Unexpected Characteristics of the IFN-γ Reporters in Nontransformed T Cells

Mice, reagents, cytokines, and Abs

DO11.10 αβ TCR-transgenic mice on wild-type and Stat4-deficient backgrounds have been described previously (27, 28). Female BALB/c mice were bred in our facility. CsA, PMA, and ionomycin were purchased from Sigma (St. Louis, MO). Recombinant human IL-2 was obtained from Takeda (Osaka, Japan), murine rIL-12 was purchased from Genetics Institute (Cambridge, MA), and murine rIFN-γ was obtained from Genentech (South San Francisco, CA). Murine rIL-4 was generated from transfected P815 mastocytoma cells as high titer culture supernatant. Monoclonal anti-mIFN-γ (H22) was obtained from Dr. R. D. Schreiber (Washington University, St. Louis, MO); monoclonal anti-IL-4 (11B11) (29) and anti-IL-12 (Tosh) (30) were previously described.

Retroviral constructs and retroviral transfection

A 3.5-kb murine IFN-γ promoter was provided by Dr. H. Fox (The Scripps Institute, La Jolla, CA) (31). A retrovirus-based reporter, hCD4-pA-GFP-RV, which uses a cytoplasmic truncated human CD4 (hCD4) to mark viral infection and green fluorescence protein (GFP) to report promoter activity, was constructed as follows. First, a polyadenylation site generated by PCR from the murine MHC class I K^b gene using EcoRI- and SalI-tailed oligonucleotides (5′pA-R1, GGAAATCGATTGAGAATGCTTAGAGGT; 3′pA-SalI, ACGCGTCGACCTGTTCACACTCAGCTG) was digested with EcoRI and SalI. A GFP cDNA was isolated from GFP-RV by NcoI digestion, blunting using Vent polymerase, and EcoRI digestion (32). GFP cDNA and the polyadenylation site were ligated, in a trimolecular reaction, into SmaI/SalI-digested pBSSK to produce the plasmid GFP-pA-BSSK. Next, CD4-RV, previously described (33), was modified by removing the internal ribosomal entry site located upstream of CD4 by digestion with BglII and NcoI, followed by blunting with Klenow and religation to produce the vector hCD4-only-RV. Then, the GFP/poly(A) cassette was removed from GFP-pA-BSSK by XbaI/XhoI, blunted with Klenow, and ligated into the Klenow-blunted BamHI site of CD4-only RV to generate the vector hCD4-pA-GFP-RV. The proper orientation of the vector was confirmed by restriction mapping, PCR, and sequencing (Fig. 3⇓).

Cytokine promoters were cloned as blunt fragments into the blunted HindIII site of hCD4-pA-GFP-RV as follows. To generate IFN-γ promoter regions by PCR, the oligonucleotide 3A-SalI, annealing to the IFN-γ promoter at +36 to +17, and the following oligonucleotides, annealing to indicated upstream IFN-γ promoter regions, were used as primers with the genomic IFN-γ template (31): 3A-SalI, ACGCGTCGACTGTCTTCTCTAGGTCAGCCG; −586 (D), GCGAAGCTTCACGTTGACCCTGAGT; −350, CCTGTGCTGTGCTCTGTGGATG; −247, TGCTTTCAGAGAATCCCACAAG; −194, CATCGTCAGAGAGCCCAAGGAG; −134, TAATGCAAAGTAACTTAGCTCC; −115, TCCCCCCACCTATCTGTCACCATCTTAA; −80, AAACCAAAAAAAAACTTGTGAAAATACGTAATCCCG; and −59 (H), GCGAAGCTTAAAATACGTAATCCCGAG. The distal element reporter was generated using the following primer sets: −134, TAATGCAAAGTAACTTAGCTCC; and distal-AS, TTTTTTTTTTTTTAAGATGGTGACAGATAGG.

The internal deletion of the GATA and AP-1 cis elements of hCD4-pA-GFP-RV-(−134) was generated using the QuickChange PCR-based strategy (Stratagene, La Jolla, CA) with the following oligonucleotides: distal-GATA-del-S, ATGCAAAGTAACTTAGCTCCCATCTGTCACCATCTTAAAAAAA; distal-GATA-del-AS, TTTTTTTAAGATGGTGACAGATGGGAGCTAAGTTACTTTGCAT; distal-AP-1-del-S, CTTAGCTCCCCCCACCTATCAAACCAAAAAAAAACTTGTGAAAATACGTA; and distal-AP-1-del-AS, TACGTATTTTCACAAGTTTTTTTTTGGTTTGATAGGTGGGGGGAGCTAAG. The 2.1-kb IL-2 promoter region was excised from IL-2-Luc (34) as a PstI fragment, Klenow-bunted, and cloned into the Klenow-blunted HindIII site of hCD4-pA-GFP-RV. The IL-4 promoter (34) was generated by PCR with the following primers: −742, CGCGGATCCGTGAATTCTCCACACTGATGCTG; and +66, CCCAAGCTTTAGCTCTGTGCCG. The oligonucleotide GFP-5′-AS internal to the GFP-coding region was used to sequence all promoter constructs: GFP-5′-AS, GTGAACAGCTCCTCGCCCTTGC.

The retroviral-based reporter, GFP-pA-rLuc-RV (Fig. 1⇓B), was constructed as follows. First the vector GFP-RV (32) was digested with BglII and NcoI to release the internal ribosomal entry site upstream of GFP, Vent-blunted, and religated to produce the intermediate GFP-only RV. The pRL-Null vector (Promega) was next modified by removing the SV40 late polyadenylation signal (which operates on both strands) by digestion with XbaI/BamHI and replacing it with the single-sided polyadenylation signal from the murine class I MHC K^b gene amplified by PCR from the genomic template with primers tailed with XbaI and BamHI sites to create the plasmid rLN-K^b-A. A BamHI fragment from rLN-K^b-A containing the Renilla luciferase-poly(A) cassette was inserted at the unique BamHI site downstream of GFP in GFP-only RV to produce GFP-pA-rLuc-RV. IFN-γ promoter regions made by PCR as described above were cloned as blunt fragments into the blunted ClaI site of GFP-pA-rLuc-RV. This vector uses GFP to mark retroviral infection and Renilla luciferase to report promoter activity.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

IFN-γ promoter is inducible by PMA/ionomycin in EL-4 lymphoma cells. EL-4 cells were transiently transfected with IFN-γ or IL-2 promoter reporter vectors inserted into pBS-Luc (A) as described previously (34 ) or GFP-pA-rLuc-RV (B) as described in Materials and Methods. After 12 h, cells were left untreated (None) or were stimulated with PMA/ionomycin (P/I) for 4 h, and luciferase activity was determined. CMV promoter-driven Renilla luciferase (A) or firefly luciferase (B) activity was used for normalization. Results were similar in three independent experiments.

Transient transfection and luciferase analysis

The pBS-Luc firefly luciferase-based reporter has been described previously (34). For transient transfections (Fig. 1⇑), EL-4 cells (10⁷) were electroporated with 20 μg IFN-γ reporter construct and 1 μg CMV-Renilla (32) luciferase construct or with 20 μg GFP-pA-rLuc-RV and 10 μg CMV-firefly luciferase (34) in 1.2 ml Iscove’s modified DMEM at (960 μF, 280 V) as described. At 12 h, cells were left untreated or were activated by 50 ng/ml PMA and 1 μM ionomycin as indicated for 4 h, and luciferase activity was measured as previously described (34). Firefly luciferase activity was normalized using the activity of Renilla luciferase of each determination as previously described (32).

Cell culture, cell sorting, ELISA, and FACS analysis

DO11.10 T cells were differentiated to Th1 and Th2 phenotypes, and retroviral infections were conducted as previously described (28, 35). Infected T cells were purified by cell sorting on day 7 after primary activation by expression of murine CD4 and either GFP (Fig. 2⇓) or human CD4 (hCD4). Cells infected with GFP-pA-rLuc-RV-base retroviral reporter constructs (Fig. 2⇓) were stained with PE-conjugated anti-murine CD4 (BD PharMingen, San Diego, CA) on ice for 30 min, washed twice, and purified by sorting for GFP and murine CD4 expression (Cytomation, Fort Collins, CO). Cells infected with hCD4-pA-GFP-RV retroviral reporters were stained with PE-conjugated anti-human CD4 (Caltag, Burlingame, CA) and CyChrome-conjugated anti-murine CD4 (BD PharMingen), purified by sorting for human and murine CD4 expression. Sorted T cells were expanded by weekly stimulation with OVA peptide (0.3 μM) at 2.5 × 10⁵ cell/ml and irradiated BALB/c splenocytes (2000 rad; 5 × 10⁶ cells/ml). Cells were harvested 7 days after the last activation, washed, and counted. To analyze cytokine production by T cells, cells were stimulated with PMA (50 ng/ml) and ionomycin (1 μM) for 24 h, and the supernatants were harvested to determine IFN-γ and IL-4 titers by sandwich ELISA as described previously (36). To analyze reporter expression in sort-purified T cells, cells were stimulated with PMA and ionomycin for 6 h and then stained with PE-conjugated anti-human CD4. After 30 min on ice, cells were washed twice with PBS, suspended in 500 μl PBS, and analyzed on a FACSCalibur (BD Biosciences, Franklin Lakes, NJ).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Integrated IFN-γ reporters show constitutive activity in Th1 and Th2 cells. A, Map of GFP-pA-rLuc-RV vector and constructs containing the −59 (GFP-pA-rLuc-RV-H) or −586 (GFP-pA-rLuc-RV-D) regions of the IFN-γ promoter in GFP-pA-rLuc-RV. B, DO11.10 T cells were activated with OVA and IL-12/11B11 (anti-IL-4) for Th1 or with IL-4/TOSH (anti-IL-12) for Th2 cell development for 2 days and were infected with either GFP-pA-rLuc-RV-H or GFP-pA-rLuc-RV-D retrovirus as indicated. GFP- and murine CD4-positive cells were sorted on day 7 and restimulated with OVA/APC for another 1 wk in the initial activation condition. Cells were left untreated (□, −) or were treated with PMA/ionomycin (▪, +), and luciferase activity was measured after 4 h. Data are the relative light units from 1 × 10⁶ cells. Similar results were obtained in two similar experiments.

EMSA

Nuclear extracts were prepared and EMSA was performed as previously described (26). Binding reactions consisted of 2 μg nuclear extract, 1 μg poly(dI-dC), 10 mM Tris-Cl (pH 7.5), 50 mM NaCl, 1 mM DTT, 1 mM EDTA, 5% (v/v) glycerol, and 5 × 10⁴ cpm ³²P-labeled probe in 20-μl reaction volumes. After incubation on ice for 30 min, the reactions were resolved by nondenaturing 4.5% PAGE for 2 h at 150 V, followed by autoradiography. The following double-stranded oligonucleotides were used: MN3, which is from −134 to −88 of the IFN-γ promoter, TAATGCAAAGTAACTTAGCTCCCCCCACCTATCTGTCACCATCTTAA; and Eα Y-box, TCGACATTTTTCTGATTGGTTAAAAGTC.

Generation of transgenic mice

The transgenic reporter construct contains a 3.4-kb IFN-γ promoter fragment cloned from a BALB/c genomic library. It contains the endogenous TATA box with the S65T-GFP cDNA inserted at the start site of translation of the first exon of IFN-γ, but with a modified Kozak consensus sequence, as previously described (37). Transgenic mice were generated as previously described, founders were identified by Southern analysis, and three independent lines were established, essentially as previously described (37). Founders were crossed onto the BALB/c background and bred with DO11.10 TCR-transgenic mice (27).

Inducible IFN-γ reporters during transient expression in transformed cells

To begin, we wished to test whether IFN-γ promoter fragments were active in transient transfection assays using standard firefly luciferase reporter pBS-Luc. Thus, we analyzed the expression of various IFN-γ reporters by transient transfection in EL-4 cells (Fig. 1⇑A). Transfected cells were stimulated with PMA/ionomycin for 4 h, and luciferase activity was determined (Fig. 1⇑A). Consistent with previous studies (10, 13), various IFN-γ promoter constructs ranging from −59 to −3489 of the transcription start site were inducible by PMA/ionomycin stimulation, with various constructs showing only slight differences in the levels of inducibility or expression. As a positive control, the IL-2 promoter-driven luciferase reporter was highly inducible by PMA/ionomycin stimulation (Fig. 1⇑A).

To allow for analysis of the IFN-γ promoter in primary T cells, we developed a bottom-strand reporter system based on delivery by recombinant retrovirus (Fig. 1⇑B). This retrovirus expresses GFP under the control of the murine stem cell virus long terminal repeat (LTR), while luciferase expression is controlled by the promoter inserted onto the retroviral bottom strand. We first tested these new retroviral report constructs in transient transfections of EL-4 cells to determine whether the promoters behave in the same manner as when contained in the pBS-Luc based luciferase vectors described above. When used in transient transfections, the retroviral IFN-γ promoter reporters were inducible by PMA/ionomycin and had generally similar levels of inducibility as in the pBS-Luc reporters. As a positive control, the IL-2 promoter was also highly inducible by PMA/ionomycin (Fig. 1⇑B). This indicates that the −59 to −1976 IFN-γ promoter regions in this novel retroviral construct function as expected when used in transient transfection of a transformed cell line.

Constitutive IFN-γ reporter activity in stably integrated nontransformed T cells

Next we asked how this retroviral reporter system would behave when stably integrated in primary CD4⁺ T cells. DO11.10 splenocytes were stimulated with OVA under Th1- or Th2-inducing conditions and infected with the GFP-pA-rLuc-RV based reporter viruses (Fig. 2⇑A) on day 2 after primary activation. On day 7, cells infected with reporter viruses were purified by cell sorting based on GFP and murine CD4 expression to >95% purity. Cells were restimulated with OVA and APC for 7 days, harvested, washed, and counted. Equal numbers of cells were stimulated with PMA/ionomycin for 4 h, and luciferase activity was measured. Surprisingly, cells infected with GFP-pA-rLuc-RV-D retrovirus showed a high constitutive luciferase activity (Fig. 2⇑B), which was not increased by PMA/ionomycin. In addition, the constitutive activation of the −586 IFN-γ promoter was not Th1 specific, but was expressed highly in both Th1 and Th2 cells (Fig. 2⇑B). Finally, the −59 IFN-γ promoter construct (GFP-pA-rLuc-RV-H) was not active in either Th1 or Th2 cells (Fig. 2⇑B), although it was inducible to the same level as the −586 construct in EL-4 cells (Fig. 1⇑B).

The retroviral reporter described in Fig. 2⇑ allows purification of retrovirally infected cells on the basis of GFP expression, but requires lysis of these cells to determine the activity of integrated promoter constructs. To generate a second, more convenient retroviral-based reporter system that could allow analysis at the single-cell level by FACS, we modified the retroviral reporter system, generating hCD4-pA-GFP-RV (Fig. 3⇓A). In this reporter system retroviral infection and purification are based on expression of the hCD4 extracellular and transmembrane domain lacking the cytoplasmic signaling domain. The activity of the test promoter within the reporter construct is determined from expression of GFP. Furthermore, because the −586 IFN-γ promoter regions behaved with unexpected constitutive expression, we also included analysis of IL-2 and IL-4 promoter constructs in this new retroviral reporter system as a verification of its capacity to permit normal inducible promoter activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Integrated IFN-γ reporters show constitutive activity at the single-cell level in Th1 and Th2 cells. A, Map of the hCD4-pA-GFP-RVreporter construct. Human CD4 was used for the infection marker, while GFP expression is under the control of the IFN-γ, IL-2, or IL-4 promoter. B, The retroviral constructs were transiently transfected into Phoenix E packaging cells. The viral supernatants were used to infect DO11.10 cell under Th1- or Th2-polarizing conditions. On day 7, human CD4 and murine CD4 double-positive cells were sorted and restimulated with OVA/APC for another week under the initial activation conditions. Cells were then harvested and stimulated with PMA/ionomycin. After 24 h of stimulation, supernatants were collected to determine IFN-γ and IL-4 titers by ELISA. C and D, hCD4-pA-GFP-RVretrovirus infected Th1 (C) or Th2 (D) cells were stimulated with PMA/ionomycin for 6 h, and the cells were stained with hCD4-PE. Numbers in the upper right quadrant represent the percentage of hCD4-positive/GFP-positive cells in the live cell gate.

Empty vector, various IFN-γ, IL-2, and IL-4 promoter regions placed into hCD4-pA-GFP-RV were used to generate infectious retrovirus as described previously (28) and were used to infect naive DO11.10 T cells activated under Th1- or Th2-inducing conditions. On day 7, retrovirally infected cells were purified by cell sorting on the basis of expression of hCD4 and murine CD4 and were re-expanded in culture for an additional 7 days. To verify that polarization of these populations had occurred as desired, T cells were stimulated on day 7 with PMA/ionomycin, and supernatants were examined by ELISA for IFN-γ and IL-4 production (Fig. 3⇑B). As expected, infection by these reporter constructs did not interfere with the cytokine-induced polarization of these cells, with all Th1-inducing conditions achieving high IFN-γ and low IL-4 production and, similarly, all Th2-inducing conditions leading to high IL-4, but low IFN-γ production (Fig. 3⇑B).

Next we analyzed the reporter activities of these various promoter regions in the purified Th1 or Th2 populations (Fig. 3⇑, C and D). Again, we analyzed two regions of the IFN-γ promoter extending to −59 (H) or −586 (D) as well as a 2.1-kb region of the IL-2 promoter and an 800-bp IL-4 promoter region. In this novel reporter system, activity is measured by expression of GFP, and infection is monitored by expression of hCD4. In Th1 cells, the IL-2 promoter was highly inducible by PMA/ionomycin treatment (Fig. 3⇑C, panels G and H), whereas the IL-4 promoter construct was not inducible (Fig. 3⇑C, panels I and J). Furthermore, the empty vector hCD4-pA-GFP-RV was inactive in expressing GFP (Fig. 3⇑C, panels A and B), as we expected. The −59 IFN-γ promoter region was inactive, driving <2% GFP-positive cells (Fig. 3⇑C, panels C and D), and consistent with the low luciferase activity observed previously in the luciferase-based retroviral reporter (Fig. 2⇑B). Furthermore, the longer IFN-γ promoter region extending to −586 bp was constitutively active, driving GFP expression in hCD4⁺ cells in both cells stimulated with PMA/ionomycin as well as cells left untreated (Fig. 3⇑C, panels E and F).

Furthermore, the activity of this −596 IFN-γ promoter region was equally active in cells driven toward the Th2 phenotype (Fig. 3⇑D, panels E and F). As controls, we found that the IL-4-containing promoter construct was significantly more active in Th2 cells (Fig. 3⇑D, panels I and J) than in Th1 cells (Fig. 3⇑C, panels I and J), indicating that this reporter system can be sensitive to subset-specific regulation of cytokine promoters. These results confirm the observations of constitutive reporter activity in the luciferase-based retroviral vector.

The first intron of the gene is also thought to contain regulatory sequences (38). We have analyzed large reporter constructs that contain the proximal promoter, first exon and intron, and part of the second exon fused to GFP, requiring splicing to generate the mature GFP reporter. To date, such constructs do not generate inducible reporter activity, perhaps due to their size or the added complexity of their splicing requirements, but they have prevented analysis of intronic sequences arranged in the native configuration with the promoter (data not shown).

Constitutive IFN-γ promoter activity in a transgenic reporter system

The constitutive activity seen here for the IFN-γ promoter, but not other cytokine promoters, could be due to a selective activation of the IFN-γ promoter by some component of the retroviral system. To test this possibility, we turned to the use of a transgenic reporter system (Fig. 4⇓). A 3.4-kb region of the murine IFN-γ promoter was used to drive a GFP reporter construct. In this construct, splicing and polyadenylation were provided by the SV40 intron and polyadenylation site, and T cell-specific expression was provided by inclusion of a downstream cassette containing the human CD2 locus control region (39) (Fig. 4⇓A). This construct is similar to our previous IL-2 reporter construct used in a transgenic study (37), which exhibited inducible GFP expression upon activation of transgenic T cells, indicating that this construct is capable of providing for T cell-specific reporter expression with appropriately inducible, rather than constitutive, activity in the context of transgenic mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Transgenic IFN-γ reporters show constitutive activity in naive, Th1, and Th2 cells. A, Map of the IFN-γP/GFP/CD2 transgene. B, Constitutive expression of the GFP reporter in DO11.10 × IFN-γP/GFP/CD2 double-transgenic mice. Spleen and lymph node CD4⁺ T cells isolated from Ag-naive DO11.10 IFN-γP/GFP or DO11.10 littermate control mice were stimulated with 10 μg/ml OVA peptide presented on RT11.B7-1 artificial APCs for 24 h and analyzed for expression of CD4 and GFP (Naive). Th1- and Th2-polarized cells were generated by culture for 7 days with irradiated BALB/c spleen-adherent cells and 10 μg/ml OVA peptide supplemented with 20 U/ml IL-12 and 10 μg/ml anti-IL-4 (Th1) or 1000 U/ml IL-4 and 10 μg/ml anti-IL-12 (Th2). T cells were recovered and restimulated for 18 h with 10 μg/ml OVA peptide presented on RT11.B7-1 fibroblasts and analyzed for expression of CD4 and GFP. The flow cytometric data are gated on lymphocytes, and the numbers in the upper right quadrants are percentages of CD4- and GFP-positive T cells. The data are from one of two similar experiments. Comparable results were obtained with three distinct IFN-γP/GFP/CD2 founder lines (data not shown). C, Lymph node cells isolated from DO11.10 IFN-γP/GFP mice were stained for murine CD4 and CD8 Abs. The CD4⁺, CD8⁺, and CD4/CD8-negative cells were gated and analyzed for GFP expression. The right panel is from the gated cells of the left panel. The number in the right panel is the percentage of GFP-positive cells.

Several transgenic lines were derived, propagated, and analyzed for reporter expression in T cells undergoing differentiation toward the Th1 or Th2 phenotype (Fig. 4⇑B). First, GFP expression is absent in nontransgenic littermate controls, as expected (Fig. 4⇑B, panels A and B). Interestingly, in the IFN-γ reporter transgenic, we observed GFP expression in even naive T cells of transgenic mice, suggesting a relaxed requirement for reporter expression. Moreover, GFP expression in naive cells was observed even in the absence of Ag-induced activation (Fig. 4⇑B, panel C) and was not significantly increased upon activation of these T cells with Ag (Fig. 4⇑B, panel D). Furthermore, when these cells were differentiated toward the Th1 phenotype, no substantial increase in the level of GFP reporter expression was observed, which was seen both in stimulated (Fig. 4⇑B, panel F) as well as unstimulated (Fig. 4⇑B, panel E) cells. Finally, similar levels of reporter expression were observed even in T cells driven toward the Th2 phenotype, in which, once again, reporter expression was constitutive rather than inducible by Ag (Fig. 4⇑B, panels G and H). Similar promoter behavior was seen in the other reporter lines as well. This constitutive and non-subset-specific expression of the IFN-γ promoter contrasts with the inducible behavior of the IL-2 promoter expressed using the same cassette (37). Furthermore, this constitutive behavior is contrasted with a previous IL-4 promoter construct analyzed in transgenic mice, in which the IL-4 promoter driving luciferase demonstrated highly inducible activity, rather than constitutive activity (40). Finally, we asked whether the GFP expression is T cell specific. Lymph node cells were stained for CD4 and CD8 expression by FACS (Fig. 4⇑C, left panel). Both CD4- and CD8-positive lymphocytes constitutively expressed GFP (Fig. 4⇑C, right panel, R2 and R3), while CD4- and CD8-negative lymphocytes were GFP negative (Fig. 4⇑C, right panel, R4). Similar results were found in splenocytes (data not shown). GFP expression was also negative in fibroblasts (data not shown).

Constitutive promoter activity resides in the distal IFN-γ promoter element, and the AP-1/GM-CSF site is critical for the activity

Based on the similarity with the results in the transgenic reporter system, we considered that the observed constitutive activity of the IFN-γ promoter regions in stable expression systems was not necessarily an artifact of the retroviral system. Therefore, we wished to identify the specific promoter regions responsible for this constitutive and non-subset-specific activity (Fig. 5⇓). To this end, we generated a series of IFN-γ promoter truncations using the hCD4-pA-GFP-RV retroviral reporter system (Fig. 3⇑A). In the first series, four additional constructs of intermediate length between D (−586) and H (−59) were used to infect CD4⁺ T cells undergoing Th1 differentiation. Retrovirally infected T cells were left untreated or were treated with PMA/ionomycin and analyzed for expression of hCD4 (to identify retrovirally infected cells) and GFP as an indication of reporter activity. Once again, each truncation including −350, −247, −194, and −134 showed activity similar to that of the original IFN-γ reporter construct (−586, D), which were all similar whether they were stimulated with PMA/ionomycin or left untreated.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

The distal IFN-γ promoter element is required for constitutive activity. Human CD4-pA-GFP-RV retrovirus-infected Th1 cells were stimulated with PMA/ionomycin for 6 h, and the cells were stained with hCD4-PE. Numbers in the upper right quadrant represent the percentage of hCD4-positive/GFP-positive cells in the live cell gate.

Therefore, the constitutive activity appeared to reside between −59 and −134 of the IFN-γ promoter. We tested an additional intermediate deletion, −80 (Fig. 5⇑B). Here, truncation of the region between −134 and −80 led to a complete loss of both induced and uninduced reporter activities, similar to the −59 promoter construct (H). As before, the IL-2 promoter as a control retained inducible activity as expected (Fig. 5⇑B, panel P). These results indicated that the constitutive activity of the IFN-γ promoter constructs in these retroviral reporters was due to an ∼50-bp region between −134 and −80 relative to the start site of transcription.

The region between −134 and −80 contains the previously described distal IFN-γ promoter element (10). To confirm whether this region could interact with nuclear factors, we isolated nuclear extracts from Th1 and Th2 cells, either left untreated or treated with PMA/ionomycin, and conducted EMSA (Fig. 6⇓). Consistent with the constitutive activity resulting from this region, we observed a constitutive complex in EMSA binding the region spanned by the MN3 probe (Fig. 6⇓). This complex was present in both Th1 and Th2 cells and was slightly augmented upon treatment with PMA/ionomycin. An inducible band of relatively high mobility was seen.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

The distal promoter element binds similar factors in Th1 and Th2 cells. Nuclear extracts were prepared after Th1 and Th2 cells were stimulated with PMA/ionomycin for 3 h. EMSA was performed with the MN3 probe containing distal IFN-γ promoter element probe or with the Eα probe as a control for the quality and quantity of the extract.

The distal IFN-γ promoter element contains three cis elements reported to bind Oct1, GATA, and AP-1/GM-CSF (10, 41). To test which of these cis elements is required for constitutive promoter activity, we generated several internal deletions affecting these sites. Deletion of Oct1 or GATA site did not significantly change the constitutive activity, while the deletion of AP-1/GM-CSF site significantly reduced the promoter activity (Fig. 7⇓). Finally, the distal element in isolation did not contain constitutive activity (Fig. 7⇓), implying that it requires cooperation with other downstream promoter elements. As a control, the −134 IFN-γ promoter containing both proximal and distal elements was constitutively active, and the −80 IFN-γ promoter, lacking the distal element, was inactive. By comparison, the IL-2 promoter was inducible by PMA and ionomycin, but did not show constitutive activity.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

The AP-1/GM-CSF binding site in the distal promoter element is critical for the constitutive activity. The internal deletion of cis elements was generated by the QuickChange PCR-based strategy using specific primers. Th1 cells were infected and stimulated with PMA/ionomycin as described in Fig. 3⇑. The left panel is the diagram of the constructs, and the right panel represents the percentage of hCD4- and GFP-positive cells after stimulation with or without PMA/ionomycin.

CsA-resistant, Stat4- and p38 mitogen-activated protein kinase (MAPK)-independent activity of the IFN-γ promoter

Because CsA inhibits TCR or PMA/ionomycin-induced IFN-γ production, we asked whether it could inhibit the constitutive activity of the IFN-γ promoter. Th1 cells were infected with different retrovirus constructs, sorted, and restimulated as described above. Some cells were incubated with CsA on day 5 after restimulation. On day 7, T cells were harvested and were either left untreated or were treated with PMA/ionomycin in the presence or the absence of CsA. The empty vector hCD4-pA-GFP-RV-infected T cells did not express GFP (Fig. 8⇓A, panels A–D), as expected. As the positive control, IL-2 promoter was inducible (Fig. 8⇓A, panels E and F), whereas the inducibility of the IL-2 promoter was inhibited by CsA (Fig. 8⇓A, compare panels F and H). However, the constitutive activity of the IFN-γ promoter was not inhibited by CsA (Fig. 8⇓A, compare panels I and J with K and L).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Constitutive activity of IFN-γ reporter activity is CsA resistant and Stat4- and p38 MAPK-independent. DO11.10 Th1 (A and C) or Stat4^−/− Th1 (B) cells were infected with the indicated retrovirus reporter constructs and sorted for hCD4 and murine CD4 expression. On day 5 after restimulation, CsA or SB 203580 was added to the culture as indicated for another 2 days. On day 7, cells were harvested and stimulated with PMA/ionomycin or IL-12/IL-18 for 6 h in the presence or the absence of CsA or SB 203580 as indicated before staining for human CD4⁺ T cells or ELISA. Data in the upper right quadrants are percentages of hCD4- and GFP-positive cells (A and B). A, DO11.10 Th1 cells; B, Stat4^−/− DO11.10 Th1 cells; C, DO11.10 Th1 cells: left panel, hCD4- and GFP-positive cells; right panel, IFN-γ titers in the supernatants.

To address whether the observed activity of the IFN-γ promoter was dependent upon Stat4 activation and regulated by the TCR signaling pathway, we tested the −134 IFN-γ retroviral reporter construct in Stat4-deficient mice and in the absence or the presence of CsA (Fig. 8⇑B). The −134 IFN-γ reporter was unaffected by the absence of Stat4, because it was expressed at similar levels and in both the presence and the absence of PMA/ionomycin stimulation (Fig. 8⇑B, panels I and J) compared with wild-type T cells (Fig. 8⇑A). We also tested whether this constitutive activity was inhibited by CsA. As a positive control, we demonstrated that the IL-2 promoter activity was completely inhibited by CsA treatment (Fig. 8⇑B, compare panels H and F). In contrast, the activity of the IFN-γ construct was unaffected by CsA, both the constitutive activity (Fig. 8⇑B, panel K) as well as activity after treatment with PMA/ionomycin (L).

IFN-γ production can be induced by treatment with IL-12/IL-18 and inhibited by the p38 MAPK inhibitor SB 203580 (26, 42). We asked whether these same treatments could regulate the constitutive reporter activity. Th1 cells were infected with various reporters, sorted, and restimulated. For this, we measured both the reporter activity and the production of endogenous IFN-γ as an internal control. Native IFN-γ was strongly induced by PMA/ionomycin and was not inhibited by SB 203580 (Fig. 8⇑C, compare lanes 3 and 4, 9 and 10, and 15 and 16, right panel). IFN-γ was also induced by IL-12/IL-18, but this was completely inhibited by SB 203580 (Fig. 8⇑ C, compare lanes 5 and 6, 11 and 12, and 17 and 18, right panel). IL-12/IL-18 did not activate the IL-2 promoter. In contrast, the −134 IFN-γ reporter was not augmented by IL-12/IL-18 and was not inhibited by SB 203580 (Fig. 8⇑C, lanes 7–10, left panel). In summary, the activity of the IFN-γ reporter shows constitutive activity that is unaffected by CsA, and independent of both Stat4 activation and p38 MAPK activation.