INTRODUCTION

The gp91^phox gene encodes the cytochrome b₅₅₈ heavy chain of the NADPH-dependent oxidase and is required for the generation of toxic free radicals and microbicidal activity by phagocytes (1). Absence of gp91^phox leads to the immunodeficiency syndrome chronic granulomatous disease (CGD)¹ (1). The gp91^phox gene is expressed nearly exclusively in terminally differentiating myeloid cells (2), is transcriptionally induced by interferon- (IFN-) (3), and provides an attractive system to study the regulation of myeloid cell gene expression and development.

Previous studies demonstrated that the 450 to +12 bp region of the gp91^phox promoter directs transcription in a subset of monocyte/macrophages in transgenic mice (4) and also in response to IFN- in stably transfected PLB985 myeloid cells (5). Complex interactions between cis elements and DNA-binding proteins occur within this promoter region (Fig. 1). Highly related if not identical DNA-binding proteins denoted BID (binding increased during differentiation) bind to three sites centered at 355 bp (DIST-BID), 225 bp (MID-BID), and 145 bp (PROX-BID) of the gp91^phox promoter. These interactions are required for IFN--induced transcription.² The MID-BID element conforms to an interferon-stimulated response element (ISRE), and is also a binding site for IFN regulatory factors (IRF).² In addition, the binding of a novel factor, HAF-1 (hematopoietic associated factor-1), to a site at 57 bp is also required for IFN- induction of the gp91^phox promoter (5). Mutations of the HAF-1 binding site have been reported in two kindreds of CGD patients who fail to normally transcribe the gp91^phox gene (6, 7). Also, a CCAAT box located at 123 to 119 bp of the gp91^phox promoter is recognized by CP1, a ubiquitous CCAAT box binding activity (8). None of these transcriptional activating factors are myeloid-specific.

The restriction of gp91^phox promoter activity to mature phagocytes requires regulated transcriptional repression mediated by the binding of CCAAT displacement protein (CDP). Four CDP binding sites have been detected centered at 350, 220, 150, and 110 bp of the gp91^phox promoter. Each CDP binding site overlaps BID or CP1 binding sites, and binding of CDP excludes these transcriptional activating factors (8, 9).² The binding of CDP to each site is down-regulated during myeloid differentiation, coincident with induction of gp91^phox expression (8, 9), and constitutive expression of CDP prevents gp91^phox induction during terminal myeloid differentiation (10). Ablation of CDP binding sites results in increased gp91^phox promoter activity in HEL, K562, and HeLa cells that do not normally express gp91^phox, a phenomenon termed de-repression (8, 9).

The 100 to +12 bp region of the gp91^phox gene, which lacks upstream repressor elements, directs transcription in nonphagocytic cells (9). Transcriptional activating elements within this region of the endogenous gp91^phox promoter are presumably interfered with by CDP binding in nonphagocytic cells and are possibly also required for appropriate transcription in mature myeloid cells. The purpose of this study was to identify cis elements and cognate DNA-binding proteins within this promoter region that are required for gp91^phox promoter activity. We describe a cis element between 100 and 85 bp of the gp91^phox promoter that is required for promoter activity in the absence of repression and which binds members of the IRF family and a factor similar to BID.

EXPERIMENTAL PROCEDURES

Individual gp91^phox promoter fragments, each of which ends at +12 bp and contains the 5-untranslated region, were obtained by utilizing the polymerase chain reaction to amplify desired regions from a 450 to +12 bp gp91^phox promoter/human growth hormone plasmid (5). Polymerase chain reaction products were digested with SalI and BamHI and subcloned into SalI/BglII-digested luciferase reporter gene vector (pXP2) (11) (the generous gift of Yu-Chung Yang, Indianapolis). Mutations of the BID or IRF binding sites were introduced into the 100 to +12 bp gp91^phox promoter by polymerase chain reaction-mediated mutagenesis. The nucleotide sequence of each gp91^phox promoter construct was confirmed by the dideoxy chain termination method using a Sequenase 2.0 kit (U. S. Biochemical Corp.). The pCO-IRF-1 (12) and pCMV-BL-IRF-2 (13) plasmids that express IRF-1 or IRF-2 under the control of the cytomegalovirus (CMV) promoter were generously provided by Stephen Goodbourn (London) and John Hiscott (Montreal), respectively. Plasmids used in transfections were prepared by two rounds of equilibrium centrifugation in CsCl ethidium bromide gradients (14).

The human cervical carcinoma cell line HeLa (15), the human erythroleukemia cell line HEL (16), and the human chronic myelogenous leukemia cell line K562 (17) were obtained from the American Type Culture Collection (Rockville, MD). The human myelomonoblastic PLB985 cell line (18) was the generous gift of Thomas Rado (Birmingham, AL). Cells were maintained in RPMI 1640 medium as described previously (9).

K562 and HEL cells were transfected by electroporation. For each transfection, 10⁷ cells were resuspended in 0.3 ml of serum-free medium and mixed with 5 µg of gp91^phox promoter/luciferase plasmid and 0.5 µg of CMV promoter-enhancer/-galactosidase (CMV--gal) plasmid at room temperature. The mixture was electroporated at 220 V, 960 microfarads, and the transfected cells were grown in 10 ml of complete medium at 37 °C and 5% CO₂. After 10 h of incubation, cells were harvested and washed twice with 1 × phosphate-buffered saline, and cell pellets were collected and resuspended in 100 µl of 1 × lysis buffer (Promega). After incubation at room temperature for 10-15 min, cell extracts were collected, and 20 µl were used to assay luciferase activity using a Promega luciferase kit and a Lumat 9210 luminometer. -Galactosidase activity in each extract was measured (14) and used to standardize luciferase values to compensate for variability in transfection efficiency between samples. For co-transfection experiments, 10 µg of expression vector was used in addition to luciferase reporter constructs. Multiple independent plasmid preparations of each promoter construct were analyzed in transfection experiments.

Nuclear extracts were isolated as described by Dignam et al. (19) from untreated HEL or K562 cells, or PLB985 cells treated with phorbol 12-myristate 13-acetate for 48 h. Nuclear extract was isolated by a minipreparation method (20) from HeLa cells treated with 1000 units/ml human IFN- (Boehringer Mannheim) for 12 h. Oligonucleotides were annealed, subcloned into pUC19, released by restriction digestion, and radiolabeled by T4 polynucleotide kinase using [-³²P]ATP or with Klenow enzyme using [-³²P]dCTP. Radiolabeled probes were resolved by polyacrylamide gel electrophoresis and eluted by the crush and soak method (14). Electrophoretic mobility shift assays (EMSA) were performed as described previously (8) with slight modification. Briefly, 3-6 µg of nuclear extract was mixed with 0.5 µg of poly(dI·dC) and competitor double-stranded oligonucleotides where indicated in a 20-µl reaction volume. All samples containing the 100 to 63 bp gp91^phox promoter probe also contain 20 ng of a high affinity CDP binding site oligonucleotide (E36) to disrupt CDP-containing complexes. The mixture was incubated on ice for 15 min prior to the addition of 15,000 cpm of probe. After another 15-min incubation on ice, samples were loaded onto a 0.5 × TBE (45 m Tris-HCl (pH 8.3), 45 m borate, 1.25 m EDTA), 6% nondenaturing polyacrylamide gel, and electrophoresis was carried out at 25 mA for 2.5 h at 4 °C. Antibodies against IRF-1, IRF-2, and ISGF-3 (p48) were purchased from Santa Cruz Biotech. (Santa Cruz, CA) and were incubated with nuclear extracts for 20 min prior to the addition of radiolabeled probes in antibody-blocking EMSA experiments.

Complementary oligonucleotides were synthesized on an Applied Biosystems model 394 synthesizer. Listed are the upper strands of double-stranded oligonucleotides corresponding to the following regions of the gp91^phox promoter (8). Mutated sequences are underlined and nomenclature used in the text is indicated in bold: PROX-BID (182 to 113 bp), 5-tttgtagttgttgaggtttaaagatttaagtttgttatggatgcaagcttttcagttgaccaatgattat-3; PROX-BID-mut, 5-tttgtagttgttgagtaagatatgttatggatcaagcagtgaccaatgattat-3; MID-BID (261 to 212 bp), 5-gttatttatctcttagttgtagaaattggtttcattttccactatgttta-3; MID-BID-mut (241 to 212 bp), 5-gttatttatctcttagttgtagaaattggtttccactatgttta-3; 105 to 63 bp, 5-aatttctgataaaagaaaaggaaaccgattgccccagggctgc-3; 100 to 63 bp, 5-ctgataaaagaaaaggaaaccgattgccccagggctgc-3; BID-mut (105 to 63 bp), 5-aatttctgataaaagaaagaaaccgattgccccagggctgc-3; IRF-mut (105 to 63 bp), 5-aatttctgataaaagaaaagccgattgccccagggctgc-3; BID + IRF-mut (105 to 63 bp), 5-aatttctgataaaagaaactgaaattgccccagggctgc-3.

Oligonucleotides corresponding to high affinity binding sites for the following transcription factors were also used as competitors or probes in EMSA (nomenclature used in text is indicated in bold): CDP (E36) (21), 5-cggatccgaattcatcgataatcgattat-3; IFN regulatory factor-1 (IRF-1) (22), 5-acggatccggcatattcaaaaccgaaaccaagtccctcgagac-3; IFN regulatory factor-2 (IRF-2) (22), 5-acggatccctcgagggaggaaagtgaaacctaaacagtg-3; IFN-stimulated gene factor-3 (p48) (ISGF-3) (23), 5-aatttcactttctagtttcactttcccttttgt-3; IFN consensus sequence binding protein (ICSBP) (24), 5-tctcctcagtttcacttctgca-3; PU.1 interaction partner (PIP) (25), 5-gaaaaagagaaataaaaggaagtgaaaccaag-3; differentiation induced factor (DIF) (26), 5-tatctgtttcaaggatttgagatgtattttcccagaaaaggaac-3.

RESULTS

Previously we demonstrated that the 100 to +12 bp region of the gp91^phox gene, which lacks multiple upstream repressor elements, exhibits significant promoter activity in cell lines such as K562 and HEL that do not normally express the gp91^phox gene (9). The level of de-repressed promoter activity is 3-fold greater than that exhibited by the SV40 early promoter in HEL cells.³ To study the proximal gp91^phox promoter in more detail, a series of progressive deletions from the 5-end of the 100 to +12 bp promoter fragment were constructed and linked to a luciferase reporter gene. These constructs were then transiently transfected into HEL and K562 cells to identify cis elements required for gp91^phox promoter activity in the absence of CDP-mediated repression. The activity of the gp91^phox promoter is drastically reduced in both HEL and K562 cells by a 16-bp truncation of the gp91^phox promoter to 84 bp (Fig. 2). This suggests that transcriptional activating element(s) are located within the 100 to 85 bp region of the gp91^phox promoter.

Functional analysis of 5-deletions of the gp91

IRF-2 and BID Bind to the 100 to 63 bp Region of the gp91^phox Promoter

The 92 to 80 bp region of the gp91^phox promoter (5-AGAAAAGGAAACC-3) conforms to an ISRE consensus sequence (RGAAANNGAAASY) (27). EMSA was performed using nuclear extracts isolated from HEL cells and a probe corresponding to the 100 to 63 bp region of the gp91^phox promoter (Fig. 3). Because this probe is recognized by CDP (data not shown), a high affinity CDP binding site (E36) was used as a competitor in these experiments. Under these conditions, three DNA-binding complexes are revealed. The two lower complexes are disrupted by oligonucleotide competitors corresponding to binding sites of the IRF family members IRF-1, IRF-2, ICSBP, PIP, and ISGF-3 (p48) (22, 23, 24, 25) (Fig. 3A, left panel). However, these complexes are not disrupted by a binding site for DIF, an IFN response factor unrelated to the IRF family (Fig. 3A, left panel) (26). None of these competitors affects the upper complex of the triplet. All three complexes are disrupted by the addition of a competitor homologous to the probe (data not shown). These data indicate that the lower two complexes exhibit a binding specificity similar to the IRF family of transcription factors. Similar DNA-binding complexes are observed using a shorter probe (100 to 79 bp) of the gp91^phox promoter that also contains the ISRE element (data not shown).

Identification of proteins that bind to the 100 to 63 bp region of the gp91^phox promoter. A, the bottom two complexes of the triplet correspond to IRF-2. EMSA was performed as described under ``Experimental Procedures'' using a nuclear extract isolated from HEL cells and a probe corresponding to the 100 to 63 bp region of the gp91<sup>phox</sup> promoter. Oligonucleotides corresponding to high affinity binding sites for the IFN response factors IRF-1, IRF-2, ISGF-3 (p48), PIP, and DIF were used as competitors (40 ng), and 2 µl of antiserum raised against IRF-1, IRF-2, or ISGF-3 were added to reactions where indicated. The doublet that contains IRF-2 is indicated by the bracket. The asterisk indicates the position of the upper complex of the triplet that does not contain IRF-2. Relative mobilities of complexes cannot be directly compared between panels. B, the upper complex of the triplet exhibits the mobility and binding site specificity of BID. EMSA was performed as described under ``Experimental Procedures'' using the 100 to 63 bp or 261 to 212 bp regions of the gp91^phox promoter as probes and a nuclear extract isolated from HEL cells. The PROX-BID oligonucleotide corresponds to the 182 to 113 bp region of the gp91^phox promoter that contains a BID binding site, while the PROX-BID-mut oligonucleotide is a mutated version of this sequence that no longer binds BID (see Footnote 2). The MID-BID oligonucleotide corresponds to the 261 to 212 bp region of the gp91^phox promoter and contains binding sites for both BID and IRF-2. The MID-BID-mut oligonucleotide corresponds to a mutated version of the BID-BID element that no longer binds either BID nor IRF-2 (see Footnote 2). The upward shift of the triplet of complexes formed by the 261 to 212 bp probe is due to the longer length of this probe compared to the 100 to 63 bp probe.

Additional experiments were performed to determine which member of the IRF family binds to the 92 to 80 bp ISRE. The two lower EMSA complexes are disrupted by antibody against IRF-2, but not by antibodies against IRF-1 or ISGF-3 (Fig. 3A, right panel), suggesting that the two lower complexes contain IRF-2. It is not unusual to detect multiple IRF-2 complexes, which may arise from proteolysis or post-translational modification (28, 29).

Previous work demonstrated that another ISRE at 233 to 222 bp of the gp91^phox promoter also produces a triplet of complexes when used as a probe in EMSA. This probe is also bound by IRF-2 and additionally by BID, a factor that interacts with several upstream gp91^phox promoter elements.² EMSA was therefore performed to evaluate whether the upper complex of the triplet generated by the 100 to 63 bp gp91^phox promoter probe contains BID. This complex is disrupted by the addition of PROX-BID oligonucleotide competitor that represents a BID binding site located at 145 bp of the gp91^phox promoter, but not by a mutated site (PROX-BID-mut) that no longer contains a BID binding site² (Fig. 3B, left panel). Furthermore, all three complexes of the triplet are disrupted by competition with the upstream ISRE-containing region from the gp91^phox promoter (MID-BID) that contains both BID and IRF-2 binding sites (Fig. 3B, left panel). However, none of the complexes are disrupted by a mutated version of this competitor (MID-BID-mut) that has lost both binding sites.

A direct comparison of the complexes generated by the two ISRE probes derived from the gp91^phox promoter reveals indistinguishable binding site specificities and relative mobilities (Fig. 3B, right panel). For both probes the upper complex is disrupted by the PROX-BID competitor, and the bottom two complexes are disrupted by the IRF-2 binding site competitor. The slightly slower mobility of the triplet generated by the upstream ISRE probe is due to the longer length of this probe compared to the downstream ISRE probe. These results suggest that the upper most complex generated by the 100 to 63 bp gp91^phox promoter probe contains BID. A conclusive examination of this hypothesis awaits the identification and/or molecular cloning of the BID activity.

The region from 100 to 76 bp of the gp91^phox promoter has been noted by Perez et al. (30) as exhibiting similarity to the myeloid activating transcription element (MATE) of the high affinity Fc receptor promoter. The MATE element has been demonstrated to be a binding site for PU.1 (31), an Ets factor largely restricted to macrophages and B cells (32). However, we found that a competitor corresponding to the 100 to 63 bp region of the gp91^phox promoter does not disrupt EMSA complexes formed with PU.1 and known PU.1 binding sites in the Fc receptor promoter (30) and SV40 enhancer (32) (data not shown). Furthermore, antiserum directed against PU.1 has no effect on complexes formed with the 100 to 63 bp gp91^phox promoter probe (data not shown). These results indicate that the 100 to 63 bp region of the gp91^phox promoter does not contain an authentic binding site for PU.1.

IRF-1 is a transcriptional activator that antagonizes the action of IRF-2 (33), is induced following IFN treatment (34), and has an identical DNA binding specificity as that exhibited by IRF-2 (22). Additional experiments were conducted to determine whether IRF-1 also binds to the 100 to 63 bp region of the gp91^phox promoter. A nuclear extract isolated from macrophages that express the gp91^phox gene generates a triplet of complexes similar to that produced by a nuclear extract isolated from HEL cells (Fig. 4A). The bottom two bands of the triplet are disrupted by antiserum directed against IRF-2, suggesting the IRF-2 activity is predominant in macrophage cell lines as well as in HEL cells. However, nuclear extracts isolated from IFN--induced HeLa cells generate an EMSA doublet that is disrupted by antibody to IRF-1, but not by antibody to IRF-2 (Fig. 4B). This indicates that the ISRE at 92 to 80 bp of the gp91^phox promoter is also an IRF-1 binding site. However, antibody against IRF-1 has no effect on complexes generated by the 100 to 63 bp probe using nuclear extract isolated from HEL cells, even following the removal of IRF-2 binding by the addition of IRF-2 antibody (data not shown). This indicates that HEL cells contain an undetectable level of IRF-1, consistent with previous observations that IRF-2 activity is predominant over IRF-1 in cells not exposed to IFN or virus (34, 35). Interestingly, IRF-2 binding activity is significantly more abundant in HEL cells than in K562 cells (Fig. 4C). This correlates with the observation that 100 to +12 bp gp91^phox promoter activity is higher in HEL cells than in K562 cells (9).

Both IRF-1 and IRF-2 bind to the 100 to 63 bp probe.

Three sets of mutations were identified that specifically ablate binding of IRF-2, BID, or both to the 100 to 63 bp gp91^phox promoter probe (Fig. 5). The IRF-mut probe specifically fails to bind IRF-2, while the BID-mut probe fails to interact with BID (Fig. 5). The slower migrating complex that appears in the BID-mut lane is a nonspecific complex, as it is disrupted by a variety of oligonucleotide competitors of unrelated sequence (data not shown). A BID + IRF-mut probe was also created that fails to interact with either BID or IRF-2. Each of these mutations was introduced individually into the 100 to +12 bp gp91^phox promoter/luciferase construct and transiently transfected into HEL and K562 cells to study the functional significance of IRF-2 and BID binding on gp91^phox promoter function.

Identification of mutations that specifically ablate the IRF or BID binding sites in the 100 to 63 bp gp91

Disruption of the BID binding site causes a 50% decrease of gp91^phox promoter activity in HEL cells, but has little effect in K562 cells (Fig. 6). Disruption of the IRF binding site causes a 93% decrease in gp91^phox promoter activity in HEL cells, and a 75% drop in K562 cells. Abolishment of both the IRF and BID binding sites decreases gp91^phox promoter activity to nearly the same level as the activity of the 84 to +12 bp promoter (Fig. 6). These data indicate that both IRF and BID binding sites are transcriptional activating elements that are necessary for gp91^phox promoter activity in the absence of transcriptional repression.

Overexpression of IRF-1 or IRF-2 Transactivates the 100 to +12 bp gp91^phox Promoter

The data presented above indicate that both IRF-1 and IRF-2 are able to bind to the ISRE at 92 to 80 bp of the gp91^phox promoter, and that IRF-2 is more abundant than IRF-1 in the cells used in this study, including macrophages that express the endogenous gp91^phox gene. This leads to the hypothesis that IRF-2 directs expression of the gp91^phox promoter in the absence of CDP-mediated transcriptional repression, a role opposite to the transcriptional repression activity commonly associated with this factor (33, 36). To directly test this hypothesis, plasmids that overexpress IRF-1 or IRF-2 were co-transfected with the wild type or IRF-mut versions of the 100 to +12 bp gp91^phox promoter/luciferase construct. Because HEL cells contain a high level of endogenous IRF-2 binding activity (Fig. 4C), K562 cells were used for the co-transfection experiments.

Overexpression of IRF-2 in K562 cells results in a 2.5-fold increase of gp91^phox promoter activity when compared to the vector control, while overexpression of IRF-1 causes a 5-fold increase in gp91^phox promoter activity (Fig. 7, solid bars). Neither IRF-1 nor IRF-2 transactivates a mutated gp91^phox promoter that lacks the IRF binding site (Fig. 7, striped bars). These data indicate that IRF-1 and IRF-2 are transcriptional activators of the gp91^phox gene promoter via binding to the ISRE site at 92 to 80 bp. The dominance of IRF-2 in HEL cells indicates that it is responsible for transcriptional activation mediated via the 92 to 80 bp IRF binding site in these cells.

DISCUSSION

This study examines the cis elements and cognate DNA-binding proteins that are required for gp91^phox promoter activity in the absence of CDP-mediated repression. Characterization of these elements should identify promoter components that are interfered with by transcriptional repression in nonphagocytic cells and which are also possibly required for appropriate gp91^phox expression in mature myeloid cells in which CDP DNA-binding activity is down-regulated.

The results demonstrate that the region between 100 and 85 bp is required for gp91^phox promoter activity. An ISRE at 92 to 80 bp serves as a binding site for IRF-1 and IRF-2, which were originally identified as important regulators of interferon and virus-induced genes (33, 35). Recent studies indicate that IRF-1 and IRF-2 are also involved in the regulation of cell proliferation, transformation, and apoptosis (37, 38, 39). Furthermore, IRF-1 and IRF-2 play critical roles during hematopoiesis. IRF-1 is involved in growth inhibition during myeloid differentiation induced by interleukin-6 and leukemia inhibitory factor (40), and deletion of IRF-1 is frequently observed in human leukemia and preleukemia myelodysplasia (41). IRF-1 is also required for the induction of the nitric oxide synthase gene in macrophages (42). IRF-1 null mice exhibit impaired development of CD8⁺ T cells (43), while B lymphopoiesis is suppressed in IRF-2 deficient mice (44).

The significance of the 92 to 80 bp IRF binding site in mediating gp91^phox promoter activity is confirmed by transient transfections in which specific ablation of this element results in decreased promoter activity in both HEL and K562 cells. The transcriptional activating activity of both IRF-1 and IRF-2 is further demonstrated by the ability of each to transactivate the gp91^phox promoter, an effect dependent on an intact ISRE. Abundant IRF-2 levels in HEL cells indicate that IRF-2, and not IRF-1, mediates transcription via the ISRE in this system. Furthermore, IRF-2 DNA-binding activity is dominant over that of IRF-1 in a macrophage cell line that expresses the gp91^phox gene, indicating that IRF-2 may also be involved in directing expression of the gp91^phox gene in mature myeloid cells. However, these findings do not exclude a role for IRF-1 in the myeloid-specific induction of gp91^phox expression following IFN- stimulation. When present in nuclear extract, IRF-1 does bind to the downstream ISRE, and transactivation studies indicate that IRF-1 induces approximately twice the expression from the 100 to +12 bp gp91^phox promoter as that produced by IRF-2.

IRF-2 has generally been described as a transcriptional repressor and functions by competing with the transcriptional activator IRF-1 for DNA-binding sites (33, 36). However, IRF-2 contains a latent transcriptional activation domain (36), and a recent study demonstrates that overexpression of IRF-2 causes a 2-fold activation of the histone H4 promoter (28). Hence, IRF-2 is similar to factors such as YY1, RAP-1, Dorsal, and Kruppel that exhibit both transcriptional activation and repression activities (45, 46, 47, 48, 49). Our findings reveal a second example of the ability of IRF-2 to function as a transcriptional activator. The relative strength of the transcriptional activating activity of IRF-2 on the gp91^phox promoter is comparable to that reported for IRF-2 action on the histone H4 promoter (28).

Subsets of myeloid cells require distinct cis elements to direct gp91^phox expression in vivo. For example, CGD patients who carry mutations in the gp91^phox promoter exhibit normal expression in 5-10% of the lineage (6, 7). Similarly, a CGD patient has been identified in whom gp91^phox transcription is restricted to myeloid cells of the eosinophil compartment (50). We have previously demonstrated that the 450 to +12 bp region of the gp91^phox gene directs myeloid-specific expression in transgenic mice, but in only a subset of monocyte/macrophages (4). We also found that CP1 binding to a CCAAT-box at 123 to 119 bp is necessary for de-repressed gp91^phox promoter activity in HeLa and K562 cells, but not in HEL cells (9). Variability in promoter element requirements may reflect heterogeneity in the complement of transcription factors present in various cell types. We hypothesize that abundant IRF-2 activity in HEL cells makes CP1 binding dispensable for gp91^phox promoter activity, hence providing a molecular mechanism to explain one example of variable cis element requirements for gp91^phox promoter activity.

Binding of a factor denoted BID to three upstream elements of the gp91^phox promoter is required for IFN--induced expression of the gp91^phox gene.² We now describe a fourth BID binding site at approximately 90 bp that directs gp91^phox promoter activity in the absence of upstream repressor elements. It is remarkable that both ISRE elements in the gp91^phox promoter serve as binding sites for BID and IRF-1/IRF-2. However, the properties of BID do not correspond to those of previously described IFN response factors, and two of the BID binding sites in the gp91^phox promoter do not conform to the ISRE consensus sequence.² Examination of the role of the newly described BID/IRF binding sites in the regulation of the gp91^phox promoter in response to IFN and more generally during terminal myeloid cell differentiation is currently under investigation.