Interbacterial Adherence Between Actinomyces viscosus and Strains

Transcription

1 INFECTION AND IMMUNITY, Apr. 1984, p /84/ $02.00/0 Copyright 1984, American Society for Microbiology Vol. 44, No. 1 Interbacterial Adherence Between Actinomyces viscosus and Strains of Streptococcus pyogenes, Streptococcus agalactiae, and Pseudomonas aeruginosa K. KOMIYAMA1 AND R. J. GIBBONS2* University of Saskatchewan, Saskatoon, Saskatchewan, Canada 57N OWO, and Forsyth Dental Center, Boston, MA Received 4 November 1983/Accepted 10 January 1984 Interbacterial adherence was sought between strains of Actinomyces viscosus indigenous to the human mouth and strains of Streptococcus pyogenes, Streptococcus agalactiae, and Pseudomonas aeruginosa. Six of nine strains of S. pyogenes, three of five strains of S. agalactiae, and two of four strains of P. aeruginosa were found to coaggregate with each of five strains of A. viscosus tested. Some coaggregation reactions were inhibited by 0.05 M lactose and were dependent upon heat- and protease-sensitive Actinomyces components. Such reactions appear to involve the galactosyl-binding adhesin previously described in type 2 fimbriae on A. viscosus. Other coaggregation reactions were dependent upon heat- and protease-sensitive components of the pathogen. That such pathogen strains possessed an adhesin(s) was further suggested by the observation that they agglutinated human erythrocytes. The ability of coaggregation-positive and -negative strains of S. pyogenes and S. agalactiae to adhere to Actinomyces-coated agarose beads was also studied. Coaggregation-positive streptococcal strains attached in higher numbers to the Actinomyces-coated beads than did strains which were coaggregation negative. Lactose (0.05 M) inhibited the attachment of those streptococcal strains which coaggregated with A. viscosus in a lactosesensitive manner. The adherence of those streptococcal strains whose coaggregation appeared to depend upon the galactosyl-binding adhesin of A. viscosus was also reduced by components of human saliva. Crude sonic extracts of coaggregation-positive streptococci or of P. aeruginosa were also effective in aggregating Actinomyces cells. The effect of lactose and of salivary components on this extract-induced aggregation of Actinomyces cells generally paralleled that observed in other assays. The apparent prevalence and diversity of adherent reactions between the pathogens studied and indigenous strains of A. viscosus suggest that some may affect host susceptibility to these infectious agents. Downloaded from Different strains and species of oral bacteria often coaggregate when mixed together (5). These reactions are highly specific, as they occur with only certain strains of a given species (see reference 1 for review). Coaggregations involving strains ofactinomyces viscosus or Actinomyces naeslundii and various oral streptococci have been particularly well studied. McIntire and co-workers (11) have showed that many coaggregation reactions between these organisms are reversible by lactose and involve a heat- and proteasesensitive lectin-like adhesin which is contained in type 2 fimbriae on the Actinomyces surface (1-3, 11). These adhesins bind to heat-stable galactosyl-containing receptors on the streptococci (1, 10). Although such lactose-reversible coaggregations are by far the most common type (8), other coaggregation reactions may occur which depend upon different adhesins and receptors on Actinomyces spp. or streptococci, respectively (1, 3). To date, six groups of Actinomyces spp. and six groups of Streptococcus sanguis or Streptococcus mitis have been delineated based upon the sensitivity of their coaggregation reactions to heat and protease treatment and to inhibition by lactose (3, 7). Coaggregation reactions have been suggested to promote the accumulation of bacteria in dental plaques on teeth (1, 6). In support of this view, McBride and Van Der Hoeven (9) have shown that much higher numbers of Veillonella spp. can be recovered from the teeth of gnotobiotic rodents who are monoinfected with Streptococcus mutans strains which specifically coaggregate with Veillonella spp. There is also * Corresponding author. 86 evidence that the attachment of bacteria to microorganisms of a different species can, in some instances, promote initial colonization. For example, oral colonization of strains of Bacteroides gingivalis, which coaggregate with Actinomyces spp. and other gram-positive bacteria, appears to be fostered by the presence of bacterial plaques which contain such organisms on teeth (12). Although it is clear that coaggregation reactions are prevalent among oral bacteria and some are likely to play a role in colonization, little is known about the potential of indigenous bacteria to coaggregate with exogenous pathogens. The present study reports that coaggregation reactions are also common between oral strains of A. viscosus and strains of Streptococcus pyogenes, Streptococcus agalactiae, and Pseudomonas aeruginosa. MATERIALS AND METHODS Cultures and cultural conditions. Four fresh clinical isolates of group A beta-hemolytic S. pyogenes (847, 9039, 9261, and 9483) and four strains of P. aeruginosa (3218, 5623, 6695, and 7360) were obtained from L. Kuntz (Massachusetts General Hospital, Boston). Five additional fresh isolates of S. pyogenes (312, 324, 376, 406, and 804) were obtained from M. Worthington (St. Elizabeth Hospital, Brighton, Mass.). Five strains of group B S. agalactiae representative of each of five serotypes (strain 090, type la; strain H36B, type lb; strain A909, type lc; strain 18RS21, type 2; and strain M732, type 3) were obtained from D. Kasper (Harvard Medical School, Boston, Mass.). A. viscosus strains T14Vi, S2, LY7, M100, and W1838 were obtained from the culture collection of the Forsyth Dental Center. on October 4, 2018 by guest

2 VOL. 44, 1984 The streptococci were grown aerobically in Todd-Hewitt broth (BBL Microbiology Systems, Cockeysville, Md.), and the Pseudomonas strains were grown aerobically in Trypticase soy broth (BBL). Strains of A. viscosus were grown in Trypticase soy broth incubated in Brewer Anaerobic Jars filled with 80% N2, 10% H2, and 10% CO2. All cultures were incubated at 35 C. Coaggregation assays. Bacteria harvested from overnight cultures were washed three times with the coaggregation buffer described by Cisar et al. (3) and suspended in this buffer at a final concentration of 2.5 x 10 cells per ml. Coaggregation assays were performed by mixing 25-pil samples of each washed bacterial suspension in microtitration plates. Controls consisted of samples of each suspension mixed with buffer. The mixtures were incubated for 1 h at room temperature with gentle shaking, and after settling, they were scored as 0 (no coaggregation) up to 4+ (marked coaggregation) with the aid of a dissecting microscope. The effect of lactose, heat, and protease treatment was studied as described by Cisar et al. (3). Hemagglutination reactions were sought by using untreated or neuraminidase-treated human erythrocytes of type A and B as described by Costello et al. (4). In these assays, phosphate-buffered saline suspensions containing 4 x 108 organisms per ml were used. Effect of saliva on adherence of pathogens to Actinomycescoated agarose beads. Components of saliva often partially inhibit the adherent interactions of bacteria (6), including those involved in coaggregations (K. Komiyama and R. J. Gibbons, Caries Res., in press). Therefore, it was of interest to determine whether salivary components affected the adherent interactions between strains of A. viscosus and the oral-pharyngeal pathogens. All strains of A. viscosus, S. pyogenes, S. agalactiae and P. aeruginosa studied were found to agglutinate strongly (2+ to 4+) when incubated with clarified whole human saliva prepared as previously described (Komiyama and Gibbons, in press). Therefore, an assay system was used which did not depend upon visual assessment of bacterial agglutination. This involved the TABLE 1. Coaggregation reactions between pathogens and A. viscosus strains Pathogen strain A. viscosus strain T14Vi S2 LY7 M100 W1838 S. pyogenes S. agalactiae 18RS H36B A M P. aeruginosa ADHERENT REACTIONS OF A. VISCOSUS STRAINS 87 TABLE 2. Characteristics of coaggregation reactions and hemagglutinating activity of pathogens Coaggregation Hemag- Coaggre- Coaggre- after heat and glutina- Coatge-wt gation in protease tion of Pathogen strain A. visco- presence treatment of: human Of 0.05 M sus erythro- lactose Patho- A. vis- cytes gen cosus cye S. pyogenes , 9039, 9483 S. pyogenes , 804, 9261 S. agalactiae A909 S. agalactiae , M732 P. aerugin osa 3218, 5623 attachment of [3H]thymidine-labeled streptococci to Actinomyces-coated agarose beads as previously described (Komiyama and Gibbons, in press). Briefly, samples (30,ul) of spermine-conjugated agarose beads (Pierce Chemical Co., Rockford, Ill.) were incubated with 50 x 108 washed Actinomyces cells in 0.5 ml of coaggregation buffer for 1 h at room temperature. The beads were then washed three times with coaggregation buffer and treated with 1% bovine serum albumin to reduce nonspecific adsorption to any uncovered areas of the beads. Microscopic examination indicated that the beads were almost entirely covered with Actinomyces cells. The beads were again washed three times with coaggregation buffer and incubated with 300,ul containing 7.5 x 107 [3H]thymidine-labeled streptococci prepared as previously described (Komiyama and Gibbons, in press). This concentration was used because it permitted reliable quantitation of the numbers of streptococci which attached to the Actinomyces-coated beads; other streptococcal concentrations were not assessed. The streptococci were suspended in coaggregation buffer alone, in buffer containing 0.05 M lactose, or in samples of clarified whole human saliva which had been depleted of streptococcal agglutinating activity by two absorptions with 1010 washed streptococcal cells per ml (Komiyama and Gibbons, in press). Saliva depleted of streptococcal agglutinins still induced aggregation of the A. viscosus strains studied. Aggregation of Actinomyces cells by sonic extracts of pathogens. The ability of sonic extracts of the pathogens to induce aggregation of Actinomyces cells was also studied. Extracts were prepared by treating suspensions containing 1010 washed bacteria per ml with sonic oscillation for 3 min, using an MSE 100-watt ultrasonic disintegrator (MSE, Inc., Cleveland, Ohio) at maximum output as previously described (Komiyama and Gibbons, in press). The suspensions were centrifuged at 10,000 x g for 10 min, and the resulting supernatant liquid was assayed for its ability to agglutinate Actinomyces cells in microtitration plates. The effect of lactose on the extract-induced agglutination was studied by incorporating this sugar in reaction mixtures at a final concentration of 0.05 M. The effect of clarified whole human saliva which had been absorbed with 1010 washed Actinomyces cells per ml to remove agglutinins for these organisms was also studied.

3 88 KOMIYAMA AND GIBBONS INFECT. IMMUN. RESULTS Coaggregation between oral Actinomyces strains and pathogens. It was found that six of the nine strains of S. pyogenes, three of the five strains of S. agalactiae, and two of the four strains of P. aeruginosa that were studied coaggregated with all five strains of A. viscosus (Table 1). Coaggregationpositive strains of S. pyogenes could be divided into two groups based upon the sensitivity of their coaggregation reactions to 0.05 M lactose (Table 2). Lactose-sensitive coaggregation reactions involving S. pyogenes 847, 9039, and 9483 were found to be dependent upon heat heat- and protease-sensitive components of the Actinomyces strains. In contrast, lactose-insensitive coaggregations involving S. pyogenes 376, 804, and 9262 were dependent upon heat- and protease-sensitive components of the streptococci (Table 2). All of the coaggregation reactions involving the three strains of group B streptococci tested were inhibited by lactose. However, coaggregation reactions involving S. agalactiae A909 were dependent upon heat- and proteasesensitive components of the Actinomyces strains, whereas the coaggregation reactions involving S. agalactiae 090 and M732 were dependent upon heat- and protease-sensitive components of the group B streptococci (Table 2). Coaggregation reactions involving the two strains of P. aeruginosa were not inhibited by lactose and were dependent upon heat- and protease-sensitive components of the gram-negative bacilli. Hemagglutination reactions. The galactosyl-binding fimbriae of A. viscosus strains are known to agglutinate neuraminidase-treated erythrocytes (1, 4). Therefore, it was of interest to determine the hemagglutinating potential of the coaggregation-positive strains of S. pyogenes, S. agalactiae, and P. aeruginosa studied. Pathogen strains which possessed heatand protease-sensitive components involved in their coaggregation reactions with Actinomyces cells also exhibited hemagglutinating activity for human erythrocytes (Table 2). These reactions were stronger when neuraminidase-treated erythrocytes were used. However, pathogen strains whose coaggregation reactions were dependent upon heat- and protease-sensitive components of A. viscosus cells did not exhibit hemagglutinating activity (Table 2). No differences were noted between human erythrocytes of blood types A and B. Effect of saliva on the adherence of S. pyogenes and S. agalactiae to Actinomyces-coated agarose beads. The results of assays of the adherence of 3H-labeled S. pyogenes and S. agalactiae cells to Actinomyces-coated agarose beads are shown in Table 3. Approximately 106 cells of coaggregationnegative S. pyogenes 406, 312, and 324 and S. agalactiae H36B and 18RS21 attached to the Actinomyces-coated beads when the streptococci were suspended in buffer (Table 3). Use of 0.05 M lactose or of streptococcal-absorbed saliva as a suspending fluid had only a small effect on the number of these strains which attached. Significantly higher numbers of coaggregation-positive S. pyogenes 9483, 847, 9039, and 9261 and S. agalactiae A909, M732, and 090 cells attached to the Actinomyces-coated beads than did coaggregation-negative strains when the streptococci were suspended in buffer (Table 3). The presence of streptococcal-absorbed saliva or 0.05 M lactose markedly reduced the number of S. pyogenes 9483, 847, and 9039 cells which attached. These streptococcal strains formed lactose-sensitive coaggregates which depended upon heat- and protease-sensitive Actinomyces components (Table 2). In contrast, neither lactose nor streptococcal-absorbed saliva greatly reduced the number of S. pyogenes 9261 cells which attached to the Actinomycescoated beads; the coaggregates formed by this streptococcal strain were not sensitive to lactose and were dependent upon heat- and protease-sensitive components of the pathogen (Table 2). It is interesting to note that S. agalactiae M732 and 090 failed to coaggregate in the presence of 0.05 M lactose, and this sugar markedly reduced the number which adsorbed to the Actinomyces-coated beads (Table 3). However, samples of absorbed saliva did not strongly inhibit the number of M732 or 090 cells which attached to the Actinomyces-coated beads (Table 3). Coaggregation reactions involving S. agalactiae M732 and 090, like those of S. pyogenes 9261 discussed above, were dependent upon heat- and proteasesensitive components of the streptococci, rather than upon the galactosyl-binding adhesin of A. viscosus (Table 2). TABLE 3. Effect of lactose and saliva on attachment of 3H-labeled S. pyogenes and S. agalactiae strains to A. viscosus T14Vi-coated agarose beads No. (x 106 ± SE) attached to A. viscosus T14Vi-coateda agarose Pathogen Coaggregate Coaggregation beads when suspended in: straln with A. in presence of viscosus 0.05 M lactose Buffer lactose Saliva' S. pyogenes ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.11 S. agalactiae H36B 0.97 ± ± ± RS ± ± ± 0.01 A ± ± ± ± ± ± 0.27 M ± ± ± 0.25 a Similar results were also obtained with A. viscosus M100 and S2. b Saliva was absorbed with the respective streptococcal strain to remove agglutinins.

4 VOL. 44, 1984 Thus, although the adherent reactions of strains M732 and 090 were inhibited by lactose, salivary components capable of interacting with the putative streptococcal adhesins would not be expected to be present in the streptococcal-absorbed saliva. This likely explains why the streptococcal-absorbed saliva did not affect the attachment of strains M732, 090, and Aggregation of Actinomyces cells by sonic extracts of pathogens. Sonic extracts containing cell wall fragments and other material from all coaggregation-positive strains of S. pyogenes, S. agalactiae, and P. aeruginosa that were studied induced aggregation of washed Actinomyces cells, but extracts from coaggregation-negative strains did not (Table 4). Lactose (0.05 M) inhibited the agglutination induced by extracts from those S. pyogenes and S. agalactiae strains which exhibited lactose-sensitive coaggregations with A. viscosus cells. However, this sugar did not affect the agglutination of Actinomyces cells by sonic extracts derived from S. pyogenes or P. aeruginosa strains which did not exhibit lactose-sensitive coaggregations (Tables 2 and 4). Therefore, the agglutination of Actinomyces cells by sonic extracts of the pathogens had features which paralleled their coaggregation properties. Samples of clarified whole human saliva which had been absorbed with Actinomyces cells also inhibited the aggregation of Actinomyces cells induced by sonic extracts of S. pyogenes 847, 9039, and 9483 and S. agalactiae A909 (Table 4). However, the absorbed saliva did not affect aggregation of Actinomyces cells by sonic extracts of S. pyogenes 376, 804, 9261, S. agalactiae A909 and M732, or P. aeruginosa 3218 and Thus, the inhibitory effect of salivary components in this assay correlated with the effects noted when studying adherence of 3H-labeled pathogens to Actinomyces-coated agarose beads (Table 4). DISCUSSION The data obtained in the present study indicate that strains of A. viscosus indigenous to the human mouth frequently coaggregate with strains of groups A and B streptococci and with P. aeruginosa. These reactions occur in a varied, but specific, manner and appear to be generally similar to those described previously between strains of A. viscosus and oral streptococci (1, 3, 7). Some coaggregation reactions involving strains of S. pyogenes and S. agalactiae depended upon heat- and protease-sensitive components of A. viscosus. These components probably consist of the galactosyl-binding adhesin contained in type 2 fimbriae on the surface of A. viscosus cells as delineated by Cisar et al. (1-3) since their reactions were inhibitable by lactose (Tables 2 and 4). In contrast, coaggregation reactions involving S. pyogenes 376, 803, and 9261, S. agalactiae A909, and P. aeruginosa 3218 and 5623 were dependent upon heat- and protease-sensitive components of the pathogen. These components probably represent adhesins of the pathogen which bind to heat-stable receptors on A. viscosus cells. This possibility was further supported by the observation that these pathogen strains also agglutinated human erythrocytes, particularly those which had been treated with neuraminidase. The prevalence and diversity of the coaggregation reactions which developed between the strains of the pathogens studied and indigenous strains of A. viscosus suggest that some of these reactions may play a role in the adherence and subsequent colonization of pathogenic organisms when they are introduced into the oral environment. The adherent interactions involving S. pyogenes 376, 804, and 9261, S. ADHERENT REACTIONS OF A. VISCOSUS STRAINS 89 TABLE 4. Effect of Actinomyces-absorbed saliva and lactose on aggregation of Actinomyces cells induced by sonic extracts of pathogens Aggregation Extract-induced Pathogen strain Coaggregate with A. visof Actinomyces cels aggregation of Actipresence ofi tract of pathogena 0.05 M Lactose Saliva S. pyogenes , 324, 406 S. pyogenes , 9039, 9483 S. pyogenes , 804, 9261 S. agalactiae - - H36B, 18RS21 S. agalactiae + + A909 S. agalactiae , M732 P. aeruginosa 6695, 7360 P. aeruginosa , 5623 a All five strains of A. viscosus behaved similarly. b Saliva had been absorbed with respective A. viscosus strains. agalactiae 090 and M732, and P. aeruginosa 3218 and 5623 may be especially important in this regard since they were not markedly affected by components of saliva in the in vitro assays used. If such interbacterial interactions involving pathogenic organisms and indigenous bacteria do prove to be of ecological significance in vivo, they would represent a novel way in which the indigenous flora affects the susceptibility of a host to infectious agents. ACKNOWLEDGMENTS This investigation was supported in part by Public Health Service grant DE from the National Institute for Dental Research and by a grant from the Colgate-Palmolive Co. LITERATURE CITED 1. Cisar, J Coaggregation reactions between oral bacteria: studies of specific cell-to-cell adherence mediated by microbial lectins, p In R. J. Genco and S. E. Mergenhagen (ed.), Host-parasite interactions in periodontal diseases. American Society for Microbiology, Washington, D.C. 2. Cisar, J. O., E. L. Barsumian, S. H. Curl, A. E. Vatter, A. L. Sandberg, and R. P. Siraganian Detection and localization of a lectin on Actinomyces viscosus T14V by monoclonal antibodies. J. Immunol. 127: Cisar, J. O., P. E. Kolenbrander, and F. C. McIntire Specificity of coaggregation reactions between human oral streptococci and strains of Actinomyces viscosus and Actinomyces naeslundii. Infect. Immun. 24: Costello, A. H., J. 0. Cisar, P. E. Kolenbrander, and 0. Gabriel Neuraminidase-dependent hemagglutination of human erythrocytes by human strains of Actinomyces viscosus and Actinomyces naeslundii. Infect. Immun. 26: Gibbons, R. J., and M. Nygaard Interbacterial aggregation of plaque bacteria. Arch. Oral Biol. 15: Gibbons, R. J., and J. van Houte Bacterial adherence and the formation of dental plaques, p In E. H. Beachey (ed.), Bacterial adherence. Chapman and Hall, London. 7. Kolenbrander, P. E., Y. Inouye, and L. V. Holdeman

5 90 KOMIYAMA AND GIBBONS New Actinomyces and Streptococcus coaggregation groups among human oral isolates from the same site. Infect. Immun. 41: Kolenbrander, P. E., and B. L. Williams Lactose-reversible coaggregation between oral actinomycetes and Streptococcus sanguis. Infect. Immun. 33: McBride, B. C., and J. S. Van Der Hoeven Role of interbacterial adherence in colonization of the oral cavities of gnotobiotic rats infected with Streptococcus mutans and Veillonella alcalescens. Infect. Immun. 33: McIntire, F. C., L. K. Crosby, and A. E. Vatter Inhibitors INFECT. IMMUN. of coaggregation between Actinomyces viscosus T14V and Streptococcus sanguis 34:,-galactosides, related sugars, and anionic amphipathic compounds. Infect. Immun. 36: McIntire, F. C., A. E. Vatter, J. Baros, and J. Arnold Mechanism of coaggregation between Actinomyces viscosus T14V and Streptococcus sanguis 34. Infect. Immun. 21: Slots, J., and R. J. Gibbons Attachment of Bacteroides melaninogenicus subsp. asaccharolyticus to oral surfaces and its possible role in colonization of the mouth and of periodontal pockets. Infect. Immun. 19: