INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Lymphocyte activation results in the transcription of a number of gene products that mobilize and maintain a functional immune response. Such gene products regulate cellular differentiation, effector function, and clonal expansion of the responding lymphocyte population. One such gene product is Fas (CD95), a cell surface protein known to induce apoptosis in activated lymphocytes upon binding to its cognate ligand (FasL). The essential roles of Fas and FasL in the maintenance of peripheral self-tolerance underscore their prominence as homeostatic regulators of the immune system (reviewed in reference 41). Within the T-cell compartment, Fas is expressed on immature thymocytes, but does not appear to be involved in thymic deletion (13, 17). Fas expression on resting peripheral blood lymphocytes is low to absent (26, 37, 41, 44). Primary T-cell activation results in upregulation of fas mRNA and protein followed by the gradual acquisition of sensitivity to Fas-mediated apoptosis (37, 44). Although the regulation of Fas-mediated apoptosis is complex, involving assembly of the intracellular proteins necessary to initiate the apoptotic cascade (7, 40) and an absence or low abundance of inhibitory proteins, such as FAP-1 (53), FLIP (22), bcl-2, and bcl-x_L (6, 23, 25), Fas receptor cross-linking is fundamentally required (24). Fas expression is maintained on activated lymphocytes for several weeks (44), and upon secondary exposure to antigen, FasL is upregulated (1, 14, 26). Autocrine or paracrine interactions of Fas and FasL result in the apoptotic elimination of responding CD4⁺ lymphocytes, a process termed "activation-induced cell death" (AICD) (1, 14, 26). AICD can also be induced in CD4⁺ T-cell lines by T-cell receptor (TCR) engagement or pharmacological mimics of TCR signaling, such as anti-CD3 antibodies or phorbol myristate acetate (PMA) plus ionomycin (P/I), respectively. In direct contrast, activated CD8⁺ T cells do not appear to undergo apoptosis as a result of Fas ligation, but rather are stimulated to undergo clonal expansion by reverse signaling through FasL (59). Thus, while Fas appears to be equivalently upregulated in both CD4⁺ and CD8⁺ T cells upon activation (37), the signals transduced by both Fas and FasL have dissimilar biological effects on the two lymphocyte subsets.

The transcriptional machinery controlling Fas expression is largely unknown. Genomic organization studies showed that human fas is a single-copy gene containing nine exons and eight introns, spanning approximately 25 kb (5). Sequences proximal to the Fas translational start site have a relatively high GC content (61%) between 590 and 1, a number of CpG dinucleotides (28 CpGs between 590 and 1), and an absence of conventional TATA and CAAT boxes (5). These are properties characteristic of a class of polymerase II-dependent promoters that include housekeeping genes (reviewed in reference 3) and other members of the tumor necrosis factor (TNF)-nerve growth factor receptor superfamily (28, 51, 52, 54). Although Fas is constitutively expressed on a variety of nonhematopoietic cells and hematopoietic cell lines (33, 43), multiple reports have documented Fas inducibility in response to such stimuli as UV irradiation, viral infection, wild-type p53, hypoxia, and chemotherapeutic agents (2, 34, 39, 45, 47, 61). To date, however, the molecular events regulating constitutive and inducible fas expression have not been elucidated.

In the present study, we have investigated the transcriptional machinery required for activation-dependent induction of Fas expression in the CD4⁺ Jurkat T-cell line. Although constitutively expressed on Jurkat cells, Fas can be markedly upregulated by P/I activation. Using deletion constructs of the fas promoter in luciferase reporter assays, we demonstrate the presence of a previously unrecognized, activation-responsive 47-bp sequence containing a composite binding site for both Sp1 and NF-B transcription factors. Characterization of this noncanonical site (located at positions 295 to 286) by electrophoretic mobility shift assay (EMSA) and supershift analyses revealed that Sp1 was constitutively bound to this site in untreated Jurkat cells, while p50-p65 NF-B heterodimers were bound after P/I activation. Site-directed mutagenesis of the critical guanine nucleotides in the B-Sp1 site, as well as inhibition of IB degradation or NF-B translocation, inhibited inducible Fas transcription. Taken together, our results indicate a critical requirement for NF-B translocation to the B-Sp1 site at positions 295 to 286 in activation-driven fas promoter induction.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Cell culture and cellular activation. The T-lymphoma cell line Jurkat-E6 was obtained from the American Type Culture Collection (Rockville, Md.) and cultured in complete medium (RPMI 1640 with 10% fetal bovine serum [FBS] and 2 mM L-glutamine) at 0.4 × 10⁶ to 0.7 × 10⁶ cells/ml. For activation studies, Jurkat cells were removed from culture and washed once with RPMI 1640, plated in 10 ml of complete medium, and rested overnight. The following day, cells were activated with PMA (50 ng/ml) and ionomycin (3 µg/ml) for the times indicated (Sigma, St. Louis, Mo.).

Flow cytometric analysis. Indirect antibody staining and flow cytometric analyses were carried out as described previously (43). Briefly, 10⁶ cells in 100 µl of Dulbecco's phosphate-buffered saline (D-PBS) containing 2% FBS and 0.1% sodium azide (fluorescence-activated cell sorter [FACS] buffer) were incubated with either 150 ng of UB2 (Kaimaya Biomedical Co., Tukwila, Wash.) or isotype-matched, control mouse anti-human immunoglobulin G1 (IgG1) antibodies (Sigma) at 4°C for 30 min. Samples were washed twice in FACS buffer, and phycoerythrin-conjugated goat anti-mouse IgG secondary antibodies were added in 100 µl of FACS buffer for 30 min at 4°C. Samples were washed twice with FACS buffer, fixed overnight in PBS containing 1% paraformaldehyde, and analyzed with a FACScan (Becton Dickinson and Co., Mountain View, Calif.). Ten thousand cells were analyzed for each sample, with the gate set to exclude dead cells and debris.

RNA isolation and RPAs. At the indicated time points, P/I-treated Jurkat cells were pelleted and washed once in D-PBS. Control Jurkat cells were similarly cultured but were not treated with P/I. Total RNA from 10⁷ cells per sample was extracted with the RNeasy kit (Qiagen, Chatsworth, Calif.) according to the manufacturer's instructions, and RNase protection assays (RPA) were performed with the HybSpeed RPA kit (Ambion, Austin, Tex.). Samples (25 µg of total RNA) were hybridized to a riboprobe spanning Fas exons 3 to 6 (cDNA sequence 510 to 750) and a control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) riboprobe. Following RNAse digestion, protected products were boiled, separated by electrophoresis on a 5% urea-polyacrylamide gel electrophoresis (PAGE) gel, and analyzed by autoradiography. Scanning densitometry (Molecular Dynamics, Sunnyvale, Calif.) was used to quantitate the fas and GAPDH band intensities of each sample. Relative fold induction was calculated by dividing each normalized fas/GAPDH ratio against the zero time point.

Generation of fas promoter constructs and site-directed mutagenesis. A human placental genomic phage library (Clonetech, Palo Alto, Calif.) was screened with a PCR-generated, [-³²P]dCTP-labeled Fas cDNA probe containing 5' untranslated exon 1 and 2 sequences: forward, 5' GGACCCGCTCAGTACGGAGTT; reverse, 5' TTCACCTGGAGGACAGGGCTTATG (Life Technologies, Grand Island, N.Y.). Plaque hybridizations were carried out as previously described (52), and filters were washed sequentially in 6× SSC (0.9 M NaCl, 0.09 M sodium citrate)-0.1% sodium dodecyl sulfate (SDS) at 42°C for 20 min, 2× SSC-0.1% SDS at 60°C for 20 min, and 1× SSC-0.1% SDS for 20 min before autoradiography. Approximately 1.35 × 10⁶ plaques were screened, followed by sequential hybridizations of the positive clones. Subsequent subcloning of SacI or HindIII inserts into pGEM11Zf(+) (Promega, Madison, Wis.) identified a 2.16-kb HindIII fragment containing 1.74 kb proximal to the Fas translational start site (FasH3). Sanger dideoxy sequencing (U.S. Biochemical Corp., Cleveland, Ohio) confirmed that FasH3 contained 5' flanking sequences identical to those previously reported for human fas (5).

429

366

305

259

19

Transient transfections and reporter assays. A DEAE-dextran-chloroquine mixture was used for transient transfection of Jurkat cells. Briefly, 5 × 10⁶ to 7.5 × 10⁶ exponentially growing cells were washed and resuspended in 1 ml of transfection solution (RPMI plus 250 µg of DEAE-dextran per ml, 0.1 M Tris [pH 7.2], and 0.1 mM chloroquine) together with 10 µg of reporter construct and 10 ng of the pRL-tk Renilla normalizing luciferase vector (Promega). The cells were incubated for 1.5 h at 37°C, washed, and plated in 10 ml of complete media. After a 20-h recovery, transfected cells were activated with P/I for 8 h. Cell extracts and luciferase reagents were prepared with the Promega Dual-Luciferase Reporter Assay System. Both firefly and Renilla luciferase activities were monitored with a luminometer (Turner Industries, Sunnyvale, Calif.). Normalized reporter activity is expressed as the firefly luciferase value divided by the Renilla luciferase value. Relative fold induction is calculated as the normalized reporter activity of the test sample divided by either the untreated FPR-Luc or 5-Luc reporter construct, as indicated below. The Rous sarcoma virus promoter-driven IB construct (RSV-IB) and vector control were obtained from Tse-Wa Tan (Baylor College of Medicine, Houston, Tex.).

Nuclear extract preparation, EMSAs, supershift analyses, and displacement studies. Nuclear extracts were prepared by a modified method of Dignam et al. (15). In brief, 7.5 × 10⁶ Jurkat cells were washed in PBS and pelleted before resuspension in 200 µl of cold buffer A (10 mM HEPES (pH 7.9), 50 mM NaCl, 1 mM dithiothreitol, 0.1 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride [PMSF]) and incubated for 20 min on ice prior to the addition of 200 µl of cold buffer B (buffer A with 0.1% Nonidet P-40). Cells were gently pipetted and returned on ice for another 20 min. Nuclei were pelleted (5,000 × g, 2 min), washed in buffer A, and pelleted again, and nuclear proteins were extracted in 25 µl of buffer D (400 mM NaCl, 20 mM HEPES (pH 7.9), 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol 1 mM PMSF). The tubes were iced for at least 30 min, followed by centrifugation at 4°C for 10 min. The supernatant was recovered, snap-frozen in liquid nitrogen, and stored at 80°C prior to use.

306

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Expression of Fas mRNA and protein is upregulated in Jurkat cells by pharmacological agents that mimic T-cell activation. While resting T cells express low to undetectable levels of the Fas protein (26, 37, 41, 44), it is rapidly upregulated upon exposure to specific antigens, mitogens, cytokines, TCR antibodies, and various pharmacological agents such as P/I (6, 37, 44). The Jurkat T-cell line has an activated T-cell phenotype, in that it constitutively expresses cell surface Fas (Fig. 1A). Because we were interested in the transcriptional mechanisms controlling Fas expression, we asked whether Fas could be upregulated in Jurkat cells upon activation. To this end, Jurkat cells were stimulated with P/I for various times, and mRNA and protein expression was determined with RPAs and FACS analyses, respectively. As shown in Fig. 1A, Jurkat cells transiently upregulated surface Fas after P/I stimulation. Increased Fas was detectable as early as 4 h poststimulation, reached maximal levels by 8 h, and declined to near-baseline levels by 24 h. A 1.5- to 2.0-fold increase in both the percentage of cells expressing Fas as well as the mean fluorescence intensity (MFI), representative of antigenic density on a per-cell basis, was observed. Correspondingly, as shown in Fig. 1B, Fas mRNA levels were upregulated by 10-fold in Jurkat cells activated for 2 h; mRNA levels declined subsequently over the time period examined (24 h). These results demonstrate an immediate-early, activation-dependent upregulation of Fas in P/I-stimulated Jurkat cells, suggesting a likely transcriptional control.

View larger version (26K): [in this window] [in a new window]   FIG. 1.   Time course induction of Fas upregulation in P/I-treated Jurkat cells. (A) Cells were harvested at the indicated time, and Fas expression was analyzed by flow cytometry. Isotype-matched control antibody staining is represented as a dashed line, and Fas staining is represented as solid lines. The percentage of specific Fas staining and the mean fluorescence intensity (MFI) of Jurkat cells at each time point are shown to the right. (B) Cells were harvested at the indicated time, total RNA was isolated, and Fas mRNA accumulation was determined by RNAse protection analysis. Scanning densitometry was used to normalize Fas mRNA levels and to calculate the relative fold induction compared to that of the untreated, time zero sample.

A 220-bp region in the Fas promoter is required for activation-dependent upregulation. To test the premise that activation-dependent Fas upregulation was modulated at the level of transcription, approximately 1.7 kb of the 5' region flanking the fas translational start site (nucleotides 1739 to 19) were isolated from a human placental genomic library and subcloned into the pGL2-Basic luciferase reporter vector in both orientations relative to the luciferase gene. The basal activity of the FPR1-Luc(+) construct was first determined by transient transfections into various T- and B-cell lines, followed by extract preparations and luciferase activity measurements. As shown in Fig. 2A, the FPR1-Luc(+) construct demonstrated basal activity in the Jurkat (13-fold over empty vector), Molt-4 (32-fold), and HSB (14-fold) T-cell lines. In B-cell lines, modest basal activity was observed only in Raji cells (fivefold over empty vector alone), but not in Daudi or Ramos cells. These results document that the 1.7-kb region upstream of the fas ATG contains a functional promoter operative in T-cell lines and certain B-cell lines.

View larger version (28K): [in this window] [in a new window]   FIG. 2.   Basal and inducible fas promoter activity and in vitro EMSA analysis. (A) FPR1-Luc(+) containing the sequence from 1739 to 19 is active in T-cell lines and the B-lymphocyte line Raji, but inactive in Daudi or Ramos cells. Cells were transfected with 10 µg of either the empty reporter (pGL2B []) or the fas promoter reporter vector (FPR1-Luc []); 10 ng of pRL-tk vector was cotransfected for normalization. Extracts were taken after 40 h, and luciferase activity was determined. Relative fold activity was calculated as the normalized FPR1-Luc(+) activity divided by the normalized pGL2B value. (B) fas promoter sequences 460 to 240 are required for inducible fas reporter activity. FPR1-Luc in both orientations (forward and reverse), 5-Luc, 6-Luc, tk-Luc, and 5/6-tk-Luc constructs were transiently transfected into Jurkat cells with pRL-tk for normalization, and the cells were then left untreated () or were activated with P/I () for 8 h before harvesting of the extracts for measurement of luciferase activity. The values shown are averages of three independent experiments (± standard deviation) and are expressed as relative fold induction over normalized luciferase values from untreated Jurkat cells containing FPR1-Luc(+). For tk-Luc and 5/6-tk-Luc, the relative fold induction was calculated separately as the normalized luciferase value divided by the untreated, normalized tk-Luc luciferase value. (C) Specific protein complexes bind the enhancer region between 460 and 236. 32P-labeled probe A (50,000 cpm) was mixed with nuclear extracts from untreated or P/I-activated Jurkat cells (left panel, lanes 1 to 3). Specific complexes (C1 and C2) are indicated. N.S., nonspecific complexes. Cold competition mixtures with a 25-fold excess of the indicated consensus elements (right panel, lanes 4 to 7) were mixed with extracts from Jurkat cells activated with P/I for 1 h prior to addition of labeled probe A.

Activation-dependent Fas promoter induction requires nucleotides 306 to 260. To analyze the region between 460 and 240 in further detail, sequential deletions of 5-Luc spanning 40- to 60-bp intervals were generated by PCR. The precise sequences contained in the various constructs are shown in Fig. 3A. Basal and inducible promoter activities in transiently transfected, unactivated, and P/I-stimulated Jurkat cells, respectively, are shown in Fig. 3B. Deletion of nucleotides 460 to 306 and 259 to 240 had little effect on the activation-dependent induction of the fas promoter (5 to 5.7, 5.8 to 6), ruling out the contribution of the consensus NF-B sites located at 421 to 412 and 252 to 243, the GAS element at 394 to 386, and the NF-ATp-AP-2 site at positions 358 to 345. When nucleotides 306 to 260 (5.7 to 5.8) were deleted, however, inducible reporter activity was reduced to that observed with the 6 construct. The sequences between 306 and 260 were devoid of any mapped transcription factor consensus sites, except for a B-like motif present at positions 295 to 286. This motif contains the sequence GGGCGTTCCC and differs from the derived consensus B p50-p65 heterodimer motif GGGRNNYYCC (20, 30, 48) at position 4 (change of R to C). Notably, however, the critical G_1-3 nucleotides necessary for p50 binding are maintained. Comparison of this motif to a consensus Sp1 site, KRGGCGKRRY (35), also revealed homology at 8 of 10 residues, indicating a possible interaction with this transcription factor. Taken together with results from our previous studies showing that complexes that could be competed with Sp1 and NF-B oligonucleotides bound to the fas promoter region at 406 to 236, it seemed likely that these transcription factors, acting alone or in combination, might be required for activation-dependent fas promoter induction.

View larger version (31K): [in this window] [in a new window]   FIG. 3.   Localization of activation-dependent fas promoter activity. (A) The fas nucleotide sequence between 460 and 230 is shown with known consensus transcription factor elements indicated (shaded areas). The location and designation of the various deletion constructs are indicated by the solid arrow. (B) P/I inducibility of the fas promoter localizes between nucleotides 306 and 260. Jurkat cells transiently transfected with 10 µg of the various deletion constructs and 10 ng of pRL-tk as the normalizing vector were left untreated () or activated with P/I () for 8 h before extracts were prepared for luciferase activity measurements. The values shown are averages of three independent experiments ± standard deviation and are expressed as the relative fold induction of the various deletion constructs over the untreated, normalized 5-Luc luciferase value.

Sp1 and NF-B conjointly bind at the B-Sp1-like motif (295 to 286) during activation-driven fas promoter induction. To further investigate the transcription factors involved in fas promoter upregulation in P/I-activated Jurkat cells, EMSA analyses of nuclear extracts were performed with a 29-bp probe spanning the noncanonical B-Sp1-like motif (306 to 278; probe B). In concert with our previous observations with probe A, spanning nucleotides 460 to 236 (Fig. 2C), we observed constitutive binding of the C2 complex in both unactivated and P/I-stimulated extracts and inducible binding of the C1 complex upon activation (Fig. 4A). C1 was induced de novo within 15 min of P/I addition (Fig. 4A, lane 2) and persisted for up to 2 h postactivation (Fig. 4A, lane 5). The specificities of C1 and C2 binding to probe B were first demonstrated by cold-competition assays using excess unlabeled probe B (Fig. 4A, lane 9). Jurkat extracts were coincubated with radiolabeled probe B and 25-fold excesses of either unlabeled AP-1- or NF-AT-specific oligonucleotides or a mutated probe B (GGG to TTT [see below]). Under these conditions, no displacement was observed in either C1 or C2 binding complexes (Fig. 4A, lanes 7, 8, and 10). To investigate Sp1-NF-B binding to probe B, Jurkat extracts were incubated with labeled probe B and increasing concentrations of unlabeled NF-B consensus oligonucleotides. With a 25-fold excess, NF-B oligonucleotides effectively competed the inducible C1 complex, but not the C2 complex. However, at 100-fold excess, both C2 and nonspecific binding were inhibited (Fig. 4A, lanes 11 to 12). Increasing concentrations of Sp1 oligonucleotides, on the other hand, specifically competed the constitutive C2 complex, but not the inducible C1 binding (Fig. 4A, lanes 13 and 14).

View larger version (29K): [in this window] [in a new window]   FIG. 4.   In vitro analysis of a potential B-Sp1 site at 295 to 286. (A) Time course analysis of complex binding to probe B (306 to 278). Jurkat cells were activated with P/I for the indicated time, nuclear extracts were harvested, and EMSA analysis was performed with 32P-labeled probe B (left panel, lanes 1 to 5). Cold competition was performed with 1-h-activated Jurkat cell nuclear extracts with a 25-fold excess of consensus oligonucleotides for the transcription factors AP-1, NF-AT, wild-type probe B (W-T), or mutated probe B (Mut.) added prior to incubation with 32P-labeled probe B (lanes 7 to 10). For NF-B and Sp1 competition, a 25-fold excess (lanes 11 and 13) and 100-fold excess (lanes 12 and 14) of consensus oligonucleotides were added to equivalent amounts of P/I-activated nuclear extracts prior to addition of 32P-labeled probe B. N.S., nonspecific binding. (B) Supershift analyses of P/I-activated Jurkat cell nuclear extracts. Nuclear extracts were subjected to EMSA analysis in the absence (lane 1) or presence of the indicated specific antisera (lanes 2 to 8), or preimmune IgG (lane 9). With two different anti-Sp1 antisera, probe B binding to the upper C2 complex was shown to be either shifted or inhibited (lanes 7 and 8). Arrows indicate the identity of the shifted complexes. (C) Mutual exclusive binding of Sp1 and NF-B p50 to probe B. Recombinant Sp1 (4 fpu [120 ng]) was preincubated with 12,500 cpm of radiolabeled probe B in the absence (lane 2) or presence of increasing amounts (0 to 40 ng) of recombinant NF-B p50 (lanes 3 to 6) prior to EMSA analysis. Brackets indicate the Sp1-probe B and p50-probe B complexes.

Mutually exclusive binding of Sp1 versus NF-B at nucleotides 295 to 286. To assess whether binding of Sp1 and NF-B to the 295 to 286 site in the fas promoter occurs concurrently or as a mutually exclusive event, EMSA analyses were performed with probe B with recombinant Sp1 and increasing amounts of recombinant p50. Consistent with our observations using extracts from unactivated Jurkat cells, a single complex is observed when probe B is mixed with Sp1 alone (Fig. 4C, lane 2). As increasing amounts of recombinant p50 are added, however, the intensity of the Sp1 complex is decreased, with a concomitant increase in the lower p50 binding complex (Fig. 4C, lanes 3 to 6). These results indicate that the fas promoter region at 295 to 286 contains a single B-Sp1 binding motif to which either Sp1 or NF-B is exclusively bound. Upon cellular activation, constitutively bound Sp1 is likely displaced by NF-B to drive fas promoter upregulation.

NF-B recruitment regulates activation-dependent Fas promoter induction. Although the above in vitro analyses document that NF-B heterodimers are rapidly recruited to the fas promoter region controlling activation-dependent induction, these studies do not address the functional consequences of the effects of NF-B recruitment on fas promoter activity. To investigate the role of NF-B recruitment in activation-driven fas upregulation, the effects of various inhibitors of NF-B function were examined. Jurkat cells transiently transfected with the 5-Luc reporter were pretreated with two pharmacological inhibitors of NF-B and then activated with P/I, and luciferase activity was determined. As shown in Fig. 5A, cyclosporin A pretreatment of Jurkat cells completely prevented P/I-dependent upregulation of the fas promoter. Because this calcineurin inhibitor can inhibit NF-AT translocation and disrupt Ca²⁺ signals leading to NF-B induction (36), the fungal metabolite gliotoxin, a specific inhibitor of IB- degradation and NF-B translocation, was also used (46). Gliotoxin, but not the inactive gliotoxin derivative bis-dethio-bis(methylthio)gliotoxin, completely abolished P/I inducibility of the 5-Luc reporter, consistent with the notion that NF-B recruitment was functionally required for activation-dependent fas promoter induction. The critical requirement for NF-B recruitment was additionally verified in Jurkat cells overexpressing wild-type IB-. The overexpression of IB-, a natural inhibitor of NF-B translocation, has been shown to effectively block B-dependent transactivation (31, 32). As shown in Fig. 5B, P/I-activated Jurkat cells transiently transfected with the 5-Luc reporter and RsV-IB- expression plasmid showed a dose-dependent decrease in fas promoter activity with increasing IB- concentration. Taken together, these results strongly support the premise that NF-B translocation is essential for activation-dependent fas promoter induction.

View larger version (15K): [in this window] [in a new window]   FIG. 5.   Inhibition of activation-dependent fas promoter upregulation. (A) Pharmacological inhibition of 5-Luc induction by P/I. Jurkat cells transiently transfected with 10 µg of 5-Luc and 10 ng of pRL-tk were preincubated in the absence (None) or presence of cyclosporin A (CsA [1 µg/ml]), gliotoxin (GT [1 µg/ml]), or bis-dethio-bis(methylthio)gliotoxin (DGT [3 µg/ml]) for 30 min prior to P/I stimulation for 8 h. Following incubation, cells were harvested for extract preparation, and luciferase activity was determined. The values shown are representative of two independent experiments ± standard deviation and are expressed as the relative fold induction over normalized untreated 5-Luc activity. (B) Dose-dependent inhibition of P/I-inducible 5-Luc reporter activity by IB. Increasing amounts of wild-type RsV-IB () or vector control () expression plasmid were cotransfected into Jurkat cells with 10 µg of 5-Luc and 10 ng of pRL-tk 40 h prior to P/I activation. Cells were harvested for extract preparation after 8 h of P/I exposure, and luciferase activity was measured. The values shown are averages from two independent experiments ± standard deviation and are expressed as percent inhibition of P/I-induced, normalized 5-Luc reporter activity.

The B-like motif at 295 to 286 controls activation-dependent fas promoter responses. While previous experiments indicated that NF-B translocation was essential for fas promoter induction, the functional significance of the B element at positions 295 to 286 remained to be tested. To determine the specific contribution of this B-Sp1 element, oligonucleotides were synthesized with the composite site altered (as underlined) (GGGCGTTCCC to TTTCGTTCCC) and with a mutagenized 5 fas reporter construct generated by PCR. This construct was designated 5M5.7-Luc. As a control, a second 5 reporter construct containing the B site at 262 to 253 was also mutagenized (5M5.8; GGGGCTTTTC to GCCYCTTTTC) and used in reporter assays. Jurkat cells transiently transfected with the wild-type 5-Luc construct, 5M5.7-Luc, or the 5M5.8-Luc control were activated with P/I, and reporter activity was measured. Similar to previous experiments with the 5 deletion constructs (Fig. 3B), both wild-type 5-Luc and 5M5.8-Luc showed an approximate 15-fold induction following activation (Fig. 6). In contrast, 5M5.7-Luc reporter activity increased only threefold after P/I treatment. These results document the functional contribution of the B-Sp1 site at positions 295 to 286, but not the consensus NF-B site at 262 to 243, in activation-dependent fas promoter induction. Interestingly, while 5M5.7-Luc induction was significantly suppressed compared to that of the wild-type 5-Luc construct (79.6% ± 10.61%, respectively) P/I inducibility was not entirely abolished. Such findings suggest that an additional site or sites, perhaps in the region from 19 to 240 (compare Fig. 3B to Fig. 6), may be required for maximal fas promoter induction during T-cell activation.

View larger version (23K): [in this window] [in a new window]   FIG. 6.   The B-Sp1 enhancer element is required for P/I-inducible 5-Luc reporter activity. (A) Nucleotide mutagenesis of 5-Luc at either the B-Sp1 element (GGGCCC; 295 to 293 [5M5.7-Luc]) or the downstream NF-B element (GGGCCY; 252 to 249 [5M5.8-Luc]) was introduced by PCR. Wild-type or mutagenized constructs were cotransfected into Jurkat cells with 10 ng of pRL-tk, and the cells were activated with P/I for 8 h prior to being harvested for extract preparation and luciferase activity measurements. The values shown are averages of three independent experiments ± standard deviations and are expressed as the relative fold increase over untreated, normalized 5-Luc luciferase activity. (B) Concatemer constructs containing either none (tk-Luc), two (/S2tk-Luc), four (/S4tk-Luc), or six (/S6tk-Luc) copies of the B-Sp1 element (295 to 286) were juxtaposed upstream of a thymidine kinase-driven luciferase reporter and cotransfected into Jurkat cells as mentioned above before activation with P/I and luciferase measurements. Inhibition of NF-B activity were analyzed by preincubation of transfected cells with either gliotoxin (GT [1 µg/ml]) or bis-dethio-bis(methylthio)gliotoxin (DGT [1 µg/ml]) as described in the legend to Fig. 5 prior to activation with P/I. Relative fold induction is calculated against the untreated, normalized tk-Luc reporter luciferase activity. (C) Specific inhibition of NF-B-dependent upregulation of the B-Sp1 motif concatemers by coexpression of RsV-IB. Multimerized constructs were cotransfected with 5 µg of either RsV-IB or vector-only expression plasmids before activation with P/I as described above. Extracts were taken for luciferase measurements, and relative inhibition is calculated as the percentage of P/I-induced luciferase activity of the corresponding reporter constructs.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

In recent years, much attention has been focused on how the immune response is downregulated. Because self-reactive T cells are ineffectively deleted by negative selection in the thymus, the preservation of immune homeostasis requires self-regulatory mechanisms for the maintenance of tolerance and the control of excessive T-cell proliferation to foreign antigens in the periphery. Apoptosis is responsible for maintaining T-cell homeostasis by at least two mechanisms-AICD that occurs after TCR ligation by antigen and passive cell death that occurs in the absence of antigen or cytokine stimulation (1, 14, 26, 63). The Fas-FasL-mediated pathway of apoptosis appears to be central in the regulation of peripheral tolerance, because mutations in either the receptor or the ligand can induce lymphadenopathy and autoimmune disease in mice and humans caused by defects in AICD and peripheral T- and B-cell deletion (18, 60, 65). Fas is constitutively expressed on most cultured T cells, whereas freshly isolated, naïve (CD45RO⁺) peripheral blood T cells have little to no surface Fas. Early after T-cell activation, Fas is rapidly upregulated, but these Fas-positive cells remain resistant to Fas-induced apoptosis (26, 37, 41, 44). With the progression of an immune response, Fas-positive T cells acquire sensitivity to Fas-induced killing (37, 44). Although Fas upregulation is known to be a prerequisite for the elicitation of Fas-FasL-mediated AICD (1, 14, 26), the molecular mechanisms underlying this process have not been well investigated. In this report, we have identified the immediate transcriptional events and cis-acting elements required for activation-induced fas expression in the human Jurkat cell line. Using transient transfection reporter assays, we have localized a 47-bp sequence (306 to 260) upstream of the translational start site that confers P/I inducibility of the fas promoter (Fig. 3B). Within this region, a critical 10-bp enhancer (295 to 286) was shown to bind Sp1 during basal transcription expression and NF-B p50-p65 heterodimers after P/I activation (Fig. 4). When binding of NF-B was inhibited by the blockade of B translocation or mutagenesis of the B-Sp1 site, activation-dependent Fas upregulation was lost. Our findings provide the first molecular evidence for differential control of constitutive and inducible Fas expression and suggest an intervention strategy for inducible Fas upregulation.

Sp1 is a zinc-finger transcription factor constitutively expressed in a variety of cell types that binds the GC-rich consensus sequences (KGGGCGGRRY or KRGGCGKRRY) present in many cellular and viral promoters. The N terminus of Sp1 contains glutamine- and serine/threonine-rich domains required for transactivation, while the C terminus is involved in synergistic activation and interaction with other transcription factors (12). Although Sp1 has been shown to independently initiate transcription in TATA-less promoters (reviewed in reference 3), it can also form homo- or heteromultimeric complexes to transactivate a number of disparate promoters, including interleukin 2, VCAM-1, and the HIV LTR (42, 50, 56). Several lines of circumstantial evidence suggest that Sp1 may be involved in basal fas transcription. First, sequences proximal to the fas translational start site have a relatively high GC content (61% between positions 590 and 1) and an absence of conventional TATA and CAAT boxes (5), typical of a subclass of polymerase II-dependent promoters controlled by the interactions of the glutamine-rich activation domains of Sp1 with TFIID components (9). Second, the TATA-less, GC-rich area of the fas promoter (590 to 1) has been shown to be functional in analyses of transcriptional start sites (5). Third, the TATA-less 5 fas promoter construct was sufficient in driving basal transcription. Fourth, Sp1 bound to the B-Sp1 site at positions 295 to 286 in resting, but not activated, Jurkat cell extracts (data not shown). Finally, multimers of the B-Sp1 element at positions 295 to 286 showed increase basal transcription in Jurkat cells (Fig. 6B). Although Sp1 may be involved in the control of fas transcription in resting Jurkat cells, it is likely that multiple Sp1 binding elements are required for optimal basal levels of Fas expression. In this regard, site-directed mutagenesis of the B-Sp1 site at 295 to 286 alone inhibited basal fas transcription by ~20%, while disruption of a second consensus Sp1 site at 140 to 131 repressed basal fas transcription by greater than 50% (data not shown).

Nuclear translocation of the NF-B-Rel transcription factor family occurs in T lymphocytes in response to TCR engagement (reviewed in reference 20). The prototypic form of NF-B is a heterodimer complex containing NF-B1-p50 or NF-B2-p52 in combination with a transactivating subunit such as c-Rel or RelA (p65) (reviewed in reference 64). Each NF-B-Rel family member contains a conserved N-terminal region responsible for decameric DNA binding, dimerization, and IB interaction; NF-B-responsive genes are transactivated by the C-terminal domain of dimeric NF-B-Rel family members. In resting T lymphocytes, various NF-B subunits are sequestered in the cytoplasm by virtue of their association with by the IB family of inhibitors (IB, IB, IB, IB, and bcl-3), which masks their nuclear localization signal (reviewed in reference 64). During T-cell activation, intracellular signaling triggers multiple kinase pathways that converge in the activation of the IB kinase complex (IKK); IKK then phosphorylates IB or IB at conserved serines, resulting in the targeting of the B inhibitor for ubiquitin-proteasome-dependent destruction and permitting the nuclear translocation of NF-B (10). Binding sites for NF-B have been found in numerous genes involved in effector T-cell function, including those regulating cytokine production, cell adhesion molecules, and apoptosis. In this report, we have demonstrated that NF-B p50-p65 mediates inducible fas promoter activation in T lymphocytes at 460 to 240 and confers the majority of P/I responsiveness at a single, noncanonical B-Sp1-like motif (GGGCGTTCCC) located at 295 to 286. Interestingly, although two consensus NF-B sites as well as GAS, NF-ATp, AP-2, and Egr-1 sites are also present within this region, the contribution of these elements to either basal or inducible fas promoter induction in Jurkat cells was insignificant (Fig. 3). Notably, however, such motifs may be functional in other cell types, because Egr-1 binding at 371 to 338 has been reported to repress fas transcription during B-cell activation (16).

The mutually exclusive binding of either Sp1 or NF-B to the novel composite B-Sp1 site at 295 to 286 in the fas promoter differs from the previously described cross-coupling of NF-B with the transcription factor AP-1 or HMG I(Y) (58, 62). Our observations are not unexpected, however, in light of reports documenting the functional interference of Sp1 and NF-B at selected NF-B binding sites (21) and crystallographic data showing that Sp1 and p50 homodimers both interact with DNA at the major groove (19, 38, 49). In this report, we provide multiple lines of evidence that NF-B p50-p65 heterodimers displace Sp1 at the composite B-Sp1 site to drive activation-dependent interaction of the fas promoter. First, P/I-dependent promoter induction requires the sequences between 306 and 260 containing the B-Sp1 site (Fig. 3B). Second, NF-B p50-65 heterodimers bound to this region of the fas promoter only during activation (Fig. 4A and B). Third, recombinant NF-B p50 displaced Sp1 binding to the B-Sp1 site in a dose-dependent manner (Fig. 4C). Fourth, recruitment of NF-B to the fas promoter region at 295 to 286 was required for activation-dependent fas promoter induction, but not for basal activity (Fig. 5). Fifth, specific mutagenesis of the NF-B site at 295 to 286 significantly diminished P/I-dependent fas promoter induction (Fig. 6A). Finally, multimers of the B-Sp1 site juxtaposed to a minimal thymidine kinase promoter were functionally responsive to P/I stimulation in a dose-dependent manner (Fig. 6B). In further support of our observations that NF-B is critically required for activation-induced fas upregulation, several inducers of NF-B, including UV irradiation, hypoxia, cytokines, and TNF, have been reported to rapidly upregulate Fas (24, 44, 61). It is interesting to speculate that Sp1, bound to the B-Sp1 site during basal transcription, may facilitate rapid activation-induced fas promoter responsiveness mediated by NF-B translocation as a consequence of its role in DNA bending (57).

If Fas is similarly regulated in Jurkat and fresh T cells, our findings may be relevant to HIV pathogenesis, in which increased Fas expression and sensitivity to FasL-induced apoptosis have been reported during infection and disease progression (2, 4, 27). Interestingly, both HIV-1 binding and gp120-mediated CD4 cross-linking can induce NF-B (8). Because NF-B can drive both HIV-1 expression (50) and fas upregulation, inhibition of NF-B may be one major strategy to suppress both viral replication (31) and premature execution of the AICD pathway resulting in T-cell depletion. Indeed, inhibitors of NF-B may also effectively repress Fas-driven apoptosis in response to metabolic stress, genotoxic insults, and viral infections.

Although our studies clearly demonstrate a critical role for NF-B in fas upregulation, other transcriptional or posttranscriptional processes may also be operative, depending upon the cell type and cellular microenvironment. For example, the proline-, glutamine-, and histidine-rich protein TDAG51 has been reported to be critical for Fas mRNA induction after TCR engagement (47). Our laboratory has demonstrated that wild-type p53 can upregulate fas expression in both hematopoietic and nonhematopoietic cells (45). Taken with our current observations, it is plausible that p53 may act with NF-B to increase fas transcription, particularly under conditions of metabolic and genotoxic stress. mRNA stability may also regulate fas expression. In this regard, the 3' untranslated region of fas contains several AU-rich elements reported to induce selective mRNA degradation (29, 55). Given the potent biological function of the Fas protein, multiple levels of gene regulation involving both cis-acting transcription and mRNA stability would not be unexpected.

In summary, our studies have identified a novel B-Sp1 enhancer motif in the fas promoter in which NF-B p50-p65 dimers bind to drive activation-dependent Fas upregulation in T cells. The critical involvement of NF-B further underscores the importance of this transcription factor in immunohomeostasis and prompts consideration for strategies aimed toward its manipulation in various pathophysiological states involving inappropriate Fas and FasL interactions, including malignant disease, autoimmunity, and AIDS.