Search

The ability of counter immunoelectrophoresis (CIE) to detect Bacteroides fragilis endotoxin in amniotic fluid in small concentrations was evaluated. A method was developed which, in combination with ultrafiltration, permits detection of B. fragilis endotoxin in amniotic fluid in a concentration of 40 ng/ml or more. The sensitivity threshold was reduced to 2 ng/ml by using a highly reactive IgG-fraction isolated from rabbit anti-B. fragilis IPL E 323 antiserum.

The influence of the presence of saliva from different salivary glands on the adherence of Streptococcus gordonii strain HG 222 to saliva-coated polystyrene surfaces was tested. In the presence of undiluted parotid saliva or diluted whole, submandibular and sublingual saliva the adherence of HG 222 was enhanced by the formation of small aggregates on the attachment surface. In the presence of undiluted whole, submandibular and sublingual saliva large aggregates were formed and the adherence to saliva-coated polystyrene surfaces was inhibited. Adherence in the presence of whole saliva compared to adherence in buffer was decreased when lower densities of bacterial suspension were used, although in this case in the presence of whole saliva smaller bacterial aggregates were formed. In conclusion, these results suggest that the presence of saliva in solution may both enhance and decrease the adherence of S. gordonii HG 222 to saliva-coated polystyrene surfaces, partly depending on the size of bacterial aggregates that are formed in the presence of saliva.

Seventeen strains of oral bacteria of the genera Actinomyces (5), Bacteroides (3), and Streptococcus (9) were tested for aggregation by the human whole salivary mucin fraction (HWSM) in comparison to three types of animal mucin preparations from submandibular glands of cow (BSM) and sheep (OSM), and from the stomach of pig (PGM). Considerable variation was seen with respect to the rate and titer of aggregation induced by these mucins. The aggregating activity of HWSM varied widely among the different bacterial strains. The Bacteroides group showed hardly any induced aggregation, whereas the final aggregation titers varied for S. sanguis (3 strains) between 12 and 48, for S. oralis (3 strains) between 6 and 48, for the S. mutans group (3 strains) between 6 and 96, and for the five Actinomyces strains even between 6 and 192. For a particular strain, similar differences in titer were seen between the four mucins. For a human salivary mucin (MG-2) it has been described that sialic acid in the sequence NeuAc (α2,3)Gal(β1,3)GalNac- was specifically involved in the interaction with S. sanguis strains, in contrast to S. rattus BHT. Our results, however, indicate that this sugar sequence is not a prerequisite for the aggregation of S. sanguis, as animal mucins, devoid of this structure, were equally well or even better capable of inducing aggregation. On the other hand, desialization of BSM and OSM largely abolished their aggregating capability towards S. rattus BHT. Moreover, it was found that BSM and OSM, which are comparable with respect to their major oligosaccharide structure, show considerable differences in aggregating activity towards the same bacterial strain. The results indicate that the interaction and aggregation of oral bacteria with mucins is not necessarily dictated by specific oligosaccharide structures of the mucins, but may be caused instead by common physico-chemical features of the mucins as well.

The contribution of human parotid (Par) and submandibular/sublingual (SM/SL) saliva and of the human whole salivary mucin fraction (HWSM) to saliva-induced bacterial aggregation was studied for S. sanguis C476, S. oralis I581, and S. rattus HG 59. The mucous SM/SL saliva showed a much higher aggregation potency towards the S. sanguis and S. oralis strain than did the serous Par saliva. The SM/SL saliva-induced aggregation was observed after 30 min, at 60 min followed by the Par saliva-induced aggregation, and showed a 4-fold higher aggregation titer of 128 for S. sanguis, and an 8-fold higher titer of 516 for S. oralis. In contrast, the Par saliva showed a slightly higher aggregation activity than the SM/SL saliva towards S. rattus as judged by a twofold higher titer of 64. Morphologically, however, the SM/SL saliva-induced aggregation of S. rattus was far more pronounced as was also found for S. sanguis. Finally, the HWSM-induced aggregation showed a 4 to 8-fold higher titer than the originating salivary source, measuring 2048 for S. oralis and 128 for S. rattus. Moreover, no difference was observed in aggregation activity between the HWSM from whole saliva of a blood group O donor and the HWSM from SM/SL saliva of a blood group A donor. All the data point to an important, though not exclusive role of the human salivary mucin fraction in the saliva-induced aggregation of these strains.

The interaction of blood group reactive substances in saliva with bacteria was investigated by testing saliva from persons with different blood groups in a bacterial aggregation assay with Streptococcus rattus HG 59, originally S. rattus BHT. For blood group A, saliva from 10 persons out of 11 aggregated S. rattus and for blood group O, saliva from 10 persons out of 16 aggregated S. rattus. For blood group B, saliva from 6 persons out of 8 aggregated S. rattus weakly and the average aggregation activity of blood group B was much lower than for blood group A or O. In addition, saliva from 3 non-secretors did not aggregate S. rattus. The role of blood group antigens in bacterial aggregation was confirmed by inhibition studies with blood group specific sugars and various other sugars. GalNAc, specific for blood group A, inhibited bacterial aggregation by saliva whereas D-galactose, specific for blood group B, and D-fucose, specific for blood group O, did not. In addition, sialic acid, a major terminal sugar residue in mucins, also inhibited the bacterial aggregation. This study shows that the blood group and secretor status of a person may influence the interaction of saliva with bacteria in the oral cavity.

Culture filtrates of B. gingivalis and B. asaccharolyticus are cytotoxic for Vero cells. It is shown that the cytotoxic effect is due to the butyrate concentrations present in the culture filtrates of these strains. This cytotoxic effect proved to be reversible. Strains of the B. melaninogenicus subspecies intermedius and melaninogenicus did not produce butyrate and did not show cytotoxic activity towards Vero Cells. The significance of the production of toxic concentrations of butyrate for the etiology of especially periodontal diseases is discussed.

The virulence of B. gingivalis strain W83 was studied in an experimental animal model. Cells grown overnight, washed and resuspended in broth, were injected intradermally or subcutaneously in the back of guinea pigs, rats and mice. This strain proved to be very virulent, causing a severe phlegmonous abscess in guinea pigs. Also in mice, which are thought to be resistant to infections with black-pigmented Bacteroides strains, the same type of infection could be induced. Rats proved to be rather insensitive. The model presented can be used as a simple virulence test for these anaerobic bacteria.

The purpose of this study was to investigate the production of various enzymes by oral and non-oral black-pigmented Bacteroides species using chromogenic substrates. The 19 substrates present in the API ZYM system did not differentiate between B. melaninogenicus, B. denticola, B. loescheii and B. levii. The asaccharolytic black-pigmented Bacteroides species showed each species specific enzyme activity, however, differences were based on one enzyme only as far as B. asaccharolyticus and B. endodontalis are concerned. An extended number of 40 chromogenic substrates were tested in order to find more species specific enzyme. With a set of 20 substrates it appeared to be possible to discriminate between all species tested. The possibility to use enzymes for the identification of black-pigmented Bacteroides is discussed.

In a comparative study 351 strains of Micrococcus subgroups 1-3 (Baird-Parker, 1963, 1965, 1974) were classified according to the Kloos and Schleifer scheme (1975b). The results showed that Micrococcus subgroups 1, 2 and 3 are heterogeneous groups in the Kloos and Schleifer scheme. Strains belonging to Micrococcus subgroups 1 and 2 were mostly classified by the Kloos and Schleifer criteria as Staphylococcus hominis, Micrococcus subgroup 3 strains from the skin as S. cohnii, while Micrococcus subgroup 3 strains from urinary infections were classified mainly as S. saprophyticus. The correlation of novobiocin resistant S. saprophyticus biotype III (Baird-Parker, 1974) with S. saprophyticus (Kloos and Schleifer, 1975a, 1975b) when isolated from urine, was 80%. Although the Kloos and Schleifer scheme provides more information about biochemical characters, doubts are expressed about the validity of some of the species so delineated.

In order to gain insight into the relative importance of several virulence factors of Bacteroides gingivalis, 8 strains with a varying virulence were studied. The virulence of B. gingivalis was determined in a mouse model. Strains HG 66, HG 76 and HG 184 were very virulent causing phlegmonous abscesses with lesions and necrosis. The strains HG 405 and HG 462 caused phlegomonous abscesses with pus. Strains HG 91, HG 94 and HG 185 were less virulent and induced gravity abscesses. In vitro strains HG 66, HG 76 and HG 184 induced low amounts of chemiluminescence by polymorphonuclear leucocytes. All other strains including HG 405 and HG 462 caused a relatively high chemiluminescence. Most strains displayed a high sensitivity to the bactericidal activity of fresh serum except for the highly virulent strains HG 66, HG 76 and HG 184. No differences in extracellular proteolytic activity on Azocoll, production of volatile fatty acids and ammonia were found between the B. gingivalis strains studied. In conclusion, differences in virulence were shown within the species B. gingivalis; the relative importance of several virulence factors was investigated.

Colicin A-insensitive mutants of Citrobacter freundii were isolated and grouped into six phenotypic classes characterized by sensitivity, insensitivity or partial insensitivity to the bacteriocins S6, DF 13 and colicin A, and sensitivity or insensitivity to deoxycholate (DOC) and ampicillin. Mapping by the gradient-of-transmission method revealed the chromosomal regions in which the responsible genes are situated. Res-3 mapped near pur between pur and thr; Tol-5 mapped between aro and ilv and Tol-4 between gal and pur; Tol-1, Tol-2 and Tol-3 are situated close to gal. All the mutations that mapped near gal rendered the bacteria more sensitive to DOC and ampicillin. Complementation analysis with E. coli plasmids showed that the three phenotypic groups that map near gal were complemented by E. coli plasmids and fall into three complementation groups. Two of these are quivalent with the tol A and tol B genes in E. coli.

In almost all infections in the oral cavity, mixed populations of bacteria are present. However, recent evidence points to a certain specificity in these infections: Streptococcus mutans is related to caries and black-pigmented Bacteroides species are suspected pathogens in periodontal disease. Periodontal diseases, endodontic infections and submucous abscesses in the oral cavity are probably mixed infections in which anaerobic bacteria together with facultatives or other anaerobes are present. In experimental mixed anaerobic infections black-pigmented Bacteroides strains have been shown to play a key role. Little is known about the pathogenic synergy between the bacteria involved in mixed infections. Important mechanisms could be nutritional interrelationships and interactions with the host defense. Within the group of black-pigmented Bacteroides B. gingivalis seems to be the most virulent species. These bacteria possess a great number of virulence factors, which might be important in the pathogenesis of oral infections.

For comparison, human whole saliva-induced aggregation was studied by phase-contrast microscopy, spectrophotometry combined with macroscopic observations, and in microtiterplate assay under identical experimental conditions for Actinomyces viscosus HG 85 (T14-V) and HG 380 (T14-AV), Bacteroides gingivalis HG 66 (W 83), Streptococcus rattus HG 59 (BHT), and Streptococcus sanguis I HG 169. The entire process of formation, extension, and sedimentation of aggregates could merely be observed by the combination of these assays. The very first stages of aggregation could only be detected and quantitated by phase-contrast microscopy. Within 2 1/2 min. 50% of the A. viscosus, S. rattus, and S. sanguis cells were aggregated, denoted as T50. In microtiterplates, however, aggregates were observed in general only after sedimentation at 30–45 min of incubation, expressed as TA. For interpretation of the spectrophotometric curves, additional microscopic and macroscopic data were a prerequisite. The small decline in absorbance during the first 30–45 min (phase 1) corresponded to the formation and extension of non-sedimenting aggregates, whereas the subsequent pronounced fall in absorbance (phase 2) was caused by the massive sedimentation of aggregates. The moment of inflexion between both phases, TI, marked the onset of sedimentation of aggregates and corresponded very well with TA, at which time already 92–98% of the cells were aggregated as quantitated by microscopy. In conclusion, only by microscopy the formation and extension of aggregates could be observed within a few minutes and quantitated in terms of aggregation rate. From 30–45 min, merely the sedimentation of aggregates was visualized in microtiterplates, whereas the time course of the overall process was recorded indirectly by spectrophotometry.

E. coli strains were isolated from urine specimens of hospitalised patients with acute pyelonephritis, acute cystitis or asymptomatic bacteriuria (ABU), and tested for virulence in an experimental mouse model. Of 12 pyelonephritis-strains 11 were shown to be virulent and 1 avirulent; of 12 cystitis-strains 4 were virulent and 8 avirulent; of 12 ABU-strains 5 were virulent and 7 avirulent. It is concluded that, while no difference in virulence was found between cystitis- and ABU-strains, pyelonephritis-strains were more often virulent than cystitis- and ABU-strains. No associations could be shown between virulence of the isolated strains and the presence of antibody-coated bacteria in the urine. Common urinary O types were not more often virulent than other O types. No relationship was seen between virulence and the presence of K antigen or the presence of particular K types.

The virulence of beta-hemolytic and non-hemolytic strains of Streptococcus mutans was studied in neonatal mice by LD50 determinations after intracerebral injection of the bacteria. Although the differences in LD50 values are small the results may indicate that beta-hemolytic S. mutans strains (mean LD50 of 6.3 X 10(7) c.f.u.) were more virulent than non-hemolytic S. mutans strains (mean LD50 of 47.7 X 10(7) c.f.u.). The LD50 values of other viridans streptococci varied between 4.8 and 10(7) c.f.u. Strains of S. pyogenes and S. agalactiae were highly virulent in this animal model with LD50 values lower than 10(5) c.f.u.

A haemolytic E. coli strain is nephropathogenic for mice. Mutagenesis by transposon insertion resulted in mutants with altered haemolytic activity. Reduction or elimination of the haemolytic activity is accompanied with loss of virulence. It is shown that this loss of virulence is due to altered haemolytic activity and not caused by the transposon insertion itself.

E. coli strains were isolated from the urine of patients with acute cystitis in general practice and from the faeces of a comparable reference group of healthy individuals. These strains were serotyped and tested for virulence in an experimental mouse model. Of 30 cystitis-strains 18 were virulent, and of 30 faeces-strains 15 were virulent. It is concluded that the cystitis-strains were not more often virulent than the faeces-strains. O antigens commonly found among urinary E. coli isolates were present in 60% of the cystitis-strains and in 37% of the faeces-strains. K antigens commonly found in urinary E. coli strains were present in 33% of the cystitis-strains and in 12% of the faeces-strains. Neither the presence of common urinary O-antigens, nor the presence of common urinary K antigens could be associated with virulence of the isolated strains. However, it is suggested that certain O and K antigens (O2,O6, K23) may be associated with virulence for the urinary tract.

Twenty-seven oral strains of the genera Actinomyces (5), Bacteroides (3), and Streptococcus (19) were tested for aggregation by human whole saliva, as well as the effect of culture medium, Ca-ions, and bacteria concentration thereupon. Of the media tested, GF-broth gave rise to less interference by autoaggregation or higher aggregation titers than BHI and TSB, and was used throughout this study. In most cases, Ca-ions (1 mM) only enhanced the rate of induced aggregation, whereas raising the bacteria concentration increased the rate of both induced- and autoaggregation. The final titers, ranging from 1-64, were hardly affected by these parameters, except those of S. rattus HG 59 and S. mutans HG 199, which were respectively increased and decreased by Ca-ions. Saliva-induced aggregation was observed for 21 strains of A. viscosus, A. naeslundii, A. israelii, B. gingivalis, B. intermedius, S. cricetus, S. mutans, S. rattus, S. sanguis, and S. sobrinus, mostly within 15 min to 3 h. Seventeen of these strains also showed autoaggregation, usually well after the onset of induced aggregation. Any potential induced aggregation of B. gingivalis HG 91 was always masked by autoaggregation, as well as that of the S. mutans strains under a particular set of conditions. The aggregation rate and titer varied considerably in a mutually unrelated and strain-dependent way. These microtiterplate data were matched by the 5 spectrophotometric patterns observed for saliva-bacterial interaction, which moreover, gave the better differentiation between induced and autoaggregation. In conclusion, most strains tested can show rapid saliva-induced aggregation in a strain-dependent way, yet strongly affected by the experimental conditions and interference from autoaggregation.

Among colicin-A-tolerant mutants of Citrobacter freundii we characterized some as Tol-2 mutants. The Tol-2 mutation results in insensitivity to bacteriocin S6 and an enhanced sensitivity to deoxycholate (DOC), ethylenediaminetetraacetic acid (EDTA) and to ampicillin. The Tol-2 mutation was mapped close near gal and the gene order pro-tol-gal was established in crosses between C. freundii Hfr tol+ donor strains and C. freundii tol- acceptor strains. In these crosses a difference was observed in phenotypic expression of the pleiotropic properties of this tol+ gene. Expression of resistance to DOC is substantially slower than the expression of EDTA-resistance. This phenomenon may play a disturbing role in those studies on cell envelope mutants, in which resistance to DOC is used as a selected marker. The differences in expression of DOC-and EDTA-resistance are discussed.

Citrobacter freundii Hfr donor strains were isolated from a C. freundii strain harbouring a temperature-sensitive factor F ts 114 lac+, by selecting for integrative suppression of the ts 114 mutation. Three Hfr strains were characterized, which transfer their chromosomes in a linear and oriented order. The first strain transfers: O-aro+-ilv+-pur+-thr+-leu+-pro+, the second: O-ilv+-pur+-thr+-leu+-pro+ and the third: O-ilv+-aro+-nad+-his+-pro+. The whole chromosome is transferred into the recipient cell within about 145 minutes. From these results we concluded that the linkage map of C. freundii is circular.