QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

doi: 10.1242/10.1242/jcs.00151

This Article Summary Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Fan, C. Articles by Engelhardt, J. F. Search for Related Content PubMed PubMed Citation Articles by Fan, C. Articles by Engelhardt, J. F.

Temporal pattern of NFB activation influences apoptotic cell fate in a stimuli-dependent fashion

¹ Molecular Biology Graduate Program, University of Iowa College of Medicine, Iowa City, Iowa, 52242 USA
² The Center for Gene Therapy, University of Iowa College of Medicine, Iowa City, Iowa, 52242 USA
³ Department of Anatomy and Cell Biology, University of Iowa College of Medicine, Iowa City, Iowa, 52242 USA

^* Author for correspondence (e-mail: john-engelhardt{at}uiowa.edu)

Accepted 4 September 2002

	Summary

Top Summary Introduction Materials and Methods Results Discussion References

 
The transcription factor NFB is a critical immediate early response gene involved in modulating cellular responses and apoptosis following diverse environmental injuries. The activation of NFB is widely accepted to play an anti-apoptotic role in cellular responses to injury. Hence, enhancing NFB activation in the setting of injury has been proposed as one potential therapeutic approach to environmental injuries. To this end, we constructed a recombinant adenoviral vector (Ad.IBAS) expressing antisense IB mRNA that is capable of augmenting NFB activation prior to and following four types of cellular injury [TNF-, UV, hypoxia/reoxygenation (H/R) or pervanadate treatment]. Biochemical and functional analyses of NFB activation pathways for these injuries demonstrated two categories involving either serine (S32/36) phosphorylation (TNF-, UV) or tyrosine (Y42) phosphorylation (H/R or PV) of IB. We hypothesized that activation of NFB prior to injury using antisense IB mRNA would reduce apoptosis. As anticipated, recombinant adenoviral IB phosphorylation mutants (Ad.IBS32/36A or Ad.IBY42F) preferentially reduced NFB activation and enhanced apoptosis following injuries associated with either serine or tyrosine phosphorylation of IB, respectively. These studies demonstrate for the first time that an IBY42F mutant can effectively modulate NFB-mediated apoptosis in an injury-context-dependent manner. Interestingly, constitutive activation of NFB following Ad.IBAS infection reduced apoptosis only following injuries associated with IB Y42, but not S32/36, phosphorylation. These findings demonstrate that the temporal regulation of NFB and the apoptotic consequences of this activation are differentially influenced by the pathway mediating NFB activation. They also provide new insight into the therapeutic potential and limitations of modulating NFB for environmental injuries such as ischemia/reperfusion and pro-inflammatory diseases.

Key words: Antisense inhibition, NFB activation, Signal transduction, Apoptosis

	Introduction

Top Summary Introduction Materials and Methods Results Discussion References

 
The NFB transcription factor family consists of five members called p50, p52, p65 (RelA), c-Rel and RelB, all of which can form various homo- and heterodimers (Karin and Ben-Neriah, 2000). NFB is normally sequestered in the cytoplasm by proteins of the IB family, IB, ß and . The predominant induced form of NFB, a p50 and p65 heterodimer, translocates to the nucleus upon activation. Common activating stimuli of the NFB pathway include interleukin-1 (IL-1), tumor necrosis factor (TNF-), lipopolysaccharide (LPS), hypoxia/reoxygenation and phorbol myristate acetate (PMA) (Piette et al., 1997).

The most commonly studied pathway of NFB activation involves the phosphorylation of IB on serine residues 32 and 36 by a recently identified IB kinase (IKK) complex, which is composed of three subunits, , ß and (Zandi et al., 1997). IKK and IKKß are the two catalytic subunits in the complex. Phosphorylation of IB by the IKK complex leads to ubiquitination and degradation of IB, which unmasks a nuclear targeting sequence on the NFB molecule. This promotes the translocation of NFB from the cytoplasm to the nucleus where it becomes an active transcription factor (Scherer et al., 1995). Regulators of IKK activity include NFB-inducing kinase (NIK), MEKK1 and NFB-activating kinase (NAK). These regulators impart signal-specific activation of NFB through the IKK complex. TNF- and IL-1 induction of NFB is mediated by NIK phosphorylation of IKK (Hirano et al., 1996). By contrast, MEKK1, which is part of the Jun N-terminal kinase/stress-activated protein kinase pathway, can induce the activation of both IKK and IKKß (Lee et al., 1997). Most recently, NAK has been shown to directly phosphorylate IKKß (Tojima et al., 2000).

An alternative, but less studied mechanism for NFB activation, involves tyrosine phosphorylation of IB leading to dissociation of NFB from IB without proteolytic degradation (Fan et al., 1999). It was discovered that stimulation of Jurkat T-cells with either pervanadate (a protein phosphatase inhibitor) or hypoxia/reoxygenation leads to phosphorylation of IB at tyrosine residue 42 and subsequent NFB nuclear translocation without IB proteolysis (Imbert et al., 1996). Similarly, in a neuron cell model, nerve growth factor treatment leads to NFB activation through IB tyrosine phosphorylation independent of proteolytic degradation (Bui et al., 2001). In support of this mechanism, our group recently demonstrated that NFB is activated without concomitant degradation of IB in a murine model of liver ischemia/reperfusion (I/R) (Zwacka et al., 1998). Furthermore, tyrosine phosphorylation of IB was increased following I/R injury in the liver, suggesting that this degradation-independent pathway of NFB activation is important in I/R injury (Zwacka et al., 1998). Although these previous studies conclusively demonstrate that Y42 of IB is phosphorylated following hypoxia/reoxygenation, the functional importance of this phosphorylation event in regulating NFB transcriptional activation has yet to be proven using a dominant-negative IB Y42 phosphorylation mutant. In the case of both I/R and hypoxia, these findings have implicated a potentially unique tyrosine kinase in the phosphorylation of IB. Furthermore, this kinase appears to be distinct from those activated by TNF-, IL-1, LPS or PMA. The mechanism that leads to IB tyrosine phosphorylation and NFB activation remains unclear; nonetheless, several key players have been identified. C-src, an osteoclast regulatory protein, has been shown to associate with IB and phosphorylate IB on tyrosine 42 in murine bone marrow macrophages (BMMs) in a cytokine-specific manner (Abu-Amer et al., 1998). Also, p85, the regulatory subunit of PI 3-kinase, specifically associates with tyrosine-phosphorylated IB through its Src homology domains both in vitro and in vivo after stimulation of T cells with pervanadate (Beraud et al., 1999). These studies suggest that PI 3-kinase might be a candidate for regulating tyrosine phosphorylation of IB.

As a pro-inflammatory transcription factor, NFB activation in the initial phase of I/R injury may trigger upregulation of cytokines, including TNF- and IL-1, and adhesion molecules, such as ICAM-1, which can mediate the ensuing subacute inflammatory response (Fan et al., 1999). However, in addition to its pro-inflammatory action, NFB also plays a protective role in acute cellular stress responses that inhibit apoptosis following TNF- treatment (Liu et al., 1996) or ionizing irradiation (Wang et al., 1996). Furthermore, it has been shown in a two-thirds hepatectomy model that inhibition of NFB by overexpression of a dominant-negative mutant form of IB increases apoptosis and liver dysfunction (Iimuro et al., 1998). It is currently unclear how the detrimental proinflammatory and the beneficial anti-apoptotic effects of NFB activation are regulated in the setting of various types of injury.

In an effort to design therapeutic approaches to enhance NFB activation and reduce apoptosis following cellular injuries such as ischemia/reperfusion, we have generated and characterized a novel adenoviral vector that expresses an antisense mRNA of IB and enhances NFB activation. Since two distinct pathways of regulating NFB through IB serine or tyrosine phosphorylation have been identified, we have compared how altering the temporal expression profile of NFB influences apoptosis following four types of injury that utilize either serine or tyrosine IB kinase activation pathways. Three recombinant adenoviral vectors expressing a serine mutant (Ad.IBS32/36A), tyrosine mutant (Ad.IBY42F) or antisense mRNA (Ad.IBAS) of IB were used to modulate NFB activation prior to injury, and the apoptotic outcomes were compared. Results from these studies suggest that the timing of NFB activation during the acute phases of injury can significantly influence apoptotic outcomes in an injury-stimuli-dependent manner. Interestingly, activating NFB prior to injury was anti-apoptotic only following stimuli dependent on tyrosine kinase activation of IB but not IKK-dependent serine phosphorylation of IB. These findings suggest that the temporal regulation of NFB following injury is an important feature that influences apoptotic cell fate in a manner which is also dependent on the pathway of NFB activation.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion References

 
Construction of recombinant adenoviruses
The IB(Y42F) cDNA was a kind gift from J. F. Peyron (Pasteur Institute, France). The plasmid was digested with HindIII to generate a fragment containing the entire coding region for the human IB cDNA. This HindIII fragment was bidirectionally cloned into the recombinant adenovirus proviral plasmid, pAd.CMV link, which contains the cytomegalovirus (CMV) enhancer/promoter and an SV40 poly-adenylation site for efficient expression of the transgenes (Davis and Wilson, 1996). Two proviral plasmids containing the sense and antisense orientation of the IB(Y42F) cDNA were generated, pAd.IB(Y42F) and pAd.IBAS, respectively. Recombinant viruses were then generated by co-transfection of an NheI-cut pAd plasmid, with ClaI-cut Ad5.GFP viral DNA. Following co-transfection, plates were overlaid with agar, and initial plaques were harvested for screening by Southern blotting. Clones containing cDNAs were replaqued two times, and viral stocks were purified by standard protocols (Davis and Wilson, 1996). The Ad.IB(S32/36A) dominant-negative mutant adenoviral construct was a kind gift from David Brenner (UNC) (Iimuro et al., 1998).

Cell culture, adenoviral infection and environmental injury
HeLa cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 100 µg/ml penicillin and streptomycin. HeLa cells were infected with various recombinant adenoviral constructs at a multiplicity of infection (moi) ranging from 1000 to 1500 particles/cell. Adenoviral infections were performed for 2 hours in DMEM without FBS, followed by the addition of an equal volume of DMEM with 20% FBS. Fresh media was applied to cells when virus was removed at 24 hours post-infection. Experiments were typically performed at 24-72 hours post-infection. Experimental stimuli used to induce NFB were performed by the protocols listed below.

UV irradiation
HeLa cells were washed with PBS three times and irradiated with UV-C at 50 J/m² in the absence of surface liquid. Fresh medium was quickly applied to the plates following irradiation. Cells were harvested 2 hours post-irradiation for EMSA and 6 hours post-irradiation for NFB activation luciferase assays. Control, mock-irradiated cells were treated in a similar fashion but were not exposed to UV.

TNF- treatment
Human recombinant TNF- (R&D systems, Minneapolis, MN) was diluted to 10 ng/ml in fresh medium just prior to use. TNF--containing medium was applied to HeLa cells at the time of stimulation and cells remained exposed to TNF- until they were harvested for analysis. Cells were harvested at 1 hour post-TNF- treatment for EMSA and 6 hours post-TNF- treatment for NFB activation luciferase assays. Control cells were fed with fresh medium without TNF-.

Pervanadate(PV) treatment
500 mM sodium orthovanadate was prepared fresh in water. 40 µl of sodium orthovanadate was then added to a mixture of 450 µl PBS and 10 µl 30% (w/w) H₂O₂. This solution was incubated for 5 minutes at room temperature prior to the addition of catalase (200 µg/ml final concentration). The final pervanadate solution (40 mM) was incubated for 5 minutes at room temperature to remove excess H₂O₂ and then immediately diluted in DMEM for application to cells. Cells were continuously exposed to pervanadate containing medium until cells were harvested for analysis. Harvesting took place at 2 hours post-pervanadate treatment for EMSA and at 6 hours post-pervanadate treatment for NFB activation luciferase assays. Control cells were fed fresh medium without pervanadate.

Hypoxia/reoxygenation (H/R) experiments
DMEM (no glucose, 0%FBS, 1%P/S) equilibrated in 5% CO₂/95% N₂ or 5% CO₂/95% O₂ was used as hypoxia medium and reoxygenation medium, respectively. HeLa cells were incubated with hypoxia medium in an airtight chamber equilibrated with 5% CO₂/95% N₂ at 37°C for 5 hours. After hypoxia was performed, the medium was replaced with reoxygenation medium and cells were incubated in a chamber flushed with 5% CO₂/95% O₂ at 37°C. Cells were harvested 3 hours after reoxygenation for EMSA and 6 hours after reoxygenation for NFB activation luciferase assays. Control cells were fed with fresh medium at the appropriate times but were exposed to 5% CO₂ in atmospheric oxygen.

Western blotting and electrophoretic mobility shift assays (EMSA)
Cytoplasmic and nuclear extracts were prepared by standard protocols (Andrews and Faller, 1991) and used for western blotting and EMSA, respectively. Briefly, 2x10⁶ HeLa cells were collected into 1.5 ml centrifuge tubes by blunt scraping with 1 ml PBS. Cells were pelleted for 1 minute and were then resuspended in 400 µl of cold buffer A (10 mM HEPES pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride). Samples were incubated on ice for 10 minutes, then vortexed for 10 seconds and finally pelleted by brief centrifugation (1 minute). The supernatant was saved as the cytoplasmic extract for western blotting, and the pellet was resuspended in 100 µl of storage buffer (20 mM HEPES, pH 7.9, 25% Glycerol, 420 mM NaCl, 1.5 mM MgCl₂, 0.21 mM EDTA, 0.5 mM dithiothreitol and 0.2 mM phenylmethylsulfonyl fluoride). Samples were then incubated on ice for another 20 minutes followed by centrifugation. Supernatants were collected for use as nuclear extracts. Protein concentrations were determined using a Bio-Rad protein assay (Bio-Rad, Hercules, CA). 5 µg of cytoplasmic extracts were resolved on a 10% SDS-PAGE and then transferred to nitrocellulose membranes. Following standard protocols (Zwacka et al., 1998), IB protein levels were determined by western blotting using monoclonal anti-IB, anti-IBß and anti-actin antibodies (Santa Cruz Biotech, Santa Cruz, California). EMSA analysis and supershift assays were performed using an NFB-specific oligonucleotide (Promega, Madison, WI). The sequence was as follows: 5'-AGTTGAGGGGACTTTCCCAGGC-3'. The double-stranded nucleotides were end-labeled with ³²P-ATP using T4 polynucleotide kinase. 5 µg of nuclear extract was used in each assay for NFB DNA binding using standard protocols (Zwacka et al., 1998). NFB antibodies used for supershift EMSA (anti-p50, anti-p52, anti-c-Rel, anti-RelB and anti-p65) were purchased from Santa Cruz Biotech.

Apoptosis assays
HeLa cells were infected with either Ad.BglII or Ad.IBAS for 24 hours at an moi of 1000 particles/cell. Both groups were treated with UV, TNF-, pervanadate and hypoxia/reoxygenation. In the TNF- treated group, HeLa cells were pretreated with proteasome inhibitor LLnL (40 µM final concentration) (Calbiochem, La Jolla, California) for 30 minutes prior to treatment to enhance apoptosis. In all experimental groups, cells were trypsinized 18 hours following treatment and stained with annexin V—FITC and propidium iodide (PI) for apoptotic analysis. Annexin-V—FITC binding was performed using a kit from Pharmingen (Palo Alto, CA) according to the manufacturer's protocol. Briefly, 1x10⁶ cells were washed twice with cold PBS and then resuspended in 1 ml 1xbinding buffer (10 mM Hepes, pH 7.4, 140 mM NaCl, 2.5 mM CaCl₂). 5 µl of Annexin-V—FITC (100 µg/ml) and 10 µl of propidium iodide (50 µg/ml) were then added to 100 µl cell suspensions. Cells were incubated for 15 minutes in darkness at room temperature. This was followed by the addition of 400 µl of 1xbinding buffer. Samples were then analyzed by FACS. Apoptotic cells were identified as an annexin-V-positive population. The percentage of apoptotic cells was calculated as the fraction of annexin-V-positive/PI-positive cells.

Luciferase indicator assays for NFB transcriptional activation
NFB transcriptional activity was assessed using an Ad.NFBLuc reporter vector. This construct contains the luciferase gene driven by four tandem copies of the NFB consensus sequence fused to a TATA-like promoter from the Herpes simplex virus thymidine kinase gene (Sanlioglu et al., 2001). HeLa cells were infected with Ad.NFBLuc 24 hours prior to treatment by UV, TNF-, pervanadate or H/R. 5 µg of total protein from each sample was used to perform luciferase assays. Luciferase activity was measured using a kit from Promega (Catalog No. E4030, Madison, WI).

IB phosphorylation assays
Tyrosine phosphorylation of IB was evaluated by immunoprecipitation from 200 µg cytoplasmic extracts in 300 µl RIPA buffer using 2 µg IB antibody (Santa Cruz Biotech, Santa Cruz, CA) overnight at 4°C. 30 µl of Protein-A agarose was then added and incubated for 4 hours at 4°C. The agarose beads were then washed four times with ice-cold RIPA buffer before they were resuspended in SDS-PAGE loading buffer and loaded onto a 10% SDS-PAGE for western blot analysis. The membrane was probed with an antibody that recognizes phospho-tyrosine residues (Santa Cruz) at a dilution of 1:1000 following the western protocol outlined above. Serine phosphorylation of IB was detected by standard western blotting of 5 µg cell lysate using a phosphospecific antibody (Santa Cruz Biotech, Santa Cruz, California), which recognized phospho-S32/S36 of IB.

	Results

Top Summary Introduction Materials and Methods Results Discussion References

 
UV and TNF- induce serine phosphorylation of IB, whereas pervanadate and H/R induce tyrosine phosphorylation
Phosphorylation of IB is a key step in the translocation of NFB to the nucleus. Phosphorylation of IB on serine residues 32 and 36 by the inhibitor B kinase (IKK) complex leads to ubiquitination and degradation of IB proteins following pro-inflammatory stimuli. By contrast, tyrosine phosphorylation of IB has been associated with the translocation of NFB to the nucleus without IB proteolytic degradation following I/R. With the goal of modulating NFB activity to enhance cell survival following I/R injury, we sought to develop model systems capable of modulating NFB activity through tyrosine phosphorylation of IB. Two such in vitro models including hypoxia/reoxygenation (H/R) and pervanadate treatment have been shown to activate NFB in T-cells with a correlated increase in Y42 phosphorylation of IB. In an effort to establish such a model for selective induction of NFB mediated by Y42 phosphorylation of IB, we felt it would be useful to establish a cell model system also capable of inducing NFB activation through the better characterized IKK-dependent S32/36 phosphorylation of IB. To this end, many cell lines were screened for their ability to induce NFB activation following four types of environmental injury, including UV, TNF-, pervanadate and H/R treatments, which are known to involve either serine or tyrosine phosphorylation of IB. HeLa cells were chosen for their ability to induce NFB DNA binding following each of these stimuli.

To experimentally establish that HeLa cells were capable of modulating NFB through these two independent serine or tyrosine IB kinase pathways, we evaluated the state of IB serine and tyrosine phosphorylation following the four types of environmental injury. Using a phosphospecific antibody that specifically recognizes the serine-phosphorylated form of IB (S32/36), western blot analysis demonstrated that both UV and TNF- treatments induced substantial levels of IB serine phosphorylation by 15 to 30 minutes (Fig. 1A). By contrast, no significant change in the baseline levels of serine phosphorylation was observed following pervanadate or H/R treatment. Using immunoprecipitation of IB followed by western blotting with an anti-phosphotyrosine antibody, we addressed the tyrosine-phosphorylated state of IB. Results from these studies demonstrated that IB tyrosine phosphorylation significantly increased at 15-30 minutes following pervanadate or H/R treatment (Fig. 1B). By contrast, no increase in tyrosine phosphorylation of IB was detected following UV or TNF- treatment. In summary, this data demonstrates that HeLa cells can be used as a cell model to study NFB activation mediated by either serine or tyrosine phosphorylation of IB.

View larger version (33K): [in this window] [in a new window]   Fig. 1. TNF- and UV treatment lead to serine phosphorylation of IB, whereas pervanadate and H/R induce tyrosine phosphorylation of IB. HeLa cells were treated with UV (50 J/m2), TNF- (10 ng/ml), pervanadate (50 µM) or H/R (5 hours hypoxia, 15 and 30 minutes reoxygenation) for the indicated times. Both untreated and treated samples were harvested for cytoplasmic extracts. (A) 5 µg of total protein was separated by 10% SDS-PAGE and transferred to a nitrocellulose membrane. Phosphorylation of IB at S32/36 was evaluated by a phosphospecific antibody. (B) 200 µg of total cytoplasmic protein was immunoprecipitated with anti-IB antibody followed by western blotting with an anti-phosphotyrosine antibody to detect tyrosine phosphorylation of IB. Samples analyzed were identical to those evaluated in A.

IB(S32/36A) and IB(Y42F) mutants selectively inhibit NFB activation in an injury-stimulus-dependent fashion
Although Y42 phosphorylation of IB has been associated with NFB activation following pervanadate and H/R treatments, functional evidence for the importance of this tyrosine phosphorylation event in the transcriptional activation of NFB using dominant-negative mutants of IB is still lacking. To this end, we generated a recombinant adenovirus encoding the IB(Y42F) mutant and evaluated its ability to inhibit transcriptional activation of an NFB-responsive luciferase reporter (also delivered via a recombinant adenovirus vector) following each of the various types of injury. As an important control for IKK-mediated pathways of NFB activation, a similar adenovirus vector encoding a dominant-negative IB(S32/36A) mutant (Iimuro et al., 1998) was also evaluated. Consistent with the finding that IB serine phosphorylation was most significantly induced by TNF- and UV, IB(S32/36A) expression significantly inhibited NFB transcriptional activation following TNF- (98.7±5.21%, P<0.001) and UV (74.2±9.1%, P<0.001) treatments (Fig. 2). By contrast, IB (S32/36A) expression did not significantly inhibit NFB activation following pervanadate (P=0.104) or H/R treatment (P=0.464) (Fig. 2). These data confirm that the predominant pathway controlling both TNF- and UV induction of NFB in HeLa cells occurs through serine phosphorylation of IB. Furthermore, they also demonstrate that serine phosphorylation of IB plays only a minor role in regulating NFB transcriptional activity following pervanadate and H/R stimuli.

View larger version (19K): [in this window] [in a new window]   Fig. 2. IB(S32/36A) inhibits NFB activation following UV and TNF- treatment. (A) HeLa cells were co-infected with Ad.IB(S32/36A) or Ad.BglII together with Ad.NFBLuc 24 hours before treatment. Cells were harvested 6 hours after UV (50 J/m2), TNF- (10 ng/ml), pervanadate (50 µM) and H/R (5 hours hypoxia, 6 hours reoxygenation) treatments, and whole cell extracts were normalized by total protein content and subjected to luciferase assays. NFB activity was determined by the relative luciferase activity (as light units). The relative NFB activation was calculated by deducting the mean NFB baseline activation for each vector group (in the absence of stimulation) from the corresponding individual values from the same vector group in the presence of stimulus. The relative NFB activation is plotted for each individual stimulus (±s.e.m., n=6). (B) Percent NFB inhibition by IB(S32/36A) for each experimental point was calculated using the following formula: percent inhibition=1—(relative activation of each Ad.IB(S32/36A) infected sample/mean relative activation of the Ad.BglII infected group). Results depict the mean (±s.e.m.) for n=6 independent data points in each group.

Using our newly constructed recombinant adenoviral vector expressing the Y42F mutant of IB, we next sought to confirm that tyrosine phosphorylation of IB was predominantly responsible for the transcriptional activation of NFB following pervanadate or H/R treatment. Consistent with the tyrosine phosphorylation data, IB(Y42F) expression significantly inhibited NFB transcriptional activation following pervanadate (71.6±6.3%, P<0.01) and H/R (73.7±10.7%, P<0.01) treatments (Fig. 3). By contrast, both TNF- and UV induction of NFB transcriptional activity was not significantly altered by expression of IB(Y42F) (P=0.359 and P=0.257, respectively). In summary, these data demonstrate that the IB(Y42F) mutant can selectively inhibit NFB transcriptional activation following pervanadate or H/R treatment. They also confirm that tyrosine phosphorylation of IB is the predominant mechanism of NFB transcriptional activation by these stimuli.

View larger version (19K): [in this window] [in a new window]   Fig. 3. IB(Y42F) inhibits NFB activation following pervanadate and H/R treatment. (A) HeLa cells were co-infected with Ad.IB(Y42F) or Ad.BglII together with Ad.NFBLuc 24 hours before treatment. Cells were treated with UV (50 J/m2), TNF-(10 ng/ml), pervanadate (50 µM) and H/R (5 hours hypoxia, 6 hours reoxygenation) then harvested 6 hours after treatment. Whole cell extracts were normalized by total protein content and subjected to luciferase assay. NFB activity was determined by the relative luciferase activity (as light units). The relative NFB activation was calculated by deducting the mean NFB baseline activation for each vector group (in the absence of stimulation) from the corresponding individual values from the same vector group in the presence of stimulus. The relative NFB activation is plotted for each individual stimulus (±s.e.m., n=6). (B) Percent NFB inhibition by IB(Y42F) for each experimental point was calculated using the following formula: percent inhibition=1—(relative activation of each Ad.IB(Y42F) infected sample/mean relative activation of the Ad.BglII infected group). Results depict the mean (±s.e.m.) for n=6 independent data points in each group.

Inhibition of NFB activation through either IB tyrosine or serine phosphorylation pathways stimulates apoptosis following environmental injuries
Although it is widely accepted that activation of NFB following cellular injury is anti-apoptotic, this phenomena has not been established for injuries that activate NFB via tyrosine phosphorylation of IB. The lack of information on this topic is probably due to the lack of efficient vectors capable of expressing high levels of a IB(Y42F) mutant in the majority of cells. However, since recombinant adenoviral vectors are capable of expressing high levels of IB mutants, we sought to conclusively address this issue by comparing the ability of either IB(S32/36A) or IB(Y42F) mutant to selectively enhance apoptosis following each of the previously described types of environmental injury (Fig. 4). Results from these experiments demonstrated that expression of the IB(S32/36A) mutant preferentially enhanced apoptosis following TNF- or UV treatment to a far greater extent than that seen following expression of the IB(Y42F) mutant (P<0.02). Thus inhibiting NFB activation with Ad.IB(S32/36A) has a stronger influence on apoptosis following UV or TNF- treatment when compared to Ad.IB(Y42F). By contrast, expression of the IB(Y42F) mutant preferentially enhanced apoptosis following H/R or pervanadate treatment to a greater extent than that seen following expression of the IB(S32/36A) mutant (P<0.03). The inhibition of NFB activation with Ad.IB(Y42F) had a stronger influence on apoptosis following pervanadate or H/R treatment compared with Ad.IB(S32/36A). Although this differential effect of tyrosine and serine mutants of IB was not absolute, these results do functionally substantiate findings for two independent pathways of NFB activation. More importantly, they also demonstrate that inhibition of NFB activation stimulates apoptosis regardless of the injury pathway responsible for NFB activation.

View larger version (20K): [in this window] [in a new window]   Fig. 4. Inhibition of either tyrosine or serine IB phosphorylation stimulates apoptosis in an injury-context specific fashion. HeLa cells were infected with Ad.BglII, Ad.IB(S32/36A) or Ad.IB(Y42F) at an moi of 1000 particles/cell for 24 hours. Cells were treated with UV (50 J/m2), TNF- (10 ng/ml), pervanadate (50 µM) or H/R (5 hours hypoxia, 18 hours reoxygenation) and then harvested 18 hours after treatment. Cells were then stained with annexin-V—FITC and propidium iodine. Results depict the mean (±s.e.m., n=3) percent of apoptotic cells as shown by FACS analysis. Ad.IB(S32/36A) more significantly increased apoptosis following UV or TNF- treatment, whereas Ad.IB(Y42F) preferentially increased apoptosis following pervanadate or H/R treatment. Paired t-test analysis was performed between Ad.BglII infected and Ad.IB(S32/36A) or Ad.IB(Y42F) infected samples for each stimulus, and a statistically significant difference is denoted by *(P<0.05) or (P<0.001). P-values for paired t-test analysis comparing Ad.IB(S32/36A) and Ad.IB(Y42F) infected samples for each stimulus are denoted above brackets.

Antisense inhibition of IB protein enhances NFB baseline activation
The anti-apoptotic nature of NFB suggests that augmenting NFB activation prior to injury might enhance the protective effects of this molecule. Indeed, such therapeutic strategies appear to be the basis for protective effects following ischemic preconditioning in the heart (Morgan et al., 1999). To this end, we sought to evaluate whether inhibition of IB protein synthesis using an antisense mRNA approach could reduce apoptosis and improve cell survival following the various types of environmental injuries. A recombinant adenovirus (Ad.IBAS) encoding IB cDNA in the reversed orientation was tested for its ability to inhibit IB protein expression and enhance NFB activation. As seen in Fig. 5A, infection of HeLa cells with Ad.IBAS significantly reduced steady-state levels of IB protein. Interestingly the steady-state level of IBß was increased in the presence of antisense IB mRNA. This probably reflects a compensatory mechanism induced by activation of NFB. These findings also demonstrated specificity of the antisense approach to inhibit IB expression. Despite the observed changes in both IB and IBß levels, no changes in the steady-state level of the actin internal control were observed. Ad.IBAS infection consequently led to a time-dependent increase in nuclear NFB DNA binding that was not observed in Ad.GFP-infected controls (Fig. 5B). EMSA supershift assay using antibodies to various NFB subunits (p50, p52, c-Rel, RelB, and p65) demonstrated that IBAS expression induced NFB complexes composed of p50/p65 heterodimers (Fig. 5C). These results demonstrate that inhibition of IB protein expression using an antisense mRNA approach significantly elevates baseline levels of NFB in the nucleus.

View larger version (37K): [in this window] [in a new window]   Fig. 5. Ad.IBAS inhibits IB protein expression and elevates baseline levels of NFB in the nucleus. HeLa cells were infected at an moi of 1000 particles/cell with Ad.IBAS or Ad.GFP. Cytoplasmic and nuclear extracts were prepared from cells collected at 24, 48 and 72 hours after infection. (A) 5 µg of cytoplasmic protein was evaluated by western blotting using anti-IB antibody, anti-IBß antibody or anti-actin antibody. Ad.IBAS infection led to a reduction in IB protein expression at all time points (lane 2-4) and an increase in IBß expression compared with Ad.GFP-infected controls (lane 1). Changes in IB expression were referenced to actin protein levels, which did not change. (B) 5 µg of nuclear protein was evaluated by EMSA to determine NFB DNA-binding activity. The induced p65-p50 heterodimer of NFB, identified by supershift assays, is marked by an arrow. Results in A and B are derived from the same experimental samples, and experimental conditions are marked above each lane. (C) 5 µg of nuclear protein from 72 hour post-infection time points with Ad.GFP (same sample as from lane 2 Panel B) and Ad.IBAS (same sample as from lane 5 Panel B) were evaluated by EMSA supershift to determine the subunit composition of the NFB DNA binding complex. Antibodies used for supershift were against p50, p52, C-Rel, Rel B and p65. The active DNA-binding complex is identified as p65-p50 heterodimers, and supershifted bands (in lanes 2, 6, 8 and 12) are marked by an asterisk to the left of the gel.

Having demonstrated that infection with Ad.IBAS enhances baseline NFB DNA-binding activity in the nucleus by inhibiting IB protein expression, we next sought to confirm that expression of IBAS mRNA enhanced NFB transcriptional activation following UV, TNF-, pervanadate or H/R treatment. In these studies, HeLa cells were infected with either a control adenovirus, Ad.BglII, or Ad.IBAS at an moi of 1000 particles/cell 24 hours prior to exposure to UV, TNF-, pervanadate or H/R. As previously reported, each of these treatments induced NFB DNA binding (Fig. 6); however, the time course of NFB activation varied. In uninfected cells, NFB DNA binding peaked 2 hours following UV irradiation or pervanadate treatment, 1 hour following TNF- treatment and 3 hours following H/R treatment (data not shown). Hence, for comparative studies of Ad.BglII- and Ad.IBAS-infected groups, samples were harvested at the peak time points of NFB DNA binding following a given stimulus. Results from this analysis clearly demonstrated that infection with Ad.IBAS increased nuclear NFB DNA binding following UV (Fig. 6A), TNF- (Fig. 6B), pervanadate (Fig. 6C) or H/R (Fig. 6D) treatment. In summary, these results demonstrate that inhibiting IB protein expression enhances activation of NFB DNA binding in the nucleus following exposure to all types of stimuli tested.

View larger version (63K): [in this window] [in a new window]   Fig. 6. IBAS enhances NFB binding activity following UV, TNF-, pervanadate or H/R treatment. HeLa cells were infected with either Ad.BglII or Ad.IBAS at an moi of 1000 particles/cell for 24 hours. Cells were treated with (A) UV (50 J/m2), (B) TNF- (10 ng/ml), (C) pervanadate (50 µM) or (D) H/R (5 hours hypoxia, 6 hours reoxygenation). Both treated and untreated cells were collected for nuclear extract preparation and EMSA analysis. Ad.IBAS-infected cells demonstrated a higher baseline of NFB DNA-binding activity (lanes 1 versus lane 4) and enhanced NFB activation following all treatments compared with Ad BglII-infected cells (lanes 2,3 versus lanes 5,6). The conditions for various experimental groups are marked above each panel.

Since DNA binding is only one indicator of NFB activation, we also sought to directly evaluate the transcriptional activity of NFB using the recombinant adenoviral reporter vector (Ad.NFBLuc) harboring a NFB-inducible luciferase gene. In these studies, HeLa cells were co-infected with Ad.NFBLuc and Ad.IBAS or Ad.BglII 24 hours prior to treatments. Whole cell extracts were harvested 6 hours following treatment and luciferase activity was measured as an indicator of NFB transcription activation. Consistent with our EMSA result on NFB DNA binding, NFB transcriptional activity was also significantly enhanced (P<0.01) in the Ad.IBAS-infected groups, as compared to the groups infected by Ad.BglII, following each of the environmental injuries (Fig. 7).

View larger version (24K): [in this window] [in a new window]   Fig. 7. IBAS enhances NFB transcriptional activity following UV, TNF-, pervanadate or H/R treatment. HeLa cells were co-infected with Ad.NFBluc (500 particles/cell) together with either Ad.BglII or Ad.IBAS at an moi of 1000 particles/cell for 24 hours. Cells were treated with (A) UV (50 J/m2), (B) TNF-(10 ng/ml), (C) pervanadate (50 µM) and (D) H/R (5 hours hypoxia, 6 hours reoxygenation). Luciferase assays were performed on samples harvested 6 hours after treatment. Results depict the mean (±s.e.m., n=6) raw unadjusted relative light units (RLU). NFB transcriptional activity was significantly enhanced (P<0.01) in the Ad.IBAS infected groups, as compared to the groups infected by Ad.BglII, following UV, TNF-, pervanadate or H/R treatment.

Activation of NFB prior to injury is pro-apoptotic following UV or TNF- treatment but anti-apoptotic following pervanadate or H/R treatment
On the basis of our findings that dominant-negative phosphorylation mutants of IB preferentially enhanced apoptosis following each of the environmental injuries tested, we hypothesized that elevating the level of NFB activation with IBAS prior to injury would have the reverse effect. Interestingly, results from these experiments demonstrated two distinct functional consequences of IBAS expression depending on the type of environmental injury evaluated (Fig. 8). Following UV treatment, IBAS expression significantly (P<0.001) increased apoptosis to 40.6%±2.2 as compared to the Ad.BglII-infected control group (15.5%±0.8). Similar observations were seen in TNF- treated cells, which had a two-fold increase (P<0.01) in apoptosis in Ad.IBAS-infected cells (8.9%±0.5) as compared to the Ad.BglII control group (4.6%±0.3). In stark contrast, following PV treatment, IBAS expression significantly decreased (P<0.01) the percentage of apoptotic cells to 16.0%±1.5 compared with the Ad.BglII-infected control group (30.4%±2.0). Similarly, following H/R, the Ad.IBAS-infected group demonstrated a significantly lower (P<0.01) apoptotic rate (4.0%±0.5) as compared with the control group (6.0%±0.7). These findings demonstrate that increased NFB activation during the acute phase of injury can have either pro-apoptotic or anti-apoptotic effects depending on the type of injury stimulus. Interestingly, this differential apoptotic effect of enhancing NFB activation prior to injury correlated with either serine or tyrosine IB phosphorylation pathway responsible for the activation of NFB following injury. Enhancing NFB activity prior to injury was pro-apoptotic for stimuli that promote IKK-mediated serine phosphorylation of IB (TNF- and UV), although it was anti-apoptotic for stimuli that promote tyrosine phosphorylation of IB (H/R and pervanadate). Although these findings suggest that enhancing NFB activity may be a viable strategy for reducing apoptosis following I/R injury, they also suggest that the temporal regulation of NFB is an important component in determining its apoptotic influence for other types of injuries.

View larger version (22K): [in this window] [in a new window]   Fig. 8. NFB activation is pro-apoptotic following UV and TNF- treatments, while anti-apoptotic following pervanadate and H/R treatments. HeLa cells were infected with either Ad.BglII or Ad.IBAS at an MOI of 1000 particles/cell for 24 hours. Cells were treated with (A) UV (50 J/m2), (B) TNF- (10 ng/ml), (C) pervanadate (50 µM) and (D) H/R (5 hours hypoxia, 18 hours reoxygenation) then harvested 18 hours after treatment. Cells are then trypsinized and stained with annexin-V-FITC and propidium iodine. Results depict the mean (±s.e.m., n=3) percent of apoptotic cells as shown by FACS analysis. Following UV and TNF- treatment, IBAS expression significantly (P<0.01) increased apoptosis compared with the Ad.BglII-infected control group. By contrast, following pervanadate and H/R treatments, IBAS expression significantly decreased (P<0.01) the percentage of apoptotic cells as compared to the Ad.BglII-infected control group.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion References

 
The manner in which NFB balances both the detrimental pro-inflammatory and beneficial anti-apoptotic consequences of its activation in a way that promotes cell survival following injury remains a paradox. Additionally, alternative pathways for activating NFB in response to injury that are independent of IKK-mediated serine phosphorylation of IB and ubiquitin-dependent degradation are being increasingly appreciated. This alternative pathway of activating NFB through tyrosine phosphorylation of IB appears to be critical following I/R injury. The goals of this study were designed to evaluate potential therapeutic strategies to enhance NFB activation and increase cell survival following I/R injury using in vitro H/R as a model. The importance of IB tyrosine phosphorylation in NFB transcriptional activation and cell survival following I/R injury has never been directly addressed in the literature owing to the lack of highly transducible vector systems expressing the IB(Y42F) mutant. Given the diversity of pathways associated with NFB activation, we also felt that comparing injury stimuli that activate NFB through either serine or tyrosine phosphorylation of IB would also be very useful in evaluating the consequences of modulating NFB activity on apoptosis. Important to this comparison was the establishment of a HeLa cell line model capable of responding to diverse injury stimuli that activate NFB through these two independent pathways. Such a model system in a single cell line assures that differences in the pathways of NFB activation are due to differences in the injury stimuli and not cell-specific differences in the abundance of signal transduction intermediates.

Studies establishing this model system confirmed that both UV and TNF- treatment leads to serine phosphorylation of IB. By contrast, hypoxia/reoxygenation and pervanadate (a tyrosine phosphatase inhibitor) treatment led to tyrosine phosphorylation of IB. Although these studies are confirmatory in nature, they do establish that this model system specifically modulates IB serine or tyrosine phosphorylation in an injury-specific manner within the same cell type. The functional specificity of either IB S32/36 or Y42 phosphorylation in promoting NFB activation in response to specific injury stimuli was also confirmed using transcriptional reporter assays and dominant-negative Y42F or S32/36A mutants of IB. These studies for the first time demonstrate that Y42, but not S32/36, phosphorylation of IB is the predominant pathway for transcriptional activation of NFB following H/R in an epithelial cell line model.

Although the present study has tried to distinguish between serine and tyrosine phosphorylation of IB following each environmental injury, it is recognized that overlap in these two activation pathways may exist. For example, it is still debated whether UV irradiation activates NFB through IKK-mediated IB serine phosphorylation. Several reports have demonstrated that the dominant mutants IKK(KM) or IB(S32/36A) can inhibit UV-induced NFB activation (Kulms et al., 2000; Li et al., 2001). However, others have reported that NFB activation following UV-C irradiation does not involve the IKK complex or IB serine phosphorylation (Bender et al., 1998; Li and Karin, 1998). Differences in these results may be cell line dependent. Although most recent results demonstrate that the majority of UV-induced NFB activation can be inhibited by IB(S32/36A), the extent of inhibition (74%) was much less than that seen for TNF- (99%). Furthermore, 10% of UV-induced NFB transcriptional activity was inhibited by the IB(Y42F) mutant, and although this inhibition was not statistically significant, it does suggest that some overlap in these two pathways may exist. Alternatively, additional non-IB-associated mechanisms controlling NFB activation, such as Akt phosphorylation of p65 (Madrid et al., 2000), may also play some minor role in these studies.

Controversies regarding the involvement of NFB in both anti-apoptotic and pro-apoptotic pathways have been difficult to reconcile (Lipton, 1997). In part, the complexity and diversity of pathways that can activate NFB has fed this controversy. Importantly, determinants of cell survival and apoptosis are, themselves, complex and not solely regulated by a single factor such as NFB. Hence, independent pathways of NFB activation may impart unique functional consequences to NFB activity by altering the composition of signaling intermediates that may act in concert to facilitate or prevent apoptosis under a given stimulus. The functional consequences of inhibiting NFB following pro-inflammatory stimuli such as TNF- have been well documented to enhance apoptosis in the presence of IB(S32/36A) expression. However, similar information is lacking for H/R injury using a IB(Y42F) mutant. Important to designing therapeutic strategies for preventing apoptosis following I/R by modulating NFB activity was to first establish whether its inhibition of NFB was pro- or anti-apoptotic in the setting of H/R. Studies using either IB(S32/36A) or IB(Y42F) mutants confirmed that inhibiting NFB under all environmental stimuli was proapoptotic. Although each of these mutants preferentially enhanced apoptosis following injuries specific for their respective pathways of IB phosphorylation, some overlap was seen. For example, although IB(Y42F) expression more significantly increased apoptosis following H/R than IB(S32/36A), the serine mutant also significantly enhanced apoptosis in comparison with Ad.BglII-infected controls. This overlap was much higher than that seen in studies evaluating the specificity of these mutants to transcriptionally activate NFB following each of the independent stimuli. We reason that the difference in the specificity of IB mutants to modulate NFB transcriptional activation or apoptosis is due to the longer time course required to achieve apoptosis in the setting of injury. Hence, although the acute stages of NFB activation may be selective for serine or tyrosine IB phosphorylation, during the later stages of injury greater overlap in these two pathways may exist owing to activation of NFB-responsive cytokines that restimulate cells through alternative NFB pathways. Nonetheless, these studies conclusively demonstrate for the first time that inhibition of NFB activation using an IB(Y42F) mutant enhances apoptosis following H/R.

Having demonstrated that activation of NFB is anti-apoptotic following H/R, we hypothesized that enhancing NFB activation prior to injury might provide a protective advantage to cells. In support of this hypothesis, infection with recombinant adenovirus expressing antisense IB mRNA significantly enhanced NFB transcriptional activation and reduced apoptosis following either H/R or PV treatment. Surprisingly, similar studies in TNF-- or UV-treated cells gave rise to the opposite result. Although expression of antisense IB mRNA similarly enhanced NFB transcriptional activation following either TNF- or UV treatment, a significant increase in apoptosis was observed with these stimuli in antisense IB-mRNA-expressing cells. Hence, elevation of NFB transcriptional activity prior to injury promoted apoptosis following TNF- or UV treatment, although it inhibited apoptosis following pervanadate or H/R treatment. These findings suggest that the temporal regulation of NFB influences apoptosis in a stimuli-dependent fashion.

One interesting feature of the differential effects of enhancing NFB on apoptosis following these various stimuli, is the fact that they fell into two distinct groups associated with a particular pathway for NFB activation. For example, both UV- and TNF--treated groups, which demonstrated enhanced apoptosis in the setting of elevated NFB activation, utilize S32/36 phosphorylation of IB as the pathway for NFB activation. By contrast, both H/R- and pervanadate-treated groups, which demonstrated reduced apoptosis in the setting of elevated NFB activation, utilize Y42 phosphorylation of IB as the pathway for NFB activation. These findings demonstrate that, depending on the type of environmental injury and the pathway for NFB activation, the consequences of enhancing NFB activation prior to injury can produce dramatically different phenotypes in relation to programmed cell death. We hypothesis that this previously unidentified relationship between IKK or protein tyrosine kinase (PTK) pathways of NFB activation, and the ultimate apoptotic consequences of this activation, is probably due to additional stimuli-specific factors induced under each of the given conditions. NFB most probably works in concert with such factors to determine apoptotic cell fates. This current hypothesis would propose that activated NFB following IKK or PTK stimuli is biochemically identical and that stimuli-specific factors linked to either IKK or PTK pathways of NFB activation uniquely alter the subset of injury response genes transcriptionally activated by N