Abstract

Underlying principles of replacement therapy

Dental caries

Otitis media

Streptococcal pharyngitis

Benefits of replacement therapy

Difficulties and possible risks

Summary and future outlook

References

Copyright

 

The individual bacterial members of our indigeneous microbiota are actively engaged in an on-going battle to prevent colonisation and overgrowth of their terrain by competing microbes, some of which might have pathogenic potential for the host. Humans have long attempted to intervene in these bacterial interactions. Ingestion of probiotic bacteria, particularly lactobacilli, is commonly practiced to promote well-balanced intestinal microflora. As bacterial resistance to antimicrobials has increased, so too has research into colonisation of human tissues with specific effector strains capable of out-competing known bacterial pathogens. Recent progress is particularly evident in the application of avirulent Streptococcus mutans to the control of dental caries, alpha hemolytic streptococci to reduction of otitis media recurrences and Streptococcus salivarius to streptococcal pharyngitis prevention.

Documentation showing that relatively harmless microorganisms could be introduced into the indigenous microbiota of humans either to enhance resistance to or to treat infection dates from the very origins of microbiology. In 1877, Pasteur and his associate Joubert, noting suppression of anthrax bacillus growth in co-cultures with 'common bacilli' (probably Escherichia coli), commented that ''these facts perhaps justify the highest hopes for therapeutics'' [1]. In the early years of the next century, there followed a series of brave attempts by physicians to afford protection against diseases such as tuberculosis, anthrax and diphtheria by dosing patients with putatively innocuous commensal bacteria [2]. However, except for the treatment of minor ailments or as supplemental therapy, the application of so-called 'bacteriotherapy' or 'bacterioprophylaxis' was largely discontinued on the spectacular advent of antibiotics. For a time it seems that both physicians and the public became rather complacent about our potential bacterial adversaries; therapeutic antibiotics were perceived to be omnipotent – the panaceas for all bacterial ills. Within the span of a single human generation many bacterial species adapted to their antibiotic-laced ecosystems and variants flourished that are capable of resisting our most potent designer antimicrobials. The medical community must now face the reality that most chemotherapeutic agents are probably destined for a relatively short half-life of effectiveness. This dilemma might increasingly encourage us to reconsider the Pasteur approach that bacteria themselves could be our most effective allies as we continue to confront the relatively small but resilient band of miscreant microbes capable of causing infections in man and other animals [3].

Ever since probiotics were made prominent by the flamboyant Eli Metchinikoff, the practice of regular ingestion of intestinal commensals (especially lactobacilli) to confer health-promoting benefits has been commonly used by humans and for various farmed animals [4]. One beneficial side-effect of long-term ingestion of probiotic lactobacilli reported recently is an apparent reduction in dental caries [5]. In general, however, probiotic principles have not been widely applied to the specific protection of other body surfaces against bacterial infection. Several small groups of dedicated adherents to the principles of replacement therapy have, however, initiated pilot studies aimed at protecting the human host against development of otitis media [6], dental caries [7] and urogenital tract infections [8] or against infection by Streptococcus pyogenes [9,10] and Staphylococcus aureus [11]. Although the outcomes have been reported to be promising, none of these studies has resulted in widespread acceptance and routine application of replacement therapy principles as a preventative regimen. One mitigating factor might be the ethical consideration that even reputedly low-virulence colonising strains might sometimes cause infections in immune-compromised individuals. Undaunted however, some of these groups have continued to explore the application of 'friendly bacteria' to infection control and it appears that momentum is now gathering, particularly for the control of three of the most common bacterial infections of childhood: dental caries, otitis media and streptococcal pharyngitis ( Table 1).

 

The accessible surfaces of the skin, oral cavity, upper respiratory tract, gastrointestinal tract and vagina of healthy vertebrates are colonized by microbes soon after birth. What generally follows is an orderly, site-specific succession of microbial acquisitions and eliminations as the populations of indigenous microbes evolve to form climax communities – highly stable menageries of microbes perfectly adapted to life in each particular ecosystem. Collectively known as the normal microflora or indigenous microbiota, these microbes are the body's first line of defense – our personal army of protectors – with a keen interest in our well-being because our healthy tissues constitute their preferred homeland.

The basis of replacement therapy is the implantation and persistence within the normal microflora of relatively innocuous 'effector' bacteria that can competitively exclude or prevent the outgrowth of potentially disease-causing bacteria, without significantly disturbing the balance of the existing microbial ecosystem [12] ( Box 1). Sometimes the mechanisms operating in situ to confer host protection in documented instances of replacement therapy or microbial interference are not known, as exemplified by the relatively harmless S. aureus strain 502A [13] and the rough colony variant strain of Streptococcus salivarius known as TOVE-R ( Fig. 1), which has anti-Streptococcus mutans action in vivo [14]. Nevertheless, it is generally considered that mechanisms contributing to microbial interference might typically include either the greater ability of the effector bacterium to adhere to surfaces and to compete with others for limited space and key nutrients or the superior capability of that bacterium to produce and be resistant to a variety of anti-competitor molecules, some relatively non-specific in their targeting (e.g. acids, hydrogen peroxide) and others (e.g. bacteriocins, bacteriocin-like inhibitory substances (BLIS) and bacteriophages) apparently principally targeted against relatively similar bacteria ( Fig. 2).

Box 1

View Box

 

On the basis of several decades of epidemiological observations and laboratory-based research, S. mutans, and to a lesser extent Streptococcus sobrinus, are now generally considered the principal etiological agents of dental caries in humans. Two attributes of these so-called mutans streptococci (MS) that are particularly relevant to their involvement in caries are their production of large quantities of acid from dietary carbohydrate and their formation from sucrose of highly-branched extracellular polysaccharides (glucans) that help trap acidic metabolites within the plaque matrix. According to the acidogenic theory of dental caries the products of bacterial fermentation, particularly lactic acid, mediate the development of caries – reducing the pH of the microenvironment of the tooth surface below the critical threshold at which dissolution of the mineral phase of enamel and dentine is initiated. The weakened surface eventually cavitates to form a clinically evident lesion.

Attempts to devise a microbial interference-based strategy to prevent dental caries first focused on identifying relatively non-pathogenic oral commensals capable of inhibiting growth of the majority of MS [15]. Unfortunately, with the notable exception of certain enterococci [16], S. salivarius [14] and Streptococcus equi subspecies zooepidemicus [17], the strongest producers of anti-MS activity appear to be other strains of MS [18].

Most progress in this field has been made by Jeffrey Hillman's group [19]. The rationale behind their studies is that relatively avirulent strains of MS are most likely to occupy the same ecological niche in plaque as their more cariogenic counterparts. Previous studies demonstrated that it is difficult to achieve persistent colonization of plaque with laboratory strains of MS, particularly in subjects already harbouring indigenous MS [20]. Indeed horizontal transfer of MS, even between close contacts, is a rare event other than for a period of several months following the onset of tooth eruption when children tend to acquire MS strains from their primary caregiver [21]. This 'window of infectivity' period is the preferred time to attempt implantation of MS effector strains [7]. Hillman's concept, however, was that the tooth surface might be more readily colonized by a naturally occurring or genetically engineered MS effector strain that was not only relatively avirulent (i.e. weakly cariogenic) but also highly competitive, owing to its production of anti-MS BLIS activity. The strains selected for further development as the effector for MS replacement therapy were S. mutans JH1000 and its tetracycline-resistant mutant JH1001 [22]. Strain JH1000 produced potent BLIS activity in vitro that was inhibitory to a wide variety of Gram-positive species, including all but one of 125 tested MS [23]. Several mutants (eg. JH1005 and JH1140) of this strain producing increased levels of BLIS were used in human subjects to demonstrate that colonization efficacy and persistence correlated directly with the level of BLIS production [24]. The epidermin-like lantibiotic mutacin 1140 has now been isolated from strain JH1140 [25] and its primary structure elucidated [26].

Strain JH1001 and its derivatives have strong colonization properties in humans. For example, strain JH1005 was apparently retained for 14 years in some subjects following a single application, and moreover appeared to competitively exclude colonization with all other S. mutans strains [27]. The major problem with the use of strain JH1001 and its BLIS-enhanced derivatives in replacement therapy is that these strains, being highly acidogenic MS, are themselves potentially cariogenic.

Parallel studies by the same group explored techniques for reducing the cariogenicity of potential effector strains. Although mutants exhibiting defects in glucan synthesis have reduced cariogenicity in animal models, they are unlikely to compete successfully with glucan-synthesizing strains for prime plaque locations [28]. Thus, attention was focused on the introduction of mutations affecting acid production [19]. Although lactate dehydrogenase (LDH) mutants of the MS species Streptococcus rattus produced less lactic acid and had reduced cariogenicity in rodent models [29], this type of mutation was lethal in S. mutans. However, insertion of a Zygomonas mobilis gene encoding alcohol dehydrogenase helped overcome the LDH deficiency. The construction of the proposed replacement therapy strain BCS3-L1 involved insertion of the Z. mobilis gene into strain JH1140, followed by deletion of virtually all of the LDH gene [19]. This produced a strain that combined the traits of low acid production and strong mutacin production. Colonization trials in rats indicated that strain BCS3-L1 performed at least as well as strain JH1140 – indicating that it could also be effective in humans.

Issues still to be addressed:

1. Strain BCS3-L1 forms more plaque when grown in vitro in the presence of sucrose than strain JH1140 does, probably reflecting a pH-related effect on glucosyl transferase activity [19]. However, small increases in glucan formation in situ are not anticipated to result in significantly more plaque formation or plaque-associated disease (e.g. gingivitis).
2. Although reversion to LDH production (resulting in a strongly competitive cariogenic strain) following natural transformation of the effector strain is considered unlikely, it nevertheless constitutes a potential objection to implantation of LDH mutants. Further engineering of strain BSC3-L1 (deletion of comE) is currently being undertaken to cripple its transformation capability [27].
3. The toxicity of mutacin 1140 has not yet been directly tested but the molecule is broadly similar to the lantibiotic nisin, which has been widely used as a food preservative for decades. Furthermore, no treatment-related lesions have been detected in the organs of rats colonized with mutacin 1140 producers [19].
4. The potent mutacin output and different fermentation profile of strain BCS3-L1 could upset plaque ecology and result in the proliferation of organisms with pathogenic potential. Interestingly, however, the initial colonization studies with mutacin 1140-producing strains indicated apparent high specific inhibitory activity of the mutacin in situ. Exclusion of other MS was reported, with little or no other detectable modification to the total composition and balance of the subjects' plaque microbiota [24]. This high specificity contrasts dramatically with the broad activity spectrum of mutacin 1140 when tested in vitro [23]. These findings indicate that BLIS-producing bacteria do not necessarily eliminate all sensitive co-inhabitants in complex ecosystems, such as dental plaque. Presumably the distinct micro-habitats found within complex biofilms, and the differing physiological growth states of the inhabitants enable the individual members of heterogeneous populations to co-exist, despite incompatibilities they might display in more homogeneous environments such as laboratory cultures.