Adhesiveness of Bacteroides fragilis Strains Isolated from Feces of Healthy Donors, Abscesses, and Blood. Carlos A. Guzma`n,1,2 Francesca Biavasco,2 Carla ...
CURRENT MICROBIOLOGY Vol. 34 (1997), pp. 332–334
An International Journal
R Springer-Verlag New York Inc. 1997
News & Notes
Adhesiveness of Bacteroides fragilis Strains Isolated from Feces of Healthy Donors, Abscesses, and Blood Carlos A. Guzma`n,1,2 Francesca Biavasco,2 Carla Pruzzo2 1Division 2Institute
of Microbiology, GBF-National Research Centre for Biotechnology, Mascheroder Weg 1, D-38124 Braunschweig, Germany of Microbiology University of Ancona, Via Ranieri Monte D’Ago, 60131 Ancona, Italy
Received: 2 October 1996 / Accepted: 21 November 1996
Abstract. Bacteroides fragilis strains attached to oral epithelial cells (ECs) and the cell line Intestine 407 and associated with human phagocytes with different efficiencies depending on their source. The 58%, 75%, and 40% of strains isolated from feces, abscesses, and blood respectively adhered to ECs with good efficiency (11–40 bacteria/cell). Of the strains from feces and abscesses, 17% and 20% exhibited a high adherence (.40 bacteria/cell); however, none of the blood isolates presented this property. Similar results were obtained with the cell line Intestine 407 and human phagocytes. Of the isolates from feces, abscesses, and blood, 20%, 56%, and 71% respectively also exhibited hemagglutination ability, indicating that this property is a virulence trait more frequently present among pathogenic isolates than in commensal strains.
In recent years several studies have tried to elucidate the physio-pathogenetic mechanisms underlying infections caused by Bacteroides fragilis, the anaerobe most frequently isolated from clinical specimens [2, 4, 7]. The attachment process has received special attention, and it has been shown that B. fragilis strains are able to bind erythrocytes and epithelial cells (ECs) [1, 3, 4, 6, 8–12, 14, 15]. Although it has not been unequivocally demonstrated, both capsule and pili seem to be involved in this process [1, 9–11, 14]. Moreover, a correlation has been observed between virulence and hemagglutination (HA) ability [4, 11]. In the present work the role of adherence in the pathogenesis of B. fragilis was further analyzed with a large number of strains isolated from different sources. We evaluated the ability of strains isolated from clinical specimens and feces of healthy donors to attach to human ECs, to agglutinate red blood cells, and to interact with human polymorphonuclear leukocytes (PMNs).
‘‘abscess isolates’’) from intraabdominal, perirectal, and pelvic abscesses, and wounds. Isolates were considered responsible for infections if they were recovered from clinical specimens as a pure culture or, if they were isolated in mixed cultures, when they were the only anaerobes present. They were identified as B. fragilis by examination of their Api 20 patterns. In all experiments, cultures were incubated for 24 h in an anaerobic glove box in Wilkins-Chalgren (Oxoid Ltd).
Materials and Methods
All B. fragilis strains, independent of the source, exhibited the ability to adhere to the tested cells. Most of the isolates deriving from feces of healthy donors efficiently attached to ECs (75% of the strains), cell line Intestine 407 (74%), and PMNs (82%) (Table 1). All HA-positive feces isolates attached to ECs and PMNs with either good
Bacterial strains and media. Seventy-six B. fragilis strains were isolated from feces of healthy donors, 48 from blood, and 63 (named
Correspondence to: C. Pruzzo
HA, adherence to ECs and Intestine 407 cells, and association with PMNs. Bacterial agglutination of human (group A) red blood cells and adherence to both cheek ECs and the cell line Intestine 407 (ATCC CCL 6) were performed as previously described [10]; adherence values were calculated by averaging the number of bacteria adhering to 20 cells. Experiments of association with human PMNs were performed on cover slips without serum at 37°C, as described by Silverblatt et al. [13]; results were expressed as binding index (total number of bacteria/100 PMNs). The significance of the results obtained between different groups of isolates was estimated by the x2 test; differences were considered significant at P # 0.05.
Results and Discussion
333
C.A. Guzma`n et al.: Adhesiveness of B. fragilis Strains Table 1. Ability to both adhere to oral and Intestine 407 cells and associate with PMNs of B. fragilis strains deriving from clinical specimens and feces No. of strains binding to Oral Strain source and no. Feces 76 Abscess 63 Blood 48
ECsb
Intestine 407 cellsb
PMNsc
No. of HApositive strainsa
Poor efficiency
Good efficiency
High efficiency
Poor efficiency
Good efficiency
High efficiency
Poor efficiency
Good efficiency
High efficiency
15 (20%) 35 (56%) 34 (71%)
19 (25%) 3 (5%) 29 (60%)
44 (58%) 47 (75%) 19 (40%)
13 (17%) 13 (20%) 0
20 (26%) 5 (8%) 30 (63%)
43 (57%) 44 (70%) 18 (37%)
13 (17%) 14 (22%) 0
11 (15%) 6 (10%) 34 (71%)
48 (63%) 45 (71%) 14 (29%)
17 (22%) 12 (19%) 0
The percentage of strains showing the studied property is indicated in parentheses. Results obtained with feces and abscess isolates were previously described [11]. b The efficiency of adherence to oral and Intestine 407 cells was arbitrarily recorded as poor, good, or high when the mean number of bacteria per cell was #10, between 11 and 40, and .40, respectively. c The efficiency of association with PMNs was arbitrarily recorded as poor, good, or high when binding index was #100, between 101 and 300, and .300, respectively. a
(n 5 5) or high (n 5 10) efficiency. The percentage of abscess isolates that efficiently bound to ECs, Intestine 407 cells, and PMNs was even higher (90–95%; Table 1). Interestingly, all HA-positive strains bound to the tested cells with either good (n 5 23) or high (n 5 12) efficiency. The difference between the percentages of strains adhering with good/high efficiencies among feces and abscess isolates was significant (P # 0.05), suggesting that the adherence ability is stronger among isolates from suppurative infections than in commensal strains. When the blood isolates were analyzed, it was found that most of them bound poorly to ECs, Intestine 407 cells, and PMNs, and none exhibited high attachment efficiency (Table 1). Although HA capacity correlates with good/high attachment efficiency among feces and abscess isolates, no correlation was observed in blood isolates. In fact, out of 34 HA-positive blood isolates, 20 attached to ECs and associated with PMNs poorly, whereas 14 exhibited a good attachment efficiency. We have previously shown that hemagglutinating strains are more frequently isolated from abscesses than from feces of healthy donors [11]. These results are further supported by the present work, where 71% of the blood isolates are HA positive, indicating that HA is a virulence trait more frequently associated with pathogenic than with commensal isolates (62% versus 20%; P # 0.01). Although the contribution of different adhesins to bacterial attachment was not analyzed here, we have previously shown that strong adhesion and HA are associated with pili [11]. This suggests that pili may be responsible for the high efficiency attachment phenotype observed only in feces and abscess isolates. This hypothesis is further supported by the work of Brook et al. [1]
showing the presence of pili in 75%, 69%, and 6% of abscesses, feces, and blood isolates, respectively. It is well known that adherence ligand synthesis can be modulated in relation to environmental conditions [5]. It is likely that during the infection cycle adherence confers B. fragilis a selective advantage at a mucosal level and in the suppurative sites. In contrast, when the strains undergo systemic dissemination via a hematogenous route, this property may not be required and may even be detrimental for bacterial survival. The downregulation of adhesin expression or their masking by the capsule [1, 6] may protect bacteria within a niche where they are more prone to engulfment and killing by phagocytic cells. ACKNOWLEDGMENT This work was supported by Ministero della Ricerca Scientifica e Tecnologica (Grant 60%).
Literature Cited 1. Brook I, Myhal LA, Dorsey CH (1992) Encapsulation and pilus formation of Bacteroides spp. in normal flora abscesses and blood. J Infect 25:251–257 2. Crabb JH, Finberg R, Onderdonk AB, Kasper DL (1990) T cell regulation of Bacteroides fragilis-induced intraabdominal abscesses. Rev Infect Dis 12 S2:S178–S184 3. Guzma`n CA, Plate` M, Pruzzo C (1990) Role of neuraminidasedependent adherence in Bacteroides fragilis attachment to human epithelial cell. FEMS Microb Lett 71:187–192 4. Namavar F, Verweij-Van Vught MAJJ, MacLaren DM (1991) A study of the candidate virulence factors of Bacteroides fragilis. J Gen Microbiol 137:1431– 1435 5. Ofek I, Mosek I, Sharon N (1980) Mannose specific adherence of Escherichia coli freshly excreted in the urine of patients with
334
6. 7.
8.
9.
10.
urinary tract infections, and of isolates subcultured from the infected urine. Infect Immun 34:708–711 Onderdonk AB, Moon NE, Kasper DL, Bartlett JG (1978) Adherence of Bacteroides fragilis in vivo. Infect Immun 19:1083–1087 Onderdonk AB, Cisneros RL, Finberg R, Crabb JH, Kasper DL (1990) Animal model system for studying virulence of and host response to Bacteroides fragilis. Rev Infect Dis 12S2:S169–S177 Oyston PCF, Handley PS (1990) Surface structures, hemagglutination and cell surface hydrophobicity of Bacteroides fragilis strains. J Gen Microbiol 136:941–948 Patrick S, Coffey A, Emmerson AM, Larkin MJ (1988) The relationship between cell surface structure expression and hemagglutination in Bacteroides fragilis. FEMS Microbiol Lett 50:67–71 Pruzzo C, Dainelli B, Ricchetti M (1984) Piliated Bacteroides fragilis strains adhere to epithelial cells and are more sensitive to phagocytosis by human neutrophils than nonpiliated strains. Infect Immun 43:189–194
CURRENT MICROBIOLOGY Vol. 34 (1997) 11. Pruzzo C, Guzma`n CA, Dainelli B (1989) Incidence of hemagglutination activity among pathogenic and non-pathogenic Bacteroides fragilis strains and role of capsule and pili in HA and adherence. FEMS Microbiol Lett 59:113–118 12. Riley TV, Mee BJ (1984) Hemagglutination of Bacteroides fragilis. FEMS Microbiol Lett 25:229–232 13. Silverblatt FJ, Dreyer JS, Schaver S (1979) Effect of pili on susceptibility of Escherichia coli to phagocytosis. Infect Immun 24:218–223 14. van Doorn J, Mooi FR, Verweij-van Vught MAJJ, MacLaren DM (1987) Characterization of fimbriae from Bacteroides fragilis. Microb Pathog 3:87–95 15. Vel WAC, Namavar F, Verweij-Van Vught MAJJ, Pubben ANB, MacLaren DM (1986) Hemagglutination by the Bacteroides fragilis group. J Med Microbiol 21:105–107
