- Email: [email protected]
Safety aspects of enterococci from the medical point of view
International Journal of Food Microbiology 88 (2003) 255 – 262 www.elsevier.com/locate/ijfoodmicro
Review
Safety aspects of enterococci from the medical point of view F.H. Kayser * Institute of Medical Microbiology, University of Zu¨rich, Gloriastrasse 30/32, P.O. Box CH-8028 Zurich, Switzerland Accepted 26 February 2003
Abstract Enterococci occur in a remarkable array of environments. They can be found in soil, food, and water, and make up a significant portion of the normal gut flora of humans (105 – 107/g of stool) and animals. As other bacteria of the gut flora, enterococci can also cause infectious diseases. Most clinical isolates are Enterocococus faecalis, which account for 80 – 90% of clinical strains. Enterocococus faecium accounts for 5 – 10% of such isolates. Typical enterococcal infections occur in hospitalised patients with underlying conditions representing a wide spectrum of severity of illness and immune modulation. Enterococci today rank second to third in frequency among bacteria isolated from hospitalised patients. They are isolated from urinary tract infections, intra-abdominal and pelvic infections, bacteremias, wound and tissue infections, and endocarditis— often as part of a polymicrobial flora. Surprisingly, little is known about the factors that contribute to the ability of enterococci to cause infections. Many strains of E. faecalis produce a cytolysin (haemolysin) exhibiting tissue-damaging capacity. Further extracellular products often observed in clinical isolates are a proteinase (gelatinase), hyaluronidase, and extracellular superoxide. Furthermore, many of the clinical isolates possess the aggregation substance on the surface and an extracellular surface protein, both contributing to the adherence to eucaryotic cells. Some strains of E. faecalis, and many E. faecium strains are resistant to multiple antimicrobials. The ultimate role of all these factors in enterococcal pathogenicity remains to be determined. It was previously thought that enterococcal infections were endogenously acquired from the patient’s own gut flora. A rather new concept that has emerged is that enterococcal disease is a two-stage process. There is an initial colonisation of the gastrointestinal tract by enterococcal strains possessing virulence traits and/or antibiotic resistance. Subsequently, this population spreads, often facilitated by antibiotic elimination of competitors. For a selected number of patients, there is subsequent tissue invasion from the gastrointestinal tract reservoir. From this concept, it can be deduced that enterococcal strains without virulence traits and antibiotic resistances exogenously transferred into the human gut via food products or probiotics will not represent any risk for immunocompetent individuals. In very severely immunocompromised patients, however, a risk for enterococcal disease by such strains cannot completely be excluded. D 2003 Elsevier B.V. All rights reserved. Keywords: Enterocococus; Infections; Pathogenicity; Antibiotic resistance; Virulence
1. Introduction Enterococci were originally classified as enteric Gram-positive cocci and later included in the genus * Tel.: +41-1-825-4554; fax: +41-1-634-4906. E-mail address: [email protected] (F.H. Kayser). 0168-1605/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0168-1605(03)00188-0
Streptococcus. In the 1930s, with the establishment of the Lancefield serological typing system (Lancefield, 1933), enterococci were classified as group D streptococci. Sherman (1937) characterised group D streptococci by their ability to grow at 10 and 45 jC, at pH 9.6 and in 6.5% NaCl broth and to survive at 60 jC for 30 min. In the 1980s, based
256
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
on genetic differences, these organisms were removed from the genus Streptococcus and placed in their own genus, Enterococcus (Schleifer and Killer-Ba¨lz, 1984). The genus Enterococcus now comprises more than 20 species. In medicine, two of these are responsible for the overwhelming number of infections, Enterocococus faecalis and Enterocococus faecium. E. faecalis had been the predominant enterococcal species, accounting for 80– 90% of clinical isolates, and E. faecium for 5– 15%. Recent data indicate (Huycke et al., 1998) that E. faecium, especially multiresistant strains, is increasing in its relative number. Other enterococcal species such as Enterocococus durans, Enterocococus avium, Enterocococus casseliflavus, Enterocococus gallinarum, Enterocococus raffinosus, and Enterocococus hirae are only occasionally found as organisms causing infections in humans (Murray, 1990).
2. Habitat Enterococci occur in a remarkable array of environments, mainly because of their ability to grow and survive under harsh conditions. They can be regularly found in soil, food, water, and in a wide variety of animals. In humans, enterococci, along with approximately 450 other aerobic and anaerobic bacterial species, are part of the normal intestinal flora (Murray, 1990). In most individuals, 105 –107 CFU of enterococci per ml of stool are found. This seems to be a high number. It is, however, only a fraction of the total number of bacteria in the stool, i.e. 1010 – 1012 per ml, which is mainly composed of anaerobic Gram-negative rods. Smaller numbers of enterococci are observed in oropharyngeal secretions, vaginal secretions, and on the skin, especially in the perineal area. Thus, enterococci can be considered as normal commensals of the human organism.
3. Clinical infections Prior to the identification of multiple antibioticresistant strains in the late 1970s, enterococci were considered harmless bacteria, with the exception of
enterococci causing endocarditis. Over the last two decades, however, enterococci have been identified as agents of infections in hospitalised patients with increasing frequency. Currently, enterococci rank third or even second in frequency of bacteria isolated from hospitalised patients (Schaberg et al., 1991; Ru¨den et al., 1995). Major risk factors for acquiring nosocomial enterococcal infections are serious underlying disease, a long hospital stay, renal insufficiency, neutropenia, transplantation (especially liver and bone marrow transplantation), the presence of urinary or vascular catheters, and treatment in an intensive care facility. An important risk factor for infection is also is a preceding antibiotic therapy for other infectious diseases [e.g. with antibiotics against which enterococcci possess a natural resistance (such as cephalosporins) or an only intermediate susceptibility (such as fluoroquinolones)]. In this connection, superinfections with enterococci can occur. The most frequent infections caused by enterococci are summarised below (for review, see Moellering, 2000). 3.1. Urinary tract infections Urine cultures are the most frequent sources of enterococci in the clinical microbiology laboratory. Most infections are associated with catheterisation or instrumentation or both in hospitalised patients. Often, enterococci are not the only cause of a urinary tract infection, but they are found as part of a polymicrobial flora. In contrast, enterococci rarely are the cause of urinary tract infections in nonhospitalised individuals without predisposing conditions. 3.2. Bacteremia In nosocomial bacteremias, enterococci are observed in approximately 6 – 7% of all bacteria isolated from the blood stream (Bodmann and Vogel, 2001). S. aureus and E. coli are the predominant organisms in this serious infection. Portals of entry for enterococcal bacteremia include the urinary tract, intra-abdominal infections, burn wounds, diabetic foot infections, or intravascular catheters.
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
3.3. Intra-abdominal and pelvic infections Enterococci often are found as part of a mixed aerobic/anaerobic flora in intra-abdominal and pelvic infections. The role of enterococci in these conditions, however, remains doubtful. Elimination of organisms other than enterococci by antibiotic therapy not active against enterococci often is enough to cure the infection. However, in patients with nephrotic syndrome or cirrhosis, enterococci can cause spontaneous peritonitis, and also have been found to cause peritonitis in patients undergoing chronic ambulatory peritoneal dialysis (CAPD). 3.4. Wound and tissue infections Enterococci alone do not cause cellulitis or deep tissue infections. They are always isolated from mixed cultures together with Gram-negative aerobic rods and anaerobes. Thus, their significance in these infections is difficult to assess. Enterococcal wound colonisation has been observed in burn patients. They are also found occasionally in diabetic foot infections. 3.5. Endocarditis Enterococci account for approximately 5– 15% of all cases of infective endocarditis (Murray, 1990). Most cases occur in elderly male patients with underlying valvular heart disease or prosthetic valves. Common sources include genitourinary and biliary portals. In drug addicts, enterococci are estimated to cause 5– 10% of endocariditis cases. 3.6. Rare enterococcal infections Enterococci have rarely been observed to cause meningitis in normal adults. Most cases occur in patients with anatomic defects, prior neurosurgery, or head trauma. Enterococci also are exceedingly unusual in respiratory tract infections. In conclusion: Enterococci possess a low intrinsic pathogenicity and virulence potential. These organisms are mostly observed as part of a mixed flora causing infections in patients with severe underlying disease. Enterococci alone can cause sepsis or endocarditis in immunocompromised persons. Thus, a prerequisite of enterococci to cause disease is defects
257
in the resistance mechanisms of the host to these organisms.
4. Factors of the guest contributing to pathogenesis Surprisingly, little is known about the factors that contribute to the ability of enterococci to cause disease in humans. Certainly, there is not a single factor responsible for virulence such as the cholera toxin is in cholera or the diphtheria toxin in diphtheria. The major determinants of virulence known so far to play some role in the pathogenesis are certain secreted products on the one hand and adherence factors on the other. Furthermore, studies with E. faecium have shown that antibiotic resistance does play an important role in the pathogenesis of enterococcal disease (for review, see Jett et al., 1994; Mundy et al., 2000). 4.1. Cytolysin The most important virulence trait in E. faecalis seems to be the cytolysin. The cytolytic capacity of enterococci is observed in vitro as a hemolytic zone on blood agar. This phenotype of many E. faecalis strains is specified by the cytolysin operon, mostly located on transmissible plasmids and occasionally also on the chromosome. The cytolysin causes rupture of a variety of target membranes, including bacterial cells (bacteriocin), erythrocytes, and other eucaryotic cells. Nucleotide sequence determination for the cytolysin operon has revealed a complex determinant encoding five genetic markers (Jett et al., 1994). The genes cylL and cylS code for two precursor proteins. The modification protein CylM is required for the transformation of the two molecules to a secretable form. Both molecules are then secreted with the help of CylB, the product of the gene cylB of the cytolysin operon. Once secreted, the cytolysins L and S are still inactive until six amino acids are removed from each amino terminus with an enzyme encoded for by the gene cylA. In several infection models, cytolysin has been found to contribute to toxicity and lethality. For instance, cytolysin has been observed to favour the appearance of enterococci in the blood stream of experimentally infected mice (Ike et al., 1984). In an experiment, two isogenic E. faecalis strains either cytolytic or non-cytolytic were
258
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
injected intraperitoneally into mice and the number of bacteria appearing in the blood stream was significantly higher in mice infected with the cytolysinpositive strain. In humans, cytolysin also seems to contribute to pathogenesis. In one study, cytolytic strains in bacteremia were determined by regression analysis to be associated with a fivefold increased risk for death compared with infections caused by noncytolytic strains (Huycke et al., 1991). 4.2. Protease Purification and characterization of an approximately 30-kDa metalloprotease was described already 38 years ago (Bleiweis and Zimmerman, 1964). A potential contribution of this protease to virulence was suggested by the observation that protease production is a common trait in E. faecalis isolated from surgical and neurosurgical intensive care patients (Ku¨hnen et al., 1988). In another study, 54% of enterococci isolated from infected patients produced this enzyme as compared to only 12% and 14% of isolates from uninfected hospitalised patients and healthy volunteers, respectively (Coque et al., 1993). Such epidemiological studies, however, can only suggest association between protease production and virulence. 4.3. Hyaluronidase No studies up to now have demonstrated the role of hyaluronidase produced by many enterococcal strains in the pathogenesis of infection. Studies in other microorganisms, however, provide an indirect basis for speculating that this enzyme may contribute to enterococcal virulence (for review, see Mundy et al., 2000). 4.4. Extracellular superoxide Recent observations suggest that a lot of E. faecalis strains, but only a few E. faecium isolates, generate extracellular superoxide. When E. faecalis isolates from patients with endocarditis and bacteremia were compared with isolates from healthy volunteers, on average, superoxide production was 60% higher among blood isolates than in commensal strains (Huycke et al., 1995). The exact role of
superoxide in the host tissue damage, however, remains unclear. 4.5. Adherence to host tissues Bacterial adherence to host tissues is a first step in the infectious process. For gastrointestinal commensals like enterococci, adhesins that promote binding to receptors of the intestinal mucosa would be expected to play a critical role for colonisation. Without specific means of attachment, enterococci would likely be eliminated through normal intestinal motility. The molecules, with which enterococci can adhere to specific receptors of host cells, still are poorly defined. It has been suggested that the aggregation substance as well as an extracellular surface protein play a role in adherence (for review, see Jett et al., 1994; Mundy et al., 2000). Aggregation substance is a pheromoneinducible surface protein of E. faecalis encoded by pheromone-responsive plasmids and expressed in response to pheromone induction (Clewell, 1993). The protein is able to bind to the so-called binding substance of receptor cells promoting mating aggregate formation of plasmid-positive donor cells and plasmid-negative receptors during bacterial conjugation. In vitro, aggregation substance is able to mediate adhesion to a variety of eucaryotic cell surfaces such as cultured renal tubular cells (Kreft et al., 1992). The substance has also been shown to promote internalisation of enterococci by cultured human intestinal cells (Olmested et al., 1994). In a rabbit model of E. faecalis endocarditis, the substance has been shown to be associated with greater vegetation size compared to vegetations caused by isogenic aggregation substancedefective strains (Chow et al., 1993). Recent studies in an endocarditis model have shown that aggregation substance together with the binding substance may lead to destruction of myocardial and pulmonary tissues (Schlievert et al., 1997). Aggregation substance also promotes intracellular survival of E. faecalis inside neutrophils by preventing or at least delaying fusion of the phagosome with lysosomes to form phagolysosomes. Aggregation substance is also induced by serum, suggesting that larger aggregates are formed in vivo and phagocytosis of enterococcal aggregates is thus aggravated (for review, see Mundy et al., 2000). As most cytolytic strains of E. faecalis also express aggregation substance—their genetic
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
determinants being part of pheromone-responsive plasmids—both factors may well work synergistically, as shown by the results of studies in endocarditis (Chow et al., 1993). Extracellular surface protein (ESP) is another trait that appears to be associated with enterococcal virulence (Shankar et al., 1999). The gene esp encodes a large surface protein, whose exact function in pathogenesis is unknown. PCR amplification detected esp in only 3% of E. faecalis stool isolates, but in 41% of endocarditis isolates. A variant of this gene was also detected in E. faecium clinical isolates. (Willems et al., 2001). 4.6. Antimicrobial resistance Enterococci have been recognised as cause of endocarditis for almost a century. In addition to this long-established role, enterococci began to be recognised as common cause of hospital-acquired infections in recent decades. One of the major reasons why these organisms have survived in the hospital environment is their resistance to a variety of antimicrobials (for review, see Huycke et al., 1998; Cetinkaya et al., 2000; Bonten et al., 2001). Enterococci exhibit intrinsically reduced susceptibility to penicillin and to the aminopenicillins with MICs 100 times higher than MICs against streptococci. In addition, enterococci have a natural low level of resistance to lincosamides. They are intrinsically low level resistant to all aminoglycoside antibiotics, due to a decreased ability of these agents to penetrate the cell wall. This poor penetration can be overcome by the addition of a cell-wall active agent, enabling the aminoglycosides to cross the cell wall and exert their bactericidal activity inside the enterococcal cell. However, this synergism only functions if the corresponding enterococcal strain is not resistant to penicillins and/or not high level resistant to aminoglycosides. Enterococci are naturally resistant to all cephalosporins. Cotrimoxazol has been shown to fail in the therapy of enterococcal infections in both animal models and in patients. Enterococci also are able to acquire resistance either by mutation of existing chromosomal genes or by transfer of preformed resistance determinants. Examples of mutational resistance are resistance to aminopenicillins through alteration of penicillin-binding proteins, resistance to fluoroquinolones, and resistance to rifampin. Examples of
259
acquired resistance through exchange of resistanceencoding genes are resistance to tetracyclines, macrolides, chloramphenicol, glycopeptides, and to high concentrations of aminoglycosides. Acquisition can be due to conjugation mediated either by conjugative transposons, pheromone-responsive conjugative plasmids, or broad host range conjugative plasmids. Of special concern is resistance to penicillin and the aminopenicillins, high level resistance to the aminoglycosides, and resistance to the glycopeptides vancomycin and teicoplanin, as these agents are the antimicrobials of choice in treating severe enterococcal infections. Resistance to penicillin/ampicillin is attributable to mutations in the gene coding for PBP-5, decreasing the affinity for penicillins to this protein. In clinical strains isolated in the USA, this type of resistance was observed in 2% of E. faecalis and in 80% of E. faecium, respectively (Huycke et al., 1998). In Europe, the respective figures were 1% and 76% (Kresken et al., 2000). Betalactamase-producing enterococci have also been observed (Murray, 1992). However, in contrast to the situation in staphylococci, production of this enzyme is constitutive, resistance is of low level and does occur only infrequently. Acquired high level resistance to the aminoglycoside streptomycin (MICs z 2000 Ag/ml) is either ribosomally mediated or due to the production of the streptomycin-inactivating enzyme adenylyl-transferase (for review, see Murray, 1990). High level resistance to gentamicin (MICs z 1000 mg/l) is predominantly the result of the inactivating bifunctional enzyme 2W-phosphotransferase/6V-acetyltransferase conferring resistance not only to gentamicin, but also to tobramycin, netilmicin, amikacin, and kanamycin. The problem with high level resistance is that penicillin –aminoglycoside synergy does not occur. There are six recognised phenotypes of glycopeptide resistance in enterococci: VanA, B, C, D, E, and G (for review, see Cetinkaya et al., 2000; McKessar et al., 2000; Bonten et al., 2001). Of these, VanA and VanB are the clinically most important in E. faecalis and E. faecium, and VanA is the most frequent type. These resistances are encoded by newly acquired gene clusters not previously found in enterococci (Arthur and Courvalin, 1993). VanA results in resistance to vancomycin and teicoplanin, whereas VanB results in resistance to vancomycin alone.
260
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
There is an important difference in respect to glycopeptide resistance between Europe and the USA (for review, see Bonten et al., 2001). In Europe, suspected reservoirs especially for VanA are in the commercially animal husbandry. The former use of avoparcin as a growth promoter in animal feeds was the contributor as avoparcin confers cross-resistance to vancomycin. Epidemiological studies have provided evidence that VRE in Europe are ubiquitous in animals, for which avoparcin was given as a supplement. Subsequent enteric colonisation of healthy humans (role of the food chain) has been documented (Klare et al., 1995). However, infections with vancomycin resistant enterococci in Europe are rare. In a multicentre study of the Paul-Ehrlich-Society for Chemotherapy from 1998 involving numerous laboratories in Germany, Austria, and Switzerland, only 1 clinical isolate among 757 E. faecalis was resistant to vancomycin (i.e. 0.1%), and 4 strains among 78 E. faecium were resistant ( = 5.1%). In another Europe-wide study, the proportion of VRE among nosocomial enterococcal infections was under 3% (Schouten et al., 2000). In the USA on the other hand, colonisation of hospitalised patients with VRE is endemic in many tertiary care hospitals and VRE increasingly cause infections (for review, see Bonten et al., 2001). Data from the National Nosocomial Infection Surveillance study of the USA reveal a rising percentage of VRE since 1989, with rates in 2000 approaching 20% of all enterococcal clinical isolates, most of them E. faecium (Anonymous, 2000). On the other hand, colonisation with VRE is absent in healthy people in the USA (no use of avoparcin in animal husbandry). In summary, what contributes to the pathogenesis of enterococcal diseases are secretion of cytolysin, a gelatinase and possibly also a hyaluronidase, extracellular superoxide, and the presence of factors such as the aggregation substance and extracellular surface protein, promoting the adherence to host cells. Antibiotic resistance is also a major contributing factor. In clinical E. faecium strains, the virulence factors usually are absent, with notable exceptions. Nearly nothing is known of the specific interactions between E. faecium and host tissue. The main factor, which seems to play a role in the pathogenesis of E. faecium infections, is antibiotic resistance, found in many clinical isolates of this species.
5. Factors of the host contributing to pathogenesis Pathogenesis of infectious processes always results as an interplay between the aggression factors of the invading organism and the host factors trying to prevent the occurrence of disease. The outcome is the result of whether the ‘‘host’’ with his immune system or the ‘‘unwanted guest’’ with his determinants of pathogenicity and virulence prevails. The risk factors for the ‘‘host’’ comprise severe underlying disease, all kinds of severe immunosuppression, a long hospital stay (mainly in tertiary care hospitals), the presence of urinary or vascular catheters, and residency in an intensive care unit. A major risk factor for the host also is antibiotic therapy with agents inactive against enterococcal invaders. However, even in persons with impaired immunoresistance, enterococci alone are often unable to cause manifest disease. Enterococci frequently are isolated from infectious processes as part of a polymicrobial flora. Elimination of one or several partners of this polymicrobial flora other than enterococci often is enough to cure the infection.
6. Pathogenesis Enterococci belong to the normal gut flora of all humans. Therefore it was previously thought that infections due to these organisms were endogenously acquired from the patient’s own flora. Analysis of enterococcal infections in recent years has shown, however, that most infecting strains appear to be exogenously acquired. As it is also the case with other nosocomial infections, nosocomial enterococcal disease occurs by a two-step process. There is an initial, usually asymptomatic colonisation of patients, mainly of the gastrointestinal tract or occasionally the skin, by strains endemic in a hospital. These strains can come from other patients, from the hospital personnel harbouring these strains in their own gastrointestinal tract, or from environmental sources in the hospital. Such nosocomial strains often possess one or more virulence traits and/or antibiotic resistance. As host immunosuppression increases, the requirements for particular traits of the infecting strain for manifest disease decrease. Subsequently, this exogenous enterococcal population expands, often
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
facilitated by elimination of competitors through antibiotics. Once colonised with enterococci, patients may carry them for months or even years. For a number of patients, however, the second step follows, i.e. tissue invasion from the enterococcal reservoir and eventually disease.
7. Safety The increasing number of enterococcal infections in recent decades has raised concern whether the exogenous transfer of enterococci present in various foods or the treatment of gastrointestinal disease with probiotics containing enterococci could result in infections with these organisms and, thus, be a health hazard for humans. From the data presented, it seems clear that enterococci used in the food industry and in probiotics ideally should possess neither any of the virulence factors known so far nor acquired antibiotic resistance. Strains with the absence of these properties would not be able to permanently colonise the intestine, which is the first step in the infectious process. They would obviously also not be able to cause disease, should they incidentally be translocated to other sites, which normally do not contain a bacterial flora. The selection of enterococci for food production or in probiotics, therefore, requires careful safety evaluations. A well-documented example of the safety of exogenous applied enterococci is E. faecium strain SF68, used in pharmaceutical preparations to treat certain intestinal tract disorders (Reid, 1999; Elmer, 2001; Marteau et al., 2001). Examination of this organism has revealed the absence of any enterococcal virulence factors (files of Cerbios-Pharma, Barbengo, Switzerland). Furthermore, the strain does not carry a pheromone responsive conjugative plasmid, which codes for the aggregation substance. In agreement with these data is the observation that this strain is not able to adhere to vascular epithelial cells and to endocardial cells (Chisari et al., 1992). Steady state levels of SF68 in the gut of piglets were not established (Bongetta et al., 1981; Havenaar and Huis in t’Veld, 1993). When 3 108 organisms of SF68 were administered to healthy volunteers, SF68 could be detected in the stools up to 72 h after administration, and after 96 h, SF68 was no longer detectable (Car-
261
bone et al., 1980). Thus, constant colonisation of the intestinal tract over a longer period of time seems not to occur. Furthermore, acquired resistance to antimicrobials in this strain is not present (files of CerbiosPharma). In agreement with all these data is the observation that no single case of infection with this organism in humans has ever been observed. Thus, based on a more than 20-year-long history of the safe association of this strain with probiotic therapy, the use of E. faecium SF68 in pharmaceutical preparations obviously seems to be no risk for human health at the present time.
References Anonymous, 2000. National nosocomial infections surveillance (NNIS) System report, data summary from January 1992 – April 2000. American Journal of Infection Control 28, 429 – 448. Arthur, M., Courvalin, P., 1993. Genetics and mechanism of glycopeptide resistance in enterococci. Antimicrobial Agents and Chemotherapy 37, 1563 – 1571. Bleiweis, A.S., Zimmerman, L.N., 1964. Properties of proteinase from Streptococcus faecalis var. liquefaciens. Journal of Bacteriology 88, 653 – 659. Bodmann, K.F., Vogel, F., 2001. Antimikrobielle Therapie der Sepsis. Empfehlungen einer Arbeitsgruppe der Paul-Ehrlich-Gesellschaft fu¨r Chemotherapie e.V. Chemotherapie-Journal, 43 – 55. Bongetta, R., Quirino, T., Ortisi, G., Privitera, G., Foschi, D., Davagna, G., Moroni, M., Rovati, V., 1981. The colonization of Streptococcus faecium in human intestinal tract after oral administration. Bollettino dell’Istituto Sieroterapeutico Milanese (Milano) 60, 381 – 385. Bonten, M.J.M., Willems, R., Weinstein, R.A., 2001. Vancomycinresistant enterococci: why are they here, and where do they come from? Lancet. Infectious Diseases 1, 314 – 325. Carbone, M., Bonina, L., Fera, M.T., 1980. Microbiological properties of Streptococcus faecium SF68 and its relationships with other microorganisms. Bollettino dell’Istituto Sieroterapeutico Milanese (Milano) 59, 591 – 598. Cetinkaya, Y., Falk, P., Mayhall, C.G., 2000. Vancomycin-resistant enterococci. Clinical Microbiology Reviews 13, 686 – 707. Chisari, G., Lo Bue, A.M., Drago, L., Abbiati, R., Gismondo, R., 1992. Capacita` aderente di Enterococcus faecium (SF68) e di Enterococcus faecalis a diversi substrati. Giornale Italiano di Chemioterapia 39, 11 – 15. Chow, J.W., Thal, L.A., Berri, M.B., Vazquez, J.A., Donabedian, S.M., Clewell, D.B., Zervos, M.J., 1993. Plasmid-associated hemolysin and aggregation substance production contribute to virulence in experimental enterococcal endocarditis. Antimicrobial Agents and Chemotherapy 37, 2474 – 2477. Clewell, D.B., 1993. Bacterial sex pheromone-induced plasmid transfer. Cell 73, 9 – 12.
262
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
Coque, T.M., Steckelberg, J.M., Patterson, J.E., Murray, B.E., 1993. Possible virulence factors of enterococci. Abstr. 1166, 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). Elmer, G.W., 2001. Probiotics: living drugs. American Journal of Health-System Pharmacy 58, 1101 – 1109. Havenaar, R., Huis in t’Veld, J.H.J., 1993. In vitro and in vivo experiments with two commercial probiotic products containing Enterococcus faecium and Bacillus toyoi. In: Jensen, F.F., Hinton, M.H., Mulder, R.W.A.W. (Eds.), Probiotics and Pathogenicity. Proceedings of a meeting held at Vedbaek, Denmark, May 13 – 15. Huycke, M.M., Spiegel, C.A., Gilmore, M.S., 1991. Bacteremia caused by hemolytic, high-level gentamicin-resistant Enterococcus faecalis. Antimicrobial Agents and Chemotherapy 35, 1626 – 1634. Huycke, M., Joyce, W., Gilmore, M.S., 1995. Augmented production of extracellular superoxide by blood isolates of Enterococcus faecalis. Journal of Infectious Diseases 172, 273 – 276. Huycke, M.M., Sahm, D.F., Gilmore, M.S., 1998. Multiple-drug resistant enterococci: the nature of the problem and an agenda for the future. Emerging Infectious Diseases 4, 239 – 249. Ike, Y., Hashimoto, H., Clewell, D.B., 1984. Hemolysin of Streptococcus faecalis subspecies zymogenes contributes to virulence in mice. Infection and Immunity 45, 528 – 530. Jett, B.D., Huycke, M.M., Gilmore, M.S., 1994. Virulence of enterococci. Clinical Microbiology Reviews 7, 462 – 478. Klare, I., Heier, H., Claus, H., Bo¨hme, G., Marin, S., Seltmann, G., Hakenbeck, R., Atanassova, V., Witte, W., 1995. Enterococcus faecium strains with vanA-mediated high-level glycopeptide resistance isolated from food-stuffs and fecal samples of humans in the community. Microbial Drug Resistance 1, 265 – 272. Kreft, B., Marre, R., Schramm, U., Wirth, R., 1992. Aggregation substance of Enterococcus faecalis mediates adhesion to cultured renal tubular cells. Infection and Immunity 60, 25 – 30. Kresken, M., Hafner, D., und die Studiengruppe, 2000. Resistenzsituation bei klinisch wichtigen Infektionserregern gegenu¨ber Chemotherapeutika in Mitteleuropa. Ergebnisse einer multizentrischen Studie der Arbeitsgemeinschaft ‘‘Resistenz’’ in der Paul-Ehrlich-Gesellschaft fu¨r Chemotherapie e. V. aus dem Jahre 1998. Chemotherapie-Journal 9, 51 – 86. Ku¨hnen, E., Richter, F., Richter, K., Andries, L., 1988. Establishment of a typing system for group D streptococci. Zentralblatt fu¨r Bakteriologie, Mikrobiologie und Hygiene. Series A 267, 322 – 330. Lancefield, R.C., 1933. Serological differentiation of human and other groups of haemolytic streptococci. Journal of Experimental Medicine 57, 571 – 595. Marteau, Ph.R., de Vrese, M., Cellier, Ch.J., Schrezenmeir, J., 2001. Protection from gastrointestinal diseases with the use of probiotics. American Journal of Clinical Nutrition 73, 430S – 436S (Suppl.). McKessar, S.J., Berry, A.M., Bell, J.M., Turnidge, J.D., Paton, J.C., 2000. Genetic characterization of vanG, a novel vancomycin
resistance locus of Enterococcus faecalis. Antimicrobial Agents and Chemotherapy 44, 3224 – 3228. Moellering, R.C., 2000. Enterococcus species. In: Mandell, G.L., Bennett, J.E., Dolin, R. (Eds.), Principles and Practice of Infectious Diseases, 5th ed. Churchill Livingstone, Philadelphia, pp. 2147 – 2156. Mundy, L.M., Sahm, D.F., Gilmore, M., 2000. Relationship between enterococcal virulence and antimicrobial resistance. Clinical Microbiology Reviews 13, 513 – 522. Murray, B.E., 1990. The life and times of the Enterococcus. Clinical Microbiology Reviews 3, 46 – 65. Murray, B., 1992. Betalactamase-producing enterococci. Antimicrobial Agents and Chemotherapy 36, 2355 – 2359. Olmested, S., Dunny, G., Erlandsen, S., Wells, C., 1994. A plasmidencoded surface protein on Enterococcus faecalis augments its internalization by cultured intestinal epithelial cells. Journal of Infectious Diseases 170, 1549 – 1556. Reid, G., 1999. The scientific basis for probiotic strains of Lactobacillus. Applied and Environmental Microbiology 65, 3763 – 3766. Ru¨den, H., Daschner, F., Schumacher, M., 1995. Nosokomiale Infektionen in Deutschland—Erfassung und Pra¨vention (NIDEPStudie). Band 56. Schriftenreihe des Bundesministeriums fu¨r Gesundheit. Nomos Verlagsgesellschaft Baden-Baden. Schaberg, D.R., Culver, D.H., Gaynes, R.P., 1991. Major trends in the microbial etiology of nosocomial infection. American Journal of Medicine 91 (Suppl. 3B), 72S – 75S. Schleifer, K.H., Kilpper-Ba¨lz, R., 1984. Transfer of Streptococcus faecalis and Streptococcus faecium to the genus Enterococcus nom. rev. as Enterococcus faecalis comb. nov. and Enterococcus faecium comb. nov. International Journal of Systematic Bacteriology 34, 31 – 34. Schlievert, P., Dunny, G., Stoehr, J., Assimacopoulos, A., 1997. Aggregation and binding substances enhance pathogenicity in a rabbit model of Enterococcus faecalis endocarditis. Advances in Experimental Medicine and Biology (New York, NY) 418, 789 – 791. Schouten, M., Hoogkamp-Korstanje, J., Meis, J., Voss, A., 2000. Prevalence of vancomycin-resistant enterococci in Europe. European Journal of Clinical Microbiology and Infectious Diseases 19, 812 – 822. Shankar, V., Baghdayan, A., Huycke, M., Lindahl, G., Gilmore, M., 1999. Infection-derived Enterococcus faecalis strains are enriched in esp, a gene encoding a novel surface protein. Infection and Immunity 67, 193 – 200. Sherman, J.M., 1937. The streptococci. Bacteriology Reviews 1, 3 – 97. Willems, R.J.L., Homan, W., Top, J., van Santen-Verheuvel, M., Tribe, D.l., Manzioros, S., Gaillard, C., Vandenbroucke-Grauls, C.M.J.E., Mascini, E.M., van Regten, E., van Embden, J.D.A., Bonten, M.J.M., 2001. Variant esp gene as a marker of a distinct genetic lineage of vancomycin-resistant Enterococcus faecium spreading in hospitals. Lancet 357, 853 – 855.
Review
Safety aspects of enterococci from the medical point of view F.H. Kayser * Institute of Medical Microbiology, University of Zu¨rich, Gloriastrasse 30/32, P.O. Box CH-8028 Zurich, Switzerland Accepted 26 February 2003
Abstract Enterococci occur in a remarkable array of environments. They can be found in soil, food, and water, and make up a significant portion of the normal gut flora of humans (105 – 107/g of stool) and animals. As other bacteria of the gut flora, enterococci can also cause infectious diseases. Most clinical isolates are Enterocococus faecalis, which account for 80 – 90% of clinical strains. Enterocococus faecium accounts for 5 – 10% of such isolates. Typical enterococcal infections occur in hospitalised patients with underlying conditions representing a wide spectrum of severity of illness and immune modulation. Enterococci today rank second to third in frequency among bacteria isolated from hospitalised patients. They are isolated from urinary tract infections, intra-abdominal and pelvic infections, bacteremias, wound and tissue infections, and endocarditis— often as part of a polymicrobial flora. Surprisingly, little is known about the factors that contribute to the ability of enterococci to cause infections. Many strains of E. faecalis produce a cytolysin (haemolysin) exhibiting tissue-damaging capacity. Further extracellular products often observed in clinical isolates are a proteinase (gelatinase), hyaluronidase, and extracellular superoxide. Furthermore, many of the clinical isolates possess the aggregation substance on the surface and an extracellular surface protein, both contributing to the adherence to eucaryotic cells. Some strains of E. faecalis, and many E. faecium strains are resistant to multiple antimicrobials. The ultimate role of all these factors in enterococcal pathogenicity remains to be determined. It was previously thought that enterococcal infections were endogenously acquired from the patient’s own gut flora. A rather new concept that has emerged is that enterococcal disease is a two-stage process. There is an initial colonisation of the gastrointestinal tract by enterococcal strains possessing virulence traits and/or antibiotic resistance. Subsequently, this population spreads, often facilitated by antibiotic elimination of competitors. For a selected number of patients, there is subsequent tissue invasion from the gastrointestinal tract reservoir. From this concept, it can be deduced that enterococcal strains without virulence traits and antibiotic resistances exogenously transferred into the human gut via food products or probiotics will not represent any risk for immunocompetent individuals. In very severely immunocompromised patients, however, a risk for enterococcal disease by such strains cannot completely be excluded. D 2003 Elsevier B.V. All rights reserved. Keywords: Enterocococus; Infections; Pathogenicity; Antibiotic resistance; Virulence
1. Introduction Enterococci were originally classified as enteric Gram-positive cocci and later included in the genus * Tel.: +41-1-825-4554; fax: +41-1-634-4906. E-mail address: [email protected] (F.H. Kayser). 0168-1605/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0168-1605(03)00188-0
Streptococcus. In the 1930s, with the establishment of the Lancefield serological typing system (Lancefield, 1933), enterococci were classified as group D streptococci. Sherman (1937) characterised group D streptococci by their ability to grow at 10 and 45 jC, at pH 9.6 and in 6.5% NaCl broth and to survive at 60 jC for 30 min. In the 1980s, based
256
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
on genetic differences, these organisms were removed from the genus Streptococcus and placed in their own genus, Enterococcus (Schleifer and Killer-Ba¨lz, 1984). The genus Enterococcus now comprises more than 20 species. In medicine, two of these are responsible for the overwhelming number of infections, Enterocococus faecalis and Enterocococus faecium. E. faecalis had been the predominant enterococcal species, accounting for 80– 90% of clinical isolates, and E. faecium for 5– 15%. Recent data indicate (Huycke et al., 1998) that E. faecium, especially multiresistant strains, is increasing in its relative number. Other enterococcal species such as Enterocococus durans, Enterocococus avium, Enterocococus casseliflavus, Enterocococus gallinarum, Enterocococus raffinosus, and Enterocococus hirae are only occasionally found as organisms causing infections in humans (Murray, 1990).
2. Habitat Enterococci occur in a remarkable array of environments, mainly because of their ability to grow and survive under harsh conditions. They can be regularly found in soil, food, water, and in a wide variety of animals. In humans, enterococci, along with approximately 450 other aerobic and anaerobic bacterial species, are part of the normal intestinal flora (Murray, 1990). In most individuals, 105 –107 CFU of enterococci per ml of stool are found. This seems to be a high number. It is, however, only a fraction of the total number of bacteria in the stool, i.e. 1010 – 1012 per ml, which is mainly composed of anaerobic Gram-negative rods. Smaller numbers of enterococci are observed in oropharyngeal secretions, vaginal secretions, and on the skin, especially in the perineal area. Thus, enterococci can be considered as normal commensals of the human organism.
3. Clinical infections Prior to the identification of multiple antibioticresistant strains in the late 1970s, enterococci were considered harmless bacteria, with the exception of
enterococci causing endocarditis. Over the last two decades, however, enterococci have been identified as agents of infections in hospitalised patients with increasing frequency. Currently, enterococci rank third or even second in frequency of bacteria isolated from hospitalised patients (Schaberg et al., 1991; Ru¨den et al., 1995). Major risk factors for acquiring nosocomial enterococcal infections are serious underlying disease, a long hospital stay, renal insufficiency, neutropenia, transplantation (especially liver and bone marrow transplantation), the presence of urinary or vascular catheters, and treatment in an intensive care facility. An important risk factor for infection is also is a preceding antibiotic therapy for other infectious diseases [e.g. with antibiotics against which enterococcci possess a natural resistance (such as cephalosporins) or an only intermediate susceptibility (such as fluoroquinolones)]. In this connection, superinfections with enterococci can occur. The most frequent infections caused by enterococci are summarised below (for review, see Moellering, 2000). 3.1. Urinary tract infections Urine cultures are the most frequent sources of enterococci in the clinical microbiology laboratory. Most infections are associated with catheterisation or instrumentation or both in hospitalised patients. Often, enterococci are not the only cause of a urinary tract infection, but they are found as part of a polymicrobial flora. In contrast, enterococci rarely are the cause of urinary tract infections in nonhospitalised individuals without predisposing conditions. 3.2. Bacteremia In nosocomial bacteremias, enterococci are observed in approximately 6 – 7% of all bacteria isolated from the blood stream (Bodmann and Vogel, 2001). S. aureus and E. coli are the predominant organisms in this serious infection. Portals of entry for enterococcal bacteremia include the urinary tract, intra-abdominal infections, burn wounds, diabetic foot infections, or intravascular catheters.
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
3.3. Intra-abdominal and pelvic infections Enterococci often are found as part of a mixed aerobic/anaerobic flora in intra-abdominal and pelvic infections. The role of enterococci in these conditions, however, remains doubtful. Elimination of organisms other than enterococci by antibiotic therapy not active against enterococci often is enough to cure the infection. However, in patients with nephrotic syndrome or cirrhosis, enterococci can cause spontaneous peritonitis, and also have been found to cause peritonitis in patients undergoing chronic ambulatory peritoneal dialysis (CAPD). 3.4. Wound and tissue infections Enterococci alone do not cause cellulitis or deep tissue infections. They are always isolated from mixed cultures together with Gram-negative aerobic rods and anaerobes. Thus, their significance in these infections is difficult to assess. Enterococcal wound colonisation has been observed in burn patients. They are also found occasionally in diabetic foot infections. 3.5. Endocarditis Enterococci account for approximately 5– 15% of all cases of infective endocarditis (Murray, 1990). Most cases occur in elderly male patients with underlying valvular heart disease or prosthetic valves. Common sources include genitourinary and biliary portals. In drug addicts, enterococci are estimated to cause 5– 10% of endocariditis cases. 3.6. Rare enterococcal infections Enterococci have rarely been observed to cause meningitis in normal adults. Most cases occur in patients with anatomic defects, prior neurosurgery, or head trauma. Enterococci also are exceedingly unusual in respiratory tract infections. In conclusion: Enterococci possess a low intrinsic pathogenicity and virulence potential. These organisms are mostly observed as part of a mixed flora causing infections in patients with severe underlying disease. Enterococci alone can cause sepsis or endocarditis in immunocompromised persons. Thus, a prerequisite of enterococci to cause disease is defects
257
in the resistance mechanisms of the host to these organisms.
4. Factors of the guest contributing to pathogenesis Surprisingly, little is known about the factors that contribute to the ability of enterococci to cause disease in humans. Certainly, there is not a single factor responsible for virulence such as the cholera toxin is in cholera or the diphtheria toxin in diphtheria. The major determinants of virulence known so far to play some role in the pathogenesis are certain secreted products on the one hand and adherence factors on the other. Furthermore, studies with E. faecium have shown that antibiotic resistance does play an important role in the pathogenesis of enterococcal disease (for review, see Jett et al., 1994; Mundy et al., 2000). 4.1. Cytolysin The most important virulence trait in E. faecalis seems to be the cytolysin. The cytolytic capacity of enterococci is observed in vitro as a hemolytic zone on blood agar. This phenotype of many E. faecalis strains is specified by the cytolysin operon, mostly located on transmissible plasmids and occasionally also on the chromosome. The cytolysin causes rupture of a variety of target membranes, including bacterial cells (bacteriocin), erythrocytes, and other eucaryotic cells. Nucleotide sequence determination for the cytolysin operon has revealed a complex determinant encoding five genetic markers (Jett et al., 1994). The genes cylL and cylS code for two precursor proteins. The modification protein CylM is required for the transformation of the two molecules to a secretable form. Both molecules are then secreted with the help of CylB, the product of the gene cylB of the cytolysin operon. Once secreted, the cytolysins L and S are still inactive until six amino acids are removed from each amino terminus with an enzyme encoded for by the gene cylA. In several infection models, cytolysin has been found to contribute to toxicity and lethality. For instance, cytolysin has been observed to favour the appearance of enterococci in the blood stream of experimentally infected mice (Ike et al., 1984). In an experiment, two isogenic E. faecalis strains either cytolytic or non-cytolytic were
258
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
injected intraperitoneally into mice and the number of bacteria appearing in the blood stream was significantly higher in mice infected with the cytolysinpositive strain. In humans, cytolysin also seems to contribute to pathogenesis. In one study, cytolytic strains in bacteremia were determined by regression analysis to be associated with a fivefold increased risk for death compared with infections caused by noncytolytic strains (Huycke et al., 1991). 4.2. Protease Purification and characterization of an approximately 30-kDa metalloprotease was described already 38 years ago (Bleiweis and Zimmerman, 1964). A potential contribution of this protease to virulence was suggested by the observation that protease production is a common trait in E. faecalis isolated from surgical and neurosurgical intensive care patients (Ku¨hnen et al., 1988). In another study, 54% of enterococci isolated from infected patients produced this enzyme as compared to only 12% and 14% of isolates from uninfected hospitalised patients and healthy volunteers, respectively (Coque et al., 1993). Such epidemiological studies, however, can only suggest association between protease production and virulence. 4.3. Hyaluronidase No studies up to now have demonstrated the role of hyaluronidase produced by many enterococcal strains in the pathogenesis of infection. Studies in other microorganisms, however, provide an indirect basis for speculating that this enzyme may contribute to enterococcal virulence (for review, see Mundy et al., 2000). 4.4. Extracellular superoxide Recent observations suggest that a lot of E. faecalis strains, but only a few E. faecium isolates, generate extracellular superoxide. When E. faecalis isolates from patients with endocarditis and bacteremia were compared with isolates from healthy volunteers, on average, superoxide production was 60% higher among blood isolates than in commensal strains (Huycke et al., 1995). The exact role of
superoxide in the host tissue damage, however, remains unclear. 4.5. Adherence to host tissues Bacterial adherence to host tissues is a first step in the infectious process. For gastrointestinal commensals like enterococci, adhesins that promote binding to receptors of the intestinal mucosa would be expected to play a critical role for colonisation. Without specific means of attachment, enterococci would likely be eliminated through normal intestinal motility. The molecules, with which enterococci can adhere to specific receptors of host cells, still are poorly defined. It has been suggested that the aggregation substance as well as an extracellular surface protein play a role in adherence (for review, see Jett et al., 1994; Mundy et al., 2000). Aggregation substance is a pheromoneinducible surface protein of E. faecalis encoded by pheromone-responsive plasmids and expressed in response to pheromone induction (Clewell, 1993). The protein is able to bind to the so-called binding substance of receptor cells promoting mating aggregate formation of plasmid-positive donor cells and plasmid-negative receptors during bacterial conjugation. In vitro, aggregation substance is able to mediate adhesion to a variety of eucaryotic cell surfaces such as cultured renal tubular cells (Kreft et al., 1992). The substance has also been shown to promote internalisation of enterococci by cultured human intestinal cells (Olmested et al., 1994). In a rabbit model of E. faecalis endocarditis, the substance has been shown to be associated with greater vegetation size compared to vegetations caused by isogenic aggregation substancedefective strains (Chow et al., 1993). Recent studies in an endocarditis model have shown that aggregation substance together with the binding substance may lead to destruction of myocardial and pulmonary tissues (Schlievert et al., 1997). Aggregation substance also promotes intracellular survival of E. faecalis inside neutrophils by preventing or at least delaying fusion of the phagosome with lysosomes to form phagolysosomes. Aggregation substance is also induced by serum, suggesting that larger aggregates are formed in vivo and phagocytosis of enterococcal aggregates is thus aggravated (for review, see Mundy et al., 2000). As most cytolytic strains of E. faecalis also express aggregation substance—their genetic
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
determinants being part of pheromone-responsive plasmids—both factors may well work synergistically, as shown by the results of studies in endocarditis (Chow et al., 1993). Extracellular surface protein (ESP) is another trait that appears to be associated with enterococcal virulence (Shankar et al., 1999). The gene esp encodes a large surface protein, whose exact function in pathogenesis is unknown. PCR amplification detected esp in only 3% of E. faecalis stool isolates, but in 41% of endocarditis isolates. A variant of this gene was also detected in E. faecium clinical isolates. (Willems et al., 2001). 4.6. Antimicrobial resistance Enterococci have been recognised as cause of endocarditis for almost a century. In addition to this long-established role, enterococci began to be recognised as common cause of hospital-acquired infections in recent decades. One of the major reasons why these organisms have survived in the hospital environment is their resistance to a variety of antimicrobials (for review, see Huycke et al., 1998; Cetinkaya et al., 2000; Bonten et al., 2001). Enterococci exhibit intrinsically reduced susceptibility to penicillin and to the aminopenicillins with MICs 100 times higher than MICs against streptococci. In addition, enterococci have a natural low level of resistance to lincosamides. They are intrinsically low level resistant to all aminoglycoside antibiotics, due to a decreased ability of these agents to penetrate the cell wall. This poor penetration can be overcome by the addition of a cell-wall active agent, enabling the aminoglycosides to cross the cell wall and exert their bactericidal activity inside the enterococcal cell. However, this synergism only functions if the corresponding enterococcal strain is not resistant to penicillins and/or not high level resistant to aminoglycosides. Enterococci are naturally resistant to all cephalosporins. Cotrimoxazol has been shown to fail in the therapy of enterococcal infections in both animal models and in patients. Enterococci also are able to acquire resistance either by mutation of existing chromosomal genes or by transfer of preformed resistance determinants. Examples of mutational resistance are resistance to aminopenicillins through alteration of penicillin-binding proteins, resistance to fluoroquinolones, and resistance to rifampin. Examples of
259
acquired resistance through exchange of resistanceencoding genes are resistance to tetracyclines, macrolides, chloramphenicol, glycopeptides, and to high concentrations of aminoglycosides. Acquisition can be due to conjugation mediated either by conjugative transposons, pheromone-responsive conjugative plasmids, or broad host range conjugative plasmids. Of special concern is resistance to penicillin and the aminopenicillins, high level resistance to the aminoglycosides, and resistance to the glycopeptides vancomycin and teicoplanin, as these agents are the antimicrobials of choice in treating severe enterococcal infections. Resistance to penicillin/ampicillin is attributable to mutations in the gene coding for PBP-5, decreasing the affinity for penicillins to this protein. In clinical strains isolated in the USA, this type of resistance was observed in 2% of E. faecalis and in 80% of E. faecium, respectively (Huycke et al., 1998). In Europe, the respective figures were 1% and 76% (Kresken et al., 2000). Betalactamase-producing enterococci have also been observed (Murray, 1992). However, in contrast to the situation in staphylococci, production of this enzyme is constitutive, resistance is of low level and does occur only infrequently. Acquired high level resistance to the aminoglycoside streptomycin (MICs z 2000 Ag/ml) is either ribosomally mediated or due to the production of the streptomycin-inactivating enzyme adenylyl-transferase (for review, see Murray, 1990). High level resistance to gentamicin (MICs z 1000 mg/l) is predominantly the result of the inactivating bifunctional enzyme 2W-phosphotransferase/6V-acetyltransferase conferring resistance not only to gentamicin, but also to tobramycin, netilmicin, amikacin, and kanamycin. The problem with high level resistance is that penicillin –aminoglycoside synergy does not occur. There are six recognised phenotypes of glycopeptide resistance in enterococci: VanA, B, C, D, E, and G (for review, see Cetinkaya et al., 2000; McKessar et al., 2000; Bonten et al., 2001). Of these, VanA and VanB are the clinically most important in E. faecalis and E. faecium, and VanA is the most frequent type. These resistances are encoded by newly acquired gene clusters not previously found in enterococci (Arthur and Courvalin, 1993). VanA results in resistance to vancomycin and teicoplanin, whereas VanB results in resistance to vancomycin alone.
260
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
There is an important difference in respect to glycopeptide resistance between Europe and the USA (for review, see Bonten et al., 2001). In Europe, suspected reservoirs especially for VanA are in the commercially animal husbandry. The former use of avoparcin as a growth promoter in animal feeds was the contributor as avoparcin confers cross-resistance to vancomycin. Epidemiological studies have provided evidence that VRE in Europe are ubiquitous in animals, for which avoparcin was given as a supplement. Subsequent enteric colonisation of healthy humans (role of the food chain) has been documented (Klare et al., 1995). However, infections with vancomycin resistant enterococci in Europe are rare. In a multicentre study of the Paul-Ehrlich-Society for Chemotherapy from 1998 involving numerous laboratories in Germany, Austria, and Switzerland, only 1 clinical isolate among 757 E. faecalis was resistant to vancomycin (i.e. 0.1%), and 4 strains among 78 E. faecium were resistant ( = 5.1%). In another Europe-wide study, the proportion of VRE among nosocomial enterococcal infections was under 3% (Schouten et al., 2000). In the USA on the other hand, colonisation of hospitalised patients with VRE is endemic in many tertiary care hospitals and VRE increasingly cause infections (for review, see Bonten et al., 2001). Data from the National Nosocomial Infection Surveillance study of the USA reveal a rising percentage of VRE since 1989, with rates in 2000 approaching 20% of all enterococcal clinical isolates, most of them E. faecium (Anonymous, 2000). On the other hand, colonisation with VRE is absent in healthy people in the USA (no use of avoparcin in animal husbandry). In summary, what contributes to the pathogenesis of enterococcal diseases are secretion of cytolysin, a gelatinase and possibly also a hyaluronidase, extracellular superoxide, and the presence of factors such as the aggregation substance and extracellular surface protein, promoting the adherence to host cells. Antibiotic resistance is also a major contributing factor. In clinical E. faecium strains, the virulence factors usually are absent, with notable exceptions. Nearly nothing is known of the specific interactions between E. faecium and host tissue. The main factor, which seems to play a role in the pathogenesis of E. faecium infections, is antibiotic resistance, found in many clinical isolates of this species.
5. Factors of the host contributing to pathogenesis Pathogenesis of infectious processes always results as an interplay between the aggression factors of the invading organism and the host factors trying to prevent the occurrence of disease. The outcome is the result of whether the ‘‘host’’ with his immune system or the ‘‘unwanted guest’’ with his determinants of pathogenicity and virulence prevails. The risk factors for the ‘‘host’’ comprise severe underlying disease, all kinds of severe immunosuppression, a long hospital stay (mainly in tertiary care hospitals), the presence of urinary or vascular catheters, and residency in an intensive care unit. A major risk factor for the host also is antibiotic therapy with agents inactive against enterococcal invaders. However, even in persons with impaired immunoresistance, enterococci alone are often unable to cause manifest disease. Enterococci frequently are isolated from infectious processes as part of a polymicrobial flora. Elimination of one or several partners of this polymicrobial flora other than enterococci often is enough to cure the infection.
6. Pathogenesis Enterococci belong to the normal gut flora of all humans. Therefore it was previously thought that infections due to these organisms were endogenously acquired from the patient’s own flora. Analysis of enterococcal infections in recent years has shown, however, that most infecting strains appear to be exogenously acquired. As it is also the case with other nosocomial infections, nosocomial enterococcal disease occurs by a two-step process. There is an initial, usually asymptomatic colonisation of patients, mainly of the gastrointestinal tract or occasionally the skin, by strains endemic in a hospital. These strains can come from other patients, from the hospital personnel harbouring these strains in their own gastrointestinal tract, or from environmental sources in the hospital. Such nosocomial strains often possess one or more virulence traits and/or antibiotic resistance. As host immunosuppression increases, the requirements for particular traits of the infecting strain for manifest disease decrease. Subsequently, this exogenous enterococcal population expands, often
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
facilitated by elimination of competitors through antibiotics. Once colonised with enterococci, patients may carry them for months or even years. For a number of patients, however, the second step follows, i.e. tissue invasion from the enterococcal reservoir and eventually disease.
7. Safety The increasing number of enterococcal infections in recent decades has raised concern whether the exogenous transfer of enterococci present in various foods or the treatment of gastrointestinal disease with probiotics containing enterococci could result in infections with these organisms and, thus, be a health hazard for humans. From the data presented, it seems clear that enterococci used in the food industry and in probiotics ideally should possess neither any of the virulence factors known so far nor acquired antibiotic resistance. Strains with the absence of these properties would not be able to permanently colonise the intestine, which is the first step in the infectious process. They would obviously also not be able to cause disease, should they incidentally be translocated to other sites, which normally do not contain a bacterial flora. The selection of enterococci for food production or in probiotics, therefore, requires careful safety evaluations. A well-documented example of the safety of exogenous applied enterococci is E. faecium strain SF68, used in pharmaceutical preparations to treat certain intestinal tract disorders (Reid, 1999; Elmer, 2001; Marteau et al., 2001). Examination of this organism has revealed the absence of any enterococcal virulence factors (files of Cerbios-Pharma, Barbengo, Switzerland). Furthermore, the strain does not carry a pheromone responsive conjugative plasmid, which codes for the aggregation substance. In agreement with these data is the observation that this strain is not able to adhere to vascular epithelial cells and to endocardial cells (Chisari et al., 1992). Steady state levels of SF68 in the gut of piglets were not established (Bongetta et al., 1981; Havenaar and Huis in t’Veld, 1993). When 3 108 organisms of SF68 were administered to healthy volunteers, SF68 could be detected in the stools up to 72 h after administration, and after 96 h, SF68 was no longer detectable (Car-
261
bone et al., 1980). Thus, constant colonisation of the intestinal tract over a longer period of time seems not to occur. Furthermore, acquired resistance to antimicrobials in this strain is not present (files of CerbiosPharma). In agreement with all these data is the observation that no single case of infection with this organism in humans has ever been observed. Thus, based on a more than 20-year-long history of the safe association of this strain with probiotic therapy, the use of E. faecium SF68 in pharmaceutical preparations obviously seems to be no risk for human health at the present time.
References Anonymous, 2000. National nosocomial infections surveillance (NNIS) System report, data summary from January 1992 – April 2000. American Journal of Infection Control 28, 429 – 448. Arthur, M., Courvalin, P., 1993. Genetics and mechanism of glycopeptide resistance in enterococci. Antimicrobial Agents and Chemotherapy 37, 1563 – 1571. Bleiweis, A.S., Zimmerman, L.N., 1964. Properties of proteinase from Streptococcus faecalis var. liquefaciens. Journal of Bacteriology 88, 653 – 659. Bodmann, K.F., Vogel, F., 2001. Antimikrobielle Therapie der Sepsis. Empfehlungen einer Arbeitsgruppe der Paul-Ehrlich-Gesellschaft fu¨r Chemotherapie e.V. Chemotherapie-Journal, 43 – 55. Bongetta, R., Quirino, T., Ortisi, G., Privitera, G., Foschi, D., Davagna, G., Moroni, M., Rovati, V., 1981. The colonization of Streptococcus faecium in human intestinal tract after oral administration. Bollettino dell’Istituto Sieroterapeutico Milanese (Milano) 60, 381 – 385. Bonten, M.J.M., Willems, R., Weinstein, R.A., 2001. Vancomycinresistant enterococci: why are they here, and where do they come from? Lancet. Infectious Diseases 1, 314 – 325. Carbone, M., Bonina, L., Fera, M.T., 1980. Microbiological properties of Streptococcus faecium SF68 and its relationships with other microorganisms. Bollettino dell’Istituto Sieroterapeutico Milanese (Milano) 59, 591 – 598. Cetinkaya, Y., Falk, P., Mayhall, C.G., 2000. Vancomycin-resistant enterococci. Clinical Microbiology Reviews 13, 686 – 707. Chisari, G., Lo Bue, A.M., Drago, L., Abbiati, R., Gismondo, R., 1992. Capacita` aderente di Enterococcus faecium (SF68) e di Enterococcus faecalis a diversi substrati. Giornale Italiano di Chemioterapia 39, 11 – 15. Chow, J.W., Thal, L.A., Berri, M.B., Vazquez, J.A., Donabedian, S.M., Clewell, D.B., Zervos, M.J., 1993. Plasmid-associated hemolysin and aggregation substance production contribute to virulence in experimental enterococcal endocarditis. Antimicrobial Agents and Chemotherapy 37, 2474 – 2477. Clewell, D.B., 1993. Bacterial sex pheromone-induced plasmid transfer. Cell 73, 9 – 12.
262
F.H. Kayser / International Journal of Food Microbiology 88 (2003) 255 – 262
Coque, T.M., Steckelberg, J.M., Patterson, J.E., Murray, B.E., 1993. Possible virulence factors of enterococci. Abstr. 1166, 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). Elmer, G.W., 2001. Probiotics: living drugs. American Journal of Health-System Pharmacy 58, 1101 – 1109. Havenaar, R., Huis in t’Veld, J.H.J., 1993. In vitro and in vivo experiments with two commercial probiotic products containing Enterococcus faecium and Bacillus toyoi. In: Jensen, F.F., Hinton, M.H., Mulder, R.W.A.W. (Eds.), Probiotics and Pathogenicity. Proceedings of a meeting held at Vedbaek, Denmark, May 13 – 15. Huycke, M.M., Spiegel, C.A., Gilmore, M.S., 1991. Bacteremia caused by hemolytic, high-level gentamicin-resistant Enterococcus faecalis. Antimicrobial Agents and Chemotherapy 35, 1626 – 1634. Huycke, M., Joyce, W., Gilmore, M.S., 1995. Augmented production of extracellular superoxide by blood isolates of Enterococcus faecalis. Journal of Infectious Diseases 172, 273 – 276. Huycke, M.M., Sahm, D.F., Gilmore, M.S., 1998. Multiple-drug resistant enterococci: the nature of the problem and an agenda for the future. Emerging Infectious Diseases 4, 239 – 249. Ike, Y., Hashimoto, H., Clewell, D.B., 1984. Hemolysin of Streptococcus faecalis subspecies zymogenes contributes to virulence in mice. Infection and Immunity 45, 528 – 530. Jett, B.D., Huycke, M.M., Gilmore, M.S., 1994. Virulence of enterococci. Clinical Microbiology Reviews 7, 462 – 478. Klare, I., Heier, H., Claus, H., Bo¨hme, G., Marin, S., Seltmann, G., Hakenbeck, R., Atanassova, V., Witte, W., 1995. Enterococcus faecium strains with vanA-mediated high-level glycopeptide resistance isolated from food-stuffs and fecal samples of humans in the community. Microbial Drug Resistance 1, 265 – 272. Kreft, B., Marre, R., Schramm, U., Wirth, R., 1992. Aggregation substance of Enterococcus faecalis mediates adhesion to cultured renal tubular cells. Infection and Immunity 60, 25 – 30. Kresken, M., Hafner, D., und die Studiengruppe, 2000. Resistenzsituation bei klinisch wichtigen Infektionserregern gegenu¨ber Chemotherapeutika in Mitteleuropa. Ergebnisse einer multizentrischen Studie der Arbeitsgemeinschaft ‘‘Resistenz’’ in der Paul-Ehrlich-Gesellschaft fu¨r Chemotherapie e. V. aus dem Jahre 1998. Chemotherapie-Journal 9, 51 – 86. Ku¨hnen, E., Richter, F., Richter, K., Andries, L., 1988. Establishment of a typing system for group D streptococci. Zentralblatt fu¨r Bakteriologie, Mikrobiologie und Hygiene. Series A 267, 322 – 330. Lancefield, R.C., 1933. Serological differentiation of human and other groups of haemolytic streptococci. Journal of Experimental Medicine 57, 571 – 595. Marteau, Ph.R., de Vrese, M., Cellier, Ch.J., Schrezenmeir, J., 2001. Protection from gastrointestinal diseases with the use of probiotics. American Journal of Clinical Nutrition 73, 430S – 436S (Suppl.). McKessar, S.J., Berry, A.M., Bell, J.M., Turnidge, J.D., Paton, J.C., 2000. Genetic characterization of vanG, a novel vancomycin
resistance locus of Enterococcus faecalis. Antimicrobial Agents and Chemotherapy 44, 3224 – 3228. Moellering, R.C., 2000. Enterococcus species. In: Mandell, G.L., Bennett, J.E., Dolin, R. (Eds.), Principles and Practice of Infectious Diseases, 5th ed. Churchill Livingstone, Philadelphia, pp. 2147 – 2156. Mundy, L.M., Sahm, D.F., Gilmore, M., 2000. Relationship between enterococcal virulence and antimicrobial resistance. Clinical Microbiology Reviews 13, 513 – 522. Murray, B.E., 1990. The life and times of the Enterococcus. Clinical Microbiology Reviews 3, 46 – 65. Murray, B., 1992. Betalactamase-producing enterococci. Antimicrobial Agents and Chemotherapy 36, 2355 – 2359. Olmested, S., Dunny, G., Erlandsen, S., Wells, C., 1994. A plasmidencoded surface protein on Enterococcus faecalis augments its internalization by cultured intestinal epithelial cells. Journal of Infectious Diseases 170, 1549 – 1556. Reid, G., 1999. The scientific basis for probiotic strains of Lactobacillus. Applied and Environmental Microbiology 65, 3763 – 3766. Ru¨den, H., Daschner, F., Schumacher, M., 1995. Nosokomiale Infektionen in Deutschland—Erfassung und Pra¨vention (NIDEPStudie). Band 56. Schriftenreihe des Bundesministeriums fu¨r Gesundheit. Nomos Verlagsgesellschaft Baden-Baden. Schaberg, D.R., Culver, D.H., Gaynes, R.P., 1991. Major trends in the microbial etiology of nosocomial infection. American Journal of Medicine 91 (Suppl. 3B), 72S – 75S. Schleifer, K.H., Kilpper-Ba¨lz, R., 1984. Transfer of Streptococcus faecalis and Streptococcus faecium to the genus Enterococcus nom. rev. as Enterococcus faecalis comb. nov. and Enterococcus faecium comb. nov. International Journal of Systematic Bacteriology 34, 31 – 34. Schlievert, P., Dunny, G., Stoehr, J., Assimacopoulos, A., 1997. Aggregation and binding substances enhance pathogenicity in a rabbit model of Enterococcus faecalis endocarditis. Advances in Experimental Medicine and Biology (New York, NY) 418, 789 – 791. Schouten, M., Hoogkamp-Korstanje, J., Meis, J., Voss, A., 2000. Prevalence of vancomycin-resistant enterococci in Europe. European Journal of Clinical Microbiology and Infectious Diseases 19, 812 – 822. Shankar, V., Baghdayan, A., Huycke, M., Lindahl, G., Gilmore, M., 1999. Infection-derived Enterococcus faecalis strains are enriched in esp, a gene encoding a novel surface protein. Infection and Immunity 67, 193 – 200. Sherman, J.M., 1937. The streptococci. Bacteriology Reviews 1, 3 – 97. Willems, R.J.L., Homan, W., Top, J., van Santen-Verheuvel, M., Tribe, D.l., Manzioros, S., Gaillard, C., Vandenbroucke-Grauls, C.M.J.E., Mascini, E.M., van Regten, E., van Embden, J.D.A., Bonten, M.J.M., 2001. Variant esp gene as a marker of a distinct genetic lineage of vancomycin-resistant Enterococcus faecium spreading in hospitals. Lancet 357, 853 – 855.