A Proteomic Approach Identified TFEB as a Key Player in the Protective Action of Novel CB2R Bitopic Ligand FD22a against the Deleterious Effects Induced by β-Amyloid in Glial Cells

Neurodegenerative diseases (NDDs) are progressive multifactorial disorders of the nervous system sharing common pathogenic features, including intracellular misfolded protein aggregation, mitochondrial deficit, and inflammation. Taking into consideration the multifaceted nature of NDDs, development of multitarget-directed ligands (MTDLs) has evolved as an attractive therapeutic strategy. Compounds that target the cannabinoid receptor type II (CB2R) are rapidly emerging as novel effective MTDLs against common NDDs, such as Alzheimer’s disease (AD). We recently developed the first CB2R bitopic/dualsteric ligand, namely FD22a, which revealed the ability to induce neuroprotection with fewer side effects. To explore the potential of FD22a as a multitarget drug for the treatment of NDDs, we investigated here its ability to prevent the toxic effect of β-amyloid (Aβ25–35 peptide) on human cellular models of neurodegeneration, such as microglia (HMC3) and glioblastoma (U87-MG) cell lines. Our results displayed that FD22a efficiently prevented Aβ25–35 cytotoxic and proinflammatory effects in both cell lines and counteracted β-amyloid-induced depression of autophagy in U87-MG cells. Notably, a quantitative proteomic analysis of U87-MG cells revealed that FD22a was able to potently stimulate the autophagy–lysosomal pathway (ALP) by activating its master transcriptional regulator TFEB, ultimately increasing the potential of this novel CB2R bitopic/dualsteric ligand as a multitarget drug for the treatment of NDDs.


Introduction
The transcription factor EB (TFEB) is a central regulator of the autophagy-lysosomal pathway (ALP) [1,2], which is a major mechanism for degrading intracellular macromolecules, including long-lived proteins, aggregated misfolded proteins, and abnormal cytoplasmic organelles, and maintaining cellular homeostasis (Figure 1).
expression is widely known to increase the number of autophagosomes, to promote the generation of new lysosomes, and to increase the autophagic flux [1,[13][14][15].Therefore, inducing intracellular clearance through the induction of TFEB activity may represent an appealing therapeutic intervention for the treatment of NDDs.Intracellular molecules, including PI3K, MAPK, AMPK, and TNFR, possess the capability to activate the mTOR pathway, a prominent negative regulator of autophagy.mTOR, functioning as a kinase, plays a crucial role in regulating the localization and activation of TFEB, a crucial transcription factor, which governs autophagy at the transcriptional level by promoting the expression of multiple lysosomal genes, influencing autolysosome production and function.Specifically, activated mTOR phosphorylates TFEB, inhibiting its activity and confining it to the cytoplasm.Inhibiting mTOR leads to TFEB dephosphorylation, enabling its nuclear translocation and transcriptional activity.Within the nucleus, TFEB regulates autophagy in two ways: promoting the expression of autophagy-related molecules for enhanced autophagy, and activating lysosomal-pathway-related molecules, especially LAMP1, to foster lysosome formation.This orchestrated process facilitates the degradation of damaged organelles, recycling their products, such as amino acids, for cellular benefit.
It is widely recognized that dysfunction in the ALP is a pathogenic feature shared by multiple adult-onset neurodegenerative disorders (NDDs), including Alzheimer's, Parkinson's, and Huntington's diseases [3][4][5].In general, the accumulation of intracellular aggregates in the brain, which leads to synaptic dysfunction and ultimately neuronal death, is a common feature for all these pathologies.
Despite extensive research efforts, the complex etiology of these diseases is not yet clear, and limited treatment options are currently available.
TFEB is widely expressed in the CNS, including in neurons and astrocytes [6][7][8].In physiological conditions, TFEB localizes to the cytosol and rests on the lysosomal surface, where upstream kinases, such as rapamycin complex 1 (mTORC1) [9], can phosphorylate it.Thus, the inhibition of mTOR, induced by starvation and lysosomal stress, promotes TFEB dephosphorylation and its nuclear translocation [1,10].Nuclear TFEB increases the transcription of genes involved in the regulation of lysosomal, autophagic, and retromer function, collectively called the coordinated lysosomal expression and regulation (CLEAR) network.Notably, nuclear TFEB localization is decreased in different NDDs in which protein aggregation takes place [6,8,11,12], thus suggesting that defective nuclear translocation of TFEB is related to impaired protein homeostasis in neurons.TFEB overexpression is widely known to increase the number of autophagosomes, to promote the generation of new lysosomes, and to increase the autophagic flux [1,[13][14][15].Therefore, inducing intracellular clearance through the induction of TFEB activity may represent an appealing therapeutic intervention for the treatment of NDDs.
While most research has investigated CB1R, which is highly expressed in nearly all brain regions, CB2R in the brain has started to attract considerable interest only in recent years, having been considered for a long time exclusively a peripheral-type receptor [22].During the last two decades numerous experimental studies have provided robust evidence that CB2R seems to be involved in the modulation of different neurological disorders characterized by neuroinflammatory processes and microglial cell activation [23], suggesting the therapeutic potential of natural and synthetic CB2R ligands in the treatment of neurodegenerative proteinopathies, such as Alzheimer's and Parkinson's diseases [24,25].While most research has investigated CB1R, which is highly expressed in nearly all brain regions, CB2R in the brain has started to attract considerable interest only in recent years, having been considered for a long time exclusively a peripheral-type receptor [22].During the last two decades numerous experimental studies have provided robust evidence that CB2R seems to be involved in the modulation of different neurological disorders characterized by neuroinflammatory processes and microglial cell activation [23], suggesting the therapeutic potential of natural and synthetic CB2R ligands in the treatment of neurodegenerative proteinopathies, such as Alzheimer's and Parkinson's diseases [24,25].
Notably, a very recent report highlighted a new mechanism allowing CB2R to regulate autophagy (ATG), lipid metabolism, and inflammation in an animal model of postopera-tive cognitive dysfunction (POCD) through the modulation of TFEB activity [26], further increasing the potential of CB2R targeting for therapeutic intervention in NDDs.
Our group has recently developed the first CB2R bitopic/dualsteric ligand, namely FD22a, which has been shown to display beneficial biological responses both in vitro and in vivo with fewer side effects [27].
Bitopic/dualsteric ligands, which are hybrid compounds composed of orthosteric and allosteric pharmacophoric units, represent one of the most promising strategies of targeting G protein-coupled receptors (GPCRs) [28].Indeed, this approach allows the exploitation of favorable characteristics of the orthosteric and the allosteric site by a single ligand molecule, including an increased affinity or selectivity for the target receptor, often associated with functional selectivity (i.e., bias signaling pathway activation), reduced off-target activity, and therapeutic resistance [29].
Notably, novel CB2R bitopic ligand FD22a met requirements typical of the bitopic ligand, such as receptor-subtype selectivity and biased signaling for cAMP inhibition versus βarrestin2 recruitment, while revealing significant neuroprotective effects, such as the ability to efficiently combat the inflammatory process in human microglial cells and to display antinociceptive activity in vivo [27].In addition, computational studies clarified the binding mode of this compound inside the CB2R, further confirming its bitopic nature [27].
To explore in more detail the potential of newly developed CB2R bitopic ligand FD22a to target neurodegeneration, in the present study we investigated the ability of FD22a to counteract the detrimental effects produced by the neurotoxic Aβ fragment 25-35 (Aβ 25-35 ) [30] in human cellular models of neurodegeneration, such as human microglial (HMC3) and human glioblastoma-astrocytoma (U87-MG) cell lines.
Moreover, protein expression profiling, combined with pathways analyses, revealed that FD22a was able to potently stimulate the ALP pathway through TFEB activation, in turn reversing Aβ 25-35 neurotoxicity by promoting intracellular clearance.
Aβ 25-35 peptide (NH 2 -Gly-Ser-Asn-Lys-Gly-Ala-Ile-Ile-Gly-Leu-Met-COOH) (A4559, Sigma-Aldrich, Milan, Italy) was initially dissolved in double-distilled water to obtain 1 mM concentration and stored at −20 • C. To form aggregated diffusible oligomers, the solution was incubated at 37 • C for 5 days [31], then diluted in medium to the indicated concentration, just prior to cell treatments.FD22a was dissolved in DMSO to obtain a 50 mM stock solution which was kept at 4 • C. Before the experiments FD22a stock solution was diluted into the cell culture medium to the desired experimental concentration, and the final DMSO concentration was maintained no higher than 0.1%.Vehicle-treated cells (0.1% DMSO) were used as control (Ctrl).
In all the experiments 24 h after seeding, cells were exposed to pretreatment with FD22a for 24 h and then exposed to Aβ 25-35 at the pertinent concentration (10 µM for U87-MG and 1 µM for HMC3).After 48 h, cells were processed according to the specific experiment protocol.

MTT (Cell Viability Assay)
The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, Milan, Italy) reagent was used to test the effect of FD22a on cell viability.Briefly, after treatment, cells were incubated with MTT (0.5 mg/mL) for 4 h at 37 • C. The formazan products were dissolved in DMSO.An automated microplate reader (BIO-TEK, Winooski, VT, United States) was used to quantify absorbance at 540 nm.Cell viability was expressed as the percentage of control cells.

Gene Expression Analysis
Total RNA was extracted using the RNeasy Mini kit (74104, Qiagen, Hilden, Germany) and a Qubit v.1 fluorometer plus Qubit RNA HS Assay Kit (Thermo Fisher Scientific, Wilmington, DE, USA) was used to extract and quantify total RNA, on the basis of manual protocol indications.
Extracted RNA (1 µg) was retrotranscribed by using the iScriptTM gDNA Clear cDNA Synthesis Kit (Bio-Rad, Milan, Italy) according to the manufacturer's instructions, and the obtained cDNA samples were quantified by real-time PCR using a SYBR Green probe and CFX Connect Real-Time PCR Detection System (Bio-Rad, Milan, Italy).The PCR cycle program consisted of an initial denaturation at 95 • C for 30 s followed by 40 cycles of 5 s of denaturation at 95 • C and 15 s of annealing/extension at 60 • C. To verify amplicon specificity and potential primer dimer formation, a final melting protocol with ramping from 65 • C to 95 • C with 0.5 • C increments of 5 s was performed.
Primer sequences (Table 1) were designed by using Beacon Designer Software v.8.0 (Premier Biosoft International, Palo Alto, CA, USA) with a junction primer strategy, whenever possible.To exclude genomic DNA contamination, a negative retrotranscription control was used.The endogenous reference gene GAPDH was quantified for each sample.
All reactions were performed in triplicate and the amount of mRNA was calculated by the comparative critical threshold (CT) method.

Proteomic Analysis
For proteomic analysis, U87-MG human glioblastoma cells were exposed to pretreatment with FD22a (1 µM) for 24 h before being exposed for 48 h with Aβ 25-35 (10 µM) as described above.After treatment, cells were rinsed with ice-cold PBS and lysed in rehydration solution (7 M urea, 2 M thiourea, 4% CHAPS, 60 mM dithiothreitol (DTT), 0.002% bromophenol blue) added with 50 mM NaF, 2 mM Na 3 VO 4 , 1 µL/10 6 cells, and protease cocktail inhibitors (Sigma-Aldrich, St. Louis, MO, USA).After stirring and sonication, cells were allowed to rehydrate for 1 h at room temperature (RT) and, thereafter, the solution was centrifuged at 16,000× g for 10 min at RT [32].Protein contents of resulting protein extracts were measured with the Pierce Protein Assay (Thermo Fisher Scientific, Waltham, MA, USA) and bovine serum albumin was used as standard.
Two-dimensional electrophoresis (2DE) was carried out as previously described [33].Briefly, one hundred fifty micrograms of proteins was loaded on Serva IPG blue strips (SERVA-German Headquarter, Heidelberg, Germany) with a linear pH 3-10 gradient.The second dimension (SDS-PAGE) was carried out by transferring the proteins to 12% polyacrylamide gels.The gels were stained with Ruthenium II tris (bathophenanthroline disulfonate) tetrasodium salt (Cyanagen Srl, Bologna, Italy) (RuBP) [34] and images were acquired by ImageQuant LAS4010 (GE Health Care, Uppsala, Sweden).The analysis of images was performed using Same Spot (v4.1, TotalLab; Newcastle Upon Tyne, UK) software [35].The spot volume ratios among the four different conditions (control, FD22a, Aβ 25-35 , FD22a + Aβ 25-35 ) were calculated using the average spot normalized volume of the six biological replicates.The software included statistical analysis calculations.

In Gel Digestion and Mass Spectrometry
The gel pieces were digested as reported by Giusti et al. 2018 [36].Peptide MS spectra were recorded manually on the AutoFlex Speed MALDI-TOF/TOF spectrometer (Bruker Daltonics, Leipzig, Germany) operated in positive reflectron mode [37].Samples unidentified by MALDI-TOF/TOF were analyzed by LC-MS/MS using an UltiMate3000 RSLCnano chromatographic system coupled to an Orbitrap Fusion Tribrid mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA), operating in positive ionization mode, equipped with a nanoESI source (EASY-Spray NG).Peptides were loaded on a PepMap100 C18 precolumn cartridge (5 µm particle size, 100 Å pore size, 300 µm i.d.× 5 mm length, Thermo Fisher Scientific, Waltham, MA, USA) and subsequently separated on an EASY-Spray PepMap RSLC C18 column (2 µm particle size, 100 Å pore size, 75 µm i.d.× 15 cm length, Thermo Fisher Scientific, Waltham, MA, USA) at a flow rate of 300 nL/min and a temperature of 40 • C, using 0.1% FA in water (eluent A) and 99.9% ACN, 0.1% FA (eluent B).The chromatographic separation was achieved by a two-step linear gradient from 5% to 30% eluent B in 40 min, and from 30% to 55% in 5 min followed by an increase to 90% in one minute, for a total runtime of 56 min.
Precursor (MS1) survey scans were recorded in the Orbitrap, at resolving powers of 240 K (at m/z 200).Data-dependent MS/MS (MS2) analysis was performed in topspeed mode with a 3 s cycle time, during which most abundant multiple-charged (2+-5+) precursor ions detected within the range of 375-1500 m/z were selected for HCD activation in order of abundance and detected in the ion trap at a rapid scan rate after fragmentation using 30% normalized collision energy.Quadrupole isolation with a 1.6 m/z isolation window was used, and dynamic exclusion was enabled for 60 s after a single scan.Automatic gain control targets and maximum injection times were standard and auto for MS1 and 150% and 70 for MS2.For MS2, the signal intensity threshold was 5.0 × 10 3 , and the option "Injection Ions for All Available Parallelizable Time" was set.
Raw data were directly loaded in PEAKS Studio Xpro software 11 (Bioinformatic Solutions Inc., Waterloo, ON, Canada) using the "correct precursor only" option.The mass lists were searched against the UniProt/SwissProt database (downloaded January 2022) restricted to Mammalia taxonomy to which a list of common contaminants was appended (67,666 searched entries).Non-specific cleavage was allowed to one end of the peptides, with a maximum of 2 missed cleavages and 2 variable PTMs per peptide.Additionally, 10 ppm and 0.5 Da were set as the highest error mass tolerances for precursors and fragments, respectively.A −10lgp threshold for PSMs was manually set to 35.

Bioinformatic Analysis
To determine the predominant canonical pathways and interaction network involved, differentially expressed proteins obtained from the comparison of FD22a + Aβ 25-35 vs. Aβ 25-35 were functionally analyzed using QIAGEN's Ingenuity Pathway Analysis (IPA, QIAGEN Redwood City, CA, USA, www.qiagen.com/ingenuity,Build version: 321501M, Content version: 21249400, accessed on 15 January 2021).Comparison of the different analyses was performed and potential regulators and downstream functions were investigated as previously described [33].

Statistical Analysis
For cell viability, ELISAs and gene expression analysis results are expressed as the mean ± standard error of the mean (SEM).Statistical analyses were performed using commercial software (GraphPad Prism, San Diego, CA, USA) using ordinary one-way ANOVA followed by Dunnett's or Tukey's post hoc tests.Differences for which p < 0.05 were considered significant.
For proteomic studies, all analyses were performed at least in triplicate and values were expressed as mean ± standard error (SD).In 2DE experiments, a comparison among the different treatments was performed.The significance of the differences of normalized volume for each spot was calculated by the software Same Spot (v4.1, TotalLab, Newcastle Upon Tyne, UK) including the analysis of variance (ANOVA test).The protein spots that showed significant differences in expression were cut out from the gel and identified by mass spectrometry analyses.The immunoreactive bands obtained in WB experiments were analyzed using ImageQuant TL (GE Health Care).The antigen-specific bands and the total proteins after RuBP staining were quantified.The volume of each band was normalized on total proteins obtained from RuBP staining.The results were expressed as a ratio of optical density.For phosphoproteins a consecutive normalization on the expression level of corresponding proteins was performed.An unpaired t-test was used to compare differences among treatments (Prism 7; GraphPad Software, San Diego, CA, USA).Differences for which p < 0.05 were considered significant.

FD22a Prevents Aβ 25-35 -Induced Cytotoxicity in HMC3 and U87-MG Cells
To investigate the potential protective activity of FD22a against Aβ 25-35 -induced cytotoxicity, two different human cell lines, namely human microglial (HMC3) and glioblastoma (U87-MG) cells, were exposed to pretreatment (24 h) with increasing concentrations (0.1-10 µM) of FD22a before being treated for 48 h with 1 or 10 µM Aβ 25-35 , respectively (Figure 3).These two β-amyloid concentrations have been chosen as they reduce cell viability by approximately 50% compared to either HMC3 or U87-MG control cells (Figure 3C,D).Moreover, comparable Aβ 25-35 concentrations have been previously used by Polini et al. in HMC3 cell culture experiments [31].Of note, in HMC3 cells no tested concentration of FD22a affected cell viability (Figure 3A), whereas in U87-MG cells, when FD22a was used at a concentration higher than 1 µM, a slight cytotoxic effect was observed (Figure 3B).Vehicle-treated cells (Ctrl) did not show any difference compared to untreated cells (NT).

FD22a Prevents Aβ25-35-Induced Cytotoxicity in HMC3 and U87-MG Cells
To investigate the potential protective activity of FD22a against Aβ25-35-induced cytotoxicity, two different human cell lines, namely human microglial (HMC3) and glioblastoma (U87-MG) cells, were exposed to pretreatment (24 h) with increasing concentrations (0.1-10 µM) of FD22a before being treated for 48 h with 1 or 10 µM Aβ25-35, respectively (Figure 3).These two β-amyloid concentrations have been chosen as they reduce cell viability by approximately 50% compared to either HMC3 or U87-MG control cells (Figure 3C,D).Moreover, comparable Aβ25-35 concentrations have been previously used by Polini et al. in HMC3 cell culture experiments [31].Of note, in HMC3 cells no tested concentration of FD22a affected cell viability (Figure 3A), whereas in U87-MG cells, when FD22a was used at a concentration higher than 1 µM, a slight cytotoxic effect was observed (Figure 3B).Vehicle-treated cells (Ctrl) did not show any difference compared to untreated cells (NT).
Cell viability was analyzed by MTT assay.As expected, Aβ25-35 significantly reduced cell viability with respect to control cells (Figure 3C,D).When used at 0.1 µM, FD22a did not counteract the deleterious effects of Aβ25-35 on cell viability in both cell lines.On the contrary, 1 µM FD22a induced a significant increase in cell viability compared to Aβ25-35treated cells (Figure 3C,D).For these reasons, the subsequent experiments were carried out using 1 µM FD22a.Cell viability was analyzed by MTT assay.As expected, Aβ 25-35 significantly reduced cell viability with respect to control cells (Figure 3C,D).When used at 0.1 µM, FD22a did not counteract the deleterious effects of Aβ 25-35 on cell viability in both cell lines.On the contrary, 1 µM FD22a induced a significant increase in cell viability compared to Aβ 25-35treated cells (Figure 3C,D).For these reasons, the subsequent experiments were carried out using 1 µM FD22a.

FD22a Inhibits β-Amyloid-Mediated Release of Proinflammatory Factors in HMC3 and U87-MG Cells
Microglia represent the first line of immune defense within the CNS, and microglial dysfunction is considered a pathogenic mechanism common to several neurological disorders [38].Studies have revealed that Aβ peptides activate microglia to release a large variety of proinflammatory factors [39,40].Therefore, a potential strategy to delay Alzheimer's disease (AD) onset and possibly prevent its progression could be suppressing the response of microglial cells to inflammatory stress [41,42].Based on these premises, in our study we used human HMC3 microglial cells to investigate the protective effect of the newly developed CB2R bitopic/dualsteric ligand FD22a against β-amyloid's cytotoxic and proinflammatory effects.Having assessed that FD22a (1 µM) can protect HMC3 cells from β-amyloid-induced cytotoxicity, we went on to evaluate the ability of FD22a to withstand the increased production of proinflammatory cytokines induced by β-amyloid.In HMC3 cells, Aβ 25-35 (1 µM for 48 h) promoted a significant increase in the release of common proinflammatory cytokines (TNFα and IL-6) (Figure 4A,B), whereas no effect on the release of anti-inflammatory cytokine IL-10 was observed (Figure 4C).In all experiments, treatment with FD22a (1 µM) alone did not show any significant change as compared to control cells.Then, we repeated the experiments, exposing HMC3 cells to pretreatment with 1 µM FD22a.Compared with the β-amyloid-treated group, pretreatment with FD22a was demonstrated to significantly counteract the β-amyloid-increased secretion of TNFα and IL-6, even though the level of both cytokines was still higher as compared to control cells (Figure 4A,B).In addition, pretreatment with FD22a followed by Aβ 25-35 treatment induced a significantly enhanced secretion of anti-inflammatory cytokine IL-10 (Figure 4C).Taken together, these findings indicated the potential of FD22a to inhibit β-amyloid-induced microglial activation.
(B) U87-MG cells were treated with FD22a (0.1-20 µM) for 24 h; (C) HMC3 and (D) U87-MG cells were pretreated with FD22a (1 µM) and after 24 h exposed respectively to 1 µM or 10 µM Aβ25-35 for 48 h.In both experimental models, cell viability was quantified by MTT assay.Each bar corresponds to the means ± SEM of at least four independent experiments.Data were analyzed by one-way analysis of variance (ANOVA) followed by Dunnett's test.§ p < 0.05 with respect to vehicle-treated cells (Ctrl).$$$ p < 0.005 respect to Ctrl; *** p < 0.005 with respect to Aβ25-35-treated cells.

FD22a Inhibits β-Amyloid-Mediated Release of Proinflammatory Factors in HMC3 and U87-MG Cells
Microglia represent the first line of immune defense within the CNS, and microglial dysfunction is considered a pathogenic mechanism common to several neurological disorders [38].Studies have revealed that Aβ peptides activate microglia to release a large variety of proinflammatory factors [39,40].Therefore, a potential strategy to delay Alzheimer's disease (AD) onset and possibly prevent its progression could be suppressing the response of microglial cells to inflammatory stress [41,42].Based on these premises, in our study we used human HMC3 microglial cells to investigate the protective effect of the newly developed CB2R bitopic/dualsteric ligand FD22a against β-amyloid's cytotoxic and proinflammatory effects.Having assessed that FD22a (1 µM) can protect HMC3 cells from β-amyloid-induced cytotoxicity, we went on to evaluate the ability of FD22a to withstand the increased production of proinflammatory cytokines induced by β-amyloid.In HMC3 cells, Aβ25-35 (1 µM for 48 h) promoted a significant increase in the release of common proinflammatory cytokines (TNFα and IL-6) (Figure 4A,B), whereas no effect on the release of anti-inflammatory cytokine IL-10 was observed (Figure 4C).In all experiments, treatment with FD22a (1 µM) alone did not show any significant change as compared to control cells.Then, we repeated the experiments, exposing HMC3 cells to pretreatment with 1 µM FD22a.Compared with the β-amyloid-treated group, pretreatment with FD22a was demonstrated to significantly counteract the β-amyloid-increased secretion of TNFα and IL-6, even though the level of both cytokines was still higher as compared to control cells (Figure 4A,B).In addition, pretreatment with FD22a followed by Aβ25-35 treatment induced a significantly enhanced secretion of anti-inflammatory cytokine IL-10 (Figure 4C).Taken together, these findings indicated the potential of FD22a to inhibit β-amyloid-induced microglial activation.Since the role of CB2R in the beneficial effects of FD22a on the inflammatory response of LPS/TNFα-treated HMC3 cells has been previously reported, we additionally examined the involvement of CB2R in mediating the anti-inflammatory properties of FD22a in Since the role of CB2R in the beneficial effects of FD22a on the inflammatory response of LPS/TNFα-treated HMC3 cells has been previously reported, we additionally examined the involvement of CB2R in mediating the anti-inflammatory properties of FD22a in β-amyloidinduced HMC3 cells.As shown in Figure 4A-C, co-administration of CB2R selective antagonist SR144528 (1 µM) almost completely abolished the protective effect of FD22a against β-amyloid-induced microglial activation.The same set of experiments was also carried out in U87-MG cells, revealing that the FD22a/CB2R system was able to efficiently suppress β-amyloid-induced enhanced production of proinflammatory cytokines, namely TNFα and IL-6 (Figure 5A,B), and to promote the secretion of anti-inflammatory cytokine IL-10 (Figure 5C).As detected in HMC3 cells, treatment with FD22a (1 µM) alone did not produce any significant effect on cytokine release in U87-MG cells.
FD22a against β-amyloid-induced microglial activation.The same set of experiments was also carried out in U87-MG cells, revealing that the FD22a/CB2R system was able to efficiently suppress β-amyloid-induced enhanced production of proinflammatory cytokines, namely TNFα and IL-6 (Figure 5A,B), and to promote the secretion of anti-inflammatory cytokine IL-10 (Figure 5C).As detected in HMC3 cells, treatment with FD22a (1 µM) alone did not produce any significant effect on cytokine release in U87-MG cells.

FD22a Prevents β-Amyloid-Induced Down-Regulation of Autophagy in U87-MG Cells
ATG is crucial for neuronal homeostasis, and its dysfunction has been directly linked to a growing number of NDDs.Accordingly, the induction of ATG may be exploited as a strategy to assist neurons to survive by clearing abnormal protein aggregates [43].The kinase mammalian target of rapamycin (mTOR) is a central modulator of ATG [44], and a marked up-regulation of mTOR is known to contribute to AD progression in humans [45].Hence, U87-MG cells, which feature a similar mTOR up-regulation leading to ATG suppression [46][47][48][49], may represent the ideal cell line for in vitro studies.
Having assessed that FD22a (1 µM) was able to protect U87-MG cells from β-amyloid-induced cytotoxicity and proinflammatory response, we went on to examine whether exposure of U87-MG cells to Aβ25-35 could lead to further depression of autophagy and whether pretreatment with FD22a could resolve, or at least attenuate, the deleterious effect of β-amyloid in such cells.

FD22a Prevents β-Amyloid-Induced Down-Regulation of Autophagy in U87-MG Cells
ATG is crucial for neuronal homeostasis, and its dysfunction has been directly linked to a growing number of NDDs.Accordingly, the induction of ATG may be exploited as a strategy to assist neurons to survive by clearing abnormal protein aggregates [43].The kinase mammalian target of rapamycin (mTOR) is a central modulator of ATG [44], and a marked up-regulation of mTOR is known to contribute to AD progression in humans [45].Hence, U87-MG cells, which feature a similar mTOR up-regulation leading to ATG suppression [46][47][48][49], may represent the ideal cell line for in vitro studies.
Having assessed that FD22a (1 µM) was able to protect U87-MG cells from β-amyloidinduced cytotoxicity and proinflammatory response, we went on to examine whether exposure of U87-MG cells to Aβ 25-35 could lead to further depression of autophagy and whether pretreatment with FD22a could resolve, or at least attenuate, the deleterious effect of β-amyloid in such cells.

Quantitative Proteomic Analysis Uncovers the Activation of TFEB in the Restoration of Autophagy by FD22a
To uncover potential molecular pathways and protein targets involved in FD22a's effect and in its potential protective action against the deleterious impact caused by βamyloid exposure, we performed a quantitative proteomic analysis of U87-MG cells.The 2DE protein maps of cellular protein extracts obtained from U87-MG cells in different treatment conditions were compared.

Quantitative Proteomic Analysis Uncovers the Activation of TFEB in the Restoration of Autophagy by FD22a
To uncover potential molecular pathways and protein targets involved in FD22a's effect and in its potential protective action against the deleterious impact caused by βamyloid exposure, we performed a quantitative proteomic analysis of U87-MG cells.The 2DE protein maps of cellular protein extracts obtained from U87-MG cells in different treatment conditions were compared.
Overall, an average of 1800 ± 70 spots were found within a linear pH range from 3 to 10.A representative gel image is shown in Figure 7A (representative 2DE images of all groups are shown in Supplementary Materials Figure S1), whereas the Venn diagram shown in Figure 7B shows the number of protein spots found significantly differentially expressed in different comparisons.Overall, an average of 1800 ± 70 spots were found within a linear pH range from 3 to 10.A representative gel image is shown in Figure 7A (representative 2DE images of all groups are shown in Supplementary Materials Figure S1), whereas the Venn diagram shown in Figure 7B shows the number of protein spots found significantly differentially expressed in different comparisons.
Regarding comparison with the control, 67 spots were modified by the treatment with FD22a and 41 by the treatment with Aβ 25-35 , whereas pretreatment with FD22a reduced protein spot changes, as suggested by the FD22a + Aβ 25-35 vs. Aβ 25-35 comparison.A volcano plot was constructed to represent fold change and p-value in protein expression for this comparison.It revealed that about 50 percent of the 20 spots found with a significant change in expression were up-regulated (Figure 7C).These spots were identified by LC-MS/MS.The name of the identified proteins, the molecular weight (MW), isoelectric point (pI), score, coverage values of MS/MS, ratio, and p-values are listed in Table 2, whereas Figure 8 shows box plots of data distribution obtained for these proteins in different treatment conditions.Compared with the Aβ 25-35 group, the combined FD22a and Aβ 25-35 treatment significantly restores to the control value the expression of HS90A/B, lamin A (LMNA), fructose biphosphate-aldolase (AL-DOA), calcyclin-binding protein (CYBP), ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCHL1), and peroxiredoxin 2 (PRDX2).Moreover, Tables 3 and 4 report the proteins that were found differentially expressed when comparing both FD22a and Aβ 25-35 treatment with control experiments, respectively.Regarding comparison with the control, 67 spots were modified by the treatment with FD22a and 41 by the treatment with Aβ25-35, whereas pretreatment with FD22a reduced protein spot changes, as suggested by the FD22a + Aβ25-35 vs. Aβ25-35 comparison.A volcano plot was constructed to represent fold change and p-value in protein expression for this comparison.It revealed that about 50 percent of the 20 spots found with a significant change in expression were up-regulated (Figure 7C).These spots were identified by LC-MS/MS.The name of the identified proteins, the molecular weight (MW), isoelectric point (pI), score, coverage values of MS/MS, ratio, and p-values are listed in Table 2, whereas Figure 8 shows box plots of data distribution obtained for these proteins in different treatment conditions.Compared with the Aβ25-35 group, the combined FD22a and Aβ25-35 treatment significantly restores to the control value the expression of HS90A/B, lamin A (LMNA), fructose biphosphate-aldolase (ALDOA), calcyclin-binding protein (CYBP), ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCHL1), and peroxiredoxin 2 (PRDX2).Moreover, Tables 3 and 4 report the proteins that were found differentially expressed when comparing both FD22a and Aβ25-35 treatment with control experiments, respectively.To explore the molecular pathways involved in FD22a action and its ability to protect against the deleterious effects of Aβ 25-35 , proteins that were found to express differently were analyzed by Ingenuity Pathway Analysis (IPA, QIAGEN Redwood City, CA, USA, www.qiagen.com/ingenuity).An involvement of the CLEAR signaling pathway (z score = 2.2) with a TFEB-dependent activation of autophagy and lysosome biogenesis emerged from the analysis of proteins differently expressed in relation to the treatment with FD22a.Accordingly, an inhibition of NF-κB was observed in the network analysis (Figure 9).
Moreover, a list of potential regulators was obtained from causal network analysis, with significant positive and negative z-scores (Supplementary Table S1): among these, the activation of Next to BRCA1 gene 1 protein (NBR1) (z-score = 3.4, p-value = 1.16 × 10 −8 ), a ubiquitin-binding autophagy adapter which acts as a receptor for selective autophagosomal degradation of ubiquitinated targets, agreed with induction of autophagy by FD22a.On the other hand, IPA of proteins differently expressed after exposure of U87-MG cells to Aβ 25-35 supported a loss of autophagic response with caspase activation and inhibition of sestrin 2 (SESN2) (z-score = −2.1;p-value = 0.00002), a stress-inducible protein, able to activate the specific autophagic machinery for degradation of mitochondria (Supplementary Table S2).
Studies have suggested that the dephosphorylation of TFEB at serine 211 (TFEB-S211) promoted the nuclear entry of TFEB, which regulates the expression of autophagy-related genes [2,50].
Since the IPA suggested that the TFEB pathway could be involved in the action of FD22a, we performed Western blot analysis to evaluate the expression of total TFEB and TFEB-S211 in different treatment conditions.Western blot analysis revealed that exposure of U87-MG cells to 10 µM Aβ 25-35 for 48h led to a significant increase in TFEB-S211 expression (4.2-fold with respect to control, p-value < 0.001), and pretreatment with 1 µM FD22a attenuated the expression of the phosphorylated form by about 26% (Figure 10).
Western blot analysis also revealed that Aβ 25-35 treatment produced an increased expression of the phosphorylated form of mTOR (p-mTOR) in comparison to control cells, whereas the addition of FD22a induced an approximately 33% reduction in mTOR phosphorylation with respect to Aβ 25-35 -treated cells (p-value = 0.048) (Figure 10), thus confirming gene expression results.A reduction of p-mTOR, the active form of mTOR (Figure 10), and an increase in phosphatase expression observed in 2DE (Table 2) may concur with the dephosphorylation of TFEB induced by FD22a.Moreover, a list of potential regulators was obtained from causal network analysis, with significant positive and negative z-scores (Supplementary Table S1): among these, the activation of Next to BRCA1 gene 1 protein (NBR1) (z-score = 3.4, p-value = 1.16 × 10 −8 ), a ubiquitin-binding autophagy adapter which acts as a receptor for selective autophagosomal degradation of ubiquitinated targets, agreed with induction of autophagy by FD22a.On the other hand, IPA of proteins differently expressed after exposure of U87-MG cells to Aβ25-35 supported a loss of autophagic response with caspase activation and inhibition of sestrin 2 (SESN2) (z-score = −2.1;p-value = 0.00002), a stress-inducible protein, able to activate the specific autophagic machinery for degradation of mitochondria (Supplementary Table S2).
Studies have suggested that the dephosphorylation of TFEB at serine 211 (TFEB-S211) promoted the nuclear entry of TFEB, which regulates the expression of autophagy-related genes [2,50].
Since the IPA suggested that the TFEB pathway could be involved in the action of FD22a, we performed Western blot analysis to evaluate the expression of total TFEB and TFEB-S211 in different treatment conditions.Western blot analysis revealed that exposure of U87-MG cells to 10 µM Aβ25-35 for 48h led to a significant increase in TFEB-S211 expression (4.2-fold with respect to control, p-value < 0.001), and pretreatment with 1 µM FD22a attenuated the expression of the phosphorylated form by about 26% (Figure 10).Western blot analysis also revealed that Aβ25-35 treatment produced an increased expression of the phosphorylated form of mTOR (p-mTOR) in comparison to control cells, whereas the addition of FD22a induced an approximately 33% reduction in mTOR phosphorylation with respect to Aβ25-35-treated cells (p-value = 0.048) (Figure 10), thus confirming gene expression results.A reduction of p-mTOR, the active form of mTOR (Figure 10), and an increase in phosphatase expression observed in 2DE (Table 2) may concur with the dephosphorylation of TFEB induced by FD22a.pression of the phosphorylated form.In accordance with gene expression results, Western blot analysis also revealed that Aβ 25-35 treatment produced an increased expression of the phosphorylated form of mTOR (p-mTOR) in comparison to control cells, whereas the addition of FD22a induced a significant reduction of p-mTOR with respect to control and Aβ 25-35 -treated cells.Notably, the detected reduction of p-mTOR and increased expression of phosphatases revealed in 2DE experiments may concur with the dephosphorylation of TFEB induced by FD22a.

Conclusions
Overall, the results of our study highlight the potential for a multitarget treatment of neurodegenerative pathologies via FD22a-mediated activation of CB2R.The observed ability of FD22a to promote TFEB nuclear entry by dephosphorylation of S211, the target serine of mTOR, followed by TFEB-mediated activation of autophagic lysosomal function, associated with the ability to prevent β-amyloid-induced cytotoxic and proinflammatory effects, may have a marked relevance in the prevention and/or treatment of AD pathology.Even though our work provides convincing evidence of the potential of FD22a to target neurodegeneration, extensively illustrating the ability of this novel CB2R bitopic ligand to efficiently counteract the deleterious effects of β-amyloid in human glial cells, further investigations on in vivo models of AD and knockout of CB2R will be necessary to corroborate the therapeutic potential of CB2R activation in slowing or reversing AD.Furthermore, a detailed investigation of FD22a's pharmacokinetic properties will also be fundamental to pursue a future therapeutic application of this novel CB2R bitopic ligand.

Figure 1 .Figure 1 .
Figure 1.Regulation of TFEB activity, a key transcription molecule regulating autophagy.Intracellular molecules, including PI3K, MAPK, AMPK, and TNFR, possess the capability to activate the mTOR pathway, a prominent negative regulator of autophagy.mTOR, functioning as a kinase, plays a crucial role in regulating the localization and activation of TFEB, a crucial transcription factor, which governs autophagy at the transcriptional level by promoting the expression of multiple lysosomal genes, influencing autolysosome production and function.Specifically, activated mTOR phosphorylates TFEB, inhibiting its activity and confining it to the cytoplasm.Inhibiting mTOR leads to TFEB dephosphorylation, enabling its nuclear translocation and transcriptional activity.Within the nucleus, TFEB regulates autophagy in two ways: promoting the expression of autophagy-related molecules for enhanced autophagy, and activating lysosomal-pathway-related

Figure 6 .
Figure 6.FD22a pretreatment prevents Aβ25-35 deleterious transcriptional effects on autophagy and inflammation.Pretreatment with FD22a significantly modulates transcriptional expression of selected autophagy (ATG)-related (A) and proinflammatory (B) genes.Each bar corresponds to the means ± SEM of at least three independent experiments.Data were analyzed by one-way analysis of variance (ANOVA) followed by Tukey's test.$$ p < 0.01 with respect to vehicle-treated cells (control cells) $$$ p < 0.005 with respect to vehicle-treated cells (control cells); * p < 0.05 with respect to Aβ25-35-treated cells; ** p < 0.01 with respect to Aβ25-35-treated cells; *** p < 0.005 with respect to Aβ25-35-treated cells.

Figure 7 .
Figure 7. (A) Representative 2DE image of U87-MG proteome.Protein extracts were separated in a linear pH 3-10 gradient.SDS-PAGE was performed using 12% acrylamide.Gels were stained with fluorescent dye and acquired by ImageQuant TL 7. (B) Venn diagram showing the number of proteins found differentially expressed in the different comparisons: FD22a vs. Ctrl, Aβ25-35 vs. Ctrl, and FD22a + Aβ25-35 vs. Aβ25-35.Both unique and overlapping proteins are reported as numbers (Venny 2.0.2). (C) Scatter plot of fold change (x-axis) against log10 p-value (y-axis) of quantified proteins obtained for FD22a + Aβ25-35 vs. Aβ25-35 comparison.Up-regulated and down-regulated proteins are colored red and blue, respectively.Only proteins that showed both p-value and q-value < 0.05 were identified.Dotted line indicates the threshold of significance.The gene names of identified proteins are shown in the scatter plot.

Figure 7 .
Figure 7. (A) Representative 2DE image of U87-MG proteome.Protein extracts were separated in a linear pH 3-10 gradient.SDS-PAGE was performed using 12% acrylamide.Gels were stained with fluorescent dye and acquired by ImageQuant TL 7. (B) Venn diagram showing the number of proteins found differentially expressed in the different comparisons: FD22a vs. Ctrl, Aβ 25-35 vs. Ctrl, and FD22a + Aβ 25-35 vs. Aβ 25-35 .Both unique and overlapping proteins are reported as numbers (Venny 2.0.2). (C) Scatter plot of fold change (x-axis) against log 10 p-value (y-axis) of quantified proteins obtained for FD22a + Aβ 25-35 vs. Aβ 25-35 comparison.Up-regulated and down-regulated proteins are colored red and blue, respectively.Only proteins that showed both p-value and q-value < 0.05 were identified.Dotted line indicates the threshold of significance.The gene names of identified proteins are shown in the scatter plot.

Figure 9 .
Figure 9. Functional network derived from QIAGEN's Ingenuity Pathway Analysis of proteins differentially expressed in the FD22a + Aβ25-35 vs. Aβ25-35 comparison.The network describes functional relationships among proteins based on known associations in the literature.Solid line: direct interaction; dotted line: indirect interaction.Red and green indicate up-and down-regulated proteins, respectively.Orange suggests an activation whereas blue suggests an inhibition.(*) This protein has been identified in many spots.The number below the protein symbol indicates the fold change value of expression.

Figure 9 .
Figure 9. Functional network derived from QIAGEN's Ingenuity Pathway Analysis of proteins differentially expressed in the FD22a + Aβ 25-35 vs. Aβ 25-35 comparison.The network describes functional relationships among proteins based on known associations in the literature.Solid line: direct interaction; dotted line: indirect interaction.Red and green indicate up-and down-regulated proteins, respectively.Orange suggests an activation whereas blue suggests an inhibition.(*) This protein has been identified in many spots.The number below the protein symbol indicates the fold change value of expression.Cells 2024, 13, x FOR PEER REVIEW 18 of 23

Figure 10 .
Figure 10.Western blot detection and quantification of TFEB, p-TFEB (S211), mTOR, and p-mTOR in U87-MG control cells, after the addition of FD22a and after treatment with Aβ25-35 alone and in the presence of FD22a.Each bar graph represents the mean ± SEM of five independent experiments.Optical density of each immunoreactive band was normalized on total protein obtained from RuBP staining.For p-TFEB and p-mTOR the expression level of TFEB and mTOR, respectively, were used as loading control.An unpaired t-test was used to compare differences among treatments (Prism 7; GraphPad Software, San Diego, CA, USA), * p < 0.05, ** p < 0.01, and *** p < 0.001.ns means not significant.

Figure 10 .
Figure 10.Western blot detection and quantification of TFEB, p-TFEB (S211), mTOR, and p-mTOR in U87-MG control cells, after the addition of FD22a and after treatment with Aβ 25-35 alone and in the presence of FD22a.Each bar graph represents the mean ± SEM of five independent experiments.Optical density of each immunoreactive band was normalized on total protein obtained from RuBP staining.For p-TFEB and p-mTOR the expression level of TFEB and mTOR, respectively, were used as loading control.An unpaired t-test was used to compare differences among treatments (Prism 7; GraphPad Software, San Diego, CA, USA), * p < 0.05, ** p < 0.01, and *** p < 0.001.ns means not significant.

Table 1 .
Primer sequences for Real-Time PCR experiments.

Table 2 .
List of proteins found differentially expressed in the comparison of U87-MG cells treated with FD22a + Aβ 25-35 vs. cells treated with Aβ 25-35 identified by LC-MS/MS.ID: SwissProt accession number, MW: molecular weight, pI: isoelectric point.

Table 3 .
List of proteins found differentially expressed in the comparison of U87-MG cells treated with FD22a vs. cells without treatment identified by LC-MS/MS.ID: SwissProt accession number, MW: molecular weight, pI: isoelectric point.

Table 4 .
List of proteins found differentially expressed in the comparison of U87-MG cells treated with Aβ 25-35 vs. cells without treatment identified by LC-MS/MS.ID: SwissProt accession number, MW: molecular weight, pI: isoelectric point.