Calcium Modulates Promoter Occupancy by the Entamoeba histolytica Ca2+-binding Transcription Factor URE3-BP*

The Entamoeba histolyticaupstream regulatory element 3-binding protein (URE3-BP) binds to the URE3 sequence of the Gal/GalNAc-inhibitable lectinhgl5 and ferredoxin 1 (fdx) gene promoters. This binding can be inhibited in vitro by addition of calcium. Two EF-hand motifs, which are associated with the ability to bind calcium, are present in the amino acid sequence of URE3-BP. Mutation of the second EF-hand motif in URE3-BP resulted in the loss of calcium inhibition of DNA binding as monitored by electrophoretic mobility shift assay. Chromatin immunoprecipitation assays revealed that URE3-BP was physically bound to thehgl5 and fdx promoters in vivo. Parasite intracellular calcium concentrations were altered by changes in extracellular calcium. Promoter occupancy was lost when intracellular calcium levels were increased by coordinate increases in extracellular calcium. Increased intracellular calcium also resulted in decreased levels of URE3-BP mRNA. Together these results demonstrate that changes in extracellular calcium result in changes in URE3-BP mRNA and in the ability of URE3-BP to bind to URE3-containing promoters. Modulation of URE3-BP by calcium may represent an important mechanism of control of gene expression inE. histolytica.

The early branching eukaryote Entamoeba histolytica is a human parasite that is the etiologic agent of amebic dysentery and liver abscess. Only one of every 10 infections leads to disease (1), and the parasite and host factors that control the outcome of infection are not well understood. Alteration in transcriptional control of certain crucial genes may contribute to the expression of a virulence phenotype. Padilla-Vaca et al. (1) demonstrated that co-cultivation of E. histolytica with Escherichia coli O55 resulted in increased virulence and a decrease in the expression of the Gal/GalNAc-inhibitable lectin light subunit. Ramakrishnan et al. (2) have shown alterations in the hgl genes transcribed in trophozoites derived from liver abscesses compared with those transcribed in established cell cultures. Bruchhaus et al. (3) have shown, in similar work, changes in the expression of over 55 other E. histolytica transcripts.
There has been considerable divergence in the mechanisms of transcription of the early branching E. histolytica from later branching eukaryotes such as Homo sapiens and Saccharomyces cerevisiae. For instance the E. histolytica core promoter for protein encoding genes consists of a novel GAAC element, in addition to a TATA and INR (4 -6) and contains short regulatory 5Ј and 3Ј sequences (7). The RNA polymerase II is also unusual for a eukaryote in that it is resistant to ␣-amanitin (8).
To increase our understanding of transcriptional regulation in this organism we investigated the mechanisms of transcriptional control of a well characterized virulence protein, the galactose-and N-acetyl-D-galactosamine-inhibitable lectin (Gal/GalNAc-inhibitable lectin), which is essential for parasite adherence and contact-mediated cytolysis. The promoter for one of the genes encoding the lectin heavy subunit (hgl5) contains five major regulatory regions (upstream regulatory elements 1-5 (URE1-5)) 1 upstream of the core promoter (4) and a GAAC sequence motif that predominantly influences the rate and site of transcription initiation (4 -6). The hgl5 UREs differ in whether they regulate transcription via the TATA or the GAAC elements, with the URE3 exerting its effect on mRNA transcription via GAAC (9). Interestingly, mutation of the URE3 motif in the hgl5 promoter leads to an increase in transcription and in the fdx promoter decreased promoter strength (10). This indicates that, as is the case for the mouse contextdependant Pax-5 transcription factor, the URE3 sequence can mediate both positive and negative control in different milieu (11). Two different genes encode the ferredoxin proteins of E. histolytica. The fdx URE3-containing promoter regulates the gene that encodes the ferredoxin 1 protein. This transcript is down-regulated in metronidazole-resistant amebae, in contrast with the level of the ferredoxin 2 transcript (fd2), which is unaltered (12). Bruchhaus et al. (3) discovered that fd2 mRNA is one of the transcripts up-regulated in ameba cultured from liver abscesses. These two papers suggest that ferredoxins may be tightly regulated in E. histolytica. Investigation of the sequence surrounding the fd2 gene showed that a URE3 motif was located 1.2 kb 5Ј of the initiation codon (International * This work was supported by National Institutes of Health Grant AI 37941. Sequence data pertaining to the URE3-BP genomic context was obtained from the Entamoeba histolytica Genome Data base. The sequences were made freely available by the Pathogen Sequencing Unit at the Sanger Institute (ftp.sanger.ac.uk/pub/pathogens/E_histolytica) and the TIGR Entamoeba histolytica Genome Project (www.tigr.org/ tdb/e2k1/eha1/). Preliminary sequence data for E. histolytica are deposited regularly into the GSS division of GenBank TM . The sequencing effort is part of the International Entamoeba Genome Sequencing Project and is supported by an award from NIAID, National Institutes of Health. 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.
ʈ Entamoeba Genome Sequencing Project) 2 , but it is unclear whether this motif would play any role in fd2 regulation.
A yeast-one-hybrid screen of an E. histolytica cDNA library using the URE3 element as "bait" identified the URE3-binding protein (URE3-BP) (14). Analysis of the URE3-BP amino acid sequence did not identify any canonical DNA binding motifs but did reveal the presence of two EF-hand motifs, which are correlated with the ability of a protein to bind to calcium (15). This suggested that the URE3-BP protein might function as both a calcium sensor and a transcription factor. The only other reported sequence-specific DNA-binding protein known to contain EF-hand motifs is DREAM. DREAM is a human neuronal protein (also known as Calsenilin and KChIP3) that functions both as a transcription factor (16,17) and as a calcium sensor (16 -20). DREAM contains four EF-hand sequence motifs. Typically, the binding of Ca 2ϩ to EF-hands induces structural changes that alter the function of the protein. In the case of DREAM, binding of Ca 2ϩ to its EF-hands leads to a reduced affinity for sequence-specific binding to its target downstream regulatory element sequence (16).
Previous analysis of the recombinant URE3-BP protein sequence-specific DNA binding by electrophoretic mobility shift assay (EMSA) showed that recognition of a URE3-containing oligonucleotide was blocked in vitro by calcium (14). In this report, we tested the in vivo role of the EF-hand motifs of URE3-BP in protein-promoter interactions.

EXPERIMENTAL PROCEDURES
Cultivation of E. histolytica-E. histolytica strain HM1:IMSS trophozoites were grown at 37°C in TYI-S-33 medium containing penicillin (100 units/ml) (Invitrogen) and streptomycin (100 g/ml) (Invitrogen) (21). In experiments examining the effect of calcium on transcription, E. histolytica was seeded at a concentration of 2-4 ϫ 10 4 /ml 18 h prior to the addition of medium containing 5 mM MgCl 2 and 5 mM EDTA to sequester the serum calcium (1.4 -2 mM Ca 2ϩ ) calculated using the dissociation constants of the chelating ligands and the CaLBuf Program (ftp.cc.kuleuven.ac.be/pub/droogmans/cabuf.zip (22). The resulting calculated free calcium was in the order of 3-5 M. The addition of 5.4 mM CaCl 2 raised Ca 2ϩ to 1 mM. Cells were harvested at the time points indicated under "Results." RNA Isolation-Approximately ϳ2.4 -5 ϫ 10 5 E. histolytica were lysed by the addition of 0.8 ml of TRIzol reagent (Invitrogen), and total RNA was purified according to the manufacturer's directions. RNA of greater than 200 nucleotides in length was isolated from total RNA by the RNeasy protocol (Qiagen) after first treating the total RNA preparation with DNase I (Roche Molecular Biochemicals) to remove contaminating genomic DNA.
Cloning and Expression of His 6 -(wt)URE3-BP and His 6 -EF(2)mut-URE3-BP-The wild-type URE3-BP His 6 -tagged fusion protein expression vector was constructed as described previously (14). Mutagenesis of the EF-hand motif (2) of URE3-BP was done by a two-stage PCR procedure. The 3Ј sequences of URE3-BP were amplified and mutated using the oligonucleotides GTTATGAACGCTAGAGCTAGAAGTG and AAATGTCGACTTATTCCAAGAGGGAAGTAACAACG that replaced the conserved first and third amino acids of the EF-hand motif with two alanine residues. The resulting PCR product was denatured and used along with an oligonucleotide AAAAGCTCTTCAAACATCA-ACCACCT-GTAGCTAATTTC to amplify the mutated copy of URE3-BP. This DNA was then subcloned into the pCRT7/NT-TOPO expression vector (Invitrogen). The mutated gene was sequenced to confirm the presence of the desired mutation. The expression of the His 6 -tagged fusion protein was induced in E. coli BL21(DE3) (F Ϫ ompT gal [dcm] [lon] hsdSB, (rB-mB-)) cells grown in Luria-Bertani medium with 100 M isopropylthio-␤-galactopyranose. The recombinant protein was affinity-purified by use of nickel-chelate resin according to manufacturer's directions (Qiagen) and then dialyzed against DNA binding buffer (10 mM Tris-HCl, pH 7.9, 50 mM NaCl, 1 mM EDTA, 20% glycerol).
Sequence and Sequence Analysis Software-The upstream and downstream non-coding sequences of URE3-BP were analyzed using the Wisconsin Package Version 10.0, Genetics Computer Group (GCG) soft-ware and the Basic Local Alignment Search Tool (BLAST) at NCBI (23,24). Sequence data pertaining to the URE3-BP genomic context were produced by the TIGR Entamoeba histolytica Genome Project (www. tigr.org/tdb/e2k1/eha1/) and the Pathogen Sequencing Unit at the Sanger Institute (www.sanger.ac.uk/Projects/E_histolytica/blast_ server.shtml), which are both part of the International Entamoeba Genome Sequencing Project. These sequences were obtained from ftp. sanger.ac.uk/pub/pathogens/E_histolytica/and the GSS division of Gen-Bank TM . The University of Virginia Biomolecular Research Facility performed all other sequencing.
Electrophoretic Mobility Shift Assay-Electrophoretic mobility shift assays with oligonucleotides hgl5-URE3 and hgl5-MUT were performed as described previously (10). To create the radiolabeled hgl5-URE3 and hgl5-MUT probes complementary oligonucleotides were annealed and then labeled with the large DNA polymerase I subunit (Klenow) and [␣-32 P]dATP. The URE3 DNA sequence was mutated to TTA-GAATTC from the wild-type sequence of TTATCTTAT in the hgl5-MUT probe. EMSA reactions contained 3 fmol of radiolabeled probe, 0.01 g/l poly(dI⅐dC), and 0.05 g/l of recombinant URE3-BP protein purified from bacterial lysates (Qiagen) in EMSA buffer (10 mM Tris-HCl, pH 7.9, 50 mM NaCl, 3 mM MgCl 2 , 1 mM EDTA, 0.5 mg/ml nonfat milk powder (Carnation), 3% glycerol, 0.05 mg/ml bromphenol blue). When indicated 1 mM CaCl 2 was added that was calculated by the CalBuf Program of Droogmans (22) to bring the Ca 2ϩ concentration to 86 M. The reaction was incubated at 4°C for 1 h prior to electrophoresis on a non-denaturing polyacrylamide gel for 2-3 h. The gel was then fixed, dried, and quantitated by PhosphorImager analysis.
Immunodetection-Mice were immunized as described previously (14), hybridoma cell lines producing anti-URE3-BP mAb were prepared, and monoclonal antibodies against URE3-BP (4D6 and 3E6) were produced. Protein samples of interest were electrophoresed on a 12% SDSpolyacrylamide gel, transferred to a polyvinylidene difluoride membrane (Millipore) and incubated for 1 h at room temperature in 5% nonfat dry milk in blot wash buffer (50 mM Tris, pH 7.4, 200 mM NaCl, 0.1% Tween 20). The blot was then incubated overnight in 2% nonfat dry milk in blot wash buffer containing 10 g/ml of mAb 4D6 and 5 g/ml of mAb 3E6. After three 5-min washes in 2% nonfat dry milk in blot wash buffer, the membranes were incubated for 1 h with horseradish peroxidase-conjugated goat anti-mouse IgG (Fc-specific) antibody (Sigma) at a dilution of 1:1500. The secondary antibody was detected using the ECL Western blotting detection system according to the manufacturer's directions (Amersham Biosciences) and was visualized by exposure of the blot to BioMax MR-1 film (Eastman Kodak Co.).
Calcium Calibration-Amebae were prepared and labeled with the acetoxy-methyl ester of the Ca 2ϩsensitive fluorescent dye indo-1 (indo/ AM; Molecular Probes) as described by Carbajal et al. (25) with the following modifications. Labeling was performed in a buffer of 10 mM Hepes-HCl, pH 7.2, 140 mM NaCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 5 mM KCl, 10 mM glucose, 0.1% bovine serum albumin, and 0.2% pluronic F-127 (Molecular Probes) buffer containing 26.7 M indo-1/AM. The acetoxymethyl ester of indo-1 passively diffuses across cell membranes, and once inside the cell this ester is cleaved by intracellular esterases to release indo-1. This dye is a cell-impermeable and UV light-excitable fluorescent Ca 2ϩ indicator. Trophozoites were incubated in the loading buffer at a concentration of 1 ϫ 10 6 amebae per ml for 30 min at 37°C followed by 30 min at room temperature. After incubation amebae were washed as described (25) and then suspended at a concentration of 2 ϫ 10 5 amebae per ml in a modified buffer, Buffer B (20 mM Hepes-HCl, pH 7.2, 140 mM NaCl, 5 mM EDTA, 5 mM MgCl 2 ) containing MgCl 2 and EDTA to generate a store of sequestrated calcium. Before measuring [Ca 2ϩ ] i cells were warmed to 37°C. Changes in [Ca 2ϩ ] i after the addition of CaCl 2 were monitored in an SLM 8100C spectrofluorometer (SLM/Aminco) using methods published previously (26 -28). CaCl 2 was added as calculated by the CalBuf Program of Droogmans (22) to give external free calcium concentrations of 7 M free Ca 2ϩ (1.2 mM CaCl 2 ), 608 M Ca 2ϩ (5 mM CaCl 2 ), and 1.1 mM Ca 2ϩ (5.7 mM CaCl 2 ). After analysis amebae were pelleted, and the fluorescence of the supernatant was measured to control for contributions to the fluorescence measurements from cell debris.
Immunoprecipitation and Chromatin Immunoprecipitation-The mouse anti-URE3-BP mAb 4D6 (described above) or a control mAb of the same isotype (anti-lectin mAb 7F4) was bound to protein G Dynabeads (Dynal). They were then used to immunoprecipitate chromatin from amebic nuclear extracts prepared by a modification of methods described previously (10). Briefly, nuclei were harvested from amebae that had been cross-linked in 1% formaldehyde in either phosphatebuffered saline or Buffer B ϩ 1 mM Ca 2ϩ (5.7 mM CaCl 2 ). Nuclear extracts were prepared from 1 ϫ 10 8 amebae preincubated in Buffer A (10 mM Hepes-KOH, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 2 mM (2S,3S)trans-epoxysuccinyl-L-leucylamido-3-methylbutane, and 2 mM 4-(2-aminoethyl) benzenesulfonylfluoride, HCl) for 20 min prior to sedimentation (900 ϫ g for 5 min) and resuspension in Buffer A containing 6% Nonidet P-40. They were then spun through a Qiashredder (Qiagen) for 1 min and then left in Buffer A containing 6% Nonidet P-40 for 5 min at room temperature before four volumes of Buffer A were added. The nuclei were collected as described previously (14) and suspended in an equal volume of 20 mM Hepes-HCl, pH 7.9, 0.42 M NaCl, 1 mM EDTA, 1 mM EGTA, 2 mM (2S,3S)-trans-epoxysuccinyl-L-leucylamido-3-methylbutane, and 2 mM 4-(2-aminoethyl) benzenesulfonylfluoride HCl. The nuclei were sonicated for 10 s and then centrifuged (10,000 ϫ g, 15 min). The supernatant was passed three times through a 30 1 ⁄2-gauge needle to sheer DNA into ϳ100 -800-bp fragments. Two volumes of Buffer A were added before the sample was split into two equal volumes for the anti-URE3-BP and control-mAb protein G Dynabead immunoprecipitation. After immunoprecipitation and DNA purification, the samples were resuspended in 100 l of 10 mM Tris-HCl, pH 8.5, and 5 l was analyzed by real time PCR.
Real Time Quantitative PCR-Real time quantitative PCR analysis of the hgl5 and fdx promoters and srehp coding region was preformed in a Bio-Rad iCycler. The fluorescent dye SYBR Green I (Molecular Probes) was used to detect double strand DNA. Continuous SYBR Green I monitoring during amplification was done according to the manufacturer's recommendations. The 5 l of the immunoprecipitated DNA was subjected to 40 amplification cycles with Qiagen's HotStar Taq. To amplify Ϫ270 to Ϫ3 bp of the hgl5 promoter (4) the primers CTACTGAAGCTTAGTAAAGAATAGTATTGA and CTACTGGGATCCT-TGAATTTCTAGTTCATTGTCT were used. To amplify the region Ϫ513 to ϩ14 of the fdx promoter (29) oligonucleotides CTACTGAAGCTTTAAAA-ATACAAACAACTACC and CTACTGTCTAGACATTAGATTTGAATGA-ATAA were used. To amplify the coding sequences ϩ120 to ϩ 663 of the srehp gene (30) the oligonucleotides GTCCTGAAAAGCTTGAAGAAGC and GGACTTGATGCAGCATCAAGGT were used. All real time amplification reactions were performed in duplicate on each ChIP experiment, and the resulting fluorescent values were averaged.
Reverse transcription followed by real time PCR was implemented to quantitate the calcium-dependant expression of fdx and URE3-BP genes and normalized to the level of the control transcript, L10. The reverse transcription was primed using random sequence hexamers and performed with the Superscript II enzyme (Invitrogen). Real time amplification was preformed as described above. To amplify the region between ϩ3 to ϩ151 of the URE3-BP cDNA the oligonucleotides AAAA-GATCTATGGGATAGTGTTAAGTAATGGGAACC and AAAgcggaagc- Trophozoites were treated with formaldehyde to cross-link DNA protein complexes. These complexes were purified by immunoprecipitation with either an anti-URE3-BP mAb (OE) or a control antibody (‚) of the same isotype. The hgl5 promoter concentration was determined by real time PCR. A, calibration of hgl5 primer efficiency. The y axis represents quantity of input DNA, and the x axis represents the cycle at which the fluorescent value exceeded the threshold value (threshold value was set at 5-fold standard deviation of base line). The equation derived from the data is displayed on the graph. B, representative ChIP/quantitative real time PCR amplification plots of the hgl5 promoter. The y axis (Sybergreen fluorescence at 530 nm) represents the change in the emission intensity of the SYBR Green I reporter dye after subtraction of the base line (i.e. early cycles of PCR prior to detectable levels of template). The x axis represents the PCR cycle number. C, amount of the hgl5 promoter immunoprecipitated utilizing the anti-URE3-BP mAb (URE) or the control mAb (ctr) (mean Ϯ S.E., n ϭ 12, p ϭ 0.0014). gatatccaccaatgcaaccacc were used. The ferredoxin 1 cDNA was amplified using the oligonucleotides TGATGACTGTGTCGCTTGC and ACTTATTCAACTTTAAGAACTCC, which amplify the DNA between the nucleotides ϩ45 to ϩ200 of the coding sequences (31). The region of the L10 cDNA ϩ6 to ϩ72 amino acids was amplified via primers CTACTGGGATCCAGGAAGATGTTATAGACTTG and CTACTGGAAT-TCTTAATTGAATACGTGCTGC. In all experiments utilizing real time PCR the cycle threshold values (C T ; the cycle number at which fluorescence exceeds the threshold value) were linked to the quantity of initial DNA after calibration of the amplification efficiency of the primer pair utilized (32).  (15). In an EF-hand loop the calcium ion is bound by the six residues in positions 1, 3, 5, 7, 9, and 12 or X, Y, Z, -Y, -X, and -Z (15). The amino acid sequence of the second EF-hand loop in URE3-BP was DRNRSGTLEPHEI (conserved residues in bold). We replaced the residues at positions 1 and 3 in the second EF-hand to alter the sequence to ARARSGTLEPHEI (changed residues underlined).

Mutation of the EF-hand Motif in URE3-BP Abrogates Calcium Inhibition of in
The resulting mutant protein was tested in an EMSA assay (Fig. 1). The mutated URE3-BP protein was able to form a protein-DNA complex with a radiolabeled double stranded hgl5-URE3 oligonucleotide and was competed by the addition of unlabeled double stranded oligonucleotide hgl5-URE3 (Fig.  1B). The DNA-protein complex was also inhibited to a lesser extent by the addition of hgl5-MUT, a double stranded DNA oligonucleotide, in which the URE3 DNA sequence was mutated to TTAGAATTC from the wild-type sequence of TTATCT-TAT. The partial inhibition of the protein-DNA complex with hgl5-MUT suggested some loss of specificity for DNA binding of the mutated URE3-BP. Strikingly the addition of 86 M free calcium had no effect on the ability of the mutated URE3-BP to form a DNA-protein complex (Fig. 1B). These experiments demonstrated that EF(2)mutURE3-BP was no longer sensitive to Ca 2ϩ inhibition.
Detection of in Vivo Promoter Occupation of URE3-BP-ChIP assays allow study of the interaction between nuclear proteins and DNA sequences in the context of the chromatin template (33). This technique was used to test whether nuclear URE3-BP was physically associated with the hgl5 gene promoter in the chromosome. Trophozoites were treated with formaldehyde to cross-link nuclear proteins with chromosomal DNA. DNA-protein complexes were then purified by immunoprecipitation. The cross-links were reversed, and associated DNA was purified. Detection and measurement of the precipitated DNA was achieved by use of real time PCR (9). Figs. 2 and 3 show representative ChIP experiments in which DNA purified by immunoprecipitation of cross-linked nuclear DNAprotein complexes with anti-URE3-BP mAb was amplified by real time PCR for the hgl5 and fdx promoters. The graphs in Fig. 2A and Fig. 3A show the calibration of the C T value of the primer pairs used to amplify the hgl5 and fdx E. histolytica promoters against DNA concentration. Each experiment was controlled with a ChIP using an irrelevant mAb. An additional control was PCR amplification of an irrelevant gene segment (the srehp coding sequence). There was no significant difference observed between the C T for the srehp DNA amplified from the control or anti-URE3-BP mAb ChIPs (data not shown). Anti-URE3-BP mAb specifically immunoprecipitated DNA containing the hgl5 and fdx E. histolytica promoters (see Figs. 2 and 3). This was consistent with URE3-BP being physically associated with the DNA-chromatin complex at those promoters (see Figs. 2 and 3).
Manipulation of Intracellular Calcium in E. histolytica Trophozoites-Because of the importance of calcium in the sequence-specific recognition of URE3 by URE3-BP, we wished to test the effect of intracellular calcium on promoter binding by URE3-BP. We estimated the level of intracellular calcium using the fluorescent dye indo-1. Trophozoites were suspended in Buffer B (25), which contained 5 mM MgCl 2 and EDTA to sequestrate calcium. The baseline internal Ca 2ϩ was measured by the fluorescence ratio (398/480 nm) after excitation at 360 nm. The effects of altering the external [Ca 2ϩ ] o concentration on internal Ca 2ϩ was followed after the addition of 1.  (27). A rapid and consistent increase in internal Ca 2ϩ was observed (as measured by the ratio of indo-1 fluorescence emission at 398/480 nm) when the external Ca 2ϩ concentration was increased to 608 M or 1.1 mM (Fig. 4). The change in indo-1 fluorescence induced by increases in extracellular calcium was not because of leakage of the dye from the trophozoites, as indo-1 fluorescence was not detected in the cell-free supernatant (data not shown). This indicates that in these conditions, changes in extracellular calcium resulted in alterations of intracellular calcium in the amebic trophozoites.

Changes in URE3-BP Occupancy of the hgl5 and fdx Promoters in Vivo upon Alteration of Calcium Levels-The in vitro
EMSA experiments (Fig. 1A) demonstrated that the sequencespecific interaction of URE3-BP with URE3 was blocked by calcium. To test the importance of calcium for URE3-BP binding to the promoter in vivo we performed ChIP after modulating intracellular calcium. Trophozoites were suspended in Buffer B ϩ 5.7 mM CaCl 2 (1.1 mM [Ca 2ϩ ] o ) before formaldehyde treatment, ChIP, and real time PCR analyses. As demonstrated in Figs. 5 and 6 the increase in intracellular calcium (caused by the addition of 5.7 mM extracellular calcium to the trophozoites) resulted in inhibition of URE3-BP binding to both the hgl5 and fdx promoters as measured by ChIP and real time PCR. The failure to immunoprecipitate the hgl5 and fdx promoters with the anti-URE3-BP mAb was not because of calcium-mediated interference in immunoprecipitation of the URE3-BP, as similar amounts of the protein were immunoprecipitated under both conditions (Fig. 6A). We therefore concluded that in the presence of elevated intracellular calcium URE3-BP was unable to occupy promoters that contained the URE3 DNA motif.
URE3-BP Genomic Sequence Context-The URE3-BP cDNA sequence was used to identify the equivalent genomic sequence in the International Entamoeba Genome Sequencing Project data bases (Fig. 7). A URE3 DNA consensus sequence was present in the 3Ј sequence 1.3 kb distal to the URE3-BP gene stop codon. We have previously found functional regulatory sequences for other Entamoeba genes more than 1 kb from the open reading frame. 3 There were no differences in genomic and cDNA sequences, which was not surprising, because most (about 80 -90%) E. histolytica genes lack introns. A sequence (ATTCG) that strongly resembled the INR consensus ATTCA was located immediately before the cDNA start site (7). The sequence GAACT, which is identical to the GAAC consensus (GAACT), was located Ϫ40 bp from the ATG start sequence (Ϫ28 bp from the previously published cDNA sequence). The sequence (TGATATAAAG) with a very low similarity to the TATA consensus (GTATTTAAA(GyC) was located Ϫ46 bp from the ATG (6). A pentanucleotide sequence (TAATT) that was identical to the 3Ј-terminal consensus sequence TA(A/T)TT was located 42 bp after the translation stop codon (7). The 3Ј noncoding sequence contained an open reading frame in the opposite orientation that terminated at ϩ73 bp after the URE3-BP stop codon and initiated at ϩ 720 bp on the opposite strand and thus intervened between the termination of the URE3-BP open reading frame and the 3Ј URE3 sequence motif. The intervening open reading frame could encode a highly basic 25-kDa protein but did not have an easily recognizable core promoter.
Calcium Effect on fdx and URE3-BP mRNA-To investigate the role of the URE3 in the transcriptosome's response to calcium in cell culture the levels of fdx, URE3-BP and L10 ribosomal protein mRNA were followed by quantitative reverse transcriptase PCR (Fig. 8). The promoters of fdx and 3Ј sequences of URE3-BP genes contain the URE3 motif whereas the putative promoter sequences of L10 do not. The L10 mRNA was utilized as a control transcript. The ratio of fdx and the L10 mRNA transcripts isolated from trophozoites in 1 mM Ca 2ϩ or Ca 2ϩ -depleted medium showed no consistent statistically significant variation. The URE3-BP transcript, however, markedly decreased in 1 mM Ca 2ϩ . The URE3-BP mRNA calcium: calcium-depleted ratio was statistically significantly different (p Յ 0.02) from that of L10 mRNA at 4, 6, and 8 h post-calcium modulation (Fig. 8). We concluded that the presence of a URE3 motif in the URE3-BP non-coding region was consistent with the gene being under autoregulation. DISCUSSION This work demonstrates that URE3-BP is a calcium-binding protein that can function as both a transcription factor and FIG. 7. Sequence context of the URE3-BP gene. Sequences were obtained from the Entamoeba histolytica Genome Sequencing Project at the Institute for Genomic Research web site at www.tigr.org and the Wellcome Trust Sanger Institute at www.sanger.ac.uk. The 5Ј sequence is Sanger Data base Ent774f10.q1c. Sequences 3Ј included sequences derived from both the Institute for Genomic Research (ENTLU68TR, ENTLU68TF) and the Sanger Institute (Ent1246d08.p1k, Ent1123h08.q1c, Ent683f12.p1c). Numbering is as the previously published cDNA sequence of URE3-BP, accession number AF291721. The proposed TATA, GAAC, INR, terminal pentanucleotide sequences, and 3Ј URE3 motifs are outlined and underlined.
calcium sensor in E. histolytica. Only one other calcium binding eukaryotic transcription factor, DREAM, has been described. DREAM (also known as Calsenilin or KChIP3) also contains EF-hand motifs and is similar to URE3-BP in that it exhibits sequence-specific binding to DNA. An increase in calcium levels in vitro blocks DNA binding by DREAM and URE3-BP. The mutation of EF(2) of URE3-BP had little effect on DNA binding but impeded the ability of Ca 2ϩ to block in vitro DNA binding to URE3. This showed that the second EF-hand motif of URE3-BP was essential for the calcium modulation of URE3-BP DNA binding.
Promoter occupancy of URE3-BP in vivo on the native chromatin of URE3-BP of the fdx and hgl5 promoters was tested in this work. To determine whether URE3-BP interacted with URE3-containing promoters in the nucleus we utilized the ChIP assay in conjunction with real time PCR. This allowed measurement of URE3-BP bound to URE3-containing promoters in intact trophozoites. Our results indicated that URE3-BP was located on both the fdx and hgl5 promoters in the nuclear environment.
Internal changes in [Ca 2ϩ ] i have been observed in response to various stimuli (25,34,35), and several important calciumbinding proteins have been discovered in E. histolytica including a protein similar to calmodulin, EhCaBP (36,37). In addition, the multidrug resistance phenotype mediated by the up-regulation of the EhPgp1 and EhPgp5 genes of E. histolytica has been shown to be reversed by calcium-channel blocker verapamil (38 -41). To investigate the role of URE3-BP as a calcium sensor protein we investigated the impact of altering [Ca 2ϩ ] i on URE3-BP promoter occupancy. A significant decrease in URE3-BP location at both the hgl5 and fdx promoters was observed upon an increase in [Ca 2ϩ ] i . Therefore, calcium not only prevented in vitro binding of URE3-BP to URE3 but also blocked the occupancy of URE3-containing promoters by URE3-BP in vivo.
Because calcium levels influenced the binding of URE3-BP to the URE3-containing promoters the effect of calcium on the steady state levels of fdx and URE3-BP mRNA was examined. The hgl5 mRNA was not analyzed, because it was not possible to design PCR primers that would amplify hgl5 mRNA without cross-hybridizing to other hgl genes that lack URE3-containing promoters (2). In the case of the fdx gene, its transcript was not significantly changed by increases in intracellular calcium. This perhaps was not surprising as changes in intracellular calcium would be predicted to have pluripotent effects on the trophozoite that could have disparate effects on fdx mRNA abundance.
In contrast to the situation for fdx mRNA, the addition of calcium sharply decreased the level of the URE3-BP mRNA. This decrease could be mediated by an autofeedback mechanism where URE3-BP is displaced by calcium from the URE3 motif located at the 3Ј sequence of URE3-BP. Mutational analysis of the URE3 motif in the 3Ј noncoding region of URE-3BP will be required to test directly whether the gene is directly autoregulated, although one might expect, as is the case for the fdx gene, that URE-3BP will fall under complex regulatory responses to changes in intracellular calcium. In either case these results indicate that increases in intracellular calcium not only decrease the ability of URE-3BP to bind DNA but also decrease its mRNA abundance. This suggests that URE3-BP is both an effector and a responder in a cascade of calciumregulated gene expression in the parasite.
One of the mysteries of amebiasis is why only a minority of infected individuals develop invasive disease. E. histolytica encounters differing levels of extracellular calcium in the human intestine where it lives. We have shown here that intracellular calcium levels in the parasite are exquisitely sensitive to changes in extracellular calcium. The extracellular calcium concentration in the intestinal lumen is influenced by the amount of calcium ingested, absorbed, and excreted. In the small intestine where the parasite excysts, luminal calcium concentrations are well above 1 mM (42). E. histolytica trophozoites colonizing the large intestine are exposed to extracellular calcium levels that vary depending upon the diet (13,42). In hosts with a high calcium diet trophozoites would be exposed to Ͼ1.25 mM calcium. We would predict from our data that these calcium levels would be sufficient to block URE3-BP binding and decrease URE3-BP synthesis. In a low calcium diet, luminal calcium concentration may drop to ϳ1.25 mM. In this circumstance calcium absorption in the large intestine is considerably enhanced, and free Ca 2ϩ will be further decreased (42). URE3-BP-mediated gene regulation would then play a significant role in the regulation of transcription. It is possible that calcium and calcium sensors including URE3-BP influence the expression of the virulence phenotype of E. histolytica in the large intestine and impact the development of disease.