Previous Article | Next Article ![]()
Journal of Virology, February 2002, p. 1578-1587, Vol. 76, No. 4
0022-538X/01/$04.00+0 DOI: 10.1128/JVI.76.4.1578-1587.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Biotechnology Laboratory,,1 Biomedical Research Centre, and Departments of Medical Genetics, Microbiology and Immunology, and Zoology, University of British Columbia, Vancouver, British Columbia V6T 1Z3,,2 Pulmonary Research Laboratory, University of British Columbia, St. Paul's Hospital, Vancouver, British Columbia V6Z 1Y6, Canada3
Received 5 June 2001/ Accepted 9 November 2001
|
|
|---|
|
|
|---|
and ß, respectively (6, 22, 83, 93). Most viral survival factors can be classified according to which host survival protein they mimic. For example, there are viral Bcl-2-like proteins, such as BHRF1 (33) and A179L (1), and others that mimic the Fadd-like interleukin-converting enzyme-inhibitory protein, such as K13, BORFE2, and MC159 (92). There are also homologues of the cellular inhibitor of apoptosis proteins (17) and serpin homologues (19), both of which inhibit caspases and/or caspase activation. Ultimately, there are viral proteins that mimic death receptors and prevent their ligation, like MT-2 (79), SFV-T2 (85), CrmB (37), CrmC, and CrmD (58).
Notable exceptions to this classification are antiapoptotic viral proteins for which no cellular homologue or mode of action has been identified. Interestingly, some of these proteins are localized in the endoplasmic reticulum (ER). For instance, a subset of the M-T2 protein, the myxoma virus homologue of the tumor necrosis factor (TNF) receptor, is ER retained, yet it is able to block apoptosis (80). MT-4, another ER-localized myxoma virus protein, has also been shown to prevent apoptosis, through an unknown mechanism (4). Similar to M-T2, UL144 is an intracellular membrane protein homologous to the TNF receptor superfamily of proteins (5). Even ER-localized cellular proteins, shown to have an effect on apoptosis, have a poorly defined mechanism. Some of these cellular factors are protective, such as DAD1 (also known as Ost2p) (88), PDI (90), Grp94 (56), calreticulin (57), and Grp78 (87), while others are proapoptotic, such as BAP31 (70) and mutant forms of PS-1 (45) and PS-2 (40), associated with familial Alzheimer's disease.
The adenovirus type 2 (Ad2) E3-6.7K protein is localized primarily in the ER, as evidenced by cellular immunofluorescence staining and by the presence of high-mannose N-linked glycan modifications in the mature protein (103). There is evidence that a small subset of this protein is found on the plasma membrane where it associates with RID ß and allows the RID complex to downregulate the DR4 and DR5 TNF-related apoptosis-inducing ligand (TRAIL) receptors (6). However, a recent study using different experimental conditions suggests that E3-6.7K is not required for the downregulation of DR4 by RID (95), yet it may still be required by RID in downregulating DR5. The RID complex also downregulates Fas (6, 22, 83, 93) and epidermal growth factor receptor (94) and prevents TNF-induced apoptosis (28), arachidonic acid release (48), and translocation of cytosolic phospholipase A2 (cPLA2) to the membrane (18). E3-6.7K, however, was not shown to be required for any of the latter effects but may play a role in augmenting RID function. We focused our studies on examining E3-6.7K in the absence of other viral factors, in order to explore whether it has an independent role in apoptosis.
Our findings indicated that the adenovirus E3-6.7K protein maintained Ca2+ homeostasis in transfected cells since its expression buffered the Ca2+ flux generated in response to thapsigargin, an inhibitor of the ER-associated Ca2+ ATPase. This led to the resistance of E3-6.7K-transfected cells to thapsigargin-induced apoptosis. The newly observed effect of E3-6.7K was independent of the RID complex and cannot be explained by its previously proposed role in downregulating DR4 and DR5. In addition, TNF-, TRAIL-, and Fas ligand (FasL)-induced apoptosis and TNF-induced release of arachidonic acid were significantly reduced in cells expressing E3-6.7K. We speculate that E3-6.7K may offer the virus a powerful and broad survival mechanism by preventing death receptor- or Ca2+ efflux-induced apoptosis and by reducing the release of inflammatory mediators.
|
|
|---|
Generation of U937 and Jurkat cell lines stably expressing E3-6. 7K. U937 human histiocytic lymphoma (89) and Jurkat E6-1 T-cell lymphoma (101) cells, obtained from the American Type Culture Collection, were maintained in RPMI 1640 containing 10% fetal calf serum, 2 mM L-glutamine, 10 mM HEPES, 100 U of penicillin per ml, and 100 µg of streptomycin per ml in an atmosphere of 5% CO2 and 100% humidity.
U937 cells were transfected with the vector pBCMGSneo (44) alone or with pBCMGSneo carrying E3-6.7K by using the DMRIE-C cationic lipid reagent (Life Technologies) and following the manufacturer's protocol. Briefly, 4 µg of DNA was mixed and incubated for 30 min with 0.5 ml of Opti-MEM medium (Life Technologies) and 12 µl of DMRIE-C. The liposome-DNA mix was then mixed and incubated with 0.5 ml of 5 x 105 cells in Opti-MEM medium for an additional 4 h. Stably transfected cells were selected with 1 mg of G-418 sulfate per ml 48 h later.
Jurkat cells were transfected by electroporation. Cells (2.5 x 107) were washed twice in Opti-MEM medium, mixed with 20 µg of the appropriate plasmid, and electroporated with a Bio-Rad Electroporator at 250 V and 950 µF. Transfected cells were selected in medium containing 0.5 µg of puromycin per ml. The pIRESpuro2 vector (Clontech) carries the encephalomyocarditis virus internal ribosome entry site (IRES), allowing cocistronic expression of E3-6.7K and the puromycin resistance gene.
Several subclones of the transfected cell lines were generated by serial dilution and examined for expression of E3-6.7K transcript by Northern blotting. The expression of E3-6.7K was very similar in all the clones examined. However, all the G-418- and puromycin-resistant U937 and Jurkat cells that survived the selection procedure were pooled and used in our experiments, in order to avoid clonal variation.
Ratiometric intracellular [Ca2+] determination.
Jurkat-neo and Jurkat-6.7K cells were washed and resuspended at a concentration of 107 cells/ml in Opti-MEM (Life Technologies). Aliquots of 100 µl of the cell suspension were incubated for 90 min at 37°C with 6.0 µg of Indo-1 AM ester per ml. Prior to analysis, 1.9 ml of Opti-MEM was added to the cell suspension. Intracellular Ca2+ levels were measured with the ratiometric Ca2+ indicator Indo-1; its emission maximum shifts from
485 nm in Ca2+-free medium to
400 nm when the dye is saturated with Ca2+. The ratio of the emission signals at 405 nm and 485 nm represents the ratio of Ca2+-bound to Ca2+-free Indo-1. The kinetic analysis examined approximately 2 x 104 cells/min with a FACS-Vantage sorter equipped with an UV laser and appropriate filters for the 405- and 485-nm wavelengths. After the establishment of a stable baseline for the first 5 min, the cells were stimulated with 5 nM thapsigargin and monitored for another 20 min.
Arachidonic acid release assays. U937-neo and U937-6.7K cells were grown at low density in 10% HyClone fetal calf serum, RPMI 1640, 2 mM L-glutamine, and 10 mM HEPES for several days and then harvested and washed twice in phosphate-buffered saline (PBS)-1% bovine serum albumin. Approximately 5 x 106 cells (5 x 105 cells/ml) were labeled for 20 h in medium supplemented with 0.4 µCi of [3H]arachidonic acid [5,6,8,9,11,12,14,15-3H(N)] (0.1 mCi/ml stock; New England Nuclear) per ml. Cells were washed twice in RPMI 1640-0.2% bovine serum albumin and incubated for 1 h in wash medium to minimize the spontaneous release of [3H]arachidonic acid. Then 2 x 105 cells were stimulated with either medium alone, 90 ng of TNF (2,000 U/ml) (Boehringer Mannheim) per ml, 2 µg of cycloheximide (CHX) per ml, or a combination of 90 ng of TNF and 2 µg of CHX per ml for 20 h. The cellular supernatant from three independent experimental samples was mixed with scintillation fluid and counted to determine the amount of released radioactivity. For each cell line, three samples were lysed in Tris-buffered saline (50 mM Tris [pH 7.5], 150 mM NaCl [TBS]) containing 2% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, and the lysate was used to determine the total amount (in counts per minute) of incorporated [3H]arachidonic acid. The amount of [3H]arachidonic acid released in experimental samples was expressed as a percentage of the total incorporated [3H]arachidonic acid.
Thapsigargin-induced apoptosis assays in transfected Jurkat cells.
Following induction of apoptosis in Jurkat-neo and Jurkat-6.7K cells with 1 or 10 µM thapsigargin for 24 h, we used the DNA-intercalating dye YO-PRO-1 (Molecular Probes) as an assay of membrane permeability changes. Cells were harvested, washed in PBS, and resuspended at a concentration of 106 cells/ml. Following staining with 0.1 µM YO-PRO-1 and 1 µg of propidium iodide (PI) per ml, the cells were analyzed by flow cytometry with FL1
530 nm- and FL2
575 nm-compensated emission readings on a FACScan (Becton Dickson). Viable, apoptotic cells showed green fluorescence, necrotic cells showed red and green fluorescence, and live, nonapoptotic cells showed very little fluorescence. The apoptotic index (AI) excluded the necrotic cell population and was calculated as (number of viable apoptotic cells/total number of viable cells) x 100.
Death receptor-induced apoptosis assays in transfected U937 and Jurkat cells. Fluorescein isothiocyanate (FITC)-conjugated annexin V (Pharmingen) was used to determine the binding of annexin V to externalized phosphatidyl serine. The protocol was based on the manufacturer's annexin V-FITC staining protocol. For each sample, 5 x 106 U937-neo or U937-6.7K cells were harvested, washed twice in PBS, and treated for 7 h with medium alone, 100 ng of TNF per ml, 10 µg of CHX per ml, or a combination of 100 ng of TNF and 10 µg of CHX per ml. The cells were stained with 5 µl of annexin V-FITC in a mixture containing 10 mM HEPES-NaOH (pH 7.4), 140 mM NaCl, and 2.5 mM CaCl2 and analyzed with a fluorescence-activated cell sorter (FACS). Viable cells were gated based on their side scatter characteristics, and apoptotic cells were distinguished from nonapoptotic cells based on their green fluorescence.
Alternatively, in the experiments examining FasL and TRAIL-induced apoptosis, 5 x 106 Jurkat-6.7K or Jurkat-neo cells each were harvested, washed twice in PBS, and treated with agonist for 12 h. The treatment consisted of either 1 µg of DX2 (anti-human Fas monoclonal antibody) (Pharmingen) per ml and 2 µg of goat anti-mouse secondary antibody per ml or a recombinant fusion protein containing the extracellular domain of TRAIL (10 ng/ml; Upstate Biotech.) and a potentiating reagent (5 µg/ml; Upstate Biotechnology) which consists of a monoclonal antibody against the tag present in purified TRAIL. In these experiments, we used both the annexin V-Alexa-488 and the previously described YO-PRO-1 apoptosis assays (Molecular Probes). Like annexin-FITC, annexin V-Alexa-488 (Molecular Probes) labels cells that externalize phosphatidyl serine. The same staining protocol was followed, with the exception that a highly fluorescent non-cell-permeating DNA-intercalating dye, 50 µM CytoxGreen (Molecular Probes), was used to label necrotic cells. We examined the green fluorescence of stained cells with a FACS (Becton Dickinson). Viable, nonapoptotic cells showed little or no green fluorescence, while viable, apoptotic cells showed intermediate green fluorescence. Necrotic cells, on the other hand, showed 10 times greater fluorescence due to CytoxGreen staining than the apoptotic cells labeled with annexin V-Alexa-488 alone.
Activation of caspase-3 and cleavage of PARP by Western blotting. To analyze the cleavage of poly(ADP-ribose) polymerase (PARP), cell lysate equivalent to 105 cells was denatured in sodium dodecyl sulfate (SDS) sample buffer (0.1 M Tris [pH 6.8], 24% glycerol, 8% SDS, 0.2 M dithiothreitol, 0.02% bromophenol blue). The protein concentration was determined with a bicinchoninic acid assay kit (Pierce). Equivalent amounts of protein from each sample were resolved on a 10% glycine-SDS-polyacrylamide gel electrophoresis (PAGE) Laemmli gel system, blotted with the Towbin vertical-transfer wet system onto a 0.45-µm-pore-size Immobilon-P polyvinylidene difluoride (PVDF) membrane (Millipore), and incubated with anti-PARP mouse monoclonal antibody (1:5,000 dilution; Pharmingen). After four washes in TBS containing 0.1% Tween 20 detergent (TBS-T), the blot was incubated with horseradish peroxidase-conjugated goat anti-mouse antiserum (1:50,000 dilution) and visualized by chemiluminescence with a SuperSignal West Pico kit (Pierce Chemical). Alternatively, to detect caspase-3 activation, the cell lysate was resolved on a 10% T, 3% C Tricine-SDS-PAGE gel without spacer, blotted with a Towbin vertical-transfer wet system on a 0.2-µm-pore-size Immobilon-P PVDF membrane (Millipore), and incubated with anti-caspase-3 rabbit polyclonal antibody (1:5,000 dilution; Pharmingen). After four washes in TBS-T, the blot was incubated with horseradish peroxidase-conjugated goat anti-rabbit antiserum (1:50,000 dilution) and visualized by chemiluminescence with a SuperSignal West Pico kit (Pierce Chemical).
|
|
|---|
and ß to downregulate the level of TRAIL receptors 1 and 2 from the surfaces of cells (6). While this identifies a role for the subset of the E3-6.7K protein expressed at the cell surface and possibly the endosomal compartment, it does not directly indicate a role for the large subset of the protein which is retained in the early secretory compartment. We began our studies by examining the effect of E3-6.7K on death receptor-induced apoptosis in transfected cells, in the absence of other viral proteins. We chose to study the effects of E3-6.7K in cells of lymphoid and monocytic origin. Both cell types have been implicated as a possible reservoir for persistent adenovirus infections in vivo (2, 3, 13, 35, 84, 99). Immune evasion and the establishment of viral persistence may be more appropriately studied in cells resembling the in vivo reservoir of adenovirus. Subgroup C Ad5 can establish persistent infection in the human monocytic cell line U937, which maintains large copy numbers of unintegrated viral genomes following infection (13). Infected cells continue to grow and produce minute amounts of mature virus a year after infection. The Ad5-infected U937 cells have been proposed to provide a model of persistent adenovirus infection in humans. The human Jurkat T-cell line, on the other hand, has been used to study the efficacy of the immunoevasive protein E3/19K in downregulating major histocompatibility complex class I in lymphoid cells (47). We examined the responses of E3-6.7K-transfected Jurkat or U937 cells to mediators of apoptosis and inflammation.
We transfected E3-6.7K cDNA amplified by PCR from a plasmid bearing the E3 region into the TNF-sensitive U937 human histiocytic lymphoma cell line. We established a stable population of U937 cells transfected with the episomal vector pBCMGSneo containing the cDNA for E3-6.7K (U937-6.7K) and another population transfected with vector alone (U937-neo). All surviving neomycin-resistant cells (in excess of 103 clones) were pooled and used in the following experiments to avoid the clonal variation known to arise in U937 cells. Expression of E3-6.7K was analyzed by Northern blotting and by immunoprecipitation. We confirmed that the E3-6.7K protein was properly synthesized by immunoprecipitating it with a polyclonal rabbit antiserum raised against an E3-6.7K C-terminus-derived peptide and examined it by SDS-PAGE (results not shown).
We assayed the response of U937-neo and U937-6.7K cells to TNF-induced apoptosis by measuring the externalization of phosphatidyl serine by apoptotic cells (61) stained with annexin V-FITC. The presence of E3-6.7K yielded a 2.2-fold reduction in the proportion of apoptotic cells in U937-6.7K compared to U937-neo, following stimulation with TNF (Fig. 1). The U937-6.7K cells showed a 2.8-fold reduction in apoptosis compared to U937-neo cells following augmented stimulation with a combination of TNF and CHX, a protein synthesis inhibitor synergistic with TNF. The presence of E3-6.7K significantly decreased the apoptotic response in U937 cells upon stimulation with TNF or a combination of TNF and CHX.
![]() View larger version (31K): [in a new window] |
FIG. 1. E3-6.7K protects against apoptosis induced by TNF in U937 cells. U937-neo (a to d) and U937-6.7K (e to h) cells were stimulated for 7 h with medium alone, 100 ng of TNF per ml, 10 µg of CHX per ml, or a combination of 100 ng of TNF and 10 µg of CHX per ml. For each cell population, a sample treatment was stained with annexin V-FITC and analyzed with a FACS. A second sample was analyzed with a FACS in the absence of annexin V-FITC in order to determine the background fluorescence, which corresponds to the fluorescence associated with the annexin V-negative cell population from each sample. The AI was calculated as (number of annexin V-positive viable cells/total number of viable cells) x 100. The results are representative of three repeat experiments.
|
We found that Jurkat cells stably transfected with E3-6.7K (Jurkat-6.7K) cells were less sensitive to apoptosis induced through Fas (Fig. 2) or TRAIL (Fig. 3) receptors than Jurkat cells transfected with vector alone (Jurkat-neo) cells. Following stimulation with FasL or TRAIL, the apoptotic response was assayed by two independent methods. In addition to annexin V labeling of externalized phosphatidyl serine, we also measured the exclusion of the dye YO-PRO-1 by viable, nonapoptotic cells. In both assays we used a viability dye to stain the necrotic and late-apoptotic cell population, commonly excluded from the analysis of induction of apoptosis. Analysis of the apoptotic response was performed in accordance with manufacturer's instructions and in accordance with published methods used in similar studies (11, 42, 78, 81, 97). Both assays demonstrated that E3-6.7K was effective in reducing TRAIL- and FasL-induced apoptosis in Jurkat cells. It is important to note that the antiapoptotic effect of E3-6.7K against TNF- and TRAIL-induced apoptosis was more pronounced than that against FasL-induced apoptosis, as observed by measuring the externalization of phosphatidyl serine (Fig. 1, 2A, and 3A).
![]() View larger version (27K): [in a new window] |
FIG. 2. E3-6.7K protects against apoptosis induced by Fas in Jurkat-neo and Jurkat-6.7K cells. To study Fas-induced apoptosis, cells were stimulated for 12 h with 1 µg of the anti-human Fas monoclonal antibody DX2 (Pharmingen) per ml and 2 µg of goat anti-mouse polyclonal antibody per ml. (A) Apoptotic cells were stained with annexin V-Alexa-488, indicating externalization of phosphatidyl serine; AI = (number of annexin V-positive, CytoxGreen-negative cells/total number of CytoxGreen negative cells) x 100. The V sector indicates the nonapoptotic, viable cell population, the A sector indicates the apoptotic, viable cell population, and the N sector indicates the necrotic cell population. (B) Alternatively, cells were stained with YO-PRO-1, indicating an increase in membrane permeability; AI = (number of YO-PRO-1-positive, PI-negative cells/total number of PI-negative cells) x 100. The lower left quadrant is the nonapoptotic, viable cell population, the lower right quadrant (A) represents the apoptotic, viable cell population, and the upper right quadrant (N) represents the necrotic cell population.
|
![]() View larger version (27K): [in a new window] |
FIG. 3. E3-6.7K protects against apoptosis induced by TRAIL in Jurkat-neo and Jurkat-6.7K cells. To study TRAIL-induced apoptosis, cells were stimulated for 12 h with 10 ng of a recombinant fusion protein containing the extracellular domain of TRAIL (Upstate Biotechnology) per ml and a potentiating reagent (5 µg/ml; Upstate Biotechnology), which consists of a monoclonal antibody against the tag present in purified TRAIL. (A) The apoptotic cells were stained with annexin V-Alexa-488, indicating externalization of phosphatidyl serine; AI = (number of annexin V-positive, CytoxGreen-negative cells/total number of CytoxGreen-negative cells) x 100. The V sector indicates the nonapoptotic, viable cell population, the A sector indicates the apoptotic, viable cell population, and the N sector indicates the necrotic cell population. (B) Alternatively, cells were stained with YO-PRO-1, indicating an increase in membrane permeability; AI = (number of YO-PRO-1-positive, PI-negative cells/total number of PI-negative cells) x 100. The lower left quadrant is the nonapoptotic, viable cell population, the lower right quadrant (A) represents the apoptotic, viable cell population, and the upper right quadrant (N) represents the necrotic cell population.
|
![]() View larger version (31K): [in a new window] |
FIG. 4. The presence of E3-6.7K results in the reduction of the thapsigargin-induced Ca2+ flux and apoptosis. (A) Effect of E3-6.7K on thapsigargin-induced apoptosis in transfected Jurkat cells. Jurkat-6.7K and Jurkat-neo cells were stimulated with medium containing 1% dimethyl sulfoxide (top),or 1 or 10 µM thapsigargin (TG) for 24 h, and then they were stained with YO-PRO-1 to examine the increase in the membrane permeability of apoptotic cells. The AI excluded the necrotic cell population and was calculated as (number of YO-PRO-1-positive, PI-negative apoptotic cells/total number of PI-negative, viable cells) x 100. The A region represents the apoptotic, viable population, and V represents the nonapoptotic, viable population; AI = A/(A+V). The experiment was repeated with similar results: 23.2% ± 1.4% and 47.8% ± 4.7% for Jurkat-neo cells stimulated with 1 and 10 µM TG, respectively, and 10.3% ± 1.3% and 21.9% ± 6.6% for Jurkat-6.7K cells stimulated with 1 and 10 µM TG, respectively (P < 0.05). (B) Kinetic analysis of the effect of E3-6.7K on the release of Ca2+ from the ER in response to TG. Jurkat-6.7K and Jurkat-neo cells were loaded for 90 min at 37°C with 0.5 µM Indo-1 AM ester and examined on a UV-equipped FACS-Vantage flow cytometer (Becton Dickinson) for 5 min to establish the baseline fluorescence, equivalent to the resting level of intracellular calcium. At 5 min, cells were treated with 5 nM TG and analyzed for 25 min for a total of 5 x 105 events. The experiment was repeated three times with similar results.
|
Thapsigargin-induced Ca2+ flux in Jurkat-neo cells resulted in an 80% increase in the level of intracellular calcium. In Jurkat-6.7K cells, however, the observed increase was restricted to 34% (Fig. 4B). The presence of E3-6.7K reduced the efflux of Ca2+ in response to thapsigargin. The observed decrease in Ca2+ homeostasis correlates with the protective role conferred by E3-6.7K against thapsigargin-induced apoptosis. Both apoptosis and Ca2+ flux assays were conducted in media containing extracellular Ca2+, which is essential for the induction of apoptosis in Jurkat cells by thapsigargin (31). As a result, the initial ER Ca2+ flux is amplified by capacitative Ca2+ entry through Ca2+ release-activated Ca2+ channels.
The presence of E3-6. 7K delays TNF-induced activation of caspase-3 and cleavage of PARP. Given the protective role of E3-6.7K against both death receptor- and Ca2+-induced apoptosis, we examined its effect on the so-called "executioner" phase of apoptosis. More specifically, we assayed its effect on the TNF-induced cleavage of procaspase-3 and of the DNA repair enzyme PARP (Fig. 5). The 17- and 11-kDa proteolytic products of procaspase-3 are subunits of the heterodimeric, active form of caspase-3. The appearance of the active forms of caspase-3 was significantly delayed in U937-6.7K versus U937-neo cells (Fig. 5A). In addition, there was a significant reduction in the amount of active caspase-3 in U937-6.7K cells versus U937-neo cells. The 85-kDa inactive form of PARP appeared in both U937-neo and U937-6.7K cell lines at 2 h after stimulation with TNF-CHX (Fig. 5B). There was, however, a noticeable difference with regard to the kinetics of PARP inactivation: the amount of inactive PARP present in the U937-6.7K cells was considerably reduced compared to that in U937-neo cells. Therefore, the presence of E3-6.7K delayed and reduced the activation of caspase-3 and inhibited PARP inactivation during TNF-induced apoptosis.
![]() View larger version (66K): [in a new window] |
FIG. 5. Effect of E3-6.7K on the induction of procaspase-3 processing and PARP cleavage during TNF-induced apoptosis in vivo. Cell extracts were obtained from U937-neo and U937-6.7K cells that had been treated with 10 ng of TNF and 0.5 µg of CHX per ml for various lengths of time. Lysates containing equivalent amounts of protein based on the bicinchoninic acid (Pierce) protein concentration assay were loaded in each lane. After electrophoresis and transfer to PVDF membranes, blots were incubated with anti-caspase-3 rabbit antiserum that recognizes the 17- and 11-kDa subunits of the active, processed protein (A) and anti-PARP mouse monoclonal antibody that recognizes both the active 116-kDa and inactive 85-kDa forms of the protein (B). The blots were developed with a secondary antibody and visualized by chemiluminescence (Pierce Chemical). Similar results were obtained in two repeat experiments.
|
We found that in the presence of E3-6.7K there was a reduction in the release of [3H]arachidonic acid from U937-6.7K cells by 50% compared to that from U937-neo cells following stimulation with TNF (Fig. 6). When the stimulus was increased by the addition of TNF and CHX, U937-6.7K cells released 60% less [3H]arachidonic acid than U937-neo cells (Fig. 6). This effect can be explained by the fact that E3-6.7K affects Ca2+ homeostasis, which indirectly affects the activity of cPLA2. Our assay of [3H]arachidonic acid release was conducted for periods ranging from 12 to 48 h after stimulation (results not shown). The amount of soluble [3H]arachidonic acid is an equilibrium determined by cycles of deacylation and reacylation. We found that in our system, the equilibrium favored deacylation of incorporated [3H]arachidonic acid for the first 24 h, reaching the peak of release of [3H]arachidonic acid at 20 h poststimulation. The release of incorporated [3H]arachidonic acid over the next 24 h was minimal as intracellular stores of [3H]arachidonic acid became depleted. As a result, we studied the effect of E3-6.7K at 20 h poststimulation. The effect of E3-6.7K on the release of arachidonic acid is consistent with its protective role against apoptosis and correlates with its effect on Ca2+ homeostasis.
![]() View larger version (29K): [in a new window] |
FIG. 6. Analysis of the effect of E3-6.7K on the inducible release of radiolabeled arachidonic acid. U937-neo and U937-6.7K cells were stimulated with medium alone, 90 ng of TNF per ml, 2 µg of CHX per ml, or a combination of 90 ng of TNF and 2 µg of CHX per ml. After 20 h, the amount (in counts per minute) of [3H]arachidonic acid released in the medium was expressed as a percentage of the total incorporated [3H]arachidonic acid. The assay was set up in triplicate; however, due to their relatively small values, the standard deviations of the values for the samples from U937-6.7K cells treated with medium alone (6.72% ± 0.037%) and U937-6.7K treated with CHX (5.74% ± 0.027%) are difficult to observe on this graph. These results are representative of six repeat experiments.
|
|
|
|---|
The promoter that controls the E3 region is active in cells of lymphoid origin in the absence of E1A (102) and is responsive to TNF (16). It is possible that the role of the E3 region proteins in lymphoid cells is to aid in the establishment of persistent infections. Our studies examine the role of E3-6.7K in protecting cells of lymphoid and monocytic origin from apoptosis, with implications for the persistent viral infection seen in vivo in tonsillar or adenoidal lymphoid cells or in cells of monocytic origin. The effect of E3-6.7K on apoptosis in these cells was examined by transfection of the cDNA coding for E3-6.7K and creation of stable cell lines. The alternate approach would have been to study the response of lymphoid cells to apoptosis following infection of these cells with wild-type virus or virus with a deletion of E3-6.7K. We favored the single-gene-transfection approach for a number of reasons. Firstly, infection of lymphoid cells with subgroup C virus is extremely inefficient (84, 96), though there is a report of productive infection of Jurkat cells by Ad5 (51). Secondly, the only published deletion virus considered to have a single deletion in the E3-6.7K gene, dl739, was recently observed to express reduced levels of RID
protein (W. S. M. Wold, personal communication). This may be due to a secondary deletion but is more likely due to the fact that any deletion in the very complex transcription unit, E3, has the potential to alter the splicing pattern and 3'-end formation of E3 transcripts. Such cis-acting sequences have been found to lead to the aberrant expression of genes not directly affected by the deletion (7, 9). This therefore precludes the assumption that viruses with deletions of various E3 proteins have a normal pattern of expression of the remaining genes and affects the conclusions derived from studies that employ such viruses.
Our results indicate that the E3-6.7K protein acted in the absence of other adenovirus proteins to protect transfected cells against TNF-, FasL-, and TRAIL-induced apoptosis and against TNF-induced arachidonic acid release. More importantly, thapsigargin-induced Ca2+ efflux and apoptosis were dramatically reduced in the presence of E3-6.7K, which suggests a possible effect of E3-6.7K on events that regulate Ca2+ homeostasis. The three proteins coded by the E3 region, E3/10.4K, E3/14.5K, and E3/14.7K, that prevent death receptor-induced apoptosis have also been shown to prevent the release of arachidonic acid following stimulation with TNF (48). Similarly, E3-6.7K is effective in reducing the release of arachidonic acid. This newest addition to the E3 region proteins that prevent both TNF-induced apoptosis and release of inflammatory mediators reflects the importance of TNF and inflammation in the immune responses against adenovirus (21, 25).
The effect of E3-6.7K on TRAIL- or TNF-induced apoptosis was comparable with its effect on thapsigargin-induced apoptosis. This suggests that its effect on death receptor-induced apoptosis, in the absence of RID
and ß, could in fact be related to its effect on downstream events such as Ca2+ release. Currently, the role of Ca2+ efflux from the ER during death receptor-induced apoptosis is not very well understood. Some studies describe an effect of Ca2+ efflux on the apoptotic response of cells to FasL (41) or TNF (46), while others disagree (62). Still other studies attribute a role to Ca2+ efflux in the late and not early events following FasL-induced apoptosis (82). Through the discovery of proteins that specifically affect Ca2+ release, through well-defined mechanisms, it will be possible to explore the role of Ca2+ flux in death receptor-induced apoptosis.
A previous study indicated that E3-6.7K is required by RID
and ß to reduce the surface level of TRAIL R1 and R2 and prevent TRAIL-induced apoptosis (6). Interestingly, we noted that many of the studies that examined the effect of RID
and ß made use of viral deletion mutants that were deficient in RID
and ß or both but retained the expression of E3-6.7K. It is then plausible that E3-6.7K acts to augment the effects of RID
and ß in death receptor-induced apoptosis by acting both alone and in combination with them. We propose, therefore, that E3-6.7K is an antiapoptotic protein with two possible mechanisms. First, as previously shown, cooperating with other viral proteins E3-6.7K results in the downregulation of TRAIL receptors. Secondly, E3-6.7K acts alone to prevent apoptosis against a variety of death receptors as well as intracellular mediators of apoptosis. Its effect, in the absence of other viral proteins, appears to involve the maintenance of cytosolic Ca2+ homeostasis.
It is common for viral proteins to have more than one mechanism or function, a reflection, possibly, of the large number of functions that need to be performed by a successful virus. Some viral antiapoptotic proteins can use two distinct mechanisms to achieve their goal, as previously described for the TNF receptor homologue M-T2 (80) coded by myxoma virus. In addition, there are also examples of viral antiapoptotic proteins that have both cooperative means, with other viral proteins, and independent ones to achieve similar goals. For example, adenovirus E4orf4 can act both independently and in combination with E1A to induce apoptosis (53, 60), while E4orf6 and E1B55K have both independent and cooperative means to prevent it (20, 34).
Other viral and cellular proteins, such as p35 and the neuronal apoptosis-inhibitory protein, have also been shown to prevent thapsigargin-induced apoptosis (63, 75); few, however, have been shown to affect Ca2+ homeostasis following exposure to thapsigargin. One of these, Bcl-2, shares intriguing similarities with E3-6.7K, encompassing both their topology and their posttranslational mode of targeting 39; A. R. Moise, unpublished data) and possibly their putative mode of action. Mitochondrion-localized Bcl-2 exerts its antiapoptotic effects by blocking the release of cytochrome c from the mitochondria (29). A subset of Bcl-2 proteins localizes to the cytoplasmic face of the ER membrane (39). Both wild-type Bcl-2 and an ER-restricted Bcl-2-cb5 fusion protein block the apoptotic cross talk between the ER and the mitochondria (30), and similarly to E3-6.7K, Bcl-2 has been shown to block thapsigargin-induced Ca2+ efflux from the ER (49). Recently, Bcl-2 was shown to decrease the free Ca2+ concentration within the ER (24, 73) by increasing its permeability to Ca2+, which results in the depletion of the mobilizable Ca2+ reserves used during agonist- or apoptosis-induced efflux. It is possible that the mechanism of E3-6.7K is related to the current model of the mechanism of action of ER localized Bcl-2. In this case, expression of E3-6.7K should be able to induce a Ca2+ "leak" in the ER membrane, draining the mobilizable Ca2+ reserves. It could achieve this by regulating an existing channel or by forming a channel of its own.
The effect of E3-6.7K on thapsigargin-induced Ca2+ flux places it upstream of many apoptotic effectors that are activated by of Ca2+ release from the ER. Some of these effectors are the death-associated protein kinase (14), protein kinase C (98), the MEF2 transcription factor (105), and calcineurin. Calcineurin, a Ca2+-regulated serine-threonine phosphatase, has been implicated in a variety of processes that result in the induction of apoptosis by activating nuclear factor of activated T-cells, which results in the expression of FasL (50), or by activating the proapoptotic Bcl-2 family member Bad (100). Recently, Ca2+-induced proteases, such as the calpain class of calcium-regulated cysteine proteases (74) and caspase-12, have been implicated in apoptosis. The recently discovered caspase-12 is activated by inducers of the ER stress response and by calpains (65) but not by other inducers of apoptosis, such as TNF (66).
Recently, the luminal environment of the ER and, more particularly, ER chaperones has been shown to play an important role in determining the sensitivity of cells to apoptosis (67). Several ER-resident chaperone proteins, such as calreticulin (10) and BiP/Grp78 (57), appear to increase the Ca2+ storage capacity of the ER (55, 67) and may regulate capacitative Ca2+ entry (64). Chaperone induction has also been associated with viral infection, as many viruses induce an ER overload response through the abundant expression of secreted viral proteins, as in the case of E3-6.7K's cocistronic partner E3/19K (71). The link between viral infection, ER stress, and sensitivity to apoptosis is also hinted at by the observation that both Gam1, an antiapoptotic protein encoded by the avian adenovirus CELO (12, 26), and E1A (69) are powerful inducers of heat shock proteins. In the case of Gam1, this was an essential event for avian adenovirus replication. We sought to examine whether the protective effect of E3-6.7K observed in our study could be attributed to an effect on chaperone induction. We examined the expression of Grp94 and Grp78 in transfected cells before and after induction with thapsigargin or heat shock. Our results (not shown) indicated that the presence of E3-6.7K did not affect the induction of these two chaperones. We are currently examining the effect of E3-6.7K on other markers of the stress response, such as c-Jun NH2-terminal kinase, calpains, Hsp70, and caspase-12.
Future studies aimed at determining the mode of action of ER localized pro-and antiapoptotic proteins will bridge the gap between the ER stress response and apoptosis. We propose that like E3-6.7K, other ER-localized viral survival factors are good candidates for the functional dissection of these common pathways.
We thank W. S. Wold for the generous gift of polyclonal antisera recognizing E3-6.7K and the plasmid bearing the Ad2 E3 region. We thank E. White, D. R. Green, and R. Lippe for helpful discussions and G. Osborne from the Multi User Flow Cytometry Core Facility for technical assistance in measuring apoptosis and calcium fluxes.
|
|
|---|
B. J. Biol. Chem. 271:30249-30255.
B activity and as a target of an adenovirus inhibitor of tumor necrosis factor alpha-induced apoptosis. Proc. Natl. Acad. Sci. USA 96:1042-1047.
B by the adenovirus E3/19K protein requires its ER retention. J. Cell Biol. 132:511-522.
cooperates with calcineurin to induce Fas ligand expression during activation-induced T cell death. J. Immunol. 163:5813-5819.
B binding motifs. EMBO J. 9:4435-4442.[Medline]
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»