The nuclear receptor NR4A1 induces a form of cell death dependent on autophagy in mammalian cells

Abstract

The control of cell death is a biological process essential for proper development, and for preventing devastating pathologies like cancer and neurodegeneration. On the other hand, autophagy regulation is essential for protein and organelle degradation, and its dysfunction is associated with overlapping pathologies like cancer and neurodegeneration, but also for microbial infection and aging. In the present report we show that two evolutionarily unrelated receptors--Neurokinin 1 Receptor (NK(1)R,) a G-protein coupled receptor, and Insulin-like Growth Factor 1 Receptor (IGF1R), a tyrosine kinase receptor--both induce non-apoptotic cell death with autophagic features and requiring the activity of the autophagic core machinery proteins PI3K-III, Beclin-1 and Atg7. Remarkably, this form of cell death occurs in apoptosis-competent cells. The signal transduction pathways engaged by these receptors both converged on the activation of the nuclear receptor NR4A1, which has previously been shown to play a critical role in some paradigms of apoptosis and in NK(1)R-induced cell death. The activity of NR4A1 was necessary for IGF1R-induced cell death, as well as for a canonical model of cell death by autophagy induced by the presence of a pan-caspase inhibitor, suggesting that NR4A1 is a general modulator of this kind of cell death. During cell death by autophagy, NR4A1 was transcriptionally competent, even though a fraction of it was present in the cytoplasm. Interestingly, NR4A1 interacts with the tumor suppressor p53 but not with Beclin-1 complex. Therefore the mechanism to promote cell death by autophagy might involve regulation of gene expression, as well as protein interactions. Understanding the molecular basis of autophagy and cell death mediation by NR4A1, should provide novel insights and targets for therapeutic intervention.

Figure 1. Expression of NR4A1 dominant negative mutants inhibits IGF1R-induced death.

HEK293 cells were co-transfected with the indicated plasmids. Co-expression of dominant negative mutants NR4A1ΔDBD or NR4A1ΔN152 significantly reduced cell death. A, Percentage of cell death was determined by trypan blue exclusion 24 hr after transfection. Bars indicate standard error. ***, p<0.001, n = 4. B, Morphology of the rescued cells observed by phase contrast microscopy.

Figure 2. NR4A1-mediated vesicular cell death has autophagic features.

A, upper panels, presence of vesicles with cytoplasmic content (some marked by arrows) 12 h after cell death induction by either NK₁R/SP or IGF1R, observed by electron microscopy. Bottom panels, examples of structures resembling autophagosomes about to close (arrows). B, Localization of endogenous LC3, examined by immunoelectron microscopy using antibody against LC3, 12 h after cell death induction by either NK₁R/SP or IGF1R. Arrows indicate examples of LC3 associated with autophagosomes. For A and B, cells were transfected with empty vector (control), NK₁R and exposed to SP, or IGF1R, as indicated. C, D, GFP-LC3 is redistributed during NR4A1-mediated vesicular cell death. HEK293 cells were co-transfected with GFP-LC3 and either the empty vector (control), NK₁R and exposed to SP, or IGF1R. Examples of cells showing GFP-LC3 punctuated distribution are shown by confocal images in C. Percentage of cells with GFP-LC3 re-distribution, among GFP positive cells, are plotted in D. 4 fields per treatment were counted, in three independent experiments. Bars represent standard deviation. As a positive reference for autophagy, L929 cells were transfected with GFP-LC3 and exposed or not to caspase inhibitor zVAD.fmk. Nuclei were stained with Propidium Iodide (PI). Scale bar, 5 µm. E, LC3-II form accumulates during NR4A1-mediated vesicular cell death. HEK293 cells were transfected with the indicated expression vectors and total protein extracts were collected 24 hr after death induction to detect LC3 by Western blot. GAPDH or Tubulin were detected as loading reference. Autophagic flux did not seem to be impaired, since in the presence of Chloroquine (CQ) LC3-II accumulated furthermore, although there was not a detectable difference between CQ alone and CQ plus cell death inducers (even at lower exposure times). F, the autophagy inhibitor kinase mTOR is not activated during IGF1R-induced death. Protein extracts were collected at the indicated times (hr) after transfection with the empty vector pcDNA3 or IGF1R and the phosphorylation of the mTOR target p70 S6K was detected by Western blot. As a loading reference, tubulin was detected in the same blot. As can be observed, the basal level of phosphorylated p70 S6K did not change.

Figure 3. PI3K-III activity is necessary for NR4A1-mediated vesicular cell death.

HEK293 cells were transfected with either the empty vector pcDNA3, NK₁R and exposed or not to SP, or IGF1R as indicated. A, Pharmacological inhibition of PI3K with LY294002 reduces cell death. DMSO was added as vehicle control. B, Specific down regulation by RNAi of PI3K-III inhibits cell death. HEK293 cells were transfected with the indicated siRNAs. Cells death was determined by trypan blue exclusion 24 hr after induction. Bars represent standard error. ***, p<0.001, n = 4. C, Western blot to show the efficiency of RNAi. Luc refer to siRNAs targeting the luciferase gene, as irrelevant siRNA. GAPDH was detected as loading reference. D, induction of NR4A1 expression upon cell death activation is upstream of autophagy activation. HEK293 cells were transfected with NK₁R expression vector and exposed or not SP, and the expression of NR4A1 was monitored by Western blot in the presence or absence of the autophagy inhibitor LY294002, as indicated.

Figure 4. Beclin 1 and Atg7 are necessary for NR4A1-mediated vesicular cell death.

A, HEK293 cells were transfected with NK₁R and exposed or not to SP, or B, with the empty vector pcDNA3 or IGF1R. Expression of Beclin 1 or Atg7 was blocked by transfecting SMARTpool siRNAs for them in both models. As control, siRNA targeting a sequence not found in the human genome was transfected. Cells death was determined by trypan blue exclusion 24 hr after cell death induction. Bars represent standard error, (n = 4, * p<0.05). C, Western blot to show the efficiency of the respective RNAi. GAPDH was detected as loading reference. Atg7 signal was particularly low in these cells, therefore the exposure for this blot was longer and the contrast was increased.

Figure 5. Inhibition of autophagy by Spautin-1 prevents cell death and doubles the clonogenic growth.

HEK293 cells were transfected with either the empty vector pcDNA3, NK₁R and exposed or not to SP, or IGF1R as indicated. Either Spautin-1 or vehicle DMSO were added as indicated. Detached cells were collected from the media and cell death was determined by trypan blue exclusion 24 hr after cell death induction. Bars represent standard error, (n = 3, * p<0.05). B, bright field image of the cells that remained attached from the same cultures as in A, that were taken for clonogenic growth assays. C, Spautin-1 increases the clonogenic growth of cells expressing NK₁R and exposed to SP, or expressing IGF1R compared with vehicle. Equal number of cells from cultures shown in B was seeded, and the number of colonies containing more than 10 cells was scored (details in methods section). The plating efficiency was calculated for control cells and included in the calculation of the survival fraction. D, Spautin- 1 promotes the degradation of both Beclin 1 and PI3K-III. Western blot of total protein extracts taken at the indicated times.

Figure 6. Expression of NR4A1 dominant negative mutants inhibits autophagic cell death induced by caspase inhibition.

Mouse fibroblasts L929 were transfected with the indicated expression vectors, and treated or not with the pan-caspase inhibitor zVAD.fmk to induce autophagic cell death. A, cells exposed to zVAD.fmk were rounded up unless they expressed the indicated mutants. Scale bar, 50 µm. B, cell death determined by trypan blue exclusion 24 hr after treatment. There is a reduction of cell death by the expression of either NR4A1 dominant negative mutant ΔDBD or ΔN152, even though the efficiency of transfection was on average 30%. (n = 3; ***, p<0.001).

Figure 7. NR4A1 localized mainly in the nucleus during autophagic cell death, and is transcriptionaly competent.

HEK293 cells were transfected with NK₁R and exposed to SP for the indicated time. A, Immunofluorescence of NR4A1 (red) showed it mainly in nuclei, which were stained with Sytox Green. 20× magnification. B, Examples of cells where NR4A1 (red) was found by immunofluorescence in both the cytoplasm and the nucleus (DAPI) after 3 hr of exposure to SP. 40× magnification. C, Luciferase essay to quantify NR4A1 transcriptional activity driven by either NBRE or NuRE containing promotor. Folds of luciferase activity after 6 hr of SP addition with respect to control cells are plotted. Bar represent standard deviation (n = 4, ***, p<0.001). D, NR4A1 interacts with p53 but not with Beclin 1. HEK293 cells were transfected with His-tagged NR4A1 and NK₁R, and exposed or not to SP. NR4A1 was pulled down by NiNTA, and the presence of Beclin 1 or p53 complexed with NR4A1 was analized by Western blot.