IEX-1L, an Apoptosis Inhibitor Involved in NF-κB-Mediated Cell Survival

See allHide authors and affiliations

Science  14 Aug 1998:
Vol. 281, Issue 5379, pp. 998-1001
DOI: 10.1126/science.281.5379.998


Transcription factors of the nuclear factor–κB/rel (NF-κB) family may be important in cell survival by regulating unidentified, anti-apoptotic genes. One such gene that protects cells from apoptosis induced by Fas or tumor necrosis factor type α (TNF),IEX-1L, is described here. Its transcription induced by TNF was decreased in cells with defective NF-κB activation, rendering them sensitive to TNF-induced apoptosis, which was abolished by transfection with IEX-1L. In support, overexpression of antisense IEX-1L partially blocked TNF-induced expression ofIEX-1L and sensitized normal cells to killing. This study demonstrates a key role of IEX-1L in cellular resistance to TNF-induced apoptosis.

Tumor necrosis factor type α (TNF), a major inflammatory cytokine, simultaneously activates a cell suicide program and an anti-death activity that results in resistance of many cancer cells to TNF-mediated killing, thus limiting its use in cancer therapy (1). TNF-stimulated anti-death activity, unlike TNF-induced cell death, depends on de novo protein synthesis and the genes involved appear to be transcriptionally activated by transcription factors of the nuclear factor–κB/rel (NF-κB) family (2, 3). Hence, cells lacking NF-κB subunit RelA (p65) or overexpressing a mutated inhibitor IκBα gene showed enhanced susceptibility to TNF-mediated killing (4). Using the mRNA differential display technique (5), we cloned a gene that appeared to be the same as a previously reported immediate-early response gene IEX-1 (6), except that it had an in-frame insertion of 111 nucleotides at position 211 of the coding region for IEX-1, and it could encode a longer polypeptide with a 37–amino acid insertion relative toIEX-1 (7). The longer IEX-1 [referred to here as IEX-1L; the original IEX-1 is referred to as IEX-1S (short)] was found to be generated fromIEX-1 in the absence of RNA splicing as it contained the entire intron sequence of IEX-1 (8).

IEX-1L protein was demonstrated in 293 cells transiently transfected with a pcDNA-HA·Tag–IEX-1L plasmid by using a monoclonal antibody (mAb) to influenza virus hemagglutinin (HA) (Fig. 1A, arrow L-HA) (9). The difference between the molecular mass of HA-IEX-1L (32 kD) and of HA–IEX-1S (28 kD) could be accounted for by a 37–amino acid insertion present in IEX-1L. Endogenous IEX-1L protein was also detected by using a polyclonal antibody (Ab) to IEX-1 (10) (Fig. 1B, arrow L), which was larger than the reported IEX-1S protein (6) (Fig. 1B, arrow S). When McF-7 cells expressed IEX-1L or IEX-1S fused to green fluorescence protein (GFP) (9), a typical pattern of fluorescence around the nuclear periphery and endoplasmic reticulum membrane was observed (Fig. 1C), which was distinct from the diffuse distribution of fluorescence visible throughout the entire cell when GFP alone was expressed (Fig. 1D). The localization of IEX-1L and IEX-1S in endoplasmic reticulum and on nuclear membrane was confirmed by immunoelectron microscopic study with a GFP-specific Ab (8). This observation is consistent with the presence of a putative transmembrane integrated region in IEX-1 proteins (6).

Figure 1

Expression of transfected (A) and endogenous (B) IEX-1L proteins. (A) Immunoprecipitation followed by immunoblotting analysis was carried out as described (16) with cell lysates (50 μg) prepared from 293 cells transfected with the indicated HA · tag–containing constructs (9) using mAb 12CA5 to HA (Boehringer Mannheim). Lane Ab Control, an irrelevant mouse mAb used as control in immunoprecipitation. (B) Jurkat cell lysates prepared at various times after treatment with phorbol 12-myristate 13-acetate (PMA) (50 ng/ml) were analyzed as in (A) with anIEX-1–specific Ab (10). Lane C, normal rabbit serum control. Molecular size markers (kD) are shown on the left (A) and right (B). IgH refers to the Ig heavy chain. (C andD) McF-7 cells were transfected with a plasmid pEGFP-IEX-1L (C) or pEGFP alone (D) and photomicrographed by inverted fluorescence microscopy after 40 hours. Bar = 20 μm. A similar pattern was also observed with cells expressing an IEX-1S–GFP fusion protein.

To investigate the function of IEX-1L and IEX-1S, we stably transfected Jurkat cells with the IEX-1L orIEX-1S coding frame inserted into a pRc/CMV plasmid. Easily detectable amounts of RNA of either IEX-1L orIEX-1S were observed in two of three randomly selected clones transfected with pRc/CMV–IEX-1L and in two of four clones carrying pRc/CMV–IEX-1S but in none of the clones receiving the parent control vector (Fig. 2B). When these cells were treated with mAb to Fas, the viability was more than 80% for clone 14/L and 70% for clone 15/L, both expressing IEX-1L after 4 hours of treatment (Fig. 2A). These percentages were significantly higher than the 35% observed in wild-type cells and in control plasmid-transfected clones (13/V and 11/V) and they were similar to that observed with Bcl-2–transfected Jurkat cells (11). In contrast, clone 9/L cells expressed undetectable IEX-1L RNA, showing little effect on cell susceptibility to Fas-mediated apoptosis compared with untransfected cells. Although IEX-1S transfectants, clones 2/S and 4/S, expressed amounts of IEX-1S mRNA similar to those transfected with IEX-1L (Fig. 2B), they offered no protection against Fas-induced killing, which strongly indicates a specific anti-death function of IEX-1L. Moreover, by using specific mAbs, we were able to show that all these clones expressed similar amounts of cell surface Fas and intracellular Bcl-2 and Bcl-x molecules, ruling out the possibility that the observed protection ofIEX-1L was a result of effects of these key anti-apoptotic or apoptotic molecules (8, 11).

Figure 2

Protective effect ofIEX-1L against Fas-mediated apoptosis. (A) Viability ofIEX-1L–transfected Jurkat cell clones. Jurkat cells were stably transfected with a plasmid pRc-CMV-IEX-1L (L), pRc-CMV-IEX-1S (S), or pRc- CMV control vector (V) and the subclones were selected by limiting dilution. Jurkat cells (JK), Bcl-2–transfected Jurkat cells (Bcl-2), and representative subclones of each transfection were treated with mAb 7C11 to Fas (1:10,000 ascites) for various times and then stained with propidium iodide (PI). Percentages of viable cells are shown as mean ± standard deviation (SD). One representative result of five independent experiments performed in triplicate is shown. (B)IEX-1 expression in Jurkat cell clones in (A) was analyzed by Northern blotting with an IEX-1 probe. EndogenousIEX-1 RNA (1.3 kb) in PMA-stimulated Jurkat cells (lane PMA/2 hr) is shown as a positive control, and transfectedIEX-1 are about 0.6 kb for IEX-1L and 0.5 kb forIEX-1S, indistinguishable in size in the blot. The Pst fragment of glyceraldehyde-3-phosphate dehydrogenase (G3PDH) is used as an equal RNA loading control. (C) Apoptotic cell death of bulk cultures of Jurkat cell transfectants. Jurkat cells were stably transfected with the same plasmids as in (A) (12). Untransfected control (C) and transfected Jurkat cells were either treated with mAb to Fas for 8 hours or left untreated. Percentages of apoptotic cells (SubG1 population) were determined by flow cytometry analysis of PI-stained cells. Data shown are means ± SD of three independent experiments using the same bulk transfectants. (Inset)IEX-1 RNA expression analyzed as in (B). (D) Protection of Fas-induced apoptosis in McF-7 cells transiently transfected with IEX-1L. McF-7 cells were cotransfected with a pEGFP–IEX-1L (L), pEGFP–IEX-1S (S), or pEGFP (V) plasmid (9) along with a pRc-Fas plasmid for 36 hours, after which the cells were treated with mAb 7C11 for 6 hours or were left untreated. Percentages (mean ± SD) of apoptotic cells (based on cell morphology) were obtained by averaging the results from triplicate wells of a six-well plate, with about 100 cells counted in each by inverted fluorescence microscopy. One representative result of three independent experiments is shown.

To avoid clonal variations, we used stably transfected bulk cultures of Jurkat cells to test IEX-1L–mediated protection (12). As indicated in Fig. 2C, Jurkat cells transfected with a control vector or IEX-1S–containing plasmid underwent Fas-induced apoptosis at degrees indistinguishable from wild-type cells (about 60%). In contrast, the Fas-mediated cell death ofIEX-lL–bearing transfectants was significantly lower (about 23%).

IEX-1L–mediated protection appeared not to be restricted in Jurkat cells. McF-7 cells transiently transfected with anIEX-1S–GFP construct or GFP vector alone, along with a pRc/CMV-Fas plasmid, underwent Fas-mediated apoptosis by 35 to 40%, whereas the percentage of cells undergoing Fas-induced apoptosis was reduced by half if they expressed an IEX-1L–GFP protein (Fig. 2D).

Interestingly, IEX-1L restored the resistance to TNF-induced cell death of p65KO3T3 cells isolated from RelA−/− mice and Jurkat IκBαM cells generated by expression of a mutated IκBα protein (4, 13). As shown in Fig. 3A, TNF-induced apoptosis increased dramatically from 10 to 70% in p65KO3T3 cells bearing the parental control vector or in untransfected p65KO3T3 cells. In contrast,IEX-1L–transfected cells were markedly less sensitive to the apoptotic effect, demonstrating an ability of IEX-1L to compensate the cells for a loss of RelA that results in them becoming sensitive to TNF-mediated apoptosis. The IEX-1L–mediated protection was highly reproducible with two independently transfected bulk cell cultures, p65KO3T3/IEX-1L(1) and p65KO3T3/IEX-1L(2), in four separate experiments. This conclusion was further strengthened by a similar protection obtained from three independently transient transfections. As shown in Fig. 3D, viability of p65KO3T3 cells in response to TNF treatment increased from 20 to 35% in the absence of IEX-1L to 65 to 72% in the presence of IEX-1L.

Figure 3

IEX-1L compensates p65KO3T3 (A) and Jurkat IκBαM (B) cells for a defect in the activation of NF-κB proteins. The cells were stably transfected with a BCMGS-Hyg–IEX-1L (IEX-1L) plasmid or BCMGS-Hyg vector alone (V) and treated with the indicated concentrations of either mouse (A) or human (B) TNF for 40 hours (9, 12). Percentages of apoptotic cells were analyzed as in Fig. 2C. One representative experiment of four (A) or three (B) experiments performed in duplicate is shown. IEX-1L expression in these transfectants is shown by immunoprecipitation, followed by immunoblotting analysis with IEX-1–specific Ab. V, L1, and L2 refer to vector alone- andIEX-1L–transfected p65KO3T3 cell culturesIEX-1L(1) and IEX-1(2), respectively (A, inset). C, V, and L represent untransfected and vector alone- orIEX-1L-transfected Jurkat IκBαM cells (B, inset). (C) Northern blotting analysis of IEX-1L expression in Jurkat IκBαM cells. Jurkat IκBαM cells and Jurkat Lxsn control cells were stimulated with TNF (1000 U/ml). Blot containing mRNA at 2 μg per lane purified at the indicated time points was hybridized with an IEX-1L–specific probe, consisting of a 111-nucleotide sequence present only in theIEX-1L gene (7). G3PDH was used as an equal mRNA loading control. (D) Transient expression ofIEX-1L protects p65KO3T3 cells from TNF-induced cell death. p65KO3T3 cells were cotransfected with equal amounts of plasmid pRc–IEX-1L (IEX-1L) or pRc empty vector (Vector), along with pCMV-LacZ (lacZ) expressing vector (Stratagene) by the lipofectamine protocol. Mouse TNF-α (1000 U/ml) was added 40 hours later for 20 hours and then the cells were stained with 5-bromo-4-chloro-3-indolyl β-d-galactopyranoside. Viable blue cells remaining after TNF treatment are shown as a percentage of untreated viable blue cells. Each represents the mean of triplicate wells of a six-well plate ± SD and cells are counted as described in Fig. 2D. Data of three independent experiments (Exp. 1, 2, and 3) performed are shown.

Similarly, the percentage of apoptotic Jurkat IκBαM cells or cells transfected with vector alone increased from 10% to more than 55% over the TNF concentrations used (Fig. 3B). However, Jurkat IκBαM cells transfected with IEX-1L showed a diminished capacity to undergo TNF-induced apoptosis; in fact, the TNF dose-response curve of IEX-1L–transfected Jurkat IκBαM cells was similar to that observed with Jurkat Lxsn cells (Jurkat cells transfected with a Lxsn vector alone) (4) (Fig. 3B).

We next examined whether TNF-induced expression of IEX-1L was defective in these cells. As shown in Fig. 3C, IEX-1L mRNA was about 14-fold lower in density in Jurkat IκBαM cells than in Jurkat Lxsn cells after 1 hour of treatment with TNF. A weak band similar to IEX-1L in size ( about 1.3 kb) was observed in wild-type 3T3 (wt3T3) but not in p65KO3T3 cells 30 min after treatment with mouse TNF (8). The difference in the amount ofIEX-1L transcript appeared to be diminished with prolonged TNF treatment.

These results suggested that an early defect in the expression ofIEX-1L was likely a cause for the increased susceptibility of p65KO3T3 and Jurkat IκBαM cells to TNF-mediated apoptosis. To investigate this directly, we expressed IEX-1L in an antisense orientation in normal Jurkat cells. This partially blocked TNF-stimulated expression of IEX-1, as evidenced by a reduction in immunofluorescent staining by IEX-1–specific antibody in cells bearing antisense IEX-1L relative to that in control cells (Fig. 4A). In spite of having normal NF-κB proteins, the antisenseIEX-1L–transfected cells underwent apoptosis at highly significant (P < 0.01) amounts at a TNF concentration of 200 units (U)/ml compared with those observed with untransfected cells or with cells transfected with vector alone (Fig. 4B), which suggests that NF-κB–mediated protection depends on expression ofIEX-1L. The effect of antisense IEX-1L became less pronounced at a TNF concentration of 500 U/ml (P < 0.05), presumably due to a larger amount of induced IEX-1L transcript that overcame the neutralizing effect of antisenseIEX-1L mRNA or additional anti-apoptotic genes being activated.

Figure 4

Effect of antisense IEX-1L. (A) Jurkat cells were stably transfected with either a construct inserted with IEX-1L in an antisense orientation (JK/pRcAIEX-1L) or pRc/CMV vector alone (Jk/pRc) (12). The transfectants and untransfected cells (JK) were treated with or without TNF (500 U/ml) for 3 hours, fixed in 1% paraformaldehyde, permeabilized in 1% digitonin, and then stained withIEX-1–specific Ab, followed by phycoerythrin-conjugated goat immunoglobulin G to rabbit. Numbers represent mean fluorescence intensity for analysis of 5000 events; dashed line represents histograms of an irrelevant Ab. (B) Cells in (A) were treated for 40 hours with indicated concentrations of TNF. Apoptosis was analyzed as in Fig. 2C. Percentages represent mean ± SD of three independent experiments with one transfection. One representative experiment of three independent transfections performed is shown. Statistic significance was analyzed by Student's ttest.

The data presented here demonstrate that cellular resistance to TNF-induced killing is directly related to the ability of cells to rapidly express IEX-1L in response to TNF stimulation. Thus, a rapid increase in the expression of IEX-1L after addition of TNF may be key to the mechanism underlying TNF-mediated protection. Indeed, all TNF killing-resistant cell lines tested including HeLa, Jurkat, U937, Sw480, H9, NIH 3T3, and Hut78 cells express IEX-1L after TNF stimulation (6, 8). In contrast, a decrease or delay in TNF-induced expression ofIEX-1L is likely to increase cell susceptibility to TNF-induced apoptosis, as was found in p65KO3T3, Jurkat-IκBαM, and Jurkat cells bearing an antisense IEX-1L (4). Our unpublished data also showed that IEX-1L was potentially regulated by the RelA/c-rel complex (8), in agreement with previous observations that overexpression of the c-rel gene protected cells from TNF-induced cell death (2, 3) and that RelA gene knockout mice died at 15 days of gestation (14). However, unlike RelA−/− mice, mice lacking the c-rel gene are developmentally healthy (15), which suggests thatIEX-1L may be only one of the NF-κB/Rel protein-regulated survival genes.


View Abstract

Navigate This Article