Vol. 55, No. 1INFECTION AND IMMUNITY, Jan. 1987, p. 49-560019-9567/87/010049-08$02.00/0Copyright © 1987, American Society for Microbiology

Shared Epitopes between Mycoplasma pneumoniae Major AdhesinProtein P1 and a 140-Kilodalton Protein of Mycoplasma genitalium

JANICE MORRISON-PLUMMER, ANNA LAZZELL, AND JOEL B. BASEMAN*Department of Microbiology, The University of Texas Health Science Center at San Antonio, San Antonio, Texas 78284

Received 14 August 1986/Accepted 27 September 1986

Previous serological data have demonstrated cross-reactive antigens between two pathogenic species ofmycoplasmas, M. pneumoniae and M. genitalium. Preliminary analysis of sera and monoclonal antibodies(MAbs) to protein antigens of these species showed an immunodominance of adhesin P1 (165 kilodaltons [kDa])of M. pneumoniae in mice and hamsters and a 140-kDa protein of M. genitalium in mice and experimentallyinfected chimpanzees. To further characterize these two proteins, we assayed multiple anti-Pl and anti-140-kDa protein MAbs by enzyme-linked immunosorbent assay, immunoblot, and radioimmunoprecipitationtechniques. The 140-kDa M. genitalium protein was shown to be surface accessible and insensitive to levels oftrypsin which readily degrade protein P1. Peptide mapping was used to identify a unique class of MAbs whichbound a cross-reactive molecule common to both the major adhesin protein P1 of M. pneumoniae and the140-kDa protein ofM. genitalium. MAbs generated against both M. pneumoniae and M. genitalium which werereactive with this determinant blocked M. pneumoniae attachment to chicken erythrocytes.

Mycoplasma pneumoniae and Mycoplasma genitaliumare pathogens which share common ultrastructural, morpho-logical, and serological features (3, 6, 9, 16, 17, 22, 25). Inrecent studies in our laboratory, various proteins of M.pneumoniae were identified as ligands mediating M. pneu-moniae attachment to host cells (1, 2, 11, 13, 20). Atrypsin-sensitive surface protein designated P1 (165 kilodal-tons [kDa]) appears to be a primary ligand mediatingcytadherence consistent with its dense clustering at theterminal nap region of the mycoplasma attachment organelleand its absence in specific nonadhering mycoplasma mutants(1, 8, 10, 12, 13). Moreover, monospecific and monoclonalantibodies (MAbs) to protein P1 of M. pneumoniae havebeen shown to block mycoplasma attachment to erythro-cytes (RBCs) and respiratory epithelium (11, 20).Recent studies by Tully et al. (22, 23) have demonstrated

that M. genitalium not only shares the flask shape of M.pneumoniae, but also possesses a predominant surface napextending distally from the tip organelle. This structureappears to mediate the adherence of M. genitalium to Veromonkey kidney cell monolayers (22), thus implying that aprocess similar to that of M. pneumoniae regulates M.genitalium cytadherence.

This study characterizes an immunodominant protein ofM. genitalium with a molecular weight of 140,000. MAbsdirected against both the 140-kDa protein and protein P1 ofM. pneumoniae were used to examine the antigenic related-ness of these two proteins. Using this approach, a cross-reactive epitope(s) on the two molecules was identified.

MATERIALS AND METHODS

Mycoplasma strains and culture conditions. Virulent, hem-adsorbing M. pneumoniae M129-B16 was grown in 32-oz(946-ml) prescription bottles in 70 ml of Edward medium (7)at 37°C for 72 h. Glass-adherent mycoplasmas were washedfour times with phosphate-buffered saline (PBS; pH 7.2) andcollected by centrifugation (9,500 x g, 20 min). M. genital-ium G37 and M30 were grown in SP-4 medium (22, 25). The

* Corresponding author.

same procedures for growth and harvesting of M. pneumo-niae were used for M. genitalium. All other strains ofmycoplasmas were a gift from J. G. Tully (National Instituteof Allergy and Infectious Diseases, Frederick Cancer Re-search Facility, Frederick, Md.). Acholeplasma laidlawiiB-PG9 was grown in a modified Edward medium containingpenicillin G (100 U ml-1) as previously described (4).

Immunization of mice for production of MAbs. BALB/cfemale mice (3 to 6 weeks old) were used for all immu-nizations. The immunization and hybridization proceduresfor fusions H5, H10, H12, and H13 have been describedpreviously (19, 20). For fusion H23, mice were injectedintraperitoneally on day 1 with 125 pug of M. genitalium G37emulsified 1:1 (vol/vol) in Freund complete adjuvant (DifcoLaboratories, Detroit, Mich.). On day 9, mice were givenintraperitoneal injections of 125 ,ug of M. genitalium. Themouse which demonstrated the highest titer by enzyme-linked immunosorbent assay (ELISA) against M. genitaliumwas sacrificed 3 days later, and the spleen was excised andprepared for hybridization (19). For fusion H37, mice wereimmunized on days 1, 15, 73, 80, and 125 with intraperitonealinjections of 100 to 175 pug of M. genitalium G37 in saline.Hybridization was performed on day 129.

Sera. Rabbit anti-M. genitalium serum was obtained bysubcutaneous and intramuscular injection of male NewZealand White rabbits with 1010 M. genitalium cells emulsi-fied in Freund complete adjuvant. Boosts were 3 weeks apartwith Freund incomplete adjuvant. Blood was collected fromthe ear.ELISA. A modified ELISA was used for detection of

MAbs to whole mycoplasma organisms. Freshly harvestedmycoplasmas were assayed for protein content by theLowry method (18) and diluted to 26.6 p.g/ml of PBS.Aliquots (75 ,ul) containing 2 ,ug, an amount previouslydetermined to be optimal, were distributed into wells ofImmulon I microtiter strips (Dynatech Laboratories, Inc.,Alexandria, Va.). Wells were dried overnight at 37°C andstored dry at 4°C.The ELISA procedure was also used to determine MAb

reactivity to sodium dodecyl sulfate (SDS)-gel-eluted proteinP1 of M. pneumoniae and the 140-kDa protein of M.

49

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

50 MORRISON-PLUMMER ET AL.

genitalium. Freshly harvested mycoplasmas (2 mg) weresolubilized in PBS and an equal volume of SP buffer (0.1 MTris [pH 6.8], 2% SDS, 20% glycerol, 2%p-mercaptoethanol,0.02% bromophenol blue), and the samples were boiled for 3min. Samples were then microcentrifuged for 5 min to pelletinsoluble proteins. Solubilized mycoplasmas were then ana-lyzed by preparative 7.5% SDS-polyacrylamide gel electro-phoresis (PAGE) (14). To determine where the proteinsmigrated, we stained a small longitudinal section of each gelwith Coomassie brilliant blue (Sigma Chemical Co., St.Louis, Mo.), and the unstained protein bands correspondingto the 165-kDa protein of M. pneumoniae and the 140-kDaprotein of M. genitalium were then sliced, macerated, andgently shaken in H20 for 48 to 60 h at 4°C. The suspensionwas centrifuged at 17,000 x g for 20 min to pellet acrylamidefragments. Supernatants were lyophilized and reconstitutedin coating buffer (26) at a final concentration of 40 ng/ml, and75 RI was placed in each well of Immulon II microtiter plates(Dynatech). Protein concentrations of the gel bands weredetermined by readings on a LKB Bromma 2202 ultrascanlaser densitometer and 3390A Hewlett-Packard integrator.

Before use, antigen-coated wells were blocked with PBSsupplemented with 1% bovine serum albumin (PBS-BSA;Sigma). The ELISA procedure has been described previ-ously (19, 26).

Competition ELISA. Competitive ELISAs were performedwith alkaline phosphatase (bovine intestine type VII-T;Sigma)-labeled antibodies. Column-purified ascites of anti-P1 MAbs H5.2G4, H13.6E7, H10.8E8, and H12.5B8 werelabeled according to the procedure of Voller et al. (26).Whole M. pneumoniae-coated microELISA wells were

blocked with PBS-BSA. Unlabeled MAb was diluted inPBS-BSA and added to the antigen-coated wells 5 minbefore the addition of alkaline phosphatase-labeled MAb.Wells were then incubated for 2 h at 37°C before threewashes with PBS and two with H20 followed by the additionof the p-nitrophenol phosphatase substrate.Immunoglobulin isotyping of MAbs. MAbs were isotyped

by ELISA as previously described (19).MAb purification. MAbs were purified by passage through

a protein A-Sepharose CL4B column (Pharmacia FineChemicals, Piscataway, N.J.) (20).Western blot analysis of MAbs. Pellets of M. pneumoniae

or M. genitalium were suspended in PBS and an equalvolume of SP buffer and then analyzed by 7.5% SDS-PAGE.Each gel contained 2.0 to 2.5 mg of total protein. Proteinswere electrophoretically transferred to nitrocellulose mem-branes (Bio-Rad Laboratories, Richmond, Calif.) by themethod of Towbin (21). Molecular weights of proteins weredetermined by coelectrophoresis of 14C-labeled molecularweight protein standards (Bethesda Research Laboratories,Inc., Gaithersburg, Md.).To block unbound sites on the nitrocellulose membranes,

nitrocellulose strips were incubated for 2 h at 25°C in 3%gelatin in Tris-buffered saline (TBS; 20 mM Tris, 500 mMNaCl, pH 7.5). They were then incubated for 1 h at 37°C orovernight at 25°C with MAb supernatants, protein A-Sepharose-purified MAb, or serum diluted in TBS-1% gela-tin. Before incubation for 2 h at 37°C with horseradishperoxidase-conjugated anti-rabbit immunoglobulin G or anti-mouse immunoglobulin (Hyclone, Logan, Utah), blots werewashed three times with TBS. The substrate-color-developing reagent 4-chloro-1-naphthol (Bio-Rad) was usedto develop the immunoblots.Three other procedures using immunoblot technology

were employed. Peptide mapping of protein P1 utilized the

technique of Cleveland et al. (5). Briefly, M. pneumoniaewas electrophoresed on 4% SDS-PAGE gels. The gels werestained with Coomassie brilliant blue, and then the P1 bandwas excised and equilibrated in buffer (0.125 M Tris [pH6.8], 0.1% SDS, 1 mM EDTA) for 30 min. P1-containing gelfragments were then loaded onto 10% SDS-PAGE gels, andvarious concentrations of trypsin in a 20% glycerol solutionwere added. Samples were electrophoresed into the stackinggel and incubated for 90 min before continued electrophore-sis through the separating gel. Gel proteins were thentransferred to nitrocellulose membranes for immunoblotting.

Immunoprecipitation-immunoblot assays were performedwith M. pneumoniae and M. genitalium solubilized inTDSET (10 mM Tris [pH 7.8], 0.2% sodium deoxycholate,0.1% SDS, 10 mM tetrasodium EDTA, 1% Triton X-100)containing 1 mM phenylmethylsulfonyl fluoride. Myco-plasma preparations were processed by procedures previ-ously described for the soluble-antigen radioimmuno-precipitation (RIP) assay (1). Briefly, solubilized antigen wasmixed with MAb and incubated for 15 min at 37°C and thenovernight at 4°C. Antigen-antibody complexes were isolatedby the addition of 50 [L of a 10% suspension of washedformaldehyde-treated protein A-bearing Staphylococcus au-reus Cowan 1. After a 90-min incubation at 4°C, the antigen-antibody-protein A complexes were pelleted and washedfour times with TDSET. Pellets and supernatants wereprepared in SP buffer for SDS-PAGE (7.5% separating gel).Gel proteins were transferred onto nitrocellulose and probedwith MAb preparations as described in the text.

In certain experiments, proteolytic digestion of M. pneu-moniae and M. genitalium was performed before SDS-PAGE. M. pneumoniae and M. genitalium organisms (400pLg each) were aliquoted (500 [lI each) into microcentrifugetubes pretreated with Sigmacote (Sigma) to reduce nonspe-cific protein absorption. Each tube received 0, 5, 50, or 100pLg of protease for 30 min at 37°C. Enzyme inhibitors (at aconcentration two-fold greater than the appropriate enzyme)were added to halt the proteolytic digestion, and tubes wereincubated at 0°C for 10 min. Samples were then centrifuged(14,000 x g) and washed once with cold PBS. Resultingpellets were solubilized in SP buffer and used as antigen inthe previously described immunoblot.

Inhibition of M. pneumoniae attachment to chicken RBCs.Attachment of M. pneumoniae to chicken erythrocytes(RBCs) was used to assess the blocking capacity of MAbs(11, 20). Briefly, [35S]methionine-labeled mycoplasmas (109organisms in 300-pA samples) were added to siliconizedmicrocentrifuge tubes containing either 300 pul of PBS orPBS plus MAb. Tubes were rocked at 37°C for 1 h followedby the addition of chicken blood (6.0% [vol/vol] final dilu-tion; Granite, Burlington, N.C.). After 1 h at 37°C, RBC-coated mycoplasmas were gently washed, and the boundradioactivity was determined.

Immunofluorescence. Mycoplasmas were grown on 12-mmglass cover slips in their respective growth media. After 48 h,the medium was aspirated, and the cover slips were washedthree times with PBS. Glass-attached mycoplasmas werethen incubated for 2 h at 37°C in 1 ml of PBS-BSA to blocknonspecific binding sites. Column-purified MAbs were di-luted to 50 ,ug/ml in PBS-BSA and added to the cover slips.After a 2-h incubation at 37°C, cover slips were washed asbefore and incubated for 1 h at 37°C with a 1:50 dilution (inPBS-BSA) of rhodamine-labeled goat anti-mouse immuno-globulin (Hyclone). Cover slips were then washed andexamined with a Leitz Ortholux II fluorescence microscope.

Chemicals. All enzymes and enzyme inhibitors were ob-

INFECT. IMMUN.

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

MYCOPLASMA SPECIES CROSS-REACTIVE ANTIGENS 51

A B C D E F

-_

200 K

92K

68K

FIG. 1. RIP of sera from experimentally infected chimpanzeesno. 992 (A and B) and no. 941 (C and D). Preinfection sera (A and C)and postinfection sera (B and D) were reacted with "S-labeled M.genitalium. Lane E, total profile of solubilized M. genitaliumproteins; lane F, molecular weight standards (K = 1,000). Arrowindicates the 140-kDa protein).

the anti-140-kDa protein MAbs which were tested boundwith the same intensity. Some appeared to bind weakly (Fig.2B) or not at all, implying that the epitope which theyrecognized was oriented in such a manner as to stericallyhinder binding or was not fully accessible. As a knownpositive control, MAbs to the P1 adhesin were incubatedwith M. pneumoniae (Fig. 2C). Figure 2D demonstrates thelack of binding by an anti-140-kDa protein MAb to M.pneumoniae.

Previous studies have demonstrated that protein P1 of M.pneumoniae is highly trypsin sensitive (2). When M.genitalium was exposed to various concentrations of tryp-sin, the intensity of the 140-kDa protein band was unaltered(Fig. 3A and B). Two small bands just below the 140-kDaprotein were bound by the anti-140-kDa protein MAb (Fig.3B), yet these bands were not present when polyclonalantiserum was used (Fig. 3A). Even at concentrations oftrypsin which were previously shown to digest protein P1readily and which clearly digested several other M. geni-talium proteins, the 140-kDa protein was unaffected.To further examine the protease sensitivity of the 140-kDa

protein, we exposed M. genitalium to several enzymes at 5,50, and 250 ,ug/500 ,g of mycoplasmas. Digestion wasperformed at 37°C for 30 min. Papainase at 50 and 250 ,ug

tained from Sigma. The following enzymes were used:trypsin (EC 3.4.21.4, from bovine pancreas) type III;papainase (EC 3.4.22.2, from papaya latex) type IV; alpha-chymotrypsin (from bovine pancreas) type VII; Streptomy-ces griseus protease type XIV. The following inhibitors wereused: trypsin inhibitor (from turkey egg white) type II-T;L-1-tosylamide-2-phenyl-ethylchloromethyl ketone; N-a-p-tosyl-L-lysine chloromethyl ketone; and phenylmethyl-sulfonyl fluoride.

RESULTS

Immunodominance of 140-kDa protein of M. genitalium.Previous studies demonstrated the immunodominance ofprotein P1 of M. pneumoniae in human covalescent andhamster sera (15). In multiple mouse hybridoma fusions withM. pneumoniae as the antigen, we showed that the majorityof the resulting MAbs were specific for protein P1 (20).Similarly, when MAbs binding to air-dried M. genitalium inthe ELISA were analyzed by immunoblot or RIP, 52 of 126antibodies were reactive. Of these, 23 hybridomas secretedMAbs which bound specifically to the 140-kDa protein. Theremaining 29 clones secreted MAbs to a variety of otherproteins with no more than four antibodies showing identicalbinding patterns. Clearly, the 140-kDa protein was highlyimmunogenic in mice immunized with M. genitalium. It wasof interest to examine sera of experimentally infected chim-panzees to determine whether the 140-kDa protein would beimmunogenic in a primate species (24). Postinfection chim-panzee sera reacted strongly with the 140-kDa protein (Fig.1). These data encouraged further analysis of the surfaceaccessibility and protease sensitivity of this immunodomi-nant 140-kDa protein.

Characterization of 140-kDa protein of M. genitalium. Todetermine whether the 140-kDa protein ofM. genitalium wasexpressed on the mycoplasma surface, M. genitalium wasincubated with anti-140-kDa protein MAbs followed withrhodamine-conjugated anti-mouse immunoglobulin and ex-amined for immunofluorescence. The 140-kDa protein isclearly surface accessible for binding by MAb H23.17H7(anti-140-kDa protein) (Fig. 2A). Interestingly, not all 13 of

FIG. 2. Immunofluorescence of M. genitalium (A and B) reactedwith H23.17H7 (A,-anti-140-kDa protein) and H23.2F8 (B, anti-140-kDa protein) and M. pneumoniae (C and D) reacted with H13.6E7(C, anti-Pl) and H23.2F8 (D, anti-140-kDa protein).

VOL. 55, 1997

-W

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

52 MORRISON-PLUMMER ET AL.

A B C D

-., *A qqlw

P w4

0 100 50C D 100 500

FIG. 3, Immunoblot of M. genitalium after exposure toO, 100, or500 ,ug of trypsin and analysis by 7.5% SDS-PAGE. Blots wereprobed with rabbit anti-M. genitalium serum at a 1:5,000 dilution (A)and H23.17H7 (anti-140-kDa protein) (B). Arrow indicates the140-kDa protein of M. genitalium.

(Fig. 4, lanes G and H) and S. griseus protease at 50 and 250 ,ug(lanes J and K) markedly digested the 140-kDa protein. Chy-motrypsin had only a minimal effect on the 140-kDa protein.

Proteolytic peptide mapping of M. pneumoniae protein P1.In a recent study, Clyde and Hu (6) reported on one of agroup of M. pneumoniae anti-Pl MAbs which bound in animmunoblot to a 100-kDa protein of M. genitalium. UsingRIP, we had previously examined several anti-Pi MAbs andwere unable to identify any which precipitated proteins ofM. genitalium (3). In an effort to understand these possiblyconflicting reports, we screened a large group of anti-PlMAbs for reactivity to M. genitalium. To effectively imple-ment this study, we categorized the anti-Pl MAbs into three

A B C D E F GH I J K*. - w _* -

3gp:gI3i:u_mjjat4d.I-

FIG. 4. Immunoblot of enzymatic digests of M. genitaliumprobed with rabbit anti-M. genitalium serum at a 1:1,000 dilution,Lanes: A, molecular weight standards; B, no enzyme; C, chymo-trypsin, 5 ,ug; D, chymotrypsin, 50 jig; E, chymotrypsin, 250 ,ug; F,papainase, 5 jig; G, papainase, 50 ,Ig; H, papainase, 250 jig; I, S.griseus protease, 5 jig; J, protease, 50 jig; K, protease, 250 ,ug.Arrow indicates the 140-kDa protein.

FIG. 5. Immunoblot of trypsin-digested protein P1 (no trypsin,lane A; 10 p.g of trypsin, lanes J3 to D). Blots were probed withanti-Pl MAbs H12.5B8 (class 2) (A and B), H10.8E8 (class 3) (C),and H13.6E7 (class 1) (D).

classes based on their immunoblot profiles against trypsin-digested protein P1. Figure 5 presents the profiles of eachMAb class. Undigested protein P1 was bound by all threeclasses of MAbs (Fig. 5A). Yet, upon tryptic digestion,MAbs in each of the three classes bound with a unique andhighly reproducible pattern. These results led us to postulatethat these MAbs bound to three separate epitopes.

Competition ELISA with MAbs. To confirm the uniquespecificities of the anti-Pl MAbs, we performed a series ofcompetition ELISAs. Anti-Pl MAbs of the class 1, 2, or 3type were conjugated to alkaline phosphatase enzyme (Table1). The labeled antibody was then competed with unlabeledanti-Pl antibodies as indicated. Inhibition occurred onlywhen MAbs were competed with enzyme-labeled MAbs ofthe same class. The results reinforced the immunoblot data

TABLE 1. Competition assay: inhibition of binding of alkalinephosphatase-conjugated anti-Pl MAbs to M. pneumoniae

Amt of inhibition at inhibiting MAb concn ofa:

Anti-Pl MAbs 2.5 0.5 0.05 0.005,uLg/well ,|Lg/well ,utg/well ,uwg/well

H13.6E7 (class 1)bH13.6E7 (class 1)C 0.025d NTe NT NTH5.2G4 (class 1) 0.055 0.079 0.323 0.598H12.21B8 (class 2) 0.686 0.651 0.709 0.646H12.5B8 (class 2) 0.672 0.672 0.737 0.665H10.8E8 (class 3) 0.706 0.660 0.639 0.621

H12.5B8 (class 2)bH13.6E7 (class 1)C 0.482 0.476 0.381 0.393H12.21B8 (class 2) 0.053 0.094 0.214 0.369H10.8E8 (class 3) 0.375 0.411 0.403 0.403

H10.8E8 (class 3)bH13.6E7 (class 1)C 1.680 1.993 1.997 1.650H12.21B8 (class 2) 2.008 1.912 1.986 1.449H10.8Eg (class 3) 0.070 0.155 0.823 1.591

a MAbs specific to P1 protein of M. pneumoniae were diluted to 2.5, 0.5,0.05, and 0.005 ,ug per microtiter well and added to triplicate wells coated withair-dried M. pneumoniae.

b Alkaline phosphatase-conjugated MAb.c Unlabeled competing MAbs.d Values are means of a minimum of triplicate assays. Optical density

readings measured at 405 nm. Standard deviation less than 20%o of mean.eNT, Not tested.

A B

INFECT. IMMUN.

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

MYCOPLASMA SPECIES CROSS-REACTIVE ANTIGENS 53

A B A B A B A B A B-7_

0

..M

--

--a.....

:W.

I w

I

a b d eC

FIG. 6. Immunoblot analysis of anti-Pl MAb classes reacted with M. genitalium and M. pneumoniae. Antibodies H5.2G4 and H13.6E7 (aand b, respectively, class 1 MAb) bind the 165-kDa protein P1 of M. pneumoniae (lane B) and cross-react with a 140-kDa protein of M.genitalium (lane A). Antibodies H12.21B8 (c, class 2), H12.5B8 (d, class 2), and H10.8E8 (e, class 3) recognize only the P1 protein in M.pneumoniae. 0, Protein P1; ~, 140-kDa protein.

demonstrating that the three distinct classes of anti-Pl MAbsbound to different epitopes on the P1 molecule.A class of MAbs to protein P1 of M. pneumoniae binds

specifically to the 140-kDa protein of M. genitalium. Onceclassified, representative anti-Pl MAbs from the threeclasses were used to determine whether protein P1 and the140-kDa protein were antigenically related. As anticipatedfrom previous experiments (3), none of the MAbs in any ofthe three classes of anti-Pl MAbs precipitated a cross-reactive protein of M. genitalium when analyzed by RIP.However, when these same antibodies were reacted byimmunoblot methodology, one of the classes bound to a140-kDa protein of M. genitalium (Fig. 6). None of theantibodies in the other two classes of anti-Pl antibodiesreacted by immunoblot with M. genitalium.MAbs to 140-kDa protein of M. genitalium identify a

cross-reactive epitope of M. pneumoniae. Identification of aclass of anti-Pl MAbs which bound the 140-kDa protein ofM. genitalium provided the rationale for examining anti-M.genitalium (140-kDa protein) MAbs for cross-reactivity withM. pneumoniae protein P1. When screened by immunoblot,a group of anti-140-kDa protein MAbs reacted with proteinP1 (data not shown). None of the anti-140 antibody prepa-rations bound protein P1 when assayed by RIP.To understand this discrepancy between immunoblot and

RIP results, we performed a series of immunoprecipitation-immunoblot experiments using column-purified anti-140-kDaprotein and anti-Pl MAbs. MAbs were reacted with unla-

beled TDSET-solubilized mycoplasmas in a manner similarto the soluble-antigen RIP (1), and the antigen-antibodycomplex was precipitated with protein A-bearing S. aureus.Pellets and supernatants were then electrophoresed andblotted onto nitrocellulose for probing with the same anti-body used in the precipitation part of the assay or analternate MAb.MAbs were able to precipitate only those proteins against

which they were generated (i.e., homologous immunogens)(Fig. 7). Anti-Pi MAb precipitated only protein P1, reflectedby the presence of P1 in the antigen-antibody-S. aureuspellet (Fig. 7a, lane A), and anti-M. genitalium 140-kDaprotein MAb precipitated only the 140-kDa protein (Fig. 7b,lane A). These same antibodies were unable to precipitatethe cross-reactive protein in the heterologous antigen prep-aration as indicated by the absence of antigen in the anti-body-S. aureus pellets (Fig. 7c, lane A, and d, lane A). Itappeared that the MAbs were only able to bind to thecross-reactive heterologous antigen when the antigen waspresented on a fixed matrix, such as nitrocellulose (Fig. 7c,lane B, and d, lane B). This experimental sequence wasperformed with both anti-Pl and anti-140-kDa protein MAbsas probes, and results were consistent with the above.

Variation in MAb affinity for homologous and heterologousantigens. Because of the apparent immunologic relationshipsbetween M. pneumoniae protein P1 and the 140-kDa proteinof M. genitalium, we decided to determine antibody affinitiesby ELISA in homologous and heterologous mycoplasma

VOL. 55, 1987

%dwomw -- - . -,

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

54 MORRISON-PLUMMER ET AL.

~~~~~~~~~~~~~~~.

"... ... .. ......... _;; ,

b c d

FIG. 7. Immunoblot of S. aureus protein A-precipitated pellets(A) and supernatants (B) of M. pneumoniae and M. genitalium.After centrifugation of antigen-antibody-S. aureus protein A com-plexes, pellets and supernatants were electrophoresed on a 7.5%SDS-acrylamide gel and transferred to nitrocellulose for immuno-blotting with H23.17H7 (anti-140-kDA protein) as the probe. Precip-itation experiments were performed as follows: a, H13.6E7 (anti-Pl)plus M. pneumoniae; b, H23.17H7 plus M. genitalium; c, H13.6E7plus M. genitalium; d, H23.17H7 plus M. pneumoniae. Arrowsindicate protein P1 (a) and the 140-kDa protein (b).

systems. Initial antigen titrations demonstrated that theoptimal antigen concentration for air-dried whole mycoplas-mas was 2 jig of protein per well. Additional antigentitrations were performed with various dilutions of SDS-PAGE-eluted proteins (P1 and 140 kDa). Optimal bindingoccurred at a concentration of approximately 3 to 5 ng perwell. Using these antigen concentrations, we screened atotal of 12 anti-140-kDa protein and three anti-Pl MAbs byELISA at antibody concentrations of 0.001 to 5.00 jig permicrotiter well. ELISAs were performed with both M.pneumoniae and M. genitalium air-dried whole antigenpreparations and gel-eluted protein P1 and 140-kDa proteins.Care was taken to coordinate the timing of the assays so thateach MAb was tested simultaneously on the four antigens,thereby eliminating test variability which might arise fromsequential assays. Figures 8A, B, and C are representative ofthe binding curves. In Fig. 8A and B, MAbs H13.6E7(anti-Pl) and H23.1F11 (anti-140-kDa protein) bound to thehomologous antigen preparations with optical density valuesup to 1,000-fold greater than that obtained with the heterol-ogous antigen. This was most clearly observed in the gel-eluted preparations. Interestingly, at high antibody concen-trations (>1 ,ug per well), 3 of the 12 anti-140-kDa proteinMAbs were shown to bind strongly to the heterologouswhole M. pneumoniae antigen preparation (Fig. 8C). Non-specific binding was not responsible for this observationsince none of the MAbs (even at 10 to 50 ,ug per well) boundto other mycoplasma species (M. salivarium, M. orale, M.hominis) or A. laidlawii (data not shown).

Anti-140-kDa protein MAbs block attachment of M. pneu-moniae to chicken RBCs. We previously demonstrated thatMAbs to proteins of M. pneumoniae block attachment ofmycoplasmas to chicken RBCs (20). The significance of thebinding curve shown in Fig. 8C was apparent when it was

E

at)

0

40-

ci

200

120

40

2001

1201

40

0

0

'-_ ~ ~ ~_ -__

5 1 .5 .1 .05 .01 .001

ug of Purified Antibody per WellFIG. 8. Diagram of binding curves of MAbs H13.6E7 (anti-

protein P1) (A), H23.1F1j (anti-140-kDa protein) (B), and H37.3C6(anti-140-kDa protein) (C) reacted with M. pneumoniae (0) or M.genitalium (0). Air-dried whole organisms (top graphs) and purifiedgel-eluted protein P1 of M. pneumoniae or the 140-kDa protein ofM.genitalium (bottom graphs) were used as antigens.

observed that a select group of anti-140-kDa protein MAbsblocked M. pneumoniae attachment to chicken RBCs (Table2). Only those anti-140-kDa protein MAbs which demon-strated the binding pattern seen in Fig. 8C inhibitedcytadherence. All other anti-140-kDa protein MAbs testedhad no significant effect on M. pneumoniae attachment.Antibody H37.5C9 which bound the 31- and 68-kDa proteinsof M. genitalium (Table 2) also had no effect on attachment,although these proteins were demonstrated to be surfaceaccessible by immunofluorescence (data not shown). Asexpected, strong surface binding of H37.3C6 (anti-140-kDa

A B

'SINEW

Pi B

.,:.:F.

..

200

120

40

200

1201

_

A B A B.. : f ... : . .... ..... ............. ... s........... .

.:> f: .. .... .*:...: s_

..

*: :.s . .:* ,'.j,B,SI,

*: :.2. .',

, :::.,.,''..w, .::':

'.'' '''.' .'

R...

a

200

to- Whole Organism Ao M pneumonioe* M genitalium

10-

Gel Purified Protein

J0, --- ---0-----

to Whl Organis

o Whole Organism

::

10

0 Gel Purified Protein

WholeOrganismC

to Gel Purified Protein

INFECT. IMMUN.

-4

1201

40,

2001

1201

401

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

MYCOPLASMA SPECIES CROSS-REACTIVE ANTIGENS 55

TABLE 2. Effect of MAbs on M. pneumoniae attachment tochicken RBCs

Mol wt (103) p.g of %MAb of antigena MAb/tubeb InhibitioncH12.5B8 165 200 82.6H23.24A3 140 200 16.8H37.15F8 140 200 15.4H23.2F8 140 200 13.5H37.5C9 31, 68 200 19.4H23.22E7 140 200 41.4

100 40.0H23.17H7 140 200 54.3

100 40.0H37.3C6 140 200 73.0

100 68.650 26.0

a All MAbs were generated against M. genitalium except H12.5B8 (anti-M.pneumoniae protein P1).

b Final volume of 0.35 ml per microcentrifuge tube.c Percent inhibition determined as 100 minus percent control. Percent

control = average counts per minute of test samples divided by averagecounts per minute of PBS control samples x 100. Duplicate wells were usedper test sample, and data represent a minimum of triplicate assays.

protein) to M. pneumoniae was observed by immunofluo-rescence, which reaffirmed the attachment data in Table 2.

DISCUSSION

In this paper we used MAbs to characterize a 140-kDasurface-accessible protein of M. genitalium which, in con-trast to M. pneumoniae protein P1, was determined to beinsensitive to trypsin digestion. The surface location of thisprotein was documented by a variety of techniques includingimmunofluorescence (Fig. 1), whole-cell ELISA (Fig. 8), andaccessibility of the protein to papain and protease digestion(Fig. 4). The relationship between the M. pneumoniae adhe-sin P1 and the 140-kDa protein of M. genitalium wasassessed by using three classes of anti-Pi MAbs which weredistinguished by their immunoblot patterns with proteolyti-cally digested P1. Partial enzymatic proteolysis of proteins isa frequently used method for mapping peptides (5), and P1 isan ideal protein for mapping since it is easily digested bytrypsin and produces a highly reproducible pattern (2). Oneof the resulting classes of anti-Pl MAbs appeared to bind across-reactive antigen located on both the P1 and 140-kDaproteins.

Similarities between P1 and the 140-kDa protein such assurface location, immunodominance in experimentally in-fected animals, and the shared cross-reactivity of anepitope(s) on these two molecules led us to propose that the140-kDa protein of M. genitalium might be functionallyactive in cytadherence. This hypothesis was supported byimmunoprecipitation-immunoblot studies and attachmentdata which showed that one group of the anti-140-kDaprotein MAbs identified a unique determinant which wascross-reactive with protein P1 and blocked attachment ofM.pneumoniae to RBCs. Unfortunately, technical difficultieswith M. genitalium prevented us from performing reproduc-ible assays to determine whether these same MAbs inhibitM. genitalium attachment to RBCs.Data presented help to explain a discrepancy between

immunoblot and RIP results. The RIP required high-affinityantibodies to precipitate antigen-antibody-S. aureus proteinA complexes. Antibody affinity was not as critical a factorwhen the antigens were presented in a manner which re-

quired simple binding (i.e., on a solid matrix such as anELISA or immunoblot) as opposed to precipitation. Yeteven in the solid matrix binding assays, it was necessary forlow-affinity antibodies to be at relatively high protein con-centrations for binding to occur.The usefulness of MAbs for molecular analysis of micro-

organisms should not be underestimated; however, ourstudies underscore both the advantages and disadvantagesassociated with these immunologic probes. Their ability todistinguish unique epitopes enabled us to demonstrate atleast one cross-reactive determinant among two im-munodominant proteins that may have functional similaritiesin M. pneumoniae and M. genitalium. Our data also dem-onstrate that an inherent characteristic of MAbs is theirvariable affinity, and multiple techniques are necessary toensure proper interpretation of experimental results.

ACKNOWLEDGMENTS

We express our gratitude to Rose Garza for her expert secretarialassistance. For technical assistance we thank Belen Puleo-Scheppke, Cheryl Cummings, and Diana Drouillard. We also thankBelen Puleo-Scheppke for preparation of the artwork in this manu-script.

This research was supported by Public Health Service grant Al18540 from the National Institute of Allergy and Infectious Diseasesand by Cistron Biotechnology, Inc.

LITERATURE CITED1. Baseman, J. B., R. M. Cole, D. C. Krause, and D. K. Leith. 1982.

Molecular basis for cytadsorption of Mycoplasma pneumoniae.J. Bacteriol. 151:1514-1522.

2. Baseman, J. B., D. L. Droufilard, D. K. Leith, and J. Morrison-Plummer. 1985. Role of Mycoplasma pneumoniae adhesin P1and accessory proteins in cytadsorption, p. 18-23. In S. E.Mergenhagen and B. Rosan (ed.), Molecular basis of oralmicrobial adhesion. American Society for Microbiology, Wash-ington, D.C.

3. Baseman, J. B., D. L. Drouiliard, D. K. Leith, and J. G. Tully.1984. Absence of Mycoplasma pneumoniae cytadsorption pro-tein P1 in Mycoplasma genitalium and Mycoplasma gal-lisepticum. Infect. Immun. 43:1103-1105.

4. Beaman, K. D., and J. D. Pollack. 1981. Adenylate energycharge in Acholeplasma laidlawii. J. Bacteriol. 146:1055-1058.

5. Cleveland, D. W., S. G. Fischer, M. W. Kirschner, and U. K.Laemmli. 1977. Peptide mapping by limited proteolysis in so-dium dodecyl sulfate and analysis by gel electrophoresis. J.Biol. Chem. 252:1102-1106.

6. Clyde, W. A., and P. C. Hu. 1986. Antigenic determinants of theattachment protein ofMycoplasma pneumoniae shared by otherpathogenic mycoplasma species. Infect. Immun. 51:690-692.

7. Edward, D. G. 1947. A selective medium for pleuro pneumonia-like organism. J. Gen. Microbiol. 1:238-243.

8. Feldner, J., U. Gobel, and W. Bredt. 1982. Mycoplasma pneu-moniae adhesin localized to tip structure by monoclonal anti-body. Nature (London) 298:765-767.

9. Hu, P. C., W. A. Clyde, and A. M. Collier. 1984. Conservationof pathogenic mycoplasma antigens. Isr. J. Med. Sci. 20:916-919.

10. Hu, P. C., R. M. Cole, Y. S. Huang, J. A. Graham, D. E.Gardner, A. M. Collier, and W. A. Clyde, Jr. 1982. Mycoplasmapneumoniae infection: role of a surface protein in the attach-ment organelle. Science 216:313-315.

11. Krause, D. C., and J. B. Baseman. 1983. Inhibition of Myco-plasma pneumoniae hemadsorption and adherence to respira-tory epithelium by antibodies to a membrane protein. Infect.Immun. 39:1180-1186.

12. Krause, D. C., D. K. Leith, and J. B. Baseman. 1983. Reacqui-sition of specific proteins confers virulence in Mycoplasmapneumoniae. Infect. Immun. 39:830-836.

VOL. 55, 1987

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from

56 MORRISON-PLUMMER ET AL.

13. Krause, D. C., D. K. Leith, R. M. Wilson, and J. B. Baseman.1982. Identification of Mycoplasma pneumoniae proteins asso-ciated with hemadsorption and virulence. Infection. Immun.35:809-817.

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

15. Leith, D. K., L. B. Trevino, J. G. Tully, L. B. Senterfit, and J. B.Baseman. 1983. Host discrimination of Mycoplasma pneumo-niae proteinaceous immunogens. J. Exp. Med. 157:502-514.

16. Lind, K. 1982. Serological cross-reaction between Mycoplasmagenitalium and M. pneumoniae. Lancet ii:1158-1159.

17. Lind, K., B. 0. Lindhardt, H. J. Schutten, J. Blom, and C.Christiansen. 1984. Serological cross-reactions between Myco-plasma genitalium and Mycoplasma pneumoniae. J. Clin. Mi-crobiol. 20:1036-1043.

18. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

19. Morrison-Plummer, J., D. H. Jones, and J. B. Baseman. 1983.An ELISA to detect monoclonal antibodies specific for lipiddeterminants of Mycoplasma pneumoniae. J. Immunol. Meth-ods 64:165-178.

20. Morrison-Plummer, J., D. K. Leith, and J. B. Baseman. 1986.Biological effects of anti-lipid and anti-protein monoclonal anti-

bodies protein on Mycoplasma pneumoniae. Infect. Immun.53:398-403.

21. 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.

22. Tully, J. G., D. Taylor-Robinson, R. M. Cole, and D. L. Rose.1981. A newly discovered mycoplasma in the human urogenitaltract. Lancet i:1288-1291.

23. Tully, J. G., D. Taylor-Robinson, D. L. Rose, R. M. Cole, andJ. M. Bove. 1983. Mycoplasma genitalium, a new species fromthe human urogenital tract. Int. J. Syst. Bacteriol. 33:387-396.

24. Tully, J. B., D. Taylor-Robinson, D. L. Rose, P. M. Furr, C. E.Graham, and M. F. Barile. 1986. Urogenital challenge of pri-mate species with Mycoplasma genitalium and characteristicsof infection induced in chimpanzees. J. Infect. Dis. 153:1046-1054.

25. Tully, J. G., R. F. Whitcomb, H. F. Clark, and D. L. Williamson.1977. Pathogenic mycoplasmas: cultivation and vertebratepathogenicity of a new spiroplasma. Science 195:892-894.

26. Voller, A., D. Bidwell, and A. Bartlett. 1976. Microplate enzymeimmunoassays for the immunodiagnosis of virus infections, p.506-512. In N. R. Rose and H. Friedman (ed.), Manual ofclinical immunology. American Society for Microbiology,Washington, D.C.

INFECT. IMMUN.

on February 1, 2016 by guest

http://iai.asm.org/

Dow

nloaded from