If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
NF-κB Inhibition Reveals Differential Mechanisms of TNF Versus TRAIL-Induced Apoptosis Upstream or at the Level of Caspase-8 Activation Independent of cIAP2
Tumor Immunology Program, German Cancer Research Center, Heidelberg, GermanyLaboratory for Experimental Oncology and Radiation, Academic Medisch Centrum UVA, Amsterdam, The Netherlands
Laboratory for Experimental Dermatology, Department of Dermatology and Venerology, Otto-von-Guericke-University Magdeburg, Leipziger Strasse 44, Magdeburg 39120, Germany
Death ligands not only activate a death program but also regulate inflammatory signalling pathways, for example, through NF-κB induction. Although tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) and TNF both activate NF-κB in human keratinocytes, only TRAIL potently induces apoptosis. However, when induction of NF-κB was inhibited with a kinase dead IKK2 mutant (IKK2-KD), TNF- but not TRAIL-induced apoptosis was dramatically enhanced. Acquired susceptibility to TNF-induced apoptosis was due to increased caspase-8 activation. To investigate the mechanism of resistance of HaCaT keratinocytes to TNF-induced apoptosis, we analyzed a panel of NF-κB-regulated effector molecules. Interestingly, the inhibitor of apoptosis protein (IAP) family member cIAP2, but not cIAP1, X-linked inhibitor of apoptosis, TNF receptor-associated factor (TRAF)-1, or TRAF2, was downregulated in sensitive but not in resistant HaCaT keratinocytes. Surprisingly, however, stable inducible expression of cIAP2 was not sufficient to render IKK2-KD-sensitized keratinocytes resistant to TNF, and reduction of cIAP2 alone did not increase the sensitivity of HaCaT keratinocytes to TNF. In conclusion, we demonstrate that inhibition of NF-κB dramatically sensitizes human keratinocytes to TNF- but not to TRAIL-induced apoptosis and that this sensitization for TNF was largely independent of cIAP2. Our data thus clearly exclude the candidates proposed to date to confer TNF apoptosis resistance and suggest the function of an unanticipated effector of NF-κB critical for the survival of HaCaT keratinocytes upstream or at the level of caspase-8 activation.
Abbreviations:
Ab
antibody
cFLIPL
c-FLICE inhibitory protein long form
EE
constitutively active
FADD
Fas-associated death domain protein
4-HT
4-hydroxytamoxifen
IAP
inhibitor of apoptosis protein
KD
kinase dead
siRNA
small interfering RNA
TNF
tumor necrosis factor
TRAF
TNF receptor-associated factor
TRAIL
TNF-related apoptosis-inducing ligand
XIAP
X-linked inhibitor of apoptosis
Introduction
Apoptosis is a highly regulated physiological process that is critical for tissue homeostasis and is initiated by a plethora of physiological and pathophysiological stimuli, such as UV irradiation, growth factor deprivation, or chemotherapeutic drugs (
). The so-called extrinsic pathway is initiated by ligation of death receptors whose ligands such as tumor necrosis factor (TNF), CD95L (Fas ligand), and TNF-related apoptosis-inducing ligand (TRAIL) are all members of the TNF superfamily (
). They have been studied intensively over the past decade, and their role in activation-induced cell death, autoimmune disorders, immune privilege, and tumor evasion from the immune system is now well established (reviewed in
). TRAIL and CD95L are mainly considered pro-apoptotic ligands that act by facilitating recruitment of the adaptor protein Fas-associated death domain protein (FADD) and activation of the initiator caspases 8 and 10 to form a “death-inducing signalling complex,” the so-called DISC. Activated initiator caspases process and activate effector caspases, which ultimately cause cell death (
). TNF binds to two different membrane-bound receptors, TNF-R1 and TNF-R2. However, TNF binding initially promotes recruitment of TNF receptor-associated death domain (TRADD), receptor-interacting protein-1 (RIP-1), and TNF receptor-associated factor (TRAF)-2 to the membrane-associated TNF-R1, which results in activation of NF-κB and pro-survival signalling (
Fas-associated death domain protein and caspase-8 are not recruited to the tumor necrosis factor receptor 1 signaling complex during tumor necrosis factor-induced apoptosis.
). Following an initial burst of pro-survival signalling, the membrane-associated signalling complex dissociates and reassembles in the cytoplasm with FADD and caspase-8, and caspase-8 becomes subsequently activated (
). In nearly all cases, the initial pro-survival signalling from the membrane-associated complex is sufficient to protect cells from the subsequent formation of the pro-apoptotic caspase-8-containing complex, which additionally contains components of the IKK complex required for activation of NF-κB, another important function of TNF (
Based on the observation that inhibition of NF-κB is required to sensitize cells to TNF-induced apoptosis, it has been widely accepted that NF-κB target genes are responsible for the maintenance of resistance to TNF-induced apoptosis. The inhibitors of apoptosis protein, cIAP1 and cIAP2, were originally identified as TNF-R2-associated proteins (
). XIAP, in particular, is a well-characterized caspase inhibitor. For these reasons, it has long been suspected and claimed that the upregulation of cIAP1/2 following TNF signalling is responsible, either partially or in its entirety, for the protection afforded by the initial burst of pro-survival signalling (
). However, owing to technical difficulties and lack of tools, it has been extremely difficult to prove or disprove this hypothesis. Therefore, the exact mechanism of resistance to TNF-mediated apoptosis has not been elucidated so far.
Keratinocytes express all the components required to execute the apoptotic program, and the intricate balance between apoptotic and nonapoptotic signals may be crucial for the outcome of death receptor triggering in the skin (
). Furthermore, the role of death receptors such as TNF-R1, CD95, TRAIL-R1, or TRAIL-R2 in keratinocytes is well established in vitro and has been confirmed in several knockout models (for review, see
). Pathological modulation of apoptosis signalling in the skin can lead to skin cancer or inflammatory diseases like psoriasis or alopecia areata (for review, see
). In this regard, it is significant that induction of apoptosis by keratinocyte-derived CD95L has been suggested to represent an important pathogenic factor for toxic epidermal necrolysis (
). Importantly, inhibition of TNF signalling shows clinical benefit for inflammatory diseases like psoriasis, indicating its importance in the pathophysiological context of skin diseases. Our previous reports have demonstrated that inhibition of NF-κB activation does not sensitize keratinocytes to apoptosis induction by the death ligand TRAIL, while effectively interfering with TRAIL-induced NF-κB activation and the induction of its target genes. These data suggested that gene induction is a distinct apoptosis-independent signal elicited by TRAIL receptors in the skin (
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
In this report, we compared the mechanisms governing the sensitivity to TNF-mediated versus TRAIL-induced apoptosis. We first show that inhibition of NF-κB dramatically sensitized HaCaT keratinocytes to TNF-induced apoptosis. We then demonstrate that TNF-induced apoptosis is inhibited upstream or at the level of caspase-8 activation as it was efficiently blocked by the long form of c-FLICE inhibitory protein (cFLIPL). Upon analysis of a larger panel of anti-apoptotic NF-κB target genes, we found a specific and significant reduction in the expression of cIAP2, whereas the levels of cIAP1, TRAF1, TRAF2, or cFLIP were not affected. Surprisingly, however, selective downregulation of cIAP2 by small interfering RNA (siRNA) did not sensitize HaCaT keratinocytes to TNF, and inducible lentiviral expression of cIAP2 in IKK2-kinase dead (KD)-expressing HaCaT keratinocytes was not sufficient to re-establish TNF resistance. In contrast to the most commonly accepted model (
), our data show that although NF-κB inhibition is a prerequisite for sensitization to TNF-induced apoptosis, the regulated expression of cIAP2 alone is not sufficient to maintain resistance to TNF in the absence of constitutive or induced NF-κB activity. Furthermore, they strongly suggest that at least one other NF-κB target gene that acts upstream or at the level of caspase-8 must be important to maintain resistance to TNF-induced apoptosis.
Results
Modulation of NF-κB activity by IKK2 mutants in human HaCaT keratinocytes
We recently showed that inhibition of NF-κB did not modify the sensitivity to TRAIL-induced apoptosis in keratinocytes (
). To test the impact of modulation of NF-κB activity on TNF responses in human skin, we generated HaCaT keratinocyte cell lines carrying either KD or constitutively active (EE) mutants of IKK2. Western blot analysis confirmed expression of the high and comparable expression of the respective molecules in HaCaT cells as compared to low levels of expression of the wild-type kinase in control-infected cells (Figure 1a). IKK2-KD expression resulted in complete inhibition of basal as well as either TRAIL- or TNF-induced NF-κB DNA binding (Figure 1b). As expected, IKK2-EE-expressing HaCaT keratinocytes showed high basal NF-κB DNA binding and no further induction upon treatment with TRAIL or TNF (Figure 1b). These bulk-infected HaCaT keratinocytes were examined for apoptosis induction following stimulation with TRAIL or TNF, measuring the percentage of cells with hypodiploid DNA content by propidium iodide staining (Figure 2). Consistent with our previous data (
), the prominent sensitivity to TRAIL-induced apoptosis was not modified by ectopic expression of mutants of IKK2 (Figure 2a), and TRAIL-treated cells showed an identical dose-dependent kill curve as vector cells (Figure 2b). In contrast, TNF triggering of vector-infected cells revealed their resistance to TNF-induced apoptosis. Strikingly, resistance to TNF-induced apoptosis is abrogated in IKK2-KD-expressing HaCaT keratinocytes, whereas expression of IKK2-EE had no effect on the extent of TNF-mediated apoptosis as compared to vector cells (Figure 2). When we used inducible expression of IKK2 mutants, very similar results were obtained, strongly arguing against a cell culture artifact occurring during the selection procedure (Figure S3). Taken together, our data suggested that sensitivity to TNF-induced apoptosis is regulated by NF-κB, but plays a negligible role in TRAIL-mediated apoptosis in HaCaT keratinocytes.
Figure 1Overexpression of IKK2 mutants in human HaCaT keratinocytes. (a) HaCaT keratinocytes were retrovirally transduced with IKK2-KD, IKK2-EE, or control vector as described in Materials and Methods. A total of 20 μg protein of total cellular lysates was separated by western blotting and subsequently analyzed for expression of IKK2. Tubulin served as a loading control. (b) Suppression and constitutive activation of NF-κB in IKK2-KD- and IKK2-EE-expressing cells. Pools of infected HaCaT keratinocytes were either left untreated or stimulated with 1 μg ml−1 TRAIL or TNF for 1 hour, harvested, and protein content was estimated in triplicate. Subsequently, nuclear extracts were analyzed for κB-specific DNA binding by electrophoretic mobility shift assay. The positions of p65/p50 heterodimers or p50 homodimers are indicated and were confirmed using supershift analysis in separate experiments (Figure S1b).
Figure 2Overexpression of IKK2-KD, but not IKK2-EE leads to sensitization to TNF-mediated apoptosis in HaCaT keratinocytes. (a) TNF-induced apoptosis of IKK2-KD-expressing HaCaT keratinocytes. Pools of infected cells were either left untreated or stimulated with 1 μg ml−1 TRAIL or 1 μg ml−1 TNF for 6 hours, harvested, and examined for the induction of apoptosis using hypodiploidy analysis. (b) Dose-dependent induction of TNF- or TRAIL-mediated cell death in IKK2 mutant-expressing HaCaT keratinocytes. Pools of infected cells were either left untreated or stimulated with increasing concentrations of TRAIL or TNF for 16–24 hours. Viability was examined by crystal violet staining. The mean±SD of a total of four independent experiments is shown.
Inducible expression of IKK2 mutants recapitulates results using constitutive expression of IKK2 mutants. (a) Inducible expression of IKK2-KD or IKK2-EE mutants in HaCaT keratinocytes. HaCaT keratinocytes were lentivirally infected with inducible IKK2-KD or IKK2-EE constructs as outlined in materials and methods. Subsequently cells were induced for 24 hours with 100 nM of 4-HT or diluent alone. Cellular lysates were assayed for expression of IKK2 by Western blot analysis. Tubulin served as a loading control. (b) Induction of IKK2 mutants leads to sensitization to TNF-induced apoptosis in IKK2-KD-, but not IKK2-EE-expressing HaCaT keratinocytes. Cell lines were treated for 48 hrs with 4-HT, and were subsequently treated for 24 hrs in the presence of the indicated concentrations of HF-TNF. In the absence of IKK2-KD mutant expression, no significant variation of TNF apoptosis sensitivity is detectable when compared to control-infected cell lines. Shown are mean +/- SD of a representative experiment of a total of two independent experiments
Activation of caspases during TNF- and TRAIL-induced cell death
To investigate the mechanism responsible for the loss of resistance to TNF-mediated cell death in IKK2-KD-expressing HaCaT keratinocytes, we analyzed the activation of caspases following TNF or TRAIL treatment (Figure 3). In control cells, as well as in IKK2-KD-expressing HaCaT keratinocytes, processing of caspase-8 was readily detectable following TRAIL stimulation within 120 minutes (Figure 3a, lanes 15–28). In contrast, caspase processing was undetectable in TNF-stimulated control cells, whereas IKK2-KD HaCaT cells showed cleavage of caspase-8, cFLIP (detectable upon prolonged exposure), and the main effector caspase-3 with delayed kinetics only after 240–360 minutes (Figure 3a, lanes 1–14). Moreover, a marked upregulation of cFLIPS and a slight increase in cFLIPL expression was detectable following TNF stimulation for 4–6 hours in an NF-κB-dependent manner (Figure 3a, lanes 1–14). It was theoretically possible that TNF induced upregulation of other death ligands such as TRAIL or CD95L and that these ligands killed IKK2-KD cells in an autocrine loop. Such autocrine induction of apoptosis could explain the 2-hour delay in caspase activation during TNF-mediated cell death (Figure 3a, lane 13) is compared to TRAIL-mediated cell death (Figure 3a, lanes 19 and 26). To test this hypothesis, we performed experiments in the presence of receptor fusion proteins (TNF-R2-Fc, CD95-Fc, and TRAIL-R2-Fc). We found that TNF-induced cell death in NF-κB-inhibited cells was fully blocked by the addition of TNF-R2-Fc. Although TRAIL-R2-Fc and CD95-Fc proteins were able to block TRAIL- and CD95L-mediated cell death and therefore were fully biologically active, they were unable to prevent TNF-induced cell death (Figure 3b), excluding the possibility that TNF-induced cell death is mediated by autocrine production of CD95L or TRAIL. It remained possible that higher death receptor expression in the IKK2-KD cells determined increased sensitivity to TNF. We therefore characterized expression of TRAIL and TNF surface receptors. Consistent with our previous reports (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
), control HaCaT cells expressed TRAIL-R1 and TRAIL-R2 and barely detectable levels of TNF-R1 and TNF-R2 on the cell surface (Figure S2a). IKK2-KD-expressing cells showed similar levels of these receptors when compared to controls (Figure S2a). IKK2-EE-expressing HaCaT cells showed higher surface expression levels of TRAIL-R1, as well as TRAIL-R2, TNF-R1, and TNF-R2, in agreement with a previous report that NF-κB induces TRAIL death receptor expression (
). Yet, IKK2-EE-expressing cells were as resistant to TNF-mediated cell death as control cells, despite higher expression levels of TNF-R1 and TNF-R2. Conversely, IKK2-KD cells, if anything, expressed slightly less TNF-R1 and TNF-R2 than control cells, but they were markedly more sensitive to TNF-induced cell death than their respective controls. Taken together, these data demonstrate that receptor expression is not the major determinant of TNF apoptosis sensitivity in HaCaT keratinocytes.
Figure 3TNF stimulation of IKK2-KD-transduced HaCaT keratinocytes leads to caspase-8 and caspase-3 cleavage with delayed kinetics. (a) TNF-induced caspase-8, caspase-10, and cFLIP cleavage. Cells expressing IKK2-KD or control vector-infected cells were incubated for the indicated times with TRAIL or TNF (both at 1 μg ml−1) and subsequently analyzed for cleavage of caspase-8 (p43/p41/p18), cFLIP (p43), and caspase-3 by western blot (WB). (b) Autocrine secretion of death ligands such as TRAIL or CD95L is not responsible for TNF-induced cell death. Pools of infected HaCaT keratinocytes were either left untreated or stimulated with TRAIL, TNF, or CD95L (all at 1 μg ml−1) for 16–24 hours in the presence or absence of TNF-R2-Fc, TRAIL-R2-Fc, or CD95-Fc (all at 10 μg ml−1). Cell viability was determined by crystal violet staining. The mean±SD of a total of two independent experiments is shown. TNF-induced cell death in NF-κB-inhibited cells was fully blocked by the addition of TNF-R2-Fc (Enbrel), whereas TRAIL-R2-Fc and CD95-Fc were ineffective. Control cultures incubated with TRAIL or CD95L showed effective inhibition of CD95L- or TRAIL-mediated cell death by these reagents.
TRAIL-R, TNF-R, and XIAP levels, as well as constitutive and TNF-induced expression of cIAP2 in IKK2 mutant-expressing HaCaT keratinocytes. (a) TNF and TRAIL receptor expression analysis of keratinocytes infected with mutants of IKK2. Control keratinocytes and keratinocytes expressing either IKK2-KD or IKK2-EE were examined for cell surface expression of TRAIL-R1-4 or TNF-R1-2 by FACScan for GFP positive/propidium iodide negative cells and compared to the respective isotype-specific control staining (open areas). One of four independent experiments is shown. (b) XIAP is absent at the protein level in HaCaT keratinocytes Total cellular lysates of three different donors of primary human keratinocytes (PK), HaCaT and SCC25 squamous cell carcinoma were separated by western blot analysis and subsequently analyzed for expression of XIAP. Shown are a 15 sec and a 3 min exposure to assay for low amounts of XIAP at the protein level. (c) TNF stimulation induces cIAP2 within 6 hours in an NF-κB-dependent manner. Cells expressing IKK2-KD or control vector-infected cells were incubated for the indicated times with TNF (1 μg/ml) and subsequently analyzed for expression of cIAP2 by Western blot (WB). Tubulin serves as a loading control.
cFLIP differentially affects TNF- and TRAIL-induced apoptosis signalling
cFLIP is a crucial inhibitor of CD95 or TRAIL-R1/2 signalling pathways, and mouse embryonic fibroblasts from cFLIP-deficient animals are highly sensitive to TNF-induced apoptosis (
). Basal levels of cFLIP were not detectably changed in IKK2-KD-expressing HaCaT cells when compared to control cells. However, cFLIPL (55 kDa) and cFLIPS (25 kDa) were slightly induced in an NF-κB-dependent manner within 4–6 hours of TNF treatment (compare Figure 3a, lanes 1–14), suggesting that cFLIP may be a crucial factor for TNF apoptosis resistance as reported (
). To further explore the impact of cFLIP on TNF-induced cell death in keratinocytes, we next sequentially expressed both cFLIPL and IKK2-KD in HaCaT keratinocytes. In line with our previous results, cFLIP efficiently inhibited TRAIL-induced apoptosis (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
). Notably, TNF-induced cell death (Figure 4a) as well as TNF-induced caspase-8 and caspase-3 cleavage was largely inhibited upon overexpression of cFLIPL (Figure 4b). Interestingly, in cells expressing cFLIPL, cleavage of cFLIPL to the p43 fragment was detectable only when basal or induced NF-κB activation was inhibited by expression of IKK2-KD, indicating that in these cells caspase activity is sufficient for cFLIPL cleavage, but not for allowing sufficient effector caspase activation required for apoptosis induction. Thus, as expected, expression of high levels of cFLIP is sufficient for TRAIL and TNF apoptosis resistance. To further explore the dynamic requirement of cFLIP for TNF resistance, we infected HaCaT keratinocytes with retroviral siRNA against cFLIP. Successfully infected cells were selected for a short time and analyzed for TRAIL- or TNF-mediated caspase-8 activation as well as apoptosis sensitivity. Of note, the expression of a retroviral control construct alone increased basal cFLIP expression in HaCaT keratinocytes (data not shown). Nonetheless, cFLIP siRNA sensitized cells for TRAIL-induced caspase-8 activation in a small but highly reproducible as well as significant manner (Figure 4c), whereas cFLIP-repressed cells remained insensitive to TNF-induced caspase activation or apoptosis induction (Figure 4d). Downregulation of cFLIP is therefore not sufficient to sensitize for TNF-induced cell death. Thus, we hypothesized that other NF-κB-dependent factors may contribute to the maintenance of TNF apoptosis resistance upstream of caspase-8 activation in HaCaT keratinocytes.
Figure 4The role of cFLIP in TNF-induced apoptosis in IKK2-KD-infected HaCaT keratinocytes. (a) Pools of cFLIPL-infected cells as indicated were either left untreated or stimulated with TRAIL or TNF (each at 1 μg ml−1) for 6 hours, harvested, and examined for the induction of TNF- or TRAIL-mediated cell death. A representative experiment of three independent experiments is shown. (b) Inhibition of caspase-8 cleavage in TNF-treated IKK2-KD/c-FLIPL-expressing HaCaT keratinocytes. cFLIPL- and IKK2-KD-expressing HaCaT keratinocytes were mock-treated or stimulated with 1 μg ml−1 of TRAIL or TNF for 3 hours and total cellular lysates were analyzed for cFLIPL, cFLIPS, caspase-8, and caspase-3. Molecular weights of full-length proteins and cleavage products are indicated. (c) Stable retroviral siRNA expression against cFLIP sensitizes HaCaT keratinocytes for TRAIL-, but not for TNF-induced apoptosis. Upon successful infection with the respective constructs, cells were mock-treated or treated with TRAIL or TNF (both at 1 μg ml−1) for 3 hours and subsequently subjected to western blot (WB) analysis for cFLIP and caspase-8. Molecular weight of full-length proteins and cleavage products are indicated. An intentionally overexposed blot is shown to detect low levels of cFLIP expression in shRNA-expressing cells. Tubulin served as a loading control. (d) cFLIP downregulation sensitizes HaCaT keratinocytes for TRAIL- but not for TNF-induced apoptosis. Pools of infected cells as shown in (c) were either left untreated or stimulated with increasing concentrations of TRAIL or TNF for 16–24 hours. Cells were then harvested and viability was examined by crystal violet staining. A representative experiment of a total of five independent experiments is shown. The sensitivity of cFLIP siRNA-infected cells to TRAIL was moderately increased in independent experiments but in a highly reproducible manner. Two-way ANOVA analysis was performed with post hoc testing (Bonferroni). Global testing demonstrated high significance (P<0.001) of treatment differences. Post hoc testing revealed that the difference between both cell types was insignificant for TNF treatment, whereas the difference between cFLIP-siRNA-treated cells and control cells was significant for TRAIL treatment.
cIAP2 is not critical for the modulation of TNF-induced apoptosis in HaCaT keratinocytes
Although overexpression of cFLIP can inhibit apoptosis induced by both TNF and TRAIL (Figure 4), we thought it unlikely that this accounted for the differential sensitivity of the IKK2-KD cells, given the fact that modulation of physiological levels of cFLIP using siRNA was effective to sensitize for TRAIL- but not TNF-induced apoptosis. We therefore investigated a panel of target genes reported to be regulated by NF-κB by RNase protection assays. The steady-state level of cIAP2 mRNA was remarkably and specifically downregulated in the IKK2-KD mutant cell line. Surprisingly, none of the other presumptive NF-κB targets, such as cIAP1, XIAP, TRAF1, or TRAF2, were downregulated (Figure 5a and b). The protein levels of cIAP2 were also markedly reduced in the IKK2-KD cell lines and, gratifyingly, specifically upregulated in the IKK2-EE mutant cells (Figure 5c). These data, obtained from three independent, polyclonal cell lines, strongly supported the concept that steady-state levels of cIAP2 are specifically regulated by NF-κB in HaCaT keratinocytes and furthermore suggested that the key difference that determines NF-κB-mediated resistance to TNF would rely on the expression of cIAP2. We next tested whether expression of IKK2-KD modulated the composition of the protein complexes recruited to TNF or TRAIL receptors (Figure 5d). Although ligand affinity precipitates of TNF in control and in NF-κB-inhibited HaCaT cells contained comparable amounts of TNF-R1, TNF receptor-associated death domain, receptor-interacting protein, TRAF2, and cIAP1, the amount of cIAP2 present in the TNF receptor complex of IKK2-KD-expressing cells was drastically reduced. Moreover, the known TRAIL DISC components TRAIL-R2, FADD, and caspase-8 were unchanged in the presence of IKK2-KD. To test whether cIAP2 is indeed crucial to maintain TNF resistance, the expression of cIAP2 was knocked down by using stably expressed siRNA. Although expression of siRNA1 efficiently downregulated cIAP2 (Figure 6a) compared to control-infected cells, or cells infected with the less efficient siRNA2, modulation of cIAP2 was not sufficient to sensitize HaCaT cells to TNF-induced apoptosis (Figure 6b). It could be argued that even though significant knockdown of cIAP2 protein was achieved, the extent of the downregulation was still not sufficient to sensitize cells to TNF. We therefore sought to test the hypothesis that cIAP2 expression is sufficient to make cells resistant to TNF by overexpressing cIAP2 in TNF-sensitive IKK2-KD cells, utilizing a 4-hydroxytamoxifen (4-HT)-inducible construct for cIAP2 that we have developed recently. We therefore performed sequential transduction of parental HaCaT cells with constructs for the 4-HT-responsive ER Gal4 VP16 transcriptional activator, the Gal4-dependent cIAP2 construct, and the retroviral IKK2-KD construct. Addition of 4-HT to these cells rapidly induced cIAP2 in a dose-dependent manner (Figure 7a), and maximal induction was obtained with 50–100 nM 4-HT within 4–24 hours. For the subsequent biochemical characterization and the apoptosis sensitivity studies, a range of 10–100 nM 4-HT was used. IKK2-KD-expressing HaCaT cells carrying the inducible cIAP2 construct in the absence of the 4-HT-responsive VP16 transactivation domain (GEV16) were highly sensitive to TNF-induced cell death independent of 4-HT (data not shown). Expression of GEV16 alone slightly increased TNF apoptosis resistance in response to 4-HT (Figure 7b, left panel). The induced upregulation of cIAP2 increased relative TNF apoptosis resistance when compared to the effect of 4-HT in GEV16-infected cells alone (Figure 7b). However, higher cIAP2 expression levels achieved by using increased concentrations of 4-HT up to 100 nM (compare Figure 7a) did not further increase resistance to TNF (data not shown), thereby excluding the possibility that the incomplete protection to TNF-induced apoptosis might be due to insufficient expression of cIAP2. In contrast, control cells with intact NF-κB activation remained completely apoptosis-resistant using TNF concentrations of up to 1 μg ml−1 (Figure 7b and data not shown).
Figure 5Analysis of NF-κB-dependent pro- and anti-apoptotic molecules in HaCaT keratinocytes. (a and b) Control HaCaT cells and HaCaT keratinocytes expressing either IKK2-KD or IKK2-EE were examined for mRNA expression of a panel of potential NF-κB-dependent target genes using multiprobe RNase protection assays. Steady-state levels of the indicated genes were determined by normalizing against the expression of the L32 housekeeping gene as assessed by PhosphoImager densitometry. (c) Steady-state protein expression of the indicated molecules was assessed by western blot analysis. Total cellular lysates were analyzed for the indicated molecules. Molecular weight markers are shown on the left. Asterisks indicate nonspecific band detected by the respective Abs. (d) TNF and TRAIL receptor signalling complexes in control or NF-κB-inhibited cells. HaCaT keratinocytes were stimulated with 2.5 μg ml−1 TNF for 15 minutes or 2.5 μg ml−1 TRAIL for 30 minutes, and subsequently receptor complexes were precipitated by ligand affinity precipitation using HF-TRAIL or HF-TNF, respectively, as described in Materials and Methods. Precipitation of TNF and TRAIL receptors following lysis (−) served as internal specificity control when compared to ligand affinity precipitates (+). Equal amounts of total cellular lysates (lysates) or ligand affinity precipitates (IP) were subsequently analyzed by western blotting for the indicated molecules.
Figure 6Downregulation of cIAP2 using siRNA expression. (a) Steady-state protein levels of cIAP2, cIAP1, and tubulin were assessed by western blot analysis. Molecular weight markers are indicated on the left. Asterisks indicate a nonspecific band detected by the respective Abs. (b) Downregulation of cIAP2 is not sufficient to sensitize HaCaT keratinocytes to TNF-induced cell death. Dose-dependent induction of TNF-mediated cell death in stable cIAP2 siRNA-expressing keratinocytes. Pools of cells infected with either control vector, partially effective siRNA to cIAP2 (siRNA2), or effective siRNA (siRNA1) were left untreated or stimulated with increasing concentrations of TNF for 16–24 hours, harvested, and examined for the induction of cell death using crystal violet assay as indicated in Materials and Methods. The mean ± SD of a total of two independent experiments is shown.
Figure 7Inducible upregulation of cIAP2 in TNF-sensitive IKK2-KD-expressing HaCaT keratinocytes is not sufficient to restore TNF apoptosis resistance. (a) Inducible expression of cIAP2. Cells were transduced with a 4-HT-inducible lentiviral construct carrying cIAP2 and subsequently incubated in the presence or absence of the indicated concentration of 4-HT for 4 (upper panel) and 24 hours (lower panel). Cells were then compared for protein expression of cIAP2, cIAP1, and tubulin by western blotting. Vector-infected (Vector) and IKK2-KD-infected cells (IKK2-KD) served as control for cIAP2 expression. Asterisks indicate the nonspecific band detected by the respective Abs. (b) Inducible upregulation of cIAP2 is not sufficient to restore resistance to TNF-induced cell death. Pools of cells infected with the inducible lentiviral construct together with GEV16 (GEV16/cIAP2) or the GEV16 construct alone (GEV16) were infected with control vector or IKK2-KD and subsequently analyzed for TNF-induced cell death in the presence or absence of 20 nM 4-HT. Cells were harvested 24 hours later and viability was assessed by crystal violet staining. The mean±SD of a total of two independent experiments is shown. (c) TNF receptor signalling complexes in control and NF-κB-inhibited HaCaT keratinocytes in the presence and absence of cIAP2. HaCaT keratinocytes transduced with the inducible lentiviral construct together with GEV16 and followed by transduction with IKK2-KD or control vector as indicated in (b) were treated with 4-HT (50 nM) or diluent alone for 4 hours. Cells were subsequently stimulated with 2.5 μg ml−1 TNF for 15 minutes, and TNF receptor complexes were precipitated as described in Materials and Methods. Precipitation of TNF receptors following lysis (−) served as an internal specificity control when compared to ligand affinity precipitates (+). Equal amounts of total cellular lysates (lysates) or ligand affinity precipitates (IP) were subsequently analyzed by western blotting for the indicated molecules. Asterisks indicate nonspecific band detected by the respective Abs. Molecular weights are indicated on the left.
To investigate the function of cIAP2 for TNF signalling in HaCaT keratinocytes in more detail, TNF receptor complexes were precipitated in NF-κB-inhibited cells inducibly expressing cIAP2. In line with the findings shown in Figure 5d and previous reports (
Fas-associated death domain protein and caspase-8 are not recruited to the tumor necrosis factor receptor 1 signaling complex during tumor necrosis factor-induced apoptosis.
), TNF receptor-associated death domain, receptor-interacting protein, and TRAF2 were found to be recruited to the TNF receptor complex irrespective of the presence or absence of cIAP2 (Figure 7c). In addition, cIAP1 and cIAP2 were specifically recruited to the complex, and cIAP2 levels in the complex were increased in cells induced with 4-HT. Of note, the increase of cIAP2 in the TNF complex was less pronounced than the increase detected in total cellular lysates (Figure 7c), suggesting that the affinity of cIAP2 for the complex is weak. Alternatively, other known or unknown adaptor proteins such as TRAF2 that are able to bind cIAP2 may be a limiting factor in this context. Future studies are required to delineate this point in detail. Finally, inducing the expression of cIAP2 did not alter recruitment or post-translational modification of any of the other proteins examined in the TNF receptor complex (Figure 7c, lanes 2, 4, 6, and 8). Our data show that although cIAP2 is induced by TNF, surprisingly, it is unlikely to play a key role in the resistance of HaCaT keratinocytes to TNF, raising the important question why cIAP2 is induced by TNF. Our studies strongly suggest that additional NF-κB-dependent factors as yet unidentified are required to preserve TNF resistance in HaCaT keratinocytes and that the search for additional factors is therefore warranted. The inducible lentiviral experimental system described in this report will prove invaluable to study the impact of additional NF-κB-dependent candidate molecules in the future.
Discussion
TNF exerts pleiotropic functions in almost all organ systems in a cell type-specific manner. In contrast to other death ligands such as CD95L or TRAIL, TNF is a major proinflammatory mediator, with the additional option to induce apoptosis. Both functions of TNF have been widely studied, and the inhibition of the proinflammatory effects of TNF by a variety of inhibitory agents is already used with marked success for patients requiring treatment of inflammatory diseases such as rheumatoid arthritis, inflammatory bowel disease, or psoriasis (
). As elimination of keratinocytes by apoptosis might severely impact the cytokine-mediated inflammatory responses of the skin, the understanding of TNF apoptosis sensitivity is of central importance for the understanding of TNF-induced skin inflammation. Over the past decade, it has been widely accepted that NF-κB proteins (such as RelA/p65, p50, and RelB) are involved in the maintenance of apoptosis resistance to a multitude of stimuli (
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
). In striking contrast, we have observed a dramatic sensitization to TNF-mediated apoptosis when NF-κB activation is inhibited by expression of a dominant-negative IKK2 mutant. These data are in line with a number of reports that have shown that the death-inducing function of TNF is masked whenever the activation of NF-κB by TNF is possible (for review, see
), we have investigated the regulatory mechanisms relevant for TNF-induced apoptosis. TNF exerts its diverse biological effects through two different receptors, and only TNF-R1 contains an intracellular death domain (
). In human HaCaT keratinocytes, both TNF receptors are expressed at low levels on the cell surface. However, there is now a broad body of evidence that TNF-induced apoptosis is largely mediated via TNF-R1 (
). We excluded that interference with NF-κB activation leads to modification of surface receptor expression. Our data showed that expression of TNF-R1 and TNF-R2 is rather decreased in IKK2-KD-expressing HaCaT cells, despite the dramatic increase in TNF sensitivity, and this change in sensitivity is in line with data using a transdominant negative form of IκBα (
). Thus, modulation of cell-surface TNF receptor expression is clearly not the cause for increased TNF apoptosis sensitivity when NF-κB activation is inhibited, despite the fact that our experimental system is unable to distinguish constitutive versus induced NF-κB activity and their differential role in sensitivity to TNF.
Our kinetic biochemical studies revealed that TNF-induced initiator caspase cleavage is blocked in HaCaT keratinocytes. Upon inhibition of NF-κB, TNF induces detectable initiator caspase activation within 4–6 hours. Animal studies have indicated that the sensitivity to TNF-induced apoptosis is regulated by cFLIP (
). Our data show that cFLIPL and cFLIPS are induced by TNF but not TRAIL within 4–6 hours in HaCaT keratinocytes, in line with genome-wide screening data in primary keratinocytes (
). This TNF-dependent induction of cFLIP was absent in NF-κB-inhibited cells, whereas basal cFLIP levels remained unchanged. These data confirm that TNF-induced NF-κB activation regulates cFLIP expression (
) and, despite similar early activation of NF-κB by TRAIL and TNF (compare Figure 1b), TNF appears to be a more potent inducer of cFLIP, indicating that signalling pathways other than NF-κB activated by TNF might be crucial for sustained gene induction. Because cFLIP is a short-lived protein (
Inhibition of death receptor-mediated gene induction by a cycloheximide-sensitive factor occurs at the level of or upstream of Fas-associated death domain protein (FADD).
), our data suggest that additional signals might be required for increasing cFLIP levels, such as mRNA stability or protein degradation. Interestingly, cFLIPL cleavage to p43 is obtained when NF-κB activation in response to TNF is blocked (Figure 4). These results indicate that whenever NF-κB is inhibited, an intracellular complex that allows cFLIP and/or caspase-8 activity can form and that it is not formed when TNF-induced NF-κB activation can proceed.
We showed, in line with many other results, that high levels of cFLIP are sufficient to protect NF-κB-inhibited cells against TNF-induced cell death. Thus, cFLIP might confer TNF resistance to HaCaT keratinocytes in line with data in primary keratinocytes (
). Further depletion of this low level of steady-state cFLIP further sensitizes cells to TRAIL but fails to sensitize cells to TNF-induced apoptosis. Thus, our data reveal a striking difference for cFLIP apoptosis protection in TRAIL- versus TNF-induced apoptosis. It will be interesting to examine additional signals activated by TRAIL and TNF receptors in keratinocytes, such as the activation of NF-κB and c-Jun-N-terminal kinase. This might be particularly interesting in the context of high or low levels of cFLIP, which may not only alter caspase-8 cleavage but may rather modulate activation of NF-κB selectively by TRAIL but not by TNF, as indicated recently (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
). Whether these signalling pathways indeed play a role in protection against apoptosis or whether they may alter gene induction by these ligands needs to be determined in future studies. Notably, NF-κB-dependent induction of cFLIP is counteracted by degradation, recently shown to be mediated by c-Jun-N-terminal kinase-mediated activation of the E3 ubiquitin ligase Itch (
). Although our study clearly demonstrates that cFLIP is not the critical regulator of TNF apoptosis resistance in human skin, future studies including organ-specific knockout models for cFLIP (
) will have to clarify the physiological importance of cFLIP deficiency for TNF and TRAIL apoptosis resistance as well as gene induction in the skin in vivo.
Current models —based on the concerted overexpression of the NF-κB target molecules TRAF1, TRAF2, cIAP1, and cIAP2—suggest that these molecules are relevant for TNF-mediated apoptosis (
). Thus, the exact role of cIAP1 or cIAP2 in TNF-induced cell death is still unclear to date. To investigate these contradictory findings in our experimental system, we examined the expression of a larger panel of potential NF-κB target genes including these four genes. Interestingly, only the cIAP2 expression level was significantly modulated at the mRNA and protein levels in TNF-sensitive HaCaT keratinocytes in which basal and death ligand-mediated NF-κB activity was blocked. Thus, cIAP2 represented a likely candidate for the regulation of TNF resistance in keratinocytes. To further test the relevance of cIAP2, we subsequently performed siRNA studies to selectively interfere with cIAP2 expression. However, these studies did not demonstrate sensitization of HaCaT keratinocytes to TNF-induced apoptosis, despite effective downregulation of cIAP2 at the mRNA and protein level. cIAP2 is induced dramatically by TNF in a number of cell types (
) as well as in HaCaT keratinocytes within 4–6 hours (Figure S2c). However, our data demonstrate that cIAP2 is not crucial for the maintenance of TNF resistance, at least in HaCaT keratinocytes. To rule out that siRNA effectiveness was not high enough for a decrease of cIAP2 mRNA levels and to circumvent potential experimental problems that could arise during clonal selection, we employed a 4-HT-inducible system to test the impact of cIAP2 expression in the context of TNF-sensitive, NF-κB-inhibited HaCaT cells (Figure 7). These findings clearly demonstrated that cIAP2 is not sufficient to maintain TNF resistance in HaCaT cells, although we were able to detect a moderate dose-dependent protection by cIAP2 expression when compared to control infected cells (compare Figure 7b). This protective effect could, however, not be further increased by higher expression levels of cIAP2 (compare Figure 7a), suggesting that cIAP2 is not critical for the modulation of TNF-induced apoptosis. It has been suggested previously that compensatory expression of cIAP1 or XIAP might explain the largely absent phenotype in XIAP knockout mice (
), we also studied whether expression of IKK2-KD with consequent loss of cIAP2 protein expression altered expression of cIAP1. However, our data excluded that compensatory expression of cIAP1 could explain TNF-induced cell death in IKK2-KD-expressing HaCaT cells (compare Figures 6 and 7). Future studies will hopefully reveal why cIAP2 is such a highly regulated protein in human keratinocytes and why cIAP2 is not sufficient to maintain TNF apoptosis resistance. Our study thus highlights an interesting conundrum and suggests a more complex role for cIAP2 than previously thought.
Our experimental system has certain limitations. First, it has been suggested that NF-κB signalling is not identical when comparing HaCaT keratinocytes to primary keratinocytes. Thus, future comparative studies with primary cells will prove to be highly interesting with respect to TNF apoptosis resistance (
). Moreover, there might be dynamic changes of NF-κB target genes that we are unable to detect owing to the fact that stable expression of IKK2-KD downregulates constitutive as well as TNF-induced target genes (compare Figure S1c). Second, the fact that inducible expression of IKK2-KD or IKK2-EE mutants in HaCaT keratinocytes recapitulates our findings with stable expression of IKK2 mutants is indicative of the physiological relevance of our data (Figure S3). Nonetheless, future studies will have to investigate what other known or unknown target genes of NF-κB are critical to maintain TNF apoptosis resistance other than cIAP2. Most likely, these studies may require a genome-wide approach or the use of inducible abrogation of NF-κB function, as performed previously in a mouse model or as feasible with inducible IKK mutants (
TNF- or TRAIL-induced NF-κB activation in HaCaT keratinocytes (a) Intentional overexposure of the blot shown in figure 1a. (b) Supershift analysis identifies p65/p50 heterodimers and p50/p50 homodimers as the major DNA binding complexes in HaCaT keratinocytes following death ligand stimulation (TRAIL, 1 μg/ml, 1 h). Supershift EMSA were performed as described previously (
HaCaT cells (control) or cells stimulated with TRAIL for 1h (TRAIL) were assayed for the presence of p65, c-Rel, RelB, p50, and p52 containing complexes. Shown are intentionally overexposed autoradiographies of 24 hours (for TRAIL stimulated cells) as well as 50 hrs (for control cells) to allow comparison. (c) IKK2-KD inhibits induction of TNF- and TRAIL-induced NF-κB DNA binding up to 6 hrs after stimulation. Nuclear extracts were analyzed for κB-specific DNA binding by EMSA. The positions of p65/p50 heterodimers or p50 homodimers are indicated.
As cIAP2 is not responsible for apoptosis resistance to TNF, what could be its physiological role in the skin? Recent reports have suggested that cIAP2 might be more important for proinflammatory signals initiated by other surface receptors such as Toll-like receptors (
) In particular, cIAP2 knockout cells that do not change sensitivity to CD95L-induced apoptosis have a markedly increased sensitivity to lipopolysaccharide-induced apoptosis, but do not show any overt phenotype under physiological conditions (
). Because contamination of lipopolysaccharide as the cause of TNF-induced cell death in our preparations is highly unlikely (Figure 3b), our data clearly indicate cell type as well as stimulus-specific differences that might dramatically alter the outcome of TNF receptor triggering in different organ systems. Our results support the hypothesis that cIAP1 or cIAP2 may have a role in the amplification of proinflammatory pathways that secondarily activate an NF-κB-dependent anti-apoptotic response rather than acting as direct anti-apoptotic caspase-inhibitory molecules. Future studies examining interaction partners of cIAP2 following TNF stimulation in keratinocytes are thus required to delineate these points in more detail. To this end, post-translational modifications of cIAP1 or cIAP2 might be crucial for their signalling capabilities in the context of TNF receptor triggering (
). The fact that cIAP2 is not sufficient to maintain apoptosis resistance to TNF in NF-κB-inhibited cells may thus stimulate future studies using the experimental system described in this report that will investigate the contribution of other NF-κB-regulated molecules for the protection against TNF-mediated apoptosis.
Materials and Methods
Materials
The following antibodies (Abs) were used: Abs to caspase-8 (C15), cFLIP (NF-6; Alexis, San Diego, CA), FADD (Transduction Laboratories, San Diego, CA), IκBα (C-21), TNF-R1 (H-5), TRAF2 (C-20; from Santa Cruz Biotechnology Inc., Santa Cruz, CA), CPP32 (kindly provided by D.W. Nicholson, Merck Frost, QC, Canada), and RIP-1 and TRADD (Becton Dickinson, Heidelberg, Germany). Abs to IKK2 (Ab 2684) were from New England Biolabs (Boston, MA), Abs to TRAIL-R2 (anti-DR5) were from Sigma (Saint Louis, MO), and Abs to cIAP2 were from R&D Systems Inc. (Minneapolis, MN). mAbs to TNF-R1 (H398) and a rabbit antiserum specific for TNF-R2 for cell-surface stainings were generously provided by H. Wajant (Institute for Molecular Medicine, University of Würzburg, Germany). Rat anti-cIAP1 mAbs were generated as described recently (
). For expression of His-FLAG-TRAIL (HF-TRAIL) and His-FLAG-TNF (HF-TNF), the complementary DNAs encoding the extracellular domain of TRAIL (aa95–281) or TNF (aa78–233) were N-terminally fused to a FLAG tag (DYKDDDDK) by PCR cloning and ligated into the BamHI- and NotI-restricted pQE32 vector (Qiagen, Hilden, Germany). The proteins were expressed in Escherichia coli M15 (pREP4) (Qiagen) at 18°C overnight and purified using a standard Ni-NTA column (Qiagen). Protein purity (>95%) and integrity were verified by SDS-PAGE and Coomassie staining. Bioactivity of HF-TRAIL and HF-TNF was confirmed by apoptosis assays in Jurkat J16 and U937 cells, respectively. Ligand-mediated cell death was completely blocked by addition of either soluble TRAIL-R2-Fc protein or TNF-R2-Fc protein, respectively. Recombinant leucine zipper TRAIL was generated as described (
). Horseradish peroxidase -conjugated donkey anti-rabbit and goat anti-mouse IgG Abs were from Pharmingen (Hamburg, Germany), and horseradish peroxidase-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, and IgG1κ were obtained from Southern Biotechnology Associates (Birmingham, AL). TRAIL-R1 (HS 101), TRAIL-R2 (HS 201), TRAIL-R3 (HS 301), and TRAIL-R4 (HS 402) mAbs for FACScan analysis of surface receptor expression were used as described (
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
) and are available from Alexis (San Diego, CA). Recombinant human TNF was obtained from Strathmann Biotech (Hannover, Germany). 4-HT was from Sigma-Aldrich (Steinheim, Germany). TNF-R2-Fc was provided by Wyeth Pharma (Münster, Germany).
Cell culture
The spontaneously transformed keratinocyte line HaCaT was kindly provided by Dr N. Fusenig (DKFZ, Heidelberg, Germany) and cultured as described (
Activation of NF-kappa B via the Ikappa B kinase complex is both essential and sufficient for proinflammatory gene expression in primary endothelial cells.
) was used for infection of HaCaT keratinocytes. Briefly, the amphotrophic producer cell line ϕNX was transfected with 10 μg of the retroviral vectors by calcium phosphate precipitation. To select transfected producer cells, 1 μg ml−1 puromycin (Sigma, Taufkirchen, Germany) or 0.5 μg ml−1 zeocin was added to the culture medium for 7–14 days to obtain >95% green fluorescent protein -positive producer cells. Cell culture supernatants containing viral particles were generated by incubation of producer cells with HaCaT medium (DMEM containing 10% fetal calf serum) overnight. Following filtration (45 μm filter; Schleicher & Schuell, Dassel, Germany), culture supernatant was added to HaCaT cells seeded in six-well plates 24 hours earlier in the presence of 1 μg ml−1 polybrene. HaCaT cells were centrifuged for 3 hours at 21°C, and viral particle containing supernatant was subsequently replaced by fresh medium. After 10–14 days recovery of bulk-infected cultures, FACS analysis for green fluorescent protein expression (data not shown) and western blot analysis (Figure 1) was performed on expanded polyclonal cells to confirm ectopic expression of the respective molecules. For sequential transduction with IKK2-KD and cFLIPL, cells were first infected with cFLIPL or control vector exactly as described recently (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
) and subsequently infected with either IKK2-KD or the respective empty construct. Aliquots of cells were used for the experiments between passages 1 and 10 postinfection for cytotoxicity assays and biochemical characterization.
Stable siRNA expression
We used stable expression of siRNA as published previously (
). For generation of the constructs, complementary DNA 64-mer oligomers containing either cIAP2 or cFLIP targeting sequences (for cIAP2, nucleotide start position +316: 5′-gatccccGAAGCTACCTCTCAGCCTAttcaagaga TAGGCTGAG AGGTAGCTTCtttttggaaa-3′ and nucleotide start position +466: 5′-gatccccGCCTTGATGAGAAGTTCCTttcaagagaAGGAA CTTCTCATCAAGGCtttttggaaa-3′ were used; for cFLIP, nucleotide start position +911: gatccccGGAGCAGGGACAAGTTACAttcaagagaTGTAA CTTGTCCCTGCTCCtttttggaaa was employed) were cloned into the pSuper.retro retroviral vector using HindIII and BglII restriction sites. The resulting vectors or control empty vector was transfected into the amphotrophic producer cell line exactly as described above (
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
). The retrovirus-containing supernatant was then used to infect HaCaT cells, and infected HaCaT cells were selected by puromycin (1 μg ml−1) for 3–14 days to obtain puromycin-resistant bulk-infected cultures for further analysis.
Lentiviral infection
To generate cells expressing 4-HT-inducible cIAP2, HaCaT cells were transduced with a lentivirus pF GEV16 Super PGKHygro, which expresses a Gal4 DNA binding domain fused to a mutant estrogen receptor and GEV16, and a lentivirus pF 5 × UAS hs cIAP2 W SV40 Puro, which expresses cIAP2 in a Gal4-dependent fashion (cIAP2). To generate lentiviral supernatants, 293T cells were transfected with 3 μg pMD2.G, 5 μg pMDlg/pRRE, and 2.5 μg pRSV-Rev of the lentiviral packaging vectors (
) together with the constructs described above. The supernatants were harvested 24 hours post-transfection, filtered (45 μm filter, Schleicher & Schuell), and concentrated by centrifugation (19,500 × g, 2 hours at 12°C). The concentrated virus was added to HaCaT cells with 5 μg ml−1 polybrene, and HaCaT cells were spin-infected. Stable cell lines were selected in hygromycin (100–150 μg ml−1), puromycin (1 μg ml−1), or both. Certain lines were subsequently infected with retrovirus containing IKK2-KD or control vector as described above and selected with zeocin (300 μg ml−1) for 14 days.
FACScan analysis
For surface staining of TRAIL receptors (TRAIL-R1 to TRAIL-R4) and TNF receptors (TNF-R1 and TNF-R2), cells were trypsinized and 2 × 105 cells were incubated with mAbs against TRAIL-R1 to TRAIL-R4, TNF-R1 and TNF-R2, or isotype-matched control IgG for 30 minutes followed by incubation with biotinylated goat anti-mouse or biotinylated goat anti-rabbit secondary Abs (BD Pharmingen, Heidelberg, Germany) and Cy5-phycoerythrin-labelled streptavidin (Caltag, Burlingame, CA) as described (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
). For all experiments, 104 cells were analyzed by FACScan (Becton Dickinson & Co, San Jose, CA).
Ligand affinity precipitation of receptor complexes
For the precipitation of the TRAIL DISC or the TNF complex, 5 × 106 HaCaT keratinocytes were used for each condition. Cells were washed once with DMEM medium at 37°C and subsequently incubated for the indicated time periods at 37°C in the presence of either 2.5 μg ml−1 HF-TRAIL pre-complexed with 5 μg ml−1 anti-FLAG M2 (Sigma, Taufkirchen, Germany) for 30 minutes or 2.5 μg/ml HF-TNF pre-complexed with 5 μg ml−1 anti-FLAG M2 for 15 minutes, or, for the unstimulated control, in the absence of ligands. Receptor complex formation was stopped by washing the monolayer four times with ice-cold phosphate-buffered saline. Cells were lysed on ice by the addition of 2 ml lysis buffer (30 mM Tris-HCl (pH 7.5) at 21°C, 120 mM NaCl, 10% glycerol, 1% Triton X-100, Complete protease inhibitor cocktail (Roche Molecular Diagnostics, Mannheim, Germany)). After 15 minutes of lysis, the lysates were centrifuged at 20,000 g for 30 minutes to remove cellular debris. A minor fraction of this lysate was used to control the input of the respective proteins. Receptor complexes were precipitated from the lysates by co-incubation with 40 μl of protein G beads (Roche) for 48 hours on an end-over-end shaker at 4°C. For precipitation of the non-stimulated receptors, 50 ng of either HF-TNF or HF-TRAIL pre-complexed with anti-FLAG M2 antibody as indicated above was added to the lysates prepared from non-stimulated cells to control protein association with non-stimulated receptor(s). Ligand affinity precipitates were washed five times with lysis buffer before the protein complexes were eluted from the dried beads by the addition of standard reducing sample buffer and boiling at 95°C. Subsequently, proteins were separated by SDS-PAGE on 4–12% NuPAGE gradient gels (Invitrogen, Karlsruhe, Germany) before detection of receptor complex components by western blot analysis.
Western blot analysis
Cell lysates were prepared essentially as described before (
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
). Subsequently, 5–20 μg of proteins was separated by SDS-PAGE on 4–12% NuPAGE Bis-Tris gradient gels (Invitrogen) and transferred to nitrocellulose or polyvinylidene difluoride membranes. Blocking of membranes and incubation with the indicated primary and appropriate secondary Abs were performed essentially as described elsewhere (
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
). The oligonucleotide probe used was a consensus NF-κB binding site derived from the κB element of the IL-2 promoter (“TCEdA>C”): 5′-CTAAATCCCCACTTTAGGGAGAACCAG-3′.
RNase protection assays
Total RNA was extracted using TRIzol reagent (Life Technologies, Gaithersburg, MD) and processed using Pharmingen's RNase protection assay system (hAPO-5, hAPO-3d) according to the manufacturer's instructions. Image data were collected with a PhosphoImager (Fuji Photo, Düsseldorf, Germany).
Apoptosis and cytotoxicity assays
Crystal violet staining of attached, living cells was performed 16–24 hours after stimulation with different concentrations of TRAIL or TNF in 96-well plates with the indicated concentrations as described (
) in duplicate or triplicate wells per condition. The optical density of control cultures was normalized as 100% to compare independent experiments. Subdiploid DNA content was analyzed as described by
. Briefly, cells from a 35 mm dish were cultured until they reached 70% confluence and were subsequently stimulated for 6 hours. Cells were then detached, washed with cold phosphate-buffered saline, and resuspended in buffer N (0.1% (w/v) sodium citrate, 0.1% (v/v) Triton X-100, 50 μg ml−1 propidium iodide). Cells were kept in the dark at 4°C for 48 hours and diploidy was measured by FACScan analysis.
Conflict of Interest
The authors state no conflict of interest.
ACKNOWLEDGMENTS
We thank P.H. Krammer for providing mAbs to caspase-8 (C-15) and cFLIP (NF-6), and D.W. Nicholson for providing CPP32 (caspase-3) antiserum. We are grateful to L. van Parijs for providing lentiviral packaging constructs and H. Wajant for providing reagents and for critical reading of the manuscript. M.R. Sprick is supported by an EMBO Long Term fellowship. We thank E. Horn, S. Pietzke, K. Garzinski, and M. Möckel for excellent technical assistance and B. Peters for statistical analysis. Part of this study was funded by grants from the Wilhelm-Sander-Stiftung (2000.092.2), Deutsche Krebshilfe (106849), Exzellenzförderung Sachsen-Anhalt (N2_OGU, TP6) and DFG (Le 953/5-1) to M. Leverkus. J. Silke is funded by an NHMRC career development grant.
Figure S1. TNF- or TRAIL-induced NF-κB activation in HaCaT keratinocytes.
Figure S2. TRAIL-R, TNF-R, and XIAP levels, as well as constitutive and TNF-induced expression of cIAP2, in IKK2 mutant-expressing HaCaT keratinocytes.
Figure S3. Inducible expression of IKK2 mutants recapitulates results using constitutive expression of IKK2 mutants.
REFERENCES
Ashkenazi A.
Targeting death and decoy receptors of the tumour-necrosis factor superfamily.
Activation of NF-kappa B via the Ikappa B kinase complex is both essential and sufficient for proinflammatory gene expression in primary endothelial cells.
Fas-associated death domain protein and caspase-8 are not recruited to the tumor necrosis factor receptor 1 signaling complex during tumor necrosis factor-induced apoptosis.
Proteasome inhibition results in TRAIL sensitization of primary keratinocytes by removing the resistance-mediating block of effector caspase maturation.
cFLIPL inhibits tumor necrosis factor-related apoptosis-inducing ligand-mediated NF-kappaB activation at the death-inducing signaling complex in human keratinocytes.
Inhibition of death receptor-mediated gene induction by a cycloheximide-sensitive factor occurs at the level of or upstream of Fas-associated death domain protein (FADD).
33862-8&id=gr1.jpg){kind=link}
33862-8&id=gr2.jpg){kind=link}
33862-8&id=gr3.jpg){kind=link}
33862-8&id=gr4.jpg){kind=link}
33862-8&id=gr5.jpg){kind=link}
33862-8&id=gr6.jpg){kind=link}
33862-8&id=gr7.jpg){kind=link}