Structural and functional analysis of the Entamoeba histolytica EhrabB gene promoter

Background The Entamoeba histolytica EhrabB gene encodes for a Rab GTPase involved in phagocytosis. It is located at a virulence locus where the Ehcp112 gene is in the complementary strand at 332 bp of EhrabB start codon, suggesting a finely regulated transcription of both genes. However, the transcription regulation in this parasite is poorly understood. Results To initiate the knowledge of EhrabB gene expression regulation, here we studied the structural characteristics of its gene promoter and its control transcription elements. In silico searches of the EhrabB 5'-flanking region revealed that it contains a motif similar to the upstream regulatory element 1 (URE1) of the E. histolytica hgl5 gene. It also has sequences with homology to C/EBP and GATA1 binding sites, and heat shock elements (HSE). Primer extension experiments revealed that EhrabB has at least four transcription initiation sites. The elements at the 5'-flanking region that drive EhrabB gene expression were detected and characterized using transitory transfected trophozoites with a plasmid carrying the CAT reporter gene. EhrabB transcription is negatively regulated by a sequence located between positions -491 to -428 with respect to the first transcription initiation site. We also showed that the URE1-like motif activates EhrabB transcription. In addition, heat shock activated the EhrabB promoter in episomal constructs and lead to an increase in de novo EhrabB transcription. Conclusion The data suggest that EhrabB transcription is controlled negatively by an unidentified sequence, but it is activated by an URE1-like motif. Our analyses also revealed the presence of activator HSE that function under stress.


Background
Entamoeba histolytica is the protozoon responsible for up to 100,000 deaths each year [1]. Only 10% of infected people present disease symptoms expressed in intestinal or extraintestinal amoebiasis. The molecular mechanisms participating in the parasite invasiveness are not completely understood. Diverse populations of E. histolytica, including clones derived from a particular strain, display different virulence phenotype [2][3][4]. In addition, longtime cultured trophozoites, which show poor capacity to produce liver abscesses in experimental animals, recover their virulence after incubation with cholesterol or with certain types of bacteria, or after their passage by hamster livers [5][6][7][8]. This behavior could be due to changes in expression of certain genes. In fact, trophozoites isolated from infected animals present changes in gene expression with respect to trophozoites cultured in the laboratory [9,10]. It has been also shown that trophozoite interaction with human collagen type I and Ca 2+ triggers transcription of virulence related molecules as the amoebapore C, the cysteine proteinase 5 and a cdc48-like protein [11,12].
In contrast to other eukaryotes, transcription of E. histolytica protein-coding genes is insensitive to high concentrations of alpha-amanitin [13]. In addition, certain promoters of protein-coding genes present three unusual motifs in their 5'-flanking region [14]: (a) the GTATT-TAAA(G/C) sequence, which corresponds to the TATA box; (b) the AAAGAACT sequence, named GAAC element; and (c) the initiator AAAATTCA sequence (Inr). Furthermore, five upstream regulatory elements (URE1 to URE5) were identified in the promoter of the galactose-inhibitable lectin heavy subunit gene hgl5 [14]. Additionally, the structural and functional characterization of promoters driving the expression of EhPgp1 and EhPgp5 genes, whose products are implicated in the multidrug-resistance event, revealed that they contain motifs with homology to cisregulatory elements of eukaryotic cells [15][16][17][18][19].
Vesicular trafficking plays an important role in the pathogenic mechanism of E. histolytica, because it participates in secretion of virulence factors and in the internalization of host cells. In eukaryotes, vesicular trafficking is regulated by Rab GTPases, which by cycling between active (GTPbound) and inactive (GDP-bound) conformations act as molecular switches that integrate several events at each step of vesicular transport [20]. Thus, Rab proteins may be involved in the E. histolytica pathogenicity. EhRabB is a Rab GTPase located in small vesicles that in wild-type trophozoites are translocated to plasma membrane and to phagocytic mouths during phagocytosis [21], whereas, in trophozoites deficient in phagocytosis most of these vesicles remain in the cytoplasm [22]. In addition, Marion et al. [23], using a proteomic approach, described that EhRabB is among the proteins identified during early phagocytosis. Although EhRabB was not identified in independent phagosome protomics studies using carboxylated latex beads [24,25], those results suggest that EhRabB participates in phagocytosis, an event related to the E. histolytica pathogenicity [26].
The EhrabB gene is located close to the Ehcp112 and Ehadh112 genes, whose products form the EhCPADH complex, also involved in the parasite pathogenic mechanism [27]. These three genes spans a 4500 bp region named virulence locus (VI) [28], providing a good model to study gene transcription regulation of virulence-related genes. The EhrabB gene is situated 332 bp upstream of the Ehcp112 gene, but in the complementary strand [21]. Some of the cis-elements that control the Ehadh112 and Ehcp112 transcription are located inside the open reading frame of Ehcp112 and EhrabB, respectively [28,29]. However, the cis-acting sequences which might be relevant to drive the EhrabB gene expression have not been identified yet.
To elucidate the molecular mechanisms controlling EhrabB gene expression, we studied here the structural and functional characteristics of its promoter. Our analyses revealed that it has several transcription initiation sites. We also detected the presence of one region that represses and one motif that activates EhrabB transcription. Interestingly, the gene promoter has regulatory heat shock elements that under heat shock stress activate EhrabB transcription.

Structural analysis of the 5'-flanking region of EhrabB
The proximity of EhrabB and Ehcp112 genes and their opposite transcription, suggest a finely regulated gene expression of these two genes. To initiate the comprehension of the mechanisms involved in their transcription, we studied here the EhrabB promoter region, comprising from the start codon to -755 bp upstream of this site. In this region we looked for: (a) consensus sequences described in E. histolytica core promoters [14]; (b) sequences with similarity to the URE1 to URE5 elements [14]; and (c) consensus sequences described for binding of eukaryotic transcription factors.
Neither E. histolytica Inr elements nor TATA-box sequences were identified in the EhrabB promoter region. In contrast, we detected a sequence similar to the GAAC element at 22 to 29 nucleotides upstream of the start codon (Fig. 1). The absence of Inr and TATA box consensus sequences suggests that there is a particular unknown transcription initiation mechanism for EhrabB. Our search also revealed a sequence located at -423 to -413 that is 81.8% identical to the URE1 sequence described as a transcriptional activator of the hgl5 gene [14] (Fig. 1). The analysis also showed the presence of three C/EBP, one GATA-1 and seven heat shock elements (HSE) consensus sequences in this region (Fig. 1).

Determination of the EhrabB transcription start site
The EhrabB gene transcription initiation site was mapped by primer extension experiments using a [ 32 P]-labeled 18bp antisense oligonucleotide that spans from 85 to 68 nucleotides inside the gene open reading frame. As often found in TATA-less promoters [30], EhrabB gene exhibited several transcription initiation sites located at -72, -266, -300 and -360, with respect to its start codon ( Fig. 2A). The adenine residue located at the start of the shortest EhrabB transcript was arbitrarily designed as position +1 (Fig. 1). The AAAAGAAGA +1 sequence, where this transcript starts, differs from the consensus sequence described for the transcription initiation site (AAAATTCA +1 ) of the majority of E. histolytica genes [14]. On the other hand, the GAAC element, a C/EBP site and two HSE sequences found by in silico searches appeared situated downstream of the first transcription initiation site (Fig. 1), but we do not know yet their relevance in transcriptional activity.
Most of the E. histolytica genes examined so far have a short (0-21 bp) 5' untranslated region (5'UTR) [31]. EhrabB as other few E. histolytica transcripts [27,32,33] have long 5'UTR, which could have a specific role in translation regulation [34]. The shortest 5'UTR of EhrabB is composed of 72 nucleotides, whereas the longest has 360 nucleotides and the Ehcp112 transcript, which runs in opposite direction, has a 280 nucleotides 5'UTR [27]. On the other hand, the EhrabB start codon is located at 332 bp upstream of the Ehcp112 start codon, thus, both transcripts overlap in at least 20 nucleotides (Fig. 2B). This act supports the hypothesis that expression of EhrabB and Ehcp112 genes are co-regulated.

Delimitation and characterization of the EhrabB gene promoter and its relevant motifs
The active regions of EhrabB gene promoter were characterized in trophozoites transfected with the pBSCAT-ACT vector [15] in which we cloned DNA fragments containing 96 bp downstream of the first transcription initiation site and distinct lengths fragments of the upstream region. The larger fragment (-683 to +96; pRab683), containing the GATA-1 (-195 to -184), C/EBP (-260 to -243 and -220 to -208), URE1 (-423 to -413), and HSE (-624 to -608, -577 to -543, -488 to -458, -403 to -388, and -15 to -2) sequences, showed only 15 ± 1% CAT activity of the one exhibited by actin gene promoter [35], used here as a control (Fig. 3B). Interestingly, in pRab428 plasmid (-428 to +96), where three out of five HSE motifs present were Primary structure of the EhrabB promoter gene Figure 1 Primary structure of the EhrabB promoter gene. Nucleotide sequence of the 755 bp upstream from the ATG (lower case). The transcription initiation sites are marked by arrows. URE1-like sequence is boxed. Putative HSE sequences are shown by dots. Putative C/EBP sequences are underlined. Putative GATA-1 sequence is double underlined. Putative GAAC sequence is in bold. Numbering of nucleotides is from +1 at the first transcription start site.
Due to the distance between EhrabB and Ehcp112 genes and to their transcription in opposite directions ( Fig. 2B) is probable that our constructs could display transcription from the opposite strand to the reporter gene, producing an antisense RNA that possibly affect the promoter assays, and differences in CAT activity between pRab428 and pRab683 could be due to variations in the antisense level. To investigate if antisense RNA of the cat reporter gene is transcribed in the transfected trophozoites, we carried out strand-specific RT-PCR assays. In both constructs (pRab683 and pRab428), RT-PCR assays produced the expected fragment when we used oligo(dT) or antisense primers to synthesize cDNA, whereas no amplification was obtained when a sense primer was utilized (Fig. 3C), indicating that antisense RNA is not produced in the transfectants. These results revealed that transcriptional differences observed in transfected trophozoites are due to promoter activities, and they are according with those obtained on the transcriptional analysis of the Ehcp112 gene [28], which demonstrated that the 332 bp intergenic region between Ehcp112 and EhrabB is not able to promote the Ehcp112 transcription, and that sequences driving the Ehcp112 transcription are located inside the coding region of EhrabB (between 681 and 833 bp upstream of the Ehcp112 start codon). In our constructs this region was replaced by the cat reporter gene, consequently the sequences driving the transcription of antisense RNA are absent.

Delimitation of the repressor region in EhrabB gene promoter
According with our analyses, in the DNA fragment that inhibits EhrabB transcription (-428 to -683) we did not identify consensus transcription repressor sequences described in other systems. Thus, to delimitate the cis-elements that negatively control EhrabB transcription, we Transcriptional start sites of the EhrabB gene performed sequential deletions between position -683 and -428. Different length fragments were again cloned into the pBSCAT-ACT vector and transfected into trophozoites, then, CAT activities were measured (Fig. 4A,B). In construction comprising from -491 to +96 nt (pRab491) CAT activity was 50 ± 3% of pRab428 (Fig. 4B) and the pRab555 (-555 to +96) showed a dramatic drop of activity (25 ± 11% of pRab428) (Fig. 4B). These experiments indicated that the 67 bp region located between positions -491 to -428 contains transcription inhibitors. In this region we only detected a HSE element (-488 to -458).
EMSA experiments using a 63 bp DNA probe containing nucleotides form -491 to -428 confirmed the presence of nuclear proteins interacting with this region. Results showed the formation of three DNA-protein complexes (Fig. 4C) that were competed by 150-fold excess of the same cold-probe, but remained in the presence of 350fold excess of the nonspecific competitor (Fig. 4C), indicating that DNA-protein interactions were specific. The complexes formation strengthened the hypothesis that this region has a functional repressor activity in EhrabB gene promoter.

Delimitation of the activator region in EhrabB gene promoter
To delimitate the cis-elements that activate the EhrabB transcription, we generated constructions with sequential deletions between position -428 and -257 that were cloned into the pBSCAT-ACT vector (Fig. 5A,B). Activity exhibited by extracts from trophozoites transfected with plasmids pRab369 (-369 to +96), pRab321 (-321 to +96), and pRab273 (-273 to +96) decreased to 33 to 44% of activity displayed by the pRab428 plasmid (Fig. 5B). Constructions with additional different deletions did show similar CAT activity than pRab369. These experiments indicated that cis-elements that activate EhrabB gene transcription are located between positions -428 to -369 nt. In this region we identified a sequence (-423 to -413) with 81.8% identity to URE1 sequence described for the hgl5 gene promoter [14] and one putative HSE motif (-403 to -388) (Fig. 5A).
Promoter activity of different fragments from the 5'-flanking region of the EhrabB gene EMSA experiments using nuclear extracts and the 59 bp DNA fragment comprising from -428 to -369 revealed that two DNA-protein complexes are formed in this region. Complexes were competed by 150-fold excess of the same cold-sequence, but not by the presence of 350fold excess of the nonspecific competitor (Fig. 5C). Complexes formation supported the hypothesis that in this DNA fragment are situated cis-elements involved in EhrabB transcription activation.

Role of the URE1-like sequence in the EhrabB transcription
The 59 bp fragment between -428 and -369 activates transcription, forms complexes with some nuclear proteins and contains a sequence that has nine out eleven base pairs identical to the URE1 motif of hgl5 (81.8%). To analyze the role of the URE1-like sequence in EhrabB promoter activity, we deleted 13 bp containing this sequence and cloned a DNA fragment from -415 to +96 in front of the CAT reporter gene (pRab415) (Fig. 6A,B). This construct only exhibited 33 ± 4% CAT activity of pRab428 that includes the URE1-like sequence (Fig. 6B). Our results show that the URE1-like sequence located from -423 to -413 bp is a cis-activating element of EhrabB transcription.
We also carried out EMSA experiments using the DNA fragment containing from -428 to -412 bp as a probe. Two DNA-protein complexes were detected in these assays (Fig. 6C). Both complexes were competed by a 150-fold excess of the same cold sequence, but they were not competed by a 150-fold excess of a nonspecific competitor with similar base composition to URE1, but mingled sequence (Fig. 6C). Interestingly, when we used as competitor a fragment containing the URE1 sequence from the hgl5 promoter, complexes were competed and they disappeared (Fig. 6C, lane 4). Moreover, two DNA-protein complexes with similar mobility were detected when this sequence was used as a probe (Fig. 6C, lane 6). These results evidenced that similar nuclear factor(s) recognized the URE1 sequence of the hgl5 gene promoter and the URE1-like sequence present in EhrabB gene promoter.

Effect of heat shock on EhrabB gene transcription
Comparative analyses of the 5'-flanking region of genes induced by heat shock have revealed that they contain tracts with three or more nGAAn or nTTCn sequences, which are essential for heat induction response [36]. These tracts are called HSE and the presence of multiple HSE can act in a cooperative way. As mentioned before, in the EhrabB 5'-flanking region we found seven putative HSEs, five of them situated upstream of the first start transcription site (-624 to -608, -577 to -543, -488 to -458, -403 to -388, and -15 to -2) and the two other at +20 to +39 and +44 to +60 (Fig. 1). To prove that they have a functional role in the EhrabB transcription, we used the construction pRab683 (-683 to +96) that includes all these sequences (Fig. 7A) to transfect E. histolytica. Its ability to drive CAT expression was measured 2 and 4 h after trophozoites incubation at 42°C. CAT enzymatic activity augmented approximately twice on heat shocked trophozoites with respect to the activity displayed by cells maintained at 37°C (Fig. 7B). Our results indicate that HSE motifs present into the EhrabB gene promoter could be functional under heat shock stress.
On the other hand, to verify the effect of heat shock in EhrabB gene transcription, we performed real-time RT-PCR assays using RNA isolated from trophozoites maintained at 37°C or from trophozoites incubated 4 h at 42°C. In concordance with CAT assays, trophozoites under heat shock presented a five-fold increase of EhrabB transcription in comparison with those grown at 37°C (Fig. 7C). These results support the hypothesis that under heat shock of trophozoites the HSE sequences found in the EhrabB gene promoter regulate the transcription of this gene. It is possible that these elements also regulate the gene transcription under other stimuli such as oxidative stress, during invasion, or stress induced by anti-E. histolytica drugs. However, this assumption needs to be proved.
Gene expression analysis by microarray assays revealed that heat shock treatment of E. histolytica trophozoites induce the expression of virulence-related genes such as the cysteine proteases 4 and 6, and specific alleles of the Gal/GalNac lectin [37]. In addition, some Rab GTPases are upregulated under stress conditions in other organ- isms [38][39][40], indicating that these proteins may play an important role in cell adaptation to stress, beyond their housekeeping function in intracellular vesicle trafficking. In Saccharomyces cervisiae the HSE sequences are also involved in the transcription induction not only of heat shock proteins (HSP) but also of genes encoding proteins involved in diverse cellular processes [41].

Conclusion
We showed here that the EhrabB gene is transcriptionally activated by heat shock and that the EhrabB gene promoter contains cis-elements that repress and activate the gene transcription on normal conditions. The cis-activating element corresponds to a URE1-like sequence, whereas de cis-repressive element remains to be identified.

Sequence analysis of the 5'-flanking region of EhrabB
The 5'-flanking region of the EhrabB gene, which contains its putative promoter, was previously isolated from a genomic clone that includes the Ehadh112, Ehcp112 and EhrabB genes [GeneBank: AF172320]. The EhrabB upstream sequence is identical to the sequence found in the E. histolytica genome project [43]. Prediction of putative cis-acting elements in EhRabB gene promoter was done using the TRANSFAC database [44]. For searching cis-acting elements found in other E. histolytica genes (TATA box, GAAC element, Inr and URE sequences) we used the LALIGN program [45].

Primer extension, RT-PCR and real-time RT-PCR assays
Total E. histolytica RNA was obtained using the TRIZOL reagent (GIBCO BRL) according to the manufacturer recommendations. For primer extension assays, 40 µg of RNA were hybridized to an 18 bp [ 32 P]-labeled primer (5'-GGTCACAATACTGACGAA-3') complementary to nucleotides 85-68 downstream to the start codon of the EhrabB gene. Annealing was performed as previously described [15]. Nucleic acids were phenol-chloroform extracted, ethanol precipitated, and separated by electrophoresis on 6% urea-polyacrylamide gels. A sequencing ladder was generated using the above oligonucleotide and the plasmid containing the 5'-flanking region of the EhrabB gene as template. Gels were visualized with a phosphoimager analyzer (BIO-RAD).

Role of URE1 in the transcriptional activity of EhrabB
For strand-specific RT-PCR of the cat reporter gene we used specific sense (5'-ATGGAGAAAAAAATCACTGGA-TATA-3') or antisense (5'-ATAGGCCAGGTTTTCACCG-TAACAC-3') oligonucleotides for the cDNA synthesis, and then PCR assays were performed using the same primers. Amplification conditions were as follows: 94°C for 5 min; then 30 cycles at 94°C for 40 s, 57°C for 35 s and 72°C for 40 s; and a final extension at 72°C for 7 min.

Plasmids construction
Different regions of the EhrabB promoter were inserted into the pBSCAT-ACT plasmid in front of the CAT reporter gene followed by the 3'-flanking sequence of the E. histolytica actin gene [15]. All fragments contained 24 bp of the 5'-end coding region of the EhrabB gene and were generated by PCR using Pfx DNA polymerase (Invitrogen), specific primers (Table 1), and the genomic clone containing the EhrabB upstream flanking region as template. The orientation and sequence of each construct were confirmed by DNA sequencing. As a positive control we used the pA5'A3'CAT vector containing the E. histolytica actin gene promoter [35] and as negative control the reporter vector without promoter region (pBSCAT-ACT).

Transient transfection and CAT assays
Transfection was carried out by electroporation as described [35]. CAT activity was determined by two-phase diffusion assays [13] using 100 µg of trophozoite extracts, 200 µl of 1.25 mM chloramphenicol and [ 14 C]-butyryl CoA (NEN Life Science Products). The background activity displayed by trophozoites transfected with the promoter-less vector was subtracted from activity obtained from cells transfected with each construct. Activities were expressed as relative activity with respect to that obtained from trophozoites transfected with the pA5'A3'CAT plasmid. CAT activity for each construct was assayed at least three times by duplicate. Data are expressed as mean of three independent experiments ± S.D. for each group. Statistical significance was determined by Student's t-test and a difference of p < 0.05 was considered significant.
To determine the effect of heat shock on promoter activity, trophozoites transfected with the construct containing 683 bp of the EhrabB gene promoter were incubated at 37°C for 48 h, and then at 42°C during 2 and 4 h before performing the CAT assays.
Effect of heat shock on EhrabB transcription Figure 7 Effect of heat shock on EhrabB transcription. (A) Schematic representation of the EhrabB promoter region and its putative cis-acting consensus sequences (boxes). Arrows indicate the transcription initiation sites. (B) The 683 bp DNA fragment from the 5'-flanking region of EhrabB was cloned upstream of the cat reporter gene into the pBSCAT-ACT vector. E. histolytica trophozoites were transfected with this construction and incubated at 37°C for 48 h. Then, trophozoites were incubated for 2 and 4 h more at 37°C or at 42°C and CAT activities were determined. Activities are relative to that obtained with the actin promoter gene at 37°C, used as positive control. Each bar corresponds to the average of CAT activities ± S.D. representative of three independent experiments performed by duplicate. (C) Real-time RT-PCR assays using RNA from trophozoites grown at 37°C (A) or after 4 h of heat shock treatment (A-HS) and specific primers for EhrabB. Amplification of the E. histolytica18S rRNA was used as normalizer.