293 cells. Second, the exposure of HEK-293 LRRK2 knockout (KO) cells to ROT displays an almost-normal phenotype. Indeed, KO cells showed neither H2O2, DJ-1Cys106SO3, p-Ser395 LRRK2, p-Ser129α-Syn, nor CC3 but displayed high ΔΨm, reduced GCase activity, and the accumulation of lysosomes and autophagolysosomes. Similar observations are obtained when HEK-293 LRRK2 wild-type (WT) cells are exposed to the inhibitor GCase conduritol-β-epoxide (CBE). Taken together, these observations imply that the combined development of LRRK2 inhibitors and compounds for recovering GCase activity might be promising therapeutic agents for PD.


Introduction
Parkinson's disease (PD) is a neurodegenerative disorder characterized by motor symptoms such as bradykinesia, rigidity, resting tremor, and gait disturbance [1]. PD is mainly caused by the progressive loss of dopaminergic (DAergic) neurons in the substantia nigra and the intraneuronal presence of Lewy bodies (LBs), composed of aggregates of alphasynuclein (α-Syn) [2]. Although initially described in six subjects [3], PD has reached pandemic proportions [4]. Indeed, it is projected that 12 million people will suffer from the neurologic disorder by 2040 [5], mainly affecting the population under 65 years of age [6]. Unfortunately, no curative treatment is available yet. To further aggravate matters, the etiopathogenesis of the disorder is still unresolved. Despite this drawback, mitochondrial damage, oxidative stress (OS), and alteration in the autophagy-lysosomal pathway (ALP) have been among the most known genetic risk factors for the development of PD [34]. Indeed, there is accumulating evidence that a buildup of GlcCer due to the dysfunction of GCase can also increase the accumulation of α-Syn [35,36]. Furthermore, LRRK2 kinase has been suggested to regulate GCase levels and enzymatic activity differently depending on the cell type in PD [37]. Since the GBA1 pathway might be convergent to LRRK2 and α-Syn, GBA1 has become a potential therapeutic target to slow PD [38]. However, the mechanism by which LRRK2 and α-Syn are associated with dysfunctional GCase, autophagy, and cell death has not yet been fully established.
Autophagy is a complex process that involves the fusion of autophagosomes and lysosomes to form the autophagolysosome to remove superfluous and damaged organelles (e.g., dysfunctional mitochondria) and cytosolic proteins [39]. Autophagy appears as a protective mechanism in response to stress [40,41], and it may or may not be associated with cell death, depending on the intensity of the insult. Such dynamic flux in the formation of the autophagy-lysosome can be modulated by inhibitors such as bafilomycin A1 (BafA1), which inhibits vacuolar-type H + -ATPase [42], and chloroquine (CQ), which blocks autophagosome fusion with the lysosome and slows down lysosomal acidification [43,44]. Interestingly, it has been reported that ROT blocks autophagic flux prior to inducing cell death [45]. However, it is not entirely clear if LRRK2 kinase is involved in such altered processes in cells under OS. Moreover, the inhibition of LRRK2 kinase activity results in increased GCase activity in DAergic neurons derived from PD patients with either LRRK2 or GBA1 mutations [46]. Yet, it is not yet known whether null LRRK2 may have a similar effect on HEK-293 cells exposed to ROT.
To acquire an understanding of these issues, the present investigation aimed to investigate the effect of the inhibitor of GCase conduritol-β-epoxide (CBE) and ROT on HEK-293 related to LRRK2, α-Syn, autophagy, and apoptosis. To achieve this aim, HEK-293 LRRK2 WT cells and HEK-293 LRRK2 knockout (KO) cells were used. By means of different techniques of biochemistry, immunofluorescence microscopy, and flow cytometry, we found that CBE and ROT inhibited the enzymatic activity of GCase to a similar extent. Interestingly, ROT and CBE induced a high accumulation of lysosomes and autophagolysosomes in HEK-293 cells, but ROT diminished ∆Ψ m , induced p-Ser 935 LRRK2 concomitantly with p-Ser 129 α-Syn, and induced DJ-1Cys 106 SO 3 and CC3 in those cells. On the other hand, ROT inhibited GCase in HEK-293 LRRK2 KO cells. Consequently, it induced a high accumulation of lysosomes and autophagolysosomes but was ineffective in triggering damage to ∆Ψ m , p-Ser 395 LRRK2p-Ser 129 α-Syn, DJ-1Cys 106 SO 3 , and CC3. Taken together, these results suggest that ROT decreases the activity of GCase, induces mitochondrial damage, phosphorylates LRRK2, which, in turn, phosphorylates α-Syn, triggers the concomitant accumulation of lysosomes and autophagolysosomes, and causes signs of OS and apoptosis in HEK-293 cells. Of note, ROT impairment of ALP and the induction of apoptosis occur in an LRRK2-independent and LRRK2-dependent fashion, respectively, supporting the idea that LRRK2 is a pro-apoptotic kinase in cells under OS stimuli. Therefore, LRRK2 has become a therapeutic target for the treatment of PD.

Rotenone (ROT) Inhibits Glucocerebrosidase (GCase) Activity by Mimicking the Inhibitor Conduritol-β-Epoxide (CBE) in HEK-293 Cells
CBE is a cyclitol epoxide that covalently and irreversibly reacts with the catalytic nucleophile of the lysosomal enzyme GCase and, thus, irreversibly inactivates the enzyme. We, therefore, first evaluated whether CBE inhibits GCase in HEK-293 cells. Effectively, Figure 1 shows that the enzymatic activity of GCase decreased by −62% and −87% in HEK-293 exposed to (10 µM) and (50 µM) CBE, respectively, compared to untreated cells ( Figure 1A). An in silico molecular docking simulation analysis [47] revealed that CBE binds to a pocket in GCase (Protein Data Bank, PDB #6T13, Vina score: −6.0) interacting with at least 15 residues (Table 1), wherein the residue Glu 340 of the protein GCase is the catalytic nucleophile critical for covalent binding with the inhibitor (Figure 1B and inset) [48].

Rotenone (ROT) but Not Conduritol-β-Epoxide (CBE) Induces the Oxidation of Stress Sensor Protein DJ-1 and Cleaved Caspase 3 (CC3) in HEK-293 Cells
It is known that the oxidation of the stress sensor protein DJ-1Cys 106 -SH (sulfhydryl group) into DJ-1Cys 106 -SO 3 (sulfonic acid) is a specific target of the non-radical ROS H 2 O 2 [52]. We, thus, determined whether CBE or ROT can generate H 2 O 2 and induce the generation of CC3. Therefore, HEK-293 cells were exposed to CBE (10 µM) or ROT (10 µM) for 24 h. As shown in Figure    Quantitative analysis of CC3 mean fluorescence intensity (J). The data are expressed as mean ± SD; * p < 0.05; *** p < 0.001; ns-not significant. The histograms, bars, and photomicrographs represent 1 out of 3 independent experiments (n = 3). Image magnification, 20×. White line (area) represents magnification of broken line (area).

Rotenone (ROT) but Not Conduritol-β-Epoxide (CBE) Induces Phosphorylation of Alpha-synuclein (α-Syn) and LRRK2 Kinase in HEK-293 Cells
We wanted to assess whether ROT and CBE trigger the phosphorylation of α-Syn concurrently with LRKK2 in HEK-293 cells. As shown in Figure 5, ROT but not CBE induced a statistically significant increase in p-Ser 129 α-Syn, as detected by flow cytometry ( Figure 5A, +580%), and IMF ( Figure 5B-E, 44.5-f.i.). Previously, it was shown that ROT induced the phosphorylation of LRRK2 in nerve-like cells [26]. Therefore, we evaluated whether ROT and CBE cause p-Ser 935 LRKK2 in HEK-293 cells. Figure 5F shows that ROT induced p-Ser 935 LRKK2 by +1580% in HEK-293 cells, as evaluated by flow cytometry. In contrast, CBE was unable to induce any effect on LRRK2 ( Figure 5F). Similar data were documented by IMF ( Figure 5G whether ROT and CBE cause p-Ser 935 LRKK2 in HEK-293 cells. Figure 5F shows that ROT induced p-Ser 935 LRKK2 by +1580% in HEK-293 cells, as evaluated by flow cytometry. In contrast, CBE was unable to induce any effect on LRRK2 ( Figure 5F). Similar data were documented by IMF ( Figure 5G    The data are expressed as mean ± SD; *** p < 0.001; ns-not significant. The histograms, dot graphs, and photomicrographs represent 1 out of 3 independent experiments (n = 3). Image magnification, 20×.

Rotenone (ROT) Does Not Induce the Phosphorylation of LRRK Kinase in HEK-293 LRRK2 Knockout (KO) Cells
The above finding that ROT induced the phosphorylation of LRRK2 prompted us to expand our observation by inquiring whether ROT could induce p-Ser 395 LRRK2 in HEK-293 LRRK2 KO cells. We, therefore, first confirmed the protein expression status of LRRK2 in both wild-type (WT) HEK-293 and HEK-293 LRRK2 KO cells. Figure 6 shows the expression of LRRK2 in WT HEK-293 cells, but its expression was almost completely reduced in HEK-293 LRRK2 KO cells, according to flow cytometry analysis (−90%, Figure 6A). When both WT and KO cells were exposed to ROT, it was observed that the neurotoxin induced p-Ser 395 LRRK2 by +820% in WT HEK-293 cells ( Figure 6B,C), but this effect was drastically reduced by −96% in HEK-293 LRRK2 KO cells when compared to treated WT cells ( Figure 6B (G″-I″), and merge (G-I) of untreated HEK-293 cells (I) or cells treated with (10 µM) CBE (H) or µM) ROT (I). Quantitative analysis of pS 935 LRRK2 mean fluorescence intensity (J). The data expressed as mean ± SD; *** p < 0.001; ns-not significant. The histograms, dot graphs, and ph micrographs represent 1 out of 3 independent experiments (n = 3). Image magnification, 20×.

Rotenone (ROT) Does Not Induce the Phosphorylation of LRRK Kinase in HEK-293 LRR Knockout (KO) Cells
The above finding that ROT induced the phosphorylation of LRRK2 prompted u expand our observation by inquiring whether ROT could induce p-Ser 395 LRRK2 in HE 293 LRRK2 KO cells. We, therefore, first confirmed the protein expression status of LRR in both wild-type (WT) HEK-293 and HEK-293 LRRK2 KO cells. Figure 6 shows the pression of LRRK2 in WT HEK-293 cells, but its expression was almost completely duced in HEK-293 LRRK2 KO cells, according to flow cytometry analysis (−90%, Fig  6A). When both WT and KO cells were exposed to ROT, it was observed that the neu toxin induced p-Ser 395 LRRK2 by +820% in WT HEK-293 cells ( Figure 6B,C), but this eff was drastically reduced by −96% in HEK-293 LRRK2 KO cells when compared to trea WT cells ( Figure 6B,C). Similar observations were obtained by IMF analysis ( Figure  H). The data are expressed as mean ± SD; ** p < 0.01; *** p < 0.001. The histograms, bars, and photomicrographs represent 1 out of 3 independent experiments (n = 3). Image magnification, 20×.

ROT Inhibits the Enzymatic Activity of GCase Equally in Both WT and HEK-293 KO Cells
In parallel experiments, we wanted to determine whether ROT affects the enzymatic activity of GCase in KO cells. To achieve this end, WT and HEK-293 LRRK2 KO cells were first exposed to ROT for 24 h and then quantified for the percentage of enzymatic activity. As shown in Figure 6I, ROT inhibited the activity of GCase to a similar extent in WT and HEK-293 LRRK2 KO cells.

ROT Inhibits the Enzymatic Activity of GCase Equally in Both WT and HEK-293 KO Cells
In parallel experiments, we wanted to determine whether ROT affects the enzymatic activity of GCase in KO cells. To achieve this end, WT and HEK-293 LRRK2 KO cells were first exposed to ROT for 24 h and then quantified for the percentage of enzymatic activity. As shown in Figure 6I, ROT inhibited the activity of GCase to a similar extent in WT and HEK-293 LRRK2 KO cells.

ROT Induces the Accumulation of Lysosomes and Reduces Mitochondrial Potential in HEK-293 KO Cells
Then, we wondered whether ROT could alter the lysosomal system and damage ΔΨm in KO cells. Flow cytometry analysis revealed that untreated WT and KO cells showed no statistical difference in the percentages of lysosomal accumulation ( Figure 7A  . The data are expressed as mean ± SD; * p < 0.05; ** p < 0.01; *** p < 0.001; ns-not significant. The smooth dot plots, bars, histograms, and photomicrographs represent 1 out of 3 independent experiments (n = 3). Image magnification, 20×.

Rotenone (ROT) Induces An Increase in Autophagosomes Independently of LRRK2
To determine whether ROT affects the autophagosome flux in HEK-293 LRRK2 KO cells, WT and KO cells were exposed to ROT for 24 h. As shown in Figure 8, ROT provoked the accumulation of lysosomes ( Figure 8B) and increased autophagolysosomes ( Figure 8G

Rotenone (ROT) Induces An Increase in Autophagosomes Independently of LRRK2
To determine whether ROT affects the autophagosome flux in HEK-293 LRRK2 KO cells, WT and KO cells were exposed to ROT for 24 h. As shown in Figure 8, ROT provoked the accumulation of lysosomes ( Figure 8B) and increased autophagolysosomes ( Figure  8G

ROT Neither Induces the Phosphorylation of α-Syn, the Oxidation of DJ-1, Nor the Activation of Caspase 3 (CC3) in HEK-293 LRRK2 KO Cells
We wanted to determine the effect of ROT in KO cells regarding α-Syn, DJ-1, and CC3. As shown in Figure 9, ROT was highly efficient, inducing p-Ser 129 α-Syn by +1000% ( Figure 9A,B), oxidizing DJ-1 by +2500% ( Figure 9H,I), and the production of CC3 by +1800% ( Figure 9O,P) in WT HEK-293 cells. In sharp contrast, ROT was strongly ineffective in triggering the phosphorylation of α-Syn ( Figure 9A

ROT Neither Induces the Phosphorylation of α-Syn, the Oxidation of DJ-1, Nor the Activation of Caspase 3 (CC3) in HEK-293 LRRK2 KO Cells
We wanted to determine the effect of ROT in KO cells regarding α-Syn, DJ-1, and CC3. As shown in Figure 9, ROT was highly efficient, inducing p-Ser 129 α-Syn by +1000% ( Figure 9A,B), oxidizing DJ-1 by +2500% ( Figure 9H,I), and the production of CC3 by +1800% ( Figure 9O,P) in WT HEK-293 cells. In sharp contrast, ROT was strongly ineffective in triggering the phosphorylation of α-Syn ( Figure 9A

Discussion
We confirm that CBE significantly reduced the enzymatic activity of GCase (e.g., by −87% at 50 µM) via binding to Glu 340 of the enzyme, as verified by in silico molecular docking analysis [48,53]. Therefore, a deficiency in GCase activity might lead to diminished performance of the enzyme in the lysosome and the disruption of lysosomal targeting. Interestingly, CBE induces the accumulation of lysosomes to a similar extent as the lysosome inhibitors, CQ and BAF, in HEK-293 cells. These observations suggest that CBE is an effective compound to dysregulate the autophagy pathway [54]. Under the present experimental conditions, we found that CBE causes neither mitochondrial potential alteration, triggers OS, as reflected by the non-oxidized sensor protein DJ-1Cys 106 -SOH, impairs ∆Ψ m , induces α-Syn and LRRK2 phosphorylation, nor produces CC3 in HEK-293 cells. However, we do not discard the possibility that CBE under prolonged incubation might reverse the fate of these cellular proteins and mitochondria [55]. Therefore, we were able to separate malfunctioning autophagy processes (i.e., lysosome from autophagosome fusion) from OS and apoptosis. Surprisingly, ROT diminishes the activity of GCase (by −48% at 10 µM ROT), according to the enzymatic GCase test, most probably through its binding to the critical catalytic residue Glu 340 , as predicted by molecular docking analysis. Accordingly, ROT interacts with 14 out of 15 residues similar (94% similarity) to the ones reported for CBE in the catalytic pocket of GCase. Although ROT displays a much more negative Vina score, which is indicative of a strong binding affinity (e.g., −9.2 ROT versus −6.0 CBE), CBE is much more specific towards GCase than ROT, according to the biochemical reduction in GCase activity. Despite this drawback, ROT proves to be highly effective, provoking, in a simultaneous fashion, a significant increase in the oxidation of DJ-1Cys-SH into DJ-1Cys 106 SO 3 , CC3, the phosphorylation of α-Syn and LRKK2 kinase, the accumulation of both lysosomes and autophagolysosomes, and a significant decrease in ∆Ψ m . These data imply that ROT triggers (i) the accumulation of lysosomes and autophagolysosomes, (ii) mitochondrial-dependent OS damage, and (iii) apoptosis in HEK-293 cells.
However, how does ROT link these three processes? Our findings suggest that ROT triggers two alternative and complementary mechanisms, involving the interaction between ROT and GCase and ROT and mitochondrial complex I, which eventually converge on apoptosis. Mechanistically, ROT functions as a strong inhibitor of complex I of the mitochondrial respiratory chain [18] via the inhibition of electron transfer from the iron-sulfur centers in complex I to ubiquinone, leading to a blockade of the I Q site [56], and an overreduction of complex I causes electrons to leak and produce ROS superoxide anion radical (O 2 − ). The O 2 − radical can dismutase into non-radical reactive H 2 O 2 [19], which, in turn, via signaling mechanisms [57,58], oxidizes the stress sensor protein DJ-1, leading to the overproduction of DJ-1Cys 106 SO 3 [52]. Interestingly, oxidized DJ-1 has been proposed as a possible biomarker of PD [59,60]. Alternatively, H 2 O 2 might activate LRRK2 kinase activity by directly enhancing its autophosphorylation, e.g., at Tyr 1967 [61], Ser 2032 , and Tyr 2035 [62,63], or indirectly, via the phosphorylation of Ser 910 and Ser 935 by the inhibition of the nuclear factor-κB (IκB) kinase (IKK) complex [64]. Of note, we found p-Ser 935 LRRK2positive cells in HEK-293 cells exposed to ROT. Interestingly, it has been found that H 2 O 2 increases LRRK2 kinase activity and enhances LRRK2 cell toxicity in HEK-293T cultured cells, mouse primary cortical neuronal cultures [65], and nerve-like cells [26]. Therefore, phosphorylated LRRK2 kinase might directly or indirectly damage mitochondria and trigger the activation of several cell-death-related proteins: (i) LRRK2 directly interacts with dynamin-like protein 1 (DLP1), a key mitochondrial fission protein, increasing its mitochondrial targeting and, thus, promoting mitochondrial fragmentation [66]; (ii) increases the phosphorylation of peroxiredoxin 3, exacerbating OS-induced cell death [67]; (iii) phosphorylates both the activation of apoptosis-signal-regulating kinase 1 (ASK1) [68] and MKK4/MAPK kinase [69], thereby triggering the JNK/c-JUN/PUMA death pathway; and (iv) phosphorylates the pro-apoptotic transcription factor TP53 [70], thus triggering downstream apoptosis signaling. Taken together, these observations suggest that phosphorylated LRRK2 might work as a pro-apoptotic kinase under OS stimuli [26,71]. Furthermore, (v) LRRK2 kinase phosphorylates α-Syn at Ser 129 [72], which is the major component of pathological deposits in PD [73]. In line with this, we found p-Ser 935 LRRK2 concomitant with p-Ser 129 α-Syn-positive cells in HEK-293 cells exposed to ROT. Although the exact mechanism by which α-Syn causes dopamine neuronal loss is unclear, α-Syn has been suggested to interfere with mitochondrial dynamics and promote mitochondrial fragmentation through α-Synand DLP1-dependent mechanisms or by the direct binding of α-Syn to mitochondria, occurring independently of proteins involved in mitochondrial dynamics [66,74]. All these effects lead to an acceleration in the disposal of damaged mitochondria through mitophagy [75] and/or the activation of the pro-apoptosis protein caspase 3 (this work). Additionally, aggregated α-Syn might disrupt phagosome and lysosome fusion [76]. Indeed, α-Syn has been shown to disrupt intracellular trafficking and the lysosomal activity of GCase [77]. Taken together, these observations suggest a self-propagating positive feedback process in which elevated levels of toxic α-Syn lead to a depletion of lysosomal GCase, resulting in a progressive accumulation of GlcCer that promotes the additional formation of toxic α-Syn [78]. However, we found that there is no p-Ser 129 α-Syn in CBE-exposed cells. In agreement with others [79], this observation suggests that the chemical inhibition of GCase activity per se is not sufficient to provoke the phosphorylation and accumulation of α-Syn, and that the inhibition of GCase has to be accompanied by an additional stimulus, e.g., mitochondrial-derived OS, necessary to trigger the phosphorylation of α-Syn via LRRK2. Indeed, α-Syn can directly inhibit lysosomal GCase activity [80,81], or it can indirectly reduce GCase activity by inhibiting its transport from the endoplasmic reticulum to the lysosomes [82]. As expected, however, CBE only increased intracellular granular content and acidic particles, i.e., lysosomes, and augmented autophagy-lysosome fusion.
We hypothesized that HEK-293 cells carrying an LRRK2 null gene would revert to normal the cytotoxic effects associated with ROT-induced mitochondrial damage, lysosomal dysfunction, impaired autolysosome formation, OS marker DJ-1, and apoptosis marker CC3 in HEK-293 LRRK2 KO cells. Effectively, we found that HEK-293 LRRK2 KO cells exposed to ROT exhibit an accumulation of lysosomes and autophagolysosomes only. Several observations support these findings. First, ROT reduced the levels of GCase activity to similar percentage values in WT and KO cells. Second, no significant changes in ∆Ψ m were observed in LRRK2 WT and KO cells. Third, ROT induced almost neither p-Ser 129 α-Syn, oxD-1Cys 106 SO 3 , nor CC3-positive cells in KO cells. Finally, ROT induced a significant increase in the accumulation of lysosomes and autophagolysosomes in WT and LRRK2 KO cells, as reflected by the accrual of the LC3-II marker. Taken together, these results suggest that HEK-293 LRRK2 null cells become resilient to ROT-induced OS and apoptosis, but the mutant cells still suffer from lysosomal and autophagy dysfunctionality, which, under prolonged incubation, may lead to cell death. Interestingly, the pathologic phenotype of HEK-293 LRRK2 KO cells exposed to ROT resembles the phenotype of WT HEK-293 cells exposed to CBE only. Our findings suggest that the dysfunction of lysosomal activity and/or the disturbance of autophagy-lysosome fusion is independent of LRRK2 kinase activity. Thus, these results support the notion that LRRK2 is a critical kinase in the apoptosis pathway and α-Syn is a major mediator of LRRK2-induced toxicity. In contrast to others [46,83], our findings suggest that LRRK2 kinase might not represent a direct regulator of lysosomal GCase, lysosomal function, or autophagy-lysosome fusion. However, we do not discard the possibility that LRRK2 kinase activity affects both the levels and catalytic activity of GCase in a cell-type-specific manner [37]. Further investigation is, therefore, needed to clarify this issue.
HEK-293 cells have been extensively used as an in vitro model system to study PD due not only to their easy handling, reliable growth, and propensity for transfection but also to their amenability to stringent quantitative assessments. Most notably, HEK-293 cells express the typical features of immature neurons, such as the neurofilament (NF) subunits NF-L, NF-M, and NF-H, α-internexin, vimentin, and keratins 8 and 18, and also reveal the expression of mRNAs specific for numerous other genes normally expressed in neuronal lineage cells [84]. Therefore, HEK-293 might qualify as a human neuronal cell line model. Indeed, HEK293 cells provide a reasonable approximation for addressing numerous questions of basic biology in PD. Indeed, HEK-293 cells have demonstrated a clear pro-apoptotic transcriptional response profile similar to that in neurons undergoing apoptosis [85]. Moreover, HEK-293 cells display elements of the autophagy-lysosome that are mechanistically similar to those expressed in DAergic neurons [50]. Interestingly, since LRRK2 is a well-conserved evolutionary gene [86], HEK-293 cells have been used to identify molecular substrates of this kinase [87] and to study LRRK2 mutations' functional analysis [88,89]. HEK-293 cells have also been used to demonstrate that, depending on its concentration, the neurotoxin ROT can induce sublethal and/or lethal effects. For instance, it has been shown that ROT (10 nM) might induce the cytosolic production of H 2 O 2 only in HEK-293 cells [90], whereas, at higher concentrations (e.g., 10 µM), it induces autophagy and apoptosis ( [45,91] and this work). Last but not least, HEK-293 cells have been used to ectopically express not only the human dopamine transporter (hDAT) to study, e.g., the toxic effect of MPTP [92], but also dopaminergic receptors, e.g., D1 [93] and/or D5 [94]. Taken together, all these biological features suggest that HEK-293 cells are a highly promising cellular model to reveal the molecular aspects, as described in the present investigation, of the insidious degenerative disorder PD.

HEK-293 Cell Line
The HEK-293, a specific immortalized cell line derived from a human embryonic kidney, was purchased from AcceGen Biotech (cat #ABC-TC0008, AcceGen Biotech, Fairfield, NJ, USA), and the HEK-293 LRRK2 knockout (OK) cell line was kindly provided by Dr.

GCase Activity Assay
Cellular GCase activity was determined using the Beta-Glucosidase Assay Kit (cat #ab272521, Abcam, Boston, MA, USA) according to the manufacturer's recommendations. Briefly, cell lysates were incubated with p-nitrophenyl-α-D-glucopyranoside, which is hydrolyzed specifically by β-glucosidase into a yellow-colored product (maximal absorbance at 405 nm). The rate of the reaction was directly proportional to the enzyme activity.

Characterization of Lysosomal Complexity
To analyze lysosomal complexity, cells were incubated with the cell-permeable, nonfixable, green, fluorescent dye LysoTracker Green DND-26 (50 nM, cat #L7526, Thermo Fisher Scientific, Waltham, MA, USA) for 30 min at 37 • C. Cells were then washed, and LysoTracker fluorescence was determined by analysis of fluorescence microscopy images in a Floid Cells Imaging Station microscope (Cat# 4471136, Life Technologies, Carlsbad, CA, USA), or flow cytometry using a BD LSRFortessa II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA). The experiment was conducted 3 times, and 10,000 events were acquired for analysis. Flow cytometry analysis for LysoTracker/SSCA was performed by selecting, in the FL-1 channel, all cells with LysoTracker reactivity (>99%), in order to perform the analysis of the total LysoTracker-positive population. SSCA parameter was adjusted to the mean fluorescence of the control (UNT; 40 K ± 3.5 K) plus two standard deviations (i.e., values above 47 K). Quantitative data and figures were obtained using FlowJo 7.6.2 Data Analysis Software (BD Biosciences, Franklin Lakes, NJ, USA).

Analysis of Mitochondrial Membrane Potential (∆Ψm)
The assessment of ∆Ψm was performed according to Ref. [95]. Briefly, cells were incubated for 20 min at RT in the dark, with a deep-red MitoTracker (20 nM final concentration) compound (cat #M22426, Thermo Scientific, Waltham, MA, USA). Cells were analyzed using fluorescence microscopy Floid Cells Imaging Station microscope (cat# 4471136, Life Technologies, Carlsbad, CA, USA), or a BD LSRFortessa II flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA). The experiment was conducted 3 times, and 10,000 events were acquired for analysis. MitoTracker highly positive cells were selected located between 10 4 and 10 6 . No discrimination by complexity was made. Quantitative data and figures were obtained using FlowJo 7.6.

Autophagy Assay
The autophagy assay was performed according to the manufacturer's recommendation (cat #MAK138, Sigma-Aldrich, Saint Louis, MO, USA). Briefly, cells under different treatments were incubated with 1X stain reagent for 20 min. Then, the fluorescence intensity (ex 360/em 520 nm) was measured using a BD LSRFortessa II flow cytometer (BD Biosciences). Twenty thousand events were acquired, and the acquisition analysis was performed using FlowJo 7.6.2 Data Analysis Software (BD Biosciences, Franklin Lakes, NJ, USA).

Molecular Docking
For in silico molecular docking analysis, we used the X-ray diffraction crystallography protein structure of glucocerebrosidase (GCase; Protein Data Bank (PDB) code: 6T13). Blind molecular docking was performed with CB-Dock version 2 [47], a cavity detection-guided protein-ligand blind docking web server that uses Autodock Vina (version 1.1.2, Scripps Research Institute, La Jolla, CA, USA). The SDF structure files of the tested compounds (conduritol-β-epoxide (CBE): PubChem CID 119054; rotenone (ROT): PubChem CID 6758) were downloaded from PubChem. Molecular blind docking was performed by uploading the 3D structure PDB file of GCase into the server with the SDF file of each compound. For analysis, we selected the docking poses with the strongest Vina score in the catalytical pocket. The generated PDB files of the molecular docking of each compound were visualized with the CB-Dock2 interphase.

Data Analysis
In this experimental design, a vial of HEK-293 (WT and LRRK2 KO cells) was thawed and cultured, and the cell suspension was pipetted at a standardized cellular density of 2 × 10 4 cells per cm 2 into different wells of a 24-well plate. Cells (i.e., the biological and observational units) [96] were randomized to wells by simple randomization (sampling without replacement method), and then, wells (i.e., the experimental units) were randomized to treatments using a similar method. Experiments were conducted in triplicate. The data from individual replicate wells were averaged to yield a value of n = 1 for the experiment, and this was repeated on three occasions blind to the experimenter and/or flow cytometer analyst for a final value of n = 3 [96]. Based on the assumptions that the experimental unit (i.e., the well) data comply with independence of observations, the dependent variable is normally distributed in each treatment group (Shapiro-Wilk test), and there is homogeneity of variances (Levene's test); the statistical significance was determined by one-way analysis of variance (ANOVA) followed by Tukey's post hoc comparison calculated with GraphPad Prism 5.0 software (https://www.graphpad.com/; accessed on 5 February 2023). Differences between groups were only deemed significant when a p-value of 0.05 (*), 0.001 (**), or 0.001 (***). All data were illustrated as the mean SD.