Experimental Parasitology 119 (2008) 364–372

Contents lists available at ScienceDirect

Experimental Parasitology

journal homepage: www.elsevier .com/ locate /yexpr

Giardia duodenalis: Adhesion-deficient clones have reduced ability to establishinfection in Mongolian gerbils

Javier Hernández-Sánchez a, Rocío Fonseca Liñan a, María del Rosario Salinas-Tobón b,Guadalupe Ortega-Pierres a,*

a Departamento de Genética y Biología Molecular, Centro de Investigación y de Estudios Avanzados del IPN, Ave. IPN 2508, C.P. 07360 Mexico, DF, Mexicob Departamento de Inmunologia, Escuela Nacional de Ciencias Biológicas, Instituto Politécnico Nacional Prolongación de Carpio y Plan de Ayala S/N, C.P. 11340 Mexico, DF, Mexico

a r t i c l e i n f o

Article history:Received 14 November 2007Received in revised form 20 March 2008Accepted 22 March 2008Available online 29 March 2008

Index Descriptors and Abbreviations:Giardia duodenalisAdhesion-deficient clonesSurface proteinsMonoclonal antibodiesInfectivityMongolian gerbils

0014-4894/$ - see front matter � 2008 Elsevier Inc. Adoi:10.1016/j.exppara.2008.03.010

* Corresponding author. Fax: +52 55 5061 3931.E-mail address: gortega@cinvestav.mx (G. Ortega-

a b s t r a c t

The role of Giardia duodenalis surface molecules in the attachment of trophozoites to epithelial cells hasbeen established through the dual strategies of characterizing G. duodenalis clones with deficient adhesionand blocking experiments with surface-specific monoclonal antibodies. Also, the infectivity of the ana-lyzed clones was tested using Mongolian gerbils as experimental model. Two adhesion-deficient G. duo-denalis clones, C6 and C7, were isolated from the wild type C5 clone which in turn was obtained fromthe WB strain. The adhesion efficiencies of C6 and C7 clones (48.2 ± 4.9 and 32.6 ± 2.4, respectively) weresignificantly lower as compared with WB strain or C5 clone (82.8 ± 6.4 and 79.9 ± 7.9). Analysis of radio-label surface proteins by 1D and 2D SDS–PAGE revealed prominently labelled 28 and 88 kDa componentsin C6 and C7 clones and a major 200 kDa protein in the C5 clone and the WB strain. The 88 and 200 kDacomponents are acidic proteins by two-dimensional electrophoretic analyses. The most striking differencebetween wild-type and adhesion-deficient Giardia trophozoites was the reduced expression of a 200 kDasurface protein in the latter. Significantly, a mAb (IG3) specific for the 200 kDa protein that reacted withmore than 99% of WB and C5 trophozoites and less than 1% of C6 and C7 trophozoites as determined byindirect immunofluorescence inhibited the adhesion of trophozoites from WB and C5 clone to MadinDarby Canine Kidney cells by 52% and 40.9%, respectively, suggesting a participation of this antigen inadherence. Finally, the functional relevance of trophozoite adhesion to epithelial cells was indicated bythe reduced capacity of the adhesion-deficient clones to establish the infection in Mongolian gerbils.

� 2008 Elsevier Inc. All rights reserved.

1. Introduction

The flagellate parasite, Giardia duodenalis, is the most commonintestinal parasite in humans worldwide (Wolfe, 1992). In humanand animal giardiasis, a number of different mechanisms of patho-genesis have been proposed to explain the observed intestinal dys-function. These include malabsorption of glucose, sodium andwater and reduced epithelial absorptive surface area (revised inBuret, 2007). Also, in humans chronic giardiasis may cause hyperse-cretion of chloride (Troeger et al., 2007). In giardiasis, disruption ofcellular F-actin and tight junctional ZO-I together with the resultingincrease in transepithelial permeability seem to be modulated atleast in part by myosin-light-chain kinase and pro-apoptotic cas-pase 3 (Scott et al., 2002; Chin et al., 2002). Further, studies by Troe-ger et al. (2007) using TUNEL labelling reported that epithelialbarrier dysfunction in patients with choric giardiasis is associatedwith increased rates of enterocyte apoptosis. The loss of epithelialbarrier function will, in turn, allow the activation of host im-

ll rights reserved.

Pierres).

mune-dependent pathological pathways. A common factor in-cluded in most of the pathogenic mechanisms postulated is theattachment of trophozoites to the intestinal epithelial cells. Fur-thermore, adhesion of Giardia trophozoites to the intestinal epithe-lium is crucial to both the initial colonization and the maintenanceof the infection, since parasites that do not attach or move againstthe flow of intestinal fluid will be expelled (Crouch et al., 1991).Among the theories that have been proposed to explain G. duode-nalis attachment are (1) a suction force beneath the ventral discgenerated by the movement of ventral flagella and lateral crest(Holberton, 1974; Hansen et al., 2006), (2) a mechanical process re-lated to contractile protein elements of the ventral disc and ventro-lateral flange (Crossley and Holberton, 1983; Feely et al., 1982;Campanati et al., 2002) and (3) a combination of the two mecha-nisms (Sousa et al., 2001). More recently Erlandsen et al. (2004)suggested that the adhesive nature of the ventrolateral flange mightbe involved in the attachment of G. duodenalis trophozoites to asubstratum. However G. duodenalis, in common with some bacteriaand other protozoa, has surface molecules with lectin activity, aproperty that has been related to adherence with mammalianerythrocytes and rat epithelial cells (Farthing et al., 1986; Inge

J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372 365

et al., 1988; Magne et al., 1991; Sousa et al., 2001). In addition, sur-face molecules of 64, 66 and 88 kDa have been implicated in theattachment of G. duodenalis trophozoites to target cells in vitro (Ingeet al., 1988; Gillin et al., 1990). Moreover, the activity of certain se-creted proteinases explains a role in the trophozoite–epithelial cellinteraction (Rodríguez-Fuentes et al., 2006). Not surprisingly, sur-face molecules have also been widely implicated in the adhesionand pathogenesis of other protozoa as Entamoeba histolytica. Themajor adhesion molecule of this parasite is the immunodominantgalactose/N-acetylgalactosamine (Gal/GalNAc) inhibitable surfacelectin (McCoy et al., 1994; Petri et al., 2002). This lectin plays a cen-tral role in adhesion of the trophozoites to intestinal mucins and toenterocytes (Tse and Chadee, 1991) and is a major virulence factor(Petri et al., 2002). Further a coordinated action of E. histolytica cyto-skeleton and surface adhesion molecules explains the adhesion andpathogenesis of E. histolytica trophozoites (Tavares et al., 2000,2005). Likewise the importance of surface molecules in adherencehas been established for other parasitic protozoan through the iso-lation of adhesion and virulence deficient clones. For example,clones isolated from a single strain of Leishmania infantum dis-played significant differences in infectivity for mouse peritonealmacrophages which correlated with the ability to infect hamsters(Méndez et al., 2001). Adhesion-deficient Babesia bovis clones withvariable adhesion to cultured epithelial cells and virulence in Hol-stein cattle have also been isolated (Canto et al., 2006). In addition,the isolation of adhesion-deficient clones of E. histolytica has alsosuggested the dual importance of the cytoskeleton and surface mol-ecules in various processes essential for pathogenesis like adhesionto substrata (Gallegos et al., 1986; Rodriguez and Orozco, 1986; Ar-royo and Orozco, 1987). Thus, to study the relevance of surface mol-ecules in adhesion and pathogenesis of G. duodenalis we haveisolated Giardia clones with deficient adhesion properties and a re-duced expression of a surface antigen of 200 kDa. Moreover, adhe-sion of wild type Giardia was inhibited by a monoclonal antibody(mAb) specific for the 200 kDa surface molecule suggesting thatthis antigen may be involved in the adherence of the parasite toMDCK epithelial cells. The importance of Giardia attachmentin vivo was further suggested by the reduced capacity of theseclones to colonize the intestine of Mongolian gerbils.

2. Materials and methods

2.1. Parasites, cells and media

Giardia duodenalis, WB strain corresponds to the strain ATCC30957 from the American Type Culture Collection. Trophozoiteswere axenically cultivated in screw cap borosilicate glass tubes(16 � 125 mm) at 37 �C in modified TYI-S-33 medium (Diamondet al., 1978; Keister, 1983) containing 10% heat-inactivated bovineserum (HyClone), penicillin 50 lg/ml, streptomycin 50 lg/ml and0.5 g of bovine bile per liter (Sigma). Log-phase cultures were har-vested by chilling on ice followed by agitation to dislodge attachedcells. Trophozoites were collected by centrifugation at 500g for10 min at 4 �C and washed three times with phosphate-bufferedsaline (PBS, pH 7.2). P3x63Ag8.653 mouse myeloma (Köhler andMilstein, 1975) and Madin–Darby canine kidney (MDCK) cell lineswere grown in RPMI 1640 and Dulbecco’s modified Eagle’s med-ium (DMEM) (GIBCO), respectively, both supplemented with 10%foetal calf serum (HyClone).

2.2. Experimental animals

Six- to ten-week-old male Mongolian gerbils (Meriones ungui-culatus) were used to assess the infectivity of G. duodenalis clones.The gerbils were bred at our animal facility using animals from

Tumblebrook Farm Inc. (West Brookfield, MA, USA). At least 10days before experimental infection, individual animals were placedin cages and treated per os for three consecutive days with a solu-tion (90 mg/kg/day) of secnidazol (Secnidal, Rhöne–Poulec Phar-ma), which was administered by gavage. This treatment ensuredthat the gerbils were free from Giardia infections, as determinedby examinations of faeces and the small intestine of randomly se-lected treated and untreated gerbils. Animals were housed one percage and had access to food (Purina Lab Chow) and water ad lib.

2.3. Cloning methods and selection of trophozoites with deficiency inadhesion to epithelial cells

Two cloning methods were used. G. duodenalis (WB strain) wasinitially cloned twice in agarose (Gillin and Diamond, 1980)obtaining the C2 clone. Further cloning was carried out by limitingdilution following the method by Baum et al. (1988). C5 clone wasisolated after the third cloning using this method. Adhesion-defi-cient G. duodenalis C6 and C7 clones were isolated by limiting dilu-tion from C5 clone. The initial lag growth phase of these clones waslonger than the one observed in the wild type phenotype but theexponential phase was comparable for all clones (data not shown).The adhesion and growth characteristics remained constant for upto three months and then slowly reverted to the wild typephenotype.

2.4. Adhesion to MDCK cells

Giardia duodenalis trophozoites (7.5 � 105) suspended in 300 llDMEM without serum were added to wells (24 well-plates, Linbro,McLean) containing confluent MDCK cells and incubated 90 min at37 �C. After the interaction, the DMEM medium containing unat-tached trophozoites was collected. Remaining trophozoites at-tached to the cells were collected by incubating the wells withPBS at 4 �C for 30 min. Both fractions were pelleted, fixed andcounted and the percentage of adherent trophozoites was calcu-lated considering as 100% the number of trophozoites (7.5 � 105)used in the assays. To determine the effect of mAbs in the adhesionof trophozoites to MDCK cells, trophozoites were incubated usingsubagglutinating concentrations of mAbs at 4 �C for 60 min previ-ous to the attachment assay.

2.5. Experimental infection

Each of five gerbils per G. duodenalis clone or WB strain was in-fected orally with 1 � 106 trophozoites by gavage as previously de-scribed (Argüello-García and Ortega-Pierres, 1997). Fifteen dayslater when, as previously reported a maximum number of tropho-zoites is observed in the small intestine (Belosevic et al., 1983), thepresence and number of trophozoites in the small intestine weredetermined by the method of Roberts-Thomson et al. (1976) andAggarwal et al. (1983). Briefly, the first 10 cm of the upper intestinewere scraped into 10 ml of ice-cold PBS and the scrapings werecentrifuged at 500g for 5 min. The sediment was resuspended in500 ll–5 ml of PBS according to the number of trophozoites foundin each animal and the number of trophozoites was counted usinga hemocytometer. Results were calculated by taking the mean tro-phozoite counts from all animals inoculated with the same G. duo-denalis clone or WB strain.

2.6. Production of monoclonal antibodies specific to G. duodenalissurface proteins

Female BALB/c mice were injected intraperitoneally with1 � 107 trophozoites three times at weekly intervals. Animals wereboosted intravenously a month later with 1 � 107 trophozoites and

366 J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372

4 days later the animals were sacrificed. Fusion to P3x63Ag.653myeloma cells in a ratio of 1:1 was done following the procedureof Fazekas de St. Groth and Scheidegger (1980). Hybrids producingantibodies and giving positive ELlSA results with crude extractwere tested for G. duodenalis surface-specific antibodies by indirectimmunofluorescence and strongly positive cultures were ex-panded and cloned twice by limiting dilution. Immunoglobulinsubclass was determined by ELlSA using anti-mouse isotype-spe-cific antibodies (Cappel, Laboratories). NIM-M1 (Ortega-Pierreset al., 1984) specific for tyvelose moiety expressed in Trichinellaspiralis muscle larvae surface antigens was used as an isotypematched negative control for the IgM mAbs produced. The IgG3C1 mAb which recognizes a G. duodenalis surface epitope not ex-posed on trophozoites from WB strain and on the clones obtainedin this work was the negative control for IgG mAbs.

2.7. ELlSA

The ELlSA assays were performed as described by Voller et al.(1979). Crude G. duodenalis extracts (10 lg/ml in carbonate buffer,pH 9.6) were used to evaluate antibody titers in immunized mousesera, to detect positive hybridoma clones and to determine the iso-type of mAbs. One hundred microliters of this antigen were addedto each well of a 96-well flat-bottomed polystyrene microtiterplate (Dynatech) and incubated at 4 �C overnight. After blockingnonbinding sites with 0.5% BSA, 0.5% Tween 20 in PBS, hybridomasupernatants were added to each well in triplicate. Plates wereincubated at 37 �C for 90 min with shaking and then washed threetimes with PBS–Tween. One hundred microliters of horseradishperoxidase-labelled goat anti-mouse IgG or IgM at 1:2000 dilutionin PBS–Tween were added. The plates were washed three timesand then reacted with 0.1 ml of substrate solution (0.04% O-phen-ylenediamine, 0.03% H2O2, 0.1 M citric acid and 0.2 M sodiumdiphosphate, pH 5) at room temperature for 10–15 min. The plateswere read at a 492 nm on a Titertek multiscan plate (Minireader11, Dynatech Laboratories). A cut off value of 0.2 was establishedfrom the mean plus 3 standard deviations from negative serumcontrol absorbance. Wells with absorbances above 0.5 were con-sidered positive. Serum from immunized and non immunized miceor the culture medium used to grow hybridoma cells (10% FCS,DMEM) were used as positive and negative controls, respectively.

2.8. Immunofluorescence

Monoclonal antibodies directed to surface antigens were de-tected by indirect immunofluorescence using live trophozoites.To reduce background immunofluorescence, trophozoites werefirst incubated with 1% bovine serum in PBS at room temperaturefor 1 h. Trophozoites were washed with PBS three times by centri-fugation at 500g for 10 min, after this were incubated with undi-luted hybridoma supernatant at 37 �C for 1 h and then washed 3times. The cells were then incubated with a 1:500 dilution of fluo-rescein isothiocyanate-labelled rabbit anti-mouse immunoglobu-lins (Cappel, Laboratories). After incubation at room temperaturefor 1 h, trophozoites were washed three times and reactivity ofmAbs in trophozoites as detected by fluorescence was examinedwith a Zeiss fluorescence microscope.

2.9. Radioiodination

This was performed according to Ortega-Pierres et al. (1984).Briefly a reaction mixture was prepared in disposable 13 �100 mm ice-cold glass tubes, by adding 0.5 mCi of Na 125I (Amer-sham), 11 ll of 270 mM KI–PBS, 100 ll of lactoperoxidase (1 mg/ml in PBS) and 120 � 106 trophozoites suspended in PBS to complete1 ml. Two volumes of 25 ll 0.3% H2O2 in PBS were added. Trophozo-

ites were washed 4–5 times with 0.1 M KI–PBS by centrifuging at750g for 10 min and solubilized for antigen preparation as describedbelow. Microscopic examination of the trophozoites at the end of thelabelling period revealed that nearly all were motile and retainednormal morphology, and no cells were stained with trypan blue.

2.10. Antigen preparation

Antigens used in ELlSA and SDS–PAGE were prepared by sus-pending trophozoites in 0.5% Triton X-100 (Sigma) in 10 mM Tris,pH 8.3 containing the protease inhibitors phenylmethylsulfonylfluoride (1 mM) and N-ethyl maleimide (25 mM) and sonicated 6times using 15-s bursts at 12 microns on ice (Ultrasonic Generator,MSE) alternating with 30-s intervals. The sonicate was centrifugedat 15,600g for 30 min (4 �C) and the supernatant was collected andthe protein content was estimated by a modified Lowry assay (Dul-ley and Grieve, 1975). Freshly prepared antigen was used immedi-ately or stored in liquid nitrogen.

2.11. Gel electrophoresis and immunoblotting

Gradient (7.5–15%) acrylamide slab gel electrophoresis was per-formed by the method of Laemmli (1970) with the following mod-ifications. A Mini Protean II electrophoresis unit (Bio-Rad) was used.Soluble antigen, prepared as described above, was boiled in samplebuffer for 3 min with or without b-mercaptoethanol. Then40,000 cpm of radioiodinated or 40 lg of cold antigen in each lanewere normally loaded. Proteins of G. duodenalis and molecularweight standards (Pharmacia) were electrophoresed simulta-neously. Two-dimensional analysis was performed as describedby O’Farrell (1975) using a 3–10 isoelectric point range in the firstdimension and in the second dimension a continuous 7.5–15% poly-acrylamide gradient under reducing conditions. After electrophore-sis, gels containing 125I-labelled proteins, were dried and exposed at�70 �C to Kodak X-Omat XK-1 film with an intensifying screen.

In the immunoblot assays, the proteins separated by electro-phoresis were transferred from the gel to nitrocellulose paperusing the method of Towbin et al. (1979). Control strips containingmolecular weight standards and parasite proteins were stainedwith amido black (Fisher). Protein blots to be reacted with mAbswere first incubated in PBS containing 0.25% (wt/vol) gelatine (type1, Sigma), and 5 mM EDTA at 37 �C for 60 min. The strips wereincubated with hybridoma culture supernatants, polyclonal anti-bodies or DMEM (negative control) at 37 �C for 3 h. After washing3 times with PBS–0.05% Tween 20, the strips were incubated at37 �C for 3 h with horseradish peroxidase-conjugated goat anti-mouse Igs diluted 1:2000. After incubation with secondary anti-bodies, strips were washed and incubated with 4-Cl-1-a-naphtoland 0.03% H2O2 for 2 to 5 min.

2.12. Statistics

Adhesion, inhibition of adhesion and animal colonization datawere analyzed by using the two-tailed Student’s t-test. Values ofp < 0.05 were considered significant.

3. Results

3.1. Isolation of G. duodenalis clones with deficient adhesion

Adhesion-deficient G. duodenalis C6 and C7 clones were isolatedby limiting dilution from C5 clone which in turn was derived fromWB strain after successive steps of limiting dilution and cloning inagarose as described in Section 2. These two clones differ in theiradhesion capacity and considering the adhesion of WB strain to

J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372 367

MDCK cells as 100%, C7 clone showed the lowest adhesion effi-ciency (39%) compared with 58.2% and 96.5% for C6 and the wildtype C5 clones, respectively. The rates of adherence of C6 and C7clones were significantly lower compared with that of wild typeWB strain or C5 clone (p < 0.01, Table 1).

3.2. Infection of mongolian gerbils with G. duodenalis clones

To analyze whether the adhesion deficiency of the G. duodenalisclones isolated correlated with their capacity to colonize gerbilintestines and establish the infection, the animals were inoculatedintragastrically as indicated in Section 2 and 15 days later thenumber of parasites in intestinal contents was determined. While

Table 1Adherence of G. duodenalisa clones to MDCK cells

G. duodenalis strain or clone Percent adherencea (mean ± SE)

WB 100C5 96.5 ± 7.9b

C6 58.2 ± 4.9c

C7 39.4 ± 2.4c

a Adhesion of WB strain was taken as 100% (absolute value for WB was82.8 ± 6.4), n = 15 for each clone and is the result of five experiments in triplicate.

b Not significant compared with WB.c p < 0.01 compared with WB or C5.

Table 2Infectivity of G. duodenalis clones for Mongolian gerbils

G. intestinalisstrain or clone

Ratio of gerbils positive atnecropsy to gerbils inoculated

Percentages of trophozoitesa

(mean ± SE)

PBS 0/0 0/0WB 5/5 100C5 5/5 94.7 ± 0.83b

C6 2/5 8.7 ± 0.51c

C7 2/5 1.6 ± 0.07c

Five gerbils per clone were inoculated intragastrically with 1 � 106 trophozoitesand 15 days later the number of parasites in the small intestine was determined.

a Expressed as a percentage of the number of trophozoites collected from theanimal group inoculated with WB strain (absolute number for WB strain is3.32 � 106 ± 0.91).

b Not significant compared with WB.c p < 0.01 compared with WB or C5.

Fig. 1. Electrophoretic analysis of radioiodinated proteins. (A) A Triton X-100-soluble exta gradient 7.5–15% SDS–PAGE and stained with Coomassie blue. Molecular weights and pand used for autoradiography. Molecular weight markers are indicated. WB, C5, C6 and

all gerbils (5 animals/clone/experiment, the results from one repre-sentative experiment are shown in Table 2) inoculated with WBstrain or C5 clone became infected, only 40% (2 out of 5) of the ani-mals inoculated either with C6 or C7 clones were positive for intes-tinal parasites. Animal groups inoculated with C6 and C7 clonesshowed significantly lower (p < 0.01) numbers of trophozoites incomparison with animals infected with trophozoites from WBstrain or C5 clone. Only 8.7% and 1.6% of trophozoites were deter-mined in animals infected with C6 or C7 clones, respectively, whencompared with the number of trophozoites collected from animalsinoculated with WB (Table 2) or 9.2% and 1.7% trophozoites wererecovered from animals infected with C6 or C7 clones, respectively,when compared with the number of trophozoites collected fromanimals inoculated with C5 clone (data not shown in the table).These findings suggest a correlation between in vitro adhesion defi-ciency of C6 and C7 clones and its deficient capacity to colonize thesmall intestinal epithelium of gerbils.

3.3. Surface protein analysis

Cell surface proteins of the G. duodenalis clones were identifiedand characterized using the lactoperoxidase procedure as de-scribed in Section 2. SDS–PAGE analysis of iodinated proteins re-vealed a prominent high molecular weight component of200 kDa in WB strain and C5 clone (Fig. 1A) and bands of 28 and88 kDa in C6 and C7 clones as well as a band of 82 kDa in cloneC7 (Fig. 1B). The two-dimensional electrophoretic analysis indi-cated that the 200 and 88 kDa components are composed of acidicproteins with different isoelectric points (Fig. 2B). The 200 kDaprotein was markedly less labelled in C6 and C7 clones. Other lessheavily labelled components were detected at 23, 39 and 88 kDabands in WB strain and C5 clone. The reduced expression of the200 kDa protein in C6 and C7 clones was corroborated by the Coo-massie blue-stained profiles (Fig. 1A); however, no additional dif-ferences were observed.

3.4. Surface-specific monoclonal antibodies

To further characterize the 200 kDa G. duodenalis surface proteinas well as other surface antigens, and to determine if they play anyrole in the attachment of trophozoites to MDCK cells, several sur-

ract from labelled trophozoites with 125I was analyzed under reducing conditions byositions of standards are indicated on the left. (B) The gels shown in (A) were driedC7 correspond to WB strain and C5, C6 and C7 clones, respectively.

368 J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372

face antigen-specific monoclonal antibodies against G. duodenalistrophozoites were isolated. Five stable cell lines were preparedfrom BALB/c mice immunized with axenic trophozoites from WBisolate. When hybridoma culture supernatants were assayed byindirect immunofluorescence using WB strain trophozoites threemAbs (1G3, 3E8 and 2B12) reacted with surface membrane compo-nents, giving uniform fluorescence staining of the cell surface andflagella on living trophozoites (Fig. 3A only reactivity of mAb 1G3is shown). Other two mAbs were isolated; one (1E11) also reactedwith living cells, giving a patchy fluorescence pattern of the cellbody but not the flagella (Fig. 3B). The other mAb (2A12) gave onlya faint uniform fluorescence of the body (Fig. 3C). The intensity ofthe immunofluorescent staining was unchanged by the presenceof complete growth medium or its components, indicating thatthe antibodies were specific for trophozoite antigens. In addition,control normal mouse sera or the control mAbs 3C1 (IgG) or NIM-M1 (IgM) showed no reactivity with WB strain or Giardia clones(data not shown) some of the mAbs properties are summarized inTable 3. When these antibodies were tested with G. duodenalisclones, mAbs 1G3, 3E8 and 2B12, reacted with more than 99% oftrophozoites from WB strain and C5 clone and with less than 1%of trophozoites from C6 and C7 clones (data not shown).

Fig. 2. Two-dimensional electrophoretic analysis of radioiodinated proteins. (A) Soluconditions by two-dimensional electrophoresis as described by O’Farrell (1975) using a cblue. (B) The gels used in (A) were dried and autoradiographed. pI stands for isoelectric

3.5. Protein specificity of the monoclonal antibodies

To examine the surface protein specificities of the monoclonalantibodies, immunoblot experiments were performed on TritonX-100 extracts of trophozoites. Hybridoma supernatants contain-ing mAbs 1G3, 2B12 and 3E8 recognized a similar 200 kDa band(Fig. 4, only reaction with mAb 1G3 is shown). A control immuno-blot incubated with an irrelevant monoclonal isotype-matchedantibody (NIM-M1) directed against T. spiralis gave no reaction.No reactivity was found with normal mouse serum or with med-ium (DMEM, 10% CFS). The mobility of the 200 kDa protein onSDS–PAGE was the same under reducing or non reducing condi-tions indicating that it is a single chain protein with few, if any,intrachain disulfide bonds susceptible to reducing agents.

3.6. Effect of monoclonal antibody 1G3 on trophozoite attachment toMDCK cells

In these assays the monoclonal antibody 1G3 was selectedbased on the strong fluorescent signal observed in trophozoitesincubated with this mAb as well as in the strong reaction withGiardia extract as detected by Western blot analysis Trophozoites

ble proteins from labelled trophozoites with 125I were analyzed under reducingontinuous 7.5–15% polyacrylamide gradient. The gels were stained with Coomassiepoint.

Fig. 2 (continued)

J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372 369

were incubated at 4 �C for 60 min in TYI-S medium, pre-immuneand immune mouse serum, or in subagglutinating concentrationsof 1G3 monoclonal antibodies. After incubation, the cells werewashed with PBS, counted and checked for viability by trypan blue

Fig. 3. Reactivity of monoclonal antibodies on G. duodenalis trophozoites determined bhybridoma culture supernatants from mAbs IG3 (A), 1E11 (B) or 2A12 (C). Reactivity of thrabbit anti-mouse immunoglobulins. Scale bar: 10 lm.

exclusion before interaction with MDCK cells. This antibody par-tially inhibited the adherence of trophozoites from WB strain andC5 clone to MDCK cells (52% and 40.9%, respectively, Table 4,p < 0.005) compared with the control isotype matched mAb NIM-

y indirect immunofluorescence. Trophozoites from WB strain were incubated withese monoclonal antibodies was visualized using a fluorescein isothiocyanate-labeled

Table 4Inhibition of adhesiona of Giardia trophozoites to MDCK cells

G. duodenalis strain or clone Monoclonal antibody

1G3 b3C1 cNIM-M1

WB d52.0 ± 6.8 90.4 ± 0.8 91.6 ± 2.0C5 d40.9 ± 8.6 93.0 ± 1.9 92.9 ± 2.0C6 93.9 ± 2.9 94.5 ± 6.7 94.5 ± 4.5C7 105.6 ± 11.9 91.7 ± 2.5 92.1 ± 7.8

a Expressed as a percentage of the adherence observed by the correspondingtrophozoites not exposed to antibody. (Absolute adherence for WB, C5, C6 and C7were 81.9 ± 5.6, 79.8 ± 2.1, 55.6 ± 8.06 and 31.9 ± 9.5, respectively). Trophozoites(7.5 � 105) were incubated with subagglutinating concentrations of mAbs at 4 �Cfor 60 min. They were washed with PBS and allowed to adhere to MDCK confluentmonolayers at 37 �C for 90 min. Both trophozoites not attached to MDCK cells andthose remaining adherent, were collected, fixed and counted. n = 9 and is the resultof three experiments by triplicate.

b mAb (lgG) which is directed against a G. duodenalis surface epitope not exposedon trophozoites of the WB strain or clones analysed in this study.

c mAb (lgM) against Trichinella spiralis surface proteins.d p < 0.01 compared with mAb NIM-M1.

Table 3Properties of five surface-specific monoclonal antibodies directed against G. duode-nalis trophozoites

Antibody Isotype Trophozoite aggregation aMW of proteins recognized

1G3 IgM ++++ 200 kDa3E8 IgG ++ 200 kDa2B12 IgG ++ 200 kDa1E11 IgM ND Multiple banding2A12 ND ND Multiple banding

ND, not determined.a Determined by immunoblotting.

370 J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372

M1 or IgG 3C1 mAb. The significant inhibition of adherence of tro-phozoites from WB strain and C5 clone by mAb 1G3, in which the200 kDa component is highly expressed, suggests that this antigenmay be involved in the adhesion phenomenon.

4. Discussion

In the present work, we describe the isolation of adhesion-defi-cient G. duodenalis trophozoites from C6 and C7 clones from WBstrain. These clones showed a reduced adhesion to MDCK cellsand a defective infectivity for Mongolian gerbils indicating a corre-lation of virulence in vivo and adhesion deficiency in vitro. Surfaceradiolabelling revealed the prominent 28 and 88 kDa bands inthese clones, in contrast to the prominent labelling of a 200 kDacomponent in trophozoites from wild type C5 clone and WB strain.In addition, the mAbs 1G3, that stained more than 99% of tropho-zoites from WB and C5 clone and less than 1% of trophozoites fromC6 and C7 clones, significantly inhibited the adhesion of trophozo-

Fig. 4. Western blot analysis of G. duodenalis protein recognized by monoclonalantibody 1G3. Trophozoite proteins were solubilized in Triton X-100 by sonicationand separated by gradient (7.5–15%) SDS–PAGE and transferred to nitrocellulosepaper (NC). NC was incubated with mAb 1G3 and developed with a peroxidase-coupled anti-mouse IgGs. Lane 1; standards of the indicated molecular weight. WB,C5, C6 and C7 correspond to trophozoites extracts from WB strain and C5, C6 and C7clones, respectively.

ites to MDCK cells from WB strain and from C5 clone, suggestingthe participation of this antigen in the adhesion process. Two-dimensional analysis of radiolabel surface proteins indicated thatthe 88 and 200 kDa components are acidic proteins. The stronglylabelled 200 kDa protein component clearly corresponded to asimilar 200 kDa Coomassie blue-stained protein from WB strainand C5 clone trophozoites. In contrast, the 28 and 88 kDa proteinswere expressed at similar levels by all Giardia clones, indicatingthat the differential labelling of these proteins in C6 and C7 clonesis due to a differential exposure of radioiodinable residues at thetrophozoite’s surface. Unlike the polydispersed pattern of radioio-dinated components ranging from 94 to 225 kDa reported by Nashet al. (1983), we detected a prominently radiolabel 200 kDa proteinand the less heavily labelled 23, 39 and 88 kDa proteins in the wildtype C5 clone and the WB isolate of G. duodenalis. Later, the samegroup (Nash et al., 1990) reported variable expression of surfaceantigens of 72 and 200 kDa in clones derived from the GS/M-85isolate of G. duodenalis. Earlier studies on the characterization ofGiardia surface proteins also failed to detect a polydispersed pat-tern (Einfeld and Stibbs, 1984; Edson et al., 1986). It is possible thatthe radioiodinable proteins of 170 and 180 kDa previously identi-fied (Nash and Aggarwall, 1986; Torian et al., 1984; Einfeld and Sti-bbs, 1984) correspond to the 200 kDa protein we have detected.The 88 kDa protein is the most frequently identified surface mole-cule in G. duodenalis. Also an 82 kDa surface antigen has been de-tected by Einfeld and Stibbs (1984), Kumkum et al. (1988) andOrtega-Pierres et al. (1988). The identity of these antigens alongwith the strongly radiolabel 88 kDa protein identified in adhe-sion-deficient C6 and C7 clones remains to be established.Although in the present report, the role of the 88 kDa protein inthe attachment phenomenon was not investigated, a previousstudy has suggested that 88 and 85 kDa surface antigens partici-pate in the adhesion of G. duodenalis trophozoites to target cells(Inge et al., 1988; Gillin et al., 1990). Thus, these antigens and othermembrane lectins specific for D-mannosyl and glucosyl residuesmay partially influence parasite adherence to epithelial cells (Ingeet al., 1988; Magne et al., 1991).

The isolation of adhesion-deficient Giardia clones indicateseither the heterogeneous composition of the trophozoites popula-tion or the production of new phenotypic clones. Different studieshave revealed heterogeneity within the population of different par-asitic protozoan including Giardia (Adam et al., 1988; Thompsonand Meloni, 1993; Méndez et al., 2001; Argüello-García et al.,2004). In addition, several studies demonstrate that Giardia isolatesfrom humans differ in surface antigens and other biological charac-teristics (Nash and Keister, 1985; Smith et al., 1982; Bertram et al.,

J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372 371

1983), and that these differences may explain some of the variabil-ity in the clinical features and infectivity noted in humans, miceand gerbils (Nash et al., 1987; Aggarwal et al., 1983; Aggarwaland Nash, 1987). Thus, the isolation of the adhesion-deficient Giar-dia clones in this work further corroborates the heterogeneitywithin Giardia cell populations and their reduced ability to estab-lish the infection in Mongolian gerbils may be partially caused bythe differences displayed in the surface protein profile and thein vitro adhesion deficiency. The isolation of clones deficient inadhesion and virulence from other parasitic protozoan has alsoproved to be useful to ascertain the relationship of surface mole-cules to adherence and other aspects of the pathogenesis (Rodri-guez and Orozco, 1986; Orozco et al., 1987; Méndez et al., 2001;Nevils et al., 2000). For example, B. bovis clones with phenotypicdifferences in adhesion to cultured epithelial cells and virulencein Holstein cattle have been isolated (Canto et al., 2006). Similarly,adhesion-deficient clones of E. histolytica have established the roleof surface molecules in the adherence and virulence (Arroyo andOrozco, 1987). Thus, adhesion-deficient mutants altered in theexpression of a 112 kDa surface protein are not recognized bymAbs which inhibit partially wild-type trophozoite adherence (Ar-royo and Orozco, 1987; Rodriguez et al., 1989; García-Rivera et al.,1997).

The selection of surface-specific monoclonal antibodies hasbeen successfully used to assess the role of surface molecules inthe adhesion of other protozoan parasites (Meza et al., 1987; Rav-din et al., 1986; Arroyo and Orozco, 1987; García-Rivera et al.,1997,1999; Canto et al., 2006). In our study the surface-specificmonoclonal antibodies (1G3, 2B12 and 3E8) may be directedagainst different epitopes of the same �200 kDa protein as theyare not of the same isotype and the mAb 1G3 gave the brightestmembrane fluorescence and the strongest reaction by immunob-loting. Furthermore, mAb 1G3 agglutinated live G. duodenalis tro-phozoites, however under subaglutinating titer, this mAbsignificantly inhibited the adherence of trophozoites. Some reportssuggest that G. duodenalis like some bacteria and other protozoahave surface proteins with lectin activity which has been relatedto adherence with mammalian erythrocytes and rat epithelial cells(Farthing et al., 1986; Inge et al., 1988). Also inhibition studies withantibodies have suggested that the surface molecules with molec-ular weights of 64, 66, 85 and 88 kDa may be involved in theattachment of G. duodenalis trophozoites to epithelial cellsin vitro (Inge et al., 1988; Gillin et al., 1990).

Finally, it was suggested for sometime that the force responsi-ble for Giardia attachment was a negative pressure below the ven-tral chamber of the adhesive disk (VAD). However recent studiesby Erlandsen et al. (2004) using microfabricated pillars to studytrophozoites attachment have indicated that adhesion of trophozo-ites occurs via adhesive forces which are not related to the VAD. In-stead they suggest that the ventrolateral flange (VL) which in livingtrophozoites undergoes rapid dynamic changes in forming focalcontacts might play an important role in orienting trophozoitesattachment via the VAD. This process could be influence by eitherlectin-mediated interactions or surface molecules. In our study theobservation that the mAb directed against the 200 kDa proteininhibited the adhesion of WB and clone 5 to target cells suggestthat indeed surface molecules might influence cell adhesion. Fur-thermore, the reduced ability of adhesion-deficient Giardia clones,which do not express this surface molecule, to establish the infec-tion in Mongolian gerbils may be partially caused by the differ-ences displayed in the surface protein profile. Recent studieshave reported that surface molecules might also interact withcytoskeleton proteins in the attachment of E. histolytica to entero-cytes (Tavares et al., 2000, 2005). In this context it will be interest-ing to determine if molecules of these two compartments interactin the adhesion of G. duodenalis to target cells.

Although adherence and in vivo colonization are very complexphenomena in the host-parasite interplay, this work demonstratesthe value of the dual strategy of using surface-specific monoclonalantibodies and adhesion-deficient G. duodenalis clones. It will beinteresting to further analyse the role of the 200 kDa antigen de-tected in this study in Giardia adherence and virulence. Further-more, the purification of each of the surface proteins identified inthis study is necessary to establish their contribution in adhesion,pathogenesis and host immunity. To this aim, proteomic analysisof the proteins from the clones used in this study will be helpfulin defining more precisely the factors involved in adhesion of tro-phozoites to target cells.

Acknowledgments

We are grateful to Michael Parkhouse, Raúl Argüello-García andRosa María Bermúdez for critically reading this manuscript. Wealso thank René López Bolaños and Blanca Herrera Ramírez fortechnical assistance and Arturo Perez-Taylor for technical assis-tance and the photographic work.

References

Adam, R.D., Aggarwall, A., Lal, A.A., De la Cruz, V.F., McCutchan, T., Nash, T.E., 1988.Antigenic variation of a cysteine-rich protein in Giardia lamblia. Journal ofExperimental Medicine 167, 109–118.

Aggarwal, A., Nash, T.E., 1987. Comparison of two antigenically distinct Giardialamblia isolates in gerbils. The American Journal of Tropical Medicine andHygiene 36 (2), 325–332.

Aggarwal, A., Bhatia, A., Naik, S.R., Vinayak, V.K., 1983. Variable virulence of isolatesof Giardia lamblia in mice. Annals of Tropical Medicine and Parasitology 77 (2),163–167.

Argüello-García, R., Cruz-Soto, M., Romero-Montoya, L., Ortega-Pierres, G., 2004.Variability and variation in drug susceptibility among Giardia duodenalisisolates and clones exposed to 5-nitroimidazoles and benzimidazoles in vitro.Journal of Antimicrobial Chemotherapy 54, 711–721.

Argüello-García, R., Ortega-Pierres, M.G., 1997. Giardia duodenalis: analysis ofhumoral immune response in experimentally infected gerbils (Merionesunguiculatus). Archives of Medical Research 28 (2), 171–178.

Arroyo, R., Orozco, E., 1987. Localization and identification of an Entamoebahistolytica adhesin. Molecular and Biochemical Parasitology 23, 151–158.

Baum, K.F., Berens, R.L., Jones, R.H., Marr, J.J., 1988. A new method for cloningGiardia lamblia, with a discussion of the statistical considerations of limitingdilution. Journal of Parasitology 74, 267–269.

Belosevic, M., Faubert, G.M., MacLean, J.D., Law, C., Croll, N.A., 1983. Giardia lambliainfections in Mongolian gerbils: an animal model. The Journal of InfectiousDiseases 147 (2), 222–226.

Bertram, M.A., Meyer, E.A., Lile, J.D., Morse, S.A., 1983. A comparison of isoenzymesof five axenic Giardia isolates. The Journal of Parasitology 69 (5), 793–801.

Buret, A.G., 2007. Mechanisms of epithelial dysfunction in giardiasis. Gut 56 (3),328–335.

Campanati, L., Holloschi, A., Troster, H., Spring, H., De Souza, W., Monteiro-Leal, L.H.,2002. Video-microscopy observations of fast dynamic processes in theProtozoon Giardia lamblia. Cell Motility and the Cytoskeleton 51, 213–224.

Canto, G.J., Figueroa, J.V., Ramos, J.A., Rojas, E.E., García-Tapia, D., Alvarez, J.A., Allred,D.R., Carson, C.A., 2006. Evaluation of cattle inoculated with Babesia bovis clonesadhesive in vitro to bovine brain endothelial cells. Annals of the New YorkAcademy of Sciences 1081, 397–404.

Chin, A.C., Teoh, D.A., Scott, K.G., Meddings, J.B., Macnaughton, W.K., Buret, A.G.,2002. Strain-dependent induction of enterocyte apoptosis by Giardia lambliadisrupts epithelial barrier function in a caspase-3-dependent manner. Infectionand Immunity 70 (7), 3673–3680.

Crossley, R., Holberton, D.V., 1983. Characterization of proteins from thecytoskeleton of Giardia lamblia. Journal of Cell Science 59, 81–103.

Crouch, A.A., Seow, W.K., Whitman, L.M., Smith, S.E., Thong, Y.H., 1991. Inhibition ofadherence of Giardia duodenalis by human neutrophils and monocytes.Transactions of the Royal Society of Tropical Medicine and Hygiene 85, 375–379.

Diamond, L.S., Harlow, D.R., Cunnick, C.C., 1978. A new medium for the axeniccultivation of Entamoeba histolytica and other Entamoeba. Transactions of theRoyal Society of Tropical Medicine and Hygiene 72 (4), 431–432.

Dulley, J.R., Grieve, P.A., 1975. A simple technique for eliminating interference bydetergents in the Lowry method of protein determination. AnalyticalBiochemistry 64, 136–141.

Edson, C.M., Farthing, M.J.G., Thorley-Lawson, D.A., Keush, G.T., 1986. An 88,000-MrGiardia lamblia surface protein which is immunogenic in humans. Infection andImmunity 54 (3), 621–625.

Einfeld, D.A., Stibbs, H.H., 1984. Identification and characterization of a majorsurface antigen of Giardia lamblia. Infection and Immunity 46 (2), 377–383.

372 J. Hernández-Sánchez et al. / Experimental Parasitology 119 (2008) 364–372

Erlandsen, S.L., Russo, A.P., Turner, J.N., 2004. Evidence for adhesive activity of theventrolateral flange in Giardia lamblia. The Journal of Eukaryotic Microbiology51 (1), 73–80.

Farthing, M.J.G., Pereira, M.E.A., Keusch, G.T., 1986. Description and characterizationof a surface lectin from Giardia lamblia. Infection and Immunity 51 (2), 661–667.

Fazekas de St. Groth, S., Scheidegger, D., 1980. Production of monoclonal antibodies:strategy and tactics. Journal of Immunological Methods 35, 1–21.

Feely, D.E., Schollmeyer, J.V., Erlandsen, S.L., 1982. Giardia spp. Distribution ofcontractile proteins in the attachment organelle. Experimental Parasitology 53,145–154.

Gallegos, B., De la Garza, M., Meza, I., 1986. Mutantes de movilidad de Entamoebahistolytica resistentes a citocalasina D. Archivos de Investigación Medica 17(suppl.), 87–93.

García-Rivera, G., Avila, A., Ayala, P., Arroyo, R., Rigothier, M.C., Orozco, E., 1997.Identification and location of the cell-binding domain in the 112 kDa adhesingene of Entamoeba histolytica. Archives of Medical Research 28 (Spec), 164–165.

García-Rivera, G., Rodriguez, M.A., Ocádiz, R., Martínez-López, M.C., Arroyo, R.,González-Robles, A., Orozco, E., 1999. Entamoeba histolytica: a novel cysteineprotease and an adhesin form the 112 kDa surface protein. MolecularMicrobiology 33 (3), 556–568.

Gillin, F.D., Diamond, L.S., 1980. Clonal growth of Giardia lamblia trophozoites in asemisolid agarose medium. Journal of Parasitology 66 (2), 350–352.

Gillin, F.D., Hagblom, P., Harwood, J., Aley, S.B., Reiner, D.S., McCaffery, M., So, M.,Guiney, D.G., 1990. Isolation and expression of the gene for a major surfaceprotein of Giardia lamblia. Proceedings of the National Academy of Sciences USA87, 4463–4467.

Hansen, W.R., Tulyathan, O., Dawson, S.C., Cande, W.Z., Fletcher, D.A., 2006. Giardialamblia attachment force is insensitive to surface treatments. Eukaryotic Cell 5(4), 781–783.

Holberton, D.V., 1974. Attachment of Giardia—a hydrodynamic model based onflagellar activity. Journal of Experimental Biology 60, 207–221.

Inge, P.M., Edson, C.M., Farthing, M.J., 1988. Attachment of Giardia lamblia to ratintestinal epithelial cells. Gut 29, 795–801.

Keister, D.B., 1983. Axenic culture of Giardia lamblia in TYI-S-33 mediumsupplemented with bile. Transactions of the Royal Society of TropicalMedicine and Hygiene 77 (4), 487–488.

Köhler, G., Milstein, C., 1975. Continuous culture of fused cells secreting antibody ofpredefined specificity. Nature (London) 256, 495–497.

Kumkum, Khanna, R., Khuller, M., Mehta, S., Vianayak, V.K., 1988. Plasma membraneassociated antigens of trophozoites of axenic Giardia lamblia. Transactions of theRoyal Society of Tropical Medicine and Hygiene 82, 439–444.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature (London) 227, 680–685.

Magne, D., Favennec, L., Chochillon, C., Gorenflot, A., Meillet, D., Kapel, N., Raichvarg,D., Savel, J., Gobert, J.G., 1991. Role of cytoskeleton and surface lectins in Giardiaduodenalis attachment to Caco2 cells. Parasitology Research 77, 659–662.

McCoy, J.J., Mann, B.J., Petri Jr., W.A., 1994. Adherence and cytotoxicity of Entamoebahistolytica or how lectins let parasites stick around. Infection and Immunity 62,3045–3050.

Méndez, S., Fernández-Pérez, F.J., Santín, M., De la Fuente, C., Cuquerella, M.,Gómez-Muñóz, M.T., Alunda, J.M., 2001. Correlation between in vitro and in vivoinfectivity of Leishmania infantum clones. The Journal of EukaryoticMicrobiology 48 (6), 616–621.

Meza, I., Cazares, F., Rosales-Encina, J.L., Talamas-Rohana, P., Rojkind, M., 1987. Useof antibodies to characterize a 220-kilodalton surface protein from Entamoebahistolytica. The Journal of Infectious Diseases 156 (5), 798–805.

Nash, T.E., Keister, D.B., 1985. Differences in excretory–secretory products andsurface antigens among 19 isolates of Giardia. The Journal of Infectious Diseases152 (6), 1166–1171.

Nash, T.E., Herrington, D.A., Losonsky, G.A., Levine, M.M., 1987. Experimentalhuman infections with Giardia lamblia. The Journal of Infectious Diseases 156(6), 974–984.

Nash, T.E., Gillin, F.D., Smith, P.D., 1983. Excretory–secretory products of Giardialamblia. The Journal of Immunology 131 (4), 2004–2010.

Nash, T.E., Herrington, D.A., Levine, M.M., Conrad, J.T., Merritt Jr., J.W., 1990.Antigenic variation of Giardia lamblia in experimental human infections. TheJournal of Immunology 144 (11), 4362–4369.

Nash, T.E., Aggarwall, A., 1986. Cytotoxicity of monoclonal antibodies to a subset ofGiardia isolates. The Journal of Immunology 136 (7), 2628–2632.

Nevils, M.A., Figueroa, J.V., Turk, J.R., Canto, G.J., Le, V., Ellersieck, M.R., Carson, C.A.,2000. Cloned lines of Babesia bovis differ in their ability to induce cerebralbabesiosis in cattle. Parasitology Research 86 (6), 437–443.

O’Farrell, P.H., 1975. High resolution two-dimensional electrophoresis of proteins.Journal of Biological Chemistry 250, 4007–4021.

Orozco, E., Rodriguez, M.A., Murphy, C.F., Salata, R.A., Petri, W.A., Smith, R.D., Ravdin,J.I., 1987. Entamoeba histolytica: cytopathogenicity and lectin activity ofavirulent mutants. Experimental Parasitology 63, 157–165.

Ortega-Pierres, G., Chayen, A., Clark, N.W., Parkhouse, R.M., 1984. The occurrence ofantibodies to hidden and exposed determinants of surface antigens ofTrichinella spiralis. Parasitology 88, 359–369.

Ortega-Pierres, M.G., Lascurain, R., Argüello García, R., Coral Vázquez, R., Acosta, G.,Santos, J.I., 1988. The response of humans to antigens of Giardia lamblia.Advances in Giardia Research, 177–180.

Petri Jr., W.A., Haque, R., Mann, B.J., 2002. The bittersweet interface of parasite andhost: lectin-carbohydrate interactions during human invasion by the parasiteEntamoeba histolytica. Annual Reviews of Microbiology 56, 39–64.

Ravdin, J.I., Petri, W.A., Murphy, C.F., Smith, R.D., 1986. Production of mousemonoclonal antibodies which inhibit in vitro adherence of Entamoeba histolyticatrophozoites. Infection and Immunity 53 (1), 1–5.

Roberts-Thomson, I.C., Stevens, D.P., Mahmoud, A.A., Warren, K.S., 1976. Acquiredresistance to infection in an animal model of giardiasis. Journal of Immunology117, 2036–2037.

Rodríguez-Fuentes, G.B., Cedillo-Rivera, R., Fonseca-Liñán, R., Argüello-GarcÍa, R.,Muñoz, O., Ortega-Pierres, G., Yépez-Mulia, L., 2006. Giardia duodenalis: analysisof secreted proteases upon trophozoite–epithelial cell interaction in vitro.Memórias do Instituto Oswaldo Cruz 101 (6), 693–696.

Rodriguez, M.A., Orozco, E., 1986. Isolation and characterization of phagocytosis-and virulence-deficient mutants of Entamoeba histolytica. The Journal ofInfectious Diseases 154 (1), 27–32.

Rodriguez, M.A., Hernandez, F., Santos, L., Valdez, A., Orozco, E., 1989. Entamoebahistolytica: molecules involved in the target cell–parasite relationship.Molecular and Biochemical Parasitology 37, 87–99.

Scott, K.G., Meddings, J.B., Kirk, D.R., Lees-Miller, S.P., Buret, A.G., 2002. Intestinalinfection with Giardia spp. reduces epithelial barrier function in a myosin lightchain kinase-dependent fashion. Gastroenterology 123 (4), 1179–1190.

Smith, P.D., Gillin, F.D., Kaushal, N.A., Nash, T.E., 1982. Antigenic analysis of Giardialamblia from Afghanistan, Puerto Rico, Ecuador, and Oregon. Infection andImmunity 36 (2), 714–719.

Sousa, M.C., GonÇalves, C.A., Bairos, V.A., Poiares-da-Silva, J., 2001. Adherence ofGiardia lamblia trophozoites to Int-407 human intestinal cells. Clinical andDiagnostic Laboratory Immunology 8 (2), 258–265.

Tavares, P., Sansonetti, P., Guillén, N., 2000. The interplay between receptor cappingand cytoskeleton remodelling in Entamoeba histolytica. Archives of MedicalResearch 31, S140–S142.

Tavares, P., Rigothier, M.C., Khun, H., Roux, P., Huerre, M., Guillén, N., 2005. Roles ofcell adhesion and cytoskeleton activity in Entamoeba histolytica pathogenesis: adelicate balance. Infection and Immunity 73 (3), 1771–1778.

Thompson, R.C., Meloni, B.P., 1993. Molecular variation in Giardia. Acta Tropica 53,167–184.

Torian, B.E., Barnes, R.C., Stephens, R.S., Stibbs, H.H., 1984. Tubulin and high-molecular-weight polypeptides as Giardia lamblia antigens. Infection andImmunity 46 (1), 152–158.

Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of proteins frompolyacrylamide gels to nitrocellulose sheets: procedure and some applications.Procedures of the National Academy of Sciences (USA) 76, 4350–4354.

Troeger, H., Epple, H.J., Schneider, T., Wahnschaffe, U., Ullrich, R., Burchard, G.D.,Jelinek, T., Zeitz, M., Fromm, M., Schulzke, J.D., 2007. Effect of chronic Giardialamblia infection on epithelial transport and barrier function in humanduodenum. Gut 56 (3), 328–335.

Tse, S.K., Chadee, K., 1991. The interaction between intestinal mucus glycoproteinsand enteric infections. Parasitology Today 7, 163–172.

Voller, A., Bidwell, D.E., Bartlett, A., 1979. The Enzyme Linked ImmunosorbentAssay, ELlSA Dynatech Europe, Guernsey, United Kingdom.

Wolfe, M.S., 1992. Giardiasis. Clinical Microbiology Reviews 5 (1), 93–100.