Entamoeba histolytica is a human pathogen which can grow using different sources of iron such as free iron, lactoferrin, transferrin, ferritin or haemoglobin. In the present study, we found that E. histolytica was also capable of supporting its growth in the presence of haem as the sole iron supply. In addition, when trophozoites were maintained in cultures supplemented with haemoglobin as the only iron source, the haem was released and thus it was introduced into cells. Interestingly, the Ehhmbp26 and Ehhmbp45 proteins could be related to the mechanism of iron acquisition in this protozoan, since they were secreted to the medium under iron-starvation conditions, and presented higher binding affinity for haem than for haemoglobin. In addition, both proteins were unable to bind free iron or transferrin in the presence of haem. Taken together, our results suggest that Ehhmbp26 and Ehhmbp45 could function as haemophores, secreted by this parasite to facilitate the scavenging of haem from the host environment during the infective process.
- Entamoeba histolytica
- haem-binding protein
- iron starvation
To satisfy their need for iron during cell growth, bacteria release molecules named siderophores . Siderophores pick up free iron and deliver it to the cellular membrane, where a specific receptor protein is responsible for binding and transporting it. In humans, iron is found attached to proteins called metalloproteins, because free iron is highly toxic. In particular, haemoproteins are successful at storing iron as they can contain protoporphyrin rings with a central iron atom attached. Often, haemoglobin is also used by pathogens as an iron source. The haemoglobin molecule is an assembly of four globins and each subunit contains a haem. To acquire this element from this plentiful supply, pathogens have developed an efficient mechanism that consists of the expression of proteins capable of binding haemoglobin or haem (haemoglobin receptors and haemophores) . Haemophores are proteins released by cells to uptake haemoglobin or haem, such as HasA from Serratia marcescens  and HupO in Vibrio fluvialis . Generally, haemoglobin receptors are inserted into the outer cellular membrane and they are necessary to bind external haem or haemoglobin [5–7]. Moreover, there are also receptors attached to internal cellular membranes such as HbR (haemoglobin receptor) of Leishmania donovani . Both haemophores and haemoglobin receptors bind the iron source using the conserved motifs FRAP (Phe-Arg-Ala-Pro) and NPNL (Asn-Pro-Asn-Leu) . However, haemophores present higher affinity for the iron source than receptors, since they have to scavenge iron during starvation, whereas receptors only have to wait for the iron supply . The molecular mechanisms of iron acquisition have been studied in bacteria; however, in protozoa, these processes are still poorly understood.
Entamoeba histolytica is a protozoan that infects humans. Under laboratory conditions, this pathogen can grow axenically and maintain its growth using different provisions of iron such as: free iron, lactoferrin, transferrin, ferritin or human haemoglobin [10–13]. In addition, we previously reported that, under iron starvation condition, this parasite expressed two genes encoding the Ehhmbp26 and Ehhmbp45 proteins. Analysis of their amino acid sequences revealed that they had the motifs necessary for haemoglobin binding. Furthermore, in vitro functional assays confirmed that both were capable of binding haemoglobin [14,15]. Consequently, they were related to the mechanism of iron acquisition when E. histolytica uses haemoglobin exclusively as its iron source. In the present study, we found evidence that E. histolytica can support its growth in the presence of haem as the only iron supply. Interestingly, both Ehhmbp26 and Ehhmbp45 were secreted into the medium, possibly to bind haem. These proteins also presented an elevated binding affinity for haem. Our results suggest that Ehhmbp26 and Ehhmbp45 could function as haemophores secreted by this parasite to facilitate scavenging of haem from the host environment during the infective process.
Trophozoites of E. histolytica strain HM1-IMSS were grown axenically in TY1-S-33 medium  supplemented as reported previously . To analyse the effect of iron, trophozoites were harvested by centrifugation at 500 g for 5 min, the cellular pellet was washed with PBS and suspended (106 trophozoites/ml) in medium (TYI: trypticase/yeast extract/iron) without iron. Cells were iron-starved for 2 h [10,17]. Then trophozoites (104) were inoculated in 10 ml of TYI medium, which was supplemented with equimolar quantities (10 mM iron) of haem, haemoglobin or ammonium ferric citrate. Cell viability was monitored by microscopy using the Trypan Blue dye-exclusion test.
Spectrophotometric determination of haem
Cells were iron-starved for 2 h, thereafter 10 mM haemoglobin was added. Every 5 min, an aliquot of 2 ml was collected during 60 min. Each sample was centrifuged at 500 g for 10 min to separate medium and trophozoites. Trophozoites were washed twice with PBS and were lysed using the freeze–thaw method. In the medium and lysate fractions, the absorbance (397 nm) of haem was determined by spectrophotometric analysis, which is widely used for this purpose [18,19]. Absorbance values were adjusted by subtraction of values obtained for similar assays performed without addition of the haemoglobin molecule.
Overexpression of Ehhmbp26 and Ehhmbp45 proteins
Escherichia coli BL21(DE3) pLysS bacteria were transformed with pehhmbp26  or pehhmbp45  plasmids, which contained the ehhmbp26 and ehhmbp45 genes respectively. Transformed bacteria were grown at 37 °C in LB (Luria–Bertani) medium with added ampicillin (100 μg/ml) until the culture reached a D600 of 0.6. Protein expression was induced by adding 1 mM IPTG (isopropyl β-D-thiogalactopyranoside) at 37 °C over 3 h. Protein expression was visualized by SDS/PAGE (12% gels) and staining with Coomassie Blue and corroborated by Western blotting using anti-GST (glutathione transferase) monoclonal antibodies (1:8000 dilution).
Purification of the GST–Ehhmbp26, GST–Ehhmbp45 proteins and GST digestion
Induced bacteria were harvested and suspended in PBS (pH 7.3) plus lysozyme (1 mg/ml). Then, cells were lysed by sonication at 4 °C for 1 min, with pulses of 10 s, using 60% amplitude. Triton X-100 was added up to final concentration of 1%. After 1 h, the mixture was centrifuged at 10000 g for 20 min. Supernatant was subjected to glutathione–Sepharose affinity chromatography (GE Healthcare). The flowthrough was collected and the column was washed five times with PBS. The remaining proteins were released with elution buffer [75 mM Hepes (pH 7.4), 150 mM NaCl, 15 mM reduced glutathione, 0.5 mM DTT (dithiothreitol) and 0.1% Triton X-100]. Each fraction was analysed by SDS/PAGE (12% gels) followed by staining with Coomassie Blue. To corroborate the presence of the Ehhmbp26 and Ehhmbp45 proteins coupled to GST, Western blotting was performed using anti-GST antibodies (1:8000 dilution). Fractions containing the GST–Ehhmbp26 or GST–Ehhmbp45 proteins were dialysed at 4 °C for 12 h with PreScission cleavage buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA and 1 mM DTT) and treated for 12 h at 4 °C with PreScission Protease to remove the GST polypeptide [14,15]. Finally, proteins were recovered using a glutathione–Sepharose affinity column and dialysed overnight at 4 °C against PBS at pH 6.8.
Production of specific antibodies raised against Ehhmbp26 and Ehhmbp45 proteins
Purified Ehhmbp26 and Ehhmbp45 were used as antigens to immunize rabbits (New Zealand white strain). Immunization was performed, intravenously, four times a day for 8 days, using 150 μg of proteins resuspended in PBS (350 μl) and complete Freund's adjuvant (350 μl). Thereafter, antisera were obtained and tested using total E. histolytica extracts or both Ehhmbp26 and Ehhmbp45 purified proteins.
Western blotting assay
Proteins fractions from E. histolytica cultures were resolved by SDS/PAGE (12% gels) and transferred on to nitrocellulose membranes (Bio-Rad Laboratories). Membranes were blocked with 5% (w/v) non-fat dried milk powder in PBST (PBS with 0.002% Tween 20) and then the serum specific for either Ehhmbp26 or Ehhmbp45 proteins was added and incubated for 2 h, at room temperature (25 °C). After three washes (45 min) with PBST, the membranes were incubated with horseradish-peroxidase-conjugated anti-(rabbit IgG) secondary monoclonal antibodies (Zymed) (1:20000 dilution). As internal and fractionation controls, we used anti-actin polyclonal antibodies (1:1000 dilution)  and anti-EhADH112 monoclonal antibodies (1:9 dilution) . The results were visualized using the ECL (enhanced chemiluminescence) plus detection system (GE Healthcare).
To test the capability of the Ehhmbp26 and Ehhmbp45 proteins to bind haem, binding assays were performed. First, purified proteins (20 μM) were suspended in binding buffer (250 mM Tris/HCl, pH 8.0, 5 mM EDTA and 10% glycerol) and 20 μM haem was added. The mixture was incubated at 37 °C for 30 min. Thereafter the complex was separated by native PAGE (7% gels) . To identify haem interaction, native gels were revealed with hydrogen peroxide (3.5%), and proteins were identified by Western blotting assays using specific antibodies.
Haem and haemoglobin assays to estimate the binding affinity
In order to estimate the affinity of the Ehhmbp26 and Ehhmbp45 proteins by haem or haemoglobin, binding assays were performed by UV–visible spectrophotometry. Purified proteins (20 μM) in binding buffer (250 mM Tris/HCl, pH 8.0, 5 mM EDTA and 10% glycerol) was titrated with increasing concentrations of haem (1, 2, 7, 15, 46, 77, 150, 310, 460, 610, 770 and 930 μM) or haemoglobin (1, 2, 7, 15, 46, 77, 150, 310, 460, 610 and 770 μM). The absorbance of the UV–visible spectrum between 200 and 800 nm was measured using a spectrophotometer. Spectra were recorded 1 min after the addition of each haem or haemoglobin aliquot in triplicate on three separate occasions. The intensity of Soret peaks at 250 nm of the haem–Ehhmbp26, at 412 nm of haemoglobin–Ehhmbp26, at 255 nm of haem–Ehhmbp45 and at 415 nm of haemoglobin–Ehhmbp45 were used to monitor complex formation. To generate the binding isotherms, absorbance values were plotted against the haem or haemoglobin concentration (M) respectively. Data were fitted to a one-binding-site model using the non-linear regression function to determine the dissociation constant (Kd) assuming that binding follows the Law of Mass Action. Statistical analysis, binding stoichiometry and affinity estimations were performed using GraphPad Prism version 5.00 for Windows (http://www.graphpad.com).
Competitive binding assays
To analyse whether free iron or transferrin have the capacity to compete for Ehhmbp26 or Ehhmbp45, haem-binding and competitive binding assays were performed. Proteins (20 μM) were suspended in binding buffer with haem (20 μM) and increasing concentration of citratic ferric (1, 2, 3 or 4 mM) or transferrin (2, 3 or 4 mM) were added. Mixtures were incubated at 37 °C for 30 min. Thereafter, the complexes were separated by native PAGE (7% gels).
Fractionation of E. histolytica cultures
To explore the location of Ehhmbp26 and Ehhmbp45, we separated proteins from cultures grown under iron-absence conditions. Fractions were obtained for EPs (extracellular proteins), CPs (cytoplasmic proteins) and MPs (membrane proteins). After collection of medium, cells were washed to remove traces of secreted proteins. Then, trophozoites were suspended in PBS containing PHMB (p-hydroxymercuribenzoate) protease inhibitor (100 mM for each 1.2×106 cells), and they were broken mechanically using a homogenizer. Samples were centrifuged at 500 g for 10 min at 4 °C to eliminate unbroken cells and debris. To obtain cytoplasmic and membrane fractions, samples were ultracentrifuged at 38000 rev./min for 1 h at 4 °C using an SW40Ti rotor (Beckman). The integrity of proteins was tested using SDS/PAGE (12% gels) followed by staining with Coomassie Blue. The fractions obtained were used in Western blotting assays using specific antibodies raised against the Ehhmbp26 and Ehhmbp45 proteins. Results were expressed as the percentage of total protein in each fraction. As fractionation controls, we performed Western blotting for actin and EhADH112 proteins. All experiments were performed in triplicate.
Haem supports the cellular growth of E. histolytica
E. histolytica is a pathogenic protozoan that can maintain its iron requirements using haemoglobin. This haemoprotein is composed of four globins, each one attached to a haem, which is capable of binding iron. We studied the growth kinetics after iron starvation, using medium supplemented with haem, haemoglobin or free iron. Cell viability was monitored by microscopy using a Trypan Blue dye-exclusion test. When the medium was maintained without an iron source, cells began to die after the third day. Interestingly, haem, more than haemoglobin, supported the cell growth for almost 8 days indicating that haem was more efficient than haemoglobin and that E. histolytica has a preference for this as an iron supply (Figure 1).
E. histolytica can internalize haem from haemoglobin
Both haem and haemoglobin support the cellular growth of E. histolytica, but starved trophozoites have a preference for haem. This fact suggests that haemoglobin may be being degraded, and the released haem is being used as source of iron. To investigate this hypothesis, we performed experiments in which the absorption spectrum at 397 nm of haem was monitored in the culture medium and within cells, after haemoglobin was added to cultures of starved cells. The absorbance values were adjusted to dismiss the contribution of the haem that could come from the parasites' own haemoprotein. We found that, during the first 25 min, haem increased in both the medium and the trophozoites by similar proportions (Figure 2). After 30 min, the amount of haem in the medium remained constant, whereas the amount of haem inside the cells increased up to 40 min before remaining essentially constant. These results clearly show that, under these conditions, the release of haem from the haemoglobin molecule was induced, and consequently the iron–protoporphyrin ring was internalized into the cells.
Ehhmbp26 and Ehhmbp45 proteins are expressed by E. histolytica
It has been reported that, under iron starvation, E. histolytica expresses the ehhmbp26 and ehhmbp45 transcripts [14,15]. In order to identify the Ehhmbp26 and Ehhmbp45 proteins in this pathogen, specific antibodies raised against the purified proteins were generated in rabbits as mentioned in the Experimental section. Both proteins were detected by Western blotting in total extracts of E. histolytica from cultures, grown under iron-absence conditions (Figures 3A and 3B, lanes 4). Their identity was confirmed using particular purified protein from a heterologous system (Figures 3A and 3B, lanes 5). This result indicates that both proteins are translated in E. histolytica and they probably have an important contribution to the biology of this parasite.
Ehhmbp26 and Ehhmbp45 purified proteins bind haem
The capacity of E. histolytica to support its cellular growth using haem as iron supply and the presence of haem inside the trophozoites suggest that a specific mechanism to utilize it must be orchestrated. To correlate the use of haem as an iron source with the synthesis of the Ehhmbp26 and Ehhmbp45 proteins by this protozoan, we evaluated the in vitro haem-binding activity for these proteins. In order to approximate our experimental system to a model for the in vivo infection process in E. histolytica, we performed all binding assays at pH 8.0, which is close to that found in the ulcerative intestine, where plenty of haemoglobin, loaded with haem, becomes accessible to this parasite. Both proteins were capable of binding haem (Figure 4A, lanes 3 and 6). Migration of haem–protein complexes was slower compared with free haem and purified proteins. As expected, the unrelated protein GST, used as negative control, was not able to bind haem (Figure 4A, lane 9). The identity of each protein was confirmed by Western blotting using specific antibodies (Figures 4B–4D, lanes 1 and 3). We reported previously that Ehhmbp26 and Ehhmbp45 function as haemoglobin-binding proteins [14,15]. Furthermore, our new findings indicate that both proteins are capable of binding haem too.
Ehhmbp26 and Ehhmbp45 proteins bind haem and haemoglobin with dissimilar affinity
Haem and haemoglobin are used as iron sources by E. histolytica. Even when both proteins bind haemoglobin and haem, the binding behaviour could be different. To explore this assumption, we calculated the Kd for both proteins using haemoglobin or haem as a ligand. Binding assays for both purified proteins confirmed that they have haem- and haemoglobin-binding activity. Plotting absorbance values of the complexes formed against the haem or haemoglobin concentration yielded saturation curves for both proteins (results not shown), indicating that Ehhmbp26 and Ehhmbp45 bind haem and haemoglobin in a 1:1 stoichiometry. In addition, we found that the affinity of Ehhmbp26 and Ehhmbp45 for haem was stronger than that for haemoglobin. Calculations showed Kd values in the micromolar range for haem binding, whereas for haemoglobin, these values were in the millimolar order. Ehhmbp26 had the best affinity for haem (Kd=0.48 ± 0.13 μM), whereas Ehhmbp45 binding affinity for this ligand was lower (Kd=1.39 ± 0.10 μM). On the other hand, the haemoglobin concentration needed to achieve a half-maximum binding at equilibrium for Ehhmbp26 and Ehhmbp45 to haemoglobin were 0.13 ± 0.05 mM and 0.025 ± 0.011 mM respectively, indicating that these proteins predominantly bind haem.
Ehhmbp26 and Ehhmbp45 purified proteins do not bind either free iron or transferrin
Haem is responsible for attaching the iron atom. To test whether Ehhmbp26 and Ehhmbp45 proteins bind iron directly, we performed competitive binding experiments. In both cases, haem was not replaced by free iron, even when a 200-fold molar excess was used (Figures 5A and 5B, lanes 4–7). Transferrin, another iron source, was also employed to test the binding specificity. As observed for free iron, this metalloprotein did not compete with haem either (Figures 5C and 5D lanes 4–6). This result clearly demonstrated that the Ehhmbp26 and Ehhmbp45 proteins bind the iron–protoporphyrin ring in spite of the presence of high amounts of free iron or transferrin.
Ehhmbp26 and Ehhmbp45 proteins are secreted by E. histolytica under iron starvation
Expression of the Ehhmbp26 and Ehhmbp45 proteins was induced by iron starvation in E. histolytica [14,15]. To corroborate whether these proteins are haemophores and are secreted by E. histolytica, we cultivated trophozoites under these conditions and the presence of these proteins in cytoplasmic, membrane and extracellular fractions was analysed. Over 99% of total Ehhmbp26 and Ehhmbp45 proteins were strongly detected by Western blotting in the culture medium (EP) (Figures 6A and 6B). Some traces of these proteins were found in the soluble fraction (CP) (Figures 6A and 6B); they could correspond to the traffic of proteins after their synthesis. Moreover, they were not recognized in the membrane fraction (MP) (Figures 6A and 6B). Also, as expected for a major component of cytoskeleton, the predictable 42 kDa band for actin was immunodetected in the CP fraction and not in the MP or EP fractions. In the same way, anti-EhADH112 antibodies recognized a proper protein only in the MP fraction (Figure 6C). All of our results suggest that the Ehhmbp26 and Ehhmbp45 proteins are secreted by E. histolytica, probably to scavenge iron from haem or haemoglobin. This would resemble a haemophore-like function during the invasive process.
E. histolytica is a pathogenic protozoan. To establish infection in the human intestine, it requires nutrients from the host. For instance, this micro-organism can acquire iron from lactoferrin, ferritin or haemoproteins . Haemoglobin is a necessary haem source for pathogens, because it provides four iron–protoporphyrin rings. In the present study, we found evidence that haem supported the cellular growth of E. histolytica more than haemoglobin. In addition, when starved trophozoites were supplemented with this haemoprotein as the only iron supply, haem was detected both in the medium and inside cells (Figure 2). Moreover, under iron scarcity, this parasite secreted the Ehhmbp26 and Ehhmbp45 proteins which were capable of binding haem with high affinity.
We established the experimental conditions to confirm that E. histolytica uses free haem as an iron source. They consisted of adding 10 mM free haem to the medium. Higher concentrations were not required. Since traces of iron from medium were not sufficient to support E. histolytica cellular growth (Figure 1), it was unnecessary to use a chelating agent to deplete this metal. These findings are, in fact, in contrast with those reported previously by Serrano-Luna et al. , who showed that E. histolytica was unable to use haem as a sole iron source, in spite of the fact that haem was not toxic to cells. Our results also indicate that free haem supported cell growth better than haemoglobin. It is possible that haem becomes less accessible when it is inside the haemoglobin molecule. In bacteria, the complete mechanism involves the secretion of proteases to digest haemoproteins and, consequently, to release the haem . If a similar process occurs in E. histolytica, the liberated haem could well be bound by Ehhmbp26 or Ehhmbp45, and thereafter transported to the cellular membrane components, in order to be internalized. E. histolytica secretes proteases, such as EhCP112, that digest haemoglobin . This antecedent would explain why haem was observed in the medium, under conditions in which haemoglobin was added as the only iron supply (Figure 2). In addition to this, the binding affinities were lower for haemoglobin (Ehhmbp26 Kd=0.13 ± 0.05 mM; Ehhmbp45 Kd=0.025 ± 0.011 mM) in contrast with haem (Ehhmbp26 Kd=0.48 ± 0.13 μM; Ehhmbp45 Kd=1.39 ± 0.10 μM), and the last one becomes the major target of Ehhmbp26 and Ehhmbp45 proteins. Low affinity constants, in the micromolar range, have also been reported for Photobacterium damselae HutB protein, Serratia marcenscens TonB-dependent haem receptor HasR, Porphyromonas gingivalis outer membrane haem receptor HmuR and for HasA (haemophore) from Serratia marcenscens [24–26].
Our results suggest that Ehhmbp26 or Ehhmbp45 could be actively engaged as haem-binding proteins. We observed that haem–Ehhmbp26 and haem–Ehhmbp45 complexes were highly specific, since the addition of free iron or transferrin, the main extracellular protein in human serum, did not compete for binding (Figures 4 and 5). Haemophores are molecules secreted to bind specifically haem or haemoproteins, [22,27,28]. It has been reported that several bacteria, under iron scarcity, secrete siderophores to scavenge free iron [1,29,30]. Our experimental data support the hypothesis that both proteins did not bind free iron and that they specifically bind haem, consequently they could be haemophores more than siderophores.
This finding correlates with the transcriptional activation of both genes observed under these conditions [14,15]. Perhaps, in other situations such as iron sufficiency, a different cellular distribution for these proteins could be promoted. Taken together, our results provide the first evidence of a mechanism to obtain iron from haemoglobin and haem, which involves Ehhmbp26 and Ehhmbp45 secretion to scavenge haem. This event could be essential during the infective process in the intestine, where bleeding ulcers are produced by E. histolytica colonization. After that, if trophozoites reach the bloodstream, erythrophagocytosis is the main mechanism to satisfy the iron and haem requirements of this pathogen .
Areli Cruz-Castañeda performed the experiments. Areli Cruz-Castañeda, Mavil López-Casamichana and José Olivares-Trejo designed the experiments, analysed the data and wrote the paper.
This work was supported by Consejo Nacional de Ciencia y Tecnologia (CONACyT) [grant number 54212].
We thank Dr Norma Velázquez (Hospital Infantil Federico Gómez, Mexico City, Mexico) and Dr Máximo Martínez (Centro de Vigilancia Epidemiologica, Distrito Federal, Mexico City, Mexico) for their help in the generation of specific antibodies used in this work. We also thank Dr Guillermina Garcia (Infectomica y Patogenesis Molecular, Centro de Estudios Avanzados IPN, Mexico City, Mexico) and Dr Elisa Azuara (Posgrado en Ciencias Genomicas, Universidad Autonoma de la Ciudad de Mexico) for the donations of anti-actin and anti-EhADH112 antibodies utilized as controls in this study.
Abbreviations: CP, cytoplasmic protein; DTT, dithiothreitol; EP, extracellular protein; GST, glutathione transferase; MP, membrane protein
- © The Authors Journal compilation © 2011 Biochemical Society