JOURNAL OF CLINICAL MICROBIOLOGY, Nov. 1992, p. 2807-28130095-1137/921112807-07$02.00/0Copyright © 1992, American Society for Microbiology

Vol. 30, No. 11

A Monoclonal Antibody for Distinction of Invasive andNoninvasive Clinical Isolates of Entamoeba histolytica

ARMANDO GONZALEZ-RUIZ,1* RASHIDUL HAQUE,2 TAYYAB REHMAN,1 AURA AGUIRRE,3CARLOS JARAMILLO,3 GUADALUPE CASTANON,4 ANDREW HALL,2t FELIPE GUHL,3GUILLERMO RUIZ-PALACIOS,4 DAVID C. WARHURST,' AND MICHAEL A. MILES1

Department ofMedical Parasitology, London School ofHygiene and Tropical Medicine, Keppel Street,London WCIE 7HT, United Kingdom1; International Centre for Diarrhoeal Diseases Research,

Bangladesh, Dhaka 1,000, Bangladesh2; Laboratorio de Microbiologia y Parasitologia,Departamento de Ciencias Biol6gicas, Universidad de los Andes, Apartado

Ae'reo 4976, Bogotd, Colombia3; and Departamento de Infectologia,Instituto Nacional de la Nutricion "Salvador Zubirdn,"

C P. 14000, MeXico City, Mexico4

Received 8 June 1992/Accepted 4 August 1992

Approximately 10% of the world population is infected with Entamoeba histolyfica, but only 10% of thecarriers develop symptomatic amebiasis. This discrepancy could be explained by the genotypic differencesbetween the morphologically indistinguishable invasive and noninvasive strains of E. histolytica currentlyidentified by zymodeme analysis, a technique that is unsuitable for routine diagnostic laboratories. Here wereport the production of a monoclonal antibody against E. histolytica and its use in an immunofluorescenceassay to identify invasive isolates cultured from stool samples of infected patients in several regions whereamebiasis is endemic: Bangladesh, Colombia, and Mexico. After testing a total of 88 E. histolytica isolates, thecorrelation between zymodeme characterization and the immunofluorescence assay with the invasive isolate-specific monoclonal antibody was 100%o. The epitope detected by the invasive isolate-specific monoclonalantibody resides in a previously undescribed internal protein with molecular masses of 84 and 81 kDa in axenicand polyxenic E. histolytica strains, respectively.

Entamoeba histolytica is the etiological agent of humanamebiasis. The infection is distributed worldwide, althoughthe majority of cases are found in developing countrieswhere fecal-oral transmission of E. histolytica cysts is facil-itated by poor hygiene and the lack of clean water. Thespectrum of clinical manifestations observed in infectedindividuals ranges from asymptomatic infection to intestinaland extraintestinal invasive disease. In 1984, it was esti-mated that approximately 10% of the world population wasinfected with E. histolytica and that only 10% of the infectedsubjects progressed to invasive amebiasis (47). The discrep-ancy between the number of infected persons and theoccurrence of invasive amebiasis seems to be due to theexistence of two genetically different but morphologicallyindistinguishable parasites within the species. Brumpt (5)made this suggestion after infecting kittens with E. histolyt-ica strains from asymptomatic carriers and observing nodevelopment of disease. He proposed a new nomenclaturefor the noninvasive and the invasive strains: Entamoebadispar and Entamoeba dysenteriae, respectively. A fewyears later, Simic (36) infected kittens with E. histolyticastrains from asymptomatic carriers and then passed thesestrains on to human volunteers. Neither the kittens nor thehumans developed symptoms attributable to the infection.More recently, Martinez-Palomo et al. (21) showed that

only invasive strains of E. histolytica agglutinated in the

* Corresponding author.t Present address: Department of Pure and Applied Biology,

Imperial College, London SW7 2BB, United Kingdom.

presence of the lectin concanavalin A, giving evidence for aphenotypic difference residing on the parasite membrane.Later, Sargeaunt and coworkers provided strong support forBrumpt's suggestion by showing, with more than 3,000 E.histolytica isolates, that zymodeme analysis clearly distin-guished isolates cultured from asymptomatic patients andthose cultured from patients with symptomatic amebic in-fection (31-33).On the basis of zymodeme characterization, other mark-

ers, such as resistance to complement-mediated lysis byhuman serum (28), release of a neutral thiol proteinase (27),overexpression of and structural differences between homol-ogous cysteine proteinases (42), and recognition by specificmonoclonal antibodies (MAbs) (22, 24, 38, 40) and byspecific DNA probes (4, 7, 13, 39, 41, 42), have beendescribed for invasive strains. Restriction fragment lengthpolymorphisms and sequencing of the small subunit of therRNA of invasive and noninvasive E. histolytica strains haveshown an estimated sequence divergence of 2.2% (6), evi-dence which reaffirms Brumpt's suggestion that the twoparasites are sufficiently genetically distinct to merit thestatus of independent species.Zymodeme analysis is time-consuming and expensive, and

so simpler diagnostic techniques are needed for the identifi-cation of invasive strains of E. histolytica. Here we reportthe production and applications of an invasive isolate-spe-cific MAb against E. histolytica for the identification ofinvasive clinical isolates from several regions where theorganism is endemic and for the preliminary characterizationof the corresponding antigen, which has not previously beendescribed.

2807

2808 GONZALEZ-RUIZ ET AL.

MATERIALS AND METHODS

Parasite strains and culture conditions. The axenic E.histolytica strains HK9 (14), NIH:200 (43), and HM-1:IMSS,formerly known as ABRM (8) (ATCC 30015, 30458, and30459, respectively), which were originally isolated frompatients with invasive intestinal amebiasis, were cultured inDiamond's medium TPS-1 (9) at 37°C. Entamoeba mosh-kovskii CST (ATCC 30131), a free-living ameba isolated fromsewage, the reptilian ameba Entamoeba invadens TRM (23),and a clinical isolate of Pentatrichomonas hominis, anintestinal commensal flagellate (1), were cultured in biphasicRobinson's medium (30) at room temperature (RT). Blasto-cystis hominis, an organism of uncertain taxonomic statusassociated with diarrhea in humans (1), was cultured inbiphasic Robinson's medium at 37°C. Giardia intestinalisPortland 1 was cultured in Diamond's medium TYI-S-33 (9)at 37°C. The polyxenic E. histolytica invasive isolates TEand SI and the noninvasive isolates C29, 8672, and SAW1734 were cultured from patients with symptomatic andasymptomatic amebic infections, respectively (38). All theclinical isolates of E. histolytica cultured from patients inareas where amebiasis is endemic were initially isolated andmaintained in biphasic Robinson's medium at 37°C. Someisolates were also mass cultured in liquid Robinson's me-dium (37) at 37°C. These isolates were from the InternationalCentre for Diarrhoeal Diseases Research, Bangladesh (ICD-DRB), in Dhaka, Bangladesh, the Laboratorio de Microbio-logia y Parasitologia de la Universidad de los Andes (UNI-ANDES) in Bogota', Colombia, and the Instituto Nacional dela Nutricion "Salvador Zubiran" (INNSZ) in Mexico City,Mexico. Isolates from ICDDRB came either from patients ofall age groups seen at the hospital for diarrheal diseases or

from children less than 14 years old living in an urban slumarea in Dhaka. Isolates from UNIANDES were obtainedfrom patients in areas where amebiasis is endemic or frompatients referred to UNIANDES for parasitological diagno-sis. Isolates from INNSZ were recovered from a suburbancommunity on the outskirts of Mexico City and, within thatcommunity, mainly from children 5 years of age and under.

Stool samples. Recently collected stools were examinedmacroscopically for the presence of blood and mucus; asmear of feces in 0.9% saline was examined microscopicallyfor the presence of erythrocytes, leukocytes, and E. histolyt-ica trophozoites; the presence of erythrocytes within E.histolytica trophozoites was also recorded. Feces wereconcentrated by the formalin-ether concentration technique(29) for parasite cysts and ova, and the concentrate wasexamined microscopically. Feces were also inoculated intoRobinson's medium within 6 h of collection, and E. histolyt-ica-positive cultures were subcultured every 48 h.Zymodeme characterization. The zymodeme characteriza-

tion of the E. histolytica isolates was performed in thecountry where the isolates were obtained. At ICDDRB,cellulose acetate plates and a Zip Zone chamber with a TitanPlus electrophoresis power supply (no. 1501; Helena Labo-ratories, Beaumont, Tex.) were used as previously described(15); at UNIANDES, thin-layer starch-gel electrophoresiswas employed according to the methods of Sargeaunt andWilliams (32) and Sargeaunt et al. (33) for phosphoglucomu-tase (EC 2.7.5.1), L-malate:NADP+ oxidoreductase (oxalo-acetate decarboxylating) (EC 1.1.1.4.0), and glucose phos-phate isomerase (EC 5.3.1.9) and according to the methodsof Farri et al. (10) for hexokinase (EC 2.7.1.1); and atINNSZ, the method was that of Kollaritsch et al. (18) withautomated isoenzyme isoelectrofocusing on PhastSystem

equipment (no. 18-1600-01; Pharmacia, Uppsala, Sweden).Briefly, E. histolytica trophozoites were harvested from2-day, well-grown Robinson's cultures when the number oforganisms was approximately 5 x 104/ml (31), washed twicewith sterile 0.9% saline, resuspended in 500 pl of distilledwater containing enzyme stabilizers (1 mM EDTA-1 mMdithiothreitol-1 mM I-aminocaproic acid), lysed by freezingand thawing (three times), and centrifuged at 14,000 x g for1 h at 4°C. The supematants were collected, aliquoted, andstored in liquid N2 until used. After electrophoresis of thelysates, the enzymes were developed in the dark by using anagar overlay incorporating substrate solutions according tostandard methods (10, 34).

Production of E. histolytica NP-40 protein extract. Axenicand polyxenic E. histolytica strains were harvested fromliquid Robinson's medium by chilling the culture tubes inice-water for 15 min, washed (three times) in cold phosphate-buffered saline (PBS), pH 7.2, and centrifuged at 250 x g for5 min at 4°C. The pelleted cells were lysed by the addition of3 volumes of lysis buffer (20 mM Tris-HCl [pH 8.0], 150 mMNaCl, 0.75% [wt/vol] Nonidet P-40 [NP-40] [no. 56009; BDHChemicals, Poole, Dorset, United Kingdom]) containingprotease inhibitors (1 mM phenylmethylsulfonyl fluoride[no. P7626; Sigma Chemical Co., Poole, Dorset, UnitedKingdom], 1 mM iodacetamide [no. 441812H; BDH], and 1,g of tosyl-L-lysine chloromethyl ketone [no. T7254; Sigma])per ml with continuous stirring on ice for 15 min. Themixture was centrifuged at 16,000 x g for 15 min at 4°C, andthe supernatant was recovered and stored at -20°C after theprotein content was measured by the BCA protein assayreagent method (no. 23225; Pierce, Rockford, Ill.).

Production of MAbs. MAbs were produced according tothe method of Galfre and Milstein (12) with some modifica-tions. Two BALB/c mice were immunized intraperitoneallywith 50 ,ug of NP-40 protein extract ofE. histolytica NIH:200emulsified with an equal volume of complete Freund'sadjuvant. Mice were given intraperitoneal booster injectionsof the same extract and incomplete Freund's adjuvant (threetimes) at 2-week intervals. A final booster of 50 ,ug of proteindiluted in 200 ,ul of PBS was injected into the tail vein. Theantibody response was assessed by indirect enzyme-linkedimmunosorbent assay (ELISA) (46) by coating the mi-croplate wells with the E. histolytica antigen used for immu-nization. Four days after the final booster, the spleen of onemouse was removed and the spleen cells were fused with theSP2/O-Agl4 myeloma cell line (35) by using 45% polyethyl-ene glycol 4000 (no. 9727; Merck, Darmstadt, Germany)(11), pH 8.1 (48). Hybridomas secreting MAbs against wholeE. histolytica antigen were screened by ELISA and clonedtwice by limiting dilution by using BALB/c peritoneal mac-rophages as feeder cells. A third screening of nine selectedhybridoma cell lines (see Results) was performed by anindirect immunofluorescence assay (IFA) (see below) withreference invasive and noninvasive E. histolytica polyxenicstrains as antigens. Ascites were produced by the inocula-tion of hybridomas intraperitoneally into pristane-primedmice. MAbs were isotyped by ELISA with a commercial kit(Sigma ISO-2). The specificity of MAbs against other intes-tinal parasites (see Results) was determined by ELISA andIFA.

Immunofluorescence analysis of fixed parasites. Axenic andpolyxenic parasite strains cultured in Diamond's or liquidRobinson's medium were harvested by chilling culture tubesin ice-water for 15 min and washed twice with cold PBS, and30 ,ul of a suspension of 105 cells per ml was spotted ontoeach well of multispot polytetrafluorethene-coated slides

J. CLIN. MICROBIOL.

A MONOCLONAL ANTIBODY FOR INVASIVE E. HISTOLYTICA 2809

(no. PH-001; Loughton, Hendley, Essex, United Kingdom).Strains grown in biphasic Robinson's medium were har-vested after 2 days by pipetting out the bottom starch layerfrom the 5-ml container and spotting out 30 ,ul of the mixturecontaining parasites, bacteria, and starch as describedabove. Slides were air dried, fixed in methanol for 5 min, andstored at -20°C. Slides were warmed at RT for 30 min, andtwofold serial dilutions of MAb-containing culture superna-tant or mouse ascites were prepared in PBS-2% casein(PBS/C). A 40-,ul sample of each dilution was spotted perwell across the slides in duplicate, and the slides wereincubated at RT for 30 min in a humid chamber. After theslides were washed twice in a PBS bath for 5 min at RT withcontinuous rocking, 40 ,lI of a 1/50 dilution in PBS/C of goat(at the London School of Hygiene and Tropical Medicine[LSHTM]) or rabbit (at ICDDRB) anti-mouse immunoglob-ulins conjugated to fluorescein isothiocyanate (no. F1010;Sigma, and Dakopatt A/S, Copenhagen, Denmark, respec-tively), with 1/10,000 (wt/vol) Evans blue for counterstain-ing, was spotted per well and the slides were incubated foranother 30 min at RT. The slides were washed as describedabove, mounted in 50% glycerol in PBS, and examined atICDDRB in an Olympus fluorescent microscope (modelBH-2; Olympus Optical Co. Ltd., Tokyo, Japan) and atLSHTM in a Leitz Diaplan microscope with epifluorescentlight (Leitz, Wetzlar, Germany). The titers of positive slideswere recorded. Initially, preimmune and immune mouse serawere used as negative and positive controls, respectively, ata 1/200 dilution. Later, slides with known invasive andnoninvasive strains were also included as controls whenunknown clinical isolates were tested.

Immunofluorescence analysis of living parasites. Immuno-fluorescence analysis of live E. histolytica trophozoites wascarried out as previously described by Blakely et al. (2) withsome modifications. Briefly, axenic HM-1:IMSS trophozo-ites were harvested and washed in cold 0.9% saline with0.2% glucose and 2 mM CaCl2-1 mM ascorbic acid (no.10303; BDH) and 6 mM L-cysteine (no. C-7880; Sigma).Aliquots of 105 cells were tested with mouse ascites contain-ing invasive isolate-specific MAb diluted 1/20 in the samesaline solution for 1.5 h at 37°C, and duplicate aliquots werekept at 4°C. After being washed (twice), the parasites wereexposed to a 1/20 dilution of fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin (no. F1010;Sigma) for 30 min at 37°C, with duplicate samples exposed at4°C. After being washed again, trophozoites were examinedas wet preparations on a Leitz Diaplan fluorescence micro-scope. Controls included cell aliquots fixed with 1% gluta-raldehyde and tested as described above and trophozoitestested with buffer alone, without conjugate, or with nonim-mune ascites.Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

(SDS-PAGE) and Western blotting (immunoblotting). NP-40protein extracts from axenic and polyxenic strains culturedin liquid Robinson's medium were mixed with sample buffer(62.5 mM Tris-HCl [pH 6], 2% SDS, 5% P-mercaptoethanol,10% glycerol, 0.002% bromophenol blue). Polyxenic E.histolytica strains cultured in biphasic Robinson's mediumwere harvested and mixed with sample buffer, and parasiteproteins were solubilized by boiling for 4 min. After boiling,samples were centrifuged at 8,500 x g for 1 min at RT andsupernatants were loaded for electrophoresis by using 25 pgof protein from the parasites cultured in liquid medium and30 ,ul of the mixture from parasites cultured in biphasicmedium per gel track. A 5-,ul sample of a solution ofmolecular mass standards (as follows [molecular masses in

kilodaltons]: myosin, 200; phosphorylase b, 97.4; bovineserum albumin, 69; ovalbumin, 46; carbonic anhydrase, 30;trypsin inhibitor, 21.5; lysozyme, 14.3; no. RPN.756; Amer-sham International plc, Amersham, Buckinghamshire, En-gland) was mixed with an equal volume of sample buffer,boiled as described above, and loaded in one track of eachgel. SDS-PAGE was performed according to Laemmli'smethod (19) with 12% separating and 4% stacking gels in aMini-Protean II Dual Slab Cell (no. 165-2940; Bio-Rad Lab-oratories Ltd., Hemel Hempstead, Hertfordshire, UnitedKingdom) at 200 V for 45 min. Electrophoresed proteinswere transferred to nitrocellulose (no. 401-196; Schleicher &Schuell, Dassel, Germany) by the procedure of Towbin et al.(45) using a semidry electroblotting apparatus (SartoblotII-S, no. SM 17556; Sartorius Ltd., Epsom, Surrey, UnitedKingdom) (4 mA/cm2, 30 min). Replica gels were stainedwith Coomassie brilliant blue. Nitrocellulose was blockedwith PBS-0.05% Tween 20-2% casein (PBST/C) (60 min atRT), washed (three times for 10 min each at RT in PBS-0.05% Tween 20), and probed with supernatant (1/2 dilutionin PBST/C) or mouse ascites (1/100 dilution 1 h, RT) con-taining invasive isolate-specific MAb. After repeated wash-ing, as described above, the membrane was exposed to a1/2,000 dilution of peroxidase-conjugated rabbit anti-mouseimmunoglobulin (no. 315-036-003; Jackson ImmunoresearchInc., Philadelphia, Pa.) in PBST/C (1 h, RT) and washedagain (three times), and the protein bands were developed in50 mM Tris-HCl (pH 7.4)-0.2 M NaCl-0.8% diaminobenzi-dine (no. 13033; BDH)-5% 4-chloro-1-naphthol (no. C-8890;Sigma)-0.006% H202 for 2 to 3 min.

RESULTS

Selection of an invasive isolate-specific MAb against E.histolytica. Nine of seventeen hybridoma cell lines initiallyselected by ELISA detected one or more bands on Westernblots (Fig. 1A). These nine cell lines were screened by IFAon fixed E. histolytica trophozoites, two of which (NIH:200and SI) were from an invasive zymodeme and two of which(C29 and 8672) were from noninvasive zymodemes. MAb20/7D, an immunoglobulin Gl mouse immunoglobulin,which detected a single band on the Western blots, in IFArecognized only the invasive E. histolytica trophozoites.This prompted us to test a larger number of reference(HM1:IMSS and SAW 1734) and clinical isolates from dif-ferent areas of endemicity (Table 1). MAb 20/7D reactedwith all the invasive E. histolytica isolates and with none ofthe noninvasive isolates, including those isolates which werecharacterized by zymodeme and the four isolates whichwere characterized on clinical grounds only. The fluorescentpatterns observed varied from patchy to homogeneous in thecytoplasm and were more intense in the nucleus and itsmembrane (Fig. 2). The culture conditions (axenic or poly-xenic isolates and liquid or biphasic medium) and the differ-ent methods of IFA slide preparation (parasites harvestedfrom liquid medium or spotted directly from biphasic Rob-inson's medium with its accompanying mixture of starch andbacteria) did not affect the recognition of invasive E. his-tolytica by MAb 20/7D (Table 1). Interestingly, with somepolyxenic isolates, such as BOG-3, not all trophozoites onthe slide reacted with the MAb, suggesting the presence of aheterogeneous cell population infecting the patient (Fig. 2).MAb 20/7D did not react in IFA with trophozoites of E.moshkovskii, E. invadens, G. intestinalis, P. hominis, and B.hominis (data not shown).

Preliminary characterization of the antigen recognized by

VOL. 30, 1992

2810 GONZALEZ-RUIZ ET AL.

A B12 3 4 5

200 . _97.4 F69 .4630 --21.5 _

1 2 3

200°i 97.4-F

69 Lr463021.5

FIG. 1. Western blot of axenic and polyxenic clinical invasiveisolates of E. histolytica. NP-40 protein extract of the axenic strainHK9 and the mixture of starch, bacteria, and parasites from thebottom layer of well-grown polyxenic Robinson's cultures were

subjected to SDS-PAGE and transferred to nitrocellulose. (A) Nitro-cellulose was cut into individual strips and probed with either culturesupernatant diluted 1/2 or serum from the mouse that was used for thefusion diluted 1/200. Lanes: 1 and 2, two different MAbs which were

not specific for invasive E. histolytica and which represent the ninehybridoma lines selected by Western blotting for screening by immu-nofluorescence with invasive and noninvasive E. histolytica tropho-zoites; 3, invasive isolate-specific MAb 20/7D; 4, polyclonal mouse

serum; 5, secondary anti-mouse peroxidase-conjugate only. (B) Ni-trocellulose was probed with ascites containing the invasive isolate-specific MAb 20/7D diluted 1/100. Lanes: 1, axenic invasive E.histolytica HK9; 2 and 3, polyxenic invasive clinical isolates BOG-2and BOG-3 cultured from patients in Colombia with amebic dysen-tery. Numbers on the left, molecular weight markers (in thousands).

MAb 20/7D. MAb 20/7D reacted with axenic and polyxenicinvasive E. histolytica trophozoites fixed with methanol(Fig. 2) and with HM-1:IMSS trophozoites fixed with glutar-aldehyde but failed to react with live HM-1:IMSS trophozo-

ites (data not shown). This suggests that the epitope recog-nized by MAb 20/7D is not exposed on the surface of theparasite.

In SDS-PAGE and Western blotting, MAb 20/7D recog-nized a single band with estimated molecular masses of 84kDa in axenically grown and 81 kDa in polyxenically growninvasive E. histolytica (Fig. 1B). This band appeared clearlyin samples from all the invasive isolates tested and as adefined or very faint band in samples from the noninvasiveisolates SAW 1734, BOG-1, and C29 (data not shown).

DISCUSSION

We have produced a MAb capable of distinguishing inva-sive strains of E. histolytica from noninvasive strains byIFA. So far, we have tested 88 E. histolytica isolates,comprising 6 reference and 82 clinical isolates, 78 character-ized by zymodeme and 4 by clinical information only (seebelow). The clinical isolates were recovered from patientswith symptomatic or asymptomatic infections from Bang-ladesh, Colombia, and Mexico, countries in which amebicinfection is highly endemic (47) and which are widely sepa-rated geographically. Overall, the sensitivity and specificityof the IFA based on the invasive isolate-specific MAb 20/7Dwere both 100%.The 78 clinical isolates characterized by zymodeme rep-

resent three different zymodemes, two invasive and onenoninvasive, from a total of the six parent zymodemesreclassified by Blanc and Sargeaunt (3). These authorssuggested such a reclassification after showing that theamount of starch in the culture medium produced variabilityin the isoenzyme electrophoresis patterns; this variabilityproduced second-order zymodemes derived from the parentzymodemes. This variability together with the cost and timeneeded for the zymodeme analysis supports the use of

TABLE 1. Reactivity of MAb 20/7D with reference and clinical isolates of E. histolytica

Tiype and Culture No. with diagnosisZeodemed IFAno. of Origin' conditions' D ACeme titerisolatesADDC

Reference3 U.K. Ax (2), Xe (1) 3 II 1:1,6003 U.K. Xe; L, B (1) 3 I Negative

Clinical2 Ban Xe 2 XIV 1:1,2808 Ban Xe 7 1 XIV 1:6403 Ban Xe 3 XIV 1:3203 Ban Xe 2 1 II 1:1,280

16 Ban Xe 6 9 1 II 1:6404 Ban Xe 1 3 II 1:3201 Col Xe; L, B 1 II 1:3,2001 Col Xe 1 II 1:1,6002 U.K. Xe 2e ND 1:1,6001 Col Xe 1 ND 1:1,600

31 Ban Xe 27 4 I Negative2 Col Xe, L, B 2 I Negative7 Mex Xe 7 INegative1 Col Xe 1 ND Negative

a Ban, Bangladesh; Col, Colombia; Mex, Mexico; U.K., United Kingdom. The axenic strains HK9 and HM-1:IMSS were isolated in Korea and Mexicorespectively, and maintained in culture at LSHTM.

b Ax, axenic isolates cultured in Diamond's medium; Xe, polyxenic or monoxenic isolates with Escherichia coli (all isolates cultured in biphasic Robinson'smedium unless otherwise stated); L, liquid medium; B, biphasic medium; numbers in parentheses, numbers of isolates cultured under the conditions stated.

c AD, amebic dysentery (erythrocytes observed inside E. histolytica trophozoites by microscopy); D, diarrhea (loose or liquid stools); AC, asymptomaticcarrier (formed stools).

d XIV and II, invasive zymodemes; I, noninvasive zymodeme; ND, not done.e One isolate recovered from the stools of a patient suffering from an amebic liver abscess.

J. CLIN. MICROBIOL.

A MONOCLONAL ANTIBODY FOR INVASIVE E. HISTOLYTICA 2811

simpler laboratory techniques such as MAb-based IFA forthe characterization of clinical isolates of E. histolytica (38,40). Recently developed objective attachments allow fluo-rescence microscopy to be performed with a standard lightmicroscope in field conditions (20, 26). We have also adaptedMAb 20/7D to a capture ELISA for the detection of E.histolytica antigen in feces (14a). Thus, if the IFA assayresults are confirmed worldwide with all the zymodemes,zymodeme analysis can be reserved for reference laborato-ries.The nonuniform reactivity of some clinical polyxenic

isolates in IFA agrees with the concept that some patientscan be infected with mixtures of E. histolytica strains, whichrequires cloning of the cultured parasites for reliable zymo-deme analysis (31). Direct visualization in the IFA eliminatesthe need for cloning. Alternatively, this nonuniform reactiv-ity could be the result of differential or developmentalregulation of the expression of the epitope by invasivestrains in culture.

Considering the evolutionary distance between invasiveand noninvasive strains (6), it is not surprising to findanother antigenic difference that separates them. Neverthe-less, MAb 20/7D recognizes an epitope with interestingfeatures. The 81- and 84-kDa protein bearing the epitope isnot equivalent to those already described (2, 22, 24, 27, 40,44). Furthermore, the fact that it is present in both invasiveand noninvasive isolates, as shown by Western blotting, butdifferentially recognized by IFA suggests that this epitope isexpressed and processed in such a way that it is accessible toMAb 20/7D only in nondenatured E. histolytica antigens ofthe invasive zymodemes. Once the protein has been dena-tured in SDS-PAGE, the epitope in noninvasive strains alsobecomes accessible to the MAb. Similar examples of crypticepitopes unmasked by solubilization of proteins have beenseen with the Rho(D) antigen of human erythrocytes (25) andmembrane antigens of the parasites Leishmania tropica (16)and Trypanosoma cruzi (17).MAb 20/7D has been used to isolate a clone from a cDNA

expression library of the E. histolytica invasive strain NIH:200 (14a). This finding suggests that the reactivity of theepitope with the MAb does not depend on glycosylation ofthe protein, although glycosylation differences could explainthe observed difference in molecular weights between theaxenic and the polyxenic isolates studied. The binding ofMAb 20/7D might reflect different biological properties suchas pathogenicity, although it does not imply major structuraldifferences, as the epitope is present in both strains. Quan-titative studies on the expression of the protein in invasiveand noninvasive E. histolytica strains and further compari-sons of these antigens could give a better insight into thebasis of their distinction by the 20/7D MAb.The conservation of the epitope recognized by MAb 20/7D

FIG. 2. Immunofluorescent photomicrographs of methanol-fixedE. histolytica trophozoites probed with mouse ascites containing theinvasive isolate-specific MAb 20/7D diluted 1/50, followed by goatanti-mouse fluorescein-conjugated antiserum diluted 1/50 containingEvans blue diluted 1/10,000 for counterstaining. (A) Axenicallygrown invasive strain HM-1:IMSS; (B) polyxenically grown inva-sive clinical isolate BOG-3; (C) polyxenically grown noninvasiveclinical isolate BOG-1. Axenic and polyxenic trophozoites werecultured in TPS-1 (Diamond's) and Robinson's media, respectively.The polyxenic invasive and noninvasive clinical isolates were iso-lated in Colombia from a patient with amebic dysentery and anasymptomatic cyst carrier, respectively.

VOL. 30, 1992

2812 GONZALEZ-RUIZ ET AL.

in the axenic strains HK9 (14) and NIH:200 (43), isolated 39and 43 years ago, respectively, and in the polyxenic culturesrecently isolated from infected patients suggests that theprotein bearing this epitope may be essential for the organ-ism.The invasive isolate 2788, from Bangladesh, was recov-

ered from an asymptomatic carrier who could be a source fordisease transmission if left untreated. Prospective epidemi-ological studies using MAb 20/7D would define more clearlythe geographical distribution and the patterns of transmis-sion of invasive strains and might allow focused interventionstudies for particular regions, communities, or individuals.Although the IFA results obtained with the clinical iso-

lates for which zymodemes had not been determined alsoaccorded with the invasive isolate-specific reactivity of MAb20/7D, it is necessary to perform zymodeme analysis withcultured E. histolytica trophozoites from a more extensiveseries of isolates. Field studies are in progress to examinesuch a series of isolates including all known zymodemes.A limited number of direct fecal smears from patients with

dysentery have been examined by fluorescence microscopyafter being probed with MAb 20/7D, with encouraging re-sults (data not shown). If perfected, this technique wouldobviate the need to culture the parasite prior to identificationof invasive organisms.

ACKNOWLEDGMENTS

A.G.R. is a recipient of a scholarship from the Consejo Nacionalde Ciencia y Tecnologia (CONACYT), Mexico, and A.A. is partiallysupported by a scholarship from the Sir Patrick Manson Bequest.This work has been supported by the Nestle Nutrition ResearchGrant Programme (grant no. 87/34), a grant from the Commission ofthe European Communities (contract no. STD-2 0206-UK), theBritish Council in Colombia and the United Kingdom, and the WHOProgramme on Intestinal Parasitic Infections.We are very grateful to the following people for providing us with

parasite strains: L. S. Diamond, E. histolytica HK9, NIH:200, andHM-1:IMSS; J. P. Ackers, E. histolytica isolates TE, SI, C29, and8672; D. Mirelman, E. histolytica SAW 1734; J. E. Williams, E.moshkovskii, E. invadens, P. hominis, and B. hominis; and A.Goldin, G. intestinalis Portland 1. We also thank I. Frame for hiscomments on the manuscript.

REFERENCES1. Ash, L. R., and T. C. Orihel (ed.). 1990. Protozoans, p. 82-88.

Atlas of human parasitology American Society of ClinicalPathologists, Chicago.

2. Blakely, P., P. G. Sargeaunt, and S. L. Reed. 1990. An immu-nogenic 30-kDa surface antigen of pathogenic clinical isolates ofEntamoeba histolytica. J. Infect. Dis. 162:949-954.

3. Blanc, D., and P. G. Sargeaunt. 1991. Entamoeba histolyticazymodemes: exhibition of a and 8 bands only of glucose phos-phate isomerase and phosphoglucomutase may be influenced bystarch content in the medium. Exp. Parasitol. 72:87-90.

4. Bracha, R., L. S. Diamond, J. P. Ackers, G. D. Burchard, and D.Mirelman. 1990. Differentiation of clinical isolates of Ent-amoeba histolytica by using specific DNA probes. J. Clin.Microbiol. 28:680-684.

5. Brumpt, E. 1925. Etude sommaire de l"Entamoeba dispar' n.sp., amibe a kystes quadrinuclees parasite de l'homme. Bull.Acad. Med. 94:942-952.

6. Clark, C. G., and L. S. Diamond. 1991. Ribosomal RNA genesof 'pathogenic' and 'nonpathogenic' Entamoeba histolytica aredistinct. Mol. Biochem. Parasitol. 49:297-302.

7. Cruz-Reyes, J. A., W. M. Spice, T. Rehman, E. Gisborne, and J.Ackers. 1992. Ribosomal DNA sequences in the differentiationof pathogenic and non-pathogenic isolates of Entamoeba his-tolytica. Parasitology 104:239-246.

8. De la Torre, M., L. Landa, and B. Sepulveda. 1970. Avances en

los metodos para el cultivo de Enu;moeba histolytica. Arch.Invest. Med. (Mexico) 1(Suppl.):S9-Sl4.

9. Diamond, L. S. 1983. Lumen dwelling protozoa: Entamoeba,trichomonads and Giardia, p. 65-109. In J. B. Jensen (ed.), Invitro cultivation of protozoan parasites. CRC Press, Inc., BocaRaton, Fla.

10. Farri, T. A., P. G. Sargeaunt, D. C. Warhurst, J. E. Williams,and R. Bhojnani. 1980. Electrophoretic studies of the hexoki-nase of Entamoeba histolytica groups I to IV. Trans. R. Soc.Trop. Med. Hyg. 74:672-673.

11. Fazekas de St. Groth, S., and D. Scheidegger. 1980. Productionof monoclonal antibodies: strategy and tactics. J. Immunol.Methods 35:1-21.

12. Galfre, G., and C. Milstein. 1981. Preparation of monoclonalantibodies: strategies and procedures. Methods Enzymol. 73:1-46.

13. Garfinkel, L. I., M. Giladi, M. Huber, C. Gitler, D. Mirelman,M. Revel, and S. Rozenblatt. 1989. DNA probes specific forEntamoeba histolytica possessing pathogenic and nonpatho-genic zymodemes. Infect. Immun. 57:926-931.

14. Geiman, Q. M., and C. E. Becker. 1953. In vitro growth andmetabolism of Endamoeba histolytica. Ann. N.Y. Acad. Sci.56:1048-1056.

14a.Gonzalez-Ruiz, A., et al. Unpublished data.15. Haque, R., A. Hall, and S. Tzipori. 1990. Zymodemes of

Entamoeba histolytica in Dhaka, Bangladesh. Ann. Trop. Med.Parasitol. 84:629-632.

16. Jaffe, C. L., and D. McMahon-Pratt. 1983. Monoclonal antibod-ies specific for Leishmania tropica. I. Characterization of anti-gens associated with stage- and species-specific determinants. J.Immunol. 131:1987-1993.

17. Kirchhoff, L. V., S. Hieny, M. Shiver, D. Snary, and A. Sher.1984. Cryptic epitope explains the failure of a monoclonalantibody to bind to certain isolates of Trypanosoma cruzi. J.Immunol. 133:2731-2735.

18. Kollaritsch, H., H. Stemberger, M. Binder, and G. Wiedermann.1989. Suitability of the Phast system for discrimination ofEntamoeba isolates by isoenzyme isoelectrofocusing. Trans. R.Soc. Trop. Med. Hyg. 83:211-212.

19. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

20. Makler, M. T. 1992. Fluorescent microscope objective. Trans.R. Soc. Trop. Med. Hyg. 86:108.

21. Martinez-Palomo, A., A. Gonzalez-Robles, and M. De la Torre.1973. Selective agglutination of pathogenic strains of Ent-amoeba histolytica induced Con A. Nature (London) New Biol.245:186-187.

22. Mufioz, M. D. L., E. Lamoyi, G. Le6n, R. Tovar, J. Perez-Garcia, M. de la Torre, E. Murueta, and R. M. Bernal. 1990.Antigens in electron-dense granules from Entamoeba histolyticaas possible markers for pathogenicity. J. Clin. Microbiol. 28:2418-2424.

23. Neal, R. A. 1988. Phylogeny: the relationship of Entamoebahistolytica to morphologically similar amebae of the four-nucle-ate cyst group, p. 13-26. In J. Ravdin (ed.), Amebiasis. Humaninfection by Entamoeba histolytica. John Wiley & Sons, NewYork.

24. Petri, W. A., Jr., T. F. H. G. Jackson, V. Gathiram, K. Kress,L. D. Saffer, T. L. Snodgrass, M. D. Chapman, Z. Keren, and D.Mirelman. 1990. Pathogenic and nonpathogenic strains of Ent-amoeba histolytica can be differentiated by monoclonal antibod-ies to the galactose-specific adherence lectin. Infect. Immun.58:1802-1806.

25. Plapp, F. V., M. M. Kowalski, L. Tilzer, P. J. Brown, J. Evans,and M. Chiga. 1979. Partial purification of Rho(D) antigen fromRh positive and negative erythrocytes. Proc. Natl. Acad. Sci.USA 76:2964-2968.

26. Polsuwan, C., B. Lumlertdaecha, W. Tepsumethanon, and H.Wilde. 1992. Using the UV ParaLens adapter on a standardlaboratory microscope for fluorescent rabies antibody detec-tion. Trans. R. Soc. Trop. Med. Hyg. 86:107.

27. Reed, S. L., W. E. Keene, and J. H. McKerrow. 1989. Thiol

J. CLIN. MICROBIOL.

A MONOCLONAL ANTIBODY FOR INVASIVE E. HISTOLYTICA 2813

proteinase expression and pathogenicity of Entamoeba histolyt-ica. J. Clin. Microbiol. 27:2772-2777.

28. Reed, S. L., P. G. Sargeaunt, and A. I. Braude. 1983. Resistanceto lysis by human serum of pathogenic Entamoeba histolytica.Trans. R. Soc. Trop. Med. Hyg. 77:248-253.

29. Ritchie, L. S. 1948. An ether sedimentation technique forroutine stool examinations. Bull. U.S. Army Med. Dep. 8:326.

30. Robinson, G. L. 1968. The laboratory diagnosis of humanparasitic amoeba. Trans. R. Soc. Trop. Med. Hyg. 62:285-294.

31. Sargeaunt, P. G. 1988. Zymodemes ofEntamoeba histolytica, p.370-387. In J. Ravdin (ed.), Amebiasis. Human infection byEntamoeba histolytica. John Wiley & Sons, New York.

32. Sargeaunt, P. G., and J. E. Williams. 1978. Electrophoreticisoenzyme patterns of Entamoeba histolytica and Entamoebacoli. Trans. R. Soc. Trop. Med. Hyg. 72:164-166.

33. Sargeaunt, P. G., J. E. Williams, and J. D. Greene. 1978. Thedifferentiation of invasive and noninvasive Entamoeba histolyt-ica by isoenzyme electrophoresis. Trans. R. Soc. Trop. Med.Hyg. 72:519-521.

34. Sargeaunt, P. G., J. E. Williams, and R. A. Neal. 1980. Acomparative study of Entamoeba histolytica (NIH:200, HK9,etc.), "E. histolytica-like" and other morphologically identicalamoebae using isoenzyme electrophoresis. Trans. R. Soc. Trop.Med. Hyg. 74:469-474.

35. Shulman, M., C. D. Wilde, and G. Kohler. 1978. A better cellline for making hybridomas secreting specific antibodies. Na-ture (London) 276:269-270.

36. Simic, T. 1931. Infection experimentale de l'homme par Ent-amoeba dispar Brumpt. Ann. Parasitol. Hum. Comp. 9:385-391.

37. Spice, W. M., and J. P. Ackers. 1990. Large-scale production ofEntamoeba histolytica trophozoites in polyxenic culture. Trans.R. Soc. Trop. Med. Hyg. 84:693-694.

38. Strachan, W. D., P. L. Chiodini, W. M. Spice, A. H. Moody, andJ. P. Ackers. 1988. Immunologic differentiation of pathogenicand nonpathogenic isolates of Entamoeba histolytica. Lanceti:561-563.

39. Tachibana, H., S. Ihara, S. Kobayashi, Y. Kaneda, T. Takeuchi,and Y. Watanabe. 1991. Differences in genomic DNA sequencesbetween pathogenic and nonpathogenic isolates of Entamoebahistolytica identified by polymerase chain reaction. J. Clin.Microbiol. 29:2234-2239.

40. Tachibana, H., S. Kobayashi, Y. Kato, K. Nagakura, Y. Kaneda,and T. Takeuchi. 1990. Identification of a pathogenic isolate-specific 30,000Mr antigen of Entamoeba histolytica by using amonoclonal antibody. Infect. Immun. 58:955-960.

41. Tannich, E., R. D. Horstmann, J. Knobloch, and H. H. Arnold.1989. Genomic DNA differences between pathogenic and non-pathogenic Entamoeba histolytica. Proc. Natl. Acad. Sci. USA86:5118-5122.

42. Tannich, E., H. Scholze, R. Nickel, and R. D. Horstmann. 1991.Homologous cysteine proteinases of pathogenic and nonpatho-genic Entamoeba histolytica. Differences in structure andexpression. J. Biol. Chem. 266:4798-4803.

43. Tobie, J. E. 1949. Experimental infection of the rabbit withEndamoeba histolytica. Am. J. Trop. Med. Hyg. 29:859-870.

44. Torian, B. E., S. A. Lukehart, and W. E. Stamm. 1987. Use ofmonoclonal antibodies to identify, characterize, and purify a96,000-dalton surface antigen of pathogenic Entamoeba histolyt-ica. J. Infect. Dis. 156:334-343.

45. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretictransfer of proteins from polyacrylamide gels to nitrocellulosesheets: procedure and some applications. Proc. Natl. Acad. Sci.USA 76:4350-4354.

46. Voller, A., and D. Bidwell. 1986. Enzyme-linked immunosorbentassay, p. 99-109. In N. R. Rose, H. Friedman, and J. L. Fahey(ed.), Manual of clinical laboratory immunology, 3rd ed. Amer-ican Society for Microbiology, Washington, D.C.

47. Walsh, J. A. 1988. Prevalence of Entamoeba histolytica infec-tion, p. 91-105. In J. Ravdin (ed.), Amebiasis. Human infectionby Entamoeba histolytica. John Wiley & Sons, New York.

48. Westerwoudt, R. J. 1985. Improved fusion methods. IV. Tech-nical aspects. J. Immunol. Methods 77:181-196.

VOL. 30, 1992