Evidence for involvement of ERK, PI3K, and RSK in induction of Bcl-2 by valproate

Abstract

Valproate, an anticonvulsant and mood stabilizer, up-regulates Bcl-2, a neurotrophic/neuroprotective protein. In this study, we investigated the molecular mechanism through which Bcl-2 is up-regulated by valproate using cultured human neuron-like cells. Valproate, within therapeutically relevant ranges, induced time- and concentration-dependent up-regulations of both Bcl-2 messenger RNA and protein implicating an underlying gene transcriptional-mediated mechanism. Bcl-2 up-regulations were associated with ERK1/2 and PI3K pathway activations and elevated levels of activated phospho-RSK and phospho-CREB, convergent targets of the ERK1/2 and PI3K pathways. Valproate increased transcriptional activity of a human bcl-2 promoter-reporter gene construct. This effect was attenuated, but not blocked, by mutation of a CREB DNA binding site, a CRE site in the human bcl-2 promoter sequence. ERK and/or PI3K pathway inhibitors and RSK1 small hairpin RNA knockdown reduced, but did not abolish, baseline and valproate-induced promoter activities and lowered Bcl-2 protein levels. These data collectively suggest that valproate induces Bcl-2 regulation partially through activations of the ERK and PI3K cascades and their convergent kinase, RSK, although other unknown mechanism(s) are likely involved. Given the known roles of Bcl-2 in the central nervous system, the current findings offer a partial yet complex molecular mechanistic explanation for the known neurobiological effects of valproate including neurite growth, neuronal survival, and neurogenesis.

Figure 1

VPA increases bcl-2 protein levels in SH-SY5Y cells in a time- and concentration-dependent manner. SH-SY5Y cells were grown in DMEM medium supplemented with 10% FBS, antibiotics, and sodium pyridine to 80–90% confluence and treated thereafter for 8, 24, 48, 72, or 96 h with VPA (0.8 mM; VPA8, 24, 48, 72, or 96, respectively) or with vehicle control (C8, 24, 48, 72, or 96, respectively). A, C Cells were harvested and processed for Western blot analysis. B, D Densitometric results for Western blot analysis are displayed as percent of control, which was calculated as a mean of all of the controls. Data in the bar graph are means±SE from four sets of replicates. ***p<0.001; **p<0.01

Figure 2

VPA increases bcl-2 mRNA in a time-dependent manner. RT-PCR was utilized to quantitate bcl-2 mRNA levels from SH-SY5Y cells treated with 0.8 mM VPA for 1, 2, or 3 days (VPA1, 2, or 3, respectively) or with vehicle control (C1, 2, 3, respectively). Total cellular RNA was isolated from VPA-treated cells using Trizol reagent (Invitrogen) and purified using RNeasy mini kit (Qiagen). The RNA concentrations were determined in duplicates spectrophotometrically, and the quality of RNA was evaluated by visual inspection after electrophoresis of an aliquot of the RNA samples. Results are displayed as fold differences from respective controls for that day. Data in the bar graph are means±SE from three or more sets of replicates. ***p<0.001; **p<0.01, *p<0.05

Figure 3

VPA enhances phospho-ERK1/2 and phospho-Akt levels in SH-SY5Y cells in a concentration-dependent manner. SH-SY5Y cells were grown as previously described and treated with VPA (0.8 mM) for 48 h, collected and processed for Western blot analysis, and probed with either phospho-ERK1/2 (Thr202/Tyr204) or phospho-Akt (Ser473) antibody (A, C). Densitometric results for Western blot analysis are displayed as percent of control, which was calculated as a mean of all of the controls (B, D). The phospho-Akt/Akt ratios were also increased by VPA treatment in a concentration-dependent manner (E). Data in the bar graph are means±SE from four sets of replicates

Figure 4

VPA enhances phospho-RSK1 and phospho-CREB levels in SH-SY5Y cells in a dose-dependent manner. SH-SY5Y c1ells were grown as previously described and treated with VPA (0.8 mM) for 48 h, collected and processed for Western blot analysis, and probed with either phospho-RSK1 (Thr359/Ser363) or phospho-CREB (Ser133) antibody (A, C). Densitometric results for Western blot analysis are displayed as percent of control, which was calculated as a mean of all of the controls (B, D). Data in the bar graph are means±SE from four sets of replicates

Figure 5

VPA enhances activity of a bcl-2 promoter construct in SH-SY5Y cells in a dose-dependent manner and is partially dependent on a viable CRE site within the bcl-2 promoter. A luciferase reporter gene construct driven by a promoter containing a human bcl-2 gene promoter fragment was transfected into SH-SY5Y cells which were treated with various doses of VPA for 48 h to produce a dose–response curve for the promoter activity. A luciferase assay system was used to assess the promoter activity. Additionally, a single base pair mutation was made in the CRE site of the human bcl-2 promoter fragment and transfected into SH-SY5Y cells which were treated with the same doses of VPA for 48 h to produce a dose–response curve for the promoter activity. Data in the bar graph are means±SE from three or more sets of replicates. A two-way ANOVA of the data indicates that there is a highly significant group effect (WT vs mutant, F=57, p<0.0001), a highly significant concentration effect (F=117.3, p< 0.0001), and a highly significant interaction effect between group and concentration effects (F=14.88, p<0.0001)

Figure 6

Chemical inhibition of either the ERK or PI3K pathway drastically reduces activity of a human bcl-2 promoter–reporter gene construct and inhibition of both pathways virtually abolishes the activity. A SH-SY5Y cells were grown as before, transfected with the promoter construct, and treated with PD98059 (50 μM) or LY294002 (10 μM) or both for 48 h and subsequently collected and processed for luciferase assay. B The same experiment was set up again with VPA (0.8 mM final concentration) added to half of the cells from each treatment condition 2 h after inhibitor treatment. Cells were collected and a luciferase assay system was again used to assess the promoter activity. Data in the bar graph are means±SE from three or more sets of replicates. ***p<0.001, **p<0.01 from DMSO controls; ###p< 0.001, ##p<0.01 from DMSO/VPA

Figure 7

RSK1 knockdown attenuates activity of a human bcl-2 promoter–reporter gene construct. Several RSK1-specific shRNAs, designed using online tools, as well as a LacZ control that coded for β-galactosidase, were cloned into a U6 promoter-driven vector and tested for RSK1 knockdown efficiency. Additionally, a 1-bp mutation of the most efficient shRNA (sequence 23) was designed (23mut) to rescue or reverse the knockdown and to show that the knockdown was not due to off-target effects. A, B SH-SY5Y cells were grown as before and transfected with the shRNA constructs for 2 days before cells were collected and processed for RSK1-specific immunoblotting and GAPDH as an off-target control. C Sequence 23 RSK1-specific shRNA and the LacZ control shRNAs were transfected into SH-SY5Y cells as above. Cells were collected 2 days later and processed for bcl-2 promoter activity. RSK1-specific shRNA significantly reduced promoter activity (D). The previous experiment was set up again with VPA (0.8 mM final concentration) added 2 h after the shRNA transfections. Cells were collected and processed as before. RSK1-specific shRNA significantly reduced VPA-induced enhancement of promoter activity compared to that of the LacZ shRNA. Data in the bar graph are means±SE from three or more sets of replicates. ***p< 0.001, **p<0.01 from control; ###p<0.001 from Control/VPA

Figure 8

Chemical inhibition of the ERK or PI3K pathway or both or RSK1 knockdown reduces Bcl-2 protein levels. SH-SY5Y cells were grown as before and treated with PD98059 (50 μM) or LY294002 (10 μM) or DMSO vehicle control with or without VPA (0.8 mM) for 3 days. A Cells were collected and processed for immunoblot analysis with anti-Bcl-2 or GAPDH load control. Densitometric measurements were obtained as before and represent means±SE of four independent experiments. C and D SH-SY5Y cells were grown to confluence and transfected with the (sequence 23) RSK1-specific shRNA construct for 2 days. Cells were collected and harvested for immunoblot analysis with anti-Bcl-2. Data in the bar graph are means±SE from three or more sets of replicates. **p<0.01, *p<0.05, ##p<0.01 compared to VPA treatment alone

Figure 9

VPA regulates Bcl-2 in part by activating both the ERK and PI3K pathways. Lines with arrowheads indicate activation of designated target. Lines with filled-in circles indicate inhibition of designated target