It has been shown in a wide variety of cells that X-rays exposure, UV irradiation and chemicals activate cellular responses to DNA repair 20. To initiate the study of the mechanisms involved in DNA repair in E. histolytica, we used UV-C light irradiation to induce DNA damage in trophozoites. Our experiments showed that during the first 12 h after irradiation with 254 nm UV-C (150 J/m²), cell survival was not significantly affected (Fig. 1A). Using the same experimental conditions, we analyzed the presence of 3'-hydroxyl DNA ends by TUNEL and FACS assays. In untreated trophozoites, FACS analysis evidenced the presence of <1% TUNEL positive cells; meanwhile, 30 min after treatment, 57.4 ± 2.74% of UV-C irradiated cells showed DNA fragmentation (Fig. 1B, upper panels). DNA damage reduction was observed at 3, 6 and 12 h after treatment (27.11 ± 4.84, 8.79 ± 3.36 and 0.77 ± 2.59%, respectively). Propidium iodide stained cells were checked under the fluorescence microscope to confirm the absence of cytoplasmic stain (Fig. 1B, lower panels).

The comet assay (single-cell gel electrophoresis) is widely used to measure DNA damage and repair. Results obtained through neutral comet assay (Fig. 1C) confirmed the induction of DSBs in trophozoites by UV-C treatment. Typical comet-like structures were observed at 30 min and 3 h, while a reduction of the DNA tails was observed at 6 h after UV-C treatment. As expected, 12 h after the genotoxic insult, DNA migration was similar to the control untreated cells (No UV-C). Taking altogether, these data indicated that UV-C irradiation efficiently induced DNA damage and consequently, repair mechanisms were activated to restore DNA integrity allowing cell survival.

DNA DSBs induce early phosphorylation of yeast H2A (major H2A closer to mammalian H2AX) and human H2AX histones on a conserved serine residue located in the SQ motif at C terminus, producing γH2A and γH2AX, respectively 21. As in yeast, E. histolytica seems to have replaced the canonical H2A with H2AX 22. Two genes (locus EHI_126210 and EHI_188960) that encode putative proteins with 55 and 57% identity (e-value 2e-27and 2e-28) to yeast H2A and human H2AX histones, respectively, were found in the E. histolytica genome. These genes predict two 17.6 kDa conserved paralogous H2AX proteins that share 93% identity. Notably, both contain the H2AX exclusive SQ motif with the potentially phosphorylable serine residue (S156) (Fig. 2A).

Taking advantage of the high conservation between H. sapiens and E. histolytica H2AX C-terminus, we performed Western blot assays using the anti-human γH2AX antibody to detect serine-phosphorylated EhH2AX homologues (γEhH2AX) in cytoplasmic (CE) and nuclear (NE) extracts of trophozoites. Protein amount and integrity were confirmed on Coomassie blue stained-gels (data not shown). In NE from non-irradiated cells, we identified a 17-kDa weak band, which corresponds to the expected molecular weight of γEhH2AX histones (Fig. 2B, lane 2). Interestingly, 10 min after UV-C irradiation, this band was five-fold more intense, suggesting an increase in the amount of nuclear γEhH2AX, and 30 min after treatment no band was found (Fig. 2B, lanes 4 and 6). However, these assays did not allow us to distinguish whether one or both EhH2AX proteins were phosphorylated. In contrast, no signals were observed in CE (Fig. 2B, lanes 1, 3 and 5). We used as an integrity control an anti-EhPAP serum, which recognized the 63-kDa EhPAP protein 23 in non-irradiated and irradiated trophozoites (Fig. 2B, middle panel). In addition, an anti-actin monoclonal antibody, used as control for cell fractionation, strongly detected the expected 42-kDa band in CE and a slight signal in NE, as expected for a major component of cytoskeleton (Fig. 2B, lower panel). These data showed that UV-C irradiation of trophozoites is a useful model to generate DNA DSBs and study DNA repair in E. histolytica.

In order to investigate the presence of RAD52 epistasis group related genes in E. histolytica genome, we surveyed the parasite Pathema database (Table 1). We found Ehmre11, Ehrad50 and Ehnbs1 genes, which could encode the E. histolytica putative MRE11-RAD50-NBS1 protein complex that functions as the primary sensor of DNA DSBs in other organisms 9. Both EhMRE11 and EhRAD50 proteins exhibited 32 to 23% identities (e-values from 3e-49 to 9e-36) with S. cerevisiae and H. sapiens orthologous proteins, respectively; whereas the EhNBS1 sequence appears to be more divergent (17 to 24% identity and e-values from 0.003 to 0.002). E. histolytica also contains genes encoding the putative recombinase EhRAD51 and its paralogous protein EhRAD51C. EhRAD52, EhRAD54, EhRAD54B and EhRAD59 (EhRAD52/22 in Pathema database) predicted proteins are also encoded in the E. histolytica genome. As in yeast, rad51 paralogs (rad51b, rad51d, xrcc2 and xrcc3) that participate in HR in vertebrates were not found in E. histolytica (Table 1). In conclusion, E. histolytica genome contains a conserved set of repair genes, which suggests that it is skilled to perform recombinational DNA repair.

As a first step towards establishing the role of the E. histolytica RAD52 epistasis group related genes, we evaluated their mRNA expression by semi-quantitative RT-PCR using the UV-C irradiation model described above. Most genes exhibited a differential mRNA expression profile before and after irradiation (Fig. 3). Ehmre11, Ehrad51, Ehrad51c and Ehrad52 genes were transcribed at a very low level in non-irradiated trophozoites; meanwhile mRNA expression was induced from 30 min to 12 h after genotoxic damage. Particularly, the Ehrad51 mRNA expression was 16-, 11- and 4-fold increased at 30 min, 3 h and 12 h, respectively, after UV-C irradiation, when compared with untreated cells (Fig. 3A and 3B). On the other hand, the Ehnbs1, Ehrad54 and Ehrad59 genes were abundantly transcribed in untreated trophozoites; however, mRNA levels were down-regulated after UV-C treatment. The Ehrad50 gene expression showed the highest steady-state mRNA levels in non-irradiated trophozoites. At 30 min after UV-C irradiation, Ehrad50 transcript levels dropped drastically; 3 h later, they moderately increased, and at 12 h they diminished again. In contrast, Ehrad54b gene did not seem to be expressed under the experimental conditions tested here (Fig. 3A and 3B). We observed minimal changes in the 25S rRNA expression used as control (Fig. 3A, lower panel). These data showed that E. histolytica RAD52 epistasis group related genes were differentially expressed in response to DNA damage.

Since RAD51 recombinases are considered as key enzymes in HR and DNA repair processes in many organisms 24, we focused on the characterization of the E. histolytica EhRAD51 protein. Ehrad51 is an intron-less 1101 bp gene, which encodes a 367 amino acids (aa) polypeptide (40.3-kDa). Sequence similarity searches by BLAST showed the lowest e-values (from 3e-29 to 2e-20) and high identity (from 59 to 75%) with many eukaryotic RAD51 proteins, from plants to human, including protozoan parasites. Moreover, EhRAD51 showed 51% and 36% identity with Methanococcus voltae RADA and Escherichia coli RECA bacterial recombinases, respectively (Additional file 1). Amino acid sequence alignment of EhRAD51 protein with yeast and human RAD51 orthologs revealed that these proteins share functional and structural conserved motifs (Fig. 4A). EhRAD51 contains the putative polymerization motif (110–113 aa residues), which tethers individual subunits to form quaternary assemblies in human RAD51 protein 24 (Additional file 2). We also identified the ATPase Walker A or phosphate binding loop (P-loop: 152–159 aa residues) and Walker B motifs (240–249 aa residues), the ssDNA binding loops L1 (255–264 aa residues) and L2 (293–311 aa residues), as well as the ATP-stacking motif or ATP cap (342–350 aa residues) at the C-terminus, which are essential for nucleofilament assembling and ATP hydrolysis in RAD51/RECA-like recombinases 2627. Remarkably, the EhRAD51 N-terminus has a low-complexity region of 34-aa highly enriched in glutamic residues, which is not present in homologous proteins (Fig. 4A). Phylogenetic relationships among EhRAD51 and RAD51/RECA related proteins from diverse organisms, revealed a progressive evolution from eubacteria to eukaryotes, being EhRAD51 more related to protozoan recombinases (Fig. 4B).

The recombinant EhRAD51 protein (rEhRAD51) was expressed in E. coli BL21 (DE3) plysS strain as a 6x His-tagged fusion polypeptide and subsequently purified by affinity chromatography (Fig. 5A, lanes 3 and 4). By Western blot assays using monoclonal anti-6xHis tag antibodies, the purified rEhRAD51 was detected as a single 47 kDa band, which was slightly higher than the 44.1 kDa expected weight (Fig. 5B, lane 2). Then, rEhRAD51 was used to generate rabbit polyclonal anti-EhRAD51 antibodies. These antibodies recognized the 47 kDa rEhRAD51 band (Fig. 5B, lane 4), whereas the preimmune serum, used as negative control, did not detect any signal (Fig. 5B, lane 3). To evaluate the expression of the native EhRAD51 in E. histolytica, we performed Western blot assays using anti-EhRAD51 antibodies and protein extracts from irradiated and non-irradiated trophozoites. Antibodies reacted with a weak 46 kDa band in CE from non treated trophozoites, but not signal was detected in NE (Fig. 5C, higher panel, lanes 1 and 2). Meanwhile, at 30 min after UV-C irradiation, antibodies strongly detected the expected 41 kDa endogenous EhRAD51 in CE, but not in NE (Fig. 5C, higher panel lanes 3 and 4). Intriguingly, antibodies also detected a 46 kDa band in both NE and CE from UV-C irradiated trophozoites, which may correspond to a modified form of the 41 kDa protein. The specificity of anti-EhRAD51 antibodies was confirmed performing a similar Western blot assay using anti-EhRAD51 antibodies previously pre-incubated with purified rEhRAD51 protein and the recognition of both 46 and 41 kDa proteins was specifically inhibited (data not shown). In addition, the use of anti-EhPAP and anti-actin antibodies confirmed protein integrity and cell fractionation of CE and NE (Fig. 5C, middle and lower panels). Our findings showed that EhRAD51 was overexpressed in response to UV-C irradiation, and distributed in both nuclear and cytoplasmic compartments.

In order to better characterize the EhRAD51 expression and function, we investigated its subcellular location in trophozoites through immunofluorescence and laser confocal microscopy. In agreement with the Western blot results, EhRAD51 was detected at low levels in the cytosol of non-irradiated trophozoites (Fig. 6, panels A-D), whereas at 30 min after UV-C irradiation we noted a dramatic accumulation of cytoplasmic EhRAD51 protein. Interestingly, we also observed a scattered distribution of EhRAD51 typical foci-like structures in the nucleus (Fig. 6, panels E-H). Three hours later, the cytoplasmic signal diminished, while nuclear foci-like structures remained (Fig. 6, panels I-L). At 12 h after genotoxic damage, both cytoplasmic and nuclear EhRAD51 signals were very weak, being EhRAD51 foci-like structures scarce (Fig. 6, panels M-P). Quantification of nuclear foci like-structures by statistical microscopic analysis showed that about 60% of the cells contained at least one focus at 30 min after UV-C irradiation (Fig. 6Q). These findings confirmed that EhRAD51 was up-regulated after UV-C irradiation and suggested that it was redistributed into the nucleus during the first 3 h after DNA damage.

In silico analysis of the EhRAD51 aa sequence evidenced the presence of two putative DNA binding domains. To verify that EhRAD51 is a DNA binding protein, we performed EMSA using increasing amounts of purified rEhRAD51 protein and a fixed concentration of radiolabeled 50-bp ssDNA or 270-bp dsDNA fragments as probes. In order to discard interactions of contaminant E. coli proteins with DNA probes, we used mock purified fractions obtained from untransformed bacteria as a negative control. Results showed that incubation of rEhRAD51 with ssDNA and dsDNA probes resulted in five DNA-protein complexes (C_I-C_V) formation, suggesting that alternative populations of RAD51 protomers were associated to each DNA probes (Fig. 7A and 7B, lanes 2 to 4). The fastest migration ssDNA-protein complex C_I that was also formed with the mock fraction was considered as unspecific (Fig. 7A, lanes 5 to 7). No complexes were formed in the EMSA control performed with the dsDNA probe (Fig. 7B, lanes 5 to 7). Notably, the abundance of slow migration DNA-protein complexes appeared to increase in the presence of the highest rEhRAD51 amount (Fig. 7A and 7B, lanes 2 to 4). These results showed that rEhRAD51 was able to efficiently bind both ssDNA and dsDNA substrates in vitro.

In order to evaluate the homologous DNA strand transfer function of the rEhRAD51 protein, we performed a pairing assay involving the D-loop structure formation as described in Experimental procedures. Results revealed that rEhRAD51 was able to shift the electrophoretic mobility of the radioactive-labeled 200-bp ssDNA probe incubated with homologous circular dsDNA (Fig. 7C, lanes 2 to 4). This indicated that rEhRAD51 was able to catalyze ssDNA transfer to homologous dsDNA forming the three-stranded pairing molecule or D-loop structure, which has a reduced electrophoretic mobility in comparison with the ssDNA probe. The D-loop formation specificity was confirmed by incubation of rEhRAD51 and ssDNA probe in the absence of homologous dsDNA substrate (Fig. 7C, lane 5), and in the presence of a heterologous dsDNA substrate (Fig. 7C, lane 6), since no complex was observed. In addition, we did not observe any D-loop structure in the absence of rEhRAD51 (Fig. 7C, lane 1). Densitometric analysis of radioactive products showed that D-loop structure formation using 7.5 μg of rEhRAD51 was 3.6 and 1.7-fold higher than with 2.5 and 5 μg of rEhRAD51, respectively (Fig. 7D). These results indicated that EhRAD51 protein was able to catalyze specific DNA paring and exchange between DNA homologous strands in vitro.

Trophozoites of E. histolytica clone A (strain HM1: IMSS) were axenically cultured in TYI-S-33 medium 36 at 37°C and harvested during exponential growth phase.

Trophozoites (2 × 10⁶) grown in culture bottles were transferred into glass dishes and incubated at 37°C for 30 min. Medium and floating cells were discarded, and adhered trophozoites were irradiated with 254 nm UV-C light at 150 J/m² for 8 s using a UV Stratalinker 1800 device (Stratagene). After treatment, cells were incubated in fresh TYI-S-33 medium at 37°C for 0.5, 3, 6 and 12 h to be used in different experiments. Non-irradiated cells were used as a control in all experiments. Cell viability was monitored by microscopy using a trypan blue dye exclusion test. Assays were done three times by duplicate.

Trophozoites (2 × 10⁶) were harvested at 0.5, 3, 6 and 12 h after UV-C irradiation, washed with PBS 1× and fixed with 1% paraformaldehyde. After cell permeabilization with 70% ethanol, DNA damage was quantified using the APO-BrdUTP TUNEL Assay Kit (Molecular Probes) in order to detect 3'-hydroxyl ends in DNA. Permeabilized trophozoites were incubated at 37°C for 1 h in the DNA-labeling solution, which contains terminal deoxynucleotidyl transferase enzyme (TdT) and deoxythymidine analog 5-bromo-2'-deoxyuridine 5'-triphosphate (BrdUTP). Then, cells were washed twice and suspended in antibody staining solution (Alexa Fluor 488 dye-labeled anti-BrdU antibody) at room temperature for 1 h. After that, cells were incubated in propidium iodide/RNase A staining buffer at room temperature for 30 min. Samples were analyzed by flow cytometry in a BD FACS Calibur system and fluorescence data were plotted with the FloJo software.

Trophozoites (5 × 10⁴) were harvested at 0.5, 3, 6 and 12 h after UV-C irradiation. Neutral comet assay were performed using protocols from Tice and co-workers 37. Briefly, cells were mixed with agarose and spread over a warmed, precoated microscope slides. Agarose was allowed to solidify at 4°C, followed by immersion in cold lysis fresh solution (2.5 M NaCl, 100 mM EDTA, 10 mM Tris, pH 7) overnight. Next, electrophoresis was carried out in neutral buffer for 20 min at 1.5 V/cm (measured electrode to electrode) in the dark at 4°C. Finally, the slides were completely dried and ethidium bromide-stained DNA was observed at 400× magnification using an epifluorescence microscope (Leica DMIL).

The two Ehh2ax genes, which are homologous to the human h2ax gen, had been previously reported 22. Their existence in the E. histolytica Pathema database 38 were confirmed by BLAST using yeast H2A and human H2AX protein sequences as queries. The presence of phosphorylated forms of EhH2AX histone (γEhH2AX) in E. histolytica protein extracts obtained 10 or 30 min after UV-C irradiation was evaluated by Western blot assays using the anti-phospho-Histone H2AX (pSer¹³⁹), which was developed in rabbit using a synthetic phosphorylated peptide corresponding to 134–142 aa residues (including the phosphorylated Ser) of human H2AX histone C-terminus (Sigma). Subcellular fractionation to obtain CE and NE from clone A trophozoites was performed as described 39. Proteins were separated by 10% SDS-PAGE, transferred to nitrocellulose membranes (BioRad) and blocked with 1% BSA/PBS solution. Then, filters were incubated at room temperature for 2 h with the anti-human γH2AX polyclonal antibody (1:7000 dilution), washed with PBS 1× 0.05%Tween and incubated at 37°C for 1 h with goat anti-rabbit IgG horseradish peroxidase secondary antibody (Zymed) at 1:10000 dilution. Bands were revealed by ECL Plus Western blotting system (Amersham). As internal controls, we used polyclonal antibodies (1:1000 dilution) raised against the E. histolytica poly(A) polymerase EhPAP and anti-actin antibodies.

RAD52 epistasis group related genes were identified in E. histolytica Pathema database using both yeast and human protein sequences as queries. Putative E. histolytica orthologous proteins were selected from BLAST analysis according to the following criteria: (i) at least 20% identity and 35% homology to the query sequence; (ii) e-value lower than 0.002; and (iii) absence of stop codons in the coding sequence. Predicted aa sequences were aligned by the ClustalW software 40. Functional domains were predicted by the Prosite program 41. Phylogenetic inference was performed using the Neighbor-joining distance method 42 as implemented in the Molecular Evolutionary Genetics Analysis (MEGA version 3.1) software 43. Tree robustness was established by bootstrapping test, involving 1000 replications of the data based on the criteria of 50% majority-rule consensus.

Total RNA was obtained using Trizol reagent (Invitrogen) from trophozoites of clone A grown in basal culture conditions or after UV-C treatment. Semi-quantitative RT-PCR was performed as previously described 44 using 1 μg of total RNA and specific primers for each gene (Table 2). As a control, we amplified a 25S rRNA gene internal sequence. Products were separated by 6% PAGE, stained with ethidium bromide and submitted to densitometric analysis in a Gel doc 1000 apparatus (BioRad) using the Quantity One software. Data are the mean of three independent assays.

The 1098-bp full-length Ehrad51 gene was PCR-amplified from genomic DNA of clone A trophozoites using Ehrad51-S (5'-CG

GGATCC

AAGCTT

E. coli BL21 (DE3) pLysS bacteria were transformed with pRSET -Ehrad51 plasmid and grown at 37°C in 2-TY medium containing 100 μg/ml ampicillin and 34 μg/ml chloramphenicol to an OD_600nm of 0.6. The expression of rEhRAD51 was induced with 1 mM isopropyl beta-D-thiogalacto pyranoside (IPTG) at 37°C for 3 h. Cells were harvested, resuspended in lysis buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole, pH 8.0) and lysed by sonication at 4°C. Soluble rEhRAD51 was purified near to homogeneity under denaturing and native conditions through Ni²⁺-NTA affinity chromatography according to the manufacturer recommendations (Qiagen). Purified rEhRAD51 identity and integrity were confirmed by 10% SDS-PAGE and Western blot assays using anti-6xHis tag antibodies (Roche) at 1:5000 dilution and the ECL Plus Western blotting detection system (Amersham).

Purified rEhRAD51 was submitted to preparative 10% SDS-PAGE, electroeluted from Coomassie stained-gels and subsequently used as antigen to immunize a New Zeland male rabbit. An initial dose of 200 μg of rEhRAD51 in complete Freund's adjuvant (Sigma) was subcutaneously inoculated into the animal, and then three doses of 100 μg in incomplete Freund's adjuvant were injected every 15 days. One week after the last immunization, the rabbit was bled and polyclonal antiserum was obtained. IgGs were purified through protein G sepharose chromatography and tested for reactivity against rEhRAD51 protein by Western blot assays.

Western blot assays were performed using CE and NE proteins obtained before or 30 min after UV-C irradiation, and the membranes were incubated with anti-EhRAD51 polyclonal antibodies (1:1000 dilution) and goat anti-rabbit IgG horseradish peroxidase secondary antibody (Zymed)(1:10000 dilution). Immunodetected proteins were revealed with the ECL Plus Western blotting system (Amersham). The specificity of the anti-EhRAD51 antibodies was confirmed using anti-EhRAD51 antibodies pre-incubated with purified rEhRAD51 protein. As internal controls, we used polyclonal antibodies raised against the E. histolytica EhPAP 23 and actin proteins.

Trophozoites were grown on sterile coverslips, fixed with 4% paraformaldehyde at 37°C for 1 h, permeabilized with acetone and blocked with 1% BSA/PBS. Next, cells were incubated with anti-EhRAD51 polyclonal antibodies (1:200 dilution) at 37°C for 2 h, followed by the anti-rabbit fluoresceinated monoclonal antibody (1:100 dilution) at 37°C for 1 h. Then, trophozoites were washed three times with PBS 1× at room temperature and DNA was counterstained with propidium iodide (5 μg/ml) for 7 min. Light optical sections were obtained through a Nikon inverted microscope attached to a laser confocal scanning system (Leica) and analyzed by Confocal Assistant software Image J 45.

For the electrophoretic mobility shift assay (EMSA), we used two DNA probes: a 50-nt ss oligonucleotide (adh50) from the Ehadh112 gene 46, which was [γ-³²P]dATP (500 μCi/mmol) 3'-end labeled by T4 polynucleotide kinase at 37°C for 30 min, and a 270-bp dsDNA fragment (pgp270) of the 3'-UTR EhPgp5 gene [50] that was [α-³²P]dATP (200 μCi/mmol) uniformly labeled by PCR. EMSA was carried out in a 25 μl final volume reaction in binding buffer (50 mM Tris-HCl pH 7.8, 1 mM DTT, 10 mM MgCl₂, 1 mM ATP) in the presence of increasing amounts of native rEhRAD51 (0, 2.5, 5 and 7.5 μg). Reactions were started by addition of adh50 or pgp270 radiolabeled probes (10000 cpm) at 37°C for 15 min. Control assays were performed substituting purified rEhRAD51 by the mock purified fraction obtained from untransformed bacteria. DNA-protein complexes were resolved on 6% non-denaturing TBE polyacrylamide gels, vacuum-dried and exposed to Phosphor Imager screen (BioRad).

The EhRAD51 homologous DNA strand transfer activity was evaluated by the D-loop formation assay according to the described procedure 47. A ssDNA fragment of 200 bases (pgp200), which is complementary to the 3'-UTR Ehpgp5 gene cloned in the dsDNA plasmid 44, was [γ-³²P]dATP (500 μCi/mmol) 3'-end labeled by T4 polynucleotide kinase at 37°C for 30 min. Increasing amounts of rEhRAD51 (0, 2.5, 5 and 7.5 μg) were pre-incubated in reaction buffer (50 mM Tris-HCl pH 7.8, 1 mM DTT, 10 mM MgCl₂ and 1 mM ATP) with the pgp200 probe (10,000 cpm) at 37°C for 15 min. Then, homologous dsDNA plasmid (1 μM) was added and the mixture was incubated at 37°C for 30 min. A non-related plasmid was used as heterologous dsDNA control. Reactions were stopped by addition of 0.1% SDS. To prevent that EhRAD51 binds and shifts the pgp200 probe, samples were deproteinized with proteinase K (1 mg/ml) at 37°C for 10 min. Then, they were fractionated by 1% agarose gel electrophoresis in TAE 1× buffer and transferred to a nylon membrane (Amersham) in SSC 20× solution overnight. Homologous DNA strand transfer activity of rEhRAD51 was evaluated through the visualization of radioactive DNA products in a Phosphor Imager (BioRad) and quantified by densitometry using the Quantity one software (BioRad).