Pharmacological GCase Activity Enhancement Inhibits Tau Accumulation

A slow decline in the autophagy-lysosomal pathway is a hallmark of the normal aging brain. Yet, an acceleration of this cellular function may propel neurodegenerative events. In fact, mutations in genes associated with the autophagy-lysosomal pathway can lead to Parkinson’s disease. Also, amyloidogenic protein deposition is observed in lysosomal storage disorders, which are caused by genetic mutations representing risk factors for Parkinson’s disease. For example, Gaucher’s disease GBA1 mutations leading to defects in lysosomal sphingolipid metabolism cause α-synuclein accumulation. We observed that increased lysosomal Tau accumulation is found in human dermal fibroblasts engineered for inducible Tau expression. Inhibition of the GBA1 product GCase augmented Tau-dependent lysosomal stress and Tau accumulation. Here, we show increased Tau seed-induced Tau accumulation in Gaucher’s fibroblasts carrying GBA1 mutations when compared to normal fibroblasts. Pharmacological enhancement of GCase reversed this effect, notably, also in normal fibroblasts. This suggests that boosting GCase activity may represent a therapeutic strategy to slow down aging-dependent lysosomal deficits and brain protein deposition.


Introduction
Neurodegenerative tauopathies are characterized by the progressive accumulation of neurofibrillary hyperphosphorylated Tau protein in the brain [1][2][3] .Tau pathology can occur secondarily to the deposition of other amyloidogenic proteins forming e.g., βamyloid plaques in Alzheimer's disease or α-synuclein Lewy bodies in Parkinson's disease 4 .Protein deposition is linked to a deterioration of cellular functions eventually leading to cell death.A slow decline in cell functions is a hallmark of normal aging as observed e.g., for the autophagy-lysosomal pathway (ALP).This important cellular pathway is dedicated to the elimination of intracellular organelles and macromolecules as well as protein aggregates 5,6 .Not surprisingly, ALP malfunction contributes to the neurodegenerative process.Indeed, the deposition of amyloidogenic proteins has been reported in lysosomal storage disorders (LSD) 7,8 .LSD are caused by inherited defects in lysosomal or non-lysosomal proteins resulting in aberrant buildup of lysosomal substrates and deleterious ALP function 9 .The correlation between protein accumulation and ALP malfunction is better documented in Parkinson's disease.

Whereas monoallelic mutations in ALP genes may represent risk factors for
Parkinson's disease, homozygous mutations cause LSD 10,11 .An important example is given by the GBA1 gene encoding for glucocerebrosidase (GCase), a lysosomal enzyme metabolizing glucosylceramide.In fact, biallelic GBA1 mutation causes Gaucher's disease, a mendelian LSD disorder affecting several organs and tissues due to cells accumulating fatty substances.Yet, monoallelic GBA1 variants stand for the main genetic risk for Parkinson's disease, suggesting that GCase malfunction may be linked to protein accumulation in the brain 12 .Loss of GCase activity in GBA1-Parkinson's patients has been observed in brain, blood, and cells such as dermal fibroblasts or dopaminergic neurons derived from pluripotent stem cells [13][14][15] .Indicating a driving role of ALP malfunctionand possibly also of agingin the pathogenesis of Parkinson's disease, GBA1 mutation carriers show defects in sphingolipid metabolism and α-synuclein accumulation 16,17 .Noteworthy, lysosomal GCase malfunction appears to occur also in the idiopathic Parkinson brain 18 and fibroblasts 19 .Decreased GCase and buildup of glucosylceramide in degradative organelles (DOs) can inhibit autophagy 20 contributing to α-synuclein accumulation, which in turn disturbs the targeting of GCase to lysosomes 21 and interferes directly with GCase activity 22,23 .
We recently showed that this noxious circle of events also occurs in relation to Tau accumulation 24 .When treated with extracellular Alzheimer's brain-derived Tau seeds, a dramatic increase in Tau accumulation in DOs was observed in primary human dermal fibroblasts engineered for inducible fluorescently tagged Tau expression.In this cellular system, Tau was shown to be a substrate of macroautophagy and to accumulate in DOs causing lysosomal stress monitored by nuclear TFE3 translocation and accumulation of lactosylceramide, a product of glucosylceramide metabolism.
Notably, pharmacological GCase inhibition augmented Tau-dependent lysosomal stress and Tau accumulation 24 .The use of primary dermal fibroblasts allows for exploring whether the GBA1 genotype may affect Tau-dependent pathological phenotypes.So, in the current study, we show increased seed-induced Tau accumulation in Gaucher's fibroblasts carrying the GBA1 L444P/L444P biallelic mutation when compared to GBA1 WT fibroblasts.Gain Therapeutics has applied its drug discovery platform to develop pharmacological chaperones that allosterically bind GCase and enhance its activity, such as GT-02216 25,26 .Tau accumulation was rescued by GT-02216 on the GBA1 L444P/L444P and on the GBA1 WT genotype.

GT-02216 binds to the GCase protein
Gain Therapeutics has applied its proprietary SEE-Tx® drug discovery platform 25,27,28 to identify a druggable allosteric site of GCase distinct from the active pocket and for performing virtual compound screening.This methodology starts from the highresolution structure of native human GCase (PDB code: 2V3F 29 ) to find novel sites and binding hotspots for guiding molecular docking 30 .The virtual screening was performed with a non-redundant library of 4.8 million commercial drug-like compounds.The best scoring compounds were experimentally validated and served as a starting point for a medicinal chemistry program that led to the discovery of GT-02216 26 (Fig. 1A).
The direct binding of GT-02216 to GCase was confirmed by surface plasmon resonance (SPR).SPR allows the study of the strength and kinetics of molecular interactions in real time 31 .Performing SPR experiments at pH 5.0 and pH 7.4, can provide insight into how molecular interactions behave under different physiological conditions, mimicking e.g., acidic organelles or the endoplasmic reticulum, respectively 32 .GT-02216 binds to GCase protein in a dose-dependent manner at both acidic pH 5.0 and neutral pH 7.4 (Fig. 1B) with similar binding affinity (Fig. 1C).

GT-02216 enhances GCase activity in primary human fibroblasts
The presence of GBA1 mutations has a deleterious effect on GCase activity.This was first confirmed in fibroblasts using an enzyme activity assay.In this assay, the fluorescent product 4-methylumbelliferone (4MU) is released by GCase activity from the synthetic substrate 4-methylumbelliferyl-β-D-glucopyranoside.Basal GCase activity was found to be reduced for all mutant GBA1 fibroblasts analyzed in a panel of dermal fibroblasts derived from Gaucher's patients (Fig. 2A).
Treatment of the cells with the pharmacological chaperone GT-02216 for 4 days resulted in a significant enhancement of GCase activity in mutant GBA1 fibroblasts as well as in GBA1 WT(XY) fibroblasts (Fig. 2B), except for the GBA1 L444P/L444P(I) β cell line which showed undetectable GCase activity (Fig. 2A and B).GT-02216 elicited maximal, ~2-fold increased activity on the fibroblast lines harboring the homozygote GBA1 L444P/L444P mutation.Consequently, the GBA1 L444P/L444P(I) α line and a GBA1 WT(XX) line were chosen as the cellular models for subsequent studies.Treatment of these two fibroblast lines with increasing concentrations of GT-02216 for 4 days resulted in a dose-dependent rescue of GCase activity (Fig. 2C and D).
Reduced GCase activity is expected to induce an accumulation of the primary physiological GCase substrates glucosylceramide (GlcCer) and glucosylsphingosine.
In fibroblasts, GlcCer is usually evaluated through the measure of hexosylceramide (HexCer), a racemic mixture composed of GlcCer and galactosylceramide.Indeed, treatment of GBA1 L444P/L444P(I) α fibroblasts with conduritol B epoxide (CBE), a wellknown potent irreversible inhibitor of GCase 33 , led to a massive accumulation of HexCer as previously described in the literature 34 (Fig. 3A).Notably, GT-02216 reduced HexCer in the same mutant fibroblast line (Fig. 3B).In line with these data, the amount of HexCer measured in GBA1 WT(XX) fibroblasts was much lower compared to that present in GBA1 L444P/L444P(I) α fibroblasts (Fig. 3C).GT-02216 treatment decreased HexCer also in the wild-type fibroblasts (Fig. 3D), consistent with the enhancement of wild-type GCase measured in the 4MU activity assay (Fig. 2A and D).
Next, we confirmed the data obtained for the parental human dermal fibroblasts in the two lines, GBA1 L444P/L444P(I) α and GBA1 WT(XX) , engineered for inducible Tau-mCherry expression.Upon induction of ectopic Tau expression with doxycycline, Tau-GBA1 L444P/L444P(I) α fibroblasts presented a 17.5% of the GCase activity measured in Tau-GBA1 WT(XX) fibroblasts (Fig. 4A).We confirmed the ability of the pharmacologic chaperone GT-02216 to rescue GCase activity in a dose-dependent manner in Tau-GBA1 L444P/L444P(I) α fibroblasts (Fig. 4B).GT-02216 boosted GCase activity in Tau-GBA1 WT(XX) fibroblasts independently of the absence or presence of Tau expression (Fig. 4C).GT-02216 also elicited a dose-dependent increase in GCase activity also in Tau-GBA1 WT(XX) fibroblasts (Fig. 4D).The dose-response curves differed somehow from those obtained in parental fibroblasts (Fig. 2C and D), and one explanation could be that the experiments were conducted in different laboratories.

Increased Tau accumulation in mutant GBA1 fibroblasts is reduced by GT-02216
We reported previously Tau accumulation in DOs of human primary fibroblasts treated with extracellular brain-derived Tau seeds.Moreover, pharmacological GCase inhibition with CBE increased Tau accumulation and induced lysosomal stress, suggesting a synergistic effect between lysosomal stress and the presence of accumulated Tau in DOs 24 .So, next we asked whether the presence of mutant GCase might affect Tau accumulation in DOs of primary human fibroblasts.Tau accumulation at basal conditions was 3.5-fold higher in Tau-GBA1 L444P/L444P (I) α fibroblasts compared to Tau-GBA1 WT(XX) fibroblasts (Fig. 5A).Moreover, when Tau-GBA1 L444P/L444P(I) α fibroblasts were treated with Alzheimer brain-derived Tau seeds, Tau accumulation increased by a further 2.7-fold (Fig. 5B), but the same treatment did not affect GCase activity (Fig. 5C).Tau-GBA1 WT(XX) fibroblasts displayed a lower, ~1.6-fold seedinduced Tau accumulation when compared to basal conditions (Fig. 5D), in the absence of altered GCase activity (Fig. 5E).These data show that genetic GCase impairment potentiated seed-induced Tau accumulation.However, treatment of Tau-GBA1 L444P/L444P(I) α fibroblasts for 4 days with GT-02216 resulted in a dose-dependent reduction of Tau accumulation both at basal (no seeds) or in the presence of Tau seeds (Fig. 5F).GT-02216 reduced Tau accumulation also in Tau-GBA1 WT(XX) fibroblasts although the dose-dependency showed reduced potency (Fig. 5G).Overall, the data indicate that intracellular Tau accumulation was dependent on lysosomal function that was affected by the enzymatic activity of a single lysosomal enzyme.
We then questioned whether Tau accumulation may somehow affect DOs.First, to confirm that Tau-mCherry accumulation occurred in LAMP1-positive DOs in a GT-02216-dependent manner, we determined for each single fibroblast the Pearson's Correlation Coefficient (PCC) of mCherry biofluorescence with LAMP1 fluorescent staining 35 .Whereas Tau seed-treatment of Tau-GBA1 L444P/L444P(I) α fibroblasts significantly increased Tau in LAMP1-positive DOs, the presence of GT-02216 reversed this seed-dependent effect (Fig. 6A).Similar, but less pronounced data, were obtained in Tau-GBA1 WT(XX) fibroblasts (Fig. 6B).Consistent with our previous findings in normal fibroblasts 24 , Tau seed-treatment increased the number of DOs present in Tau-GBA1 L444P/L444P(I) α (Fig. 6C) and Tau-GBA1 WT(XX) (Fig. 6D) and again, GT-02216 reversed this effect.Possibly, the presence of mutated GCase (compare the light gray bars in Fig. 6C and D), or seed-induced Tau accumulation in DOs, led to lysosomal malfunction followed by a cellular response in the form of increased DOs biogenesis.
On the other hand, the effect of GT-02216 on GCase activity improved lysosomal function, which favored Tau degradation and decreased Tau accumulation in DOs.

GT-02216 protects rat hippocampal primary neurons challenged with Tau oligomers
To further explore the effects of GT-02216 in a different cellular context, we extended our investigation to a neuronal model.Given that Tau oligomers (TauO) are known to exert neurotoxic effects, we assessed whether GT-02216 could confer neuroprotection in primary rat hippocampal neurons challenged with TauO.Incubation of these cells with TauO at a concentration of 5 μmol/L resulted in a significant decrease in cell viability, consistent with previous findings 36,37 .Recombinant Brain Derived Neurotrophic Factor (BDNF), known to be protective against Tau-related neurotoxicity in several neurodegenerative disorders 38 , was used as a positive control in the assay.BDNF significantly attenuated TauO-induced neuronal death in our model validating the experimental set-up (Fig. 7A).GT-02216 treatment also effectively rescued cell viability in the TauO-challenged hippocampal neurons, demonstrating its neuroprotective potential against TauO-induced neurotoxicity (Fig. 7A).

Discussion
Our study was an initial exploration of the therapeutic potential of the pharmacological GCase-enhancing chaperone GT-02216 in tauopathies.GT-02216 was selected starting from a novel druggable allosteric GCase site identified with the proprietary Gain Therapeutics SEE-Tx® drug-discovery platform 25,27,28 .In vitro assays showed GT-02216 binding to GCase that increased GCase activity and reduced endogenous GCase substrates in fibroblasts.The SEE-Tx® computational technology was previously successfully applied to other enzymes linked to LSD 39,40 .Examples are alpha-L-iduronidase deficiency causing mucopolysaccharidosis type I, an inherited lysosomal disease 40 , and mutant β-galactosidase 39,41 triggering GM1 gangliosidosis and mucopolysaccharidosis type IVB, also known as Morquio B disease 39,41 .
GCase enzymatic dysfunction leads to defects in metabolism and lysosomal stress.
We selected several human dermal fibroblasts derived from healthy donors and donors carrying GBA1 mutations as our model.We first validated the model showing that the activity of GCase was reduced in GBA1 mutation-carrying fibroblasts when compared to normal fibroblasts, which led to increased amounts of intracellular sphingolipids, the substrates of lysosomal GCase.We then showed that GT-02216 enhanced GCase activity and reduced HexCer accumulation in GBA1 L444P/L444P mutant and wild-type fibroblasts.Fibroblasts carrying the homozygous L444P/L444P mutation, which is associated with severe neuronopathic Gaucher's disease and increased dementia risk 42 , showed the strongest decrease in GCase activity and were therefore selected for the whole study.To evaluate the pathological accumulation of Tau in this model, we took advantage of a previous study reporting evidence of aberrant Tau accumulation in DOs linked to lysosomal stress in primary human fibroblasts expressing various fluorescent forms of Tau 24 .This adverse process was further propagated by incubating the cells with Alzheimer's brain-derived Tau seeds as well as in the presence of CBE, a cell-active pharmacological irreversible inhibitor of lysosomal GCase 24 .In agreement with these data, we observed increased seedinduced Tau accumulation in DOs of Gaucher's fibroblasts expressing defective GCase when compared to wild-type fibroblasts.These data stress the potential implication of GCase deficiency and the accumulation of GCase substrates in abnormal Tau accumulation in DOs.Of relevance in this context, treatment of the cells with GT-02216 efficiently reduced Tau accumulation.Moreover, the rescuing effect was also observed in cells with the wild-type GCase background.We concluded that lysosomal dysfunction associated to GCase deficits in combination with internalized Tau seeds may together contribute to the accumulation of Tau on route to degradation within DOs.However, Tau accumulation and impairment of lysosomal function and lipid metabolism may reciprocally engage in a sequence of harmful events.Improving lysosomal function through pharmacological GCase enhancement may represent a viable strategy to reduce aging-associated cellular stress by indirectly targeting Tau accumulation.The reduction of Tau accumulation in basal and seed-induced conditions after treatment with GT-02216 observed in fibroblasts, acquired further relevance from the finding that GT-02216 rescued the viability of (endogenously Tau expressing) rat hippocampal neurons challenged with exogenous Tau oligomers.This highlighted the critical balance maintained by ALP in neuronal health and showed the potential of GT-02216 as a neuroprotective agent targeting Tau-related neurotoxicity with an effect comparable to that of the neuroprotectant BDNF.
Our study further confirms the emerging crosstalk between ALP dysfunction and neurodegeneration.Mice deficient in autophagy exhibit neuronal accumulation of aggregate-prone proteins and neurodegeneration, demonstrating the crucial role of autophagy in neuronal homeostasis 43,44 .The role of the ALP in neurodegenerative diseases has been extensively studied in recent years.Late-onset neurodegenerative disorders such as Parkinson's, Huntington's, and Alzheimer's diseases are characterized by the accumulation of intracellular aggregates in the brain.Clearance of these aggregates is typically associated with improvement of symptoms 45 , indicating that ALP impairment and accumulation of pathogenic protein forms may together contribute to disease progression.Tau lesions are detected in LSD models such as Gaucher's disease, Niemann-Pick, Sanfilippo syndrome type B, Christianson syndrome and Fabry's disease [46][47][48][49] .Moreover, restoring mutated GCase activity with the chaperone Ambroxol or through ectopic expression of wild-type GCase delayed Tau and α-synuclein accumulation 50,51 .Thus, correcting ALP defects are appealing therapeutic interventions.
Considering the data showing efficacy of GT-02216 also in wild-type cells, our study prompts additional investigations to further underpin that boosting GCase activity could represent a viable therapeutic strategy to slow-down aging-dependent protein deposition in the general patient population not affected by GBA1 mutations.Despite the fact that GBA1 mutations are not linked to Alzheimer's disease or other neurodegenerative proteinopathies besides Parkinson's disease 52,53 , altogether, our findings suggest that GT-02216 shows promise as a potential disease-modifying treatment option for aging, Alzheimer's disease or other tauopathies beyond those directly associated with GBA1 mutations.

Human dermal fibroblasts
Human primary GBA1 WT(XX) fibroblasts were isolated from a dermal biopsy obtained from a healthy 30-year-old female 24 .GBA1 L444P/L444P(I) α Gaucher type I fibroblasts (Coriell, #GM10915) were originally isolated from a 7-year-old male and are homozygote for a T>C transition at nucleotide 1448 in exon 10 of the GBA1 gene.

Fibroblast treatments
Tau expression was routinely induced for 4 days with 0.3 µg/mL doxycycline (Sigma-Aldrich, #D9891).GT-02216 was incubated with cells for 4 days.Independently of the compound concentration, the vehicle DMSO was kept constant.Tau seeds were enriched from frozen Alzheimer's disease frontal cortex samples as described 24 .The tissues were obtained from The Netherlands Brain Bank, Netherlands Institute for Neuroscience, Amsterdam (www.brainbank.nl).Anonymized donors signed a written informed consent for brain autopsy and further use of tissue and clinical information for research purposes.Tau seeds were directly supplemented to the cell culture media for 1 day before the cells were gently washed with PBS (Gibco, #10010023), cell nuclei were stained with 2.5 μg/mL Hoechst 33342 (Invitrogen, #H3570) for 10 min, 37°C, followed by gentle washes in complete medium and PBS, before fixation for further analysis.
For Tau accumulation analysis, fibroblasts seeded at 30x10 3 cells/well on poly-D-lysine (Sigma-Aldrich, #P6407) coated 8-well microscope slides (Ibidi, #80826-IBI) were fixed in 2% formaldehyde in PBS for 15 min at room temperature, washed three times with 500 mmol/L glycine in PBS and then three times with PBS.
For LAMP1 immunostaining analysis, fibroblasts were fixed in 4% formaldehyde in PBS for 5 min at 37°C followed by cold 2% formaldehyde in 6.25% methanol (Sigma-Aldrich, #32213) in PBS for 1 min at room temperature.Cells were washed three times with 500 mmol/L glycine in PBS and two times in PBS.Unspecific binding was blocked with 5% NGS (Sigma-Aldrich, #N2513) in PBS for 1 h at room temperature and three washes in PBS.Staining with 0.4 µg/ml LAMP1 primary antibody (SantaCruz Biotechnology, Inc., #SC20011) was done in 0.5% NGS in PBS for 1 h at room temperature, followed by three washes in PBS and the addition of 2 µg/ml AlexaFluor TM 488 (Invitrogen, #A11001) secondary antibody and 1 µg/ml DAPI (Sigma-Aldrich, #D9542) for 1 h at room temperature.After the last incubation, cells were washed three times and stored in PBS.
Image analysis and processing of the raw data were performed with Fiji/ImageJ v1.54 software.Pearson's Correlation Coefficient (PCC) was determined with ImageJ JACoP v2.0 plugin for dual-color colocalization of Tau and LAMP1.Acquisition settings were kept constant across all conditions analyzed and acquired images were processed with the exact same parameters, including background subtraction.Quantification of the percentage of cells with a Tau puncta phenotype was normalized with the total number of DAPI positive nuclei.Single Tau puncta were manually identified and analyzed with the ImageJ software 24,35 .

Hexosylceramide Assay
Primary human GBA1 WT (XX) fibroblasts were seeded at 2.5x10 5 cells/T25 flasks (Falcon ® , #353808) in DMEM+GlutaMAX TM containing 15% FBS and 1% PenStrep and incubated at 37°C, 5% CO2 for 1 day.Cell media were then replaced with fresh medium containing the desired concentration of GT-02216.After 10 days, cells were detached, collected into two 2 mL tubes for each flask, and placed in ice.After centrifugation at 13000 rpm at 4°C for 5 min, the cell pellets were rinsed with 1 mL cold PBS and collected into a single tube for each sample.Samples were centrifuged at 13000 rpm at 4°C for 2 min and cells pellets stored at -80°C until further analysis.

Statistics and reproducibility
Statistical analysis was performed with GraphPad Prism version 10.2.3 with at least three independent biological replicates.Most quantifications are reported as fold over control/untreated conditions, unless otherwise indicated in the graphs.Quantification of total HexCer in GBA1 L444P/L444P(I) α compared to GBA1 WT(XX) fibroblasts (mean ± SD normalized on GBA1 WT(XX) , n= 3).Unpaired t test, p .0006.D. As in B. for GBA1 WT(XX) (10 mmol/L stock solution in DMSO) was measured by SPR at pH 7.4 in 10 mmol/L HEPES, 5 mmol/L EDTA, 150 mmol/L NaCl and 0.01% Tween-20 or at pH 5.0 in 20 mmol/L Na phosphate, 2.7 mmol/L KCl, 137 mmol/L NaCl, 5 mmol/L tartrate, 0.01% Tween-20.Empty, activated, and deactivated parallel channels on the same SPR sensor were used as reference channels.Raw SPR signals monitored on the active channel with immobilized GCase protein were subtracted with signals monitored on the reference channel (empty sensor surface) and further subtracted with the signal monitored for the running buffer (double referenced), and finally corrected for DMSO signal mismatch between sample and running buffer.To extract binding affinity values, the plotted SPR data were further fitted with the four-parameter logarithmic dose-response equation without constraint (GraphPad Prism version 10.2.3).

Figure 1 .
Figure 1. A. Molecular structure of the GCase pharmacological chaperone GT-02216.B. SPR dose-response for GT-02216 binding to immobilized human GCase protein monitored at acidic (pH 5.0) and neutral (pH 7.4) conditions.C. SPR binding properties determined at the indicated pH values.

Figure 7 .
Figure 7. Viability of rat hippocampal primary neurons treated for 2 days with GT-02216 at the indicated concentration or with BDNF before being challenged with 5 µmol/L Tau oligomers for 1d.Cell viability was assessed with the MTT assay (mean ± SD, n= 3-4 wells).Ordinary 1way ANOVA (p <.0001) and Dunnett's multiple tests against the control, **** p < .0001.