NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Brogden KA, Guthmiller JM, editors. Polymicrobial Diseases. Washington (DC): ASM Press; 2002.

Cover of Polymicrobial Diseases

Polymicrobial Diseases.

Show details

Chapter 9Abscesses


Author Information

Abscesses that develop as a result of introduction of the normal endogenous flora into a normally sterile body site are often polymicrobial in nature. These flora can gain access to the sterile site by direct extension or secondary to laceration or perforation. Because of the uniqueness of the normal endogenous flora at the various body sites, the microbiology of such abscesses is generally predictable. This chapter describes the specific microbiology of polymicrobial abscesses that occur at various body sites. It also reviews the data that demonstrate the synergy between the aerobic and anaerobic components of these abscesses, and highlights the role of the bacterial capsule as a virulence factor that enhances the formation of an abscess.

The Role of Normal Flora in Polymicrobial Abscesses

In most mucus membranes, anaerobes outnumber aerobic and facultative bacteria in ratios ranging from 10:1 to 10,000:1, with anaerobic gram-negative bacilli (AGNB) microorganisms predominating (12, 44). Species of the Bacteroides fragilis group that colonize the gastrointestinal tract are usually isolated in intra-abdominal and rectal abscesses; pigmented Prevotella, Porphyromonas, and Fusobacterium spp. that colonize the oral cavity are present mainly in oral cavity abscesses, and Prevotella bivia and Prevotella disiens that predominate in the cervical canal are most often recovered in pelvic abscesses. The predominant aerobes and facultative organisms in abdominal and rectal abscesses are Enterobacteriaceae and staphylococci, and Neisseria gonorrhoeae are common in pelvic abscesses (Fig. 1; Table 1).

Figure 1. Distribution of organisms in abscesses, wounds, burns, and decubitus ulcers.

Figure 1

Distribution of organisms in abscesses, wounds, burns, and decubitus ulcers.

Table 1. Microbiologic characteristics of 676 cutaneous abscesses.

Table 1

Microbiologic characteristics of 676 cutaneous abscesses.

The bacterial flora of the gastrointestinal tract (GIT) is very dynamic, and changes in the flora influence the type and severity of post-perforation infection. The stomach and upper bowel flora contain 104 organisms or fewer/g, the lower ileum contains up to 108 organisms/g, and the colon contains up to 1011 organisms/g, most of which are anaerobes (46). The low number of organisms in the stomach is believed to be caused by the detrimental effect of the low pH of the stomach on the organisms ingested from the oropharynx. The contents of the gut slowly become alkaline at the lower intestine. This change, the effect of bile, and the decrease in oxygen tension in the lower intestine allow for selection of bile-resistant organisms and increase in the number of strict anaerobes. Numerous organisms in the upper GIT can, however, be found in patients with decreased stomach acidity or with a shorter GIT or anastomosis.

The variations in the number of bacteria in the GIT account for the differences that are observed in cultures of the peritoneal cavity after perforations. Three different isolates per specimen and about 107 organisms/g were recovered from perforation of the small intestine, while 26 different bacterial isolates and 1012 organisms/g were isolated from specimens of colonic perforation (71). This high load of microorganisms is believed to account for the higher frequency (50%) of the infections that follow colonic injury, compared with that after chest injuries (18%), found by Dellinger et al. (43). The higher number of organisms in the distal part of the colon also explains why infection developed in 45% of patients with descending-colon injuries, compared with about 13% in the other sites of the colon.

The Microbiology of Abscesses of Endogenous Origin

Gram-positive anaerobic cocci are normal skin inhabitants and part of the normal fecal flora (44). These cocci are also isolated from intra-abdominal abscesses (6, 1921, 2325). They were isolated as frequently as AGNB from abscesses of the perineal region, and they were also frequently isolated from nonperineal cutaneous abscesses.

Organisms belonging to the B. fragilis group, which predominate in the feces, were cultured most frequently from abscesses of the perirectal area (46). Prevotella melaninogenica, which occurs in stool and in the oral cavity (44), also was recovered from this site and from the head.

The microbiology of intra-abdominal abscesses that develop following perforation of viscera is made of similar patterns of organisms and is made up of the gastrointestinal flora at the level of the perforation (Table 2). The pre-dominant anaerobic bacteria are the B. fragilis group, Peptostreptococcus spp., and Clostridium spp., while the most commonly isolated aerobic and facultative bacteria are Enterobacteriaceae and group D enterococci. These organisms were recovered from a variety of intra-abdominal (6), retroperitoneal (20), visceral (44, 46) [e.g., pancreatic (21), hepatic, and splenic (23)], and perirectal abscesses (25) [after diverticulitis rupture (19) and subphrenic (24)]. Similarity also exists in the microbiology of pelvic, vulvo-vaginal (27), and prostatic (5) abscesses that originate from the rectal and cervical flora (44, 46). The predominant anaerobic bacteria are P. bivia, P. disiens, and peptostreptococci, while the common aerobic and facultative bacteria include Enterobacteriaceae, N. gonorrhoeae, and group B streptococci (Table 2).

Table 2. Predominant isolates in abscesses at various body sites.

Table 2

Predominant isolates in abscesses at various body sites.

The microbiology of dental, orofacial, and neck abscesses is mainly made of oral flora organisms (Table 3) (9). These include periton-sillar (29), retropharyngeal (10), parotic (28), and cervical (22) lymph glands. The main anaerobes are pigmented Prevotella, Porphyromonas, Fusobacterium, and Peptostreptococcus spp. The most commonly isolated aerobes and facultative bacteria are Streptococcus pyogenes and Staphylococcus aureus.

Table 3. Predominant isolates from abscesses.

Table 3

Predominant isolates from abscesses.

The microbiology of skin and soft tissue abscesses is also related to their location (Fig. 1) (9, 17, 18, 26, 55). S. pyogenes and S. aureus that colonize the skin over the entire body can be recovered at all locations. The location of the abscess is of paramount importance in the emergence of the other organism(s) that may also be involved in the infection. Under appropriate conditions of lowered tissue resistance, almost any of the common bacteria can initiate an infectious process. Cultures from lesions frequently contain several bacterial species; as might be expected, the organisms found most frequently are the "normal flora" of these regions (Table 1).

Aspirates from abscesses of the perineal and oral regions tend to yield organisms found in stool or mouth flora (17, 18, 55). Conversely, pus obtained from abscesses in areas remote from the rectum or mouth contain primarily constituents of the microflora indigenous to the skin such as S. pyogenes and S. aureus. Multiple anaerobic organisms usually are recovered from the perineal region, whereas only about one aerobe per abscess is present at other sites (18) (Table 1). Anaerobes also are more often recovered alone, without aerobes, from the perineal area. Mixed aerobic and anaerobic infections are more prevalent in the perirectal, head, finger, and nail bed areas. The similarities in the rates of isolation of mixed aerobic and anaerobic flora and the high rate of recovery of anaerobes in these areas are of particular interest. These similarities and the high rate of recovery can be caused by the introduction of mouth flora, which is predominantly anaerobic, onto the fingers by sucking or nail biting, which are common activities among children. This is parallel to the acquisition of infection following human bites and clenched fist injuries in which anaerobic mouth flora was the source of most bacterial isolates (68).

The polymicrobial nature of abdominal, pelvic, and skin and soft tissue (proximal to the oral or rectal areas) abscesses is apparent in most patients, where the number of isolates in an infectious site varies between 2 and 6 (12, 44) (Tables 1 and 3). The average number of isolates is 3.6 in skin and soft tissue infections (2.6 anaerobes and 1.0 aerobe) per specimen (17, 18, 55), 5 in intra-abdominal infection (3.0 anaerobes and 2.0 aerobes) per specimen (6, 1921, 2325), and 4 in pelvic infections (2.8 anaerobes and 1.2 aerobes) per specimen (5, 27, 69). Polymicrobic infections are more pathogenic for experimental animals than those involving single organisms (1). The number of isolates in these polymicrobial abscesses varies from 2 to 6 (Tables 1 and 3), and generally is higher when reported in studies in which stricter methods for collection, transportation, and cultivation of anaerobic organisms are used (2, 10, 29, 68).

Virulence of Anaerobic Bacteria

Although more than 400 bacterial species colonize in the colon, and more than 200 reside in oral cavities, the average number of bacterial species in infections associated with colonic perforation is five (44). The dominant anaerobic bacteria in this type of disease include the B. fragilis group, Prevotella and Porphyromonas spp. (previously called the Bacteroides melaninogenicus group), Fusobacterium nucleatum, Clostridium perfringens, and Peptostreptococcus spp. Thus, from the multiple anaerobic bacteria that make up the normal flora, only a few are common in the septic process; it is likely that their virulence is an important factor in their selection. The ability to produce a capsule (encapsulation) has been observed in all these anaerobic bacteria and may serve as an important virulence factor (48). Of all the anaerobes, B. fragilis group isolates are the most frequently encountered organisms in intra-abdominal abscesses or anaerobic bacteremia. Members of the B. fragilis group have several virulence factors and resist β-lactam antibiotics through production of the enzyme β-lactamase (12, 44), possession of a capsule that inhibits phagocytosis (70), and the production of other enzymes and metabolic by-products. Succinic acid is an important metabolic by-product that can reduce poly-morphonuclear migration (66). In addition to a capsule, anaerobic bacteria possess other important virulence factors. These include the production of superoxide dismutase and catalase, immunoglobulin proteases, coagulation-promoting and -spreading factors (such as hyaluronidase, collagenase, and fibrinolysin), and adherence factors (44, 48). Other factors that enhance the virulence of anaerobes include mucosal damage, oxidation-reduction potential drop, and the presence of hemoglobin or blood in an infected site.

Encapsulation of Anaerobic Bacteria in Mixed Infection and Abscesses

Encapsulation of anaerobic bacteria defined by production of extracellular polysaccharide, termed glycocalyx (42), has been recognized as an important virulence factor. Several studies demonstrated the pathogenicity of encapsulated anaerobes and their ability to induce abscesses in experimental animals even when inoculated alone. Onderdonk et al. (60) correlated the formation of intra-abdominal abscesses in mice and rats by B. fragilis strains with the presence of capsule. Encapsulated B. fragilis strains or purified capsular polysaccharide alone induced abscesses, whereas nonencapsulated strains seldom caused abscesses unless they were combined with an aerobic organism. Simon et al. (67) showed that encapsulated Bacteroides strains resisted neutrophil-mediated killing, compared with nonencapsulated strains. Encapsulated B. fragilis strains adhered better to rat mesothelium than did nonencapsulated strains (62).

The susceptibility of pathogenic bacteria to phagocytosis and killing by polymorphonuclear leukocytes and macrophages is of major importance in determining the outcome of the host-pathogen interaction. Bjornson et al. (3) demonstrated phagocytosis and killing of B. fragilis by human leukocytes in vitro. Phagocytosis of B. fragilis in the presence of serum occurred in aerobic and anaerobic conditions. Ingham et al. (49) investigated the effect of Bacteroides spp. on the phagocytic killing of facultative species. Killing of B. fragilis and Proteus mirabilis in mixtures in vitro was impaired when the concentration of B. fragilis was greater than 1 × 107 CFU/ml in the phagocytic system. Tofte et al. (70) and Jones and Gemmell (50) reported that both phagocytic uptake and killing of facultative species were impaired at high concentrations of encapsulated Bacteroides. This inhibitory effect of Bacteroides could be related to the effect of capsule on phagocytosis.

The presence of capsule in B. fragilis was shown to provide the organism with growth advantage in vivo over nonencapsulated isolates (65). Furthermore, encapsulated strains survived better in vitro than nonencapsulated variants when they were grown in an aerobic environment (65), and suppression of bacterial growth by the antimicrobial clindamycin was reduced in an encapsulated strain, as compared with a nonencapsulated one (30). Another recently described mechanism of protection independent of encapsulation is the inhibition of polymorphonuclear migration due to the production of succinic acid by nonencapsulated Bacteroides spp. (66).

The importance of the capsular polysaccharide of B. fragilis as an immunogen was demonstrated when antibodies against it protected animals from early bacteriemia (52). However, prevention of the formation of intra-abdominal abscess by this organism was found to be mediated by T cells (61).

The presence of a polysaccharide capsule, as defined by the electron microscopic visualization of a Ruthenium red-stained structure external to the cell wall, has been documented for all members of the B. fragilis group (4, 41), and pigmented Prevotella and Porphyromonas (31), Fusobacterium spp. (37), and anaerobic cocci (40). However, because Ruthenium red stains some acidic polysaccharides as well as some lipopolysaccharides, some investigators consider immunochemical methods to be more reliable for defining a true polysaccharide capsule (51).

The capsule of B. fragilis was studied more than any other anaerobic bacterium and was found to consist of two chemically distinct polysaccharides (A and B). Polysaccharide A is neutral at pH 7.3, but negatively charged at pH 8.6, and it contains predominantly galactose. Polysaccharide B is negatively charged at both pH 7.3 and 8.6 and contains fructose, galactose, quinovosamine, galacturonic acid, and galactosamine. These polysaccharides can give a complex multiprecipitin profile when reacting with homologous antiserum in an immunoelectrophoretic assay. The dual polysaccharide motif is a common feature of B. fragilis strains (63) that may result in different affinities for capsular detection stains such as Ruthenium red. This could also explain the diversity seen in electron microscopy analysis of B. fragilis cells (15, 64).

Encapsulated Anaerobic Bacteria in Clinical Infections and Abscesses

Aspirates obtained from infections next to mucous membrane surfaces generally contain a complex bacterial population consisting of several species (12, 44). Although anaerobes are components of mixed infections, their role and relative importance in the disease process are not always considered.

In an attempt to define the important pathogens among the isolates recovered from clinical specimens, we studied the virulence and importance of encapsulated bacterial isolates recovered from 13 clinical abscesses (35). This was done by injecting each of the 35 isolates (30 anaerobes and 5 aerobes) subcutaneously (s.c.) into mice alone or in all possible combinations with the other isolates recovered from the same abscess. We then observed their ability to induce and/or survive in a subcutaneous abscess. Sixteen of the isolates were encapsulated; 15 of them were able to cause abscesses by themselves and were recovered from the abscesses even when inoculated alone. The other organisms, which were not encapsulated, were not able to induce abscesses when inoculated alone. However, some were able to survive when injected with encapsulated strains. Therefore, the possession of a capsule by an organism was associated with increased virulence, compared with the same organism's nonencapsulated counterparts, and might have allowed some of the other accompanying organisms to survive. We found this phenomenon to occur in Bacteroides spp., Prevotella spp., anaerobic gram-positive cocci, Clostridium spp., and Escherichia coli. Detection of a capsule in a clinical isolate may therefore suggest a pathogenic role of the organism in the infection.

Three studies support the importance of encapsulated anaerobic organisms in respiratory infections (13, 32, 34). The presence of encapsulated and abscess-forming organisms that belong to the pigmented Prevotella and Porphyromonas spp. in 25 children with acute tonsillitis and 23 children without tonsillar inflammation (control) was investigated (32). Encapsulated pigmented Prevotella and Porphyromonas were found in 23 of 25 children with acute tonsillitis, compared with 5 of 23 controls (P < 0.001). Subcutaneous inoculation into mice of the Prevotella and Porphyromonas strains that had been isolated from patients with tonsillitis produced abscesses in 17 of 25 instances, compared with 9 of 23 controls (P < 0.05). These findings suggest a possible pathogenic role for pigmented Prevotella and Porphyromonas spp. in acute tonsillar infection, and also suggest the importance of encapsulation in the pathogenesis of the infection.

In another study (13), the presence of encapsulated AGNB (pigmented Prevotella and Porphyromonas spp. and Bacteroides spp.) and anaerobic gram-positive cocci was investigated in 182 patients with chronic orofacial infections and in the pharynx of 26 individuals without inflammation. Forty-nine of the patients had chronic otitis media, 45 had cervical lymphadenitis, 37 had chronic sinusitis, 24 had chronic mastoiditis, 10 had peritonsillar abscesses, and 12 had periodontal abscesses. Of the 216 isolates of pigmented Prevotella and Porphyromonas, B. fragilis group, and anaerobic cocci, 170 (79%) were found to be encapsulated in patients with chronic infections, compared with only 34 of 96 (35%) controls (P < 0.001).

The presence of encapsulated and piliated AGNB (mostly B. fragilis group and pigmented Prevotella and Porphyromonas) was investigated in isolates from blood, abscesses, and normal flora (34). Of the strains of AGNB isolated, 45 of 54 (83%) recovered from blood and 31 of 40 (78%) found in abscesses were encapsulated. In contrast, only 7 of 71 (10%) similar strains isolated from the feces or pharynx of healthy persons were encapsulated (P < 0.001). Pili were observed in 3 of 54 (6%) of strains isolated from blood, 30 of 40 (75%) of those recovered from abscesses (P < 0.001), and 49 of 71 (69%) of those found in normal flora (P < 0.001) (Fig. 2; shows only B. fragilis group). The pre-dominance of encapsulated forms in all strains of AGNB in blood and in abscesses suggests an increased virulence of encapsulated forms compared with nonencapsulated isolates. In contrast, the presence of pili in AGNB recovered mostly from abscesses and normal flora suggests that this structure may play a role in the ability of these organisms to adhere to mucous membranes and may interfere with their ability to spread systematically. These findings illustrate the morphological differences that may be observed in AGNB from various anatomical sites.

Figure 2. Dynamics of pili and capsule of B.

Figure 2

Dynamics of pili and capsule of B. fragilis group.

The predominance of encapsulated Bacteroides, Prevotella, and Porphyromonas spp. recovered from blood and abscesses compared with their rate of encapsulation in the normal flora of the pharynx and feces suggests an increased virulence of these strains as compared to nonencapsulated strains. In contrast to the emergence of encapsulated Bacteroides spp. in blood and abscesses, the presence of pili was less frequent in such strains recovered from blood. The rate of piliated strains was high among those recovered from abscesses.

Since most Bacteroides, Prevotella, and Porphyromonas spp. recovered from infected sites probably originate from the predominantly nonencapsulated endogenous flora of mucous membranes, they may express their capsules only during the inflammatory process. The frequent recovery of encapsulated AGNB in such conditions illustrates their increased virulence as compared with their nonencapsulated counterparts.

Complete eradication of experimental Bacteroides infection by means of metronidazole was not achieved when these organisms were encapsulated (11). Once the organisms become encapsulated, eradication of Bacteroides infection becomes difficult. Treatment of infections involving nonencapsulated Bacteroides spp., however, was more efficacious. Early treatment of anaerobic infections may therefore prevent the emergence of encapsulated AGNB, and subsequent bacteremia.

The recovery of a larger number of encapsulated anaerobic organisms from orofacial infections, abscesses, and blood of patients provides support for the potential pathogenic role of encapsulated organisms. Early and vigorous antimicrobial therapy, directed at both aerobic and anaerobic bacteria present in these mixed infections, may abort the infection before the emergence of encapsulated strains that contribute to the chronicity of the infection.

Capsule Formation in Experimental Abscesses

The ability of the aerobic component in mixed infections to enhance the appearance of encapsulated anaerobic bacteria in these infections was studied in a s.c. abscess model in mice. The anaerobic bacteria with which they were inoculated were those commonly recovered in mixed infections.

Pigmented Prevotella and Porphyromonas spp. (31), P. bivia (14), B. fragilis group (16), and anaerobic and facultative gram-positive cocci (AFGPC) (40) did not induce abscess when isolates that contained only a few encapsulated organisms (< 1%) were inoculated. However, when these relatively nonencapsulated isolates were inoculated, mixed with abscess-forming viable or nonviable bacteria ("helpers"), the Bacteroides, Prevotella, Porphyromonas, and AFGPC survived in the abscess and became heavily encapsulated (> 50% of organisms had a capsule). Thereafter, these heavily encapsulated anaerobic isolates were able to induce abscesses when injected alone (Fig. 3). Of interest is the observed appearance of pili along with encapsulation in the B. fragilis group after coinoculation with Klebsiella pneumoniae (16).

Figure 3. Encapsulation cycle of B.

Figure 3

Encapsulation cycle of B. fragilis group after passage in mice. Helper is viable bacteria or formolized bacteria or capsular material.

Most of the helper strains were encapsulated; however, several of the strains were not encapsulated, but they were able to induce abscesses when inoculated alone. The helper organisms used in conjunction with pigmented Prevotella and Porphyromonas spp. and AFGPC were S. aureus, S. pyogenes, Haemophilus influenzae, Pseudomonas aeruginosa, E. coli, K. pneumoniae, and Bacteroides spp. (31, 40). For the B. fragilis group, these organisms were E. coli, K. pneumoniae, S. aureus, S. pyogenes, and Enterococcus spp. (16). N. gonorrhoeae was chosen as a helper for B. fragilis, and Prevotella and Porphyromonas spp. (14). Of interest is the observed inability of N. gonorrhoeae strains to survive in intra-abdominal abscesses and also their disappearance from s.c. abscesses within 5 days of inoculation with Bacteroides spp. and P. bivia (14).

The virulence of Fusobacterium spp. was also associated with the presence of a capsule. Only encapsulated strains of F. nucleatum, Fusobacterium necrophorum, and Fusobacterium varium were able to induce abscesses when inoculated alone (37). However, after passage in animals of nonencapsulated strains, none of these organisms acquired a capsule.

The presence of a thick granular cell wall (300 to 360 Å) before animal passage was associated with virulence of Clostridium spp. (38). Such structure was observed before inoculation into animals, only in C. perfringens and Clostridium butyricum, the only organisms capable of inducing an s.c. abscess when inoculated alone. This structure was observed in other Clostridium spp. only after their coinoculation with encapsulated Bacteroides spp. or K. pneumoniae. However, other undetermined factors may also contribute to the induction of an abscess, because most isolates of Clostridium difficile were not able to produce an abscess even though they possessed a thick wall.

The selection of encapsulated Bacteroides spp. and AFGPC with the assistance of other encapsulated or nonencapsulated but abscess-forming aerobic or anaerobic organisms may explain the conversion into pathogens of non-pathogenic organisms that are part of the normal host flora or are concomitant pathogens. Although such a phenomenon was not observed in Fusobacterium spp., the presence of a capsule in these organisms was a prerequisite for induction of s.c. abscesses. Some Clostridium spp. also manifested cell wall changes after animal passage that could be associated with increased virulence. Although the exact nature and chemical composition of the capsule or external cell wall may be different in each of the anaerobic species studied, the changes that were observed tended to follow similar patterns.

Significance of Anaerobic Bacteria in Abscesses Mixed with Other Flora

Although anaerobic bacteria often are recovered mixed with other aerobic and facultative flora, their exact role in these infections and their relative contribution to the pathogenic process are unknown. The relative importance of the organisms present in an abscess caused by two bacteria (an aerobe and an anaerobe) and the effect of encapsulation on that relationship were determined by comparing the abscess sizes in (i) mice treated with antibiotics directed against one or both organisms and (ii) nontreated animals (3639).

As judged by selective antimicrobial therapy, the possession of a capsule in most mixed infections involving AGNB generally made these organisms more important than their aerobic counterparts. In almost all instances, the aerobic counterparts in the infection were more important than nonencapsulated AGNB (39). Encapsulated members of the pigmented Prevotella and Porphyromonas spp. were almost always more important in mixed infections than their aerobic counterparts (S. pyogenes, Streptococcus pneumoniae, K. pneumoniae, H. influenzae, and S. aureus). Encapsulated B. fragilis group organisms were more important than or as important as E. coli and enterococci, and less important than S. aureus, S. pyogenes, and K. pneumoniae.

In contrast to Bacteroides spp., encapsulated AFGPC were more often found to be less important than their aerobic counterparts (36). Clostridium and Fusobacterium spp. were less or equally important to enteric gram-negative rods (37, 38). Although Fusobacterium, AFGPC, and Clostridium spp. were generally equal to or less important than their aerobic counterpart, variations in the relationship existed. However, as determined by the abscess size, most of the anaerobic organisms enhanced mixed infection.

Synergy between Anaerobic and Aerobic or Facultative Anaerobic Bacteria in Abscesses

Several studies documented the synergistic effect of mixtures of aerobic and anaerobic bacteria in experimental infections. Altermeier (1) demonstrated the pathogenicity of bacteria isolates recovered from peritoneal cultures after appendiceal rupture. Pure cultures of individual isolates were relatively innocuous when implanted s.c. in animals, but combinations of facultative and anaerobic strains showed increased virulence. Similar observations were reported by Meleney et al. (56) and Hite et al. (47).

We have evaluated the synergistic potentials between aerobic and anaerobic bacteria usually recovered in mixed infections (33). Each bacterium was inoculated s.c. into mice alone or mixed with another organism, and synergistic effects were determined by observing abscess formation and animal mortality. The tested bacteria included encapsulated Bacteroides spp., Prevotella spp., Porphyromonas spp., Fusobacterium spp., Clostridium spp., and anaerobic cocci. Facultative and anaerobic bacteria included S. aureus, P. aeruginosa, E. coli, K. pneumoniae, and P. mirabilis. In many combinations, the anaerobes significantly enhanced the virulence of each of the five aerobes (Tables 4 and 5). The most virulent combinations were between P. aeruginosa or S. aureus and anaerobic cocci or AGNB.

Table 4. Synergy between anaerobic and aerobic and facultative bacteria.

Table 4

Synergy between anaerobic and aerobic and facultative bacteria.

Table 5. Synergy between anaerobic bacteria.

Table 5

Synergy between anaerobic bacteria.

Enhancement of the growth of each bacterial component in mixed infection was evaluated by studying the relative growth of each bacterial component. This was done by comparing (i) the growth of each organism in an abscess when present with another organism with (ii) the growth of those bacteria when inoculated alone (7, 8, 38, 40).

S. pyogenes, E. coli, S. aureus, K. pneumoniae, and P. aeruginosa were enhanced by B. fragilis, P. melaninogenica (7, 8), Peptostreptococcus spp. (7), Fusobacterium spp. (37), and Clostridium spp. (38) except C. difficile. Although mutual enhancement of growth of both aerobic and anaerobic bacteria was noticed, the number of aerobic and facultative bacteria was increased many times more than their anaerobic counterparts. Encapsulated Bacteroides spp. were able to enhance the growth of aerobic and facultative anaerobic bacteria more than nonencapsulated organisms (Table 6) (7, 11). Exceptions to the mutual enhancement were noticed in combinations of organisms that generally are not recovered together in mixed infections, such as enterococci and P. melaninogenica. The observations above suggest that the aerobic and facultative bacteria benefit even more than do the anaerobes from their symbiosis.

Table 6. Average numbers of encapsulated and nonencapsulated Bacteroides spp. in mixed abscesses with E. coli.

Table 6

Average numbers of encapsulated and nonencapsulated Bacteroides spp. in mixed abscesses with E. coli.

The mutual enhancement between aerobic and facultative organisms and B. fragilis group was apparent in numerous combinations (8). An increase in the number of aerobic and facultative bacteria when combined with three members of the B. fragilis group was illustrated in almost all instances except for the one between H. influenzae and B. fragilis group (Fig. 4 and 5; Table 7). However, this is not surprising since this combination is rarely seen in clinical infections.

Figure 4. Number of B.

Figure 4

Number of B. fragilis and E. coli (log10 CFU) in s.c. abscesses induced by single and combined bacteria in mice. A mutual significant enhancement of growth of both E. coli and B. fragilis (B. frag) was noted in mixed infection compared with the number (more...)

Figure 5. Change in number of aerobes when mixed in an abscess with B.

Figure 5

Change in number of aerobes when mixed in an abscess with B. fragilis. The increase in number of aerobic or facultative bacteria is illustrated by this bar graph (8).

Table 7. Average numbers of facultative and aerobic bacteria in subcutaneous abscesses induced by each organism alone combined with B. fragilis group.

Table 7

Average numbers of facultative and aerobic bacteria in subcutaneous abscesses induced by each organism alone combined with B. fragilis group.

Mechanisms of Synergy

Several hypotheses have been proposed to explain microbial synergy in mixed infection (55). When this phenomenon occurs in mixtures of aerobic and anaerobic flora, it may be due to protection from phagocytosis and intra-cellular killing (49), production of essential growth factors (53), and lowering of oxidation-reduction potentials in host tissues (53). Obligate anaerobes can interfere with the phagocytosis and killing of aerobic bacteria (49). The ability of human polymorphonuclear leukocytes to phagocytose and kill P. mirabilis was impaired in vitro when the human serum used to opsonize the target bacterium was pre-treated with live or dead organisms of various AGNB (50). Porphyromonas gingivalis cells or supernatant culture fluid was shown to possess the greatest inhibitory effect among the Bacteroides spp. (59). Supernatants of cultures of B. fragilis group, pigmented Prevotella and Porphyromonas, and P. gingivalis were capable of inhibiting the chemotaxis of leukocytes to the chemotactic factors of P. mirabilis (58).

Bacteria may also provide nutrients for each other. Klebsiella produces succinate, which supports Porphyromonas asaccharolytica (54), and oral diphtheroids produced vitamin K1, which is a growth factor for P. melaninogenica (45).

Another possible mechanism that explains the synergistic effect of aerobic-anaerobic combinations is the lowering of local oxygen concentrations and the oxidation-reduction potential by the aerobic bacteria. The resultant physical conditions are appropriate for replication and invasion by the anaerobic component of the infection. Such environmental factors are known to be critical for anaerobic growth in vitro and may apply with equal relevance to in vivo experimental animal studies. Mergenhagen et al. noted that the infecting dose of anaerobic cocci was significantly lowered when the inoculum was supplemented with chemical reducing agents (57). A similar effect may be produced by facultative bacteria, which may provide the proper conditions for establishing an anaerobic infection at a previously well-oxygenated site.

Demonstration of the synergistic potentials of anaerobic bacteria (such as the B. fragilis group, Porphyromonas and Prevotella spp., Fusobacterium spp., Clostridium spp., and anaerobic cocci) mixed with various aerobic and anaerobic bacteria further indicates their pathogenic role. Further studies are needed to investigate the exact mechanisms by which such synergy occurs, and the mode by which capsular material enhances it.


Altermeier W. A. The pathogenicity of the bacteria of appendicitis. Surgery. 1942;11:374–378.
Bennion R. S., Baron E. J., Thompson, Jr J. E., Downes J., Summanen P., Talan D. A., Finegold S. M. The bacteriology of gangrenous and perforated appendicitis-revisited. Ann. Surg. 1990;211:165–171. [PMC free article: PMC1357960] [PubMed: 2405791]
Bjornson A. B., Altemeier W. A., Bjornson H. S. Comparison of the in vitro bactericidal activity of human serum and leukocytes against Bacteroides fragilis and Fusoacterium mortiferum in aerobic and anaerobic environments. Infect. Immun. 1976;14:843–847. [PMC free article: PMC420963] [PubMed: 965101]
Bjornson A. B., Bjornson H. S., Ashraf M., Lang T. J. Quantitative variability in requirements for opsonization of strains within the Bacteroides fragilis group. J. Infect. Dis. 1983;148:667–675. [PubMed: 6631060]
Brook I. Anaerobic bacteria in suppurative genitourinary infections. J. Urol. 1989;141:889– 893. [PubMed: 2926882]
Brook I. A 12 year study of aerobic and anaerobic bacteria in intra-abdominal and post surgical abdominal wound infections. Surg. Gynecol. Obstet. 1989;169:387–392. [PubMed: 2683149]
Brook I. Enhancement of growth of aerobic, anaerobic, and facultative bacteria in mixed infections with anaerobic and facultative gram-positive cocci. J. Surg. Res. 1988;45:222–227. [PubMed: 3043109]
Brook I. Enhancement of growth of aerobic and facultative bacteria in mixed infections with Bacteroides sp. Infect. Immun. 1985;50:929–931. [PMC free article: PMC261172] [PubMed: 4066037]
Brook I. Microbiology of abscesses of the head and neck in children. Ann. Otol. Rhinol. Laryngol. 1987;96:429–433. [PubMed: 3619288]
Brook I. Microbiology of retropharyngeal abscesses in children. Am. J. Dis. Child. 1987;141:202–204. [PubMed: 3812388]
Brook I. Pathogenicity of encapsulated and non-encapsulated members of Bacteroides fragilis and melaninogenicus groups in mixed infection with Escherichia coli and Streptococcus pyogenes. J. Med. Microbiol. 1988;27:191–198. [PubMed: 2903934]
Brook, I. 2002. Pediatric Anaerobic Infection: Diagnosis and Management. Marcel Dekker, New York, N.Y.
Brook I. Recovery of encapsulated anaerobic bacteria from orofacial abscesses. J. Med. Microbiol. 1986;22:171–176. [PubMed: 3462395]
Brook I. The effect of encapsulation on the pathogenicity of mixed infection of Neisseria gonorrhoea and Bacteroides spp. Am. J. Obstet. Gynecol. 1986;155:421–428.
Brook I. The role of encapsulated anaerobic bacteria in synergistic infections. FEMS Microbiol. Rev. 1994;13:65–74. [PubMed: 8117468]
Brook I., Coolbaugh J. C., Walker R. I. Pathogenicity of piliated and encapsulated Bacteroides fragilis. Eur. J. Clin. Microbiol. 1984;3:207–209. [PubMed: 6147247]
Brook I., Finegold S. M. Aerobic and anaerobic bacteriology of cutaneous abscesses in children. Pediatrics. 1981;67:891–895. [PubMed: 7232053]
Brook I., Frazier E. H. Aerobic and anaerobic bacteriology of wounds and cutaneous abscesses. Arch. Surg. 1990;125:1445–1551. [PubMed: 2241555]
Brook I., Frazier E. H. Aerobic and anaerobic microbiology in intra-abdominal infections associated with diverticulitis. J. Med. Microbiol. 2000;49:827–830. [PubMed: 10966232]
Brook I., Frazier E. H. Aerobic and anaerobic microbiology of retroperitoneal abscesses. Clin. Infect. Dis. 1998;26:938–941. [PubMed: 9564479]
Brook I., Frazier E. G. Microbiological analysis of pancreatic abscess. Clin. Infect. Dis. 1996;22:384–385. [PubMed: 8838210]
Brook I., Frazier E. H. Microbiology of cervical lymphadenitis in adults. Acta Otolaryngol. 1998;118:443–446. [PubMed: 9655225]
Brook I., Frazier E. H. Microbiology of liver and spleen abscesses. J. Med. Microbiol. 1998;47:1075–1080. [PubMed: 9856643]
Brook I., Frazier E. H. Microbiology of subphrenic abscesses: a 14-year experience. Am. Surg. 1999;65:1049–1053. [PubMed: 10551755]
Brook I., Frazier E. H. The aerobic and anaerobic bacteriology of perirectal abscesses. J. Clin. Microbiol. 1997;35:2974–2976. [PMC free article: PMC230099] [PubMed: 9350771]
Brook I., Frazier E. H., Gher M. E. Aerobic and anaerobic microbiology of periapical abscess. Oral Microbiol. Immunol. 1991;6:123–125. [PubMed: 1945488]
Brook I., Frazier E. H., Thomas R. L. Aerobic and anaerobic microbiologic factors and recovery of beta-lactamase-producing bacteria from obstetric and gynecologic infection. Surg. Gynecol. Obstet. 1991;172:138–144. [PubMed: 1989117]
Brook I., Frazier E. H., Thompson D. H. Aerobic and anaerobic microbiology of acute suppurative parotitis. Laryngoscope. 1991;101:170–172. [PubMed: 1992268]
Brook I., Frazier E. H., Thompson D. H. Aerobic and anaerobic microbiology of peritonsillar abscess. Laryngoscope. 1991;101:289–292. [PubMed: 2000017]
Brook I., Gillmore J. D. Increased resistance of encapsulated Bacteroides fragilis to clindamycin. Chemotherapy. 1994;40:16–20. [PubMed: 8306810]
Brook I., Gillmore J. D., Coolbaugh J. C., Walker R. I. Pathogenicity of encapsulated Bacteroides melaninogenicus group, Bacteroides oralis, and Bacteroides ruminicola in abscesses in mice. J. Infect. 1983;7:281–286. [PubMed: 6141205]
Brook I., Gober A. E. Bacteroides melaninogenicus, its recovery from tonsils of children with acute tonsillitis. Arch. Otolaryngol. 1983;109:818–820. [PubMed: 6139104]
Brook I., Hunter V., Walker R. I. Synergistic effects of anaerobic cocci, Bacteroides, Clostridia, Fusobacteria, and aerobic bacteria on mouse mortality and induction of subcutaneous abscess. J. Infect. Dis. 1984;149:924–928. [PubMed: 6429254]
Brook I., Myhal L. A., Dorsey C. H. Encapsulation and pilus formation of Bacteroides spp. in formal flora abscesses and blood. J. Infect. 1992;25:251–257. [PubMed: 1361937]
Brook I., Walker R. I. Infectivity of organisms recovered from polymicrobial abscesses. Infect. Immun. 1983;42:986–989. [PMC free article: PMC264396] [PubMed: 6642674]
Brook I., Walker R. I. Pathogenicity of anaerobic gram-positive cocci. Infect. Immun. 1984;45:320–324. [PMC free article: PMC263223] [PubMed: 6746091]
Brook I., Walker R. I. Pathogenicity of Fusobacterium species. J. Med. Microb. 1986;21:93–100. [PubMed: 2936888]
Brook I., Walker R. I. Pathogenicity of some Clostridium species with other bacteria in mixed infection. J. Infect. 1986;13:245–253. [PubMed: 2878957]
Brook I., Walker R. I. Significance of encapsulated Bacteroides melaninogenicus and Bacteroides fragilis groups in mixed infections. Infect. Immun. 1984;44:12–15. [PMC free article: PMC263451] [PubMed: 6142862]
Brook I., Walker R. I. The role of encapsulation in the pathogenicity of anaerobic gram-positive cocci. Can. J. Microbiol. 1985;31:176– 180. [PubMed: 3986716]
Burt S., Meldrum S., Woods D. R., Jones D. T. Colonial variation capsule formation and bacteriophage resistance in Bacteroides thetaiotaomicron. Appl. Environ. Microbiol. 1978;35:439–443. [PMC free article: PMC242850] [PubMed: 637541]
Costerton J. W., Irvin R. T., Cheng K. J. The role of bacterial surface structures in pathogenesis. Crit. Rev. Microbiol. 1981;8:303–338. [PubMed: 6115736]
Dellinger E. P., Oreskovich M. R., Wertz M. J., Hamasaki V., Lennard E. S. Risk of infection following laparotomy for penetrating abdominal injury. Arch. Surg. 1984;119:20–27. [PubMed: 6689870]
Finegold, S. M. 1977. Anaerobic Bacteria in Human Disease. Academic Press, Inc., New York, N.Y.
Gibbons R. J., MacDonald J. B. Hemin and vitamin K compounds as required factors for the cultivation of certain strains of Bacteroides melaninogenicus. J. Bacteriol. 1960;80:164–170. [PMC free article: PMC278836] [PubMed: 13827907]
Gorbach S. L. Intestinal microflora. Gastroenterology. 1971;50:1110–1116. [PubMed: 4933894]
Hite K. E., Locke M., Hesseltine H. C. Synergism in experimental infections with nonsporulating anaerobic bacteria. Infect. Dis. 1949;84:1–9. [PubMed: 18109246]
Hofstad T. Virulence factors in anaerobic bacteria. Eur. J. Clin. Microbiol. Infect. Dis. 1992;11:1044–1048. [PubMed: 1363545]
Ingham H. R., Sisson P. R., Tharagonnet D., Seldon J. B., Codd A. A. Inhibition of phagocytosis in vitro by obligate anaerobes. Lancet. 1977;ii:1252–1254. [PubMed: 73952]
Jones G. R., Gemmell C. G. Impairment by Bacteroides species of opsonization and phagocytosis of enterobacteria. J. Med. Microbiol. 1982;15:351–361. [PubMed: 6126594]
Kasper D. L. Immunochemical characterization of two surface polysaccharides of Bacteroides fragilis. Infect. Immun. 1991;59:2075–2082. [PMC free article: PMC257968] [PubMed: 2037368]
Kasper D. L., Onderdonk A. B. Infection with Bacteroides fragilis pathogenesis and immunoprophylaxis in an animal model. Scand. J. Infect. Dis. 1982;31(Suppl.):28–33. [PubMed: 6954635]
Lev M. K., Kridell K. C., Milford A. F. Succinate as a growth factor for Bacteroides melaninogenicus. J. Bacteriol. 1971;108:175–178. [PMC free article: PMC247047] [PubMed: 4941554]
Mayrand D., McBride B. G. Ecological relationships of bacteria involved in a simple mixed anaerobic infection. Infect. Immun. 1980;27:44–50. [PMC free article: PMC550719] [PubMed: 6987176]
Meislin H. W., Lerner S. A., Graves M. H., McGehee M. D., Kocka F. E., Morello J. A., Rosen P. Cutaneous abscesses. Anaerobic and aerobic bacteriology and outpatient management. Ann. Intern. Med. 1977;87:145–149. [PubMed: 196534]
Meleny F., Olpp S., Harvey H. D., Jern H. Peritonitis. Synergism of bacteria commonly found in peritoneal exudates. Arch. Surg. 1932;25:709–721.
Mergenhagen S. E., Thonard J. C., Scherp H. W. Studies on synergistic infections. Experimental infections with anaerobic streptococci. J. Infect. Dis. 1958;103:33–44. [PubMed: 13575859]
Namavar F. A., Verweij M. J. J., Bal M., van Steenbergen T. J. M., de Graaf J., MacLaren D. M. Effect of anaerobic bacteria on killing of Proteus mirabilis by human poly-morphonuclear leukocytes. Infect. Immun. 1983;40:930–935. [PMC free article: PMC348141] [PubMed: 6133837]
Namavar F., Verweij-van Vught A. M. J. J., Vel W. A. C., Bal M., MacLaren D. M. Polymorphonuclear leukocyte chemotaxis by mixed anaerobic and aerobic bacteria. J. Med. Microbiol. 1984;18:167–172. [PubMed: 6387126]
Onderdonk A. B., Kasper D. L., Cisneros D. L., Bartlett J. G. The capsular polysaccharide of Bacteroides fragilis as a virulence factor: comparison of the pathogenic potential of encapsulated strains. J. Infect. Dis. 1977;136:82–89. [PubMed: 886206]
Onderdonk A. B., Markham R. R., Zalenzik D. F., Cineros R. L., Kasper D. L. Evidence for T cell-dependent immunity to Bacteroides fragilis in an intra-abdominal abscess model. J. Clin. Invest. 1982;69:9–16. [PMC free article: PMC371162] [PubMed: 6976357]
Onderdonk A. B., Moon N. E., Kasper D. L., Bartlett J. G. Adherence of Bacteroides fragilis in vivo. Infect. Immun. 1978;19:1083–1087. [PMC free article: PMC422299] [PubMed: 640723]
Pantosti A., Tzianabos A. O., Reinap B. G., Onderdonk A. B., Kasper D. L. Bacteroides fragilis strains express multiple capsular polysaccharides. J. Clin. Microbiol. 1993;31:1850–1855. [PMC free article: PMC265644] [PubMed: 8349763]
Patrick S., Reid J. H., Coffey A. Capsulation of in vitro and in vivo frown Bacteroides species. J. Gen. Microbiol. 1986;132:1099–1109. [PubMed: 3760820]
Patrick S., Reid J. H., Larkin M. H. The growth and survival of capsulate and non-capsulate Bacteroides fragilis in vivo and in vitro. J. Med. Microbiol. 1984;17:237–246. [PubMed: 6726783]
Rotstein D., Pruett T. L., Fiegel V. D., Nelson R. D., Simmons R. L. Succinic acid, a metabolic byproduct of Bacteroides species, inhibits polymorphonuclear leukocytes function. Infect. Immun. 1985;48:402. [PMC free article: PMC261327] [PubMed: 3886546]
Simon G. L., Kelmpner M. S., Kasper D. L., Gorbach S. L. Alteration in opsonophagocytic killing by neutrophils of Bacteroides fragilis associated with animal and laboratory passage: effect of capsular polysaccharide. J. Infect. Dis. 1982;145:72–79. [PubMed: 7054319]
Talan D. A., Citron D. M., Abrahamian F. M., Moran G. J., Goldstein E. J. Bacteriologic analysis of infected dog and cat bites. Emergency Medicine Infection Study Group. N. Engl. J. Med. 1999;340:85–92. [PubMed: 9887159]
Thadepalli H., Gorbach S. L., Keith L. Anaerobic infections of the female genital tract: bacteriology and therapeutic implementation. Am. J. Obstet. Gynecol. 1973;117:1034–1039. [PubMed: 4586019]
Tofte R. W., Peterson P. U., Schmeling D. Opsonization of four Bacteroides species: role of the classical complement pathway and immunoglobulin. Infect. Immun. 1980;27:784–792. [PMC free article: PMC550840] [PubMed: 6103873]
Weinstein W. M., Onderdonk A. B., Bartlett J. G., Louie T. J., Gorbach S. L. Antimicrobial therapy of experimental intraabdominal sepsis. J. Infect. Dis. 1975;132:282–286. [PubMed: 1159331]
Copyright © 2002, ASM Press.
Bookshelf ID: NBK2497