Dithiocarbamates trigger differentiation and induction of CD11c gene through AP-1 in the myeloid lineage.

It has recently been shown that the alteration of the cell-redox status affects the transcription factor expression and activity. Dithiocarbamates (DTCs) are potent antioxidant agents that can switch the expression of genes dependent on the activation of the transcription factors AP-1 and NF kappa B. In this study, we show that these agents triggered the expression of genes involved in myeloid differentiation of the promonocytic U-937 cell line. DTCs promoted differentiation-associated changes that included the surface up-regulation of beta 2-integrins (CD11a-c/CD18), cell growth arrest concomitant with transferrin receptor (CD71) down-modulation, induction of the nonspecific esterase enzyme, and a rapid drop in the mRNA levels of c-myc. A further analysis, focused on the molecular mechanisms leading to the activation of CD11c expression, revealed that the pyrrolidine derivative of DTC (PDTC) increased CD11c mRNA levels and augmented its gene promoter activity. Transfection experiments with reporter constructs harboring different promoter regions of CD11c gene, indicated the presence of a functional DTC-responsive region located between positions -160 and +40 of the promoter. Gel retardation assays revealed that the PDTC-induced DNA-protein complexes were restricted to members of the Fos and Jun families that bound to an AP-1 site located at position -60 from the transcription start site. A role for this site was confirmed by in vitro mutagenesis experiments that indicated the functional importance of this site for the CD11c gene transcriptional activation in response to PDTC. The effect of DTCs on myeloid cell differentiation supports a possible role for these agents in the therapy of some bone marrow-derived malignancies.

It has recently been shown that the alteration of the cell-redox status affects the transcription factor expression and activity. Dithiocarbamates (DTCs) are potent antioxidant agents that can switch the expression of genes dependent on the activation of the transcription factors AP-1 and NFB. In this study, we show that these agents triggered the expression of genes involved in myeloid differentiation of the promonocytic U-937 cell line. DTCs promoted differentiation-associated changes that included the surface up-regulation of ␤2-integrins (CD11a-c/CD18), cell growth arrest concomitant with transferrin receptor (CD71) down-modulation, induction of the nonspecific esterase enzyme, and a rapid drop in the mRNA levels of c-myc. A further analysis, focused on the molecular mechanisms leading to the activation of CD11c expression, revealed that the pyrrolidine derivative of DTC (PDTC) increased CD11c mRNA levels and augmented its gene promoter activity. Transfection experiments with reporter constructs harboring different promoter regions of CD11c gene, indicated the presence of a functional DTC-responsive region located between positions ؊160 and ؉40 of the promoter. Gel retardation assays revealed that the PDTC-induced DNA-protein complexes were restricted to members of the Fos and Jun families that bound to an AP-1 site located at position ؊60 from the transcription start site. A role for this site was confirmed by in vitro mutagenesis experiments that indicated the functional importance of this site for the CD11c gene transcriptional activation in response to PDTC. The effect of DTCs on myeloid cell differentiation supports a possible role for these agents in the therapy of some bone marrow-derived malignancies.
Differentiation of myeloid cells is a complex process that involves the regulated expression of an array of genes responsible for a series of cellular changes that drive precursor cells to different functional and phenotypic cell stages. Several leukemic cell lines arrested at different steps of the myeloid differentiation program, such as the human U-937 and HL-60 cell lines (Sundström and Nilsson, 1976;Collins, 1987), have widely been used to dissect the molecular events accompanying this process. A variety of differentiation inducers such as phorbol esters, 1,25-dihydroxyvitamin D 3 , interferon-␥, and tumor necrosis factor-␣ have been employed to drive these leukemic cell lines toward the macrophage differentiation pathways (Hass et al., 1989;Dood et al., 1983;Ralph et al., 1983;Schü tze et al., 1988). During this process, the expression of adhesion molecules implicated in intercellular and cellular-extracellular matrix interactions is tightly regulated (Hynes, 1992). Thus, the expression of different members of adhesion receptors, such as the components of the ␤2 leukocyte integrins subfamily LFA-1 (CD11a/CD18; ␣L/␤2), Mac-1 (CD11b/CD18; ␣M/␤2), and p150,95 (CD11c/CD18; ␣X/␤2), together with the member of the immunoglobulin superfamily ICAM-1 (CD54), are upregulated, whereas VLA-4 (CD49/CD29, ␣4␤1), a member of the ␤1 integrin subfamily, is typically down-regulated during differentiation (Miller et al., 1986;Dustin et al., 1986;Ferreira et al., 1991). Other cell surface molecules such as the transferrin receptor (CD71) are also down-modulated (Hass et al., 1989). Additional changes commonly linked with myeloid differentiation and often interrelated are cell growth arrest and changes in the expression of the c-myc proto-oncogene. Thus, the HL-60 and U-937 cell lines exhibit high basal levels of c-myc which are thought to be involved in the high proliferation rate of these cells (Collins, 1987;Cotter et al., 1994). A drop in the c-myc levels, accompanied by cell growth arrest, takes place when cells are induced to differentiate with a variety of agents, which appears to occur in normal bone marrow myeloid progenitors undergoing terminal differentiation (Collins, 1987;Rius et al., 1990).
Recently, reactive oxygen intermediates (ROIs) 1 such as H 2 O 2 , superoxide (O 2 Ϫ ), and hydroxyl radical (OH ⅐ ), have been described as common signal transduction mediators in a number of activation pathways. The involvement of these intermediates in such processes has been analyzed on the basis of their effects on the activities of the transcription factors NFB and AP-1. Thus, NFB behaves as an oxidative-stress responsive factor that can be directly activated by H 2 O 2 (Schreck et al., 1991;Meyer et al., 1993) or through stimuli that increase intracellular ROIs levels (Schreck et al., 1992b). These prooxidant stimuli can also induce AP-1 activation as in the case of UV light and H 2 O 2 (Devary et al., 1991), proinflammatory cytokines (Brenner et al., 1989;Muñ oz et al., 1992) and ␥-radiation (Datta et al., 1992). Strikingly, whereas the activation of NFB can be prevented by several antioxidant compounds with ROIs scavenging properties, including DTCs (Meyer et al., 1993;Schenk et al., 1994;Schreck et al., 1992a;Staal et al., 1990), these agents per se induce both AP-1 DNA binding activity and AP-1-dependent transcriptional activation by mechanisms that involve de novo transcription of c-fos and c-jun (Meyer et al., 1993;Schenk et al., 1994). An additional redox mechanism that modulates AP-1 activity is based on the presence of a critical redox-sensitive cysteine residue that regulates the DNA binding of Fos and Jun proteins (Abate et al., 1990). Strikingly, AP-1 is activated early on when an antioxidative state is induced in cells under hypoxic conditions (Yao et al., 1994). Hence, AP-1 can be activated by signals generated under prooxidant and antioxidant conditions. Cellular differentiation is mainly regulated by selective expression of genes largely controlled at the transcriptional level. Since AP-1 can behave as an antioxidant-responsive transcription factor and exogenous antioxidants can potentially switch on genes through the activation of this transcription factor, we analyzed herein whether the DTCs affect the expression of genes involved in myeloid differentiation. Interestingly, we found that DTCs trigger the expression of myeloid differentiation antigens, as well as other changes associated to the differentiation of U-937 cells. To characterize the molecular mechanisms involved in this process, we analyzed the effect of these agents on the transcription of p150,95 ␣-subunit (CD11c) gene. This analysis indicated that DTCs stimulate CD11c gene promoter and identified the transcription factor AP-1 as a central target implicated in this activation.
Flow cytometry analysis was performed as described previously (Postigo et al., 1991). Briefly, after different treatments, cells were collected by centrifugation, resuspended in phosphate-buffered saline, and incubated for 30 min with 100 l (1 g) of purified RR1/1 or 100 l of the hybridoma culture supernatants indicated above. Antibodies were incubated in the presence of 50 g/ml ␥-globulin to prevent binding through the Fc portion of the antibodies. Cells were washed, and bound antibodies were detected using fluorescein isothiocyanate-conjugated rabbit F(ab)Ј2 anti-mouse IgG (Dako A/S, Denmark), and finally resuspended in 300 l of propidium iodide (2 ng/l). The supernatant from the myeloma P3X63 was included as a negative control. Samples were analyzed by flow cytometry in a fluorescein-activated cell sorter cytofluorometer (Becton Dickinson, Mountain View, CA). Analysis was performed over viable cells (typically higher than 95%) as determined by staining with the fluorochrome propidium iodide. Cell proliferation was determined by counting in a hematocytometer the number of viable cells by trypan blue exclusion.
For cytochemical staining, 10 5 cells were centrifuged in a Cytospin (Shandon, Shandon Southern Products Ltd.) over microscope slides and stained for the detection of the nonspecific esterase (NSE) following the instructions of a NSE kit (Sigma). Cell preparations were observed in a Nikon Labophot-2 photomicroscope, and more than 300 cells were counted to score the proportion of strongly positive stained cells.
RNA Analysis-Cells were incubated with different concentrations of PDTC or PMA and harvested after various time points. Total cellular RNA was isolated from each sample by using either the guanidinium isothiocyanate-CsCl ultracentrifugation method (Sambrook et al., 1989) or the Ultraspec TM system (Biotecx Laboratories, Inc.). Denaturated RNA (5-10 g) was blotted onto nitrocellulose membranes following the instructions of Bio-Dot® SF Microfiltration Apparatus (Bio-Rad). After UV cross-linking, the filters were hybridized with the following specific probes: a 0.8-kb PstI fragment from the CD11c cDNA (Corbí et al., 1987), a 1.4-kb HindIII-EcoRI fragment from c-myc cDNA, a 0.8-kb BglII-NcoI fragment from the c-fos cDNA, a 0.8-kb HindIII-PstI fragment from the c-jun cDNA, and a 0.6-kb HindIII-BamHI fragment of the ␤-actin cDNA for comparative purposes.
For transfection experiments, a total of 35 ϫ 10 6 U-937 cells were electroporated with 50 g of the different constructs, as described previously (López-Cabrera et al., 1993), and cultured either in the absence or presence of PDTC, 50 M, or PMA, 20 ng/ml. After 13 h, cells were collected by centrifugation, and luciferase activity was quantified following instructions described in a luciferase assay kit (Promega, Madison, WI). To determine transfection efficiency, 12 g of pCMV␤-gal (Clontech Laboratories Inc., Palo Alto, CA), which contains the cytomegalovirus promoter upstream to the ␤-galactosidase gene, were included in each transfection. ␤-Galactosidase activity was measured following the protocol previously described (Sambrook et al., 1989).
Nuclear Extracts-Small scale nuclear extracts from untreated or PDTC-treated U-937 cells were obtained as described previously (Osborn et al., 1989) with some modifications. Briefly, cells were collected by centrifugation and washed once with phosphate-buffered saline and twice with buffer A (10 mM Hepes, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM dithiothreitol). The cell pellet was then resuspended in buffer A, 0.1% Nonidet P-40, and incubated for 10 min. Lysed cells were microcentrifuged, and nuclear pellet was extracted with 3 volumes of buffer C (20 mM Hepes pH 7.9, 25% v/v glycerol, 0.4 M KCl, 1.5 mM MgCl 2 , 0.2 mM EDTA) in the presence of dithiothreitol and protease inhibitors. Nuclear pellet volume was estimated to correct the final KCl concentration to 400 mM. After 30 min of incubation, nuclei were microcentrifuged for 30 -45 min, and the supernatant was diluted with 4 volumes of modified buffer D (20 mM Hepes pH 7.9, 20% v/v glycerol, 50 mM KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol) (Osborn et al., 1989), and frozen at Ϫ70°C. All steps were performed on ice. Protein concentrations were determined by Bradford assay.
Electrophoretic Mobility Shift Assays (EMSAs)-Gel retardation assays were performed as described (Redondo et al., 1991) with minor modifications. A total of 2 g of nuclear protein were incubated with 1 g of poly(dI-dC) DNA carrier, 2.5 l of 5 ϫ DNA binding buffer (polyvinylethanol, 10% w/v; glycerol, 12.5% v/v; 50 mM Tris, pH 8.0, EDTA, 2.5 mM; dithiothreitol, 2.5 mM) in a final volume of 10.5 l for 10 min on ice. The incubation was performed in the presence or absence of 30-fold molar excess of homologous unlabeled oligonucleotide. Then 5 ϫ 10 4 cpm (5 ϫ 10 7 cpm/g) of 32 P-labeled double-stranded oligonucleotides (2 l) were added and incubated at room temperature for 20 min. DNA-protein complexes were resolved from free probe using a 4% polyacrylamide gel in TBE buffer. The serological characterization of the AP-1 complex resolved by EMSA was performed by incubating the antisera (1 l) and the nuclear extracts prior to the addition of labeled oligonucleotides. The rabbit antisera used against Fos family (RR26/8) and Jun family (6 36/6) members were kindly provided by Dr. Rodrigo Bravo (Bristol-Myers Squibb, Princeton, NJ).

Differentiation-associated Changes Induced by Dithiocarbamates in U-937
Cell Line-Promonocytic U-937 cells can be induced to differentiate toward macrophages in the presence of phorbol esters (Minta and Pambrum, 1985). During this process, a regulated expression of important cell surface molecules takes place. To explore the effects of DTCs on U-937 differentiation, these cells were treated with 50 M of PDTC for 24 h, and changes in the expression of a number of differentiation markers were analyzed by flow cytometry. PDTC treatment results in up-regulation of cell surface expression of integrins ␤2 (CD11a-c/CD18) and ICAM-1 (CD54), and down-modulation of the ␤1 integrin VLA-4 (CD49d/CD29) and the transferrin receptor (CD71). These changes were qualitatively similar to those elicited by PMA (Fig. 1A) (Miller et al., 1986;Dustin et al., 1986;Ferreira et al., 1991;Hass et al., 1989). Furthermore, other DTCs, such as DDTC and its disulfide-linked form, disulfiram, induced a cell surface antigen expression pattern comparable to that triggered by PDTC in the U-937 cell line (Fig. 1B), as well as in the promyelocytic HL-60 cell line (data not shown).
Since cell surface changes induced by dithiocarbamates in U-937 and HL-60 cells were characteristic of differentiation processes, we performed experiments directed to determine whether DTCs exerted additional changes associated to the myeloid differentiation program using PDTC. The appearance of the acid ␣-naphthyl acetate esterase (nonspecific esterase) enzymatic activity is a typical marker of myeloid differentiation toward monocyte-macrophages cells induced by PMA (Rovera et al., 1979;Collins, 1987). Cytochemical experiments performed in U-937 revealed that PDTC significantly induced the expression of NSE at 24 h ( Fig. 2A), although to lower extent than that elicited by PMA (data not shown). At this time, treatment with the antioxidant produced cell growth inhibition at doses as low as 20 M, being maximal with concentrations ranging from 35 to 50 M ( Fig. 2A). Additionally, dot blot experiments revealed that the levels of c-myc RNA markedly declined after 6 h of treatment with 100 M of PDTC and remained at undetectable levels by 16 h, as occurred with PMA ( Fig. 2B). At lower doses of DTC, c-myc levels were also down-modulated after 6 h, and later reinduced (data not shown). Altogether, these results demonstrate that PDTC not only induces cell surface differentiation markers but also affects the proliferation of U-937 cells as well as other differentiation-associated events.
Since PDTC exerted similar differentiation-associated changes than those triggered by PMA, we performed experiments to determine whether protein kinase C (PKC) was mediating the effect of PDTC. As expected, treatment of the cells with the PKC-specific inhibitor bisindolylmaleimide resulted in a dramatic inhibition of CD11c and CD54 surface expression induced by PMA. By contrast, the induction of these markers by PDTC remained unaffected in the presence of the same concentrations of the inhibitor (Fig. 3). Furthermore, microscopic examination of cultures of U-937 cells revealed additional differences in the signals displayed by both inducers. Thus, PMA-mediated differentiation was accompanied with strong cell-aggregation and cell adherence to plastic surfaces, whereas PDTC, failed to induce such changes (not shown). These data indicate that the signaling pathway triggered by the dithiocarbamate is not dependent on PKC activity and differs from that elicited by PMA.
Effect of PDTC on CD11c Gene Transcription-The expression of the integrin p150,95 (CD11c/CD18), a marker of myeloid differentiation, is regulated during the myeloid development (Hogg et al., 1986;Miller et al., 1986). Previous works have demonstrated that the transcriptional activation of ␣-subunit (CD11c)-encoding gene directs the expression of the CD11c/ CD18 heterodimer during PMA-induced differentiation (Bellón et al., 1994;López-Cabrera et al., 1993). To search for the mechanisms responsible for the phenotypic changes induced by DTCs, we analyzed their effect on the expression of CD11cencoding gene. Dot blot analysis revealed that in PDTC-treated cells CD11c mRNA levels increased after 6 h of incubation, and reached its highest levels between 16 and 24 h (Fig. 4A). PMA, which induced a stronger surface expression of the p150,95 molecule, also triggered a higher mRNA induction (Bellón et al., 1994) (data not shown).
To further investigate the effect of PDTC on CD11c gene transcription, we transfected U-937 cells with reporter plasmids containing the regions of the CD11c promoter spanning from Ϫ361 to ϩ40, Ϫ253 to ϩ40, and Ϫ160 to ϩ40 (López-Cabrera et al., 1993). All of these constructs displayed basal promoter activity that was clearly increased in PDTC-treated transfected cells (Fig. 4B). This inducible activity was qualitatively similar although less potent than that exerted by PMA, which was used as a positive control. The transfection of longer fragments of the CD11c gene promoter spanning from Ϫ640 to ϩ40 and Ϫ960 to ϩ40 yielded similar levels of induction (data not shown) suggesting that the cis-acting elements responsible for the transcriptional activation mediated by PDTC were located within the Ϫ160 to ϩ40 proximal region of the CD11c promoter.
Nuclear Factor Binding to the CD11c Gene Proximal Promoter Region in Response to PDTC-Data base analysis of the nucleotide sequences within the Ϫ160 to ϩ40 region indicated the presence of consensus motifs for binding of different transcription factors, such as AP-1, Sp1, AP-2, PU.1, and RARE sites (López-Cabrera et al., 1993). To identify the nuclear factors responsible for the activation of the CD11c promoter by PDTC, we used overlapping double-stranded oligonucleotides spanning through this region as probes for mobility shift assays. Sequences from Ϫ68 to Ϫ46, Ϫ136 to Ϫ99, Ϫ81 to Ϫ55, Ϫ39 to Ϫ8, Ϫ54 to Ϫ23, Ϫ19 to ϩ11, and Ϫ109 to Ϫ81 of the CD11c promoter were analyzed for the binding of nuclear extracts from U-937 cells treated or not with PDTC by EMSA (Fig. 5A, and data not shown). We have found that the complex retarded by probe A (Ϫ68 to Ϫ46), which contains a functional AP-1 site, was substantially augmented, in a dose-dependent manner, in response to PDTC (Fig. 5A). This retarded complex was specifically competed by addition of an excess of unlabeled homologous oligonucleotide to the binding reaction, whereas an excess of heterologous oligonucleotide failed to compete this binding (data not shown). Specific protein-DNA complexes formed by incubation of nuclear extracts from untreated cells with the rest of oligonucleotide probes containing either Sp1 (Fig. 5A, probe B) or the putative binding sequences for Sp1 (probe C), PU.1 (probes D, E, F, and G), and RARE (E) remained unaffected when they were tested using nuclear extracts from PDTC-treated cells (data not shown). Finally, the most conclusive evidence showing that PDTC-induced nuclear factors bound to the AP-1 site were members of the Fos and Jun families of transcription factors was obtained by supershift assays adding specific antisera directed against members of both families to the binding reaction. Antibodies to Jun and Fos efficiently inhibited the formation of the specific complex (Fig. 5B).
In order to better characterize the mechanism involved in the induction of AP-1 DNA binding activity mediated by PDTC, we examined the effect of the antioxidant on the steady state mRNA levels of c-jun and c-fos. We performed dot-blot analysis using RNA from cells treated with PDTC or PMA. PDTC strongly increased c-jun mRNA levels as early as 30 min of treatment, reaching its highest levels between 4 and 7 h. Similarly, the dithiocarbamate also induced an early expression of c-fos mRNA that declined after 2 h of treatment (Fig. 5C).
Role of AP-1 on CD11c Gene Promoter Activation by PDTC-To further determine the functional contribution of the AP-1 site in the activation of CD11c gene promoter by PDTC, we used the plasmids pCD11C160(-60mut)-Luc and pCD11C361(Ϫ60mut)-Luc, which contained a 3-base pair substitution within the AP-1 binding site (Ϫ61 to Ϫ54) in the context of the Ϫ160/ϩ40 and Ϫ361/ϩ40 fragments of the CD11c promoter (Fig. 6A). This 3-base pair substitution affected AP-1 binding activity, since oligonucleotides containing the mutated sequence failed to compete with the specific AP-1 complex generated with the radiolabeled wild type probe in EMSA (data not shown). In addition, the activities of the pCD11C160(-60mut)-Luc and pCD11C361(-60mut)-Luc mutated constructs in response to PDTC were significantly lower than those obtained with the wild type constructs (Fig. 6B). However, the mutated constructs displayed a similar fold induction than did the wild type constructs in response to butyrate, a very potent activator of CD11c promoter. 2 Hence, mutations of the AP-1 site did not indiscriminately block the transcriptional responsiveness of the CD11c promoter that remained inducible by other differentiation inducer. Taken together, our data indicated that the DTC-induced CD11c gene expression is regulated at the transcriptional level involving the functional activation of AP-1.

DISCUSSION
Alteration of the redox status of the cells by antioxidant agents such as DTCs, has previously been shown to interfere with the activity and expression of the transcription factors NFB and AP-1, which appear to be involved in cell differentiation and activation processes (Baeuerle and Henkel, 1994). The results presented in this report show that DTCs simultaneously induce the expression of myeloid differentiation markers in U-937 cells and inhibit cell proliferation. To characterize the mechanisms responsible for DTC-mediated cell differenti-

FIG. 2. Effects of PDTC on cell growth, nonspecific esterase staining, and c-myc RNA levels in U-937 cells.
A, cells were treated with different doses of PDTC for 24 h and processed for ␣-naphthyl acetate esterase (NSE) staining. Percents of strongly stained cells and cell growth are represented. Cell number was determined by seeding 3 ϫ 10 5 cells in the presence and the absence of PDTC at the concentrations indicated. After 24 h, cells were counted in a hematocytometer. Cell viability, determined by trypan blue exclusion, was higher than 95% for the different treatments. B, dot blot analysis was performed with total RNA isolated from cells untreated or treated with PMA, 20 ng/mL, and PDTC, 100 M, for 6 or 16 h. RNA was hybridized with cDNA probes of the human c-myc and ␤-actin genes.
ation, we have analyzed as a model the activation of CD11c gene transcription. These studies suggest that PDTC induces the transcriptional activation of the CD11c gene through the activation of the AP-1 transcription factor.
The comparative analysis of differentiation markers induced by DTCs and PMA indicated that both elicit a similar pattern of cell surface antigen expression. In addition, PDTC and PMA induced the NSE enzymatic activity, inhibited cell growth, down-modulated c-myc RNA levels, and increased those of c-fos and c-jun. However, several lines of evidence indicate that both inducers act through different signaling mechanism. (i) Whereas the changes in the expression of myeloid surface antigens by PMA have been shown to be dependent on the activity of PKC (Bellón et al., 1994) (Fig. 3), those mediated by PDTC were not affected by the specific PKC inhibitor bisindolylmaleimide. In accordance, it has been previously shown that the treatment of Jurkat T cells with PDTC does not affect the activity of PKC (Meyer et al., 1993). (ii) PMA triggers strong homotypic aggregation, cell adhesion, and spreading to plastic surface (Collins, 1987;Miller et al., 1986;Nueda et al., 1995). In contrast, PDTC induced only a weak homotypic aggregation. These results also support that PKC-independent signals are mediating the effect of DTCs, since the aggregation and adhesion processes induced by PMA have been proposed to be dependent on a PKC-induced conformational change of the ␤2 integrin molecules (Keizer et al., 1988). (iii) Flow cytometry analysis using dichlorofluorescein dye revealed that PMA induces an increase of the intracellular H 2 O 2 levels in U-937 cells FIG. 5. Effects of PDTC on binding activity to the ؊68/؊46 and ؊136/؊99 regions of CD11c promoter and on the RNA levels of c-fos and c-jun. A, nuclear extracts prepared from U-937 cells untreated and treated for 6 h with PDTC (50 and 100 M) were incubated with probes including the Ϫ68 to Ϫ46 and Ϫ136 to Ϫ99 regions (probes A and B) of the CD11c gene promoter. DNA-protein complexes were resolved by EMSA. A 30-fold molar excess of unlabeled homologous oligonucleotide was used as competitor in binding reactions. B, probe A, containing an AP-1 consensus site was incubated with nuclear extracts from PDTC (100 M) treated cells in the presence of preimmune rabbit antiserum or rabbit antisera against Fos family (RR 26/8) or Jun family (6 36/6) members. C, Dot blot analysis was performed with total RNA isolated from U-937 untreated (C) or treated with PDTC 50 M and PMA 20 ng/ml for the time points indicated. RNA was hybridized with cDNA probes of c-fos, c-jun, and ␤-actin.
FIG. 6. Role of AP-1 in CD11c gene promoter activity in response to PDTC. A, sequences of the wild type (white box) and mutated regions (black box) (Ϫ68 to Ϫ46) of the CD11c promoter. The substitutions introduced within the AP-1 site (Ϫ61 to Ϫ54) are indicated. B, recombinant wild type plasmids, containing Ϫ361/ϩ40 and Ϫ160/ϩ40 regions of CD11c gene promoter and the respective mutated plasmids, were transiently transfected into U-937 cells, and after 13 h of treatment with PDTC, luciferase activity was measured. Relative luciferase activity was expressed as fold induction over the activity of the promoterless plasmid pXP2 in untreated and treated cells. Results are representative of three independent experiments. whereas PDTC has the opposite effect. 3 Several antioxidants have been shown to influence the activity of the inducible transcription factors NFB and AP-1 (Beg et al., 1993;Meyer et al., 1993;Schreck et al., 1992aSchreck et al., , 1992bStaal et al., 1990). In HeLa cells PDTC, N-Acetyl-L-Cysteine (NAC) and the antioxidative enzyme ADF/thioredoxin potently activates the transcription factor AP-1 involving induction of c-fos and c-jun (Meyer et al., 1993). Our results indicate that PDTC stimulates the surface expression of CD11c/CD18 (p150,95), increasing its ␣-subunit (CD11c) mRNA steady state levels by transcriptional mechanisms acting on the CD11c gene promoter. The induction of c-fos and c-jun mRNA levels by PDTC, as well as the presence of an AP-1 site at Ϫ60, the elimination of which greatly decreases the PDTC responsiveness of the CD11c promoter, supports an important role of AP-1 transcription factor in the activation and differentiation process elicited by DTCs. In this context, this transcription factor has been suggested to play a role in the myeloid differentiation triggered by different inducers such as PMA or etoposide (Bellón et al., 1994;Pérez et al., 1994). Furthermore, PMA has been shown to induce AP-1 DNA binding activity to the functional AP-1 site at Ϫ60 on the CD11c promoter, concomitant with the appearance of Fos protein detected by EMSA, while only Jun components, that account for the basal levels of AP-1, are detected in undifferentiated U-937 cells. 2 Thus, AP-1 may represent a transcription factor that can integrate different signals elicited by several differentiation agents that converge at this point. However, in order to explain the distinct patterns of differentiation observed, additional signaling different to AP-1 activation must be produced by the different inducers.
A thorough study of the nuclear factors that bind to the responsive region of the CD11c promoter spanning from Ϫ160 to ϩ40, revealed that PDTC restrictively induced DNA binding activity to the AP-1 site. Hence, AP-1 is stimulated in a specific manner and can function as an antioxidant-responsive transcription factor in U-937 cells. Similarly, the genes of the antioxidative enzymes NAD(P)H:quinone oxidoreductase, EC 1.6.99.2, previously known as DT-diaphorase and the glutathione S-transferase Ya subunit, are controlled at the transcriptional level through antioxidant responsive elements in their promoter regions. These elements are required for the transcriptional activation mediated by several antioxidants (Rushmore et al., 1991). Since antioxidants can induce a hypoxic status in the cells, it is noteworthy that the DT-diaphorase gene is transcriptionally activated under hypoxic conditions in human colon cancer cells. Moreover, this effect involves early induction of jun family genes followed by a latter elevation of c-fos m-RNA levels (Yao et al., 1994). Interestingly, the gene promoter of the DT-diaphorase contains an AP-1 motif that has been shown to be functionally involved in the transcriptional induction of the gene by polycyclic aromatic hidrocarbons and phenolic antioxidants (Li and Jaiswal, 1992). Mutation of this site results in the loss of basal and induced transcriptional activity of the promoter. As shown in this paper, the AP-1 site of the CD11c promoter behaves in a similar fashion to that of the DT-diaphorase promoter and could, therefore, be considered an antioxidant responsive element. In this regard, it would be interesting to study whether the activation of the AP-1 transcription factor under hypoxic conditions could trigger the expression of CD11c.
Finally, DTCs are drugs that have already been pharmacologically used in heavy metal poisoning and AIDS treatment (Lang et al., 1985;Sunderman et al., 1967). It will be important to determine whether the effects of the DTCs that we have described in U-937 and HL-60 cells are also operative in leukemic cells derived from patients. In this regard, retinoic acid has been used in the treatment of patients with acute promyelocytic leukemia yielding a high rate of transient clinical remissions (Chen et al., 1991;Warrel et al., 1991). However, after prolonged periods of treatment, resistance to RA appears and the disease is reactivated (Delva et al., 1993). The effects of DTCs on myeloid cell differentiation supports a therapeutical potential of these agents in bone marrow-derived malignancies. Future experiments on fresh leukemic cells will aid in the elucidation of this interesting matter.