Characterization of an Iron-responsive Promoter in the Protozoan Pathogen Trichomonas vaginalis *

Iron has been shown to regulate transcription in the protozoan pathogen Trichomonas vaginalis. In this study, a DNA transfection system was developed to monitor ap65-1 promoter activity in response to changing iron supply. In conjunction with electrophoretic mobility shift assay, iron-induced transcription of the ap65-1 gene was shown to be regulated by multiple closely spaced DNA elements spanning an iron-responsive region (−110/−54), including an iron-responsive DNA element (−98AGATAACGA−90), which overlaps with a 3′-MYB-like protein binding sequence (−95TAACGATAT−87), and three nearby T-rich sequences (−110ATTTTT−105,−78ATTATT−73, and−59ATTTTT−54). 5′- and 3′-flanking sequences of the iron-responsive region were shown to regulate basal transcription. A distal DNA regulatory region was shown to enhance both basal and iron-induced transcription. These findings delineate the DNA regulatory elements and nuclear proteins involving in iron-induced transcription of the ap65-1 gene, which provide useful tools for the future study of transcriptional regulation in T. vaginalis.

Human infection by the protozoan pathogen Trichomonas vaginalis causes one of the most common sexually transmitted diseases throughout the world (1). Although this protozoan infection usually manifests itself as self-limiting in males, it can impose serious health problems for female patients especially during pregnancy, and it is also implicated as a risk factor for cervical cancer and as a predisposition to human immunodeficiency virus contagion (2)(3). As one of the deepest branches of the eukaryotic lineage, this organism exhibits interesting features that deviate from higher eukaryotes and represents an important model system in phylogenetic studies (4). With the recent advent of gene transfer techniques for T. vaginalis (5), in-depth molecular and cellular research can be carried out in this organism.
T. vaginalis trophozoites colonize the epithelial surface of the human urogenital tract in which they obtain nutrients, multiply, and face a constant challenge from host immune surveillance. Iron, which is an essential nutrient for almost every organism, is particularly important for T. vaginalis as it regulates growth rate, metabolic activities, and the expression of certain virulence phenotypes such as cytoadherence and resistance to complement lysis (6 -11). At the molecular level, iron has been shown to up-regulate the expression of a number of cellular proteins including a group of putative adhesin molecules, the identities of which are still controversial (8,9,(12)(13)(14)(15)(16)(17)(18)(19). Iron has also been shown to regulate the phosphorylation level of a major surface immunogen P270, which may be responsible for immune evasion (20). These observations suggest that iron is a key modulator in the versatile cellular activities in T. vaginalis. Because the expression of some of the putative adhesin proteins can be inhibited by actinomycin D (8), ironinduced gene expression may be regulated at the transcription level.
The knowledge of transcriptional regulation in T. vaginalis is still very limited. T. vaginalis has been shown to use a metazoan initiator-like element spanning the transcription initiation site(s) to initiate transcription of messenger RNA (21). Further analysis of the ␣-succinyl Co-A synthetase gene (␣scs) 1 promoter also revealed two novel DNA elements within the Ϫ98/Ϫ69 region as essential for transcription initiation (22). However, the TATA boxes and other distal DNA regulatory elements commonly used in higher eukaryotes to regulate the basal transcription of messenger RNA have not been identified in T. vaginalis (22). These findings suggest that transcription machinery in T. vaginalis may deviate significantly from the well known machinery operating in higher eukaryotes.
In this study, the ap65-1 gene, which encodes a 65-kDa protein reputed to be one of the surface adhesin proteins (12), was selected as a model system to study iron-induced gene expression in T. vaginalis. The DNA regulatory elements in the ap65-1 promoter were characterized by promoter analysis in vivo in conjunction with DNA-protein interaction assays in vitro. The DNA regulatory elements distributed within Ϫ110/ Ϫ54 were found to regulate iron-induced gene expression, whereas those flanking this region were found to regulate basal transcription. A distal region was found to activate both basal and iron-induced transcriptional activities of the ap65-1 promoter. These findings provide a useful model system for future investigations of basal as well as iron-induced transcriptional regulation in T. vaginalis.

EXPERIMENTAL PROCEDURES
Culture-T. vaginalis axenic cultures were maintained at 37°C in TYI-S33 medium as described previously (23). The medium was supplemented with 10% heat-inactivated bovine calf serum without iron-* This work was supported in part by National Science Council Grant NSC89-2314-B001-011 and a grant from Academia Sinica, Taiwan, ROC. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM  fortification (Hyclone), and the iron concentration in this medium was estimated to be 1 M. Ferrous ammonium sulfate was added to the desired concentrations as described previously (8).
Molecular Cloning of the DNA Sequences Flanking the ap65-1 Gene-Sequences of the oligonucleotides used to clone the flanking regions of the ap65-1 genes and to construct the DNA transfection vectors are listed in Table I. An automatic DNA-sequencing method as described by the supplier (ABI) was used to verify the DNA sequences.
A Sau3AI genomic DNA library derived from T. vaginalis JH32A#4 was constructed in pBluescriptIIKSϩ (Stratagene). The sequence flanking 5Ј of the ap65-1 gene was amplified from the library using a gene-specific 3Ј-primer ap1 and either T7 or T3 on the vector as 5Ј primer by PCR amplification. The PCR products were then cloned into a pGEM-T vector (Promega), and a positive clone pAP5Ј with a 0.25-kb DNA sequence spanning the Ϫ235/ϩ35 region of the ap65-1 gene was obtained (Fig. 1A). An overlap DNA sequence spanning the Ϫ217/ϩ217 region was amplified from T. vaginalis T1 genomic DNA by a second PCR amplification using primers pϪ217 and ap2 and cloned into pGEM-T to produce pAP(Ϫ217/ϩ217). The 3Ј-untranslated region of the ap65-1 gene was amplified by PCR using the primer pair ap3 and ap4 from genomic DNA and cloned into pGEM-T to produce pAP3Ј.
Primer Extension-Cellular RNA was extracted using the UltraSpec RNA reagent (Biotek). Primer extension was performed as described previously (24), with the exception that the ␥-32 P-labeled oligonucleotide was purified using a NAP-5 gel filtration column (Amersham Biosciences, Inc.) and that primer extension was performed at 52°C by Moloney murine leukemia virus reverse transcriptase SuperScript II (Invitrogen). Oligonucleotides ap5 and tub3 were used to prime ap65-1 messenger RNA and tubulin messenger RNA, respectively (Table I).
Plasmid Construction-pAP5Ј was digested by BglII/SalI and fused together with a 1.7-kb lucϩ fragment excised from pSPlucϩ (Promega) by BglII/XhoI digestion, resulting in pAPlucϩ3Ј⌬. The insert from pAP3Ј was excised by EcoRI/NdeI digestion and cloned into EcoRI/NdeIdigested pAPlucϩ3Ј⌬ to produce pAPlucϩ ( Fig. 2A). The sequence spanning the Ϫ234/ϩ399 region of the ␤-tubulin gene was amplified from genomic DNA by PCR amplification using primers tub1 and tub3Ј (25) and cloned into pGEM-T to produce pTUB5Ј. The region spanning the Ϫ234/ϩ32 region of pTUB5Ј was then amplified by primers tub1 and tub 2 and cloned into pGEM-T. The insert was excised by digestion with SacII and BglII, and the resulted DNA fragment was cloned into SacII/ BglII-digested pAPlucϩ to produce pTUBlucϩ (Fig. 2B).
5Ј deletion mutants with the exception of pϪ114 were constructed by amplifying DNA from pAPlucϩ using one of the 5Ј primers at the defined site (Table I) and a 3Ј primer luc344R derived from the lucϩ gene (24). The PCR products were cloned into pGEM-T. The inserts were then excised by SacII/NarI and cloned into pAPlucϩ to replace the original SacII/NarI sequence. The SacII/EcoRV fragment was removed from pAPlucϩ by restriction enzyme digestions, and the resulting DNA was treated with Klenow DNA polymerase before ligation to produce pϪ114. All deletion constructs are named according to the location of the 5Ј end relative to the transcription start site (Fig. 3).
Targeted mutagenesis was performed by PCR to create mutations in pAPlucϩ. A restriction enzyme site was designed on oligonucleotides for each region to be mutated (Table II). To create mutations within the Ϫ230/Ϫ192 region, a PCR product was amplified from pAPlucϩ using a 5Ј primer with clustered mutations at the target site and luc344R as the 3Ј primer. To create two point mutations in the initiator region, a PCR product was amplified from pAPlucϩ using ap5Ј as the 5Ј primer and an antisense 3Ј primer m(ϩ1/ϩ3)-3Ј. To create mutations within the Ϫ187/Ϫ3 region with the exception of the Ϫ109/Ϫ102 and Ϫ101/Ϫ96 regions, a 5Ј-PCR product was amplified from pAPlucϩ using ap5Ј as the 5Ј primer and an antisense 3Ј primer at the target site, and a 3Ј-PCR product was amplified using a 5Ј primer at the target site and luc344R as the 3Ј primer. The PCR products were cloned into pGEM-T. The inserts were excised by appropriate enzymes and ligated with SacII/NarI-digested pAPlucϩ to produce a series of mutant constructs (see Fig. 4). To create mutations in the Ϫ109/Ϫ102 and Ϫ101/Ϫ96 regions, a PCR product was amplified from pAPlucϩ using a 5Ј primer at the target site and luc344R as the 3Ј primer. The PCR products were cloned into pGEM-T. The inserts were excised by EcoRV and NarI and ligated with EcoRV/NarI-digested pAPlucϩ to produce respective mutant constructs (see Fig. 4).
DNA Transfection and Luciferase Assay-T. vaginalis T1 cells grown to 1.5 ϫ 10 6 trophozoites ml Ϫ1 were diluted 10-fold with fresh medium and incubated overnight until cell density reached 1.5 ϫ 10 6 trophozoites ml Ϫ1 . Cells were harvested from cultures by centrifugation at 900 ϫ g for 10 min (GPR centrifuge, Beckman) and resuspended in fresh medium at a final concentration of 10 8 trophozoites ml Ϫ1 . The cells were passed through a 23-gauge needle gently four times using a 5-ml syringe to disperse cell clumps. An aliquot of 300 l of cell suspension was mixed with 60 g of plasmid DNA in a 0.4-cm gap ice-cooled electroporation cuvette (Invitrogen). Electroporation was performed at 300 V, 1000 microfarads, and 720 ohms using a BTX Electro Cell Manipulator 600 (BTX). Cells were kept on ice for 15 min after electroporation and divided into two tubes with fresh medium. A preliminary study using pTUBlucϩ (see below) to transfect cells from various T. vaginalis isolates revealed that transfection efficiency of cells from the T1 isolate was at least 100-fold higher than cells from JH32A#4, NIH-C1, or T068II isolates under our experimental conditions. 2 The T. vaginalis T1 isolate was therefore selected for the transfection experiments performed in this report. Luciferase activity of transfected cells was performed as described previously (24), with the exception that 200 g ml Ϫ1 N ␣ -p-tosyl-L-lysine chloromethyl ketone was used as protease inhibitor to replace aprotinin and that the cell lysate was directly assayed without pretreatment at Ϫ70°C. The data were analyzed by one-way analysis of variance (ANOVA) in SPSS software version 8.0 (1997).
Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay-Nuclear extract was prepared as described previously (24), with the exception that 200 g ml Ϫ1 N ␣ -p-tosyl-L-lysine chloromethyl ketone was used as protease inhibitor instead of aprotinin and that a Douncetype homogenizer (Wheaton) was used to homogenize cells. A BCA protein quantification kit was used to determine protein concentration in nuclear lysate as described by the supplier (Pierce). Probe labeling and electrophoretic mobility shift assay were performed as described previously (24).

RESULTS
Mapping the Start Site of ap65-1 Messenger RNA-The 5Јflanking sequence of the ap65-1 gene was first cloned from the T. vaginalis JH32A#4 and T1 isolates by two separate PCR amplifications, and an identical 0.25-kilobase pair DNA sequence was obtained ( Fig. 1A) (GenBank TM accession number AF364546), indicating that the ap65-1 gene is conserved between the two isolates. The transcription start site of ap65-1 messenger RNA was then mapped by primer extension using RNA extracted from T. vaginalis T1 cells grown in 12 M iron. A major extension product of 71 nucleotides and a minor extension product of 72 nucleotides were consistently produced in reactions priming 50 g of cellular RNA with ␥-32 P-labeled ap5 (Fig. 1B, lane 1). The major extension product is mapped to an adenosine 14 upstream of the translation start site, indicating that the initiator-like sequence most proximal to the translation start site serves as the initiator element of the ap65-1 promoter. The adenosine residue in this initiator element is defined as ϩ1 in the text.  Primer extension reactions were then performed to analyze transcriptional regulation of the ap65-1 gene by iron using RNA samples from a cloned T. vaginalis T1 cell line. In each of these reactions, 50 g of cellular RNA was primed with ␥-32 Plabeled ap5 or tub3. Overall extension signals of ap65-1 messenger RNA from cells treated with 250 M iron were 3-fold higher than those from cells treated with 1 M iron (Fig. 1B,  lanes 2 and 3, respectively). Consistent with previous findings (25), the primer extension of ␤-tubulin messenger RNA resulted in a major product of 58 nucleotides, and the intensity of this band only changed slightly in response to changing iron supply (Fig. 1B, lanes 4 and 5).
Promoter Assay-Two luciferase expression plasmids, pAPlucϩ and pTUBlucϩ (Fig. 2), were used to monitor transcriptional activities of the ap65-1 and ␤-tubulin promoters, respectively, in T. vaginalis T1 cells. In these experiments, luciferase activity in cells transfected with pSPlucϩ was taken as background. A low level of luciferase activity (ϳ80-fold above background) was first detected in pAPlucϩ-transfected cells at 11 h post-transfection. The activity increased steadily until reaching an optimal level (ϳ750-fold above background) at 32 h post-transfection and declined slowly as the stationary phase of cell growth was reached. Luciferase activity of pAPlucϩ-transfected cells measured at 30 h post-transfection exhibited an iron-dependent increase from 1 to 500 M iron and leveled off at a higher concentration ( Fig. 2A). Iron concentration below 1 M was not tested, because it requires the addition of the iron-chelator 2-2Ј-dipyridyl, which retards the growth of transfected cells in our experimental conditions. Luciferase expression in pTUBlucϩ-transfected cells was 370-fold and nearly 30,000-fold above background at 13 and 28 h posttransfection, respectively. However, luciferase activity in pTUBlucϩ-transfected cells was independent of iron concentration (Fig. 2B). Whereas transcriptional activity of the ap65-1 promoter was enhanced by 15-fold in the presence of 250 M iron, it was rather insensitive to other divalent metal ions such as Ca 2ϩ , Co 2ϩ , Cu 2ϩ , Mg 2ϩ , Mn 2ϩ , and Zn 2ϩ at a similar concentration (Fig. 2C), indicating that this promoter is specifically responsive to ferrous ion.
Mapping the Regulatory Regions of the ap65-1 Promoter-In subsequent experiments, luciferase activities in cells grown in 1 and 250 M iron measured at 30 h post-transfection were  taken as basal and inducible transcriptional activities, respectively. Luciferase activity measured in pAPlucϩ-transfected cells from low iron cultures was taken as the original activity (100%).
The regulatory regions in the ap65-1 promoter were first mapped by testing the transcription efficiency of a series of pAPlucϩ 5Ј deletion mutants in transfected cells (Fig. 3). In these experiments, pAPlucϩ-transfected cells exhibited an average ϳ16-fold induction with iron treatment. Basal luciferase activity was reduced to ϳ35%, ϳ24%, and ϳ18% original level with deletions to Ϫ217, Ϫ204, and Ϫ184, respectively, and iron-induced luciferase expression was reduced to ϳ5.5-, ϳ3.3-, and ϳ3-fold, respectively. Basal luciferase activity remained at ϳ18% original level with deletions to Ϫ133, Ϫ114, and Ϫ80, but the induction level with iron treatment reduced to 2.2-, 1.3-, and 0.9-fold, respectively. Basal luciferase activity dropped to 11 and 7% original level with deletions to Ϫ40 and Ϫ12, respectively, without obvious iron-induced gene expression. With the deletion of the 3Ј-untranslated region (pAP lucϩ3Ј⌬), luciferase activity decreased to 10% original level, but the induction level remained at 14-fold with iron treatment. These findings suggest that iron-inducible gene expression is primarily regulated by the DNA element(s) distal to the transcription start sites.
The 5Ј-flanking sequence in pAPlucϩ was further studied by scanning mutagenesis (Fig. 4). Significant reduction in basal luciferase expression was observed in mutants with clustered mutations spanning the Ϫ121/Ϫ102 (Fig. 4, pmϪ121/Ϫ118, pmϪ114/Ϫ111, and pmϪ109/Ϫ102) and Ϫ52/Ϫ39 (Fig. 4, pmϪ52/Ϫ48 and pmϪ44/Ϫ39) regions. Their activities were reduced to ϳ30% original level. The most severe reduction in basal transcription was seen in a mutant with two point mu-tations in the reputed initiator region (Fig. 4, pmϩ1/ϩ3), which was only ϳ2% original level. On the other hand, a significant reduction in iron-induced luciferase expression was seen in mutants with clustered mutations spanning the Ϫ109/Ϫ56 region (Fig. 4, pmϪ109/Ϫ102, pmϪ101/Ϫ96, pmϪ95/Ϫ81, pmϪ80/Ϫ66, and pmϪ61/Ϫ56). These findings suggest that the basal transcription of the 65-1 gene is regulated by the DNA elements spanning the Ϫ121/Ϫ102 and Ϫ52/Ϫ39 regions in concert with the proximal initiator sequence, and iron-induced gene expression is regulated by the DNA element(s) spanning the Ϫ109/Ϫ56 region. In conjunction with the deletion mapping experiments (Fig. 3), these results also suggest that the Ϫ230/ Ϫ184 region may contain DNA regulatory elements essential for optimal transcriptional activity of the ap65-1 promoter.
Binding of Nuclear Proteins to the Iron-responsive Region-Nuclear proteins interacting with the DNA regulatory elements in the iron-responsive region of the ap65-1 promoter were then studied by electrophoretic mobility shift assays.
A major DNA-protein complex was detected in 8% polyacrylamide gel testing for binding of nuclear proteins to 32 P-labeled Ϫ68/Ϫ45 (Fig. 5). Similar banding patterns were observed in reactions using nuclear lysate from cells treated with either low or high iron (data not shown). Complex formation was greatly inhibited by 250ϫ molar excess of (Ϫ68/Ϫ45) but was only slightly inhibited by 1000ϫ molar excess of (Ϫ107/Ϫ85) (Fig. 5A). Further competition assays were performed using 1000ϫ molar excess of mutated sequences m(Ϫ68/Ϫ45) series, each with 3-bp mutation within the Ϫ68/Ϫ45 region (Fig. 5B). The results showed that Ϫ59 ATTTTT Ϫ54 is a nuclear protein binding site. Mutation of the adenosine residue to a guanosine or cytosine residue in this binding site resulted in less efficient competition (Fig. 5C), indicating that a sequence with five contiguous thymine residues is a potential nuclear protein binding site and that the adenosine residue preceding the thymine residues is preferred for optimal DNA-protein interaction. A similar protein-DNA complex was also formed in reactions using 32 P-labeled (Ϫ85/Ϫ68), and the protein binding site was localized to Ϫ78 ATTATT Ϫ73 (data not shown). Visual examination reveals a third T-rich sequence Ϫ110 ATTTTT Ϫ105 , which may also serve as a nuclear protein-targeting site. These T-rich sequences are referred to as T-boxes subsequently.
On the other hand, two major DNA-protein complexes were detected in 8% polyacrylamide gel testing for binding of nuclear proteins to 32 P-labeled (Ϫ107/Ϫ85) (Fig. 6A). Similar banding patterns were observed in reactions using nuclear lysate from cells treated with either low or high iron (data not shown). The formation of both complexes was abolished by 250ϫ molar excess of (Ϫ107/Ϫ85) but not by 250ϫ molar excess of (Ϫ133/ Ϫ110). Further competition assays were performed using 250ϫ molar excess of the mutated sequences m(Ϫ107/Ϫ85) series, each with a 3-bp mutation within the Ϫ107/Ϫ85 region (Fig.  6B). The results showed that Ϫ95 TAACGATAT Ϫ87 contains a nuclear protein binding site. The sequence Ϫ95 TAACGATAT Ϫ87 is similar to the DNA binding sequences of the MYB-family transcription factors (consensus (c/t)AACG(g/t)) in higher eukaryotes (26 -29). In contrast to another distinct MYB-like protein binding site ( Ϫ44 TATCGT Ϫ39 ) in the ap65-1 promoter, 3 the sequence Ϫ95 TAACGATAT Ϫ87 interacting with larger nuclear MYB-like proteins is referred to as the tvMYBl binding site.
The detection of potential iron-induced nuclear protein-DNA complex was further explored by using a 32 P-labeled m7(Ϫ107/ Ϫ85) probe in which the 3Ј moiety of the tvMYBl binding sequence is mutated (see sequence in Fig. 6B). As expected, nuclear protein-DNA complexes targeting to the tvMYBl binding site were abolished in binding reactions using nuclear proteins from cells without iron treatment (Fig. 7A, lane 1). By contrast, two major DNA-protein complexes, which migrated slower than the complexes targeting to the tvMYBl binding site, formed in binding reactions using nuclear proteins from iron-treated cells (Fig. 7A, lane 2). These complexes were displaced efficiently by 50ϫ molar excess of m7(Ϫ107/Ϫ85) (Fig.  7A, lanes 3-5) but not by up to 500ϫ molar excess of (Ϫ133/ Ϫ110) (Fig. 7A, lanes 6 -8). Further competition assays were performed using 200ϫ molar excess of (Ϫ107/Ϫ85) and the mutated sequences m(Ϫ107/Ϫ85) series. The DNA-protein complexes were completely displaced by m7(Ϫ107/Ϫ85) (Fig.  7B, lane 10). They were not displaced by m5(Ϫ107/Ϫ85) and m6(Ϫ107/Ϫ85) but were displaced to a lesser extent by m4(Ϫ107/Ϫ85) than by any other mutated sequences or (Ϫ107/ Ϫ85) (Fig. 7B). These observations indicate that the DNA sequence centered at Ϫ98 AGATAACGA Ϫ90 contains a targeting site for iron-induced nuclear proteins, and its flanking sequences may also contribute to binding affinity. This possibility remains to be investigated. The DNA sequence Ϫ98 AGATA-ACGA Ϫ90 is referred to as the iron-responsive DNA element. In conjunction with Fig. 6, these observations indicate that the 3Ј moiety of the iron-responsive DNA element overlaps with 5Ј moiety of the tvMYBl binding site. DISCUSSION Transcriptional regulation has been implicated as one of the major regulatory mechanisms in modulating expression of certain T. vaginalis virulence phenotypes in response to changing iron supply (8). In this study, the ap65-1 gene was selected as a model system to investigate iron-mediated transcriptional regulation in T. vaginalis. Using primer extension and transient luciferase expression assays (Figs. 1B and 2), we found that steady-state ap65-1 messenger RNA as well as transcriptional activity of the episomal ap65-1 promoter in T. vaginalis T1 cells are positively regulated by iron ( Fig. 2A). We also found that the transcriptional activity of the ap65-1 promoter is insensitive to other divalent metal ions (Fig. 2C). On the other hand, ␤-tubulin messenger RNA and transcriptional activity of the episomal ␤-tubulin promoter are independent of changing iron supply (Fig. 2B). These findings are consistent with previous results describing the expression features of the ap65 gene family in other T. vaginalis isolates (8,12,15) and show that our experimental system is suitable for the study of iron-regulated expression of the ap65-1 gene.
Primer extension experiments and mutational analysis of the ap65-1 promoter in pAPlucϩ revealed that the basal transcription of the ap65-1 gene in T. vaginalis T1 cells is primarily regulated by a conserved initiator element closest to the translation start site (Figs. 1B and 4). In good agreement with the properties of the metazoan initiators (30,31), this initiator is both essential and sufficient to confer nearly 7% basal transcriptional activity (Figs. 3 and 4). This minimal transcriptional activity can be activated nearly up to 15-and 250-fold in low and high iron environments, respectively, in concert with distinct sets of distal DNA elements grouped into overlapping basal and iron-responsive regions at the proximal site and a discrete activation region at the distal site (Figs. 3 and 4). Unlike the distal DNA regulatory elements identified in the ␣-scs promoter (22), the mutation of any one of these distal DNA regulatory elements only resulted in at most 3-fold reduction in basal transcriptional activity (Fig. 4). Common DNA regulatory elements used by the ␣-scs and ap65-1 promoters  4 -11), each with a 3-bp mutation, were included in the reactions. In C, a 1000ϫ molar excess of (Ϫ68/Ϫ45) (lane 3) and the mutated sequences m(Ϫ68/Ϫ45) series, each with a single point mutation at Ϫ59 (lanes 4 -6), were included in the reactions. The reaction mixtures were separated in 8% polyacrylamide gel by electrophoresis. The DNA sequence of (Ϫ68/Ϫ45) (uppercase letters and hyphens) and its mutated sequences (lowercase letters) are listed. The nuclear protein targeting sequence is underlined.
were not found other than the conserved initiator elements.
The most intriguing feature of the ap65-1 promoter is the presence of multiple closely spaced DNA regulatory elements spanning Ϫ110/Ϫ54 to regulate iron-induced transcription. These DNA elements include an iron-responsive DNA element overlapping with a tvMYBl binding site and three flanking FIG. 6. Binding of nuclear proteins to the (؊107/؊85) region of the ap65-1 promoter. 10 g of nuclear extract was incubated with ␥-32 P-labeled (Ϫ107/Ϫ85) (lane 2) for 20 min at room temperature. In A, 10ϫ (lanes 3 and 6), 50ϫ (lanes 4  and 7), and 250ϫ (lanes 5 and 8) molar excesses of (Ϫ107/Ϫ85) (lanes 3-5) and (Ϫ133/Ϫ110) (lanes 6 -8) were included in the reactions. In B, 250ϫ molar excess of (Ϫ107/Ϫ85) (lane 3) and the mutated sequences m(Ϫ107/Ϫ85) series (lanes 4 -11), each with a 3-bp mutation, were included in the reactions. The reaction mixtures were separated in 8% acrylamide gel by electrophoresis. The DNA sequence of (Ϫ107/Ϫ85) (uppercase letters and dashes) and its mutated sequences (lowercase letters) are listed. The nuclear proteintargeting sequence is underlined.  6), 250ϫ (lanes 4 and 7), and 500ϫ (lanes 5 and 8) molar excesses of m7 (Ϫ107/Ϫ85) (lanes 3-5) and (Ϫ133/Ϫ110) (lanes 6 -8) were included in the reactions. In B, 200ϫ molar excess of (Ϫ107/Ϫ85) (lane 3) and the mutated sequences m(Ϫ107/Ϫ85) series (lanes 4 -11) (sequences as shown in Fig. 6B) were included in the reactions. The reaction mixtures were separated in 8% acrylamide gel by electrophoresis. In C, the DNA sequence of the iron-responsive region is listed. The targeting sites for constitutively expressed nuclear proteins (tvMYBl binding sequence and T-boxes) are underlined. The targeting site for iron-induced nuclear proteins (ironresponsive DNA element) is indicated by a line above the sequence.
T-boxes as summarized in Fig. 7C. It appears that the inaccessibility of the iron-responsive DNA element to iron-induced nuclear proteins in electrophoretic mobility shift assays is resulting from preoccupation of the site with constitutively expressed MYB-like nuclear proteins (Figs. 6 and 7), and these two distinct types of nuclear DNA-binding proteins may compete for same binding site in vivo to fine tune the transcription of the ap65-1 gene in response to environmental stimuli. This possibility remains to be examined. It is tempting to speculate that the constitutively expressed nuclear proteins targeting to the flanking T-boxes may actively participate in the formation of iron-induced transcriptional complex surrounding the ironresponsive DNA element, because mutation of any one of these three T-boxes resulted in a significant loss of iron-responsiveness (Fig. 4). Whether iron induces de novo biosynthesis of certain transcription factor(s) or modifies certain existing transcription factor(s) to interact with the iron-responsive DNA element has yet to be determined. The iron responsive region alone is also insufficient to confer iron-inducible gene expression without the distal activation region (Fig. 3), indicating close interactions of potential transcription factors targeting to each DNA regulatory element in these regions. It is clear that iron-mediated transcriptional regulation in T. vaginalis is distinct from other iron-mediated transcriptional regulations of iron acquisition, storage, or detoxification in prokaryotes (32)(33)(34), yeasts (35,36), fungi (37), and plants (38). Because the sequences in the 5Ј-and 3Ј-untranslated region of ap65-1 messenger RNA are not involved in the iron-induced transcriptional activity of pAPlucϩ (Fig. 3), the iron-regulated ap65-1 gene expression is unlikely to occur at a post-transcriptional or translation level similar to the iron-induced gene expression in more complex organisms (39).
In summary, our study of the ap65-1 promoter provides an excellent model system for future investigations on basal transcription as well as iron-induced transcription in T. vaginalis, one of the earliest diverging eukaryotic single cells.