Register for Mechanobiology 2016!

Arginine deiminase has multiple regulatory roles in the biology of Giardia lamblia
Maria Carolina Touz, Andrea Silvana Rópolo, Maria Romina Rivero, Cecilia Veronica Vranych, John Thomas Conrad, Staffan Gunnar Svard, Theodore Elliott Nash


The protozoan parasite Giardia lamblia uses arginine deiminase (ADI) to produce energy from free L-arginine under anaerobic conditions. In this work, we demonstrate that, in addition to its known role as a metabolic enzyme, it also functions as a peptidylarginine deiminase, converting protein-bound arginine into citrulline. G. lamblia ADI specifically binds to and citrullinates the arginine in the conserved CRGKA tail of variant-specific surface proteins (VSPs), affecting both antigenic switching and antibody-mediated cell death. During encystation, ADI translocates from the cytoplasm to the nuclei and appears to play a regulatory role in the expression of encystation-specific genes. ADI is also sumoylated, which might modulate its activity. Our findings reveal a dual role played by ADI and define novel regulatory pathways used by Giardia for survival.


Giardia lamblia is a ubiquitous unicellular parasite of humans and other vertebrates that commonly causes diarrhea and gastrointestinal upset (Adam, 2001). G. lamblia trophozoites undergo fundamental biological changes in response to adverse environmental conditions. To survive outside the host's intestine, Giardia trophozoites differentiate into infective cysts, which are then released with the feces and are responsible for the transmission of the disease among susceptible hosts. Another mechanism of adaptation is antigenic variation (a process by which the parasite continuously switches its major surface molecules), allowing the parasite to evade the host's immune response while residing in the small intestine (Nash, 2002). Only one variant-specific surface protein (VSP), from a pool of approximately 250 genes present in the genome of the parasite, is expressed on the surface of G. lamblia trophozoites at any point in time (Morrison et al., 2007). These antigens are membrane proteins with common characteristics, such as their variable N-terminal extracellular region and a well-conserved C-terminus – which consists of a hydrophobic intramembranous domain that anchors the VSPs to the parasite surface as well as a perfectly conserved five-amino-acid (CRGKA) C-terminal tail located in the cytoplasm (Mowatt et al., 1991; Mowatt et al., 1994). However, the biological purpose of the conserved elements in VSPs and their role in antigenic variation are still unknown.

The function of the CRGKA tail of VSPs is unclear, but recent evidence indicates that it is important in VSP biology. It has been previously shown that many VSPs are palmitoylated (Hiltpold et al., 2000; Papanastasiou et al., 1997b), and we recently established that palmitic acid binds specifically to the Cys of the CRGKA tail of the VSPs, thus controlling the segregation of these proteins to detergent-resistant domains at the plasma membrane (Touz et al., 2005). Consequently, VSP mutants unable to bind palmitic acid show abnormal membrane segregation and avoid VSP-specific-antibody cytotoxicity, performing an essential function in the control of VSP-mediated signaling and processing (Touz et al., 2005). These results, taken together, prompted us to investigate whether the CRGKA cytoplasmic tail contains other post-translational modifications or interacting proteins involved in the control of signaling.

In this work, we found that G. lamblia arginine deiminase (ADI, encoded by ArcA) bound specifically to the CRGKA sequence. Prior studies in G. lamblia determined that ADI catalyzes the irreversible catabolism of arginine to citrulline in the arginine dehydrolase (ADH) pathway and serves as an important source of energy (Schofield et al., 1990). This pathway has been regarded as being restricted to prokaryotic organisms. By contrast, higher eukaryotes use nitric-oxide synthase (NOS) to convert free L-arginine into citrulline and nitric oxide (NO), and use peptidylarginine deiminases (PADs) to deiminate protein-bound arginine and convert it to citrulline (citrullination) by a Ca2+-dependent mechanism. This post-translational modification has a major impact on the structure and function of the target protein (Vossenaar et al., 2003). We discovered that the function of ADI goes beyond energy production because it also functions as a PAD. We present evidence indicating that ADI plays an essential role in the control of antigenic variation via VSP citrullination and influences the process of encystation. Our findings on the participation of ADI in unusual post-translational modification mechanisms during G. lamblia survival are discussed.


ADI interacts with VSPs

Almost all VSP genes encode the type-I integral membrane proteins that have a five-amino-acid cytoplasmic tail, CRGKA. Because this cytoplasmic sequence is highly conserved, we examined the possibility of CRGKA interaction with regulatory proteins. Pull-down assays using the synthesized peptide NH2-HHHHHHCRGKA-COO (His6-CRGKA) and further liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis showed an associated protein corresponding to the enzyme arginine deiminase (ADI; GGD 112103) in both isolate GS clone H7 (GS/H7) and isolate WB clone 1267 (WB/1267) (Fig. 1A). The 85 kDa band corresponded to the modified whole protein (see below), and the ∼13 kDa band to the protein C-terminus (Fig. 1A). In addition, using the yeast two-hybrid system, we determined that the entire VSPH7 [H7-BD (binding domain)] and VSP1267 (1267-BD) proteins, but not VSP1267 lacking the cytoplasmic tail CRGKA (1267-tail-BD), interacted with ADI [ADI-AD (activation domain)], thus validating the VSP-ADI interaction found by the peptide pull-down assays (Fig. 1B).

Fig. 1.

G. lamblia ADI is associated with VSPs through their cytoplasmic tail. (A) Left: representation of the pull-down assays. Right: SDS-PAGE and Coomassie staining show an ∼85 kDa and an ∼13 kDa protein from both GS/H7 (GS) and WB/1267 (WB) isolates. Identification of these bands as being ADI was performed by LC/MS-MS. Controls without peptide fail to pull down any protein. (B) Protein-protein interactions were detected by the ability of yeast cells (AH109) to grow on selective plates. In the upper-left panel, the expression of the entire VSPH7 protein (H7-BD), the entire VSP1267 protein (1267-BD) and VSP1267 lacking its cytoplasmic tail (1267-tail-BD) with ADI (ADI-AD) is revealed by the presence of white colonies in minimal medium lacking leucine and tryptophan (–L/–T medium). Interaction of ADI-AD with both VSPH7-BD and VSP1267-BD is shown in the bottom-left panel by the growth of yeast colonies in plates lacking tryptophan, leucine and histidine [TDO (triple-dropout medium) plates]. No interaction between ADI and 1267-tail-BD is observed. Controls of the methodology include ESCP-BD–MuA-AD interaction (+) and emptyBD–ADI-AD vector (–). (C) Schematic representation of wild-type VSPH7 and transgenic VSPH7 proteins. The VSPH7 ORF contains a signal peptide, an extracellular domain, a transmembrane domain and a cytoplasmic tail. H7-HA has a HA epitope sequence at the C-terminus. H7-tail possesses the HA epitope right after the transmembrane domain of H7. ΔR-H7 is VSPH7 containing a point mutation of the R residue of its cytoplasmic tail. (D) H7-HA and ΔR-H7, but not H7-tail, co-immunoprecipitate with ADI. Antibody against ADI was used to immunoprecipitate comparable amounts of protein from WB transgenic cells. Western blotting of original lysate was stained with anti-HA mAb labeled with alkaline phosphatase (Lysate, bottom panel). UT, untransfected cells; STD, molecular weight standard.

To confirm the interaction of ADI with the tail of VSPs, we created WB-strain transgenic trophozoites that stably expressed different VSPH7 versions [H7-HA is VSPH7 with a hemagglutin (HA) epitope sequence at the C-terminus; H7-tail is VSPH7 with a HA epitope right after its transmembrane domain; and ΔR-H7 is VSPH7 containing a point mutation of the R residue of its cytoplasmic tail] (Fig. 1C) and performed immunoprecipitation (IPP) assays. VSPH7 is not genetically codified nor expressed in the G. lamblia WB/1267, thereby allowing detection of the extracellular portion of VSPH7 using VSPH7-specific monoclonal antibody (mAb) G10/4 and allowing detection of the C-terminus by using anti-HA mAb. Our results showed that H7-HA and ΔR-H7, but not H7-tail, co-immunoprecipitate with ADI polyclonal antibody (pAb) (Fig. 1D). This suggests that the whole tail of the VSPs rather than one amino acid is necessary for ADI-VSP interaction, as expected from an enzyme-substrate association.

ADI acts as a PAD

Using a C-terminal HA-tagged ADI (ADI-HA) and immunofluorescence assay (IFA), we found that the subcellular localization of ADI-HA is cytoplasmic (Fig. 2A). We next examined whether ADI localizes close to the tail of VSPs. Confocal imaging showed that ADI partially colocalizes with VSP9B10 close to the plasma membrane (Fig. 2, merge, top panel). Similar results were found when anti-ADI pAb, raised against recombinant ADI protein (Ringqvist et al., 2008) was used, thereby confirming the cytoplasmic localization of ADI as well as its partial colocalization with VSP9B10 (Fig. 2A, bottom panels).

Two observations suggested that ADI possesses PAD activity. First, by sequence analysis and 3D-structure prediction, we found that G. lamblia ADI contains a `Cys-His-Glu' active site similar to the PADs (data not shown). Second, we discovered that ADI binds to the cytoplasmic tail of VSPs, a finding not expected of an enzyme involved exclusively in the energy pathway. To determine whether VSPs are citrullinated, western blotting using extracts from cloned trophozoites expressing VSP1267, VSP9B10 (both from isolate WB), or VSPH7 (from isolate GS) was performed. Fig. 2B shows that only the full-length VSP1267, VSP9B10 and VSPH7 proteins are citrullinated. It was previously shown that VSPs are processed by deletion of the conserved C-terminus (Papanastasiou et al., 1997a; Touz et al., 2005). This event explains why the processed N-terminus (Fig. 2B, lower band of VSP1267 and VSPH7), which is also recognized by their specific mAbs, is not citrullinated.

To demonstrate that ADI possesses PAD activity, we mixed active ADI-HA, purified from transgenic G. lamblia trophozoites, and His6-CRGKA. After His6-CRGKA purification and western blotting, we were able to demonstrate that ADI, but not a non-related enzyme, citrullinates the His6-CRGKA peptide (Fig. 2C). The requirements for PAD activity of ADI to citrullinate the VSP tail were determined by analysis of transfected VSPH7 mutants for the presence of citrulline. Citrullination was found in H7-HA, but not in H7-tail or ΔR-H7, indicating that citrullination could not occur without the presence of the crucial arginine in the CRGKA tail (Fig. 2D). As shown in the bottom panel of Fig. 2D, equal amounts of transgenic VSPs were detected after IPP.

Fig. 2.

VSP citrullination is probably mediated by the PAD activity of ADI. (A) Top panels: confocal direct IFA was performed on permeabilized cells, showing a cytoplasmic distribution of ADI-HA (green) by using FITC-labeled anti-HA mAb and its partial colocalization (yellow in merge) with VSP9B10 (red) underneath the plasma membrane of the transgenic trophozoite. Nuclei are stained with DAPI (blue). Bottom panels: the same result was obtained using Alexa-Fluor-488–anti-ADI (green) in wild-type cells. Texas-Red–9B10 mAb was used to visualize VSP9B10 (red). Scale bars: 10 μm. (B) Western blotting of G. lamblia homogenates expressing different VSPs that reacted with anti-citrulline pAb (left panel). The same filter membranes were stripped, cut and reacted with 5C1, 9B10 and G10/4 mAbs against VSP1267, VSP9B10 and VSPH7, respectively (right panel), indicating that these VSPs are citrullinated. STD, molecular weight standard. (C) Dot-blotting to detect citrullination of the H6-CRGKA peptide after incubation with the purified recombinant ADI-HA (ADI-HA). A non-related purified enzyme ESCP-HA was used as a negative control. Dot-blotting to detect H6-CRGKA was performed using anti-H6 mAb. (D) Top panel: specific citrullination of the CRGKA tail is shown. Western blotting using anti-citrulline pAb performed after immunoprecipitation with anti-HA mAb of H7-transgenic trophozoites. Bottom panel: the presence of H7-HA, H7-tail and ΔR-H7 after immunoprecipitation was analyzed using anti-HA mAb labeled with alkaline phosphatase. UT, untransfected cells.

ADI participates in the control of antigenic variation

Antibodies directed against VSPs either induce antigenic variation in, inhibit the growth of or kill recognized trophozoites, allowing the repopulation of trophozoites expressing other VSPs. In vitro, results depend either on the concentration of the antibodies added to G. lamblia cultures or the exposure time. It has been observed that a high concentration of the antibodies cause immediate immobilization as well as detachment and aggregation of the trophozoites. By contrast, it was reported that, when a low concentration of anti-VSP antibody is added to the culture (or during a short time), no cytotoxicity is observed, but a rapid accumulation of parasites expressing VSPs that are different to the one that was originally expressed arise (Aggarwal et al., 1989; Stager et al., 1997; Touz et al., 2005).

Fig. 3.

ADI participates in the control of cell death, probably by altering VSP switching. (A) Cytotoxicity assays of G. lamblia trophozoites WB/1267, GS/H7, WB/1267 transfected with H7-HA or WB/1267 transfected with ΔR-H7 (from left to right on x axis) analyzed after 24 hours post-addition of the anti-VSPH7-specific antibodies G10-4 and anti-GS-VSPs pAb by estimating the number of adherent viable parasites. Controls include cells without treatment (w/o mAb) and the use of a non-related mAb (8G8 mAb). Data represent the means ± s.d. for n=2 of three independent experiments. (B) Progenies were analyzed by addition of goat anti-mouse FITC-conjugated antibody in IFA. Positive cells correspond to trophozoites expressing VSPH7. DIC, differential interference contrast. Scale bars: 10 μm. (C) VSP expression is established in WB/9B10 wild-type and WB/9B10-ADI+ transgenic trophozoites after a short time exposure to specific VSP9B10 mAb. Controls had no exposure to the mAb (–). Data represent the means ± s.d. for n=3 of two independent experiments. (D) VSP9B10 citrullination is analyzed by western blotting in WB/9B10 and WB/9B10-ADI+ trophozoites after a short time exposure to the anti-VSP9B10 mAb. The membrane labeled with anti-citrulline pAb (right) was stripped and re-blotted with anti-VSP9B10 mAb (left). A similar amount of loaded VSP9B10 is observed. Molecular mass of VSP9B10 is indicated on the right.

To determine whether citrullination of VSPs abrogates antibody-mediated cytotoxicity, we used specific VSPH7 Abs and WB/1267 transfected trophozoites, which expressed ∼100% VSPH7 mutants (Fig. 1C). Controls for cytotoxicity effects included a GS/H7 clone expressing ∼100% VSPH7 (+ control), and wild-type WB/1267 expressing VSP1267 and a non-specific antibody (8G8 mAb) (–controls). An equal surface expression of H7-HA and ΔR-H7 was controlled by selecting trophozoites with the same level of surface fluorescence using a fluorescence-activated cell sorter (FACS) and anti-VSPH7 antibodies (data not shown). Fig. 3A shows that, after addition of anti-VSPH7 mAb (G10/4) or pAb (anti-GS/VSP), immobilization, aggregation, detachment and complement-independent cytotoxicity was observed for the native GS/H7 clone and for H7-HA-transgenic trophozoites. These demonstrated cell-survival rates lower than 5% after 24 hours. By contrast, recovery of the non-citrullinated ΔR-H7 mutant expressing trophozoites exposed to the anti-VSPH7 Abs showed no cytotoxicity after 24 hours compared to controls. As expected, WB/1267 trophozoites exposed to anti-VSPH7 Abs and the samples exposed to 8G8 mAb showed no cytotoxicity (Fig. 3A).

Analysis of the progeny revealed that the surviving GS/H7 (not shown) and H7-HA trophozoites were nearly all VSPH7 negative, whereas the surviving ΔR-H7 trophozoites were almost 100% VSPH7 positive (Fig. 3B). These findings strongly suggest that, similar to palmitoylation, VSP citrullination plays an essential role in G. lamblia survival (Touz et al., 2005) (this study).

Because an inability to citrullinate VSPs results in decreased antigenic switching under Ab pressure, increased citrullination might result in enhanced antigenic switching in the same situation. To test this hypothesis, wild-type isolate WB clone 9B10 trophozoites (WB/9B10) and WB/9B10 transgenic trophozoites overexpressing active ADI (WB/9B10-ADI+) were exposed to the specific anti-VSP9B10 mAb (not used for controls) for only 10 minutes, with the cytotoxicity as well the VSP switching being analyzed at 24 hours. As expected, no cytotoxicity was observed. However, a significant increase of VSP switching was seen in the transfected trophozoites, with this being accompanied by enhanced citrullination (Fig. 3C,D). Similar results were also obtained when wild type (WB/1267wt) and WB/1267-ADI+ were tested (not shown). The VSP switching was relatively frequent and even small populations of recently cloned organisms contained parasites expressing different VSPs. Related to this, we observed that trophozoites that highly expressed ADI switched faster than wild-type cells under antibody pressure. These results could not be explained by differences in growth rates because transfected and non-transfected trophozoites multiply at the same speed.

ADI is post-translationally modified

ADI breakdown products were reported previously (Palm et al., 2003). In this work, analysis of protein expression of either ADI-HA or the native ADI showed that this protein is highly unstable and/or post-translationally modified, because several bands were observed, including a major 85 kDa band instead of the predicted 64 kDa band (Fig. 4A). Analysis of ADI failed to find probable glycosylation, myristolation, prenylation or glycoinositolphospholipid-binding sites. However, phosphorylation was predicted in several positions and we also found that ADI possessed two motifs with a high probability of modification by the 11 kDa SUMO1 protein (small ubiquitin-related modifier, also called Sentrin) at positions K101 (LKYE) and K387 (IKAD). Western blot assays using anti-SUMO1 mAb detected an 85 kDa protein corresponding to the ADI higher band (Fig. 4B). In addition, anti-SUMO1 mAb was able to immunoprecipitate the 85 kDa band of ADI, thus confirming the ADI-SUMO1 interaction (Fig. 4C). These results suggest that G. lamblia SUMO1 binds to ADI, probably to the K101 and K387 residues, thereby explaining why the ADI molecular mass increased from the predicted 64 kDa to 85 kDa. Nevertheless, the role of sumoylation on the modification of ADI and the biological function of this modification during G. lamblia growth and differentiation are the subject of on-going experiments.

ADI might play an additional role during G. lamblia encystation

Post-translational modifications of proteins occur and are important in many cellular processes, including cell-cycle progression, apoptosis, cellular proliferation and development, and have more recently been found in cell differentiation (Dohmen, 2004; Leight et al., 2005; Poulin et al., 2005; Shalizi et al., 2006). These findings prompted us to investigate whether ADI is involved in the process of encystation, the developmental process that a trophozoite goes through to turn it into a cyst. This process is initiated when trophozoites are deprived of cholesterol or challenged with high bile concentrations, resulting in the appearance of large secretory granules called encystation-specific secretory vesicles (ESVs). These vesicles transport cyst-wall materials (e.g. CWP1 and CWP2) for the subsequent release and extracellular assembly of the rigid cyst wall that protects the parasite outside its host (Fig. 5A). First, we characterized ADI by analyzing the level of adi mRNA and protein expression in both wild-type (WB/1267wt) and ADI-transgenic (WB/1267-ADI+) trophozoites at 0, 6, 24 and 48 hours of encystation. Slot-blot assays showed that there was a minor increase in adi mRNA during encystation, compared with cwp2 and the house-keeping gdh, in WB/1267wt (Fig. 5B, left panel). By contrast, a higher expression of adi mRNA and a lower expression of cwp2 mRNA were observed, whereas gdh mRNA levels remained equal during encystation of WB/1267-ADI+ cells (Fig. 5B, right panel).

Fig. 4.

Post-translational modifications of ADI. (A) Western blotting using anti-HA mAb shows ADI-HA in ADI-transgenic trophozoites. Multiple bands are also obtained using specific anti-ADI pAb in wild-type cells. The band of 64-66 kDa corresponding to the predicted protein sequences is observed in both cases (arrowhead), together with degradation products (gray arrows) including the low-weight band (gray arrow with asterisk) found in the peptide pull-down (see Fig. 1). Also, for both HA-tagged and native ADI, an increase in molecular mass from a 64-66 kDa to a 85 kDa band is shown (black arrow), suggesting that ADI undergoes post-translational modification. STD, molecular weight standard. (B) Western blot assays using anti-SUMO1 mAb recognize an ∼85 kDa band (arrow) in both WB/1267 (WB) and GS/H7 (GS) G. lamblia clones. The same filter membrane was stripped and re-blotted with anti-ADI pAb, showing a perfect match with the higher band that is recognized by the anti-SUMO1 mAb. To confirm the lack of residual primary antibodies after stripping, only the secondary antibody was added to the stripped blots, showing no signal (Control). (C) Western blotting using biotin-conjugated anti-ADI pAb was performed to detect ADI bands after immunoprecipitation with anti-SUMO1 mAb. (a) ADI in lysate before IPP; (b) ADI after immunoprecipitation by using 0.1 μg of anti-SUMO1 mAb (arrow); (c) ADI after immunoprecipitation by using 1 μg of anti-SUMO1 mAb (arrow); (d) control using a non-related anti-HA mAb; and (e) supernatant of sample c.

Semi-quantitative reverse transcriptase (RT)-PCR assays corroborated the data obtained by slot-blotting. gdh mRNA expression showed no variation during encystation for either WB/1267wt or WB/1267-ADI+ trophozoites. Furthermore, no variation in mRNA levels was detected for adi during encystation in either case. However, its expression was found to be 2.5-times higher in transgenic cells. cwp2 mRNA dramatically increased at 24 hours in wild-type cells, but not in ADI-transgenic trophozoites, suggesting that ADI inhibited cwp2 expression and therefore might participate in the regulation of the cwp2 gene (Fig. 5C).

Fig. 5.

ADI expression increases during G. lamblia differentiation. (A) IFA and confocal microscopy show the G. lamblia encystation process. CWP2 (red) is synthesized and transported in ESVs in encysting trophozoites (ET, arrowhead). At the end of the encystation process, CWP2 is found in mature cyst walls (Cyst). Nuclei are stained with DAPI (blue). Scale bars: 10 μm. (B) Slot-blotting qualitatively shows gdh, adi and cwp2 gene expression at 0, 6, 24 and 48 hours of encystation in wild-type cells (WB/1267wt) and transgenic trophozoites (WB/1267-ADI+). The assay was performed in triplicate. (C) Analysis of gdh, adi and cwp2 gene expression by RT-PCR and optical density quantified by densiometry comparing both wild-type (WB/1267wt) and ADI-transgenic (WB/1267-ADI+) trophozoites. Data represent the means ± s.d. for n=4 of two independent experiments. (D) Western blotting using specific antibodies shows ADI, modified citrulline (Cit), VSP1267 and CWP2 protein expression in both wild-type (WB/1267wt) and ADI-transgenic (WB/1267-ADI+) trophozoites. Non-encysting trophozoites (NT, 10 μg) and 24-hour-encysting trophozoites (ET, 10 μg) were used for each sample.

Western blotting using specific antibodies showed that, in WB/1267wt and WB/1267-ADI+ trophozoites, ADI levels increase after 24 hours of encystation. The expression of VSP1267 was equivalent in both cell types (VSP1267) but the amount of citrulline detected in WB/1267-ADI+ trophozoites was higher than in WB/1267wt trophozoites (also demonstrated in Fig. 3D). Surprisingly, although ADI expression increased, VSP1267 citrullination was not modified during encystation (Fig. 5D, Cit). As expected, CWP2 rose in WB/1267wt trophozoites during encystation (Fig. 5D, left panel). Conversely, there was a decrease in CWP expression to almost undetectable levels in WB/1267-ADI+ trophozoites, validating a role of ADI in the regulation of CWP2 (Fig. 5D, right panel). Generally, WB/1267 trophozoites differentiate into cysts in about 4-8% of total cells after 48 hours in encysting medium. Our results showed that overexpression of ADI reduced the number of cysts produced (approximately 0.1%), compared with wild type (not shown). Even though the amount of formed cyst is drastically reduced in transgenic trophozoites that highly express ADI, in cells expressing ADI at low levels, the kinetics of differentiation is unaltered and the formed cysts are resistant to harsh environmental conditions (C.V.V. and A.S.R., unpublished results).

Because Lys residues are also the target for ubiquitylation, it is possible that ADI sumoylation functions as an inhibitor of ADI proteolysis during differentiation. Another possibility is that the addition of SUMO1 proteins to ADI allows cytoplasm-nuclei translocation (Ishov et al., 1999). To test this last hypothesis, the presence of sumoylated ADI was analyzed in nuclear and cytoplasmic fractions both before and during encystation, by western blotting. In wild-type trophozoites, the ADI 85 kDa band increased in the nuclear fraction during encystation (Fig. 6A). To test for cytoplasmic contamination of nuclear fractions, the presence of VSP1267, a protein not found in the nucleus, was investigated by western blotting and was undetected in these fractions (Fig. 6A).

Confirmation of the nuclear translocation of ADI was obtained by IFA in WB/1267wt (Fig. 6A,B) and ADI-HA-transfected trophozoites (Fig. 6C), both demonstrating a shift of ADI to the nuclei during encystation. Wild-type organisms only showed the presence of nuclear ADI late during encystation (when the cells contained a large number of ESVs) (Fig. 6A,B). Interestingly, those transgenic organisms showing a clear translocation of ADI-HA to the nuclei had few, if any, ESVs (Fig. 6C). The failure of ESV development together with the decrease in cwp2 mRNA in transfected encysting cells suggest that ADI has a role in the control of the encystation process.


Previous studies determined that ADI plays a role in the energy metabolism in G. lamblia. In the present work, we show an additional ADI action that affects or controls essential biological activities of the organism. There are four major findings. First, we establish that ADI acts as a PAD in addition to its known function as a metabolic enzyme. Second, ADI influences important processes in G. lamblia. Specifically, by way of its ability to citrullinate VSPs, ADI alters the VSP biology, the response to cytotoxic antibodies and the antigenic variation. Moreover, ADI appears to play an important, if not controlling, role in encystation. Lastly, because ADI is sumoylated, it is likely that this modification alters its activity, degradation and/or localization.

Fig. 6.

During encystation, ADI is translocated to the nuclei and inhibits CWP expression. (A) Results of western blotting that was performed after cytoplasm- and nuclear-fractionation assays using anti-ADI pAb at 24 and 48 hours post-encystation induction in wild-type (ADI) cells. L, cell lysate previous fractionation; C, cytoplasmic fraction; N, nuclear fraction (upper panel). Middle panel: anti-VSP1267 mAb is used to detect cytoplasm- or nuclear-fraction contamination. Lower panel: a representative time course of IFA shows ADI (red) distribution during encystation (ESVs in green). (B) Confocal microscopy and IFA using wild-type cells show that ADI (red) is translocated to the nuclei when the encysting cell is filled with ESVs (arrowheads). CWP1 is stained in green. (C) Top panels: representative differential interference contrast (DIC) and DAPI-staining images show a stable ADI-transgenic culture (Merge). Middle panels: direct IFA shows that trophozoites highly overexpressing ADI-HA (green) in the nuclei do not express CWP2 (red) during encystation (arrowheads in Merge). Bottom panels: closer analysis of three ADI-transgenic cells demonstrated that those expressing ADI-HA (green) in the nuclei (arrowheads in Merge) do not reveal CWP2 (red) expression after 24 hours of encystation, whereas those with low ADI-HA expression do (arrow in Merge). Scale bars: 10 μm.

The evidence supporting that VSPs are citrullinated by ADI is compelling. ADI binds specifically to the CRGKA tail and localizes close to the VSPs found on the plasma membrane of the trophozoite. Using specific antibodies to modified citrulline, VSPs were shown to be citrullinated specifically, at the arginine residue (R) of the CRGKA cytoplasmic tail. Also, purified ADI citrullinates the arginine in the conserved tail in vitro, verifying the PAD activity of ADI. This modification has profound effects on the biology of the parasite.

The most commonly cited biological role of antigenic variation in pathogenic microorganisms is immunological escape, in which the host antibodies produced against a dominant antigen destroy those organisms bearing it, resulting in the organisms being replaced by ones that possess a variant form of the antigen. However, the function of antigenic variation among organisms differs and, in some cases, its analysis is complex. In G. lamblia, antibodies directed against VSPs – including VSP-specific surface-reacting mAbs (Nash and Aggarwal, 1986), immune lactogenic IgA (Stager et al., 1998) and serum from infected humans (Nash et al., 1990a; Nash et al., 1990b) or animals – either inhibit the growth of or kill recognized trophozoites (Hemphill et al., 1996; Stager and Muller, 1997), thus allowing the repopulation of trophozoites expressing other VSPs. The in vitro effects range from little or no inhibition of growth to cytotoxicity, and the effects appear to depend on the concentration (direct correlation), affinity, and even perhaps the nature of the antibody and target epitopes (Hemphill et al., 1996; Nash and Aggarwal, 1986; Stager and Muller, 1997). By contrast, monovalent F(ab′) of the same antibodies exhibited none of these cytological effects (Hemphill et al., 1996; Nash and Aggarwal, 1986), although the VSP switching was not analyzed in these instances

Our results support the findings that exposure of trophozoites to a high level of specific VSP Abs results in cell death and the emergence of trophozoites expressing an antigenically different VSP. Most significantly, this event is strongly linked to deimination (citrullination) of the cytoplasmic tail of VSPs, because mutation of the amino acid arginine led to survival of targeted trophozoites and failure to switch. Conversely, we have now shown that an increase in VSP citrullination by overexpression of ADI in the presence of the specific Ab causes deregulation of VSP switching, probably due to an amplification of a signal-transduction event (Touz et al., 2005) (this study). Therefore, it is now clear that post-translational modifications are important for the control of the antigenic variation in G. lamblia. These studies are consistent with the hypothesis that both immunological and biological factors act in concert to select which VSPs are expressed in a particular host.

The functional significance of ADI in cell survival appears not to be restricted to its role in energy production and antigenic variation. We found that the subcellular localization of ADI in G. lamblia is cytoplasmic and is significantly close to the plasma membrane but that it is upregulated and translocated to the nuclei when the trophozoites are induced to encyst. At least two possibilities arise from this event: (1) as an arginine deiminase, ADI might be sequestered from the cytoplasm to enter in a `stand-by' process during encystation, because the requirement of energy during this process is lower than that needed during active growth; (2) as a PAD, ADI might be directed into the nuclei for histone modification and transcription regulation (Wang et al., 2004). However, the results presented in this work suggest that the second possibility is more likely. Post-translational histone modifications, such as phosphorylation, acetylation, methylation and citrullination, regulate a broad range of DNA- and chromatin-templated nuclear events, including transcription (Jenuwein and Allis, 2001; Wang et al., 2004). Because acetylation was shown to be a G. lamblia histone post-translational modification (Kulakova et al., 2006), we propose that ADI modifies histones by citrullination, with this post-translational modification being involved in the downregulation of the encystation process. This hypothesis is based on the fact that translocation of the functionally overexpressed ADI to the nuclei avoids cyst formation. We also suggest that ADI causes genes in the CWP family to turn off as an essential requirement to successfully complete the encystation. This effect takes place early on in ADI-transgenic cells, in which the high overexpression of ADI together with its nuclear translocation avoids CWP expression and cyst formation. Nevertheless, we cannot exclude the possibility that ADI is also involved in the control of its major surface antigen when its life cycle is completed, as was demonstrated by Svard et al. (Svard et al., 1998). Differentiation-associated switching of surface antigens could be one reason for the common occurrence of repeated infections (Gilman et al., 1988) due to the downregulation of the major VSP expressed during encystation and the appearance of new VSPs after excystation.

ADI possesses two predicted nuclear localization signals (PQRRREQ and RRGIVMGQFQAPQRRRE) (SignalIP program) (Le Panse et al., 1997), suggesting that these motifs are involved in the translocation of this protein to nuclei during encystation. There is, however, no evidence of the participation of these motifs in the nuclear protein translocation of any protein in this parasite. Thus, further analyses are needed to show how this parasite uses the nuclear signaling motifs that are highly conserved during evolution. Another possibility is that ADI is translocated to the nuclei by means of SUMO1-protein binding. Recently, it was shown that conjugation of SUMO1 could lead to protein stabilization and protection from degradation, but an increasing body of evidence implicates sumoylation in the targeting of certain proteins to the cell nucleus and to the subnuclear structures (Dohmen, 2004). The underlying mechanisms, however, are still largely unknown. In G. lamblia, the molecular biology of enzymes of the SUMO system has never been addressed. In fact, only three putative proteins involved in sumoylation have been posted in the Giardia database: SUMO1 (or Sentrin, GGD 7760), the dimer of the SUMO-activating enzyme E1 (Uba2, GGD 6288) and the SUMO ligase (GGD 17386). Our results from `in silico' data, western blotting and immunoprecipitation using anti-SUMO1 mAb indicate that ADI is a sumoylated protein. This protein modification might be essential in the nuclear translocation of ADI in order to accomplish its function during encystation, because the 85 kDa ADI was mainly located in the nuclear subfractions. Further research into the possibility that ADI function could be regulated by sumoylation might provide insight into the biology of this important intestinal parasite and also contribute to the understanding of the evolution of protein modification in eukaryotic cells. Taking earlier studies and this work together, we have been able to unravel the multiple functions of ADI in the survival of G. lamblia (Fig. 7). One of these concerns is the ability to obtain energy by using free arginine as the preferential fuel under anaerobic conditions. During the proliferative stages of growth, ADI converts arginine into citrulline, with ATP production occurring at the final enzymatic step of the ADH pathway (Schofield et al., 1992) (Fig. 7A). In addition to this function, it was reported that G. lamblia uses ADI as a competitor to NOS from the host cell for the free arginine, thereby reducing the production of NO and the host defense mechanism against microbial infection (Eckmann et al., 2000). Supporting this finding, it was confirmed that ADI is released from G. lamblia to the extracellular space as a 66 kDa protein when the trophozoites are in contact with human colon epithelial cells (Ringqvist et al., 2008) (Fig. 7B). Also, during growth, ADI acts as a PAD on the cytoplasmic tail of VSPs, probably by functioning as a regulator of antigenic variation (Fig. 7C). During encystation, ADI is located in the nuclei and cyst formation is reduced (Fig. 7D). This translocation event is most probably associated with ADI sumoylation, but the possibility of the participation of ADI nuclear-localization signals cannot been discarded.

Fig. 7.

Schematic representation of the functions of ADI during growth and encystation. (A) During growth, the trophozoites acquire free arginine from the extracellular medium. Inside the cell, cytoplasmic ADI converts arginine into citrulline, with ATP production occurring at the final enzymatic step of the ADH pathway. (B) ADI is released to the extracellular space when the trophozoites are in contact with human colon epithelial cells and compete with the host NOS for the free arginine, thereby reducing the production of NO. (C) Under low exposure to specific antibodies, ADI acts as a PAD on the cytoplasmic tail of VSP, inducing VSP switching. (D) During the last step of encystation, ADI is translocated from the cytoplasm to the nuclei, turning encystation-specific genes off and ending the process.

Changes in the environment (anaerobiosis, the presence of Abs, the presence of the host cell or depletion of cholesterol) define which functions of ADI need to be performed at one point in time. The underlying mechanisms, however, are still largely unknown. We believe that the analysis of the protein modifications associated with differentiation, along with the mediators of these activities reported here, represent an important contribution to our understanding of the control of gene expression in parasitic protozoa.

Materials and Methods

G. lamblia cultivation and transfection

Trophozoites of the isolate WB, clone 1267 (WB/1267), 9B10 (WB/9B10) and from the isolate GS, clone H7 (GS/H7) were cultured in TYI-S-33 supplemented with 10% fetal bovine serum and bovine bile (Keister, 1983). Trophozoites of the WB/1267 clone were transfected by electroporation and selected with puromycin (Singer et al., 1998; Yee and Nash, 1995). The transfection of stable WB/1267 cells was 100%, as determined by IFA and flow cytometry. A two-step encystation procedure was used by increasing the medium pH and by addition of porcine bile as reported by Boucher and Gillin (Boucher and Gillin, 1990).

Pull-down assays

G. lamblia clones GS/H7 and WB/1267 were grown, harvested and suspended in 1 ml of lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazol, pH 8.0, 1% Triton X-100, and protease inhibitors) for 3 hours at 4°C. After mild sonication using a Branson sonifier 250 (Branson, CT) with an output control of 3 and a 50% duty cycle (sonication complex), the lysate was centrifuged at 7500 g, for 30 minutes at 4°C. The supernatant was then mixed with 1 mg of His6-CRGKA peptide dissolved in 10 μl of DMSO overnight at 4°C. Controls of specific binding had no His6-CRGKA peptide. The lysate and peptide were mixed with 200 μl of Ni-agarose beads (QIAGEN, Valencia, CA) and were incubated for 4 hours at 4°C. Beads were spun down at 700 g and washed four times with wash buffer (50 mM NaH2PO4, 300 mM NaCl, pH 8.0, 0.1% Triton X-100, and protease inhibitors). Bound proteins were eluted four times with 100 μl elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazol, pH 8.0, 0.1% Triton X-100, and protease inhibitors). Peptide-bound proteins were analyzed by SDS-PAGE, stained with Coomassie G-250 (Fig. 1A). The detected bands were cut out and submitted to the Research Technologies Branch for Protein Identification (NIAID, NIH) for LC-MS/MS analysis. After two independent experiments, one protein that was associated with the His6-CRGKA peptide was identified.

Yeast two-hybrid assay

The MATCHMAKER two-hybrid system was used following the protocol suggested by the company (Clontech, Palo Alto, CA). The two-hybrid pGBKT7(TRP1) vector (GAL4 DNA-binding domain, BD) containing the genes for vsph7, vsp1267 or vsp1267 without the 3′ (TGCAGAGGCAAGGCG) gene sequence were used as bait while the adi gene was ligated to the pGADT7-Rec(LEU2) vector (GAL4 transcription activation domain, AD), yielding H7-BD, 1267-BD, 1267-tail-BD and ADI-AD vectors, respectively. AH109 transformants were cultured at 30°C for 4-5 days on plates with minimal medium lacking leucine and tryptophan (–L/–T) to test for positive transformation, or in the absence of leucine, tryptophan and histidine (TDO – triple dropout medium) to study specific protein interactions as previously described (Touz et al., 2004). Controls of the methodology include ESCP-BD–MuA-AD interaction (Touz et al., 2004) and the emptyBD–ADI-AD vector.

Expression of VSPH7 variant and ADI-HA in WB/1267 trophozoites

The construction of VSPH7 and H7-tail was detailed previously (Touz et al., 2005). ΔR-H7 was prepared using a site-directed mutagenesis kit (QuikChange, Stratagene) and the presence of the mutation was confirmed by using dye terminator cycle sequencing (Beckman Coulter). For ADI-HA construction, the vsph7 gene was replaced by the adi gene using the forward 5′-GCCTCCATGGCTGACTTCTCCAAGGATAAAGAG-3′ and 5′-GATGGTTAACCTTGATATCGACGCAGATGTCAGC-3′ oligonucleotides yielding the ADI-HA vector. ADI-HA, H7-HA, H7-tail and ΔR-H7 vectors were used to transfect the WB/1267 clone in order to generate stable puromycin-selected cells.


Anti-HA mAb (Sigma), anti-HA mAb labeled with alkaline phosphatase (Sigma), anti-ADI pAb (Ringqvist et al., 2008), anti-H6 mAb (Invitrogen), anti-CWP1 FITC-labeled mAb (Warerborne, New Orleans, LA) and CWP2–Texas-Red (Touz et al., 2003) were used. Citrullinated proteins were detected by using a polyclonal antibody against a chemically modified version of citrulline (anti-MC) (Senshu et al., 1992). To detect sumoylated proteins, mouse anti-GMP1 (anti-SUMO1) mAb (ZYMED Laboratories, Invitrogen) was used. The mAbs 5C1, G10/4 and 9B10 were used to detect VSP1267, VSPH7 and VSP9B10, respectively (Mowatt et al., 1991; Nash et al., 2001; Nash and Mowatt, 1992). Anti-ADI pAb was biotinylated by using the EZ-Link NHS-SS-Biotin (Pierce, Rockford, IL) and detected in western blotting by using alkaline-phosphatase-labeled streptavidin. For direct IFA, anti-HA FITC-labeled mAb (Sigma) VSP9B10–Texas-Red and anti-CWP2–Texas-Red (Touz et al., 2003) were used.

Western blotting and IFA

Total protein (10 μg) from parasite lysates was incubated with sample buffer, boiled for 10 minutes and separated on 10% Bistris gels. Samples were transferred to nitrocellulose membranes, blocked with 5% skimmed milk and 0.1% Tween 20 in TBS, and incubated with primary antibody diluted in the same buffer. After incubation with conjugated secondary antibody, proteins recognized by antibodies were visualized with SuperSignal West Pico chemiluminescent substrate (Pierce) and autoradiography, or by using BCIP/NBT substrate (Sigma). Controls included omission of the primary antibody, use of an unrelated antibody or assays using non-transfected cells. In Fig. 3D, the AutoDeblur v9.3 program was used (AutoDeblur v9.3, AutoQuant Imaging, NY).

For IFA of fixed cells, trophozoites cultured in growth medium were harvested and processed as described (Touz et al., 2005). Briefly, parasites were attached to slides for 30 minutes and then fixed with fresh 4% formaldehyde for 40 minutes. The cells were incubated sequentially with blocking solution (10% goat serum and 0.1% Triton X-100 in PBS) at 37°C for 30 minutes and antibodies conjugated with fluorescent dyes for 1 hour. The cells were then washed three times with PBS before the addition of mounting medium (VECTASHIELD Mounting Medium). Images were collected on a Leica TCS-NT/SP confocal microscope (Leica Microsystems, Exton, PA) using 40× or 63× oil-immersion objectives (NA 1.32, zoom X). Fluorochromes were excited using an argon laser at 488 nm for FITC and a krypton laser at 568 nm for Texas Red 568. DAPI was excited using a 364 nm argon laser. Detector slits were configured to minimize any crosstalk between the channels. Differential interference contrast (DIC) images were collected simultaneously with the fluorescence images by use of a transmitted-light detector. Images were processed using Leica TCS-NT/SP (version 1.6.587), Imaris 3.1.1 (Bitplane AG, Zurich, Switzerland) and Adobe Photoshop 5.5 (Adobe Systems) software.

Immunoprecipitation assay (IPP)

Cultured G. lamblia trophozoites were harvested and resuspended in 1 ml of lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 1% Triton X-100, and protease inhibitors) for 1 hour at 4°C. The lysate was centrifuged at 10,000 g for 10 minutes at 4°C, and the supernatant mixed with anti-ADI, anti-HA (1 μg) or anti-SUMO1 (0.2 or 1 μg) and incubated overnight at 4°C. Protein A-agarose beads (50 μl; Qiagen, Valencia, CA) was added to each sample before incubation for 4 hours at 4°C. Beads were pelleted at 700 g and washed four times with wash buffer (50 mM NaH2PO4, 300 mM NaCl, pH 8.0, 0.1% Triton X-100, and protease inhibitors). Beads were suspended in sample buffer and boiled for 10 minutes before western blot analysis.

Citrullination assay

Non-encysting and trophozoites encysted at different time points were collected, washed in PBS, resuspended in 20 μl of lysis buffer without protease inhibitors (50 mM Tris, pH 8.0, 120 nM NaCl, 1% Triton X-100), and incubated with 5 mM dithiothreitol for 30 minutes at 37°C prior to being subjected to SDS-PAGE and nitrocellulose blotting using the anti-citrulline (modified) detection kit (Upstate) as was previously described (Senshu et al., 1995). Controls included omission of the anti-citrulline antibody or absence of the chemically modifying citrulline step.

ADI-activity assay

ADI activity was assayed by measuring the modification of His6-CRGKA peptide to His6-CcitGKA. First, trophozoites expressing ADI-HA or ESCP-HA (Touz et al., 2004) were cultured, collected and sonicated in PBS buffer at 4°C until there were no intact cells before being centrifuged for 10 minutes at 700 g. ADI-HA and ESCP-HA were purified using anti-HA mAb and protein-A/G–agarose as previously described (Touz et al., 2004). 5 mM of soluble His6-CRGKA peptide (10 mg/ml of ADI activity buffer: 50 mM Tris-HCl pH 7.6, 2 mM DTT and 5 mM CaCl2) was mixed with either purified ADI-HA or ESCP-HA for 16 hours at 50°C. After samples were boiled for 5 minutes, the peptide was purified by using Ni-agarose beads, and was subjected to dot-blot and citrullination assays.

Cytotoxicity and switching assays

Cytotoxicity was assayed as we previously reported (Touz et al., 2005). Briefly, G. lamblia trophozoites expressing native VSPH7 or transgenic VSPH7 at the same intensity as surface VSPH7 were selected and harvested using CELLQuest software and a FACScalibur flow cytometer (Becton Dickinson, San Jose, CA). These clones were grown until confluence before a second IFA using G10-4 mAb was performed to test for 100% expression of VSPH7. A WB/1267 clone expressing almost 100% VSP1267 was used as a negative control. After addition of G10-4 mAb (specific for VSPH7), mouse polyclonal antiserum (anti-GS/VSP pAb, raised against GS/H7 Giardia clone) or anti-CWP2-8G8 mAb (generated against ESVs) (Lujan et al., 1995) in 1:20 dilution, we analyzed immobilization, aggregation, detachment and complement-independent cytotoxicity. Cytotoxicity was determined in triplicate after 24 hours by estimating the number of adherent viable parasites. The progeny were analyzed by growing the assayed trophozoites for 24 hours and by labeling the remaining H7-positive viable trophozoites by using goat anti-mouse FITC-conjugated antibody (Cappel). For the VSP switching assay, Giardia ADI-transgenic and non-transgenic parasites expressing approximately 100% VSP9B10 or VSP1267 were maintained in vitro before the addition of anti-VSP9B10 or anti-VSP1267 (5C1) mAbs, respectively, for only 10 minutes in a 1:20 dilution. VSP expression was analyzed by FACS as previously described (Nash et al., 2001). Controls consisted of identical cultures not exposed to mAb.

Slot-blot assay

Total RNA (2 μg) extracted from growing and encysting trophozoites was immobilized onto a Nytran SuperCharge nylon membrane (Schleicher & Schuell, Keene, NH) by using the MINIFOLD II slot-blot system, according to the manufacturer's instructions. Membranes were first hybridized with antisense probes specific for the adi gene, and then stripped and reprobed by using an antisense specific for gdh or cwp2 as the controls. Oligonucleotides used: GDH (forward 5′-ATGCCTGCCCAGACGATCGA-3′, reverse 5′-GAGCCAGAAAGAAGGACGTT-3′), ADI (forward 5′-CACTTGTGGAAATTACGTCT-3′, reverse 5′-CTTGATATCGACGCAGATGT-3′), ADI-HA (forward 5′-CACTTGTGGAAATTACGTCT-3′, reverse 5′-CTATGCATAGTCTGGTACAT-3′) and CWP2 (forward 5′-ATGATCGCAGCCCTTGTTCT-3′, reverse 5′-ATCCATCTCTCTCGAGAGTT-3′).

Semi-quantitative RT-PCR

The total RNA from wild-type and ADI-transgenic trophozoites was isolated using RNA STAT-60 reagent (Tel-Test, Friendswood, TX) followed by digestion with 50 U DNAse (Roche Diagnostics, Indianapolis, IN) for 15 minutes at 25°C, and a final purification was performed by using SV Total RNA Isolation System (Promega, Madison, WI). For reverse transcription and PCR amplification, a one-step RT-PCR kit (Qiagen, Valencia, CA) was used with dilutions of the total RNA from 0.2 ng to 20 ng per reaction in a final reaction volume of 50 μl. The reverse-transcription reaction was done at 50°C for 30 minutes followed by inactivation of the reverse transcriptase at 95°C for 15 minutes. For PCR, 30 cycles (30 seconds at 94°C, 30 seconds at 50°C and 1 minute at 72°C) were used followed by 10 minutes at 72°C for final extension. DNA-contamination control was carried out by adding ADI oligonucleotides at the PCR step of the RT-PCR reaction. Aliquots (5 μl) of the RT-PCR reaction were size-separated on 1.2% agarose gel in TAE (E-Gel, Invitrogen Corporation, Carlsbad, CA) pre-stained with ethidium bromide. Amplification products were directly quantified by densitometric scanning of the fluorescence intensity under UV light, using EagleSight software for image acquisition, documentation and analysis (Stratagene, La Jolla, CA). The oligonucleotides used were the same as the ones described for slot-blotting.

Subcellular fractionation

Non-encysting and encysting Giardia trophozoites were suspended in buffer A (20 mM Tris-HCl, pH 7.6, 5 mM MgCl2, 1.5 mM KCl, 1 mM phenylmethanesulfonyl fluoride, 2 mM dithiothreitol and 0.1% Nonidet P-40) and disrupted with sonication. The homogenate was centrifuged at 760 g at 4°C for 10 minutes to separate it into supernatant (cytoplasmic) and pellet (nuclear) fractions. The supernatant and pellet fractions were subjected to western blotting.


Adrian Hehl is acknowledged for his collaboration in the production of anti-ADI pAb, and Alfredo Caceres and laboratory members for providing numerous reagents. This work was partially supported by the Agencia Nacional para la Promoción de la Ciencia y Tecnología FONCyT Jóvenes PICT2004. This research was also supported in part by the Intramural Research Program of the National Institutes of Allergy and Infectious Diseases, National Institutes of Health.


  • * These authors contributed equally to this work

  • Accepted June 16, 2008.


View Abstract