Zbl. Bakt. 277, 340-344 (1992) © Gustav Fischer Verlag, StuttgartlNew York
Sialic Acid as Receptor of Bacteroides fragilis Lectin-like Adhesin REGINA MARIA C. P. DOMINGUES, SiLVIA MARIA B. CAVALCANTI, ARNALDO F. B. ANDRADE, and MARIA CANDIDA S. FERREIRA Instituto de Microbiologia, Centro de Ciencias da Saude, Universidade Federal do Rio de Janeiro, Ilha do Fundao, 21944 Rio de Janeiro, RJ, Brasil
Received December 6, 1991 . Revision received February 25, 1992 . Accepted March 12, 1992
Summary It was observed that sialic acid and macromolecules rich in this sugar were able to inhibit the hemagglutination activity (HA) of Bacteroides fragilis strains in low concentrations. Reversion of the HA and also of the adsorption to beads of Sepharose coupled to bovine submaxillar mucin, by this sugar residue corroborated the recognition capacity of the bacterial lectin-like adhesin. However, when erythrocytes were treated with clostridial neuraminidase, an increase in the HA of some strains was observed. Protease treatment of erythrocytes abolished the HA, indicating that cell receptors of B. fragilis are probably a glycoprotein maiety.
Zusammenfassung Sialinsiiure und sialinsiiurereiche Makromolekiile hemmen in niedriger Konzentration die Hiimagglutinationsaktivitiit von Bacteroides fragilis. Hemmung der Hiimagglutination mit Kohlenhydraten sowie Absorptionsexperimente mit "bovine submaxillar mucin" -beschichteten Sephadex-Siiulen bestiitigten die Vermutung, daB es sich urn lektiniihnliche Adhiisionsmolekiile handelt. Inkubation der Erythrozyten in (Clostridium-)Neuraminidase bewirkte eine verstiirkte Hiimagglutination einiger Bacteroides-Stiimme; Proteasebehandlung hob die Hiimagglutination auf. Diese Effekte deuten darauf hin, daB die zelluliiren Rezeptoren fiir die Adhiision von B. fragilis offensichtlich durch Glykoproteinstrukturen reprasentiert werden.
Introduction The mechanism of Bacteroides fragilis adherence to erythrocytes and epithelial cells has been the subject of many investigations in recent years (3,5,6,8,9,10, 14, 15). Nevertheless, the molecular basis of the interactions involved in this process and its
Bacteroides fragilis Adhesin Recognition
341
importance in the virulence of this microorganism have remained unresolved until now. Rogemond and Guinet (11) have already identified lectin-like adhesins in B. fragi/is strains presenting affinity for D-galactosamine and D-glucosamine, but, other authors did not confirm these findings (6). As lectin specificity is generally expressed through various saccharides or macromolecules which are tested for their capacity to inhibit the interaction processes, we decided to favour, in this study, a hemagglutination model, followed by the adsorption to Sepharose beads in order to verify the capacity of sugar binding of B. fragilis adhesins.
Materials and Methods Strains and cultural conditions. Strains of B. fragilis isolated from clinical infections and normal flora were used. They were identified according to Holdeman et al. (2). A reference strain obtained from the American Type Culture Collection (Rockville, MD) - 25285 - was also included in the study. The organisms were grown in brain-heart infusion broth (Merck) pre-reduced and sterilized anaerobically for 18 h at 37°C. Following incubation, the cells were pelleted by centrifugation and resuspended in phosphate-buffered saline 0.01 M pH 7.2 (PBS) to give a final concentration of approximately 3.0 x 10 lD CFU/ml (optical density of 1.0 at 620 nm). This was the bacterial suspension used throughout the investigation. Hemagglutination activity (HA) studies. The hemagglutination assay was done using a suspension of human A-type erythrocytes (1%) with a microtiter plate as described by Okuda et al. (4). Two strains (ATCC 25285, MC2) were used in the hemagglutination inhibition (HI) tests, in which various hapten inhibitors were used (D-arabinose, D-fucose, fructose, Dgalactose, L-galactose, D-galactosamine, N-acetyl-D-galactosamine, D-glucose, Dglucosamine, N-acetyl-glucosamine, maltose, D-mannose, melibiose, D-raffinose, D-rhamnose, L-rhamnose, sucrose, stachiose, D-xylose, N-acetylneuraminic acid, bovine submaxillar mucin (BSM), fetuin, orosomucoid, hog gastric mucin, Salmonella LPS). The sugar residues defined as the HA inhibitors were then used in reversion tests: They were added to the mixtures of hemagglutination reactions, in their minimum inhibitory concentrations. Adsorption tests. BSM was coupled to cyanogen bromide activated Sepharose 4B (Pharmacia) (BSM-Sepharose) as described by Pereira and Kabat (7). The bacterial suspension (10 lD CFU/ml) was added to the mixture which was incubated at room temperature for 5 min and examined under the light microscope. Reversion tests were performed as cited above. Treatment of erythrocytes. Human A-type erythrocytes (5%) were incubated separately with 0.05 mg/ml of trypsin and pronase E in PBS and 0.1 U/ml of Clostridium perfringens neuraminidase (type X) in phosphate-buffered saline 0.01 M, pH 6.0, for 30 min at 37°C. All enzymes were purchased from Sigma. After treatment, erythrocytes were sedimented, washed extensively and resuspended in PBS to give a final concentration of 1 %.
Results and Discussion N-acetylneuraminic acid and macromolecules rich in this sugar residue (BSM, fetuin, orosomucoid and hog gastric mucin) were capable of inhibiting the HA of the B. fragilis strains tested, in low concentrations. Salmonella LPS which has a KDO that is accepted to be structurally similar to the sialic acid, was also an inhibitor of that activity. D-galactosamine and D-glucosamine, already described by other authors (11)
342
R. M. C. P. Domingues et al.
as residues recognized by one B. fragilis adhesin, also inhibited the HA, but at higher concentrations. These results have been summarized in Table 1. In the HA reversion tests performed using N-acetylneuraminic acid, D-galactosamine and D-glucosamine, the two first ones were capable of reverting completely the HA and D-glucosamine resulted only in a partial reversion. Assisted by a light microscope, we could observe the adsorption of the two strains to BSM-Sepharose beads and the same reversion pattern of the HA was also observed. These results suggested a specific recognition capacity of the adhesin for sialic acid. Sharon et al. (13) had already established such procedures as some of the prerequisites for the postulation of a lectin-like adhesin detection. It is also interesting to mention that after a brief period of bacterial adsorption to BSM-Sepharose, an elution process could be observed. In order to check the involvement of sialic acid as a receptor of B. fragilis lectin-like adhesin, erythrocytes were treated with a neuraminidase of C. perfringens and, unexpectedly, this treatment led to an increase in the hemagglutinating titres of 6 out of 12 strains analysed. Three strains (1M 108306, 1M 118310, 030M) did not present altered hemagglutinating titres, but we could macroscopically verify a more intense reaction. These results are reported in Table 2. Table 1. Minimal Inhibitory Concentration (MIC) of HA of B. fragilis strains Inhibitors Bovine submaxillary mucin Fetuin Hog gastric mucin Orosomucoid Salmonella LPS N-acetylneuraminic acid D-glucosamine D-galactosamine a
Strains/MICa (mg/ml) ATCC 25285 MC 2 0.008 0.06 0.5 0.8
0.008 0.06 0.5 0.8
1.25
1.25
0.75 5.3 43.0
0.75 10.7 43.0
Average of three assays.
Table 2. Hemagglutination titres of B. fragilis strains with untreated (UE) and neuraminidase-treated erythrocytes (NTE) Strains ATCC 25285 MC2 1M 097R041 20656-2-1, 1M 048210, 1M 018218 1M 108306, 1M 118310,030 M MC 1, MC 3, FF14 a Average of three assays. Increase of intensity reaction, macroscopically.
Hemagglutination titres a UE NTE 16 4 4 2
2
o
128 64 16 8 2"
o
Bacteroides fragilis Adhesin Recognition
343
We believe that the split-off of sialic acid from the erythrocyte surfaces by an exogenous neuraminidase, could cause an increase in the B. fragilis HA due to a reduction of the electrostatic repulsion, a notion that has been widely accepted in the literature (12). On the other hand, the hypothesis of the exposure of receptors and the participation of more than one adhesin cannot be overlooked. Guzman et al. (1) stated this possibility in a recent work and the authors suggested a possible action of the B. fragilis sialidase in the HA. Further investigations are being undertaken in our laboratory to confirm this hypothesis. Finally, in regard to erythrocyte protease treatments, these procedures abolished completely the HA of all the strains tested, indicating that the cell receptors of B. fragi/is adhesins have probably a glycoprotein moiety. Acknowledgements. We thank Jayme Angluster and Manuel Gil Esteves for constructive criticism and Fernando Magalhaes and Joaquim Santos Filho for the technical assistance. This work was supported by grants of the following national institutions: Conselho Nacional de Desenvolvimento Cientffico e Tecnol6gico (CNPq) and Funda
Sialic Acid as Receptor of Bacteroides fragilis Lectin-like Adhesin REGINA MARIA C. P. DOMINGUES, SiLVIA MARIA B. CAVALCANTI, ARNALDO F. B. ANDRADE, and MARIA CANDIDA S. FERREIRA Instituto de Microbiologia, Centro de Ciencias da Saude, Universidade Federal do Rio de Janeiro, Ilha do Fundao, 21944 Rio de Janeiro, RJ, Brasil
Received December 6, 1991 . Revision received February 25, 1992 . Accepted March 12, 1992
Summary It was observed that sialic acid and macromolecules rich in this sugar were able to inhibit the hemagglutination activity (HA) of Bacteroides fragilis strains in low concentrations. Reversion of the HA and also of the adsorption to beads of Sepharose coupled to bovine submaxillar mucin, by this sugar residue corroborated the recognition capacity of the bacterial lectin-like adhesin. However, when erythrocytes were treated with clostridial neuraminidase, an increase in the HA of some strains was observed. Protease treatment of erythrocytes abolished the HA, indicating that cell receptors of B. fragilis are probably a glycoprotein maiety.
Zusammenfassung Sialinsiiure und sialinsiiurereiche Makromolekiile hemmen in niedriger Konzentration die Hiimagglutinationsaktivitiit von Bacteroides fragilis. Hemmung der Hiimagglutination mit Kohlenhydraten sowie Absorptionsexperimente mit "bovine submaxillar mucin" -beschichteten Sephadex-Siiulen bestiitigten die Vermutung, daB es sich urn lektiniihnliche Adhiisionsmolekiile handelt. Inkubation der Erythrozyten in (Clostridium-)Neuraminidase bewirkte eine verstiirkte Hiimagglutination einiger Bacteroides-Stiimme; Proteasebehandlung hob die Hiimagglutination auf. Diese Effekte deuten darauf hin, daB die zelluliiren Rezeptoren fiir die Adhiision von B. fragilis offensichtlich durch Glykoproteinstrukturen reprasentiert werden.
Introduction The mechanism of Bacteroides fragilis adherence to erythrocytes and epithelial cells has been the subject of many investigations in recent years (3,5,6,8,9,10, 14, 15). Nevertheless, the molecular basis of the interactions involved in this process and its
Bacteroides fragilis Adhesin Recognition
341
importance in the virulence of this microorganism have remained unresolved until now. Rogemond and Guinet (11) have already identified lectin-like adhesins in B. fragi/is strains presenting affinity for D-galactosamine and D-glucosamine, but, other authors did not confirm these findings (6). As lectin specificity is generally expressed through various saccharides or macromolecules which are tested for their capacity to inhibit the interaction processes, we decided to favour, in this study, a hemagglutination model, followed by the adsorption to Sepharose beads in order to verify the capacity of sugar binding of B. fragilis adhesins.
Materials and Methods Strains and cultural conditions. Strains of B. fragilis isolated from clinical infections and normal flora were used. They were identified according to Holdeman et al. (2). A reference strain obtained from the American Type Culture Collection (Rockville, MD) - 25285 - was also included in the study. The organisms were grown in brain-heart infusion broth (Merck) pre-reduced and sterilized anaerobically for 18 h at 37°C. Following incubation, the cells were pelleted by centrifugation and resuspended in phosphate-buffered saline 0.01 M pH 7.2 (PBS) to give a final concentration of approximately 3.0 x 10 lD CFU/ml (optical density of 1.0 at 620 nm). This was the bacterial suspension used throughout the investigation. Hemagglutination activity (HA) studies. The hemagglutination assay was done using a suspension of human A-type erythrocytes (1%) with a microtiter plate as described by Okuda et al. (4). Two strains (ATCC 25285, MC2) were used in the hemagglutination inhibition (HI) tests, in which various hapten inhibitors were used (D-arabinose, D-fucose, fructose, Dgalactose, L-galactose, D-galactosamine, N-acetyl-D-galactosamine, D-glucose, Dglucosamine, N-acetyl-glucosamine, maltose, D-mannose, melibiose, D-raffinose, D-rhamnose, L-rhamnose, sucrose, stachiose, D-xylose, N-acetylneuraminic acid, bovine submaxillar mucin (BSM), fetuin, orosomucoid, hog gastric mucin, Salmonella LPS). The sugar residues defined as the HA inhibitors were then used in reversion tests: They were added to the mixtures of hemagglutination reactions, in their minimum inhibitory concentrations. Adsorption tests. BSM was coupled to cyanogen bromide activated Sepharose 4B (Pharmacia) (BSM-Sepharose) as described by Pereira and Kabat (7). The bacterial suspension (10 lD CFU/ml) was added to the mixture which was incubated at room temperature for 5 min and examined under the light microscope. Reversion tests were performed as cited above. Treatment of erythrocytes. Human A-type erythrocytes (5%) were incubated separately with 0.05 mg/ml of trypsin and pronase E in PBS and 0.1 U/ml of Clostridium perfringens neuraminidase (type X) in phosphate-buffered saline 0.01 M, pH 6.0, for 30 min at 37°C. All enzymes were purchased from Sigma. After treatment, erythrocytes were sedimented, washed extensively and resuspended in PBS to give a final concentration of 1 %.
Results and Discussion N-acetylneuraminic acid and macromolecules rich in this sugar residue (BSM, fetuin, orosomucoid and hog gastric mucin) were capable of inhibiting the HA of the B. fragilis strains tested, in low concentrations. Salmonella LPS which has a KDO that is accepted to be structurally similar to the sialic acid, was also an inhibitor of that activity. D-galactosamine and D-glucosamine, already described by other authors (11)
342
R. M. C. P. Domingues et al.
as residues recognized by one B. fragilis adhesin, also inhibited the HA, but at higher concentrations. These results have been summarized in Table 1. In the HA reversion tests performed using N-acetylneuraminic acid, D-galactosamine and D-glucosamine, the two first ones were capable of reverting completely the HA and D-glucosamine resulted only in a partial reversion. Assisted by a light microscope, we could observe the adsorption of the two strains to BSM-Sepharose beads and the same reversion pattern of the HA was also observed. These results suggested a specific recognition capacity of the adhesin for sialic acid. Sharon et al. (13) had already established such procedures as some of the prerequisites for the postulation of a lectin-like adhesin detection. It is also interesting to mention that after a brief period of bacterial adsorption to BSM-Sepharose, an elution process could be observed. In order to check the involvement of sialic acid as a receptor of B. fragilis lectin-like adhesin, erythrocytes were treated with a neuraminidase of C. perfringens and, unexpectedly, this treatment led to an increase in the hemagglutinating titres of 6 out of 12 strains analysed. Three strains (1M 108306, 1M 118310, 030M) did not present altered hemagglutinating titres, but we could macroscopically verify a more intense reaction. These results are reported in Table 2. Table 1. Minimal Inhibitory Concentration (MIC) of HA of B. fragilis strains Inhibitors Bovine submaxillary mucin Fetuin Hog gastric mucin Orosomucoid Salmonella LPS N-acetylneuraminic acid D-glucosamine D-galactosamine a
Strains/MICa (mg/ml) ATCC 25285 MC 2 0.008 0.06 0.5 0.8
0.008 0.06 0.5 0.8
1.25
1.25
0.75 5.3 43.0
0.75 10.7 43.0
Average of three assays.
Table 2. Hemagglutination titres of B. fragilis strains with untreated (UE) and neuraminidase-treated erythrocytes (NTE) Strains ATCC 25285 MC2 1M 097R041 20656-2-1, 1M 048210, 1M 018218 1M 108306, 1M 118310,030 M MC 1, MC 3, FF14 a Average of three assays. Increase of intensity reaction, macroscopically.
Hemagglutination titres a UE NTE 16 4 4 2
2
o
128 64 16 8 2"
o
Bacteroides fragilis Adhesin Recognition
343
We believe that the split-off of sialic acid from the erythrocyte surfaces by an exogenous neuraminidase, could cause an increase in the B. fragilis HA due to a reduction of the electrostatic repulsion, a notion that has been widely accepted in the literature (12). On the other hand, the hypothesis of the exposure of receptors and the participation of more than one adhesin cannot be overlooked. Guzman et al. (1) stated this possibility in a recent work and the authors suggested a possible action of the B. fragilis sialidase in the HA. Further investigations are being undertaken in our laboratory to confirm this hypothesis. Finally, in regard to erythrocyte protease treatments, these procedures abolished completely the HA of all the strains tested, indicating that the cell receptors of B. fragi/is adhesins have probably a glycoprotein moiety. Acknowledgements. We thank Jayme Angluster and Manuel Gil Esteves for constructive criticism and Fernando Magalhaes and Joaquim Santos Filho for the technical assistance. This work was supported by grants of the following national institutions: Conselho Nacional de Desenvolvimento Cientffico e Tecnol6gico (CNPq) and Funda
