INFECTION AND IMMUNITY, OCt. 1970, p. 392-397Copyright (C) 1970 Ameriican Society for Microbiology

Vol. 2, No. 4Prinited int U.S.A.

Physicochemical Properties of Neisseria meningitidisGroup C and Y Polysaccharide Antigens

MICHAEL A. APICELLA' AND JOHN A. ROBINSON

Aerospace Mlledical Laborcatory (Cliniical), Wilfordl Hall USAF Meclical Ce&tter, aniid USAF EpidemiologicalLaboratorY, Lacklanid Air Force Base, Texas 78236

Received for publication 1 May 1970

Chemical and physical studies of the Neisseria meningitidis group C and Y antigensindicated that these preparations are acidic polysaccharides. The group C antigencontains predominately n-acetyl neuraminic acid, with small amounts of glucose,glucosamine, and mannose, whereas the group Y antigen contains glucose, glucosa-mine, and mannose. The molecular weights of the group C and Y antigens were

found to be 88,000 and 197,000, respectively. The group C antigen was susceptible toneuraminidase digestion. Finally, the immunological reactivity of neither antigenwas inhibited in hemagglutination or immunodiffusion systems by neutral or sub-stituted sugars.

Recent interest in the development of meningo-coccal polysaccharide antigens for use as humanimmunogens necessitates detailed studies of theirimmunological, biological, chemical, and physicalproperties (3). The group C and Y organisms arepresently of clinical interest because of the pre-dominance of the former in clinical disease andthe latter in carrier rate studies (17, 18, 27). Theisolation and immunochemical properties of thegroup C and Y polysaccharide antigens have beenpreviously described (20). We have completedchemical and physical studies of these antigensby colorimetric analysis, gas chromatography,infrared spectroscopy, electrofocusing, sedimen-tation velocity, and sedimentation equilibrium.The results indicate that these antigen prepara-tions are heterosaccharide polymers with distinctphysical and chemical differences.

MATERIALS AND METHODS

Antigens. Clinical isolates of N. mnieuiingitiidis groupC and primary nasopharyngeal isolations of group Yorganisms were used for antigen preparation by themethod of Robinson and Apicella (20).

Antisera. Rabbit antisera against N. meniingitidisgroups C and Y were obtained from the NationalCommunicable Disease Center (NCDC), Atlanta,Ga., or by immunization of New Zealand albinorabbits as described previously (20).

Chemical analysis. Antigen samples for analysiswere dialyzed against distilled water, lyophilized, anddried over P205. Total nitrogen determinations were

1 Present address: Department of Medicine, Stalte Uniiversity ofNew York, Buffalo, N.Y. 14215.

done on an F and M carbon-hydrogen-nitrogenanalyzer with glucosamine hydrochloride as referencestandard. Colorimetric assays for pentoses were doneby the method of Dische (7), 6-deoxyhexoses by thecysteine-sulfuric acid reaction (9), hexoses by theanthrone method (23), and heptoses by the method ofDische (8). Hexosamines were estimated by themethod of Rondle and Morgan (21), with glucosaminehydrochloride treated under similar conditions asstandard. N-acetyl hexosamines were estimated by themethods of both Aminoff, Morgan, and Watkins (1)and Reissig, Strominger, and LeLoir (19), with N-acetyl glucosamine (Sigma Chemical Co.) as stand-ard. Total uronic acids were determined by the car-bazole method (6), with a D-glucuronic acid as stand-ard. Total sialic acids were assayed by the modifiedresorcinol technique (4). 2-Keto-3-deoxy-sugars weredetermined by the thiobarbituric acid assay of Warren(29), with crystalline N-acetyl neuraminic acid(NANA; Sigma Chemical Co.) as standard.

Paper chromatography. Acid hydrolysates of neu-tral and amino sugars were dried in vacuo over P205and NaOH, separated on Whatman no. I paper by theascending method with ii-butyl alcohol-pyridine-water (6:4:3) or pyridine-ethyl acetate-acetic acid-water (5:5:1:3), and detected with aniline diphenyla-mine, Morgan-Elson, or 1% ninhydrin spray reagents.Sialic acids were detected as previously described(24).

Gas-liquid chromatography. Gas-liquid chroma-tography was done on a Barber Colman model 5000chromatograph equipped with a hydrogen flameionization detector, electrometer with l0-1 ampsensitivity, Honeywell recorder, and disc integrator.Unless otherwise stated, determinations were doneisothermally at an oven temperature of 160 C with a

392

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from

N. MENINGITIDIS POLYSACCHARIDE ANTIGENS

nitrogen carrier gas flow of 45 ml/min, on 6-ft (1.8-meter) U-shaped glass columns of 1% (w/w) OV-17Polymer on 100 to 120 mesh Gas Chrom-Q (RegisChemical). The O-trimethylsilyl ether derivatives ofboth sample and standards were prepared by themethod of Sweely and Walker (25). The methanolysisconditions for both 2- to 4-mg samples and standardswere 2 N methanolic HCl for 6 to 24 hr at 100 C. Hy-drolysates were passed over amberlite CG-45 andDowex 50W-X4 (to remove amino sugars) columns(10 by 1 cm) and dried under N2 prior to silylation.Trimethylsilyl ether derivatives of group C hydroly-sates (0.5 N methanolic-HCl, 80 C, 60 min) for detec-tion of sialic acids were chromatographed on 10%SE-30 on 80 to 100 mesh Gas Chrom-S columns withthe use of a programmed temperature increase of 2.0C/min (180 to 260 C) and were compared with au-thentic NANA under identical conditions.The methyl glycosides were identified both by rela-

tive retention times with reference to an internalstandard (L-fucose) and designation of an arbitraryvalue of 1.0 to a-glucose, and by co-chromatographywith standards derived under identical conditions. In-jection sample volume was 1 to 4 ,uliters.

Infrared absorption spectra. Infrared absorptionspectra were determined with a Beckman IR-9 spec-trophotometer by use of 1.5-mg samples of standard orantigen pressed into potassium bromide discs andscanned at 80 cm-' for 3.2 min.Enzyme studies. Group C antigen samples were hy-

drolyzed with Clostridiwn perfringens neuraminidase(type V, Sigma Chemical Co.), with an activity of0.08 unit/mg of n-acetyl neuraminic-lactose substrate.Antigen samples (500 ,ug) were incubated at pH 5.2for 48 hr at 37 C with a total of 0.08 unit of enzymeand were assayed for free sialic acid by the thiobarbi-turic acid (TBA) technique.

Inhibition studies. Studies were undertaken in anattempt to inhibit interaction between the group Cand Y antigens and their specific antisera by simpleand substituted sugars in immunodiffusion (16) andhemagglutination inhibition (10) systems. The tech-nique used was a modification of the method ofUllman and Cameron (26) and employed crystallineNANA, glucose, fructose, mannose, sucrose, glucosa-mine, galactose, galactosamine, and glucuronic acidin concentrations up to 5 mg/ml as inhibitors.

Sedimentation velocity. Sedimentation velocity stud-ies for calculation of s20,o were performed by use of amodel E ultracentrifuge with schlieren optics (5). Sam-ples were centrifuged at 201,366 X g at a constanttemperature of 20 C in phosphate-buffered saline,pH 7.2 (PBS). Plates were measured by use of a two-dimensional Gaertner microcomparator. Partial spe-cific volumes and viscosities of both antigens weremeasured at 15, 20, and 25 C at concentrations of 1, 4,and 8 mg/ml in PBS (22).

Molecular weights. Molecular weights were de-termined by the sedimentation equilibrium method ofLamm (14) with the use of schlieren optics and a short-column equilibrium cell. The schlieren phase plateangle was 80°. Rotor speeds of 10,589 and 11,272rev/min were used at a temperature of 20 C. The

columns were filled with 0.010 ml Dow 550 silicone oiland 0.11 ml of sample or 0.12 ml of solvent. Speci-mens were spun until equilibrium was obtained. Thiswas achieved at 24 hr. Measurements were made onplates obtained at 24 and 42 hr. Plates were measuredon a two-dimensional Gaertner microcomparater, andmolecular weights were calculated according to thefollowing formula:

1 dc2RT 2.303 d log- -

)(1-)2 d(r2)

Electrofocusing. The pl of the group C and Y anti-gens was determined by electrofocusing, by the methodof Vesterberg and Svenson (28). Ampholite buffer (pH3 to 5) was used in a sucrose density gradient with theanode at the bottom of the column. Voltage wasraised stepwise over 24 hr to 700 v and was maintainedat that level for an additional 24 hr. The sample sizewas 2 mg, and the sample was placed in the center ofthe column. Elution volume was 3 ml. The pH wasmeasured with a unipolar electrode with a BeckmanZeromatic pH meter. Antigen was localized by directhemagglutination and immunodiffusion with the useof specific NCDC antisera, by sialic acid analysis ofthe eluate for the group C antigen, and by the Elson-Morgan test for hexosamines for the group Y antigen.

Acrylamide gel electrophoresis. Electrophoresiswas done by use of 4% acrylamide-0.2% bisacrylamidegel in a pH 11.5 formate buffer according to themethod of Hilborn and Anastassiadis (12). Twentylambda samples containing 10 mg of antigen per mlwere placed on the gels and subjected to electrophore-sis at 2 mamp/gel for 4 hr. The gels were stained with0.5% Alcian Blue in 3% acetic acid.

RESULTS

Chemical studies. The group C antigen wasfound to be a heterosaccharide polymer contain-ing predominantly NANA and small amounts ofglucosamine, glucose, and mannose. The group Yantigen consisted of mannose, glucose, and glu-cosamine. Table 1 contains the results of thechemical analysis of both preparations. Neitherpreparation contained protein, pucleic acid, pen-toses, 6-deoxyhexoses, heptoses, uronic acid, orN-acetylated hexosamines. The optimal hy-drolysis conditions for the hexosamine analysisby the Elson-Morgan test were 2 to 6 hr in 2 NHC1 at 100 C. The paper chromatography andgas-liquid chromatography confirmed the colori-metric analysis of both antigens. An isothermalstudy of both group C and Y neutral hexoses canbe seen in Fig. 1. Co-chromatography with hexosestandards confirmed antigen peaks that had beenidentified by relative retention times. Optimalhydrolysis of the group C antigen with dilutesulfuric acid (0.1 N, 80 C, 60 min) resulted in 57%

VOL. 2, 1970 393

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from

APICELLA AND ROBINSON

TABLE 1. Chemical composition of Neisseriameningitidis group C and Y polysaccharide

antigens

Amt (pg/100 ;ig of dry antigen)Component

Group C Group Y

Glucose ............... 6.6 78.0Mannose .............. 2.0 0.6Glucosaminea ......... 4.2-7.3 7.6Sialic acidb............ 70.0 NPcNitrogend ............. 5.34 0.75

Expressed as free base.b Expressed as N-acetyl neuraminic acid.¢ Not present.d Total nitrogen by pyrolysis.

TBA-reactive material regardless of the group Cstrain studied. However, C. perfringens neura-minadase cleavage of the same group C strainsyielded 70% TBA-reactive material. Total neu-raminic acid content was identical in all group Cstrains studied and absent in all group Y strainsby both resorcinol and TBA techniques, indi-cating the lack of substituted NANA or other2-keto-3-deoxy-sugars in either antigen. TBA-reactive material was identified as NANA byelution at 245 C by programmed gas-liquidchromatography on 10% SE-30 and by migrationagainst NANA standards on paper chromatog-raphy. Infrared spectroscopy of two group Cstrain antigens and a crystalline NANA standardare shown in Fig. 2. The marked similarity of thetwo antigens to the NANA standard is apparent.In addition, three other group C antigens hadsimilar spectra. The infrared absorption spectrumof group Y antigens reveals strong banding at2,940 cm-' (assignment-CH2-), suggesting repeti-tion of a single-CH2-configuration in polymerform.Neuraminadase cleavage of the group C anti-

gen for 48 hr abolished the antigen's ability to in-hibit the hemagglutination of group C sensitizedsheep erythrocytes. Crystalline NANA or glucosa-mine at concentrations up to 5 mg/ml did not in-hibit hemagglutination with either antigen. Simi-larly, mannosamine, glucuronic acid, galactosa-mine, glucosamine, and their parent hexoses, aswell as sucrose and fructose in the same concen-trations, did not inhibit hemagglutination of thegroup C and Y sensitized sheep erythrocytes.Immunodiffusion studies done with 1 % agar gelscontaining the same sugar concentrations did notinhibit specific immunoprecipitation of eitherantigen.

Physical studies. The pl of the group C and Y

44z0

TiME (MINUTES)FIG. 1. Gas chromatography of group C and Y neu-

tral hexoses. Samples (2-,A liter) of trimethylsilyl etherderivatives of antigen hydrolysates were chromato-graphed at 160 C on 6-ft glass columns of 1% OV-17.The a- and 3-methyl glycoside of L-fucose are peaks1 and 2; peaks 3 and 4 represent mannose; and 5, 6, 7represent glucose. In the case of the group Y antigen,attenuation was increased during elution, so that allpeaks would be shown on the chromatogram.

antigens was 1.9 and 2.4, respectively. Figure 3shows the results of typical elution patterns of thegroup C and Y antigens after electrofocusing de-termined by hemagglutination. Studies were alsodone in which immunodiffusion and TBA reac-tivity were used to localize the group C peak, andthese coincided with the hemagglutination peaks.The group Y antigen precipitated at the point ofpolymer electroneutrality. Colorimetric analysisby the Elson-Morgan test for hexosamine waspositive, and the amino sugar was definitivelyidentified as glucosamine by paper chromatog-raphy. The dry weight of the precipitate accountedfor 95% of the antigen applied to the column.

394 INFEC. IMMUN.

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from

N. MENINGITIDIS POLYSACCHARIDE ANTIGENS

WAVENGTH IN AM NS

160

86ST

WAVEIUMNR IN CM

WAVELENGTh IN MICOS

WAVELENGTH IN MICRONS

WAVSMJtR IN CM1

WAeU N Ca

40

10

w'U

FIG. 2. Infrared spectra of crystalline NANA andgroup C and Y antigens.

Studies of partial specific volumes indicated thatat 4 mg/ml and 20 C the partial specific volumeof the group C antigen was 0.76 and of the groupY antigen was 0.69. In sedimentation velocity ex-periments, each antigen sedimented as a singlehomogeneous peak. The s2°0w for the group Cantigen was 3.36, and for the group Y antigen was3.84. Molecular weights of both antigens were

determined at two rotor speeds. The averagemolecular weight of the group C antigen was

88,000, and that of the group Y antigen was

197,000. Analytic acrylamide gel electrophoresisconfirmed this difference in molecular size, as

noted by the greater migration of the group Cantigen into 4% acrylamide gels (Fig. 4). In addi-tion, the group C antigen demonstrated bandingin 5, 6, and 7% gels, whereas the group Y antigenwas excluded from 6 and 7% gels.

GRP CGRP Y-

..2I6 2.0 2. 4 2 26 3 .2 34 3

FIG. 3. Electrofocusing ofthe group C and Yantigenin a pH 3 to 5 ampholyte buffer system at 700 voltsfor 48 hr. Antigenic activity was determined by thehemagglutination titer ofeluate-sensitizedsheep erythro-cytes.

DISCUSSIONThe study of acidic polysaccharides in recent

years has revealed their widespread presence innature and their potential biological importance(13). The meningococcal group C and Y antigensare acidic polysaccharide polymers. The group Cantigen contains predominately n-acetyl neu-raminic acid with small amounts of glucose,glucosamine, and mannose. The use of gas-liquidchromatography with its high degree of sensi-tivity has allowed detection of small amounts ofneutral sugars not previously found in this anti-gen. Watson (30) also detected hexosamine in hisgroup C preparation, and Gotschlich (11) foundonly n-acetyl neuraminic acid. The group C anti-gen, despite the presence of neutral and aminosugars, was susceptible to neuraminidase digestionwhich destroyed the antigenic activity of thepreparation. This would indicate that the neutraland amino sugars are present either terminally oras side chains to the repeating units of n-acetylneuraminic acid. The pI of this antigen indicatesa high degree of negative charge which is due tothe numerous polar groups on the n-acetyl neu-raminic acid molecule. These molecule elute atthe void volume of G-200 despite molecularweights of 88,000. This may be attributed to thehydrophilic character of the carboxyl group whichcreates hydration shells around the molecule, ex-cluding it from the molecular sieve (2).The group Y antigen is unique because of the

large amount of glucose present in the molecule.It also contains glucosamine and mannose but isdevoid of n-acetyl neuraminic acid by both TBAand resorcinol reactivity, by paper chromatog-raphy, and by gas-liquid chromatography. It has a

- ~ j -E

I1...,.<A?

395VOL. 2, 1970

2.5 10 x

so

S T 60

40 .

20 .

4

ST

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from

APICELLA AND ROBINSON

high negative charge as demonstrated by electro-focusing and is excluded from G-200 despite amolecular weight of 197,000, probably by thesame phenomena discussed above for the group Cantigen. The large amount of glucose in this anti-gen preparation gave us concern that we might behydrolyzing dextran eluting from the G-200 itself.Colorimetric and gas chromatographic studies ofG-200 column blanks failed to detect any neutralor substituted sugar. Neutral and substitutedsugars did not inhibit the interaction of theseantigens and their specific antibody. These resultsare similar to those of Mergenhagen (15) andsuggest that either chain length of the polymer orsome configurational site, or a combination of..... .. ..

_.. _*: . . ..

....: ...... _* ..... ::: :

. y\FIG. 4. Acrylamide gel electrophoresis of the group

C and Y antigens into 4% acrylamide-0.2% bis-acryla-mide gels.

both, is important in maintaining the antigenicdeterminant.

Because of the absence of heptoses and keto-deoxy-octonic sugars in these preparations, it isunlikely that they contain endotoxin. Studieswhich will be published shortly indicate thesepreparations have unique biological effects dis-tinct from those of endotoxin.

ACKNOWLEDGMENTS

We appreciate the cooperation and assistance of Jack Wallace,Sheldon Ladd, Richard Rerucha, Robert Booth, and JoAnn Homn.

LITERATURE CITED

1. Aminoff, D., W. T. J. Morgan, and W. M. Watkinis. 1952.Studies in immunochemistry. 11. Action of dilute alkalie onthe N-acetylhexosamine and specific blood group mucoids.Biochem. J. 51:379-389.

2. Apffel, C. A., and J. H. Peters. 1969. Tumors and serumglycoproteins. The "symbodies." Progr. Exp. Tumor Res.12:1-54.

3. Artenstein, M. S., R. Gold, J. G. Zimmerly, F. A. Wyle, H.Schneider, and C. Harkins. 1970. Prevention of meningococ-cal disease by Group C polysaccharide vaccine. N. Engl. J.Med. 282:417-421.

4. Cassidy, J. T., G. W. Jourdian, and S. Roseman. 1965. Thesialic acids. VI. Purification and properties of sialidase fromClostridiuni perfringenis. J. Biol. Chem. 240:3501-3506.

5. Chervenka, C. H. 1969. A manual of methods for the analyti-cal ultracentrifuge. Beckman Instruments Inc., Palo AltoCalif.

6. Dische, Z. 1947. A new specific color reaction of hexuronicacids. J. Biol. Chem. 167:189-198.

7. Dische, Z. 1949. Spectrophotometric method for the de-tennination of free pentose and pentose in nucleotides. J.Biol. Chem. 181:379-392.

8. Dische, Z. 1953. Qualitative and quantitative colorimetricdetermination of heptoses. J. Biol. Chem. 204:983-997.

9. Dische, Z., and L. B. Shettles. 1948. A specific color reactionof methyl pentoses and a spectrophotometric micromethodfor their determination. J. Biol. Chem. 175:595-602.

10. Gold, E. R., and H. H. Fudenberg. 1967. Chromic chloride; a

coupling reagent for passive hemagglutination reactions. J.Immunol. 99:859-866.

11. Gotschlich, E. C., T. Y. Liu, and M. S. Artenstein. 1969. Hu-man immunity to the meningococcus. III. Preparation andimmunochemical properties of the Group A, Group B andGroup C meningococcal polysaccharides. J. Exp. Med.129:1349-1365.

12. Hilborn, J. C., and P. A. Anastassiadis. 1969. Acrylamide gelelectrophoresis of acidic mucopolysaccharides. Anal. Bio-chem. 31:51-58.

13. Jeanloz, R. W. (ed.). 1969. The amino sugars, vol. IA, Chem-istry of amino sugars, p. vii. Academic Press Inc., NewYork.

14. Lamm, 0. 1929. Zur Throrie und Methodik der Ultracentri-fugenig. Z. Phys. Chem. (Leipzig) A143:177-182.

15. Mergenhagen, S. E., G. R. Martin, and E. Schiffman. 1963.Studies on an endotoxin of a Group C Neisseria men-ingitidis. J. Immunol. 90:312-317.

16. Ouchterloniy, 0. 1962. Diffusions-in-gel methods for immuno-logic analysis. Progr. Allergy. 6:30-154.

17. Pierce, W. 1969. Epidemiology of meningococcus at GreatLakes. Report of Meningitis Workshop, Naval Medical Re-search Unit No. 4, p. 19-21.

18. Reinarz, J. A. 1969. Group C meningococcal disease: Itsclinical presentation and problems in diagnosis and therapy.Report of Meningitis Workshop, Naval Medical ResearchUnit No. 4, p. 42-53.

396 INFEC. IMMUN.

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from

N. MENINGITIDIS POLYSACCHARIDE ANTIGENS

19. Reissig, J. L., J. L. Strominger, and L. F. LeLoir. 1955. Amodified colorimetric method for estimation of N-acetylamino sugars. J. Biol. Chem. 217:959-966.

20. Robinson, J. A., and M. A. Apicella. 1970. Isolation andcharacterization of Neisseria mnenintgitis group A, C, X, andY polysaccharide antigens. Infec. Immun. 1:8-14.

21. Rondle, C. J. M., and W. T. S. Morgan. 1955. The determina-tion of glucosamine and galactosamine. Biochem. J. 61:586-589.

22. Schachman, H. K. 1957. Ultracentrifugation, diffusion, andviscometry, p. 32-103. In S. P. Colowick and N. 0. Kaplan(ed.), Methods in emzymology, vol. 4. Academic Press Inc.,New York.

23. Scott, T. A., and E. H. Melvin. 1953. Determination of dextranwith anthrone. Anal. Chem. 25:1656-1661.

24. Svennerholm, E., and L. Svennerholm. 1958. Quantitativepaper partition chromnatography of sialic acids. Nature(London) 181:1154-1155.

25. Sweely, C. C., and B. Walker. 1964. Determination of carbo-hydrates in glycolipides and gangliosides by gas chroma-tography. Anal. Chem. 26:1465-1470.

26. Ullmann, W. W., and J. A. Cameron. 1969. Immunochemistryof the cell walls of Listeria nmonocyfogenes. J. Bacteriol. 98:486-493.

27. U. S. Department of Health, Education and Welfare, 1970.Morbidity and Mortality Report, Vol. 19, No. 9, p. 89-90.

28. Vesterberg, O., and H. Svensson. 1966. Isoelectric fractiona-tion analysis and characterization of ampholytes in naturalpH gradients. IV. Further studies on the resolving power inconnection with separation of myoglobulins. Acta Chem.Scand. 20:820-834.

29. Warren, L. 1959. The thiobarbituric acid assay of sialic acids.J. Biol. Chem. 234:1971-1975.

30. Watson, R. G., G. B. Marinetti, and H. W. Scherp. 1958. Thespecific hapten of Group C (group Ila) meningococcus. II.

Chemical nature. J. Immunol. 81:337-344.

VOL. 2, 1970 397

on August 28, 2019 by guest

http://iai.asm.org/

Dow

nloaded from