FOXO transcription factors directly activate bim gene expression and promote apoptosis in sympathetic neurons - PubMed (original) (raw)

FOXO transcription factors directly activate bim gene expression and promote apoptosis in sympathetic neurons

Jonathan Gilley et al. J Cell Biol. 2003.

Abstract

Developing sympathetic neurons die by apoptosis when deprived of NGF. BIM, a BH3-only member of the BCL-2 family, is induced after NGF withdrawal in these cells and contributes to NGF withdrawal-induced death. Here, we have investigated the involvement of the Forkhead box, class O (FOXO) subfamily of Forkhead transcription factors in the regulation of BIM expression by NGF. We find that overexpression of FOXO transcription factors induces BIM expression and promotes death of sympathetic neurons in a BIM-dependent manner. In addition, we find that FKHRL1 (FOXO3a) directly activates the bim promoter via two conserved FOXO binding sites and that mutation of these sites abolishes bim promoter activation after NGF withdrawal. Finally, we show that FOXO activity contributes to the NGF deprivation-induced death of sympathetic neurons.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

PI3-K signaling regulates bim gene expression in sympathetic neurons. (A) PI3-K inhibition by LY294002 induces BIM protein levels. Neurons cultured with (+) or without (−) NGF, were treated with 50 μM LY294002 (+LY) or vehicle (DMSO) for 16 h and BIM expression assessed by immunoblotting. A representative blot is shown. Changes in BIM protein levels were determined by densitometry after normalization to tubulin to control for loading. Values represent ±SEM of three experiments. (B) PI3-K inhibition induces bim mRNA levels. Neurons were treated as in A and bim and neurofilament-M (NF-M) mRNA levels were assessed by RT-PCR. Representative gel images are shown. Changes in bim mRNA levels were determined as changes in band intensity on ethidium bromide–stained 2.5% agarose gels after normalization to NF-M to control for sample input. Values represent ±SEM of five experiments.

Figure 2.

Figure 2.

NGF regulates the phosphorylation and localization of FKHRL1 in sympathetic neurons. (A) NGF regulates FKHRL1 phosphorylation via PI3-K. Neurons were cultured with or without NGF (±NGF), or treated with 100 μM LY294002 (+LY) for 16 h. FKHRL1 expression and phosphorylation was then assessed by immunoblotting using anti-FKHRL1 and antiphospho (Thr32)-FKHRL1 antibodies. Representative blots are shown (asterisk indicates the phospho-specific antibody detects a nonspecific band of this size). (B) NGF regulates FKHRL1 localization. Neurons were injected with an expression construct encoding 0.1 mg/ml wild-type FKHRL1 together with guinea pig (gp) IgG as a marker and were cultured with or without NGF (±NGF) for 16 h. Neurons were then fixed and stained with Hoechst dye to visualize nuclei, and antibodies to gp IgG and the HA epitope to identify injected cells and detect HA-tagged FKHRL1. Representative images are shown. FKHRL1 localization was classified as cytoplasmic when nuclei showed no significant signal with anti-HA staining, and nuclear when the signal was greater in the nucleus than cytoplasm. In many neurons the staining pattern was ambiguous. Percentages representing ±SEM of three experiments are shown. Bar, 10 μm.

Figure 3.

Figure 3.

FOXO transcription factors induce bim gene expression and promote a BIM-dependent death of sympathetic neurons. (A) FKHR(ADA) expression increases endogenous BIM protein and bim mRNA levels. Neurons were infected with a control adenovirus (LacZ) encoding β-galactosidase (β-Gal) or an adenovirus encoding FKHR(ADA). BIM, FKHR, and β-Gal expression was assessed by immunoblotting (IB), and bim mRNA levels were assessed by RT-PCR. Tubulin and NF-M are loading and sample input controls. Images are representative of at least three experiments. (B) BIM expression can be selectively inhibited by bim antisense oligonucleotides. Naïve PC12 cells were cotransfected with bim antisense (AS) or missense (MS) oligonucleotides and an expression construct encoding GFP, and were sorted for GFP expression by FACS® after 24 h. BIM, BAX, BCL-2, BCL-XL, and procaspase-3 protein levels in GFP-positive cells were then assessed by immunoblotting. Tubulin represents a loading control. Blots are representative of three experiments. (C) Neurons were injected with 0.2 mg/ml of an FKHRL1(A3) expression construct or empty vector together with 0.03 mg/ml of bim antisense (AS) or missense (MS) oligonucleotide mixtures, and Texas red dextran as a marker. Cells were left to recover for 4–6 h (time 0) and survival of the injected neurons were determined as the percentage of viable cells at each time point relative to time 0. Values represent ± SEM of four experiments. FKHRL1(A3) expression significantly increased death of neurons injected with missense oligonucleotides (P ≤ 0.05, t tests, FKHRL1(A3) + MS vs. vector + MS), and this was blocked in neurons injected with the bim antisense oligonucleotides (FKHRL1(A3) + AS). The antisense oligonucleotides had no significant independent effect upon neuronal survival (vector + AS).

Figure 4.

Figure 4.

The bim promoter is activated by NGF withdrawal, PI3-K inhibition, and FKHRL1(A3) expression in sympathetic neurons. (A) Structure of the 5′ end of the rat bim gene and the _bim_-LUC reporter. Exons are shown as boxes; clear regions represent 5′ untranslated region and shaded regions coding sequence. An arrow indicates the major transcription start site (+1), as determined by 5′ RACE. The _bim_-LUC reporter consists of the 5.2-kb region 5′ to the bim initiator codon in pGL3-Basic. (B) Luciferase immunostaining reveals that the bim promoter is activated after NGF withdrawal in neurons injected with the _bim_-LUC reporter. Neurons were injected with 0.01 mg/ml of the _bim_-LUC reporter and gp IgG as a marker and were cultured with or without NGF (±NGF) for 20–24 h. Cells were then fixed and stained with Hoechst dye to visualize nuclei, and antibodies to gp IgG and luciferase to identify injected cells and to assess luciferase expression. Typical images are shown. Bar, 10 μm. Neurons were scored as expressing luciferase when staining was significantly greater than background staining in uninjected neurons. The data represent ±SEM of six experiments (*, P < 0.05, t test, +NGF vs. −NGF). Bar, 10 μm. (C) Assaying for luciferase activity confirms bim promoter activation after NGF withdrawal. Neurons were injected with 0.01 mg/ml of the _bim_-LUC reporter and 0.005 mg/ml of the Renilla luciferase expression construct pRL-TK, to control for injection, and were cultured with or without NGF (±NGF) for 18–20 h after which luciferase assays were performed. The relative fold change in Firefly activity was determined after normalization to Renilla activity. The data represent ±SEM of four experiments. (*, P < 0.002, t test, +NGF vs. −NGF). (D) The bim promoter is activated by LY294002. Neurons injected as in B were treated with 50 μM LY294002 or DMSO (−) for 20–24 h in the presence of NGF and the percentage of injected cells expressing luciferase determined as in (B). The data represent ±SEM of five experiments (*, P < 0.001, t test, LY294002 vs. DMSO). (E) The bim promoter is activated by FKHRL1(A3). Neurons were injected with 0.01 mg/ml of the _bim_-LUC reporter, gp IgG, and either 0.01 or 0.1 mg/ml of wild-type (wt) FKHRL1 or FKHRL1(A3) expression construct or equimolar amounts of empty vector (−) and were maintained in the presence of NGF. After 20–24 h, the percentage of injected cells expressing luciferase was determined as in B. The data represent ±SEM of at least three experiments (*, P < 0.02 and **, P < 0.005, t tests, FKHRL1(A3) vs. vector).

Figure 5.

Figure 5.

FKHRL1 binds to FOXO binding sites located close to the bim promoter. (A) The consensus binding site for the FOXO subfamily of Forkhead transcription factors and a known FOXO binding site in the Fas ligand promoter (FasL) are compared with the conserved bim1 and bim2 sites located close to the bim promoter. Deviation from the consensus at positions 1, 6, and 8 is not uncommon. Overlapping FasL FOXO sites are underlined. Mutated versions of the bim1 and bim2 sites (mut1 and mut2) are also shown (mutations are boxed). The bim2 site spans the bim exon 1/intron 1 splice site so the mut2 mutations were designed to minimize the chance of disrupting splicing (mut2 fits the GTRAGT donor splice consensus). (B) Purified GST-FKH protein was tested for binding to double-stranded 32P-labeled oligonucleotides containing the FasL site and the wild-type and mutant bim sites listed in A by electrophoretic mobility shift assay. A representative gel is shown. An arrow indicates the mobility of specific complexes between GST-FKH and the labeled oligonucleotides. GST-c-Jun was incubated with the FasL oligonucleotide (lane 6) as a negative control. GST-FKH binding to the FasL oligonucleotide was competed with a 500-fold excess (500×) of unlabeled FasL, bim1, or bim2 oligonucleotide (lanes 3–5).

Figure 6.

Figure 6.

FOXO binding sites are required for bim promoter activation by FKHRL1(A3) and NGF withdrawal. (A) Structure of _bim_-LUC reporters containing mutations in the bim1 and bim2 FOXO sites. Reporters are represented as in Fig. 4 A. Wild-type sites are represented by shaded circles and mutated sites by crosses. The location of primers used for RT-PCR (D) is also shown. (B) Mutation of the bim1 and bim2 sites substantially reduces bim promoter activation by FKHRL1(A3). Neuronal PC12 cells were transiently transfected with the indicated reporters together with varying concentrations of the FKHRL1(A3) expression construct pECE-FKHRL1(A3) or empty vector and the Renilla luciferase expression construct pRL-TK to control for transfection efficiency. Luciferase activity was determined after 24 h and Firefly activity was normalized to Renilla activity. The fold change in normalized Firefly luciferase activity of cells transfected with pECE-FKHRL1(A3) relative to empty vector is plotted. The data represent ±SEM of at least three experiments. pGL3-Promoter (pGL3-P), an SV40-LUC reporter, is included for reference. Activation of _bim_-LUC(dm) was significantly reduced at both concentrations compared to _bim_-LUC (P < 0.02 at 0.33 μg and P < 0.002 at 1 μg, t tests). (C) The bim1 and bim2 mutations abolish bim promoter activation after NGF withdrawal. Sympathetic neurons were injected with 0.01 mg/ml of the _bim_-LUC or _bim_-LUC(dm) reporters together with gp IgG and were cultured with (+) or without (−) NGF for 20–22 h. Reporter activities were then determined by luciferase immunostaining as in Fig. 4 B. The data represent the mean of six experiments ± SEM (*, P < 0.002, t test, _bim_-LUC +NGF vs. −NGF). (D) Sympathetic neurons were injected with 0.02 mg/ml of the _bim_-LUC or _bim_-LUC(dm) reporters together with 0.01 mg/ml of pRL-TK to act as an injection control. Cells were then cultured with (+) or without (−) NGF for 16 h and _bim_-LUC and Renilla luciferase (Renilla) mRNA levels were determined by RT-PCR. A _bim_-specific and _luciferase_-specific primer spanning bim intron 1 (A) were used for _bim_-LUC amplification to prevent amplification from endogenous bim transcripts and to confirm that mutation of the bim2 site did not affect splicing. Relative changes in bim-LUC levels after NGF withdrawal were then calculated as changes in band intensities on ethidium-stained 2.5% agarose gels after normalization to Renilla levels. The data represent the mean of three experiments ± SEM (*, P < 0.025, t test, wild-type _bim_-LUC mRNA levels +NGF vs. −NGF). Representative gel images are also shown.

Figure 7.

Figure 7.

Inhibition of FOXO activity delays the NGF withdrawal–induced death of sympathetic neurons. (A) Structure of FKHRL1 and FKH(DBD). Key regulatory phosphorylation sites are shown for FKHRL1. FKH(DBD) is the DNA-binding domain (DBD) of FKHRL1 (amino acids 141–268) with the S253A mutation found in FKHRL1(A3) and an NH2-terminal FLAG epitope tag. (B) FKH(DBD) localizes to the nucleus. Neurons were injected with 0.05 mg/ml of the FKH(DBD) expression construct and gp IgG. After 24 h, neurons were fixed and stained with Hoechst dye, to visualize nuclei, and antibodies to gp IgG and the FLAG epitope to identify injected cells and detect FLAG-tagged FKH(DBD). Representatives images are shown. Bar, 10 μm. (C) FKH(DBD) is a specific inhibitor of FOXO transcriptional activity. Neurons were injected with either 0.01 mg/ml of the _bim_-LUC reporter or 0.001 mg/ml of the _c-jun_-LUC reporter together with expression vectors (or the corresponding empty vectors) for 0.01 mg/ml of FKHRL1(A3), 0.1 mg/ml of MEKK1, and 0.05 mg/ml of FKH(DBD) as indicated, and gp IgG as a marker. 20–24 h later the percentage of cells expressing luciferase was determined (as in Fig. 1). The averages of at least three experiments ±SEM are shown. FKHRL1(A3) and MEKK1 significantly increased luciferase expression in cells injected with the _bim_-LUC and _c-jun_-LUC reporters respectively (*, P < 0.02, **, P < 0.05, t test), but FKH(DBD) only inhibited activation of the _bim_-LUC reporter by FKHRL1(A3) (§ indicates P < 0.005, t test). (D) Sympathetic neurons were injected with expression constructs encoding FKH(DBD) or FLAGΔ169 (a dominant-negative c-Jun mutant), or empty vector (all at 0.05 mg/ml) together with Texas red dextran as a marker. Cells were allowed to recover overnight and were then deprived of NGF. The number of viable injected neurons was determined at 0, 24, 48, and 72 h after NGF withdrawal. Survival is expressed as a percentage of the number of viable injected neurons at time 0. Experiments were performed in a blinded manner and the average of three experiments ±SEM is shown. Survival was significantly increased by FKH(DBD) compared with empty vector (P < 0.005 at 24 h, P < 0.02 at 48 h, P < 0.01 at 72 h, t tests).

Figure 8.

Figure 8.

Regulation of bim gene expression by NGF in sympathetic neurons. FOXO transcriptional activity is inhibited by NGF via PI3-K signaling and Akt/SGK-mediated phosphorylation. NGF withdrawal and LY294002 inhibit PI3-K signaling, resulting in FOXO activation and subsequent activation of the bim promoter via the bim1 and bim2 FOXO binding sites. The JNK pathway is repressed in the presence of NGF. PI3-K signaling could account for some of this repression via Akt-mediated inhibition of MLK activity but PI3-K–independent pathways may also be involved. NGF withdrawal and LY294002 activate the JNK pathway via mixed lineage kinase 3 (MLK3), resulting in c-Jun activation. Dominant-negative c-Jun (DN c-Jun) can partially inhibit BIM induction after NGF withdrawal, which suggests that c-Jun/AP-1 may also directly regulate the bim promoter. Furthermore, our results do not rule out the possibility that other transcription factors/signaling pathways might also be involved in regulation of bim promoter activity or that NGF might regulate BIM expression post transcriptionally.

References

    1. Barthwal, M.K., P. Sathyanarayana, C.N. Kundu, B. Rana, A. Pradeep, C. Sharma, J.R. Woodgett, and A. Rana. 2003. Negative regulation of mixed lineage kinase 3 by protein kinase B/AKT leads to cell survival. J. Biol. Chem. 278:3897–3902. - PubMed
    1. Brunet, A., A. Bonni, M.J. Zigmond, M.Z. Lin, P. Juo, L.S. Hu, M.J. Anderson, K.C. Arden, J. Blenis, and M.E. Greenberg. 1999. Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell. 96:857–868. - PubMed
    1. Burgering, B.M., and G.J. Kops. 2002. Cell cycle and death control: long live Forkheads. Trends Biochem. Sci. 27:352–360. - PubMed
    1. Deshmukh, M., and E.M. Johnson, Jr. 1997. Programmed cell death in neurons: focus on the pathway of nerve growth factor deprivation-induced death of sympathetic neurons. Mol. Pharmacol. 51:897–906. - PubMed
    1. Dijkers, P.F., R.H. Medema, J.W. Lammers, L. Koenderman, and P.J. Coffer. 2000. Expression of the pro-apoptotic Bcl-2 family member Bim is regulated by the forkhead transcription factor FKHR-L1. Curr. Biol. 10:1201–1204. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources