INFECTION AND IMMUNITY, Apr. 1983, p. 284-291 Vol. 40. No. 10019-9567/83/040284-08$02.00/0Copyright © 1983, American Society for Microbiology

Identification of Immunogenic and Antibody-BindingMembrane Proteins of Pathogenic Trichomonas vaginalis

JOHN F. ALDERETE

Department of Microbiology, The University of Texas Health Science Center at San Antonio, San Antonio,Texas 78284

Received 4 August 1982/Accepted 23 November 1982

Characterization of immunogenic Trichomonas vaginalis membrane proteinswas accomplished by using extrinsically and intrinsically labeled organisms and ahighly sensitive and specific radioimmunoprecipitation procedure. Intact motiletrichomonads were compared with detergent extracts as a source of antigen inradioimmunoprecipitation experiments. Approximately 20 proteins accessible toantibody were identified and ranged in molecular weight from 200,000 to 20,000.Localization on the parasite surface of the highly immunogenic membraneproteins was attempted by using, as the indicator system, formaldehyde-fixedprotein A-bearing Staphylococcus aureus pretreated with the various antiserumreagents and incubated with live, motile parasites. Also, indirect immunofluores-cence with fluorescein isothiocyanate-anti-rabbit immunoglobulin was also em-ployed after incubation of organisms with either control serum or antiserum fromimmunized rabbits or after treatment of trichomonads with the immunoglobulin Gfraction from each respective serum. No immunoglobulin G antibody appeared tobe directed at the anterior trichomonal flagella or the posterior axostyle, whereasstrong fluorescence was detected throughout the rest of the T. vaginalis surface.The biological significance of these data is discussed.

Trichomonal vaginitis is a sexually transmit-ted disease responsible for significant morbidityin both men and women. In symptomatic fe-males, the disease is characterized by pro-nounced inflammation of the vagina, foul-smell-ing vaginal discharge, severe discomfort, andtissue cytopathology (10, 13, 18, 20). Althoughmost men appear to be asymptomatic, diseasemanifestations are increasingly being docu-mented (19, 20). The causative agent, Trichomo-nas vaginalis, is the most frequently acquiredprotozoan infestation; however, few data areavailable on key areas of this host-parasite inter-action which would yield information on possi-ble mechanisms of disease pathogenesis.

Current clinical diagnosis of trichomoniasis istedious, time consuming, highly inadequate, andbased on microscopic detection of the parasite(10, 20). These limitations in diagnosis exacer-bate already limited medical care in rural healthclinics in our country and the world. Thus, basicresearch is necessary to address relevant issuessuch as development of sensitive, accurate sero-logical assays for screening of symptomatic aswell as asymptomatic patients and perhaps mon-itoring disease progression. The development ofstrategies to identify potential vaccinogen candi-dates is equally important and necessary.The emergence of trichomoniasis as a major

sexually transmitted disease (10, 13, 18, 20),therefore, has necessitated the need for researchdirected toward identification of possible viru-lence factors. The characterization of the sur-face of T. vaginalis has not been accomplished.Furthermore, the use of conventional immuno-logical methods has failed to identify specificvirulence determinants or antigens (10, 11, 21,22). Because one of our aims is to assess hostimmunological responses to parasite antigens.the nature of antibody responses from immu-nized or infected experimental animals directedtoward immunogenic parasite proteins was re-cently demonstrated (1). In this report, numer-ous highly immunogenic proteins which are ex-posed on the membranes of intact T. vagintalisand reactive with antibody while the membraneproteins are in their native state are identified.The importance of these data is discussed.

MATERIALS AND METHODSGrowth and radiolabeling of T. vaginalis organisms.

T. vaginalis 286 (16) was a generous gift from MiklosMuller, Rockefeller University, New York, N.Y. Par-asites were maintained in vitro in an air atmosphere asdescribed elsewhere (17) with Diamond Trypticase(BBL Microbiology Systems, Cockeysville, Md.)-yeast extract-maltose medium (pH 6.2) supplemented

284

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

IMMUNOGENIC T. VAGINALIS MEMBRANE PROTEINS 285

with 10% heat-inactivated horse serum (Kansas CityBiologicals, Inc., Lenexa, Kans.). Only trichomonadsat the late-log stage (24 h) of growth (17) were utilizedfor experiments, and they represented a density of 2.5x 10' to 5.0 x 106 actively motile parasites per ml.

Intrinsic radiolabeling with [35S]methionine (specificactivity, 1,500 Ci/mmol; Amersham Corp., ArlingtonHeights, Ill.) has been previously described (1, 3, 17).Pellets containing 1.4 x 107 to 3.5 x 107 trichomonadswere frozen at -70°C until use or employed immedi-ately for the radioimmunoprecipitation (RIP) assaywith whole cells as the antigen as described below.Electrophoretic protein patterns offrozen trichomonalpellets were identical to those of freshly harvested,radiolabeled organisms.

Specific radioiodination of membrane proteins wasaccomplished by standard chloramine-T labeling tech-niques (5). One hundred milliliters containing 2.5 x 108motile organisms was washed at least three times inphosphate-buffered saline (PBS) to remove serum pro-teins contaminating trichomonal surfaces (17). Tricho-monads were suspended to 1 ml in PBS followed bythe sequential addition of 3 mCi of Na125I (Amersham)and 100 ,ul of 0.4% chloramine-T (Sigma ChemicalCo., St. Louis, Mo.). After a 5-min incubation at roomtemperature with constant gentle stirring, 100 ,ul of0.4% NaHSO3 was added to terminate the reaction.Trichomonads remained motile under these experi-mental conditions and were washed at least threetimes further with phosphate-buffered Nal (same asPBS, except Nal was substituted for NaCl) beforealiquoting for freezing at -70°C or immediate use inthe RIP with intact, motile organisms.Serum reagents. Prebled control serum and antise-

rum against T. vaginalis strain 286 were generated inNew Zealand white rabbits injected subcutaneouslyand intramuscularly with a mixture of 0.5 ml of tricho-monal suspension containing 2.5 x 107 organisms and0.5 ml of Freund complete adjuvant (GIBCO Labora-tories, Grand Island, N.Y.). Booster injections wereperformed 2 weeks later with equal amounts of orga-nisms in Freund incomplete adjuvant. Parasites wereextensively washed three times in PBS to removecontaminating medium macromolecules adsorbed ontoparasite surfaces (17). Rabbits were bled 14 days afterbooster immunization, and the sera were stored at-70°C until use. Adsorbed antiserum was generatedby suspension of 1.7 x 107 trichomonads washed threetimes in PBS in 0.5 ml of antiserum, and the incubationwas continued for 30 min at 37°C. The serum wasremoved after pelleting of the parasites, and the proc-ess was repeated 10 times. No lysis of organisms wasdetected based on cell counts with an improved Neu-bauer counting chamber. Alternate immunization pro-tocols such as intravenous injection of formaldehyde-fixed organisms before or after the booster mentionedabove neither modified nor amplified the antibodyresponse as analyzed with these experimental proce-dures.RIP assay. Zwittergent 3-12 (Z3-12) (Calbiochem-

Behring Corp., La Jolla, Calif.) extracts of [35S]me-thionine- or Na125I-labeled T. vaginalis proteins wereused as a source of antigen in the reaction mixture (1).One hundred microliters of a Z3-12-solubilized, radio-labeled preparation of strain 286 was mixed with 50 p.lof serum reagent. After incubation at 4°C overnight, a100-,ul volume of protein A-bearing Staphylococcus

auireus (12) was added to bind the antibody-antigencomplexes. After extensive washing of the protein A-bearing S. aureus radiolabeled antigen(s) were re-leased by suspending pelleted protein A-bearing S.aureus organisms in 100 p.1 of electrophoresis dissolv-ing buffer (3, 17) and boiling for three min. The proteinA-bearing S. aureus cells were then pelleted, and thesupernatant was electrophoresed as described belowfor fluorographic detection of immunogenic trichomo-nal proteins.The analysis of exposed membrane proteins on

intact, motile T. vaginalis accessible to antibody byRIP was performed immediately after intrinsic orextrinsic radiolabeling of parasites proteins. This pro-cedure has been recently and successfully employedfor a bacterial model system (7, 8) and has beendefined as a whole cell (WC) RIP. Approximately 1.25x 108 trichomonads were pelleted after intrinsic label-ing with [35S]methionine (1, 17) or chloramine-T-medi-ated radioiodination and suspended with 100 p.1 ofappropriate heat-inactivated control serum or heat-inactivated antiserum reagent. The suspension wasincubated for 120 min at 37°C with occasional gentlestirring followed by washing of the trichomonads noless than twice in PBS. T. vaginalis organisms werethen suspended in 200 p.1 ofNET (150 mM NaCl, 5 mMEDTA, 50 mM Tris-hydrochloride, pH 7.2) buffer andsolubilized by the addition of 25 p.l of 10% Z3-12detergent (4, 6). The preparation was gently homoge-nized, and an additional 200 p.l of NET buffer wasadded. The extract was then centrifuged at 35,000 rpm(Beckman SW50.1 rotor) to remove insoluble debris.Regardless of the type of radiolabeling employed forparasites in this WC RIP, greater than 80% of theradioactivity was recovered in the supernatant aftercentrifugation. Finally, 100 p.l of 10% (vol/vol) proteinA-bearing S. aureus cells prepared as previously de-scribed (2, 3, 12, 17) was added to the clarified,solubilized extract, and the mixture was again incubat-ed at 370C with occasional shaking. The protein A-bearing S. aureus cells were then pelleted and washed,and the adsorbed radiolabeled antigen was processedwith dissolving buffer as described above.Sodium dodecyl sulfate-polyacrylamide gel electro-

phoresis. The procedures outlining preparation of totalT. vaginalis proteins for sodium dodecyl sulfate-poly-acrylamide gel electrophoresis and fluorography wereas outlined elsewhere (1, 2, 17). Stacking and separat-ing gels consisted of 3 and 7.5% acrylamide (Bio-RadLaboratories, Richmond, Calif.), and electrophoresiswas carried out with a Protean Dual 16-cm slab cellapparatus (Bio-Rad) with a constant current of 15 mAper gel. After penetration of the bromophenol bluetracking dye into the separating gel, the current was

increased to 30 mA per gel. Gels were stained withCoomassie brilliant blue (Sigma), destained, and proc-essed for fluorography as described previously (2, 17).

Agglutination of trichomonads and indirect immuno-fluorescence. Specific agglutination of T. vaginalisorganisms by antibodies directed against membraneproteins was performed as recently outlined (17).Extensively washed trichomonads were incubatedwith a suspension of formaldehyde-fixed protein A-bearing S. aureus pretreated with the serum reagentalso used for the RIP assay (1, 17). After incubation forvarious lengths of time at 370C, samples of pelletscontaining agglutinated parasites and supernatants

VOL. 40, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

286 ALDERETE

having individual motile organisms were evaluatedwith dark-field optics with a Leitz Ortholux II micro-scope. The specificity of all sera used is as indicated inthe data generated in RIP assays, and the extent ofagglutination was recorded as previously described(17).The detection of antibody to T. vaginalis surfaces

was also attempted by using fluorescence microscopy.To correlate with results from RIP and protein A-bearing S. aureus-mediated agglutination, however,control normal rabbit serum and antiserum from im-munized rabbits were each fractionated with proteinA-Sepharose (Pharmacia Fine Chemicals, Piscataway,N.J.) chromatography. A 0.1-ml sample containingapproximately 2 x 106 organisms washed twice withPBS was then mixed with an equal volume of purifiedimmunoglobulin G (IgG) fraction (protein concentra-tion, 3 mg/ml) or serum reagents. After incubation for20 min at 37°C with continuous gentle shaking, theparasites were washed twice with PBS and suspendedin 0.1 ml of fluorescein isothiocyanate-goat anti-rabbitimmunoglobulin G (Cappel Labs, Cochranville, Pa.)diluted 1:5 in PBS buffer. The suspension was placedat 4°C in the dark for 20 min before gentle washing asabove. The trichomonads were finally suspended in0.2 ml of PBS and evaluated with a Zeiss IM35microscope equipped with fluorescence and darkfieldoptics. T. vaginalis organisms remained motilethroughout this procedure.

RESULTSIdentification of immunogenic membrane pro-

teins. Initial characterization of T. vaginalismembrane proteins was accomplished with a Z3-12 detergent extract of radioiodinated trichomo-nads as a source of antigen in the RIP assay.This was compared with a Z3-12 preparation of[35S]methionine-labeled parasite proteins em-ployed in the antibody-antigen reaction mixture.Efficient radiolabeling of most or all T. vaginalisproteins with [35S]methionine has been accom-plished and used successfully to assess the na-ture and extent of IgG from immunized rabbitserum for reactivity with specific immunogenicproteins (1).

Solubilized, 125I-labeled trichomonal surfaceproteins immunoprecipitated with antiserumfrom immunized rabbits are shown in Fig. 1A(lane b). The absence of any radioiodinatedproteins reactive with prebled control serum(lane a) demonstrates the specificity of the meth-odology employed. In addition, the use of anti-serum extensively adsorbed with live parasitesin the RIP assay resulted in decreased or absentprotein bands after electrophoresis (lane c), sug-gesting that antibody reactive with intact, motiletrichomonads was effectively removed and rein-forcing the idea of the surface location for theobserved protein bands (lane b). Approximately17 proteins were detected and ranged in molecu-lar weight (MW) from 200,000 (200K) to 20Kunder these experimental conditions. Alternate

Aa b

MW

200K-

B TCAc a b 286

-20-19

18 -

17-

120KK -1692.5K 4 -14

- -1312 4

68K - - *

9

44K---- 78-6

31K- -43_-1

FIG. 1. Fluorogram of RIP with Z3-12-solubilized,radioiodinated (A) or 35S-labeled (B) trichomonal pro-teins incubated with normal rabbit serum obtainedbefore immunization (lane a), antiserum generated inrabbits immunized with T. vaginalis strain 286 (lane b),and antiserum adsorbed against intact, motile orga-nisms as described in the text (lane c). TCA 286 refersto total trichloroacetic acid-precipitated trichomonalproteins used for comparative purposes.

radioiodination procedures such as lactoperoxi-dase-catalyzed labeling did not efficiently tagexposed proteins with Na125I (data not shown),but did result in similar electrophoretic patternsafter very extended exposure of X-ray film dur-ing autoradiography.Comparative profiles of immunogenic pro-

teins were obtained by using intrinsically labeledparasites. Incubation of 3 S-labeled Z3-12 286extract with the same antiserum resulted in theimmunoprecipitation of comigrating and othertrichomonal proteins (Fig. 1B, lane b). In con-trast, control prebled serum did not immunopre-cipitate any labeled parasite proteins (lane a). Ofinterest was the similarity in the antiserum reac-tivity to both 1251- and 35S-labeled proteins,suggesting a parasite origin for the radioiodi-nated immunogenic membrane proteins of T.vaginalis. The pronounced intensity of proteinbands with apparent MWs of 92K (protein 14)and 60K (protein 10) may be indicative of largenumbers of tyrosine residues for these polypep-tides. Alternately, the intensity of protein bandshas been shown to correlate with the antibodytiter in reactive serum (2), and intense bandsobtained with radioiodinated detergent extracts

INFECT. IMMUN.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

IMMUNOGENIC T. VAGINALIS MEMBRANE PROTEINS 287

a b c dSOL TCA.vg. .vg.

a-20-19

-18

-i1- -' -VW

_151

|

9 4* fi

IW -e

-2

FIG. 2. WC RIP fluorograms with 125I-labeled (lanes a and b) versus [35S]methionine-labeled (lanes c and d)intact, motile T. vaginalis cells as the antigen in the reaction mixture. Organisms were incubated with sera andsolubilized with Z3-12 detergent, and antibody-binding membrane proteins were immunoprecipitated withprotein A-bearing S. aureus and processed as described in the text. Lanes a and c represent 1251- and 35S-labeledtrichomonads, respectively, incubated with normal rabbit serum obtained before immunization. Lanes b and drepresent 125I- and 35S-labeled parasites incubated with antiserum reagent. SOL T. vg. is representative of a 35S-labeled Z3-12 extract used as a control to insure efficient solubilization of trichomonal proteins before theaddition of protein A-bearing S. aureus to the reaction mixture. TCA T. ig. is as described for Fig. 1.

may reflect the predominance of antibody to-ward these membrane proteins.A band with a MW of approximately 200K

was also obtained when 35S-labeled extract wasused in the RIP assay, but only after prolongedexposure of the X-ray film. Because parasitesradiolabeled with a 3H-amino acids mixtureyielded similar RIP data, these results suggest a

low copy number for these high-MW proteins ontrichomonal membranes. A representative con-trol profile of detergent-solubilized, 35S-labeledproteins is illustrated in Fig. 2. The proteinpattern of a Z3-12 extract of 125I-labeled surfaceproteins was not included because of the highbackground levels of either free Na1251 or otherradioiodinated components such as trichomonallipids in the acrylamide gels. Radioiodinatedproteins indicative of a profile equivalent to thatseen in Fig. IA (lane b) were, however, apparentin this extract.Antibody binding by membrane proteins with

intact trichomonads. Because investigators haverecently and successfully utilized whole orga-nisms as antigen in an RIP assay (7, 8), it wasequally important to determine whether, underour experimental conditions, the same or differ-ent parasite proteins would be immunoprecipi-tated with the IgG reagent. A successful WCRIP would greatly enhance our knowledge of

highly immunogenic T. vaginalis surface pro-teins capable of binding antibody while in theirnative membrane orientations.

Intrinsically or extrinsically labeled trichomo-nads were incubated with prebled serum orantiserum from an immunized rabbit and solubi-lized with Z3-12 before immunoprecipitation ofimmune complexes with protein A-bearing S.aureus. Antiserum reacted with the identicalreportoire of immunogenic proteins in a WC RIPwith 125I-labeled T. vaginalis (Fig. 2, lane b).Again, approximately 20 proteins with a MWspectrum of 200K to 20K were detected. Note-worthy were intense bands of apparent MW of92K and 65K consistent with data obtained withpresolubilized radiolabeled parasites as antigen(Fig. 1). In contrast to the antiserum reactivity,control prebled serum did not immunoprecipi-tate any radioiodinated trichomonal proteins(Fig. 2, lane a).The almost identical protein pattern obtained

with [35S]methionine-labeled T. vaginalis in theWC RIP (lane d) confirmed the surface-exposednature of these immunogenic proteins and dem-onstrated the specificity and sensitivity of thistechnology. The gel profile detected in this WCRIP (Fig. 2, land d) differed considerably fromthat obtained when presolubilized, 35S-labeledorganisms were employed (Fig. 1B, lane b). The

MW

200K -

120K -92.5 K -

68K -

44K -

31K -21.5K -

VOL. 40, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

288 ALDERETE

absence of immunoprecipitation of any 35S-la-beled proteins with control serum was also ob-served (Fig. 2, lane c). The demonstration of acomplete protein complement in the Z3-12 ex-tract used in the reaction mixture served as anadditional control to demonstrate that all para-site proteins were always present in the RIPreaction mixture (Fig. 2). Because host plasmaproteins are known to avidly bind to intact T.vaginalis organisms (17), it was important toestablish the trichomonal origin of precipitated,radioiodinated macromolecules. Since theseparasites were grown in a horse serum-supple-mented medium, a radioiodinated serum proteinband (Fig. 1 and 2) would imply the presence ofanti-equine serum antibody. This possibility waseliminated by lack of precipitin bands after im-munoelectrophoresis of horse serum and use ofanti-trichomonal rabbit serum as probe. Further-more, RIP with intrinsically and extrinsicallylabeled parasites and antiserum against eitherTrypticase-yeast extract-maltose medium orhorse serum did not result in the immunoprecip-itation of any radiolabeled polypeptides (unpub-lished observations).

Localization of surface-directed antibody. Be-cause trichomonads possess specialized surfacestructures directly observable with dark-fieldoptics, it was important to demonstrate theimmunogenic potential of these readily detect-able organelles. Also, because serum from im-munized rabbits has been reported to agglutinatetrichomonads (14, 15, 23), we felt that a highlysensitive and specific agglutination procedurepreviously developed in this laboratory (17)would allow us to examine whether IgG charac-terized by these RIP experiments was directedat these surface structures.

Protein A-bearing S. aureus organisms pre-treated with either control serum or antiserumagainst strain 286 were incubated with a suspen-sion of washed T. vaginalis. Thus, IgG directedagainst exposed membrane proteins on tricho-monads might allow for protein A-bearing S.aureus binding and microscopic observation ofimmunogenic parasite structures. Figure 3 (pan-el iB) clearly shows the extensive agglutinationof motile trichomonads in a representative ex-periment when protein A-bearing S. aureus cellswere incubated with rabbit antiserum. No agglu-tination or protein A-bearing S. aureus bindingto intact organisms was detectable with proteinA-bearing S. aureus cells pretreated with controlprebled serum or protein A-bearing S. aureuscells alone (Fig. 3, panel 1A). Also, antibody didnot appear to be directed against parasite anteri-or flagella (Fig. 3, panels 2A and 2B) and theposterior axostyle. Extensive clumping of pro-tein A-bearing S. aureus was apparent, howev-er, throughout the remaining trichomonal sur-

INFECT. IMMUN.

face. Indications of membrane perturbationsresulting in cap-like aggregates of adsorbed pro-tein A-bearing S. aureus cells were also detectedunder these experimental conditions.

Finally, fluorescence microscopy was alsoutilized, and the results were correlated withthose of protein A-bearing S. aureus-mediatedagglutination. Strong fluorescence of trichomo-nads agglutinated after exposure to protein A-purified immunoglobulin G or antiserum fromimmunized rabbits was readily demonstrated(Fig. 4B). Representative microscopic fieldswith individual parasites and with different fluo-rescent intensities are also illustrated (Fig. 4Cthrough 4F); in each case, the trichomonal fla-gella and axostyle were not apparent with eitherrabbit antiserum or the IgG fraction. Dark-fieldmicroscopy of the same field clearly shows thepresence of parasite flagella and axostyle notdetected by indirect immunofluorescence. Nofluorescence was ever observed with IgG fromprebled control serum or after incubation ofparasites with control serum (Fig. 4, panel A),supporting previous RIP and agglutination re-sults. Thus, antibody directed at flagella andaxostyle was never detected under these experi-mental conditions.

DISCUSSIONT. vaginalis is a protozoan parasite responsi-

ble for trichomonal vaginitis and is responsiblefor one of the leading sexually transmitted dis-eases in this country and the world. A complexhost-parasite relationship appears operative inthis urogenital infection of women. For exam-ple, an intimate association between trichomo-nads and epithelium as well as extensive tissuecytopathology have been documented (10).Nonetheless, little or no information is availableon specific virulence factors mediating diseasepathogenesis. Recent efforts by this laboratoryto examine the interaction between parasite andhost macromolecules indicated that T. vaginalispossessed the ability to loosely and avidly bindplasma or tissue components (17). This informa-tion may be useful in enhancing our understand-ing of certain aspects of the biology of thisparasite, such as trichomonal circumvention ofimmune surveillance mechanisms. Since an al-ternate strategy toward increasing our knowl-edge of this microorganism would be to charac-terize the trichomonal membrane, an attemptwas made to identify highly immunogenic andexposed membrane proteins of T. vaginalis.This effort was a logical extension of recentwork in this laboratory which examined thenature and extent of antibody directed towardimmunogenic parasite proteins and produced byimmunized or infected experimental animals (1).

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

IMMUNOGENIC T. VAGINALIS MEMBRANE PROTEINS

IA IB

2A 28FIG. 3. Representative specific protein A-bearing S. aureus-mediated agglutination of live, motile trichomo-

nads (1B) after pretreatment of staphylococci with rabbit antiserum against T. vaginalis strain 286. Treatment ofprotein A-bearing S. aureus with normal rabbit serum resulted in no detectable parasite agglutination (1A).Frames 2A and 2B are enlargements of individual trichomonads after agglutination emphasizing the absence ofprotein A-bearing S. aureus binding to parasite flagella (arrow). Protein A-bearing S. aureus pretreated withrabbit anti-286 serum was also unreactive with parasite undulating membrane and axostyle as described in thetext.

Initial characterization of immunogenictrichomonal membrane proteins was accom-plished by using a detergent extract consisting ofT. vaginalis radioiodinated by chloramine-T asantigen in an RIP assay (Fig. 1). Z3-12 repre-sents a gentle and efficient detergent system (4,6) which allowed for specific antibody-antigeninteractions. At least 17 major '251-labeled mem-brane proteins were found to reside on parasitemembranes. Of interest was the pronouncedintensity of two polypeptides with apparentMWs of 92K (protein 14) and 65K (protein 10).Previous characterization of immunogenictrichomonal proteins with intrinsically labeleddetergent preparations (1) revealed a high titer ofIgG directed against these two membrane pro-teins. Additionally, RIP performed concomitant-ly with these experiments, but with 35S-labeledproteins, clearly affirmed the highly immunogen-ic nature and parasite biosynthesis of these andother trichomonal proteins (Fig. 1B, lane b). Theenhanced specific activities of these two poly-peptides after chloramine-T-mediated radioio-

dination also cannot be excluded from theseinitial observations. Confirmation of the surfaceorientation of the 125I-labeled proteins immuno-precipitated with antiserum was obtained by RIPwith antiserum adsorbed against live, motileorganisms (Fig. 1A, lane c). The absence ofprotein bands in the electrophoretic profileswhen control prebled serum or antiserum gener-ated against medium components was used inthe RIP assay demonstrated the high specificityof the procedures employed.Almost identical results were obtained when

whole radioiodinated trichomonads were em-ployed as antigen in the WC RIP assay (Fig. 2)as compared with the employment of 125I-la-beled, presolubilized preparations. The com-monality of the autoradiograms reinforced thepresence of prominent antibody-binding pro-teins on T. vaginalis membranes and the highlyimmunogenic nature of protein 14 (92K) andprotein 10 (65K). The increased intensity of agroup of proteins with MWs of about 200Ksuggested that the prior solubilization with de-

VOL. 40, 1983 289

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

290 ALDERETE

A

C D E FFIG. 4. Indirect immunofluorescence microscopy with intact, motile T. vaginalis organisms. The interaction

of trichomonads with the IgG fraction from normal rabbit serum (A) and with IgG from a rabbit immunized withstrain 286 (B through F) is as described in the text. The use of normal rabbit serum or serum from a rabbitimmunized with T. vaginalis yielded similar results. Dark-field microscopy was performed on all of the samefields to insure parasite motility and presence of flagella and axostyle.

tergent systems did not allow for proper identifi-cation of immune determinants by antibody.Alterations in protein antigen quaternary struc-ture caused by detergents might destroy anti-body-reactive sites. This fact emphasizes theneed for more than one methodology in assess-ing immunogenic parasite proteins. Especiallyimportant was the recent demonstration of pro-teins 10, 14, 16, and 20 as those antigens pos-sessing highest IgG titer in antiserum from im-munized rabbits or subcutaneously infectedmice (1).The identification of the same antibody-acces-

sible proteins in the WC RIP performed withintrinsically [35S]methionine-labeled organismsconfirmed the parasite origin of these polypep-tides (Fig. 2, lane d). Thus, antiserum immuno-precipitated 35S-labeled proteins which comi-grated with bands obtained when 125I-labeled T.vaginalis organisms were used (Fig. 2, lane b).The absence of any additional proteins whenintrinsically labeled trichomonads were utilized

in the WC RIP indicates that the same proteinsalso contain sufficient tyrosine residues for sub-sequent RIP analysis. On the other hand, agroup of proteins around 200K in MW were notreadily detectable when 35S-labeled parasiteswere employed. As with the previous experi-ment (Fig. 1), extended exposure of X-ray filmdid result in the appearance of immunoprecipi-tated proteins in this region. The generation ofsimilar data with a mixture of tritiated aminoacids for radiolabeling total trichomonal proteins(data not shown) suggests that these group ofproteins are present in very few copies per cell.Finally, as addressed above, the proteins identi-fied as immunogens exposed on T. vaginalismembranes were additionally determined to beof trichomonal origin through lack of immunecomplexes formed by (i) immunoelectrophoresisof horse serum using anti-T. vaginalis serum as aprobe and (ii) RIP of radiolabeled trichomonalproteins with antiserum raised against mediumcomponents or horse serum.

INFECT. IMMUN.

I

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.

IMMUNOGENIC T. VAGINALIS MEMBRANE PROTEINS 291

Highly immunogenic, antibody-accessiblemembrane proteins of T. vaginalis may begrouped into these possessing MWs greater than100K (proteins 16 through 20) and those pos-sessing MWs between 65K and 100K (proteins 9through 15). Lower-MW species were detected,but usually gave weaker bands after autoradiog-raphy (Fig. 2). Future studies detailing the mo-lecular organization of these proteins while with-in the membrane will be useful in ascertainingthe structure and function properties of thesesurface proteins and characterizing the impor-tance of the lower-MW immunogens. As withprevious studies in our laboratory (1), no differ-ences in the makeup of antibody-accessible pro-teins were detected in clones (9) of T. vaginalis.Thus, possible heterogeneity among T. vaginalispopulations (10, 11, 21, 22) cannot be attributedto instability of these immunogenic membraneproteins. These conclusions, however, must beinterpreted with caution until trichomonads ex-amined immediately after isolation are com-pared with the same cultures allowed to grow forextended periods of time. Thus, the antigenicvariability of in vitro-grown parasites is present-ly under investigation. Of importance would bestudies aimed at elucidating the biological func-tion of these highly immunogenic trichomonalsurface proteins.

Finally, the demonstration that immunoglob-ulin from rabbit antiserum was not directedtoward specialized parasite membrane struc-tures is extremely noteworthy (Fig. 3 and 4).Antibody reactive with either axostyle or flagel-la was not detected using two distinct indicatorsystems. These initial RIP data along with agglu-tination and fluorescence studies are importantfor dissection of IgG-targeted proteins with po-tential as vaccinogen candidates and is consist-ent with published reports on the immunology oftrichomonal infection (14, 15, 23). The highlyimmunogenic nature of T. vaginalis membraneproteins which react with antibody while theproteins are in their native state makes thempresumptive candidates for either experimentalvaccines or immunodiagnostic reagents. Thiseffort, however, is dependent upon future analy-sis of antibody responses in serum and vaginalsecretions among infected women with variedsymptomology.

ACKNOWLEDGMENTSThis work was supported by a Public Health Service grant

AI 18768 from the National Institutes of Health. I especiallythank Grace Wagner for her patient and expert secretarialassistance.

LITERATURE CITED

1. Alderete, J. F. 1983. Antigen analysis of several pathogen-ic strains of Trichomonas vaginalis. Infect. Immun.39:1041-1047.

2. Alderete, J. F., and J. B. Baseman. 1980. Surface charac-terization of virulent Treponema pallidum. Infect. Im-mun. 30:814-823.

3. Alderete, J. F., and J. B. Baseman. 1981. Analysis ofserum IgG against Treponema pallidum protein antigensin experimentally infected rabbits. Br. J. Vener. Dis.57:302-308.

4. Baron, C., and T. E. Thompson. 1975. Solubilization ofbacterial membrane proteins using alkyl glucosides anddioctanyl phosphatidylcholine. Biochim. Biophys. Acta382:276-285.

5. Garvey, J. S., N. E. Cremer, and D. H. Sussdorf. 1977.1251 or "31I-labeled proteins. p. 171-182. In D. H. Camp-bell (ed.), Methods in immunology. W. A. Benjamin, Inc.,Reading, Mass.

6. Gonenre, A., and R. Ernst. 1978. Solubilization of mem-brane proteins by sulfobetaines noval zwittergent deter-gents. Anal. Biochem. 87:28-38.

7. Gulig, P. A., G. H. McCracken, Jr., C. F. Frisch, K. H.Johnston, and E. J. Hansen. 1982. Antibody response ofinfants to cell surface-exposed outer membrane proteinsof Haemophilus influenzae type b after systemic Haemo-philus disease. Infect. Immun. 37:82-88.

8. Hansen, E. J., C. F. Frisch, R. L. McDade, Jr., and K. H.Johnston. 1981. Identification of immunogenic outer mem-brane proteins of Haemophilus influenzae type b in theinfant rat model system. Infect. Immun. 32:1084-1092.

9. Hollander, D. H. 1976. Colonial morphology of Trichomo-nas vaginalis in agar. J. Parasitol. 62:826-828.

10. Honigberg, B. M. 1978. Trichomonads of importance inhuman medicine. p. 275-454. In J. P. Kreier (ed.), Para-sitic protozoa, vol. 2. Academic Press, Inc., New York.

11. Honigberg, B. M., and M. Goldman. 1968. Immunologicanalysis by quantitative fluorescent antibody methods ofthe effects of prolonged cultivation of Trichomonas gal-linae. J. Protozool. 15:176-184.

12. Kessler, S. W. 1976. Cell membrane antigen isolation withthe staphylococcal protein A-antibody carrier. J. Immu-nol. 117:1482-1490.

13. Kreiger, J. N. 1981. Urologic aspects of trichomoniasis.Invest. Urol. 18:411-417.

14. Lanceley, F. 1958. Serological aspects of Trichomonasvaginalis. Br. J. Vener. Dis. 34:3-8.

15. McEntegart, M. G. 1952. The application of a haemmag-glutination technique to the study of Trichomonas vagina-lis infections. J. Clin. Pathol. 5:275-280.

16. Muller, M., J. G. Meingassner, W. A. Miller, and W. J.Ledger. 1980. Three metronidazole-resistant strains ofTrichomonas vaginalis from the U.S.A. Am. J. Obstet.Gynecol. 138:808-812.

17. Peterson, K. P., and J. F. Alderete. 1982. Host plasmaproteins on the surface of pathogenic Trichomonas vagi-nalis. Infect. Immun. 37:755-762.

18. Rein, M. F., and T. A. Chapel. 1975. Trichomoniasis,candidiasis, and the minor venereal diseases. Clin. Ob-stet. Gynecol. 18:73-88.

19. Soendjojo, A., and S. Pindha. 1981. Trichomonas vaginal-is infection of the median raphe of the penis. Sex. Trans.Dis. 8:255-257.

20. Spence, M. R., D. H. Hollander, J. Smith, L. McCaig, D.Sewell, and M. Brockman. 1980. The clinical and labora-tory diagnosis of Trichomonas vaginalis infection. Sex.Trans. Dis. 7:168-172.

21. Stepkowski, S., and B. M. Honigberg. 1972. Antigenicanalysis of virulent and avirulent strains of Trichomonasgallinae by gel diffusion methods. J. Protozool. 19:306-315.

22. Teras, J. K. 1966. Differences in the antigenic propertieswithin strains of Trichomonas vaginalis. Wiad. Parazytol.12:357-363.

23. Trussell, R. E. 1946. Microagglutination tests with Tricho-monas vaginalis. J. Parasitol. 32:563-567.

VOL. 40, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/i

ai o

n 14

Feb

ruar

y 20

22 b

y 82

.102

.107

.189

.