THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 267, 25. Issue … · biologic properties associated with the B. fragilis cap- sular polysaccharide. Although greatly outnumbered by other

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 267, No. 25. Issue of September 5, pp. 18230-18235, 1992 Printed in U.S.A.

The Capsular Polysaccharide of Bacteroides frugilis Comprises Two Ionically Linked Polysaccharides*

(Received for publication, February 27, 1992)

Arthur 0. TzianabosSG, Annalisa Pantostitll, Herbert Baumannll , Jean-Robert BrissonII , Harold J. Jenningsll, and Dennis L. Kasper$** From the $Charming Laboratory, Brigham and Women’s Hospital and the **Division of Infectious Diseases, Beth Israel Hospital, Harvard Medical School, Boston, Massachusetts 02115 and the ((Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario KIA OR6, Canada

Recently, we have shown that the capsular polysaccharide of Bacteroides fragilis NCTC 9343 is composed of an aggregate of two discrete large molecular weight polysaccharides (designated polysaccharides A and B). Following disaggregation of this capsular complex by very mild acid treatment, high resolution NMR spectroscopy demonstrated that polysaccharides A and B consist of highly charged repeating unit structures with unusual substituent groups (Baumann, H., Tzian- abos, A. O., Brisson, J.-R., Kasper, D. L., and Jennings, H. J. (1992) Biochemistry 31,4081-4089). Presently, we report that the capsular polysaccharide of B. fragilis represents a complex structure that is formed as a result of ionic interactions between polysaccharides A and B. Electron microscopy of immunogold-labeled organisms (with monoclonal antibodies specific for polysaccharides A and B) demonstrated that the two polysaccharides are co-expressed on the cell surface of B. fragilis. We have shown that the purified capsule complex is made up exclusively of polysaccharide A and polysaccharide B (no other macromolecular structure was detected) in a 1:3.3 ratio and that disaggregation of this complex into the native forms of the constituent polysaccharides could be accomplished by preparative isoelectric focusing. Structural analyses of the native polysaccharides A and B showed that they possessed the same repeating unit structures as the respective acid-derived polysaccharides. The ionic nature of the linkage between polysaccharides A and B was demonstrated by reassociation of the native polysaccharides to form an aggregated polymer comparable to the original complex. The distinctive composition of this macromolecule may provide a rationale for the unusual biologic properties associated with the B. fragilis capsular polysaccharide.

Although greatly outnumbered by other Bacteroides species in the normal intestinal microflora, Bacteroides fragilis is the

*This work was supported by Grants lROl AI 22807 and 2T32AI07061-llAl from the National Instititute of Allergy and In- fectious Diseases, National Institutes of Health. This paper is issued as National Research Council of Canada No. 34261. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

J To whom correspondence and reprint requests should be ad- dressed Channing Laboratory, 180 Longwood Ave., Bos- ton, MA 02115. Tel.: 617-432-1610. Fax: 617-731-1541.

Present address: Istituto Superiore Di Sanita, Via Regina Elena, 299, 00161 Rome, Italy.

anaerobe most frequently isolated from clinical infections such as intra-abdominal abscesses and anaerobic bacteremia (1, 2). The capsular polysaccharide (CP)’ of B. fragilis is among the primary virulence determinants of this organism and has been shown to mediate a number of interesting biologic processes in the infected host. Among the most im- portant virulence properties of the CP is its ability to promote the formation of intra-abdominal abscesses in experimental animal models of disease in the absence of viable organisms (3). These abscesses, although histologically identical with those formed following infection with viable bacteria, are bacteriologically sterile. Immunization with the CP affords protection against the formation of abscesses in response to challenge with B. fragilis. This immunity is T cell-dependent and involves a non-major histocompatibility complex-re- stricted T cell circuit (4-6).

Previous investigations of the CP have indicated that it is quite complex (7, 8). Immunochemical analysis revealed that the CP is isolated as an aggregate that is composed of two distinct high molecular weight polysaccharides (designated polysaccharide A and polysaccharide B), with very different physicochemical properties. Immunoelectrophoresis (IEP) of the CP under neutral and basic pH conditions demonstrated a variable multiprecipitin profile (8).

Separation of the CP into the component polysaccharides proved technically difficult but was finally accomplished following mild acid treatment and isolation of the component polysaccharides by anion exchange chromatography. High resolution nuclear magnetic resonance (NMR) spectroscopy was performed on the purified polysaccharides A and B after acid treatment, and the fine structures were elucidated. Pol- ysaccharide A (Fig. lA), which exhibited zwitterionic behavior conferred by a positive and negative charge, consisted of the following tetrasaccharide repeating unit: -3) CY-D-AAT Galp- (1+4)-[P-~-Galf-(1-+3)] a-~-GalpNAc-(1+3)-[4, 6-pyru- vate]-P-~-Galp-(l-+.

Polysaccharide B (Fig. lB) , which possessed one positive and two negative charges, was composed of the following repeating unit: +3)~-~-QuipNAc-(1+4)-[~~-~-Fucp-(l+2)-P- ~-GalpA(1+3)[ [2-AEP~]] -P-~-g lcpNAc-(1~3)] -a-~-Galp- (1+4)-a-~-QuipNAc-(l+ (9).

We sought to determine how polysaccharides A and B are bound together to form the CP. Preparative scale isoelectric focusing facilitated the disaggregation of the CP into the component native polysaccharides, and high resolution NMR

The abbreviations used are: CP, capsular polysaccharide; IEP, immunoelectrophoresis; AAT Galp, 2-acetamido-4-amino-2,4,6-tri- deoxygalactose; QuiNacp, quinovosamine; 2-AEP, Z-aminoethyl- phoshonate: mAb, monoclonal antibody; PBS, phosphate-buffered saline.

18230

Ionic Interactions Complex Two Surface Polysaccharides of B. fragilis 18231

A) I

H,C

con I

Hsc

FIG. 1. Fine structures of B. fragilis NCTC 9343 polysaccharides A and B. A, polysaccharide A consists of a tetrasaccharide with free amino and carboxyl groups that confer zwitterionic behavior to this polymer. B, polysaccharide B consists of a hexasaccharide repeating unit with a 2-aminoethylphosphonate substituent and a galacturonic acid that confer a net negative charge to this polymer.

showed that the acid and native forms of each polysaccharide had similar repeating unit structures. Immunoelectron microscopy showed that both of the polysaccharides were surface-expressed antigens existing on the surface of the same bacterial cell. These studies demonstrate that the CP of B. fragilis NCTC 9343 is composed of a novel capsular structure in which two discrete types of highly charged polysaccharides are aggregated by ionic interactions between the two polymers.

MATERIALS AND METHODS

Bacterial Strain and Isolation of Capsular Polysaccharides-B. fragilis NCTC 9343 was originally obtained from the National Collection of Type Cultures (London, UK), stored at -80 "C in peptone-yeast broth until use, and grown anaerobically as previously described (8). The capsular complex from B. fragilis NCTC 9343 was isolated by hot phenol/water extraction, and subsequent purification of the component acid-treated polysaccharides A and B was performed as previously described (8,9).

Proportion of Polysaccharides A and B in CP-To determine the proportion of polysaccharides A and B in CP following the standard mild acid treatment (5% acetic acid at 100 "C for 1 h) and separation by anion exchange chromatography (8), the mass of each recovered polysaccharide was recorded and compared with the mass of the starting material. The percentage of polysaccharides A and B that compose the CP was calculated from the relative mass of each polysaccharide.

Separation of Constituent Native Polysaccharides of the CP by Isoelectric Focusing-Preparative isoelectric focusing was employed to separate and isolate the native form of each polysaccharide. Twenty milligrams of the CP was solubilized in 55 ml of a 1% (v/v) ampholyte solution (pH range 3-10, Bio-Rad) and loaded into the focusing chamber of a Rotofor preparative isoelectric focusing cell (Bio-Rad). A constant power of 12 watts was applied to the cell, and circulating water was used to cool the system. Electrophoresis of the sample was performed until the voltage output held at one level for 1 h. At this time, the current was terminated and 20 fractions corresponding to

20 compartments of the chamber were immediately collected by aspiration. The pH of each fraction was measured to ascertain that a linear pH gradient had been established. Fractions were assayed for polysaccharide content by double diffusion in agarose against antiserum prepared to whole bacteria. Fractions giving precipitins were dialyzed extensively against dHnO, lyophilized, and analyzed by IEP with Tris buffer at pH 7.3, as described below.

Reassociation of Native Polysaccharides-The native forms of polysaccharides A and B were mixed in a predetermined ratio of 1:3.3 in which 6 mg of native A and 20 mg of native B were dissolved together in 5 ml of 50 mM Tris-HC1, pH 7.3, and sonication in a low frequency water bath for 1 h at a temperaature that did not exceed 60 "C. The mixture chromatographed on a column containing DEAE-Sephacel equilibrated in Tris buffer, pH 7.3, as described (8) and eluted with a two-step linear salt gradient. Fractions were tested by double diffusion in agarose against antiserum prepared to whole bacteria and collected into four distinct pools (Pools I-IV).

Immunologic Assays and Preparation of Polyclonal and Monoclonal Antisera-Polyclonal antiserum specific for B. fragilis NCTC 9343 and the purified CP were prepared in New Zealand White rabbits (8). mAbs CE3 and F10, specific for polysaccharides A and B, respectively, have been reported previously (8). Double diffusion in agarose was performed by the method of Ouchterlony (10) with 1% agarose in 0.15 M saline. IEP was performed in 1% agarose with 50 mM Tris- HCl buffer, pH 7.3. The proportion of polysaccharide A and polysaccharide B in CP was also determined by competitive enzyme-linked immunosorbent assay using a previously described system (11, 12). Purified capuslar polysaccharides (polysaccharide A, polysaccharide B, or the CP) were used as inhibitors for monclonal antibody (mAb) CE3 (specific for polysaccharide A) or F10 (specific for polysaccharide B) in this assay.

Nuclear Magnetic Resonance Studies of Native and Acid-treated Polysaccharides-"H NMR spectra were recorded on a Bruker AMX5OO or AMXGOO spectrometer with the use of a broad-band probe, with the 'H coil nearest to the sample. Acetone was used as an internal chemical shift reference for 'H NMR ( 6 2.225 ppm). Spectra were recorded at 338 K and 358 K in 5-mm tubes at concen- trations of 1-8 mg in 0.5 ml of D20 at neutral pH. All experiments were carried out without sample spinning and with the standard software provided by Bruker (9).

Immunoelectron Microscopy-Bacteria were grown as described above, washed extensively with 0.15 M phosphate-buffered saline, and treated sequentially with mAb CE3, protein A-gold probe (10 nm), mAb F10, and protein A-gold probe (20 nm) to effect the specific double labeling of polysaccharides A and B with 10 nm and 20 nm gold probes, respectively. Control grids were incubated with normal mouse serum and each of the respective protein A-gold conjugates. Following the labeling procedure, organisms were examined with a Hitachi H600 electron microscope operating at 75 kV in the transmission mode.

RESULTS

Heterogeneity of the CP-The CP extracted from B. fragilis strain 9343 has been shown to consist of two distinct, high molecular weight aggregated polysaccharides (8). IEP of the purified CP performed for 3 h at pH 7.3 demonstrated a complex precipitin profile in which polysaccharide A (precipitin a) formed a precipitin arc that remained neutral and polysaccharide B (precipitin b) formed a precipitin arc that migrated toward the anode. In contrast, electrophoresis of CP at pH 8.6 showed that polysaccharide A migrated toward the anode, an indication that this polysaccharide behaved as a zwitterion. Polysaccharide B retained its negative charge at this pH. A third precipitin (precipitin c), which extended between polysaccharides A and B, was also visualized at pH 7.3 and 8.6 (Fig. 2).

Further characterization of precipitin c was performed by a two-dimensional IEP technique. IEP of the CP was performed first at pH 7.3 without the addition of antiserum to the trough and then at pH 8.6 in a second perpendicular dimension. The second electrophoretic stage was performed in 1% agarose mixed with 0.5% anticapsular serum or ascites fluid containing mAb CE3 or F10. Following washing, precip-

18232 Ionic Interactions Complex Two Surface Polysaccharides of B. fragilis

FIG. 2. IEP of B. frugilis CP. A, IEP profile following electrophoresis at pH 8.6 shows three distinct precipitins (a, b, c ) migrating toward the anode (+). B, IEP profile following electrophoresis at pH 7.3 showing three distinct precipitins. Note neutral charge of precipitin a and the precipitin corresponding to the lipopolysaccharide (Ips) of this organism. Troughs contain polyclonal rabbit antiserum raised against whole bacteria.

FIG. 3. Two-dimensional IEP of B. frugzXs CP. CP samples are electrophoresed at pH 7.3 in the first dimension, rotated go", and electrophoresed in the second dimension at pH 8.6. A, IEP profiles following precipitation with polyclonal rabbit antiserum raised against purified CP. Note peaks corresponding to precipitins a, b, and c. B, IEP profiles following precipitation with mAb CE3. Only precipitins a and c are visible. C, IEP profiles following precipitation with mAb F10. Only precipitins b and c are visible. o indicates origin.

itins were stained with Coomassie Brilliant Blue. In the second dimension IEP gel, precipitins formed with

anticapsular serum and showed polysaccharide A and B ad- joined by precipitin c (Fig. 3a). Both polysaccharides A and B migrated toward the anode in the second dimension. The zwitterionic character of polysaccharide A was demonstrated at these pH values. It was noted that mAb CE3 formed a precipitin that corresponded to polysaccharide A but included precipitin c (Fig. 3b). mAb F10 formed a precipitin corresponding to polysaccharide B that also included precipitin c (Fig. 3c). The recognition of precipitin c by each of the mAbs indicated that polysaccharide A and B must contribute to the formation of this precipitin and that perhaps this precipitin does not represent a distinct antigen.

Separation and Isolation of Native Polysaccharides A and B by Isoelectric Focusing-Preparative isoelectric focusing was employed to separate and isolate the native forms of polysaccharide A and polysaccharide B from 20 mg of CP. A linear pH gradient from 2.7 (anode) to 11.1 (cathode) was established from fraction 1 (low pH) to fraction 20 (high pH) in the focusing cell. The first two fractions (low pH range) gave two precipitin lines in double diffusion against whole organism antiserum, while fractions 6, 7, and 8 (neutral pH range) gave only a single precipitin. IEP analysis at pH 7.3 of the anodal migrating material collected in fractions 1 and 2 gave profiles corresponding to precipitins b and c, but lacked precipitin a. Polysaccharides remaining in the neutral pH range collected in fractions 6, 7, and 8 corresponded to precipitin a (Fig. 4).

Comparison of Native and Acid Forms of Polysaccharides A and B-The nature of the native forms of each polysaccharide obtained from the isoelectric focusing chamber was confirmed by chromatography on a DEAE-Sephacel column equilibrated in Tris buffer at pH 7.3 (under conditions used for the separation of the acid-treated polysaccharides). Native polysaccharide A flowed through the anion exchange column at this pH, while native polysaccharide B was retained and could

FIG. 4. Isoelectric focusing of CP. IEP of Rotofor fractions a t pH 7.3 following isoelectric focusing of the capsule complex. A , CP before focusing. B, pool of fractions 6, 7, and 8 from focusing apparatus. C, pool of fractions 1 and 2 from focusing apparatus. Troughs contain polyclonal rabbit antiserum raised against whole bacteria.

be eluted with high molarity NaCl buffer solution. Therefore, native polysaccharides A and B possessed net charges similar to the respective acid-treated polysaccharides.

The native polysaccharides were analyzed by IEP at pH 7.3 separately against mAb CE3 and mAb F10. Native and acid- treated polysaccharide A formed similar precipitin arcs with mAb CE3. IEP of native polysaccharide B against mAb F10 at this pH gave a profile that included precipitins b and c. However, acid-treated polysaccharide B formed only precipitin b.

NMR studies performed on the constituent polysaccharides of the CP demonstrated that the native form of polysaccharide A had the same structure as reported for the acid-treated polysaccharide A (9) as indicated by their identical 'H NMR spectra (Fig. 5A). These studies also showed that the repeating unit structures of the acid-treated and native forms of polysaccharide B were the same, but that hydrolysis conditions used to obtain the acid-derived polymer caused minimal deg- radation of the polysaccharide (Fig. 5B). Previously, we have shown that the acid-treated and native forms of the two polysaccharides are immunochemically identical (8). NMR analysis has also shown that the CP was composed exclusively of repeating units of polysaccharides A and B and indicated that these polymers were not linked by other types of macromolecules (9).

Determination of the Relative Proportion of Polysaccharides A and B in the CP-To establish the ratio of polysaccharides A and B in the CP, we performed studies in which a known quantity of CP was treated with mild acid to separate and isolate these constituent polysaccharides (5% acetic acid, 100 "C, 1 h ) (8). The dry weight of each isolated polysaccharide was determined and the percentage recovery of the constituent acid-treated polysaccharides from the CP was calculated from three separate experiments. The average dry weight of polysaccharides A and B recovered was 11 and 32 mg, respectively, which corresponded to an average ratio of polysaccharide A to polysaccharide B of approximately 1:3. Im- munochemical determination of the relative proportion of polysaccharide A to B in the CP using enzyme-linked immunosorbent assay inhibition studies showed that this ratio was 1:3.5 (data not shown).

Reassociation of Polysaccharides A and B-After isolating the native component polysaccharides, we attempted to demonstrate that these polymers could be reassociated to form the original complex in order to confirm that electrostatic interactions hold the two highly charged polymers together in an aggregate. The native forms of polysaccharides A and B, which had been shown to elute differently by ion exchange chromatography at pH 7.3 (see above), were mixed in a predetermined ratio of 1:3.3 (6 mg of polysaccharide A to 20 mg of polysaccharide B) in 5 ml of 50 mM Tris-HC1 buffer, pH 7.3, sonicated in a water bath for 1 h, and loaded into a column containing DEAE-Sephacel equilibrated in the same buffer. The column was washed with two bed volumes of the starting buffer, and the eluate was collected and pooled (Pool


ppm 5.0 4.0 3.0

FIG. 5. 'H NMR spectra of native and acid-treated polysaccharides. A, analysis of native and acid-treated polysaccharide A show identical spectra. Signal at 4.35 ppm is due to D2O in sample. B, analysis of acid-treated polysaccharide B ( top) and native polysaccharide B (bottom) shows that these polymers have the same repeating unit structure. Signals at 4.20 and 3.75 ppm are due to D,O and Tris in acid-treated polysaccharide A sample. Signal at 4.18 ppm is due to D20 in native polysaccharide A sample.

I

2.0 1 .o

1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ppm 5.0 4.0 3.0 2.0 1.0

I). The column was then eluted with a linear step gradient of 0-0.5 M NaCl and 0.5-2 M NaCl, and all fractions were tested for the presence of CP by double diffusion in agarose. Frac- tions that eluted with low NaCl concentration and formed single precipitins were pooled (Pool 11) as were fractions that eluted with higher NaCl concentration and formed double precipitins (Pool 111), and those that eluted with the highest salt concentration and formed single precipitins (Pool IV). Each pool was dialyzed extensively against dH20, lyophilized, and analyzed by IEP at pH 7.3 (Fig. 6).

After chromatography of the the mixed polysaccharides A and B, we noted that no polysaccharide was eluted from the column with the starting buffer (Pool I). It is surprising that IEP of Pool I1 demonstrated only polysaccharide A, despite the fact that this polysaccharide had previously been shown not to be retained by this column when chromatographed under these conditions. These results suggested that some electrostatic interactions within the mixture retarded the elution of polysaccharide A from the anion exchange column. IEP of Pool I11 showed the complex precipitin profile repre-

18234 Ionic Interactions Complex Two Surface Polysaccharides of B. fragilis

OM I 0.0.5M I 05.2M NaCl I I V Pool IV "

1 P

Pool 1

0- 0 2 0 0 4 0 0 600 800 Na"ve A

Fracllon (ml)

FIG. 6. Anion exchange chromatography of reassociated native polysaccharides A and B. IEP profiles of mixed sonicated native polysaccharides A and B following elution with a two-step salt gradient. Fractions that gave positive precipitins in double diffusion in agarose are indicated by solid lines (--). Fractions were pooled and analyzed by IEP at pH 7.3 as 1 mg/ml samples except for Pool IV, which was analyzed as a 3 mg/ml sample. Troughs contain polyclonal rabbit antiserum raised against whole bacteria. Note that native A is not retained by this column when chromatographed alone in the starting buffer.

- FIG. 7. Co-expression of polysaccharides A and B by B.

frugilie. Electron micrograph shows immunogold-labeled bacterial cells. mAb CE3 (specific for polysaccharide A) and mAb F10 (specific for polysaccharide B) were labeled with 10 and 20 nm gold particles, respectively. Note dual labeling of the surface of individual bacterial cells with the small and large gold probes. Bar = 1 pm. Inset: note dual surface labeling of capsular polysaccharide. Bar = 0.25 pm.

sentative of the complete CP (Fig. 2). When IEP was performed on a 1 mg/ml solution of Pool IV, only precipitins b and c were visualized. However, when this pool was analyzed at a higher concentration (3 mg/ml), all three precipitins (a, b, and c) appeared.

When the native polysaccharides were mixed but not sonicated and then analyzed by IEP at pH 7.3, the polysaccharides did not form the complexed precipitin profile characteristic of the CP but migrated away from one another unconnected by precipitin c.

Surface Expression of Polysaccharides A and B-Immunoe- lectron microscopy of B. fragilis with the mAbs specific for polysaccharide A or B demonstrated that these polysaccharides were co-expressed on the bacterial cell surface. The double labeling procedure affected the specific detection of polysaccharides A and B on the surface of B. fragilis cells by different sized, electron-dense gold probes. Transmission electron microscopy showed that both gold labels were present on the surface of individual B. fragilis cells in association with a diffuse layer surrounding the organisms (Fig. 7).

DISCUSSION

The CP of B. fragilis exhibits unusual biologic properties that may be attributable to its unique structural composition. High resolution NMR analysis of the CP has shown that this molecule is exclusively composed of repeating units of poly-

saccharides A and B and that other macromolecules do not contribute to the composition of the CP (9). In light of the recent structural characterization of polysaccharides A and B, we hypothesized that ionic interactions rather than cova- lent or hydrophobic bonds between these highly charged repeating unit structures were responsible for their close association. Taking advantage of the differing electrophoretic characteristics of these polysaccharides, we used an isoelectric focusing technique to separate them on the basis of their different net charges at varying pH values.

The use of preparative isoelectric focusing for the disaggregation of the CP allowed for application of high voltage current through a sample solubilized in a 1% ampholyte solution. On the basis of their structural compositions, the polysaccharides should migrate away from each other, with polysaccharide A focusing in the neutral pH region of the cell and polysaccharide B migrating toward the low pH, anodal region of the cell. IEP analysis of all the fractions collected from the cell following isoelectric focusing showed that polysaccharide A had been separated from polysaccharide B and that each polysaccharide had been isolated in pure form by this technique. NMR studies demonstrated that the native polysaccharides possessed the same repeating unit structure as their respective acid-derived counterparts. However, it is unclear why native polysaccharide B gave a precipitin profile that included precipitin c, while the acid-treated polysaccharide B did not. This result may reflect the loss of a negatively charged group on some molecules of polysaccharide B or a change in molecular size following mild acid hydrolysis.

To confirm the ionic nature of the bond that complexes the polysaccharides, we attempted to reassociate the native forms of the two polysaccharides. The electrophoretically purified native polysaccharides were mixed in a ratio of 1:3.3 (predetermined as the ratio of polysaccharide A to polysaccharide B in the CP), sonicated, and loaded into an anion exchange column. On the basis of the following evidence, it was clear that the isolated native polysaccharides were able to reassociate to form the complex precipitin profile characteristic of the CP: 1) polysaccharide A did not wash through the anion exchange column with the starting buffer after anion exchange chromatography of the mixed polysaccharides at pH 7.3 and 2) IEP of the material eluted from the column with the higher concentration NaCl buffer showed a profile con- sisting of the three linked precipitin lines characteristic of the IEP profile of CP. Considered together with data from NMR studies that demonstrate that CP comprises exclusively polysaccharides A and B, these results show that ionic interactions and not other chemical bonds or constituents between the two highly charged component polysaccharides hold these polymers together in aggregate form.

It is interesting that polysaccharide A was eluted in pure form in Pool I1 from the DEAE-Sephacel column with low NaCl concentration buffer but was prevented from washing directly through the column with the starting buffer (Pool I). This finding indicates that this polymer did not completely reassociate with polysaccharide B but was held in loose association until it was eluted with low molarity salt solution from the column. In addition, when the native polysaccharides were mixed but not sonicated and then analyzed by IEP, the polysaccharides did not form the complexed precipitin profile characteristic of the CP but migrated away from one another, unconnected by precipitin c. It is possible that sonication may free the charged groups of these polymers to make them available for interchain interaction.

Antigen recovery studies and immunologic assays indicated that polysaccharides A and B contribute to the formation of


precipitin c and that they exist in a 1:3.3 ratio to form the CP of B. fragilis. Immunoelectron microscopy using different sized gold labels that corresponded to each of the mAbs specific for polysaccharides A and B showed that these polymers were co-expressed on the surface of B. fragilis cells. This result demonstrated that isolation of the CP complex from a population of organisms was not a result of co-extraction of polysaccharides A and B from different organisms expressing one or the other polysaccharide type. The finding of the two types of gold labels randomly aligned on the cell surface strongly suggests that these structures are co-expressed by B. fragilis.

The CP of B. fragilis is composed of two types of discrete, complex polysaccharides with different physicochemical properties. Few bacterial organisms produce such a complicated array of surface polysaccharides. Rhodopseudomonas capsu- lata, Rhizobium trifoli, and Rhizobium meliloti all produce two or more surface polysaccharides in addition to their lipopoly- saccharides (13-16).

It is unknown whether presentation of the CP in this complex form confers enhanced virulence or immunity in the animal models when compared to the biologic activities of the component polysaccharides. It is probable that the highly charged nature of these polysaccharides directly affects the conformation of each component as well as the complexed CP. Previous studies have clearly demonstrated that the conformation of polysaccharides is a critical parameter that modulates the antigenicity of some bacterial polysaccharides (17-20). It is possible that the charged B. fragilis polymers may interact to form the proper conformational orientation to mediate abscess formation or T cell immunity. The CP of B. fragilis NCTC 9343 represents a novel encapsulation motif that has not been described previously for bacteria pathogenic

for humans. Further studies are needed to establish the par- ticular structural aspects of the CP and constituent polysaccharides that modulate the unusual biologic attributes of this macromolecule.

Acknowledgments-We thank Drs. L. Madoff and L. Paoletti for helpful discussions and Roger Smith and Barbara Reinap for tech- nical assistance.

REFERENCES 1. Polk, B. F., and Kasper, D. L. (1977) Ann. Intern. Med. 86,569-571 2. Moore, W. E. C. (1977) in Proceedings of the International Metronidozole

Conference (Finegold, S. M., ed) pp. 106-112, Excerpta Medica, Prince- ton. N.J

3. Onderdonk, A. B., Kasper, D. L., Cisneros, R. L., and Bartlett, J. G. (1977)

4. Onderdonk. A. B.. Markham. R. B.. Zaleznik. D. F.. Cisneros. R. L.. and

"", - -

J. Infect. Dis. 136,82-89

5.

6.

7.

8.

9.

10. 11.

12.

13.

14.

15. Glazebrook, J., and Walker, G. C. (1989) Cell 56,661-672 16. Gray, J. X., and Rolfe, B. G. (1990) Mol. Microbiol. 4,1425-1431 17. Wessels, M. R., Munoz, A., and Kasper, D. L. (1987) Proc. Natl. Acad. Sci.

18. 19.

20.

-"& -""

U. S. A. 84.9170-9174

Kasper, D. L. (1982) J. Clin. Inuest. 69,9-I6 '

Shapiro, M. E., Onderdonk, A. B., Kasper, D. L., and Finberg, R. W. (1982) J. Exp. Med. 155,1188-1197

Zaleznik, D. F., Finberg, R. W., Shapiro, M. E., Onderdonk, A. B., and Kasper, D. L. (1986) J. Infect. Dis. 154 , 40-46

Kasper, D. L., Weintraub, A., Lindberg, A. A., and Lonngren, J. (1983) J. Bacteriol. 153 , 991-997

Pantosti, A,, Tzianabos, A. O., Onderdonk, A. B., and Kasper, D. L. (1991) Infect. Immun. 59,2075-2082

Baumann, H., Tzianabos, A. O., Brisson, J.-R., Kasper, D. L., and Jennings, H. J. (1992) Biochemistry 31,4081-4089

Ouchterlony, 0. (1958) Prog. Allergy. 5 , 1-78 Rubens, C. E., Wessels, M. R., Heggen, L. M., and Kasper, D. L. (1987)

Paoletti L. C., Kas er D L Michon, F., DiFabio J. Holme, K., Jennings,

Abe, M., Sherwood, J. E., Hollingsworth, R. I., and Dazzo, F. B. (1984) J.

Omar, A. S., Weckesser, J., and Mayer, H. (1983) Arch. Microbiol. 136 ,

Proc. NatL Acad. Sci. U. S. A. 8 4 , 7208-7212

H. J.,'and WesseYB, M.'R.'i1990) J. Biol. Chem.'265, 18278-18283

Baeteriol. 160 , 517-520

941 -94fi

Wessels, M. R . , and Kas er, D L. (1989) J. Ex Med. 169, 2121-2131 Jennings, H. J., Lugow&, C., 'and Kasper, D. E. (1981) Biochemistry 2 0 ,

Jennings, H. J., Roy, R., and Michon, F. (1985) J. Immunol. 134 , 2651- 4511-4518

2657

Documents

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 267, 25. Issue … · biologic properties associated with the B. fragilis cap- sular polysaccharide. Although greatly outnumbered by other