INFECTION AND IMMUNITY, Dec. 1991, p. 4349-4356 Vol. 59, No. 120019-9567/91/124349-08$02.00/0Copyright © 1991, American Society for Microbiology

Specificity of Antibodies to O-Acetyl-Positive and O-Acetyl-Negative Group C Meningococcal Polysaccharides in

Sera from Vaccinees and CarriersGAYATHRI ARAKERE* AND CARL E. FRASCH

Division ofBacterial Products, Center for Biologics Evaluation and Research, Food andDrug Administration, Bethesda, Maryland 20892

Received 26 April 1991/Accepted 6 September 1991

Most group C Neisseria meningitidis strains produce an 0-acetyl-positive polysaccharide, a homopolymer ofa-2->9-linked N-acetylneuraminic acid with 0-acetyl groups at the C-7 and C-8 of its sialic acid residues. Themajority of disease isolates have been reported to contain this polysaccharide. Some strains produce group Cpolysaccharide lacking 0-acetyl groups. The licensed vaccine contains the 0-acetyl-positive polysaccharide. Wehave measured the antibody specificities to the two polysaccharides in sera from asymptomatic group Cmeningococcal carriers and vaccinated adults by a new enzyme-linked immunosorbent assay (ELISA)procedure using methylated human serum albumin for coating the group C polysaccharide onto microtiterplates. Inhibition of binding of serum antibodies to polysaccharide-coated plates was measured by ELISA afterincubation with 0-acetyl-positive and 0-acetyl-negative group C polysaccharides. Greater inhibition of bindingof carrier sera was observed with the homologous polysaccharide. There was substantial inhibition of bindingof vaccinee sera to the 0-acetyl-positive polysaccharide-coated plate following preincubation with 0-acetyl-positive polysaccharide, but homologous inhibition on plates coated with the 0-acetyl-negative polysacchariderequired much higher concentrations of polysaccharide. Carrier sera may have a higher proportion ofantibodies with greater specificity for the 0-acetyl-negative polysaccharide, while vaccinee sera containantibodies with greater affinity for the 0-acetyl-positive polysaccharide. Studies with monoclonal antibodiesspecific for 0-acetyl-positive and 0-acetyl-negative polysaccharides reveal that the percentage of group Cmeningococcal disease caused by 0-acetyl-negative strains remains about 15%, as found over 15 years ago.

Group C Neisseria meningitidis is responsible for approx-imately 35% of meningococcal disease in the United States.N. meningitidis is carried asymptomatically in the upperrespiratory tract by about 10% of the human population.Nasopharyngeal carriage of meningococcal strains as well asconvalescence from the disease have been shown to result inthe production of bactericidal antibodies associated withprotective immunity (8, 13, 21, 28). The human nasopharynxis the only known reservoir of the organism, and most peoplewho contract the disease do not do so by contact with anindividual with meningococcal disease. Therefore, asympto-matic carriers are presumed to be the major source oftransmission of pathogenic strains (6).Many of the studies of meningococcal carriage have

involved group B, for which it has been shown that acquisi-tion and carriage of poorly pathogenic strains possessingserotype 2 surface proteins results in development of anti-bodies that could protect against a virulent group B serotype2 strain (15). Although carriage of group B serotypes mayconfer some protection against group C meningococcalinfections, little information is available on antibody re-sponses and protective immunity due to carriage of group CN. meningitidis. Vaccinated individuals receive purifiedpolysaccharide, whereas carriers and patients are exposedpresumably to the same polysaccharide when it is an integralpart of the bacteria. This provides us with an opportunity tostudy possible similarities and differences in the immuneresponses between these two groups.

* Corresponding author.

Most group C N. meningitidis strains produce an 0-acetyl-positive (OAc+) polysaccharide which is a homopolymer ofa-2--9-linked N-acetylneuraminic acid. The 0-acetyl groupsare distributed exclusively between C-7 and C-8 of its sialicacid residues. Some group C strains produce a polysaccha-ride lacking 0-acetyl groups (OAc-). About 85% of group Cdisease strains are reported to be OAc+, and the currentlyavailable vaccine contains the OAc+ polysaccharide (1, 15,18). Studies using monoclonal antibodies indicate that, al-though the polysaccharide is a homopolymer of sialic acid, itappears to express at least three different epitopes (22, 23).The present study was undertaken to examine the specificityof antibodies to OAc+ and OAc- group C polysaccharides(GCPS) in sera from group C N. meningitidis carriers,noncarriers, and vaccinated adults.Enzyme-linked immunosorbent assay (ELISA) has been

used for measuring the antibody responses against severalbacterial polysaccharides (4, 5, 11, 24). Negatively chargedpolysaccharides do not attach well to the polystyrene com-monly used in a solid-phase ELISA. Several approacheshave been used for attachment of the polysaccharide, includ-ing precoating the plates with highly charged polymers suchas poly-L-lysine (12). We, however, found poly-L-lysine togive nonspecific absorbance with some sera. In this paper wereport a new ELISA procedure which uses methylatedhuman serum albumin (mHSA) to coat the polysaccharideon the solid surface with very little nonspecific binding.Methylated albumin is positively charged and has beenextensively used in the purification of negatively chargedpolymers like nucleic acids (7, 16).

4349

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

4350 ARAKERE AND FRASCH

MATERIALS AND METHODS

Bacteria and sera. N. meningitidis group C strains Cll,MC19, BB639, BB649, BB1098, BB1122, BB1146, BB620,BB657, BB666, BB688, BB1088, and BB1148 and strainM986-NCV-1 (a noncapsulated variant of N. meningitidisgroup B strain M986) were from the investigators' collection.

Sera from 119 adults, of which 6 were meningococcalgroup C carriers, 12 were group B carriers, and the rest werenoncarriers, were provided to us for analysis by JamesThomas of the Los Angeles County Public Health Depart-ment (27). Pre- and postvaccination sera were obtained fromlaboratory personnel. Acute- and convalescent-phase serafrom children with group C meningococcal disease werefrom our laboratory collection. The two mouse monoclonalantibodies (2055.5 and 2016.3) used in this study were fromKathryn Stein and Leonard Rubinstein, Center for BiologicsEvaluation and Research, Food and Drug Administration(22, 23).

Purification of GCPS and outer membranes. OAc+ andOAc- GCPS were prepared by the procedure of Gotschlichafter strains Cll and MC19 were grown in modified Frantzmedium (14). Briefly, bacterial cultures grown overnightwere centrifuged, and 10% hexadecyl trimethyl ammoniumbromide (Sigma Chemical Co., St. Louis, Mo.) was added tothe supernatant to a final concentration of 0.1% to precipi-tate the polysaccharide. The pellet was suspended in 1 Mcalcium chloride to dissolve the polysaccharide-detergentcomplex. A series of phenol extractions, ethanol precipita-tions, and high-speed centrifugations were done to purify thepolysaccharide. OAc+ and OAc- GCPS were run on Seph-arose CL-4B, and molecular sizes and elution profiles weresimilar for both. Purity of the polysaccharides was ascer-tained by sialic acid (26) and protein (19) analyses and alsoby the presence or absence of 0-acetyl groups as determinedby nuclear magnetic resonance spectroscopy, performed byWilliam Egan, Center for Biologics Evaluation and Re-search, Food and Drug Administration. Outer membranevesicles for use in ELISA were prepared from 65-h culturesupernatants from strain M986-NCV-1 as described previ-ously (9).

Polysaccharide ELISA. mHSA was prepared by themethod of Mandell and Hershey (20) with 25% albumin (forinjection) as the starting material. A mixture of mHSA andGCPS was made for coating Immulon 1 plates (DynatechLaboratories, Inc., Chantilly, Va.) by diluting stock solu-tions of mHSA (5 mg/ml in water) and GCPS (1 mg/ml inwater) in phosphate-buffered saline, pH 7.4 (PBS). Aftervarious concentrations were tried, a mixture of mHSA andGCPS each at a concentration of 5 ,ug/ml was found to beoptimal, with minimum antibody binding to mHSA alone.The plates were coated for 6 to 7 h at 28°C and washed fourtimes with saline containing 10 mM Tris (pH 7.5) and 0.1%Brij 35 (Sigma) by using an automatic plate washer (Dyna-tech). The standard reference serum (PB-2, assigned 2,400 Uof GCPS antibody per ml) used in the assay was seriallydiluted in twofold steps, starting with an initial dilution of1:100, in PBS containing 0.1% Brij 35 and 5% newborn calfserum (serum-conjugate buffer). Test sera were diluted ap-propriately in the same buffer to yield absorbance valueswithin the linear region of the standard curve. Reference andtest sera were applied to the plate in triplicate (100 ,ul perwell) and incubated overnight at 4°C. The plates werewashed the next day, and an appropriate dilution of goatanti-human immunoglobulin-alkaline phosphatase conjugate(affinity-purified immunoglobulin G [IgG] + IgM + IgA;

Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md.)in the serum-conjugate buffer was added to the plates (100 ,u1per well) and incubated at 28°C for 2 h. The plates werewashed, and the color was developed with the addition of100 ,ul (1 mg/ml) of p-nitrophenyl phosphate (Sigma no. 104)in 1 M Tris-0.3 mM MgCl2 (pH 9.8). The plates were read at405 nm 50 to 60 min after the addition of the substrate. Therecorded absorbance value was extrapolated to 100 min fornormalization. Units of antibody in the test sera weredetermined by using a reference curve constructed by linearregression of logit-log-transformed data obtained with thePB-2 human reference serum (11). Serum PB-2 was obtainedfrom an adult by plasmapheresis after immunization withmeningococcal polysaccharide vaccine. A value of 2,400 Uof anti-GCPS antibody per ml was assigned by comparisonwith a known anti-meningococcal outer membrane referenceserum (10). Total antibodies to the outer membrane proteins(OMP) were measured by ELISA (11).

Screening of group C meningococcal strarins with monoclo-nal antibodies specific for OAc+ and OAc- GCPS. Severaldisease isolates of group C N. meningitidis recovered fromdiverse geographic locations were grown, and their polysac-charides were purified as described above. Sialic acid wasestimated in each of these preparations by the resorcinolmethod (26). Equivalent amounts of sialic acid (5 ,ug/ml)from each strain were mixed with 5 p.g of mHSA per ml, and100 RI per well was coated on Immulon 1 plates. Half theplate was incubated with an appropriate dilution of mono-clonal antibody specific for OAc+ GCPS, and the other halfwas incubated with monoclonal antibody specific for OAc-GCPS. After overnight incubation at 4°C, plates werewashed and incubated with goat anti mouse IgG3-alkalinephosphatase conjugate (Southern Biotechnology Associates,Inc., Birmingham, Ala.) for 2 h at 28°C. Plates were proc-essed as described above.

Inhibition assays. Inhibition studies were carried out byincubating twofold serial dilutions of OAc+ and OAc- GCPSstarting from 10 ,ug/ml for 4 h at room temperature with asingle serum dilution shown to be in the linear range ofabsorbance in a direct ELISA. The 4-h incubation ofpolysaccharide with the serum dilution was chosen aftercomparing incubations of 2 to 12 h. The binding appeared toreach an equilibrium by 4 h. The plates were slowly rotatedon an orbital shaker during this time, after which 100 Rl wastransferred to an OAc+ GCPS-coated plate, and another 100,ul was transferred an OAc- polysaccharide-coated plate.The plates were incubated overnight at 4°C, washed, andthen incubated with anti-human conjugate and processed asdescribed above. Inhibition studies were also done withmouse monoclonal antibodies 2055.5 and 2016.3 in an iden-tical manner, using goat anti-mouse IgG3 conjugate.

RESULTS

GCPS ELISA. Serial dilutions of sera from carriers andnoncarriers were incubated with plates coated with a mix-ture of mHSA and GCPS to measure total antibodies toGCPS as described in Materials and Methods. Typical directbinding curves of absorbance plotted against serum dilutionfor sera from a carrier and a noncarrier are shown in Fig. 1.The differences in antibody levels between the two sera areevident from the figure.Measurement of antibodies to GCPS and OMP in carriers

and noncarriers. Sera from group C meningococcal carriers,group B carriers, and noncarriers were analyzed for antibod-ies against GCPS and OMP prepared from a nonencapsu-

INFECT. IMMUN.

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

ANTIBODIES TO GROUP C MENINGOCOCCAL POLYSACCHARIDE 4351

CO

w

1--

z

8!-w

0

am

_-.

1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0 1 2 3 4 5 6

LOG 2 (1/100) SERUM DILUIlON

FIG. 1. Direct-binding curves obtained in an mHSA-polysaccha-ride ELISA. Sera from a carrier and a noncarrier were seriallydiluted and incubated on plates coated with OAc+ GCPS plusmHSA. *, noncarrier; +, carrier.

lated variant of a group B strain (Table 1). Group C carriershad significantly higher antibodies to both GCPS and OMPthan did non-group C carriers. In contrast, sera from groupB carriers showed an increase in antibodies only againstOMP. Although the overall values for noncarriers were lowfor antibodies against GCPS and OMP when compared withvalues for carriers, the range in antibody units per milliliterwas large. Figure 2 is a bar graph depicting the percentagesof noncarriers, group C carriers, and group B carriers inthree different categories based on their antibody valuesagainst GCPS and OMP. The three categories were (i) highGCPS antibody and high OMP antibody, (ii) low GCPSantibody and high OMP antibody and, (iii) low GCPS anti-body and low OMP antibody. These categories contained 12,34, and 64%, respectively, of the noncarriers. The majority(83%) of group C carriers were in the first category, and 83%of group B carriers were in second category.Measurement of antibodies to OAc+ and OAc- GCPS by

ELISA. Sera from noncarriers, group C carriers, and adultsvaccinated with meningococcal tetravalent vaccine (OAc+GCPS), as well as acute- and convalescent-phase sera frompatients with group C meningococcal disease, were exam-ined by ELISA using GCPS containing or lacking 0-acetylgroups (Table 2). Although the binding of carrier and vac-

cinee sera to OAc- GCPS-coated plates was slightly higherthan binding to OAc+ GCPS-coated plates, the differencewas not statistically significant.

Competitive inhibition assays with sera from carriers andvaccinees. Competitive inhibition ELISAs were done withsera from group C meningococcal carriers to measure theantibody responses specific for OAc+ and OAc- GCPS (Fig.

90 _

80I

L 70

1)60U.

0

wi50-

z

w

~30-

20-

10

PS>14,OMP>15 PS<14,OMP>15 PS<14,OPM<15

ANTIBODY UNITS/ML AGAINST GCPS AND OMP

FIG. 2. Bar graph of percentages of noncarriers (U) and group C(E ) and B (0) carriers, with antibody values for OAc+ GCPS(PS) and OMP in units per milliliter, belonging to the following threegroups: PS > 14, OMP > 15; PS < 14, OMP > 15; PS < 14, OMP< 15.

3). Absorption with OAc+ polysaccharide inhibited thebinding of carrier sera to the OAc+ polysaccharide-coatedplate more than absorption with OAc- polysaccharide, andcomparably greater homologous inhibition was observedwith the OAc- polysaccharide-coated plates.Although vaccinee sera bound similarly to OAc+ and

OAc- polysaccharide-coated plates in the direct ELISA(Table 2), the competitive ELISA results were not as we hadexpected. Figure 4 shows the inhibition curves for tworepresentative serum samples. The binding of serum A toOAc+ polysaccharide-coated plates was inhibited 80% with10 ,ug of OAc+ polysaccharide per ml but only 15% with 10,ug of OAc- GCPS per ml. At the same concentration, theOAc+ and OAc- GCPS inhibited the binding to the OAc-GCPS-coated plate by 80 and 70%, respectively. The inhibi-tion of binding of serum B was similar, although the amountof antibody in serum B against the OAc- GCPS was twicethat against the OAc+ GCPS in a direct ELISA; serum A hadequal binding to both polysaccharides. Thus, Fig. 4 illus-trates that OAc+ GCPS was more efficient than OAc- GCPSin inhibiting the binding of vaccinee sera to OAc- GCPS-coated plates, which was not the case with carrier sera. Theconcentrations of OAc+ and OAc- GCPS needed for 50%inhibition of binding of six carrier and four vaccinee sera toOAc+ and OAc- GCPS were calculated from several inhi-bition curves (Table 4). The amounts of OAc+ GCPS re-

quired for 50% inhibition of binding of carrier and vaccineesera to the OAc+ GCPS-coated plate were very similar, butthere were significant differences in the amount required forinhibition of binding to the OAc- GCPS-coated plate. The

TABLE 1. Comparative antibody levels in sera of adults to group C meningococcal antigens as measured by ELISA

Geometric mean antibody level, U/ml (1 SD confidence interval)Study group n

GCPS (OAc+) OMP

Noncarrier 97 2.1 (0.7-1.3) 7.7 (1.6-36.6)Group B carrier 12 1.2 (0.7-3.9); NS' 40.2 (14.7-110.3); P < 0.001Group C carrier 6 19.0 (9.7-37.3); P < 0.001 45.7 (15.7-133.3); P < 0.001

a Significance determined by Student's t test in comparison with noncarrier geometric mean antibody level. NS, not significant.

VOL. 59, 1991

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

4352 ARAKERE AND FRASCH

TABLE 2. Levels of antibodies to 0-acetyl-positive and 0-acetyl-negative GCPS in carriers, vaccinated adults, and patients

Geometric mean antibody level,Study group n U/ml (1 SD confidence interval)

OAc+ OAc-

Noncarriers 4 1.85 (0.65-2.05) 2.11 (0.53-1.93)Group C carriers 6 15.98 (9.22-37.13) 19.75 (9.16-30.10)Vaccinated adults 11 30.69 (6.94-61.09) 36.29 (7.58-89.75)Patients: acute 4 5.06 (0.50-7.68) 3.50 (0.13-9.34)Patients: convalescent 5 12.23 (2.63-13.47) 9.15 (0.94-8.60)

concentration of OAc- GCPS needed for 50% inhibition ofbinding of vaccinee sera to the OAc- GCPS-coated platewas significantly higher than that needed for inhibition ofbinding of carrier sera.

Monoclonal antibodies with specificity for OAc+ and OAc-GCPS. Figure 5 illustrates the specificity of binding ofmonoclonal antibody 2055.5 to OAc+ GCPS. In a directELISA, this monoclonal antibody did not bind to the OAc-GCPS-coated plate. Incubation of 2055.5 with OAc- GCPSdid not result in inhibition of binding to the OAc+ GCPS-coated plate either after incubation overnight at 4°C or for 4h at room temperature. Another monoclonal antibody,

10.00 5.00 2.50 1.25 0.63 0.31 0.16

CONCENTRATION OF POLYSACCHARIDE ug/mi

FIG. 3. Inhibition of binding of carrier sera 74 and 93 to OAc+ and OAc- GCPS-coated Immulon 1 plates. The sera were preincubated withOAc+ and OAc- GCPS for 4 h and then transferred to OAc+ and OAc- GCPS-coated plates. Inhibitor: O and 0l, OAc+ GCPS; A and +,OAc- GCPS.

100

z0

Iz

C.)a:CL

80ol

Bound PS: OAc-Serum 93

Serum 74

60F

401

20 [

0

INFECT. IMMUN.

N

N 46- - - - -A--

-96I...

*.-O.- - - - -0.,..

11%

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

ANTIBODIES TO GROUP C MENINGOCOCCAL POLYSACCHARIDE 4353

10.00 5.00 2.50 1.25 0.63 0.31 0.16

20 _

10.00 5.00 2.50 1.25 0.63 0.31 0.16

CONCENTRATION OF POLYSACCHARIDE ug/miFIG. 4. Inhibition of binding of vaccinee sera A and B to OAc+ and OAc- GCPS-coated Immulon 1 plates. Vaccinee sera were

preincubated with OAc+ and OAc- GCPS for 4 h and then transferred to OAc+ and OAc- GCPS-coated plates. Inhibitor: O and O, OAc+GCPS; A and +, OAc- GCPS.

2016.3, did not bind to the OAc+ GCPS-coated plate (datanot shown) and exhibited a very similar inhibition patternwith OAc- GCPS, and OAc+ polysaccharide did not inhibitthe binding of the monoclonal antibody to the OAc- GCPScoated plate (data not shown).

Screening of meningococcal strains with monoclonal anti-bodies 2055.5 and 2016.3. Thirteen recent isolates of group CN. meningitidis from various geographic regions of theUnited States were examined for the presence of 0-acetyldeterminants (Table 3). Eleven of the 13 reacted only withmonoclonal antibody 2055.5, and two reacted with bothmonoclonal antibodies 2016.3 and 2055.5. Polysaccharides

isolated from these two strains, BB 1088 and BB 1146, werefound to be 0-acetyl negative by 13C nuclear magneticresonance analysis. Whole-cell ELISA was also performedwith each of these strains, but the data were not conclusive,as some of the strains made much less polysaccharide thanothers; therefore, additional studies were done after thepolysaccharides were purified from each strain.

DISCUSSION

Several different methods for attaching polysaccharides tothe solid phase for ELISA have been attempted, such as

100

80F

601-

401-

Bound PS: OAc+ Serum A

.- SerumB

\ *vv

\ '+~~~~~~..

-4

'0.

201-

0

100

z01=-

z

zw0w0L

80 _

60 _

40 _

Bound PS: OAc-Serum A

------ Serum B

-----------

0

VOL. 59, 1991

i .--tV,

I

r

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

4354 ARAKERE AND FRASCH

100

80

z0o

z

0wIL

60

40

2.50 1.25 0.63 0.31 0.16 0.08 0.04

CONC OF POLYSAOCHARIDE ug/ml

FIG. 5. Inhibition of binding of monoclonal antibody 2055.5 to OAc+ GCPS-coated Immulon 1 plates. Monoclonal antibody 2055.5 wasincubated with OAc+ (U) and OAc- (+) GCPS for 4 h at room temperature and then transferred to OAc+ GCPS-coated plates.

tyramination of the polysaccharide and covalent attachmentof a positively charged molecule like poly-L-lysine (4). Thesemethods have the disadvantage that they can alter the nativeconformation of the polysaccharide. Precoating the platewith poly-L-lysine seemed ideal, but we found poly-L-lysineto react nonspecifically with some human sera and thereforeto be unsuitable. Use of mHSA in attaching the polysaccha-ride to Immulon 1 plates was found to alleviate much of thenonspecific binding found with poly-L-lysine (17).

Carriage of group C meningococci increased the levels ofantibodies to both GCPS and OMP (Table 1). The extent of

TABLE 3. Screening of N. meningitidis strains with monoclonalantibodies specific for OAc+ and OAc- GCPS

Binding withmonoclonal

Strain Yr State Sero Sero antibodyisolated group type

2055.5 2016.3(OAc+) (OAc-)

BB617 1987 Okla. C 2aP1.2 + -

BB620 1987 Wash. C 2aP1.2 + -

BB639 1987 Calif. C 2aP1.2 + -

BB649 1987 N.J. C 2aP1.2 + -

BB657 1987 Calif. C NTP1.2 + -

BB666 1987 Tenn. C NTP1.2 + -

BB688 1987 Calif. C 2aP1.2 + -

BB1088 1986 Mich. C NT nt + +BB1098 1986 N.J. C 2bP1.2 + -BB1122 1987 Ohio C 2aP1.2 +BB1146 1984 Nev. C 2a nt + +BB1147 1984 N.Y. C NTP1.2 + -BB1148 1984 N.Y. C 2aP1.2 +C1l C + -MC19 C - +3006 B 2bP1.2 - -M986 B 2aP1.2

the increase is probably dependent on the duration ofcarriage and the amount of polysaccharide made by thecarrier strains. The antibody response may also be depen-dent on earlier exposure to other bacteria sharing antigenicdeterminants with the GCPS. Known meningococcal carri-ers uniformly had high levels of antibody to OMP. About12% of the noncarriers had high levels of antibodies to boththe GCPS and OMP and therefore may have been recentgroup C meningococcal carriers (Fig. 2). Of the noncarriers,34% had a high level of antibody to the OMP but low levelsto GCPS and presumably were recent carriers of meningo-cocci other than group C. In contrast, a few had high levelsof antibody to the GCPS and low levels to OMP. Theseindividuals presumably had recent exposure to anotherbacteria with a structurally related polysaccharide. How-ever, vaccination induced higher levels of antibody to theGCPS than either meningococcal carriage or exposure toother cross-reactive polysaccharides.

It had been observed in earlier studies that the OAc-GCPS may be more immunogenic in children and adults thanthe OAc+ polysaccharide. Individuals vaccinated with apreparation made from OAc- GCPS had twice the amount ofanticapsular as well as bactericidal antibodies as did individ-uals vaccinated with OAc+ polysaccharide, although thedifference was not statistically significant (12, 25). In thesestudies, the antibodies were measured in the sera generatedfrom vaccination with OAc+ and OAc- GCPS by radioim-munoassay with only OAc+ GCPS. In our study, levels ofbinding of humoral antibodies from group C carriers andfrom vaccinated individuals to OAc+ and OAc- GCPS weresimilar in a direct ELISA (Table 2).

Inhibition curves for carrier sera (Fig. 3) showed betterinhibition with the respective homologous polysaccharides,as anticipated, and indicated that the heterogeneity in theinhibition patterns of OAc+ and OAc- GCPS was due to atleast in part to differences in specificity. Homologous inhi-bition of binding of antibodies from vaccinees to the OAc+

INFECT. IMMUN.

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

ANTIBODIES TO GROUP C MENINGOCOCCAL POLYSACCHARIDE 4355

TABLE 4. Concentrations of GCPS producing 50% inhibition of antibody binding to OAc+ and OAc- GCPS

OAc+ GCPS bound OAc- GCPS bound

Serum Concn (pLg/ml) yielding 50% inhibition by: Concn (,ug/ml) yielding 50% inhibition by:

OAc+ GCPS OAc- GCPS OAc+ GCPS OAc- GCPS

Carrier sera1 6.1 >10.0 >10.0 2.72 3.5 4.8 >10.0 3.83 0.6 >10.0 >10.0 4.04 1.5 >10.0 >10.0 1.55 0.8 2.3 >10.0 1.96 0.8 1.4 >10.0 2.6Geometric mean 1.5 5.0 10.0 2.61 SD confidence interval (0.6-3.8) (2.1-11.7) (10.0-10.0) (1.8-3.8)

Vaccinee sera1 2.0 10.0 5.0 4.52 3.5 >10.0 6.0 7.03 2.0 >10.0 0.31 4.04 0.7 >10.0 0.75 3.5Geometric mean 1.7 10.0 1.6 4.61 SD confidence interval 0.9-3.4 10.0-10.0 0.4-6.9 3.4-6.2Significanceb NS NS P < 0.01 P < 0.05

a Values of >10 were considered as 10 for calculation of geometric means.b Significance determined by Student's t test by comparing geometric mean concentrations (micrograms per mililiter) for cararier and vaccine sera when

inhibited by (i) OAc+ GCPS and the bound GCPS is OAc+ (ii) OAc- GCPS and the bound GCPS is OAc+, (iii) OAc+ GCPS and the bound GCPS is OAc-, and(iv) OAc- GCPS and the bound GCPS is OAc.

GCPS-coated plate was substantial. There was, however,little inhibition with OAc- GCPS. Inhibition of binding ofvaccinee sera to OAc- GCPS-coated plates was better withOAc+ GCPS than OAc- GCPS. This suggests that a largerproportion of the antibodies resulting from immunizationwith GCPS have more specificity for the OAc+ polysaccha-ride than the antibodies resulting from group C meningococ-cal carriage. The concentrations of OAc+ GCPS needed for50% inhibition of binding of serum antibodies from vaccineesand carriers to OAc- GCPS-coated plates also substantiatethe above conclusion (Table 4). Data shown in the sametable indicate that the concentration of OAc- GCPS neededfor 50% inhibition of binding of carrier sera to the OAc-GCPS-coated plate is significantly lower than the concentra-tion needed for 50% inhibition of binding of vaccinee sera.Thus, sera from carriers appear to have a greater proportionof antibodies with higher specificity for the OAc- GCPSthan do sera from vaccinees.A small study in the 1970s suggested that most disease

isolates contained the OAc+ GCPS (2). Our limited dataobtained by using monoclonal antibodies specific for OAc+and OAc- GCPS suggest that approximately 85% of recentgroup C meningococcal disease isolates from the UnitedStates express OAc+ polysaccharide (Table 3). More strainshave to be screened to confirm this preliminary observation.However, there are no data available on the proportion ofcarrier strains that are OAc+ or OAc. It would be interest-ing to see if the proportion of OAc+- and OAc--specificantibodies in a carrier serum is dependent on whether thecarrier strain produced polysaccharide with or without0-acetyl groups and if there are more carrier strains thandisease strains lacking 0-acetyl groups.The inhibition studies with sera from carriers and vacci-

nated adults and monoclonal antibodies show that the OAc+and OAc- polysaccharides share at least one commonbackbone epitope. The absence and presence of 0-acetylgroups generate at least one unique epitope on each of the

two polysaccharides (22, 23). The proportions of epitopescommon and unique to OAc+ and OAc- GCPS in carrier andvaccinee sera are probably different, as vaccinees receiveonly OAc+ GCPS. It is also possible that the OAc- GCPShas a more stable conformation on a solid surface than insolution, as indicated by a decrease in the amount of OAc-GCPS needed for 50% inhibition of binding of vaccinee serawhen incubated with OAc- polysaccharide coated on amicrotiter plate (data not shown). In this regard, we foundthat the fine specificities of some monoclonal antibodieschanged depending upon whether the GCPS was bounddirectly to the plastic (Immulon 2) or through interactionwith a protein (mHSA; unpublished results). This may be thesame effect we observed when soluble OAc- GCPS failed toeffectively inhibit binding of human antibodies to the boundmHSA-GCPS complex.The present studies thus demonstrate that exposure to the

GCPS induces antibodies of multiple specificities. The na-ture of this exposure, i.e., carriage, disease, or vaccination,has a strong impact upon the predominant antibody speci-ficity.

ACKNOWLEDGMENTS

We thank Willie F. Vann and Roberta M. Shahin (Center forBiologics Evaluation and Research, Food and Drug Administration)for a critical review of the manuscript.

REFERENCES1. Apicella, M. A. 1974. Identification of a subgroup antigen on the

Neisseria meningitidis group C capsular polysaccharide. J.Infect. Dis. 129:147-153.

2. Apicella, M. A., and H. A. Feldman. 1976. Meningococcal groupC subgroup determinant detected by immunofluorescence.Proc. Soc. Exp. Biol. Med. 152:289-291.

3. Artenstein, M. S., R. Gold, J. G. Zimmerly, F. A. Wyle, H.Schneider, and C. Harkins. 1970. Prevention of meningococcaldisease by group C polysaccharide vaccine. N. Engl. J. Med.282:417-420.

VOL. 59, 1991

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

4356 ARAKERE AND FRASCH

4. Barra, A., D. Schulz, P. Aucouturier, and J. Preud'homme.1988. Measurement of anti-Haemophilus influenzae type b cap-sular polysaccharide antibodies by ELISA. J. Immunol. Meth-ods 115:111-117.

5. Beuvery, E. C., M. H. Kayhty, A. B. Leussink, and V. Kanhai.1984. Comparison of radioimmunoassay and enzyme-linkedimmunosorbent assay in measurement of antibodies to Neis-seria meningitidis group A polysaccharide. J. Clin. Microbiol.20:672-676.

6. Broome, C. 1986. The carrier state: Neisseria meningitidis. J.Antimicrob. Chemother. 18:25-34.

7. Corneo, G., L. Zardi, and E. Polli. 1972. Elution of humansattelite DNAs on a methylated albumin kieselguhr chromato-graphic column: isolation of satellite DNA. IV. Biochim.Biophys. Acta 269:201-204.

8. Frasch, C. E. 1983. Immunization against Neisseria meningiti-dis, p. 115-144. In C. S. F. Easmon and J. Jeljaszewicz (ed.),Medical microbiology, vol. 2. Academic Press, Inc., New York.

9. Frasch, C. E., and M. S. Peppler. 1982. Protection against groupB Neisseria meningitidis disease: preparation of soluble proteinand protein-polysaccharide immunogens. Infect. Immun. 37:271-280.

10. Frasch, C. E., and J. D. Robbins. 1978. Protection against groupB meningococcal disease. III. Immunogenicity of serotype 2vaccines and specificity of protection in a guinea pig model. J.Exp. Med. 147:629-644.

11. Frasch, C. E., J. M. Zahradnik, L. Y. Wang, L. F. Mocca, andC. M. Tsai. 1988. Antibody response of adults to an aluminumhydroxide-adsorbed Neisseria meningitidis serotype 2b proteingroup B-polysaccharide vaccine. 1988. J. Infect. Dis. 158:700-708.

12. Glode, M. P., E. B. Lewin, A. Sutton, C. T. Le, E. C. Gotschlich,and J. B. Robbins. 1979. Comparative immunogenicity of vac-cines prepared from capsular polysaccharides of group C Neis-seria meningitidis 0-acetyl positive and 0-acetyl negative var-iants and Escherichia coli K92 in adult volunteers. J. Infect.Dis. 139:52-59.

13. Goldschneider, I., E. C. Gotschlich, and M. S. Artenstein. 1969.Human immunity to the meningococcus. I. The role of humoralantibodies. J. Exp. Med. 129:1307-1326.

14. Gotschlich, E. C. 1975. Development of polysaccharide vaccinesfor the prevention of meningococcal diseases. Monogr. Allergy9:245-258.

15. Jones, D. M., and J. Eldridge. 1979. Development of antibodiesto meningococcal protein and lipopolysaccharide serotype anti-

gens in healthy carriers. J. Med. Microbiol. 12:107-111.16. Kothari, R. M., V. Shankar, and M. W. Taylor. 1972. RNA

fractionation on Kieselguhr columns. J. Chromatogr. 70:341-363.

17. Leinonen, M., and C. E. Frasch. 1982. Class-specific antibodyresponse to group B Neisseria meningitidis capsular polysac-charide: use of polylysine precoating in an enzyme-linkedimmunosorbent assay. Infect. Immun. 38:1203-1207.

18. Liu, T. Y., E. C. Gotschlich, W. Egan, and J. B. Robbins. 1977.Immunological studies on the meningococcal and relatedpolysaccharides. J. Infect. Dis. 136(Suppl.):S84-S92.

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

20. Mandell, J. D., and A. D. Hershey. 1960. A fractionating columnfor analysis of nucleic acids. Anal. Biochem. 1:66-77.

21. Reller, L. B., R. R. McGregor, and H. N. Beaty. 1973. Bacteri-cidal antibody after colonization with Neisseria meningitidis. J.Infect. Dis. 127:56-62.

22. Rubinstein, L. J., and K. E. Stein. 1988. Murine immuneresponse to the Neisseria meningitidis group C capsularpolysaccharide. I. Ontogeny. J. Immunol. 141:4352-4356.

23. Rubinstein, L. J., and K. E. Stein. 1988. Murine immuneresponse to the Neisseria meningitidis group C capsularpolysaccharide. II. Specificity. J. Immunol. 141:4357-4362.

24. Siber, G. R., C. Priehs, and D. V. Madore. 1989. Standardiza-tion of antibody assays for measuring the response to pneumo-coccal infection and immunization. Pediatr. Infect. Dis. J.8:384-391.

25. Steinoff, M. C., E. B. Lewin, E. C. Gotschlich, and J. B.Robbins. 1981. Group C Neisseria meningitidis variant polysac-charide vaccines in children. Infect. Immun. 34:144-146.

26. Svennerholm, L. 1957. Quantitative estimation of sialic acids. II.A colorimetric resorcinol-hydrochloric acid method. Biochim.Biophys. Acta 24:604-611.

27. Thomas, J. C., N. S. Bendana, S. H. Waterman, M. Rathbun, G.Arakere, C. E. Frasch, V. Magsombol, J. Clark, and J. D.Wenger. 1991. Risk factors for carriage of meningococcus in theLos Angeles County men's jail system. Am. J. Epidemiol.133:286-295.

28. Zollinger, W. D., J. Boslego, L. 0. Froholm, J. S. Ray, E. E.Moran, and B. L. Brandt. 1987. Human bactericidal antibodyresponse to meningococcal outer membrane protein vaccines.Antonie van Leeuwenkhoek. J. Microbiol. 53:403-411.

INFECT. IMMUN.

on May 17, 2020 by guest

http://iai.asm.org/

Dow

nloaded from