INFECTION AND IMMUNITY, Dec. 1994, p. 5545-5549 Vol. 62, No. 120019-9567/94/$04.00+0Copyright C) 1994, American Society for Microbiology

Synthesis and Some Immunologic Properties of an O-AcetylPectin [Poly(l1->4)-C-D-GalpA]-Protein Conjugate as a

Vaccine for Typhoid FeverSHOUSUN C. SZU,1* SLAVOMIR BYSTRICKY,"12 MARINA HINOJOSA-AHUMADA,1

WILLIAM EGAN,3 AND JOHN B. ROBBINS'Laboratoty of Developmental and Molecular Immunity, National Institute of Child Health and Human Development,

and Center for Biologics Evaluation and Research, Food and Drug Administration,3 Bethesda, Maryland 20892,and Institute of Chemistry Slovak Academy of Sciences, Bratislava, Slovak Republic2

Received 7 March 1994/Returned for modification 21 April 1994/Accepted 15 September 1994

Pectin, a plant polysaccharide, is mostly a linear homopolymer of poly(1-4)-a-D-GalpA with <5% neutralsugars: its molecular size has a broad distribution around 400 kDa, and the degree of esterification is <5%.The structure of the capsular polysaccharide of Salmonella typhi (Vi) differs from pectin in that it is Nacetylated at C-2 and 0 acetylated at C-3, and has a molecular size of -2 x 103 kDa. There is no serologicalcross-reaction between pectin and Vi. Pectin, when 0 acetylated at C-2 and C-3, is antigenically identical to Viin double immunodiffusion. Unlike Vi, 0-acetylated pectin (OAcPec) is not immunogenic in mice, probablybecause of its comparatively low molecular weight. After storage at 3 to 8°C for 3 months, there was no changein the O-acetyl content or the Mr of OAcPec. At 60°C, the Mr of OAcPec declined more rapidly than that of Vi.OAcPec conjugated to tetanus toxoid elicited Vi antibodies in mice, and reinjection elicited a booster response.The levels of Vi antibodies elicited by OAcPec-tetanus toxoid conjugates were lower than those elicited by Viconjugates, but these differences were not statistically significant. OAcPec has some advantages because it canbe measured by standardized colorimetric assays and because it forms more soluble conjugates with proteinsthan does Vi. One disadvantage is that its glycosidic bond is not as stable as that of Vi. The use of a plantpolysaccharide, pectin, as an immunogen for prevention of a systemic infection caused by a capsulatedpathogen (S. typhi) provides a novel approach to improve the preparation and immunogenicity of polysaccha-ride-based vaccines.

The capsular polysaccharide (Vi) is both an essential viru-lence factor and a protective antigen of Salmonella typhi (20).Field trials in Nepal and in the Republic of South Africashowed that a single injection of Vi conferred about 70%protection against typhoid fever in older children and in adults(1, 14). Its protective action is to elicit a critical level of serumantibodies.The limitations of Vi as a vaccine are (i) there is only -70%

efficacy in individuals 5 to 45 years old, (ii) there is anage-dependent serum antibody response (Vi elicited a com-paratively short-lived antibody responses in 2- to 5-year-oldchildren and only low levels of antibodies in a fraction ofchildren <2 years old), and (iii) reinjection did not elicit abooster response (T-cell independent) (15, 20). To increase itsimmunogenicity and to induce T-cell dependence, Vi wasconjugated to proteins (24, 25, 27). A clinical trial in adults inthe United States showed that Vi conjugates elicited signifi-cantly higher levels of serum antibodies than did Vi alone (27).Vi is a linear homopolymer of (1-4)-a-D-Ga4pANAc, vari-

ably 0 acetylated at C-3 (Fig. 1) (13, 20, 25). Synthesis of Viconjugates posed several problems. First, the high Mr of Vi(-2 x 103) causes conjugates to be poorly soluble. Second,standardization of Vi conjugates has been hindered by a lack ofa colorimetric method for quantification of this polysaccharide(22). Here, we investigated pectin as an immunogen for Vi.Pectin, a common polysaccharide of plants, is a homopolymercomposed mainly of poly(1-*4)-ct-D-GalpA (Fig. 1) (19); it has<5% neutral sugars. Treatment of pectin with acetic anhydride

* Corresponding author.

results in 0 acetylation of C-2 and C-3 (6, 23). The monomericstructure of 0-acetylated pectin (OAcPec) differs from that ofVi only in that its C-2 is 0 rather than N acetylated. As a result,OAcPec can be measured by the carbazole reaction. Pectindoes not react with Vi antiserum and does not elicit Viantibodies in animals. OAcPec, in contrast, as shown bySzewczyk and Taylor, precipitates with Vi antiserum (23).Unlike Vi, OAcPec is not immunogenic in laboratory animals,probably because of its comparatively low molecular weight(-400) (17). We conjugated OAcPec to proteins according toa scheme used for Vi and other polysaccharides containing anaminohexuronic acid (10, 26). The physicochemical properties,antigenicity, and immunogenicity in mice of OAcPec and itsconjugates were compared with those of Vi.

MATERIALS AND METHODS

Reagents. Pectin (GENU pectin, type LM-1912CSZ, fromCopenhagen, Denmark) was extracted from citrus and con-tained less than 5% esterification. Pyrogen-free water (PFW)and pyrogen-free saline (PFS) for clinical use were fromBaxter, Deerfield, Wis.; N-succinimidyl 3(2-pyridyldithio) pro-pionate (SPDP) was from Pierce, Rockford, Ill.; formamideand cystamine were from Fluka, Ronkoncoma, N.Y.; pyridine,NaOH, and HCI were from Baker Chemical, Phillipsburg, N.J.;acetic anhydride, dithiothreitol (DTT), EDTA, 1-ethyl-3(3-dimethylaminopropyl) carbodiimide (EDAC), acetyl choline,bovine serum albumin (BSA), dithionitrobenzoic acid (Ellmanreagent), D-galacturonic acid monohydride (GalA), ferric chlo-ride, and tetrabutylammonium hydroxide were from Sigma, St.Louis, Mo.; carbazole was from Aldrich, Milwaukee, Wis.;

5545

on May 6, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

5546 SZU ET AL.

FIG. 1. Repeating units of Vi and pectin. For Vi, C-2 (R) is Nacetylated and C-3 (R') is 0 acetylated. For pectin, C-2 and C-3 are

hydroxylated. For OAcPec, C-2 and C-3 are 0 acetylated.

N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)was from Calbiochem, La Jolla, Calif.; bicinchoninic acid proteinreagent, Sephacryl S-1000, Sepharose CL-4B, Sephadex G-50,Superose 6, and dextrans for calibration of gel filtrationcolumns were from Pharmacia, Piscataway, N.J. All phosphate-buffered saline (PBS) was used at 0.15 M, pH 7.2. Antiserum totetanus toxoid (Ti) was donated by William Habig, Center forBiologics Evaluation and Research, Food and Drug Adminis-tration. Pseudomonas aeruginosa exotoxin A and goat anti-serum to this protein were from List Biological Laboratories,Campbell, Calif. Vi antiserum B-260 was prepared as de-scribed previously (20).

Analytic methods. The molecular sizes of polysaccharidesand conjugates were measured with a Superose 6 high-pressureliquid chromatography (HPLC) column (1 by 28 cm) in 0.02 Msodium phosphate buffer containing 0.1 M Na2SO4 (pH 7.0) at0.5 ml/min. Carboxyls were measured by the carbazole reactionwith pectin as a standard (3, 28). O-Acetyl was measured withacetyl choline as a standard, and the results are expressed asmoles per mole of GalA (12). Sulfhydryl was determined withcysteine as a standard (8). Protein was determined withbicinchoninic acid, using BSA as a standard (21), and nucleicacids were determined by A260 (29). 13C nuclear magneticresonance spectroscopy was performed with a General ElectricGN300 spectrometer at room temperature (25). Chemicalshifts are relative to external Na-3(trimethylsilyl)propionate-2,2,3,3-d.

Polysaccharides. Pectin was dissolved in PFW (10 mg/ml) at60°C for 1 h, cooled to room temperature, and adjusted to pH7.0 with 1 M NaOH. The polysaccharide was precipitated twicewith 75% ethanol and then freeze-dried. Pectin so treatedcontained less than 1% protein and nucleic acid (29). 0

acetylation of pectin was performed as described previously(6). Briefly, pectin (1 g) was suspended in formamide (20mg/ml) at 50°C for 1 h, 20 ml of pyridine was added and mixed,and the mixture was cooled to room temperature. Aceticanhydride (15 ml) was added dropwise with mixing at roomtemperature for 2 h. The reaction mixture was dialyzed at 3 to8°C against multiple changes of deionized water and freeze-dried. The degree of 0 acetylation was -50% of the theoret-ical yield (12). To obtain a higher degree of 0 acetylation, theOAcPec was subjected again to the same procedure. The finalproduct was passed through a column (2.5 by 50 cm) ofSephadex G-50 in PFW at -2 ml/min, and the void volumepeak was passed through a sterile 0.45-[Lm-pore-size mem-brane and freeze-dried. This preparation contained -1.6 molof O-acetyl per mol of GalA, or 80% yield.Vi was purified from Citrobacter freundii WR7011 (gift of

Lewis Baron, Walter Reed Army Institute of Research, Wash-ington, D.C.) (13, 25).

Conversion to the acidic form of Vi and OAcPec was done

by dialysis against H+ Dowex 50X8 (Fluka) in PFW for 2 daysat 4°C in order to avoid hydrolysis of the O-acetyl in contactwith the acidic surface of the resin (4). Polysaccharides withvarious cations were obtained by titration of the acidic formwith corresponding K+, Ca2+, and (C2H5)4N+ hydroxides (4).

Proteins. TT, a gift from the Statens Seruminstitut, Copen-hagen, Denmark, was passed through a column (5 by 95 cm) ofSepharose CL-4B in PFS, and fractions corresponding to amolecular size of 160 kDa were pooled and concentrated (26,27). P. aeruginosa recombinant exoprotein A (rEPA) waspurified from the genetically engineered Escherichia coli strainBL21(^yDE3) as described previously (9) and was provided byJoseph Shiloach, National Institute of Diabetes and Digestiveand Kidney Diseases.

Stability. OAcPec and Vi (1 mg/ml) in 0.02 M PBS (pH 7.0)were incubated at 3 to 8, 25, 37, and 60°C. Aliquots wereremoved at 1, 2, and 12 weeks and analyzed for O-acetylcontent (12) and Mr by gel filtration.

Preparation of conjugates. Conjugates were synthesized asdescribed for Vi (26). Polysaccharide (5 mg/ml) was dissolvedin 0.2 M NaCl. Cystamine (0.1 M) was added, and the pH wasadjusted to 5.0 with 0.1 M HCl. The temperature was 37°C forVi and room temperature for OAcPec. EDAC (0.1 M) wasadded, and the reaction mixture was stirred for 4 h, with thepH maintained between 4.9 and 5.1. The reaction mixture wasdialyzed against PFS with 10 mM phosphate (pH 7) at 3 to 8°Cfor 1 day and against PFW for 3 days with multiple changes andthen freeze-dried. Thiolation was measured on an aliquot ofthe polysaccharides treated with 0.1 M DTT at room temper-ature for 1 h and passage through a Bio-Gel P10 column (2.5by 35 cm; -1.5 ml/min). Void volume fractions were titratedfor sulfhydryl content and degree of derivatization expressedas percent cystamine (8).

Derivatization of proteins with SPDP. SPDP (40 mM), inabsolute ethanol, was added dropwise at room temperaturewith stirring to protein (5 mg/ml) in HE buffer (0.15 MHEPES, 0.001 M EDTA [pH 7.5]) to a final concentration of4 mM. The reaction proceeded for 1 h, and the reactionmixture was dialyzed against HE buffer overnight. The reactionmixture was passed through a column (2.5 by 35 cm) ofBio-Gel P10 in HE buffer at -1.5 ml/min. Void volume peakswere pooled, and an aliquot was treated with 0.075 M DTT atroom temperature for 2 h; its A343 was used to calculate themolar ratio of SPDP to protein (5).

Conjugation. The cystamine-derivatized polysaccharide, 10mg/ml of PBS (pH 7.4), was treated with 0.05 M DTT at roomtemperature for 2 h and passed through a column (2.5 by 35cm) of Sephadex G-50 in PBS. Its sulfhydryl content wasdetermined on an aliquot, and the remainder was mixed withan equal weight of SPDP-derivatized protein and stirred atroom temperature for 4 h and 3 to 8°C overnight. The reactionmixture was passed through a column (2.5 by 95 cm) ofSephacryl S-1000 in PFS at 3 to 8°C (-7 ml/min). For theOAcPec-TT, fractions containing protein and polysaccharidewere pooled into two batches: OAcPec-TT1 for the voidvolume peak and OAcPec-TT2 for the lower-molecular-weightfractions. Vi-rEPA was passed through a column (2.5 by 95cm) of Sephacryl S-1000 in PFS, and the void volume fractionswere pooled.

Serology. Immunodiffusion was performed in 1% agarose inPBS with antiserum B-260. Quantitative precipitation was

performed with 100 ,ul of B-260 with equal volumes of antigen,containing 1 to 100 ,ug/ml, at 37°C for 1 h and at 3 to 8°C for5 days, with occasional mixing. The precipitates were washed incold PFS three times and dissolved in 0.8% sodium dodecylsulfate (SDS), and theirA280 values were recorded (25). Serum

INFEC'T. IMMUN.

on May 6, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

O-ACETYLATED PECTIN CONJUGATE TYPHOID VACCINE 5547

x0)-oc

0)

0

2

25 ppm 20

FIG. 2. 13C nuclear magnetic resonance spectrum of acetyl methylresonances of OAcPec. The peaks at 22.9, 23.1, and 23.6 ppm arerelative to an external standard, Na-3(trimethylsilyl)propionate-2,2,3,3-d.

Vi antibodies were measured by enzyme-linked immunosor-bent assay (ELISA), using pooled hyperimmune mouse sera,quantitated by radioimmunoassay, as the standard (1).

Immunization. Female 16- to 20-g general-purpose micefrom the National Institutes of Health colony were injectedsubcutaneously one, two, or three times at 2-week intervalswith 2.5 ,ug of the polysaccharide alone or as a conjugate. Tenmice from each group were exsanguinated 2 weeks after thefirst injection and 1 week after the second and third injections.Controls included mice injected with saline, Vi, or OAcPec.

Statistical analyses. Logarithms of antibody concentrationswere used for all calculations. Antibody concentrations belowthe sensitivity of the ELISA were assigned one-half of thatvalue. Comparisons of geometric means were performed byunpaired t tests. The Statistical Analysis System was used forall data analyses.

RESULTS

Physicochemical characterization of OAcPec. 0 acetylationranged from 0.1 to 1.6 mol/mol of GalA for pectins. Unlessspecified, the OAcPec described below had 1.6 mol of 0-acetyl/mol of GalA. Assuming that the 0-acetyl groups aredistributed equally between C-2 and C-3, at least 60% of theGalA are di-O acetylated, while 20% are mono-O acetylated.We were unable to quantitate the neutral sugars in our pectinsby colorimetric methods.

13C nuclear magnetic resonance spectra of OAcPec showedmore than two signals observed with acetyl methyl resonances,indicating that mono- and diacetylated species are present;non-O-acetylated residues could, however, be present (Fig. 2).The molecular size of OAcPec, similar to that of the pectin,had a broad distribution, with the major peak -400 kDa (Fig.3). Unlike pectin, OAcPec was soluble in 0.15 M NaCl and didnot form a gel in the presence of Ca2". Molar absorbances inthe carbazole assay were 1.32 X 103 for OAcPec, 1.61 x 103 forpectin, and 1.63 x 103 for GalA. The differences betweenpectin and GalA were <5% and are probably due to neutralsugars or the esterified GalA in pectin. Vi, in contrast, did notreact in the carbazole assay.

Pectin did not react with antiserum B-260 in double immu-nodiffusion. OAcPec, in contrast, formed a line of identity withVi (Fig. 4). Precipitation of OAcPec with Vi antiserum did notchange with different counterions, including Na+, Ca2+, K+, or

OAc Pec TT

12 14 16 18 20 22

Volume (ml)

FIG. 3. HPLC of OAcPec-TT, through Superose 6 in 0.02 Msodium phosphate buffer-0.1 M Na2SO4 (pH 7.0) at a flow rate of 0.5mlmin. The refractive index is shown by the upper line, and the A280is shown by the lower line. The peaks of OAcPec and of Ti prior toconjugation are indicated by arrows.

tetrabutylammonium. At lower degrees of 0 acetylation (0.4 to0.9 mol of 0-acetyl/mol of GalA), 0-acetylated pectin alsoyielded a line of identity with the Vi (not shown). No precip-itation in double immunodiffusion was observed when the 0acetylation of pectin was s0.2 mol/mol of GalA. Quantitativeprecipitation showed that both Vi and OAcPec precipitated 2.6mg of antibody per ml from antiserum B-260 (Fig. 5).

Stability of the O-acetyls of OAcPec and Vi. The thermosta-bilities of the O-acetyls were similar for OAcPec and Vi.Following storage at 3 to 8°C for 12 weeks, there was nochange in the concentration of O-acetyls for Vi and OAcPec: at22°C, 0-acetyls declined to 93% for Vi and to 88% forOAcPec, and at 60°C, only 12 and 10% of the 0-acetylsremained on Vi and OAcPec, respectively (Fig. 6).

Stability of molecular size. The stabilities of glycosidiclinkages of the polysaccharides were studied by gel filtration.There was no change in the molecular size of OAcPec at 3 to8°C for 3 months. After storage of OAcPec at 60°C for 3months, the molecular size decreased from 400 to 30 kDa (notshown). In contrast, the Vi was more stable: little depolymer-ization was observed after incubation at 60°C for 2 weeks, andthe molecular size shifted from 2 x 103 to 500 kDa after 3months.

Characterization of the conjugates. The degrees of thiola-tion were 4% for OAcPec and 1.3% for Vi. The HPLC profileof OAcPec-TT is shown in Fig. 3. The conjugate and a small

2 11Q3

4 1FIG. 4. Double immunodiffusion. Wells: center, Vi antiserum

B-260; 1, Vi, 100 pg/ml; 2, OAcPec K+ form; 3, OAcPec Ca2+ form; 4,OAcPec C2H5N' form.

VOL. 62, 1994

on May 6, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

5548 SZU ET AL.

E0

wu

z

co0Unco

POLYSACCHARIDE (9g)

FIG. 5. Quantitative precipitin analyses of pectin (A), OAcPec (-),and Vi (0). Equal volumes (100 ,ul) of B-260 and polysaccharide were

incubated at 37°C for 1 h and at 3 to 8°C for 5 days. The precipitateswere centrifuged, washed with saline two times, and dissolved in 0.1%SDS, and their A280 values were recorded.

portion of OAcPec were eluted in the void volume. In severalexperiments (not shown), the final yields of the conjugateswere 26% for OAcPec-TT and 10% for Vi-rEPA, based uponrecovery of the saccharide. The polysaccharide/protein ratioswere 0.4 and 0.8 for OAcPec-TT and 0.2 for Vi-rEPA (Table1).

Vi antibodies. As reported previously, Vi elicited serumantibodies in mice after one injection, and reinjection did notelicit a booster response (16, 24-27). Neither pectin norOAcPec elicited Vi antibodies after any injections. After oneinjection, the conjugates elicited similar levels of antibodies.Following the second injection, the conjugates elicited a

booster response (P < 0.001), with the geometric mean

antibody levels highest for Vi-rEPA (17.1) and lower forOAcPec-TT2 (7.65) and OAcPec-TT, (5.47) (Table 2). Thesedifferences, however, were not statistically significant. Thethird injection of all three conjugates did not elicit a booster

*60

o 40

20

011 2 Weeks 12

FIG. 6. Temperature-dependent stability of O-acetyls on Vi (---)

and OAcPec ( ) at 4°C (0), 22°C (l), 37°C (A), and 60°C (0). Thedecrease in extent of 0 acetylation is depicted as the percent remain-ing after incubation at the various intervals and temperatures com-

pared with the starting material.

TABLE 1. Compositions of the OAcPec-TT and Vi-rEPA conjugates

Molecular Cysteamine/Conjugate size polysaccharide SPDP/protein protein ratiode(kDa) of ratio (%, molar ratio prtirao

polysaccharide wt/wt) (Wt/Wt)

OAcPec-Tr1 400 4.0 3.6 0.4OAcPec-Tr2 400 4.0 3.6 0.8Vi-rEPA 2 x 103 1.3 2.1 0.2

response. Lastly, there were not statistically significant differ-ences in the geometric mean Vi antibody levels elicited byOAcPec-TT, and OAcPec-TT2 after any of the three injec-tions.

DISCUSSION

The antigenicity and immunogenicity of Vi depend on itsN-acetyl at C-2 and O-acetyl at C-3 (20, 25). As shown forother polysaccharides, removal of the O-acetyls eliminatedmost of the antigenicity and all of the immunogenicity of Vi (2,25). We have been unable to remove the N-acetyl selectively.Here, we mimic the structure of Vi with OAcPec.There are three differences between the structures of Vi and

OAcPec: (i) the molecular size of Vi (-2 x 103 kDa) is greaterthan that of OAcPec (-400 kDa), (ii) the N-acetyl at C-2 in Viis replaced by an O-acetyl in OAcPec, and (iii) OAcPec has<5% neutral sugars. At 3 to 80C, the stability of OAcPec, asmeasured by its O-acetyl content and molecular size, is similarto that of Vi. At higher temperatures, the molecular size of Viis more stable than that of OAcPec, probably because of thestabilizing effect of a hydrogen bond between the N-acetyl andthe carboxyl of the adjacent residue (25). Since vaccines will bestored at c3 to 80C, the stability characteristics of OAcPec andVi can be considered similar.Although they are antigenically indistinguishable, OAcPec,

unlike Vi, is not immunogenic in mice, probably because of itslower molecular weight (17). In mice, OAcPec and Vi proteinconjugates elicited comparable levels antibodies: reinjection ofboth conjugates induced a statistically significant rise of anti-bodies (booster effect). OAcPec has several advantages over Viin preparing conjugates for prevention of the typhoid fever: (i)

TABLE 2. Vi antibodies in mice immunized with Vi, Vi-rEPA,pectin, OAcPec, and OAcPec-TT conjugatesg

Geometric mean (n = 10) Vi antibodies (1Lg ofImmunogen antibody/ml of serum)b

1st injection 2nd injection 3rd injection

Vi 0.65 0.76 NDVi-rEPA 0.85a 17.lb 12.7cPectin <0.03 <0.03 NDOAcPec <0.03 0.04 NDOAcPec-TT 0.98d 5.47e 6.29fOAcPec-Tr2 0.879 7.65h 5.29'

a Female 14- to 20-g general-purpose mice from the National Institutes ofHealth colony were injected subcutaneously with 2.5 ,ug of polysaccharide threetimes 2 weeks apart. The mice were bled 7 days after injections. Vi antibodylevels were measured by ELISA with a reference calibrated by radioimmunoas-say (25). There were no detectable Vi antibodies in mice that were injected withsaline (data not shown).

b b and c versus a, P = 0.0002; f and e versus d, P = 0.0001; h versus g, P =0.0002; i versus g, P = 0.007; b versus c, f versus e; h versus i, b versus e or h, andc versus f or i, not significant. ND, not done.

INFECT. IMMUN.

on May 6, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

O-ACETYLATED PECTIN CONJUGATE TYPHOID VACCINE 5549

pectin is easy to obtain, and its purification is simpler thanextraction of Vi from S. typhi; (ii) pectin can be measuredduring the synthesis of the conjugate and in the final containerby a colorimetric reaction; (iii) there is no solubility problem,and the yield of OAcPec conjugates is higher than with Vi; and(iv) at 4°C, the storage temperature of vaccines, the stability ofOAcPec is similar to that of Vi. The one deficiency of OAcPecthat we are aware of is its lower Mr. As with other polysaccha-rides, the molecular weight of the Vi alone and as a conjugateis related to its immunogenicity (17, 18, 24). For these reasons,we plan a clinical trial to compare the immunogenicities of ourOAcPec and Vi conjugates.

Polysaccharides from other than the homologous organisms,such as plants, could serve as vaccines. For example, the K92capsular polysaccharide of E. coli [poly(2->8)-a-D-NeupNAc,(2-*9)-a-D-NeupNAc] is an immunogen for group B menin-gococci and E. coli Kl (poly(2--8)-a-D-NeupNAc) (7). Thecross-reaction between the mucin okra with pneumococcustype 6A polysaccharide might be another example (11).

ACKNOWLEDGMENTS

David Towne provided technical assistance in HPLC analysis duringthis study. We thank Uffe Sorensen, Statens Seruminstitut, for tetanustoxoid and GENU Co., Copenhagen, Denmark, for pectin. We aregrateful to Audrey Stone and Rachel Schneerson for review of themanuscript.

REFERENCES1. Acharya, L. L., C. U. Lowe, R. Tapa, V. L. Gurubacharya, M. B.

Shrestha, M. Cadoz, D. Schulz, J. Armand, D. A. Bryla, B.Trolifors, T. Cramton, R. Schneerson, and J. B. Robbins. 1987.Prevention of typhoid fever in Nepal with the Vi capsular polysac-charide of Salmonella typhi: a preliminary report. N. Engl. J. Med.317:1101-1104.

2. Avery, 0. T., and W. F. Goebel. 1933. Chemoimmunologicalstudies on the soluble specific substance of pneumococcus. I. Theisolation and properties of the acetyl polysaccharide of pneumo-coccus type 1. J. Exp. Med. 58:731-755.

3. Bitter, T., and H. M. Muir. 1962. A modified uronic acid carbazolereaction. Anal. Biochem. 4:330-334.

4. Bystricky, S., and S. C. Szu. 1994. O-acetylation affects the bindingproperties of the carboxyl group on Vi bacterial polysaccharide.Biophys. Chem. 51:1-7.

5. Carlsson, J., H. Drevin, and R Axen. 1978. Protein thiolation andreversible protein-protein conjugation. Biochem. J. 173:723-737.

6. Carson, T. F., and W. D. Maclay. 1946. The acetylation ofpolyuronides with formamides as a dispersing agent. J. Am. Chem.Soc. 68:1015-1017.

7. Devi, S. J., J. B. Robbins, and R. Schneerson. 1991. Antibodies topoly[(2--.8)-ot-N-acetylneuraminic acid] and poly[(2-9)-a-N-acetylneuraminic acid] are elicited by immunization of mice withEscherichia coli K92 conjugates: potential vaccines for groups Band C meningococci and E. coli. Proc. Natl. Acad. Sci. USA88:7175-7179.

8. Eliman, G. L. 1959. Tissue sulfhydryl groups. Arch. Biochem.Biophy. 82:70-77.

9. Fass, R, M. van de Walle, A. Shiloach, A. Joslyn, J. Kaufman, andJ. Shiloach. 1991. Use of high density cultures of Escherichia colifor high level production of recombinant Pseudomonas aeruginosaexotoxin A. Appl. Microbiol. Biotechnol. 36:65-69.

10. Fattom, A., C. Lue, S. C. Szu, J. Mestecky, G. Schiffman, D. Bryla,W. F. Vann, D. Watson, L. M. Kimzey, J. B. Robbins, and RSchneerson. 1990. Serum antibody response in adult volunteers

elicited by injection of the Streptococcus pneumoniae type 12Fpolysaccharide alone or conjugated to diphtheria toxoid. Infect.Immun. 58:2309-2312.

11. Heidelberger, M., and P. A. Rebers. 1960. Immunochemistry of thepneumococcal types II, V, and VI. I. The relation of type VI totype II and other correlations between chemical constitution andprecipitation in antisera to type VI. J. Bacteriol. 80:145-153.

12. Hestrin, S. 1949. The reaction of acetylcholine and other carbox-ylic acid derivatives with hydroxylamine, and its analytical appli-cation. J. Biol. Chem. 180:249-261.

13. Heyns, K., and G. Kiessling. 1967. Strukturaufkldarung des Vi-antigens aus Citrobacterfreundii (E. coli) 5396/38. Carbohydr. Res.3:340-353.

14. Klugman, K. P., H. J. Koornhof, I. T. Gilbertson, J. B. Robbins,RSchneerson, D. Schulz, M. Cadoz, J. Armand, and VaccineAdvisory Committee. 1987. Protective activity of Vi capsularpolysaccharide vaccine against typhoid fever. Lancet ii:1165-1169.

15. Landy, M. 1954. Studies in Vi antigen. VI. Immunization of humanbeings with purified Vi antigen. Am. J. Hyg. 60:52-62.

16. Landy, M. 1957. Studies on Vi antigen. VII. Characteristics of theimmune response in the mouse. Am. J. Hyg. 65:81-93.

17. Martin, D. G., F. G. Jarvis, and K. C. Milner. 1967. Physicochem-ical and biological properties of sonically treated Vi antigen. J.Bacteriol. 94:1411-1416.

18. Peeters, C. A., A.-M. Tenbergen-Meekes, D. E. Evenberg, J. T.Poolman, B. J. M. Zegers, and G. T. Rijkers. 1991. A comparativestudy of the immunogenicity of pneumococcal type 4 polysaccha-ride and oligosaccharide tetanus toxoid conjugates in adult mice. J.Immunol. 146:4308-4314.

19. Pilnik, W., and A. G. J. Voragen. 1970. Pectic substances and otheruronides, p. 53-73. In The biochemistry of fruits and theirproducts, vol. 1. Academic Press, New York.

20. Robbins, J. D., and J. B. Robbins. 1984. Re-examination ofprotective role of the Vi capsular polysaccharide (Vi antigen) ofSalmonella typhi. J. Infect. Dis. 47:436-499.

21. Smith, P. K., R L. Krohn, G. T. Hermanson, A. K. Mallia, F. H.Gartner, M. D. Provenzano, E. K. Fujimoto, N. M. Goeke, B. J.Olson, and D. C. Klenk. 1985. Measurement of protein usingbicinchoninic acid. Anal. Biochem. 150:76-85.

22. Stone, A. L., and S. C. Szu. 1988. Application of optical propertiesof Vi capsular polysaccharide for quantitation of the Vi antigen invaccines for typhoid fever. J. Clin. Microbiol. 26:719-725.

23. Szewczyk, B., and A. Taylor. 1980. Immunochemical properties ofVi antigen from Salmonella typhi Ty2: presence of two antigenicdeterminants. Infect. Immun. 29:539-544.

24. Szu, S. C., X. Li, R Schneerson, J. H. Vickers, D. Bryla, and J. B.Robbins. 1989. Comparative immunogenicities of Vi polysaccha-ride-protein conjugates composed of cholera toxin or its B subunitas a carrier bound to high- or lower-molecular-weight Vi. Infect.Immun. 57:3823-3827.

25. Szu, S. C., X. Li, A. L. Stone, and J. B. Robbins. 1991. Relationbetween structure and immunologic properties of the Vi capsularpolysaccharide. Infect. Immun. 59:4555-4561.

26. Szu, S. C., A. L Stone, J. D. Robbins, R Schneerson, and J. B.Robbins. 1987. Vi capsular polysaccharide-protein conjugates forprevention of typhoid fever. J. Exp. Med. 166:1510-1524.

27. Szu, S. C., D. N. Taylor, A. C. Trofa, J. D. Clements, J. Shiloach,J. C. Sadoff, D. A. Bryla, and J. B. Robbins. 1994. Laboratory andpreliminary clinical characterization of Vi capsular polysaccha-ride-protein conjugate vaccines. Infect. Immun. 62:4440- 4444.

28. Taylor, K. A., and J. G. Buchanan-Smith. 1992. A colorimetricmethod for the quantitation of uronic acids and a specific assay forgalacturonic acid. Anal. Biochem. 201:190-196.

29. World Health Organization Expert Committee on BiologicalStandardization. 1977. Technical report series no. 610. WorldHealth Organization, Geneva.

VOL. 62, 1994

on May 6, 2018 by guest

http://iai.asm.org/

Dow

nloaded from