Taenia solium Oncosphere Adhesion to Intestinal Epithelial andChinese Hamster Ovary Cells In Vitro�
Manuela Verastegui,1 Robert H. Gilman,1,2,7* Yanina Arana,1 Dylan Barber,3 Jeanette Velasquez,1Marilu Farfan,1 Nancy Chile,1 Jon C. Kosek,4 Margaret Kosek,2 Hector H. Garcia,1,5
Armando Gonzalez,6 and the Cysticercosis Working Group in PeruDepartment of Microbiology, Universidad Peruana Cayetano Heredia, P.O. Box 5045, Lima, Peru1; Department of International Health,
Johns Hopkins University, Bloomberg School of Hygiene and Public Health, 615 N. Wolfe Street, Room W#5515, Baltimore,Maryland 212052; Faculty of Veterinary Science, University of Melbourne, 250 Princes Highway, Werribe, Victoria 3030,
Australia3; Department of Pathology, Stanford University School of Medicine, Stanford, California 943054; Cysticercosis Unit,Instituto de Ciencias Neurologicas, Jr. Ancash 1271, Barrios Altos, Lima, Peru5; Public Health Section,School of Veterinary Medicine, Universidad Nacional Mayor de San Marcos, Apartado 03-5113, Lima 03,
Peru6; and AB PRISMA, Calle Carlos Gonzales 251, San Miguel, Lima 32, Peru7
Received 26 July 2006/Returned for modification 14 September 2006/Accepted 20 June 2007
The specific mechanisms underlying Taenia solium oncosphere adherence and penetration in the hosthave not been studied previously. We developed an in vitro adhesion model assay to evaluate themechanisms of T. solium oncosphere adherence to the host cells. The following substrates were used:porcine intestinal mucosal scrapings (PIMS), porcine small intestinal mucosal explants (PSIME), Chi-nese hamster ovary cells (CHO cells), epithelial cells from ileocecal colorectal adenocarcinoma (HCT-8cells), and epithelial cells from colorectal adenocarcinoma (Caco-2 cells). CHO cells were used to compareoncosphere adherence to fixed and viable cells, to determine the optimum time of oncosphere incubation,to determine the role of sera and monolayer cell maturation, and to determine the effect of temperatureon oncosphere adherence. Light microscopy, scanning microscopy, and transmission microscopy wereused to observe morphological characteristics of adhered oncospheres. This study showed in vitro adher-ence of activated T. solium oncospheres to PIMS, PSIME, monolayer CHO cells, Caco-2 cells, and HCT-8cells. The reproducibility of T. solium oncosphere adherence was most easily measured with CHO cells.Adherence was enhanced by serum-binding medium with >5% fetal bovine serum, which resulted in asignificantly greater number of oncospheres adhering than the number adhering when serum at aconcentration less than 2.5% was used (P < 0.05). Oncosphere adherence decreased with incubation ofcells at 4°C compared with the adherence at 37°C. Our studies also demonstrated that T. solium onco-spheres attach to cells with elongated microvillus processes and that the oncospheres expel externalsecretory vesicles that have the same oncosphere processes.
Cysticercosis is a common disease in areas of the third worldwhere pigs are raised. The life cycle of Taenia solium includesthe pig as the normal intermediate host that harbors the larvalvesicles or cysticerci. Humans can also serve as accidentalintermediate hosts through ingestion of tapeworm eggs (1, 13,24). Human cysticercosis is a parasitic disease caused by larvaeof the cestode T. solium. It is acquired when eggs released bythe adult tapeworm located in the human small intestine arepassed in human feces and accidentally ingested.
Human cysticercosis is the most common helminthic para-sitic illness affecting the central nervous system. It is especiallya problem in countries where the sanitary infrastructure isdeficient. In rural areas of Latin America, between 5 and 20%of the population show circulating antibodies to T. solium (4,12, 14).
T. solium eggs contain an oncosphere that is released from
the eggs in the host small intestine and then activated by theaction of intestinal enzymes and bile salts. In order to pene-trate through intestinal cells, the oncosphere must first adhere.Little is known about the process of oncosphere adhesion,which is required for penetration and establishment of thelarval cestode in the host.
The mechanisms by which T. solium oncospheres infect hosttissues are not known. One of the initial steps of infection bymany microorganisms involves adhesion to host cells. The pro-teins involved in the adherence mechanism can be exploited astargets for developing vaccines that might inhibit parasite ad-herence and consequently infection.
Adhesion of pathogens to host cells is the first step in inva-sion of all infectious disease pathogens. For instance, Ent-amoeba histolytica, Trichomonas sp., and Trypanosoma cruziattach to mammalian cells via specific adhesins (2, 6, 10, 28).Similar mechanisms of adherence have been observed for He-licobacter pylori, Actinomyces naeslundii, and Pseudomonasaeruginosa (3, 7, 37).
In order to focus on the initial adhesion of T. soliumoncospheres, an in vitro adhesion assay model was devel-oped to measure T. solium oncosphere adhesion using threedifferent substrates: (i) porcine intestinal mucosal scrapings
* Corresponding author. Mailing address: Department of Interna-tional Health, Johns Hopkins University, Bloomberg School of Hy-giene and Public Health, 615 N. Wolfe Street, Room W#5515, Balti-more, MD 21205. Phone: (410) 614-3959. Fax: (410) 614-6060. E-mail:[email protected].
(PIMS) (cells plus mucin), (ii) porcine small intestinal mu-cosal explants (PSIME), and (iii) monolayer cells, includingChinese hamster ovary (CHO) cells (CHO-K1 cells), epi-thelial cells from ileocecal colorectal adenocarcinoma(HCT-8 cells), and epithelial cells from colorectal adeno-carcinoma (Caco-2 cells).
MATERIALS AND METHODS
T. solium oncosphere preparation. Adult T. solium tapeworms were collectedfrom newly diagnosed patients as previously reported (18, 36). Species weredifferentiated by histology (parasite morphology) and PCR-restriction fragmentlength polymorphism analysis (PCR with restriction enzyme analysis) (22). Theproglottids were collected by sieving, washed thoroughly with distilled water, andstored in 25% glycerol supplemented with penicillin (1,000 U/ml), gentamicin(100 �g/ml), streptomycin (1 mg/ml), and amphotericin B (20 �g/ml) at 4°C untilthey were used. The eggs were obtained from gravid proglottids by gentle ho-mogenization (using a manual homogenizer) in water. Eggs were then washedthree times in distilled water by centrifugation at 1,575 � g prior to hatching. Invitro hatching of oncospheres was performed using 0.75% sodium hypochlorite(Mallinckrodt Baker, Inc., Phillipsburg, NJ) in water for 10 min at 4°C aspreviously described (20, 33, 35). Oncospheres were then washed three times inRPMI 1640 medium by centrifugation at 1,575 � g and activated with artificialintestinal fluid (1 g pancreatin [Sigma Chemical Co., St. Louis, MO], 200 mgNa2CO3, and 1 ml pig bile, with the volume adjusted to 100 ml with RPMI 1640medium [pH 8.04]) by incubation at 37°C for 1 h (23). After activation, onco-spheres were washed three times in RPMI 1640 medium, resuspended in RPMI1640 medium, and counted by microscopy (magnification, �40) using 0.4%trypan blue (Sigma Chemical Co.) to determine oncosphere viability. Onco-spheres were counted using a Neubauer chamber (Fisher Scientific, Atlanta,GA). We defined nonactivated oncospheres as oncospheres that were immobileand had centrally located hooks. In contrast, we designated oncospheres acti-vated if they showed active motility. This motility was characterized by frequentand cyclic movements of all three pairs of hooks and undulating movement of theoncosphere body. Before activated oncospheres were used in adherence assaymodels, they were incubated at 37°C in binding medium (RPMI 1640 mediumcontaining 25 mM HEPES, 25 �g/ml gentamicin, and 10% fetal bovine serum[pH 6.8]) for 20 min.
Development of in vitro oncosphere adherence assay models. In vitro adhesionassay models were developed with three different substrates: PIMS (cells plusmucin), PSIME, and CHO-K1, Caco-2, and HCT-8 monolayer cells. Our goalwas to determine which model was best suited to assessing oncosphere adher-ence. All of the adhesion studies were repeated three or four times on 2 or 3separate days. The tapeworms used were from different patient sources on eachday.
(i) PIMS. A 10-cm portion of duodenum was collected from a pig that was 6to 9 months old immediately postmortem and was maintained at 4°C. It was thenwashed with cold phosphate-buffered saline (PBS) (0.01 M dibasic sodium phos-phate, 0.01 M monobasic sodium phosphate, 0.15 M NaCl; pH 7.2) to remove theintestinal contents. The small intestine was scraped, and the mucosal material(intestinal cells and mucin) was recovered and diluted in PBS. It was then fixedon eight-well slides at room temperature before oncospheres were added. Acti-vated oncospheres were added to slides (n � 1,000) in binding medium andincubated at 37°C for 90 min. The slides were then washed in binding mediumthree times in order to remove oncospheres that had not bound to PIMS, werefixed in 1% glutaraldehyde, and were stained with periodic acid-Schiff stain(PAS) and hematoxylin and eosin stain (H&E). T. solium oncospheres wereidentified by their morphological characteristics, size, hooks, and secretory ves-icles using light microscopy (magnification, �100).
(ii) PSIME. A 10-cm portion of duodenum was collected from pigs immedi-ately postmortem and washed with cold PBS to remove the intestinal contents.The duodenum was placed in a tube of Earl’s balanced salts (Invitrogen, GrandIsland, NY) containing penicillin (100 U/ml), streptomycin (100 �g/ml), andamphotericin B (20 �g/ml) and transported to the laboratory at 4°C (within 30min). Samples were cut into full-thickness intestinal explants (approximately 0.7by 0.7 cm) in a sterile 24-well cell culture dish. Explants were positioned with theepithelial surface up on a rectangular piece of extrathick filter paper (Bio-Rad,Hercules, CA.). This filter paper was placed in a sterile 24-well cell culture dishcontaining tissue culture medium consisting of Dulbecco’s modified Eagle’smedium, a high concentration of glucose (Invitrogen), and 10% inactivated fetalbovine serum to which penicillin (100 U/ml) and streptomycin (100 �g/ml) wereadded. The level of the medium was adjusted to allow saturation of the filter
paper, which rested on a wire mesh platform. Each piece of intestine wasinoculated with 3,000 activated T. solium oncospheres in 8 �l RPMI 1640 me-dium and incubated for 1.5 h at 37°C in an atmosphere consisting of 95% oxygenand 5% carbon dioxide. Control explants, which did not contain oncospheres,were treated in the same way. At the end of the incubation period, explants werewashed three times in tissue culture medium at 37°C and then either fixed in 10%formalin at room temperature or frozen slowly in a �70°C freezer.
Formalin-fixed explants were examined using histochemistry to visualize on-cosphere adherence to host cells by light microscopy (magnification, �40).Three-micrometer sections were fixed on slides and stained with PAS and H&E.
Frozen explants were sectioned and stained using immunohistochemistry byusing pig polyvalent antibodies against T. solium oncospheres to visualize onco-sphere adherence to host cells by UV microscopy. Polyvalent antibodies againstT. solium oncospheres were obtained by subcutaneous immunization of pigs with2 ml (0.028 mg/ml) of T. solium oncosphere extract antigens. Immunization wasperformed twice with a 15-day interval between the immunizations (36). Theinitial immunization consisted of oncosphere extract antigen emulsified inFreund’s complete adjuvant (Sigma Chemical Co.). The second immunizationconsisted of the same quantity of antigen emulsified in Freund’s incompleteadjuvant. Sera were collected 15 days after the second immunization as describedby Verastegui et al. (35). Oncospheres that had attached to PSIME were placedin OCT compound (Sakura Finetek, Torrance, CA) and frozen, and 8-�m sec-tions were cut with a cryostat microtome. Sections on slides were then incubatedwith pig hyperimmune sera diluted 1:25 in PBS with 1% ovalbumin at 4°Covernight. After three washes with PBS with 1% ovalbumin, the slides wereincubated for 1 h at room temperature with goat anti-pig immunoglobulin Glabeled with fluorescein isothiocyanate conjugate (Kirkegaard & Perry Labora-tories, Gaithersburg, MD) diluted 1:50 in PBS with Evans blue (0.5%). The slideswere then washed three times with PBS, mounted with buffered glycerin (pH7.2), and examined by UV microscopy (magnification, �100).
(iii) CHO-K1 cell monolayer. Established lines of adult CHO cells (CHO-K1)were obtained from ATCC (Manassas, VA). Cultures were routinely maintainedin Ham’s F-12K medium (GIBCO Laboratories, Grand Island, NY) supple-mented with 10% fetal bovine serum and 40 �g/ml gentamicin. Cells wereincubated under 5% CO2 in air at 37°C. The medium was changed every 2 days,and cells were passaged when a monolayer reached confluence. CHO cells wereharvested using trypsin-EDTA treatment (Sigma Chemical Co.) after a completemonolayer was formed in a culture flask. The cells were then seeded into aneight-well chamber slide system with 0.7-cm2 wells (Nalgene-Nunc). The T.solium oncosphere adherence assay was performed using confluent CHO cellmonolayers in eight-well chamber slides, as previously described (6, 21, 29, 32).
(a) Comparison of T. solium oncosphere adherence to viable and fixedCHO-K1 cells. Viable monolayer cells were used immediately after monolayerformation in a chamber slide. The number of oncospheres adhering to viablemonolayer CHO-K1 cells was compared to the number of oncospheres adheringto fixed monolayer CHO-K1 cells after 48 h of cell culture. The fixed monolayerCHO cells were treated with 1% formaldehyde for 1 h and subsequently storedin PBS at 4°C until they were used (34). Viable and fixed monolayer cells wereincubated with 2,500 T. solium activated oncospheres in binding medium at 37°Cfor 1.5 h. After incubation, the unbound oncospheres were rinsed three timeswith binding medium, and the oncospheres bound to the cells were fixed with 1%glutaraldehyde in PBS and stained with PAS. Adherent oncospheres were thencounted by light microscopy (magnification, �100).
(b) Standardization of optimum time of oncosphere incubation for adherence.Forty-eight-hour viable monolayer CHO cells maintained at 37°C with 5% CO2
were incubated with 2,500 activated oncospheres in binding medium for differenttimes (30 min and 1.5, 3, and 24 h) to determine the optimum time of oncosphereincubation to obtain the highest number of adherent oncospheres.
(c) Role of sera and cell maturation in oncosphere adherence. The effect ofsera on oncosphere adherence in the CHO cell model was examined. Viablemonolayer CHO-K1 cells were tested at different times during monolayer mat-uration (24, 48, and 72 h) in the chamber slide system. Monolayer cells wereincubated with 2,500 activated oncospheres in RPMI medium both with andwithout 10% fetal bovine serum (RPMI 1640 containing 25 mM HEPES and 25�g/ml of gentamicin, pH 6.8) at 37°C with 5% CO2. Oncosphere adhesion wasevaluated after 30 min and 1 h of incubation.
The effect of varying the fetal bovine serum concentration (1, 2.5, 5, 7.5, 10, 15,and 20%) on oncosphere adherence was also evaluated using viable monolayerCHO-K1 cells. The number of oncospheres bound to monolayer cells was de-termined by PAS staining and light microscopy (magnification, �100).
(d) Effect of temperature on oncosphere adherence. The effect of temperatureon oncosphere adhesion to the CHO cells was examined at different incubationtemperatures. Viable monolayer cells were incubated with 2,500 T. solium acti-
vated oncospheres in binding medium at 4, 12, 24, and 37°C for 30 min and 1 h.After incubation, slides were rinsed three times with binding medium to removeunbound oncospheres. Oncospheres bound to the cells were fixed with 1%glutaraldehyde in PBS and stained with PAS.
(e) Scanning electron microscopy (SEM). Activated T. solium oncospheresthat were adherent to formalin-fixed monolayer CHO cells were incubatedfor 2 h with 1% glutaraldehyde in PBS (pH 7.2) plus 1% sucrose. Thespecimens were then washed three times with 1% glutaraldehyde in PBS (pH7.2) plus 1% sucrose and stored at 4°C until they were used. The preparations
were postfixed in 1.5% osmium tetroxide, and serial dehydration was per-formed using 70, 95, and 100% ethanol (2 h each). Samples were thenprocessed in a critical point bomb using liquid carbon dioxide as a transitionalfluid and then sputter coated using a Denton S-II instrument with a gold-palladium target. Images were obtained using a Hitachi 2400 microscope at15 kV.
(f) Transmission electron microscopy. Pellets of activated oncospheres andactivated oncospheres adhering to PSIME were processed as described above forSEM, and then samples were postfixed in 1.5% osmium tetroxide, dehydrated in
FIG. 1. (Panel i) T. solium oncospheres adhering to PIMS. Oncosphere adhesion was visualized by light microscopy using PAS. Magnification,�40. Arrows A, oncospheres adhering to PIMS; arrows B, oncosphere hooks. (Panel ii) T. solium oncosphere binding to viable monolayer CHOcells. The arrows indicate secretory vesicles. Magnification, �40. (Panel iii) T. solium oncosphere binding to fixed monolayer CHO cells visualizedusing PAS. The arrows indicate one pair of penetration glands. Magnification, �100. (Panel iv) Light microscopy of 3-�m porcine small intestinalsection embedded in paraffin and stained with H&E (20-min incubation period). An oncosphere (arrow) is interacting with epithelial cells.Magnification, �400. (Panels v and vi) Photographs of frozen 8-�m sections, obtained using light and UV microscopy, showing oncospheres(arrows) identified by the presence of hooks and immunofluorescence (20-min incubation period; magnification, �100). A light microscopyphotograph with Evans blue (panel v) and a UV microscopy photograph (panel vi) show the same field of view. Material stained with a contrastagent appears to be blue under visible light and red under UV light, while fluorescence is green under UV light.
FIG. 2. Ninety-five percent confidence intervals for oncospheres adherent on unfixed monolayer CHO cells after 30 min and 1.5, 3, and 25 hof incubation at 37°C.
alcohol, and embedded in epoxy, and 50-nm sections were cut. Sections werethen stained serially with lead hydroxide and uranyl acetate and were examinedusing a Phillips electron microscope (Phillips Electronic Instruments, Eindhoven,The Netherlands) operating at 75 kV.
(iv) Detection of oncosphere interaction with microvilli by immunofluores-cence. Established cell lines HCT-8, Caco-2, and CHO were used to determinethe presence of microvilli in adherent oncospheres. CHO, HCT-8, and Caco-2cells were obtained from ATCC (Manassas, VA). CHO cells were cultured asdescribed above. HCT-8 cells were cultured in RPMI 1640 medium with 10%equine serum. Caco-2 cell cultures were routinely maintained in Eagle minimumessential medium with 20% fetal bovine serum. HCT-8, Caco-2, and CHO cellswere incubated under 5% CO2 in air at 37°C. The HCT-8, Caco-2, and CHO cellswere harvested using trypsin-EDTA treatment (Sigma Chemical Co.) after acomplete cell monolayer was formed in a culture flask. The cells were thenseeded into an eight-well chamber slide system with 0.7-cm2 wells (Nalgene-Nunc). The T. solium oncosphere adherence assay was performed using conflu-ent CHO, HCT-8, and Caco-2 cell monolayers in eight-well chamber slides, asdescribed above. Viable monolayer cells were incubated with 3,000 T. soliumactivated oncospheres in binding medium at 37°C for 1.5 h. After incubation,unbound oncospheres were rinsed three times with binding medium, and the
oncospheres bound to the cells were fixed with 50% acetone and methanol andstained by immunofluorescence using rabbit polyvalent antibodies against T.solium oncospheres to visualize oncosphere adherence to monolayer cells usingUV microscopy (magnification, �40).
Polyvalent antibodies against T. solium oncospheres were obtained by subcu-taneous immunization of rabbits with 2 ml (0.028 mg/ml) of T. solium activatedoncosphere extract antigens. Immunization was performed four times, with thesecond immunization 15 days after the first immunization and with the subse-quent immunizations at 1-week intervals. The initial immunization was withoncosphere extract antigen emulsified in Freund’s complete adjuvant (SigmaChemical Co.). The subsequent immunizations were with the same quantity ofantigen emulsified in Freund’s incomplete adjuvant. Sera were collected 15 daysafter the fourth immunization.
Slides containing fixed monolayer cells with adherent oncospheres were thenincubated with rabbit hyperimmune sera diluted 1:100 in PBS with 1% oval-bumin and 0.05% Tween 20 (PBS-TO) at room temperature for 1 h. After threewashes with PBS-TO, the slides were incubated for 1 h at room temperature withgoat anti-rabbit immunoglobulin G labeled with fluorescein isothiocyanate con-jugate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) diluted 1:100 inPBS-TO plus Evans blue (0.5%). The slides were then washed three times with
FIG. 3. (a) Ninety-five percent confidence intervals for the number of oncospheres adhering on unfixed CHO-K1 cells after 1.5 h of incubationat 37°C with binding medium with serum and binding medium without serum (RPMI), obtained using 24-, 48-, and 72-h mature monolayer CHOcells. (b) Curve and 95% confidence intervals for the number of oncospheres adhering on unfixed CHO-K1 cells after 1.5 h of incubation at 37°Cwith binding medium with different fetal bovine serum concentrations (0, 1, 2.5, 5, 7.5, 10, 15, and 20%), determined with 48-h mature monolayerCHO-K1 cells.
PBS, mounted with buffered glycerin (pH 7.2), and examined by UV microscopy(magnification, �100).
Measurement of T. solium oncosphere adherence in each model. In order toallow comparisons between models, the proportion of adherent oncospheres wasestimated by dividing the number of oncospheres adhering to cells by the totalnumber of oncospheres added to the monolayer surface. Means, distributions,and confidence limits (� � 0.05) of adherent oncospheres were estimated foreach model.
RESULTS
Development of in vitro oncosphere adherence assay model. (i)PIMS. The percentage of oncosphere activation was between 50and 75% after incubation in sodium hypochlorite and artificialintestinal fluids. After treatment with the artificial intestinal flu-ids, the hooks protruded from the cell and the oncosphere mem-brane was shed. An activated oncosphere was characterized by itsmotion and movement of its hooks; some hooks projected outside
the oncosphere (Fig. 1, panel i). Also, single or multiple onco-sphere secretory vesicles appeared outside the oncosphere adja-cent to the oncosphere surface (Fig. 1, panel ii).
The overall percentage of oncospheres adhering to PIMSranged from 2 to 37%. In this model, it was important to havea good-quality stain to identify the oncospheres on pig intes-tinal mucosa. PAS gave better resolution than H&E, and iden-tification of the hooks and other oncosphere morphologicalcharacteristics was easier (Fig. 1, panels i and iii).
(ii) PSIME. Initially, we assessed porcine small intestinetissue culture survival time. The histological cellular appear-ance of the epithelium and subepithelial tissue throughout thisperiod was relatively normal up to a culture time of 4.5 h; afterthis cell degeneration occurred. Thus, oncosphere adherencestudies were not done after 4.5 h of culture.
When activated oncospheres were incubated with explants, a
FIG. 4. Curve and 95% confidence intervals for the number of oncospheres adhering on unfixed CHO-K1 cells after 1.5 h of incubation atdifferent temperatures.
FIG. 5. Scanning electron micrographs of T. solium oncospheres adhering to fixed monolayer CHO-K1 cells. (a) T. solium oncospheres attachedto the surface of CHO cells by elongated microvilli. (b) Secretory vesicles (arrows) that are present outside the oncosphere membrane. Microvilliare on the surface membrane of the oncosphere and secretory vesicles. (c) Elongate microvilli that are attached to the surface of CHO cells. Hooksare present outside the oncosphere.
number of oncosphere forms (up to 10) were observed adher-ing to the epithelial lining of the intestine following histologicalsectioning of both formalin-fixed explants (Fig. 1 panel iv) andfrozen explants (Fig. 1, panels v and vi). The appearance ofsections and oncospheres was demonstrated by H&E staining(Fig. 1, panel iv).
In frozen sections treated with anti-T. solium oncosphereantibodies, both oncospheres and the surrounding materialwere visualized by immunofluorescence (Fig. 1, panel vi). Add-ing Evans blue stain to the conjugate material permitted visu-alization of oncosphere hooks within the oncospheres (Fig. 1,panel v).
(iii) Monolayer CHO-K1 cells. (a) T. solium oncosphereadherence to viable and fixed CHO-K1 cells. Activated onco-spheres adhered to both viable and fixed monolayer CHO cells(Fig. 1, panels ii and iii). The number of oncospheres bound toviable monolayer CHO cells was higher than the numberbound to fixed monolayer cells (means, 386 � 102 and 78 � 57oncospheres, respectively; P � 0.05). Oncosphere activationwas important for adhesion to monolayer cells since 95% ofadherent oncospheres were activated.
(b) Standardization of optimum time of oncosphere incuba-tion for adherence. Incubation for 1.5 h yielded the highestnumber of adherent oncospheres on viable monolayer CHOcells (Fig. 2).
(c) Role of serum and cell maturation in oncosphere adher-ence. As Fig. 3a shows, the use of binding medium with 10%fetal bovine serum resulted in a significantly greater number ofadhering oncospheres than the use of binding medium withoutserum (P � 0.05) at all times during monolayer cell matura-tion. When binding medium with serum was used, the opti-mum time of monolayer cell maturation was 48 h. Theadherence of oncospheres increased linearly as the serum con-centration in the binding medium increased from 1 to 5% andthen plateaued (Fig. 3b).
(d) Effect of temperature on oncosphere adherence. Thenumber of oncospheres adhering to viable monolayer cellsincreased when the incubation temperature was increasedfrom 12 to 37°C (Fig. 4). The numbers of adherent onco-spheres were statistically significantly different at 12, 24, and37°C (P � 0.05) (Fig. 4). The adherent oncospheres had visiblesecretory vesicles, and the numbers of these vesicles increasedas the incubation temperature increased. Secretory vesicles(Fig. 5b) were seen in about 80 to 85% of adherent onco-spheres incubated at 37°C but in only 10% of adherent onco-spheres incubated at 12 or 4°C. The sizes of the secretoryvesicles varied from 2.07 to 3.71 �m (median, 2.65 �m).
(e) SEM. SEM showed that the surface membrane or epi-thelium of oncospheres adhering to CHO cells had elongatedmicrovilli. The elongated microvilli extended and appeared toattach to the surface of the monolayer cells (Fig. 5a, 5b, and5c). Also, these microvilli were present on the secretory vesiclemembrane, similar to surface membranes of the oncosphere(Fig. 5b). The secretory vesicles appeared to be attached toboth the CHO cells and the oncospheres via their elongatedmicrovilli.
(f) Transmission electron microscopy. Transmission elec-tron microscopy showed that the first microvilli were formedunder the oncosphere membrane of the activated oncosphere
(Fig. 6a). Microvilli appeared to attach to the pig intestinalcells using the PSIME model (Fig. 6b).
(iv) Detection of oncosphere microvilli by immunofluores-cence. As determined by UV microscopy, the surface mem-branes of oncospheres adhering to CHO, HCT-8, and Caco-2cells all reacted strongly with hyperimmune rabbit sera against
FIG. 6. (a) Transmission electron microscopy of activated T. soliumoncosphere. The arrows indicate microvilli (MV) that are under theoncosphere membrane (OM). (b) Transmission electron microscopy ofactivated T. solium oncosphere (Onc) adhering to pig small intestinalcell (IC). The large arrow indicates the microvilli (MV).
T. solium activated oncospheres (Fig. 7). Oncospheres, whentested with each cell line, had elongated microvilli and secre-tory vesicles (Fig. 7a and 7b). Also, it was possible to differ-entiate between activated oncospheres with an oncospheremembrane and activated oncospheres without an oncospheremembrane (Fig. 7c). Microvillar fragments adhering to mono-layer cells were seen 50 �m from the oncosphere surface(Fig. 7a).
DISCUSSION
The present study showed that adherence of activated T.solium oncospheres can be observed in vitro and can be mea-sured objectively using the following three different substrates:PIMS (cells plus mucin), PSIME, and monolayer cells (CHO-K1, Caco-2, and HCT-8 cells). The reproducibility of T. soliumoncosphere adherence was most easily measured with CHOcells. Serum and an increased incubation temperature en-hanced adherence. Our studies also demonstrated that surfacemembranes of adherent T. solium oncospheres had elongatedmicrovilli that attached to tissue cultures and expelled externalsecretory vesicles with the same elongated microvilli. Theseelongated microvilli appeared to attach to the CHO cells.
The PIMS model was used because the mechanism of on-cosphere adherence resembled natural infection as it occurs onthe porcine small intestinal mucosal surface. This model hasthe advantage that fresh tissue is not needed for each experi-ment. Rather, the mucosal material can be fixed on slides andused for multiple studies. The PIMS model includes both in-testinal epithelial cells and mucin. In this model, the mucin andthe epithelial cells may have different receptors that cannot bedistinguished. Also, the detection of oncospheres by staining issomewhat difficult due to high background staining.
The PSIME model resembles the natural infection that oc-curs in pigs, but it has the disadvantage that it requires freshporcine small intestinal mucosa for each experiment. Nonethe-less, the PSIME model is useful for determining immunohis-tological characteristics of early host-oncosphere interactions.
The third model, using monolayer CHO cells, is valuablesince this cell line has glycosylated mutants, which can be usedin future studies to characterize the oncosphere lectin or cellcarbohydrate receptors that are involved in the adherencemechanisms. CHO cells have been employed to investigateattachment of other parasites, including E. histolytica andTrichomonas (6, 28). However, monolayer CHO-K1 cells maynot accurately mimic the natural process of adherence.
T. solium oncosphere adhesion to monolayer CHO cells wasenhanced by the presence of serum in binding medium, indi-cating that unidentified serum factors play a primary role in theoncosphere-host cell interaction. Similarly, the adherence ofEchinococcus multilocularis and Echinococcus granulosus pro-toscoleces to human endothelial cell monolayers was enhancedby sera (19). One of the serum factors known to play a leadingrole in cell-cell and cell-substratum interactions is fibronectin(27). It has been shown that fibronectin accumulates in areas ofcell-cell contact and promotes the adhesion between certaintissues and cells (for example, fibroblasts and collagen substra-tum) (8, 26). In addition, there are other known or putativemodulators of adhesion, such as proteoglycans, various colla-gen types, laminin, and vitronectin, that require furtherexamination (30).
The movement and morphology of activated T. solium on-cospheres are similar to the movement and morphology de-scribed previously for other cestodes (5, 23, 25, 31). This studydemonstrates that secretory vesicles are expelled outside theactivated T. solium oncospheres upon activation. These secre-tory vesicles may be important either for attachment to hosttissues, to facilitate penetration through the epithelium, or tohelp protect the oncospheres against digestive enzymes. How-ever, these secretory vesicles have been found previouslywithin oncospheres (16, 31). Our study is the first study show-ing secretory vesicles outside the oncosphere body. Secretoryvesicles can be seen outside oncospheres with and without thepresence of cells; however, activation of the oncospheres isrequired.
The formation of secretory vesicles that we observed may be
20 µ m20 µ m 20 µ m20 µ m 20 µ m20 µ m
a b c
FIG. 7. UV microscopy (magnification, �100) of T. solium oncospheres adhering to viable monolayer CHO cells. (a and b) Activated T. soliumoncospheres without oncosphere membrane attached to the surface of CHO cells by elongated microvilli. The arrows indicate the elongatedmicrovilli and secretory vesicles that are present outside the oncosphere membrane. (c) Activated T. solium oncosphere with oncospheremembrane adhering to a viable monolayer CHO cell.
somewhat similar to that reported for Taenia saginata, inwhich, after oncosphere activation, changes in the compositionof the cell cytoplasm of penetration glands can be observed byelectron microscopy. When the oncosphere membrane waseliminated, the vesicles could be seen on the surface of theoncosphere (31). Other authors have mentioned that the se-cretion of these vesicles is initiated and maintained by themuscular contractions of the oncosphere (9). Our findings,however, differ from those of previous reports since we did notfind the secretory vesicles separated from the oncosphere aspart of the maturation process.
Microvilli were previously described for E. granulosus (16,17), Taenia taeniaeformis (11), T. saginata (31), and Taenia ovisoncospheres (15) using transmission microscopy. However, inthe presence of CHO, HCT-8, and Caco-2 cells, the T. soliumoncospheres developed elongated microvilli that attached tothe tissue culture cells. These T. solium oncospheres with elon-gated microvilli appear to have been described previously byEngelkirk and Williams but without reference to their function(11). These authors noted that in rats infected orally with T.taeniaformis eggs, oncospheres were present in the liver 24 hpostinfection, with long microvilli that appear to be similar towhat we described. In 1987, Harris et al. also showed that after48 h of in vitro incubation in tissue culture T. ovis oncosphereshad elongated microvilli (15). Our model of T. solium onco-sphere infection in tissue culture differs in that microvilli havebeen found in association with both the oncospheres and theirsecretory vesicles.
In addition, our model strongly supports the notion thatelongated microvilli are involved in the adherence of onco-spheres to host cells. What remains to be determined is theprecise timing of the formation of the microvilli (i.e., uponstimuli received from host cells or upon direct contact). Elon-gated microvilli were shown to have a strong reaction withhyperimmune sera against oncospheres, showing that microvil-lus antigens are more immunogenic. Inhibition of this adher-ence mechanism may provide a useful method to block T.solium infection. Also still to be resolved is the precise role(s)of the secretory vesicles. Do they play a significant role inallowing the oncosphere access into and possibly through hostintestinal tissues?
This study demonstrated three models for T. solium onco-sphere adherence. It also demonstrated that immunofluores-cence is a useful tool for observing differentially activated on-cospheres with and without oncosphere membranes. The CHOcell model has the advantage of being easily modified and isable to assess different individual variables important for ad-herence. This model should now permit us to characterize inmore detail the mechanism of T. solium oncosphere adher-ence.
ACKNOWLEDGMENTS
Funding for this project came from Gates project 23981, NIHGlobal Research training grant D43 TW006581, Wellcome Trust grantGR063109MA, and the RG-ER Anonymous Fund for Tropical Re-search.
We thank Charles Sterling, Paula Maguina for administrative help,and J. B. Phu, D. Sara, and Sari-CeCe for technical assistance.
REFERENCES
1. Acha, P., and B. Cifres. 1986. Zoonosis y enfermedades transmisiblescomunes al hombre y los animals, 2nd ed., p. 763–774. Pan American HealthOrganization, Washington, DC.
2. Alderete, J. F., and G. E. Garza. 1988. Identification and properties of Trichomo-nas vaginalis proteins involved in cytoadherence. Infect. Immun. 56:28–33.
3. Aspholm-Hurtig, M., G. Dailide, M. Lahmann, A. Kalia, D. Ilver, N. Roche,S. Vikstrom, R. Sjostrom, S. Linden, A. Bakstrom, et al. 2004. Functionaladaptation of BabA, the H. pylori ABO blood group antigen binding adhe-sion. Science 305:519–522.
4. Bern, C., H. H. Garcıa, C. Evans, A. E. Gonzalez, M. Verastegui, V. C. W. Tsang,and R. H. Gilman. 1999. Magnitude of the disease burden from neurocysticer-cosis in a developing country. Clin. Infect. Dis. 29:1203–1209.
5. Bilqees, F. M. 1968. Observations on the structure of oncosphere of Taeniacrassiceps (Zeder, 1800) Rudolphi, 1810. Can. J. Zool. 46:763–765.
6. Bonilha, V. L., C. Ciavaglia Mdo, W. Souza, F. Costa, and F. Silva. 1995. Theinvolvement of terminal carbohydrates of the mammalian cell surface in thecytoadhesion of trichomonads. Parasitol. Res. 81:121–126.
7. Brennan, M. J., J. O. Cisar, A. E. Vatter, and A. L. Sandberg. 1984. Lectin-dependent attachment of Actinomyces naeslundii to receptors on epithelialcells. Infect. Immun. 46:459–464.
8. Chen, L. B., N. Maitland, P. H. Gallimore, and J. K. McDougall. 1977.Detection of the large external transformation-sensitive protein on someepithelial cells. Exp. Cell Res. 106:39–46.
9. Chew, M. W. K. 1983. Taenia crassiceps: ultrastructural observations on theoncosphere and associated structures. J. Helminthology 57:101–103.
10. Ciavaglia Mdo, C., T. de Carbalho, and W. de Souza. 1993. Interaction ofTrypanosoma cruzi with cells with altered glycosylation patterns. Biochem.Biophys. Res. Commun. 193:718–721.
11. Engelkirk, P. G., and J. F. Williams. 1982. Taenia taeniaeformis (Cestoda) inthe rat: ultrastructure of the host-parasite interface on days 1 to 7 postin-fection. J. Parasitol. 68:620–633.
12. Flisser, A. 1988. Neurocysticercosis in Mexico. Parasitol. Today 4:131–137.13. Garcia, H. H., M. Martinez, R. H. Gilman, G. Herrera, V. C. W. Tsang, J. B.
Pilcher, et al. 1991. Diagnosis of cysticercosis in endemic regions. Lancet338:549–551.
14. Garcia, H. H., R. Gilman, M. Martinez, V. C. Tsang, J. B. Pilcher, G.Herrera, F. Diaz, M. Alvarado, and E. Miranda. 1993. Cysticercosis as amajor cause of epilepsy in Peru. The Cysticercosis Working Group in Peru(CWG). Lancet 341:568.
15. Harris, A., D. D. Heath, S. B. Lawrence, and R. J. Shaw. 1987. Ultrastructureof changes at the surface during the early development phases of Taenia oviscysticerci in vitro. Int. J. Parasitol. 17:903–910.
16. Holcman, B., D. D. Heath, and R. J. Shaw. 1994. Ultrastructure of onco-sphere and early stages of metacestode development of Echinococcus granu-losus. Int. J. Parasitol. 24:623–635.
17. Holcman, B., and D. D. Heath. 1997. The early stages of Echinococcusgranulosus development. Acta Trop. 64:5–17.
18. Jeri, C., R. H. Gilman, A. G. Lescano, H. Mayta, M. E. Ramirez, A. E.Gonzalez, R. Nazerali, and H. H. Garcia. 2004. Species identification aftertreatment for human taeniasis. Lancet 363:949–950.
19. Kirkpatrick, C. J., and D. Svilenov. 1987. Studies on the adhesion of pro-toscoleces from Echinococcus multilocularis and E. granulosus to artificialsubstrates and human endothelial cells in vitro. Parasitol. Res. 74:61–68.
20. Laws, G. F. 1968. The hatching of taeniid eggs. Exp. Parasitol. 23:1–10.21. Li, E., A. Becker, and S. L. Stanley. 1988. Use of Chinese hamster ovary cells
with altered glycosylation patterns to define the carbohydrate specificity ofthe Entamoeba histolytica adhesion. J. Exp. Med. 167:1725–1730.
22. Mayta, H., A. Talley, R. H. Gilman, J. Jimenez, M. Verastegui, M. Ruiz,H. H. Garcia, and A. E. Gonzales. 2000. Differentiating Taenia solium andTaenia saginata infection by simple hematoxylin-eosin staining and PCR-restriction enzyme analysis. J. Clin. Microbiol. 38:133–137.
23. Molinari, J. L., P. Tato, R. Lara-Aguilera, and C. White. 1993. Effects ofserum from neurocysticercosis patients on the structure and viability ofTaenia solium oncospheres. J. Parasitol. 79:124–127.
24. Nash, T. E., and F. A. Neva. 1984. Recent advances in the diagnosis andtreatment of cerebral cysticercosis. N. Engl. J. Med. 311:1492–1496.
25. Ogren, R. E. 1961. The mature oncosphere of Hymenolepis diminuta. J.Parasitol. 47:197–204.
26. Pearlstein, E. 1976. Plasma membrane glycoprotein which mediates adhe-sion of fibroblasts to collagen. Nature 262:497–500.
27. Pearlstein, E., L. I. Gold, and A. Garcia-Pardo. 1980. Fibronectin: a reviewof its structure and biological activity. Mol. Cell. Biochem. 29:103–128.
28. Ravdin, J. I., and R. L. Guerrant. 1981. The role of adherence in cytopatho-genic mechanisms of Entamoeba histolytica. Study with mammalian tissueculture cells and human erythrocytes. J. Clin. Investig. 68:1305–1313.
29. Ravdin, J. I., P. Stanley, C. F. Murphy, and W. A. Petri. 1989. Character-ization of cell surface carbohydrate receptor for Entamoeba histolytica ad-herence lectin. Infect. Immun. 57:2179–2186.
30. Ruoslahti, E., E. G. Hayman, and M. D. Pierschbacher. 1985. Extracellularmatrices and cell adhesion. Arteriosclerosis 5:581–594.
31. Schramlova, J., and K. Blazer. 1982. Ultrastructure of the hatched andunhatched oncospheres of Taenia saginata. Folia Parasitol. 29:45–50.
32. Stanley, S. M., J. L. Blanchard, N. Johnson, L. Foster, C. Kunz-Jenkins, T.Zhang, K. Tian, and F. B. Cogswell. 1995. Immunogenicity of the recombi-nant serine rich Entamoeba histolytica protein (SREHP) amebiasis vaccine inthe African green monkey. Vaccine 13:947–951.
33. Stevenson, P. 1983. Observations on the hatching and activation of freshTaenia saginata eggs. Ann. Trop. Med. Parasitol. 77:399–404.
34. Treutiger, C. J., J. Carlson, C. Scholander, and M. Wahlgren. 1998. Thetime course of cytoadhesion, immunoglobulin binding, rosette formation,and serum-induced agglutination of Plasmodium falciparum-infected eryth-rocytes. Am. J. Trop. Med. Hyg. 59:202–207.
35. Verastegui, M., R. H. Gilman, A. E. Gonzalez, H. H. Garcia, C. Gavidia, N.Falcon, T. Bernal, Y. Arana, V. C. Tsang, and Cisticercosis Working Groupin Peru. 2002. Taenia solium oncosphere antigens induce immunity in pigsagainst experimental cysticercosis. Vet. Parasitol. 108:49–62.
36. Verastegui, M., R. H. Gilman, H. H. Garcia, A. E. Gonzalez, Y. Arana, C.Jeri, I. Tuero, C. M. Gavidia, M. Levine, V. C. Tsang, and CisticercosisWorking Group in Peru. 2003. Prevalence of antibodies to unique Taeniasolium oncosphere antigens in taeniasis and human and porcine cisticercosis.Am. J. Trop. Med. Hyg. 69:438–444.
37. Vishwanath, S., and R. Ramphal. 1985. Tracheobronchial mucin receptor forPseudomonas aeruginosa: predominance of amino sugars in binding sites.Infect. Immun. 48:331–335.
