Human neuroblastoma cells-SK-N-SH (HTB-11; American Type Culture Collection, Manassas, VA, USA) and HEK293 cells (as positive control) were cultured in DMEM medium supplemented with 10% FBS and 1% antibiotics (penicillin and streptomycin), grown at 37°C and subcultured every 48 h. We measured FFAR2 expression in these cells by harvesting total RNA while cells were at 70–80% confluency, using the RNAeasy kit (Qiagen). The expression of FFAR2 mRNA was quantified using real-time PCR after converting total RNA to cDNA using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems). 18S RNA was used as an internal control. The fold change in gene expression was determined using delta-delta Ct values, following our routinely used well-established and published protocols (Nagpal et al., 2018; Ahmadi et al., 2019, 2020a,b; Wang et al., 2020).

To determine the impact of FFAR2 inhibition on Aβ-induced neuronal cell toxicity, we performed the MTT viability assay (Ronicke et al., 2008). SK-N-SH cells (1 × 10³ cells per well) were plated in a 96-well plate containing DMEM medium supplemented with 10% FBS and 1% antibiotics and grown at 37°C for 24 h. Then, media was removed and cells were washed with PBS before being replenished with DMEM media containing 1% FBS and 10 μM FFAR2 inhibitor (CATPB, Tocris) and incubated at 37°C for 2 h. Then, cells were treated with Aβ 25–32 peptide (25 μM, Sigma) and incubated for 4 h. Afterward, the media was replaced with 100 μl of fresh media, and 10 μl of MTT reagent (5 mg/ml in phosphate buffer [pH 7.2]) was added and mixed in each well before an additional 4 h incubation. Then, DMSO (50 μl) was added in each well and mixed thoroughly to dissolve formazan crystals. The absorbance was measured at 570 nm.

The primary protein sequences of human and mouse FFAR2 were retrieved from the Uniprot database in FASTA format with the Uniprot IDs O15552 and Q8VCK6, respectively. The structures of 144,856 natural compounds were obtained from ZINC database (Supplementary Table S1). Through in-silico virtual screening of these compounds based on binding to human FFAR2 using AutoDock6 software, we shortlisted the top 15 compounds (Supplementary Table S2). Furthermore, the tertiary structures of both mouse FFAR2 and human FFAR2 were predicted using the threading-based online server I-TASSER (Zhang, 2008; Roy et al., 2010; Yang et al., 2015), refined using ModRefiner (Xu and Zhang, 2011), and validated using the Ramachandran plot from RAMPAGE server (Lovell et al., 2003). The natural activators, including acetate, butyrate, and the shortlisted compounds, were blindly docked with both mouse FFAR2 and human FFAR2 using AutoDock Vina to find the compounds that demonstrated activator properties (Trott and Olson, 2010). AutoGrid program was used to fix the grid box around the proteins. The grid box for the human FFAR2 binding site was set to the XYZ coordinates of 68.028, 66.787, and 54.804, respectively. The box dimensions were 28.089, 28.562, and 25.349 along the X, Y, and Z axes, respectively. For the mouse FFAR2, the grid box binding site was set to XYZ coordinates of 63.359, 66.205, and 46.924, and the box dimensions were 40.229, 41.757, and 36.283, respectively. The nine conformations along with the Vina score were used to screen the activator. The interaction of the ligands was analyzed using the 2D interaction diagrams generated by using LigPlot+ (Laskowski and Swindells, 2011) and 3D interaction images using PyMOL2.3 (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC).

To determine the FFAR2 signal activating property of selected compounds, we treated SK-N-SH cells with 10 μM of shortlisted compounds for 30 min at 37°C and harvested for total protein extraction. We harvested cells after 30 min of incubation because ERK1/2 phosphorylation upon FFAR2 activation is a quick event (Jiao et al., 2002; Bhattacharjee et al., 2010). Total protein from the cells was extracted using a lysis buffer [consisting of 10 mmol/L Tris (pH 7.6), 150 mmol/L NaCl, 10 mmol/L Sodium orthovanadate, 10 mmol/L Sodium Pyrophosphate, 100 mmol/L Sodium fluoride, 1 mmol/L EDTA, 1 mmol/L EGTA, 1% Triton X-100, 0.5% NP-40 and a cocktail of protease inhibitors (Roche); Yadav et al., 2011, 2017; Ahmadi et al., 2019, 2020a]. After sonication, cell extracts were centrifuged at 8,000× g for 10 min at 4°C, and the supernatant was separated in a fresh tube. The total protein concentration of the supernatant was measured using a BCA protein assay kit. An equal amount of total protein (45 μg) from each sample was dissolved in loading buffer and loaded in 4% stacking gel and 10% separating SDS-PAGE gel and resolved at 100 V. Proteins were transferred on a 0.22 μm PVDF membrane and blots were developed with anti-ERK1/2, anti-phosphoERK1/2, ubiquitin (P4D1), and Aβ (6E10) primary antibodies followed by their secondary antibodies and bands were visualized using the Pierce™ Fast Western Blot Kit, ECL Substrate acquired using Syngene Pxi. Band densities were also quantified using NIH ImageJ software. β-actin was used as an internal control.

FFAR2 is a Gi/o signaling receptor, which reduces cytosolic cAMP (Falomir-Lockhart et al., 2019). To determine the FFAR2 signaling activating potential of selected compounds, we used the cAMP-Glo™ Assay to measure intracellular cAMP. In brief, SK-N-SH cells (1 × 10³ cells per well) were plated in a white 96-well plate for 24 h and then treated with 10 μM compounds and 10 μM forskolin (a positive control) dissolved in induction buffer. The cAMP levels were detected using a Luminescence Microplate Reader.

FFAR2 signaling activation increases intracellular calcium (Falomir-Lockhart et al., 2019). To further determine the activation potential of selected compounds for this signaling, we plated the SK-N-SH cells (1.25 × 10³ cells per well) into a dark 96-well plate for 24 h. Then cells were treated with 10 μM selected compound(s) and 10 mM acetate (an SCFA positive control). Also, 0.01 M EDTA (Sigma) and 114 nm carbachol were used as negative and additional positive controls, respectively. After 15 min, the cells were treated with 2× fluo-4 Direct™ calcium reagent along with 250 mM probenecid and further incubated at 37 °C for 30 min. The fluorescence was measured with excitation at 494 nm and emission at 516 nm using a microplate reader.

We have used C. elegans as a model for determining the impact of selected compounds on aging and AD-like phenotypes. The wild-type (N2) and CL2122 (dvls15 [9pPD30.38) unc-54(vector) + (pCL26)mtl-2::GFP]) were propagated at 20°C. AD transgenic worms model CL2006 (dvls [pCL12(unc-54/human abeta peptide 1–42 minigene) + pRF4), CL4176 (dvls27[myo-3p::A-Beta (1–42)::let-851 3′UTR) + rol-6 (su1006] and CL2355 (dvls 50 [pCL45 9snb-1::Abeta 1–42::3′ UTR 9 log) + mtl-2::GFP]) were maintained at 16°C. All strains were cultured on solid nematode growth medium (NGM) consisting of 3 g/L NaCl, 2.5 g/L peptone, 5 mg/L cholesterol, 1 mM CaCl2, 1 mM MgSO₄, 25 mM KH₂PO₄, and 17 g/L agar and seeded with OP50 (Escherichia coli strain) as a food supply. The worms were allowed to grow up to the adult stage, and egg carrying worms were used to isolate age-synchronized eggs using the bleaching method (Porta-de-la-Riva et al., 2012). The synchronized eggs were cultured on an NGM plate or fresh S-Complete plus media with 5% OP50, and 10 μM Fenchol (Sigma) at either 20°C (N2 and CL2122) or 16°C (CL2006, CL4176, and CL2355). All of the strains used in this study were acquired from the University of Minnesota- Caenorhabditis Genetics Center, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440).

To determine the impact of selected compounds on age-related decline in brain health, we first measured the lifespan of the worms. The L1 worms were cultured in 96-well plates on S-complete media containing E. coli OP50 and antibiotics (about 10–15 animals per well). The experimental group worms were treated with 10 μM Fenchol and control groups were treated with 0.1% DMSO. The 0.6 mM Fluorodeoxyuridine (FUDR; 30 μl) was added to each well after 3 days and animals were monitored every other day for viability. The CL4176 and CL2006 worms were grown at 16 °C and N2 worms were grown at 20°C.

For paralysis assays, synchronized eggs of CL4176 were kept at 16°C for 48 h, and the worms were transferred onto NGM plates. Experimental group worms were treated with 10 μM Fenchol and control groups were treated with DMSO at 25 °C for 20 h. Counting was performed at 2 h intervals, typically after 14 h, at 23°C, and until the last worm became paralyzed. If worms did not move or only moved their head but not their body when gently touched with a platinum loop, they were scored as paralyzed. Each C. elegans experiment was conducted with at least 100 worms.

To detect amyloid plaques in the transgenic CL2006 worms, age-synchronized worms were grown on media with and without Fenchol supplementation for up to 5 days of age and then worms were fixed in 4% paraformaldehyde- PBS. Then, the worm bodies were permeabilized in 1% Triton X-100, 125 mM Tris (pH 7.4), and 5% fresh β-mercaptoethanol. After washing three times with 1% PBS-Tween-20, the worm bodies were stained with 0.125% thioflavin S in 50% ethanol for 2 min. After washing three times with ethanol in the darkroom, worms were mounted on slides for microscopy using a Zeiss 710 confocal microscope equipped with a digital camera. A total of 20 worms were counted from each group.

Both C. elegans strains (CL2122 and CL2355) were grown on media supplemented with 10 μg/ml Fenchol and DMSO (control) and OP50 lawn at 16 °C for 36 h, and then at 23 °C for another 36 h. Then the worms were washed twice with S-Basal media and transferred to an NGM plate without E. coli OP50. These plates were divided into four equal sectors, with 1 μl of 0.1% benzaldehyde as an odorant and 1 μl of 1 M sodium azide as an anesthetic to arrest the worms upon reaching the quadrant were added to the original spot. On the opposite side of the attractant, a 1 μl drop of sodium azide and 1 μl of control odorant (100% ethanol) were added onto the NGM agar plate. Around 40–45 worms were placed in the center of each plate, incubated at room temperature for 1 h, and transferred to 4°C overnight. The chemotaxis index (CI) was scored using the following formula: (number of worms at the attractant location − number of worms at the control location)/total number of worms on the plate (Margie et al., 2013).

To determine the impact of Fenchol treatment on the AD mice model, we treated 3 months old male APP/PS1 mice (Jackson Laboratory- Stock number: MMRRC Stock No: 34829-JAX) with Fenchol (80 mg/kg body weight dose by daily oral gavage) for 3 months and compared them to their DMSO treated controls. Brains were collected and immediately homogenized in proteasome lysis buffer and proteasome activity was measured as described below. All the animal procedures and experiments were approved by the institutional animal care and use committee (IACUC) of the Wake Forest School of Medicine and the University of South Florida.

The CL2006 strain, SK-N-SH cells, and brain tissues were sonicated in a lysis buffer (150 mM NaCl, 50 mM HEPES, 2 mM DTT, 20 mM NaF, 5 mM EDTA, 2.5 mM sodium pyrophosphate, 1 mM sodium orthovanadate, and 1 mM α-glycerophosphoric acid) and then centrifuged at 15,000× g at 4 °C for 15 min. Total protein was assayed using BCA kit and 20 μg protein was mixed with 140 μM suc-leu-leu-val-tyr–7-amino-4-methylcou-marin (AMC) in a proteasome activity assay buffer (150 mM NaCl, 50 mM HEPES [pH 7.4], 5 mM ATP, and 5 mM EDTA), and then the fluorescence (excitation 380 nm and emission 460 nm) was measured every 10 min over 3 h using a microplate reader. The assay was performed with and without the proteasome inhibitor MG132 (10 μM).

The lysosomal activity was measured in 1 × 10³ SK-N-SH cells per well plated into a 96-well plate and treated with 10 μM Fenchol, 25 μm Aβ 25-32, 1% DMSO, and 500 nm bafilomycin. Bafilomycin was used as an inhibitor of cellular autophagy/lysosomal activity. The lysosomal activity was assayed using the Enz^® lysozyme assay kit (E-22013, Molecular Probes) and the fluorescence was measured in a fluorescence microplate reader using excitation and emission wavelengths of 494 nm and 518 nm, respectively (Biotek Instruments, VT, USA).

Cells were treated with 10 μl Fenchol and DMSO. After overnight incubation, Aβ 25-32 (25 μM) peptide was added to the cells for 4 h. The Senescence β-Galactosidase Staining Kit (Cell Signaling, USA) was used for the detection of the expression of β-galactosidase activity, following the manufacturer’s instructions.

The Student t-test and one/two way ANOVA have been used to determine the statistical significance, as appropriate. All the assays were done in triplicate and repeated two-three times, and values were presented as the mean and standard error of means. P-values less than 0.05 were considered statistically significant.

To determine the role of FFAR2 in neurons, we first confirmed and found that the FFAR2 gene is abundantly expressed in human neuronal SK-N-SH cells, similar to HEK293 cells, which were used as a positive control (Figure 1A). As expected, we observed that Aβ- treatment significantly induced cell death in SK-N-SH cells (Figure 1B). However, the inhibition of FFAR2 using a small compound inhibitor (CATPB) further exacerbated Aβ-induced cell death in these cells (Figures 1C,D). These results indicate that FFAR2 is expressed in the neuronal cells and its inhibition exacerbates the detrimental effects of increased Aβ and decreases the survival of neuronal cells, suggesting that the activation of FFAR2 signaling may reduce Aβ-induced neuronal toxicity.

Based on the in-silico screening of >144,000 natural compounds and their binding affinity to the active site of human FFAR2, we shortlisted the top 15 compounds (Figure 2A, Supplementary Table S2). To determine the interactions of these selected natural compounds and predict their activator vs. inhibitory properties to FFAR2, their interaction profiles were compared with known FFAR2 activators, like acetate and butyrate (Supplementary Tables S3A,B). The binding site residues Ser86, Tyr90, His140, Ile145, Val179, Arg180, Leu183, Tyr238, His242, and Arg255 were identified, similar to the published literature (Schmidt et al., 2011). We observed that acetate forms an H-bond with His242 in human FFAR2 and Arg65, Trp75, Tyr90, Gln148, Tyr238 in mouse FFAR2. Further, butyrate forms an H-bond with His242 in human FFAR2 and Arg65, Trp75, Tyr90, Gln148, Tyr238, and Arg255 in mouse FFAR2. Based on these (modeling) criteria, we classified the selected 15 compounds are potential ligands/activators for FFAR2.

In addition, to determine the FFAR2 activator activity of the selected 15 compounds in SK-N-SH cells, we used three levels of biological system screening based on measuring phospho-ERK1/2 and intracellular levels of cyclic AMP and calcium. Since FFAR2 activation increases phosphorylation of tyrosine amino acid of ERK1/2 (Liu et al., 2008), we used this as a first-level biological screening; and found a descending order of phopho-ERK1/2 levels as follows: Fenchol > Octopamine > Carvamenthenol > Camphene > Pentanone > Synephrine > Leucinamide > Carene > α-Terpineol > Breneol > L-Fenchone > D-Fenchone > Norpseudoephedrine HCl (Figure 2B). The levels of phospho-ERK1/2 proteins were highest in Fenchol and comparable to acetate, a known FFAR2 activator, in comparison to the rest of the selected compounds tested here (Figure 2B, Supplementary Figure S2B). Because FFAR2 activation decreases intracellular cAMP (Falomir-Lockhart et al., 2019) and increases calcium, we also verified and shortlisted our selected compounds to stimulate FFAR2 signaling by measuring intracellular cAMP (Falomir-Lockhart et al., 2019) and intracellular calcium levels (Al Mahri et al., 2020). Interestingly, Fenchol showed the highest decrease in cAMP in SK-N-SH cells upon Forskolin-stimulation, and such decreases were shown as Fenchol < Carvamenthenol < α-Terpineol < Carene < Pentanone < Octopamine < Camphene, in ascending order (Figure 2C). We used Forskolin as an inducer of an intracellular rise in cAMP levels (Curtin et al., 2006). Fenchol also showed the highest increase in intracellular calcium levels in SK-N-SH cells (Figure 2D). These results suggest that Fenchol is a potent activator of FFAR2 signaling and is conferred by an increase in phopho-ERK1/2 and calcium levels and a suppression of cAMP pathways.

Our further bioinformatics analyses revealed that the Fenchol (ZINC01081099) forms H-bonds with the Ser86, Gln148, Glu166, Tyr238, and His242 residues of human FFAR2 and the Thr85 and Glu320 residues of mouse FFAR2; these binding patterns are similar to those of acetate. The lowest Vina scores of −3 kcal/mol, −3.2 kcal/mol and −4.3 kcal/mol, −3.8 kcal/mol were obtained for acetate and butyrate of human and mouse FFAR2, respectively, while Fenchol also showed the lowest Vina scores of −5.5 kcal/mol and −5.3 kcal/mol with human FFAR2 and mouse FFAR2, respectively (Figures 3A–D). Further, molecular modeling analyses indicated that Fenchol and acetate have similar binding to human and mouse FFAR2 (Figures 3A–D), suggesting that Fenchol binds to the active site to stimulate FFAR2 signaling.

Since the inhibition of FFAR2 exacerbates Aβ-induced neurotoxicity, we also determined the effects of Fenchol treatment against Aβ-induced SK-N-SH neuronal cell death. We show that the Fenchol treatment significantly reduced the Aβ-induced neuronal cell death in SK-N-SH cells compared to their DMSO controls during the MTT assay (Figure 3E). Interestingly, treatment with FFAR2 inhibitor and Fenchol demonstrated no significant protection in cell viability, indicating that the protective effects of Fenchol in Aβ-induced cell neurotoxicity are dependent on the activation of FFAR2 signaling. We also observed that Fenchol treatment significantly reduced Aβ-accumulation in SK-N-SH cells compared to their Aβ-only treated controls, again in an FFAR2 dependent manner (Figures 3F,G). This finding suggests that Fenchol treatment reduced Aβ-accumulation in neuronal cells by activating FFAR2 signaling, thus preventing cell death in the SK-N-SH cell culture system. However, the physiological importance of Fenchol in Aβ-accumulation and the underlying cellular and molecular mechanisms are still not known.

To demonstrate the physiological importance of Fenchol activity in reducing Aβ-induced neurotoxicity, we used C. elegans as a model of AD. Aging is one of the major risk factors for AD pathology and an increased risk of Aβ-induced neuronal cell death (Guglielmotto et al., 2010). Interestingly, we show that Fenchol treatment had no significant impacts on the lifespan of wild-type N2 worms (Figure 4A), while it significantly increased the survival/lifespan in worms overexpressing human Aβ (CL2006 and CL4176; Figures 4B,C). Fenchol treatment significantly reduced Aβ accumulation in worms overexpressing human Aβ (Figures 4D,E). These results suggest that the Fenchol treatment prolonged the lifespan of worms by reducing the Aβ-induced neurotoxicity. Further, Fenchol treatment decreased paralytic attacks in worms overexpressing human Aβ, compared to their non-treated controls (Figure 4F). Worms overexpressing human Aβ (CL2355) treated with Fenchol also exhibited significantly higher memory behavior during the chemotaxis assay compared to their non-Aβ expressing control worms (CL2122; Figure 4G). Overall, these results suggest that Fenchol treatment reduced neuronal cell death by reducing Aβ-accumulation, which in turn improved cognitive function and improved organism survival.

So far we have demonstrated that Fenchol protects neuronal cells by reducing Aβ accumulation. However, the mechanisms were not known. We show that the Fenchol treatment did not change protein ubiquitinylation in the SK-N-SH cells (Figure 5A), suggesting that the reduction in Aβ accumulation with Fenchol treatment was not mediated through ubiquitin-dependent protein degradation. Interestingly, Fenchol treatment significantly increased the proteasomal activity in both SK-N-SH cells and C. elegans (Figures 5B,C), as well as in the cortex and hippocampus of our AD mouse model (APP/PS1; Figures 5D,E). These results indicate that Fenchol treatment reduces and clears Aβ accumulation by increasing proteasomal activity in the neuronal cells, thus reducing the neurotoxicity in these cells.

To further demonstrate the mechanism of Fenchol on neuronal cell survival, we show that the Fenchol treatment significantly increased the lysosomal activity in the neuronal cells, compared to their controls (Figure 6A). This finding suggests that Fenchol increases autophagy in the neuronal cells and, thus, might be protecting these cells from Aβ-induced neurotoxicity. We also observed that the Aβ-treated cells showed a higher rate of senescence (indicated by β-galactosidase activity staining), while Fenchol treatment significantly reduced the number of senescent neuronal cells (Figure 6B). These results suggest that the mechanism of Fenchol includes improved autophagy, reduced senescence in neurons (thus, possible protection from neuronal cell death), and amelioration of AD-like pathologies in the brain.