Antibodies against the following proteins were used: insulin receptor substrate (IRS), p-IRS (Tyr612) and PTEN (from Abways Technology); PI3K, p-PI3K (p85), p-Akt (Ser473), Akt, p-mTOR (Ser2448), ERK, p-ERK (Thr202/Tyr204), Protein G Magnetic Beads, and LC-3 (Cell Signaling Technology); p-GSK (Ser9), GSK and P62 (Abcam); PP2ACα subunit, PP2ACα with phospho-Y307 (Bioword); demethylated PP2AC subunit (Millipore); Rhes (Signalway Antibody); Flag, T181-Tau, S396-Tau, AT8-Tau, and p-ribosome protein S6 (Ser235/236) (Cell Signaling Technology); p-4EBP1 (T70), p-4EBP1 (T37/46), and 4EBP1 (ABclonal Technology); β-actin was from Santa Cruz Biotechnology. DAPI Alexa Fluor-594 or Alexa Fluor-488 labeled secondary antibodies were from Invitrogen; IRDye labeled secondary antibodies were from LI-COR. Lonafarnib, MK2206, MG132, and bafilomycin A1 were purchased from MCE Company. Thermolysin was from Jinpin Bio Company.

A 293 T (human embryonic kidney cell line expressing SV40 T antigen) and SH-SY5Y (human neuroblastoma cell line) cells were obtained from the Cell Bank, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco) at 37°C in a humidified atmosphere containing 5% CO₂ and passaged every 3–4 days.

Although Tau has 86 phosphorylation sites, phosphorylation at the AT8 sites has been shown to be the most pathogenic (Sun and Gamblin, 2009). Therefore, Tau phosphorylated at AT8 for insulin signaling was analyzed in present study. Construction of 2N4R-Tau and Tau with pseudo-phosphorylated at AT8 plasmids were described by our previous report (Cao et al., 2018). AT8 epitope of Tau includes phosphorylation atS199, S202, and T205 (Jeganathan et al., 2008). To mimic Tau phosphorylation at AT8 epitope (PTau), serine and threonine residues at AT8 epitope were substituted to glutamate. To simulate AT8-unphosphorylated Tau (uPTau), serine and threonine residues at AT8 were changed to alanine, which have been described in our previous study (Cao et al., 2018). The cells were transfected with plasmids using Lipofectamine 3000 (Invitrogen) at 24 h after planting (75–80% confluency) as described previously (Cao et al., 2018).

Cells were cultured and maintained on a round slide. After treatment with insulin, cells were washed with 0.01 M PBS followed by a fixation using 4% PFA for 10 min at room temperature. The primary antibody was added to the cells followed by incubation for 48 h at 4°C. Then, the second Alexa Fluor® antibody (Invitrogen) was adopted and incubated with cells for 1 h at 37°C. The nucleic acids were stained with DAPI (Invitrogen). After mounting with anti-fade medium (Sigma), immunofluorescent images were acquired using a fluorescence microscope (Nikon).

Cells in 100-mm dishes were rinsed with PBS, scraped into 1 mL of lysis buffer containing 50 mM Tris–HCl (pH 7.5), 150 mm NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 1 mM EDTA, and protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany), which were then sonicated. Cellular debris was removed by centrifugation at 12,000 × g for 15 min at 4°C. The supernatants were incubated with anti-Flag antibody or anti-PTEN for one night at 4°C, then the solution was incubated with Protein G Magnetic Beads for 1 h at room temperature with mixing. The beads were washed with cell lysis buffer four times to remove the unbound immune complex. Then bound proteins were eluted with SDS sample buffer for immunoblot analysis.

The cellular thermal shift assay (CETSA) is used to study thermal stabilization of proteins upon ligand binding in cells (Nagasawa et al., 2020). Samples were prepared from control and drug-exposed cells. For each set, cells were treated with lonafarnib (1 μM) or DMSO as control for 24 h. Cells were washed ice-cold PBS and collected by trypsinization, neutralized by DMEM containing 10% FBS, pelleted at 1000 × g for 5 min. Cells were then incubated in lysis buffer (50 mM Tris–HCl, pH 7.5, 0.5% Triton X-100, 200 mM NaCl, 10% glycerol) containing a complete protease inhibitor cocktail tablet (Roche Diagnostics, Mannheim, Germany) on ice. Cell debris was removed by centrifugation, and the supernatant was aliquoted. Aliquots were heated to designated temperatures for 3 min, and cooled at room temperature for 3 min, and then cooled and centrifuged at 20,000 × g, 4°C for 20 min. The proteins in the supernatant were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting.

The drug affinity responsive target stability (DARTS) assay is based on the principle that the susceptibility of the target protein to proteases is reduced upon drug binding (Lomenick et al., 2009; Pai et al., 2015). In brief, cells transfected different Tau constructs were pretreated with 1 μM of lonafarnib or DMSO as vehicle control for 24 h, washed with pre-cooled PBS, and lysed in 50 mM NaCl, 10 mM CaCl₂, and 50 mM Tris–HCl (pH7.5), and containing 5% Triton X-100, with protease and phosphatase inhibitor cocktails, and then incubated at 4°C. Lysates were centrifuged for 10 min at 18,000 × g at 4°C to pellet cellular debris. Chilled TNC buffer (50 mM Tris–HCl pH 7.5, 50 mM NaCl, and 10 mM CaCl₂) was added to the supernatant of protein lysate, and protein concentration of the lysate was measured by the BCA Protein Assay kit (Biomiga, PW0104). Then thermolysin was added to cell lysates at several designated final concentrations to cell lysates. Thermolysin reaction mixtures were incubated at room temperature for 10 min. Samples were boiled immediately after adding loading buffer to stop the digestion, and analyzed by SDS-PAGE and western blotting.

Cells were collected, washed with ice-cold PBS, lysed in RIPA buffer (150 mM NaCl, 50 mM Tris–HCl, pH7.4, 1% TritonX-100, 1% sodium deoxycholate, and 0.1% each of SDS, sodium orthovanadate, sodium fluoride, and EDTA) with a mixture of protease and phosphatase inhibitors for 30 min on ice. Samples were then sonicated on ice, and the protein concentration in each sample was quantified with a bicinchoninic acid protein assay kit (PW0104, Biomiga). Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes, which were then treated with blocking buffer TBST (137 mM NaCl, 0.1% Triton X-100, 20 mM Tris–HCl, pH 7.4) containing 5% BSA for 1 h at room temperature and incubated overnight at 4°C with primary antibody. After being washed with TBST several times, the membranes were reacted with IRDye® secondary antibody for 1 h at room temperature, washed 3 times with TBST, and analyzed with the Odyssey IR Imaging System (Li-COR).

Analysis was conducted with the GraphPad Prism Version 7.0. For western blotting protein expression data, One-Way ANOVA (with Tukey Post hoc test) was performed. The outcomes are presented as bar chart with error bars representing the mean ± SEM, respectively.

293 T cells, which lack of Tau (Santa-Maria et al., 2007), were transfected with constructs expressing wild-type 2N4R-Tau (WTau) or Tau with pseudophosphorylation at AT8 epitopes (PTau), and the effects of the overexpression of both on elements in the ISP were investigated by western blotting (Figure 1).

ISP activity is reflected by the phosphorylation of insulin receptor substrate-1 at Tyr612 (p-IRS-1), PI3K at Tyr458 (p-PI3K) and Akt at Ser473 (p-Akt), and the phosphorylation of Akt downstream factors, including GSK3β at Ser9 (p-GSK3β) and mTOR at Ser 2,448 (p-mTOR; Lomenick et al., 2009; Akhtar and Sah, 2020). In this study, a comparison of 293 T cells transfected with WTau and those transfected with an empty plasmid (referred to as control cells) indicated an increased basal ISP activity. Among the ISP proteins, the increases in the levels of p-IRS (0.05), IRS (p < 0.01), p-PI3K (p < 0.05), p-IRS/IRS (p < 0.01), and p-PI3K/PI3K (p < 0.05) were statistically significant, whereas p-mTOR and mTOR also showed an increasing tendency (Figures 1A,C).

Compared with the control cells, an increased tendency for ISP elements and their phosphorylation was also observed in PTau cells, and the increases in IRS, p-IRS (Tyr612), p-IRS/IRS were statistically significant (p < 0.01, p < 0.05, and p < 0.01 respectively). In contrast, compared with the WTau cells, the PTau cells had relatively lower ISP activity; however, only p-PI3K and total mTOR were statistically significant (p < 0.05; Figures 1B,D). Therefore, Tau overexpression could elevate the basal activity of the ISP, whereas Tau phosphorylation at the AT8 epitope somewhat reduced this effect.

To investigate the effects of WTau and PTau on the insulin-induced ISP reaction, cells were treated with 200 nM insulin for 10 min (Figure 1). In general, insulin treatment caused ISP activation in the control, WTau, and PTau cells.

The insulin-treated WTau cells were found to have relatively higher levels of p-IRS (Tyr612) (p < 0.01), p-PI3K (p < 0.01), p-Akt (p < 0.05), p-IRS/IRS (p < 0.01), and p-PI3K/PI3K (p < 0.01), p-Akt/Akt (p < 0.05) as the up-elements of ISP, compared to the insulin-treated control cells, but not downstream factors, such as GSK3β and mTOR (Figures 1A,C). In addition, insulin-treated PTau cells had a significantly higher level of p-IRS (p < 0.01) than the insulin-treated control cells, but not other ISP elements. However, compared to insulin-treated WTau cells, insulin-induced activation of ISP elements tended to be decreased, with decreases in p-PI3K (p < 0.01), PI3K (p < 0.01), p-Akt (p < 0.01), p-IRS/IRS (p < 0.05), p-PI3K/PI3K (p < 0.05), and p-Akt/Akt (p < 0.01) levels been significant (Figures 1A,C). These results indicate that the overexpression of Tau could cause a higher reaction in the upstream ISP elements, whereas Tau phosphorylation weakens this effect.

To determine whether the extended effects of Tau on the insulin-induced ISP activation, the cells were treated with insulin for 30 min (Figure 2). The upregulation of the upstream portion of ISP elements, p-IRS (p < 0.05), IRS (p < 0.01), and p-Akt (p < 0.01) was still obvious in the control cells (Figures 2A,B). Compared to insulin-treated control cells, insulin-treated WTau cells had significantly higher levels of p-IRS (p < 0.05), but not insulin-treated PTau cells. In contrast, insulin-treated WTau and PTau cells showed lower levels of p-Akt (p < 0.01) than those in insulin-treated control cells.

Meanwhile, immunofluorescence staining also showed that cells overexpressing WTau or PTau contained a higher basal level of p-IRS than that in control cells. After insulin treatment, cells overexpressing WTau or PTau demonstrated more p-IRS distribution than that in insulin-treated control cells. There was no obvious difference of p-IRS staining between the cells overexpressing WTau and PTau (Figure 3). Based on the results of 10- and 30-min insulin treatment, WTau and PTau overexpression affected the insulin-induced signals transduction of ISP.

Besides IRS/PI3K/AKT is the main ISP, insulin is also able to induce Ras and ERK, which could crosstalk with the main pathway, although Ras–ERK appears to be more acitve in response to IGFIR signaling compared to IR signaling (Cai et al., 2017). We observed that the basal level of p-ERK/ERK in WTau cells tended to increase, whereas p-ERK in PTau cells tended to decrease, however both of them did not reach statistical significance. Insulin treatment causes an activation of ERK in control cells (p < 0.05), but p-ERK/ERK was declined in WTau cells (p < 0.05), whereas no change in PTau cells (Figures 1E,F). Therefore, insulin induced ERK reaction also altered in WTau or PTau cells.

The alteration of mTOR signaling, especially for factors downstream of mTOR, is known to occur early in the progression of AD (Tramutola et al., 2015) and at the severe stages of AD (Sun et al., 2014). Studies have suggested that Aβ is involved in abnormalities in the PI3K/Akt/mTOR axis in AD (Caccamo et al., 2010; Gupta and Dey, 2012; O'Neill et al., 2012). This study evaluated the effects of the overexpression of Tau and its pseudo-phosphorylated form on the phosphorylation of ribosomal protein S6 and eIF4E binding protein (4EBP), the downstream factors of mTOR, which are reportedly elevated in AD (Tramutola et al., 2015; Sun et al., 2020).

Compared with the control cells, the basal levels of p-S6 (3.31 fold, p < 0.01), p-4EBP1-T70 (1.85 fold, p < 0.05), and p-4EBP1-T37/46 (1.54 fold, p < 0.05) in WTau cells were significantly increased, and the increase in phosphorylation of S6 was the most dramatic (Figure 4A). In addition, PTau cells showed an increased levels of p-S6 (3.3 fold, p < 0.01), p-4EBP1-T70 (1.94 fold, p < 0.05) and p-4EBP1-T37/46 (1.7 fold, p < 0.01). There were no significant differences in these proteins between the WTau and PTau cells.

However, WTau and PTau overexpression significantly suppressed the insulin-induced increase in p-S6 (2.25 and 1.99 fold, respectively), in contrast to that of control cells (8.6 fold; Figure 4A). For p-4EBP1-T70, insulin treatment caused an increase for control cells (p < 0.05), but did not change WTau cells, and even caused a reduction of PTau cells. In addition, insulin-induced 4EBP1 phosphorylation at the T37/46 sites in these cells showed a pattern similar to that of p-4EBP1-T70. Therefore, these results indicated that the overexpression of Tau and its phosphorylated form could significantly elevate the basal levels of pS6 and p-4EBP1, but alleviate the insulin-induced increase in p-S6 and p-4EBP1, particularly in PTau cells.

In this study, the overexpression of WTau and PTau in 293 T cells significantly affected the basal levels of the upstream factor p-IRS and downstream factors such as p-S6 and p-4EBP1, whereas the phosphorylation of Akt, a midstream factor, was barely affected. Akt is known to be an important connecting link between the up- and the down-stream elements in ISP (Copps and White, 2012). To explore whether Akt contributes to the abnormalities in these upstream and downstream factors in WTau and PTau cells, the cells were treated with MK2206, an Akt inhibitor (Hirai et al., 2010; Xiang et al., 2017; Bjune et al., 2018). After MK2206 treatment for 12 h, the decline of p-Akt was expected. Moreover, the levels of both p-mTOR and mTOR in WTau- and PTau-transfected cells were significantly downregulated (p < 0.01 and p < 0.05, respectively); however, the decline was not significant in the control cells (Figures 5A,B). Moreover, the MK2206 treatment caused a decline in p-S6 to a similar level among these cells (Figures 5A,B). Therefore, we speculated that Akt is involved in ISP dysregulation in WTau and PTau cells.

Furthermore, MK2206 upregulated IRS levels in the control cells (p < 0.01), but tended to downregulate IRS levels in WTau and PTau cells (p < 0.05; Figures 5A,B). These results indicate that the regulation of insulin signaling in WTau- or PTau-transfected cells was different from that in control cells. Blockage of either autophagic degradation with bafilomycin A1 (Baf-A1) or proteasome degradation with MG132 in MK2206 treated PTau cells (Figures 5C,D) elevated the levels of mTOR and IRS, suggesting that Akt plays a role in the regulation of mTOR and IRS levels in Tau overexpressing cells.

The autophagic system is another downstream effector of the mTOR pathway. The effects of WTau and PTau on autophagy were then investigated by examining SQSTM1P62 and LC-3 (microtubule-associated protein 1 light chain 3)-II (Figures 4C,D). The results showed that WTau and PTau cells had higher basal levels of P62 than the control cells (p < 0.01); while the basal level of LC-3 in WTau and PTau cells also tended to be higher than that in the control cells, especially for PTau cells (p < 0.01). Therefore, the overexpression of WTau or PTau affects the basal condition of the autophagic system. However, insulin treatment for 30 min did not cause an increase in P62 in WTau- or PTau -cells as control cells, whereas it downregulated the LC-3 levels in WTau and PTau cells (p < 0.05 and p < 0.01, respectively) to a level close to that of the insulin-treated control cells.

The ISP is regulated by phosphatases, among which PP2A (protein phosphatase 2A) and PTEN (phosphatase and tensin homolog) play critical roles (Jafari et al., 2014; Marciniak et al., 2017; Li et al., 2020).

Results also showed that the basal level of the catalytic subunit of PP2A (PP2Ac) was increased in PTau cells but not in WTau cells relative to that in the control cells (Figures 6A,B). Demethylation at Leu309 (DM-PP2Ac) or phosphorylation at Tyr307 (p-PP2Ac) would reduce the activity of PP2Ac (Wang et al., 2015; Javadpour et al., 2019). The basal level of DM-PP2Ac tended to increase in WTau cells, but significantly increased in PTau cells (p < 0.05). Insulin treatment induced a slight increase of DM-PP2Ac in WTau cells however, it decreased in PTau cells (p < 0.05).

In addition, WTau or PTau transfection did not significantly affect the basal levels of pY307-PP2Ac. Insulin treatment only caused a significant increase in pY307-PP2Ac in the WTau cells (p < 0.05), but not in PTau or control cells. Based on these results, WTau and PTau overexpression may differentially affect PP2A, which may partially devote to the effects of Tau and its phosphorylation state on ISP.

There was no obvious change in PTEN in the WTau cells, but an increase in PTau cells (1.6 fold, p < 0.01), relative to that of control cells was observed. Treatment with insulin for 30 min resulted in a decrease tendency in PTEN in all these cells, among which the decline in PTau cells was significant (p < 0.05; Figures 6A,B). Besides, this study demonstrated that WTau could co-immunoprecipitate (CoIP) with PTEN, but very little with IRS1 and mTOR (Figure 6C). Furthermore, CoIP with the PTEN antibody also showed the binding between PTEN and WTau (Figure 6D). Thereafter, we evaluated the binding between PTEN and PTau, which showed that this binding was very weak (Figure 6D). Therefore, Tau phosphorylation may compromise the function of Tau in regulating PTEN function activity.

In contrast the effects of Tau on ISP on 293 T cells, the non-neuronal cells, as described in the above results, WTau or PTau overexpressing in the human neuroblastoma cell line SH-SY5Y cells also showed deregulation of ISP, with a slight increasing tendency of p-IRS the upstream factor of ISP, but a significant increase in p-S6 (p < 0.05) in basal conditions (Figures 7A,B). In addition, insulin-induced S6 phosphorylation was also suppressed in WTau or PTau overexpressed SH-SY5Y cells (Figures 7A,B), as seen in 293 T cells (Figure 1).

Previous studies have shown that the farnesyltransferase (FTase) inhibitor lonafarnib regulates ISP (Lee et al., 2004; Oh et al., 2006, 2008). Therefore, we evaluated whether it affected ISP in SH-SY5Y cells overexpressing WTau or PTau (Figure 7). SH-SY5Y cells overexpressing WTau were treated with various concentrations (1, 5, and 10 μM) of lonafarnib for 24 h (Figures 7C,D). The downregulation of Tau level and Tau phosphorylation at S396 occured after the lonafarnib treatment at 5 μM, and was significant with 10 μM lonafarnib (p < 0.01). However, phosphorylation at T181 was not affected. The signals of the Tau unphosphorylated epitope recognized by the Tau1 antibody tended to increase after treatment with 5 and 10 μM of lonafarnib. Thereafter, 1 μM lonafarnib was used to analyze ISP to minimize the effects of lonafarnib on Tau phosphorylation.

WTau cells treated with 1 μM lonafarnib for 24 h showed an increase in the phosphorylation of ISP elements, among which the increase in p-S6 was significant (p < 0.05; Figure 7I), but the level of p-Akt was decreased (p < 0.05; Figure 7E). In contrast, treatment of PTau cells with lonafarnib resulted in a decrease in ISP activity, appearing the downregulation of p-IRS (p < 0.05), p-Akt (p < 0.05; Figures 7E,F), p-mTOR (p < 0.01; Figures 7G,H), and the downstream factors of mTOR including decreased levels of p-S6 (p < 0.001), p-4EBP1-T70 (p < 0.01), and p-4EBP1-T37/46 (p < 0.01; Figures 7I,J). Therefore, the phosphorylation of Tau at the AT8 epitope could modulate the effect of lonafarnib on the ISP.

Rhes (Ras homolog enriched in striatum), one of FTase substrates, in known to regulate Akt (Bang et al., 2012; Harrison et al., 2013). We then investigated whether Tau phosphorylation could affect the possible action of lonafarnib on Rhes using methods of DARTS and CETSA (Lomenick et al., 2009; Jafari et al., 2014; Park et al., 2016).

In DARTS assay, SY5Y cells over expressing WTau, PTau, and uPTau (pseudo-non-phosphorylated at AT8 epitope) were treated with or without the lonafarnib, lysis of SY5Y cells were then digested with 0.25, 0.5 or 1 μg thermolysin/20 μg lysate (Figure 8A). The results showed that Rhes in PTau cells was more protective from thermolysin digestion than that in WTau cells. Lonafarnib treatment caused more intact or fragmented Rhes to be left after thermolysin (Figure 8A). This effect of lonafarnib was more obvious in PTau cells comparing that in WTau or uPTau cells. Therefore, lonafarnib affected the state of Rhes more efficiently in cells expressing PTau than that of WTau, which may be related to the different effects on the ISP for these cells.

With the CETSA assay (Figure 8B), lonafarnib treatment of control cells did not cause an increase in Rhes in the soluble fraction but even appeared to decrease at 55°C (Figure 8B). Moreover, the lonafarnib treatment also resulted in a reduction in Rhes in the soluble fraction in both WTau and AT8E-Tau SH-SY5Y cells (Figure 8C). Typically, small-molecule drugs increase the heat stability of their target proteins (Jafari et al., 2014; Friman, 2020). Therefore, these results do not provide sufficient evidence for the direct binding between lonafarnib and Rhes. However, the DARTS results suggest that lonafarnib could affect the biophysical chemical state of Rhes, in which Tau phosphorylation plays a regulatory role.