IMR32 cells (ATCC, Manassas, VA, USA) were grown in Gibco® modified Eagle’s medium (MEM; Life Technologies, Burlington, ON, CAN) containing 10% fetal bovine serum (Hyclone, Logan, UT, USA), sodium pyruvate (1 mM) and 50 µg/mL gentamycin (Life Technologies). For protein analysis IMR32 cells were plated at a density of one million cells per 100 mm cellbind® plate (Corning, Lowell, MA, USA) the day before treatment. To establish a time course for NGF-mediated tau expression, IMR32 cells were left untreated (control) or treated with 50 ng/mL NGF (2.5 S Harlan Bioproducts, Indianapolis, IN, USA) and harvested at 24 h intervals for a maximum of 96 h. To evaluate the mechanisms involved in the NGF-mediated activation of tau, specific NGF signaling pathway or NOS inhibitors were added to the culture media 1 h prior to the addition of NGF (50 ng/mL). The inhibitor doses included 10 µM LY294002 (Calbiochem, San Diego, CA, USA; inhibits PI-3 kinase-mediated activation of Akt), 50 µM U0126 (Calbiochem; inhibits MEK1/2), the PLC-PKC inhibitors U-73122 and bisidolylmaleimide 1 (BIS-1; Calbiochem), 10 µM and 5 µM, respectively, 20 mM N^w-nitro-Larginine methylester (L-NAME; Sigma Aldrich, St. Louis, MO, USA; non-selective NOS inhibitor) and 2 mM s-methylisothiourea (S-MIU; Sigma; iNOS selective inhibitor). These inhibitors and doses were selected based on our previous studies on NGF-mediated signaling and gene expression in PC12 cells [25] [28] [30] -[33] . Media and treatments were replaced every 48 h.

Cells were rinsed with ice cold phosphate-buffered saline (PBS) followed by lysis in 250 µL of radioimmunoprecipitation assay (RIPA) buffer (final concentration: 50 mM Tris, 150 mM NaCl, 1% NP-40, 0.25% sodium deoxycholate, 0.5% sodium dodecylsulfate (SDS), 1 mM each of EDTA, sodium fluoride and sodium orthovanadate and one complete mini EDTA free tablet (Roche Diagnostics, Laval, QC, CAN)/10 mL of buffer). Samples were sonicated and the protein content of the supernatant determined by the method of Bradford [36] . Duplicate aliquots of some samples were dephosphorylated using 10 µL alkaline phosphatase (Roche Diagnostics) and 10 µL of the accompanying 10× dephosphorylation buffer (0.5 M Tris-HCl, 1 mM EDTA, pH 8.5), incubated at 37˚C for 60 min. Protein samples (100 µg) were boiled in SDS sample buffer with dithiothreitol (Cell Signaling Technology, Danvers, MA, USA) and a human tau protein ladder (Sigma Aldrich) was run alongside protein samples to identify tau isoform expression in the IMR32 cell line. Protein samples and standards were loaded onto 10% SDS/polyacrylamide separating gels and after electrophoresis, proteins were transferred onto nitrocellulose membranes using a semi-dry unit (Bio-Rad Laboratories, Mississauga, ON, CAN) with transfer buffer (48 mM Tris, 39 mM Glycine, 0.04% SDS and 20% methanol). After transfer, membranes were rinsed in tris-buffered saline (TBS; Tris 4 mM, NaCl 27.4 mM, pH 7.6) containing 0.1% Tween-20 (TBS-T) for 5 min then blocked for 1 h in TBS-T containing 5% skim milk and incubated overnight with mouse monoclonal tau antibody (Chemicon MAB3420; EMD Millipore, Billerica, MA, USA), 1:2500 in TBS-T with 5% skim milk. Blots were rinsed and incubated for 1 h with sheep anti-mouse IgG-horseradish peroxidase-conjugated secondary antibody (NA931; Amersham/GE Health Care, Piscataway, NJ, USA) 1:2000 in TBS-T with 5% skim milk followed by enhanced chemiluminescence (ECL; GE Healthcare). Protein bands were visualized using the STORM 860 scanner (Molecular Dynamics, subsidiary of Amersham) and densitometry was performed using Imagequant software (Molecular Dynamics). Following the detection of tau, blots were stripped and reprobed with mouse monoclonal α-tubulin antibody (1:50,000) in TBS-T containing 5% skim milk as described above.

IMR32 cells were plated on a 100 mm cell culture dish and grown to 90% confluency. Media was aspirated and cells were rinsed with cold PBS prior to isolating DNA with the DNeasy tissue kit (Qiagen, Mississauga, ON, CAN) according to the manufacturer’s directions. A 550 bp fragment of the tau promoter was amplified by polymerase chain reaction (PCR) using forward and reverse primers that included a Kpn 1 and Nhe 1 restriction site, respectively. Primer sequences were forward: 5’-CAGTGGTACCGACAATCTCTAGCTGG CCCTTAAAACATTC-3’ and reverse: 5’-CATAGCTAG CCTCCTGTAGTTGGAGTCTTTGTGTCG-3’. Purified PCR product (20 µL) and 6 µg of purified PGL3 basic vector (Promega Corporation, Madison, WI, USA) were digested with the restriction enzymes Kpn 1 and Nhe 1 (2 µL each; New England BioLabs, Ipswich, MA, USA) in NEBuffer 4 with BSA at 37˚C for 1 h. Following 20 min of heat inactivation at 65˚C samples were separated on a 1.5% agarose gel. The digested vector and PCR products were extracted from the gel and purified using the montage gel purification kit (EMD Millipore) and incubated with DNA ligase (Roche Diagnostics) at room temperature overnight. DNA from colonies of transformed DH5α cells (Life Technologies) was purified by alkaline lysis and the identity of the isolated DNA verified by restriction digest followed by DNA purification and sequencing (University of Guelph, Guelph Molecular Supercenter). DNA for transfections was purified using the Invitrogen Purelink HiPure Plasmid Filter purification kit (Life Technologies) according to the manufacturer’s directions.

One day prior to transfection, IMR32 cells were plated onto 12 well dishes at a density of 300,000 cells per well in 2 mL media. Since the transfection efficiency and viability of the IMR32 cells was not as high as our previous experiments with PC12 cells, similar experiments were also conducted with PC12 cells. These cells were plated onto 12 well dishes at a density of 200,000 cells per well. On the day of transfection, media in each well was replaced with 1 mL antibiotic free media. Transfection solution containing 100 µL Opti-MEM medium (Life Technologies), 2 µL transfectin (Bio-Rad Laboratories), 1 µg tau promoter DNA and 0.5 µg β-galactosidase control vector DNA per well was prepared and incubated at room temperature for 30 min prior to its addition to the cell culture media. After 4 to 6 h of exposure to the transfection solution, the media was replaced with 2 mL of regular cell culture media.

The following day, cells were left untreated (control) or treated with NGF (50 ng/mL) or one of the inhibitors either alone or in combination with NGF. All treatments were administered to triplicate wells and inhibitors were added 1 h prior to NGF. Twenty-four h after NGF treatment cells were harvested in 200 µL 1× passive lysis buffer (Promega). Duplicate supernatant aliquots (20 µL for both luciferase and β-galactosidase analysis) were added to a 96-well plate. Prior to determining β-galactosidase activity, 25 µL Galacton, diluted 1:100 with reaction buffer diluent (Aurora Galacto-light kit, MP Biomedicals, Irvin, CA, USA) was added to each well and incubated for 30 to 60 min. The 96-well plates were read using a FLUOstar optima plate reader luminometer (BMG; Fisher Scientific, Ottawa, ON, CAN) which added 50 µL of the appropriate Accelerator reagent to each well immediately prior to luminescence quantification. For luciferase (Promega) the signal was read for 10 s following a 2 s delay and for β-galactosidase (Galacto-light) the signal was read for 1 s.

Following densitometric analysis with Image Quant, tau isoform protein levels were expressed relative to α-tubulin levels from the same sample. Luciferase values were normalized to β-galactosidase values from the same sample to control for transfection efficiency. All data are presented as the mean ± SEM and Graphpad Instat was used to analyze raw data. For most studies, sample size consisted of a minimum of 5 experiments. Statistical analysis was carried out by one-way analysis of variance (ANOVA) followed with Tukey-Kramer multiple comparisons post-hoc test. If data failed Bartlett’s test for comparable standard deviations or did not follow Gaussian distributions according to testing using the Kologorov and Smirnov method then data were assessed with the non-parametric ANOVA equivalent, the Kruskal-Wallace test, followed by Dunn’s multiple comparisons test.

A representative western blot depicting tau isoform expression in untreated (control) and NGF treated IMR32 cell lysates is shown in Figure 1. To identify the various tau isofoms, IMR32 cell extracts and dephosphorylated cell extract were compared to a tau protein ladder containing the six human tau isoforms (Sigma; data not shown). The smallest molecular weight isoform (0N3R) is the lowest band of the tau ladder and corresponds to the lowest molecular weight band detected in the upper panel of Figure 1. The other isoforms, in order from smallest to largest molecular weight, 0N4R, 1N3R, 1N4R, 2N3R, 2N4R appear above this band on the tau protein ladder and most were detected in the IMR32 cell lysates. As seen in Figure 1, the density of the tau bands detected in IMR32 cell lysates appeared to be greater for the 0N4R and 1N3R isoforms as compared to the other isoforms. The 0N3R and 1N4R isoforms were also detected consistently and could be reliably quantified. The 2N3R isoform was only detected in extracts obtained from IMR32 cells treated with NGF, indicating the expression of this isoform increased following NGF treatment, however this isoform was not detected consistently in all samples or on all blots and was therefore not quantified. The 2N4R isoform was rarely detected in cell extracts and was only observed on 50% of the blots analyzed. This isoform was also not quantifiable. Since dephosphorylated samples did not exhibit a different isoform band pattern from their unphoshorylated counterparts (not shown), all analysis of tau protein was completed using cell extract samples that were not dephosphorylated.

Densitometric analysis of tau protein revealed differential temporal regulation of tau isoforms following NGF treatment (Figure 2). Significant differences in the level of the 0N3R isoform were detected following NGF treatment (F(4, 55) = 3.979, p = 0.0066), with a significant increase observed following 72 h of NGF exposure

as compared to control and all other treatment time points (Figure 2(a); ^*p < 0.05 vs. control and all other time points). The 0N4R isoform also exhibited significant differences in expression levels (F(4, 75) = 2.533, p = 0.0472), with increased levels detected at the 72 h treatment time compared to control (Figure 2(b)). Unlike the 0N3R isoform, the relative level of 0N4R detected following 72 h of NGF treatment was not significantly different from other time points suggesting that the level of this isoform at other treatment times falls somewhere between that observed in control and 72 h NGF treated lysates. The expression of the 1N3R isoform was also increased following NGF treatment (KW = 11.179, p = 0.0246). Although expression appeared increased at 72 h (Figure 2(c)) this difference was not quite statistically significant (p = 0.06) and the only statistically significant difference was in the 48 h NGF exposed cells, where the levels of this isoform were greater than control (^*p < 0.05). The 1N4R tau isoform was also elevated following NGF treatment (F(4, 35) = 4.456, p = 0.0051) with significantly greater levels detected in cells treat with NGF for 96 h as compared to control (^**p < 0.01; Figure 2(d)).

To determine whether NGF affects the transcription of tau, a 550 bp fragment of the human tau promoter was cloned into the pGL3 basic luciferase reporter vector. Tau promoter activation in IMR32 and PC12 cells was determined by luciferase assay where luciferase activity was normalized to beta-galactosidase activity from the same sample.

In IMR32 cells (Figure 3(a)) NGF treatment did not increase luciferase activity in pGL3 basic-transfected cells relative to control cells but did increase luciferase activity in tau promoter-pGL3 basic transfected cells in comparison to control (t = 2.152, df = 14, p = 0.0494), indicating activation of the tau promoter by NGF. Since luciferase and beta-galactosidase activity was much lower than what we had detected with previous transfections using PC12 cells [28] [33] [37] tau promoter activation was also assessed in PC12 cells. In the PC12 cell line, both luciferase and β-galactosidase activities were much higher than those detected in IMR32 cells and a more robust NGF effect was observed (Figure 3(b)). One-way ANOVA revealed significant differences between treatment groups (F(5, 32) = 93.586, ^***p < 0.001) and the Tukey-Kramer multiple comparisons test confirmed that NGF treatment significantly increased luciferase activity in cells transfected with the pGL3 basic vector alone (^*p < 0.05) and in those transfected with the tau promoter construct (^***p < 0.001). The NGF-mediated increase in luciferase activity in tau promoter-transfected cells was significantly higher than the pGL3-basic NGF response, indicating that NGF activates this fragment of the tau promoter. Based on these data, the PC12 cell line was used to assess the effects of NGF-TrkA signaling pathway and NOS inhibitors on the NGF-induced increase in tau promoter activation.

First, the effects of 50 µM UO126, which inhibits the MEK-MAP kinase pathway and 10 µM LY2940002, which inhibits the PI3K-Akt pathway were assessed (Figure 4(a)). There were no significant differences between the untreated control and inhibitor alone controls. NGF treatment increased promoter activation in both UO126 (^*p < 0.001) and LY2940002 (^***p < 0.001) pretreated cells in comparison to inhibitor alone controls. The NGF effect in UO126 pretreated cells was lower than that observed following treatment with NGF either alone (+p < 0.05) or pretreated with LY2940002 (p < 0.01) suggesting that the MAP kinase, but not Akt, signaling pathway is involved in the regulation of tau transcription.

The phospholipase C-PKC inhibitors U-73122 (10 µM) and BIS-1 (5 µM) were then used to investigate the involvement of the PKC signaling pathway. The effect of these inhibitors on NGF-induced tau promoter activation is shown in Figure 4(b). Statistical analysis revealed significant differences between treatments (F(5, 26) = 49.885, p < 0.0001). With both inhibitors, basal luciferase activity was not changed relative to control activity and for both control and inhibitor-pretreated cells, NGF resulted in a significant increase in relative luciferase activity (^***p < 0.001). In comparison to treatment with NGF alone, in BIS-1 pretreated cells the NGF effect was significantly lower (+p < 0.05).

The effect of NOS inhibitors on human tau promoter activation is shown in Figure 4(c). Statistical analysis revealed significant differences between treatments (F(5, 28) = 96.971, p < 0.0001). In comparison to untreated cells, basal luciferase activity was not different in cells treated with 2 mM S-MIU or 20 mM L-NAME. In the presence or absence of NOS inhibitor, NGF treatment significantly increased luciferase activity relative to control (^***p < 0.001), and there was no difference in the NGF response between control and NOS inhibitor pretreated samples.