Probiotics to prevent Staphylococcus aureus disease?

ABSTRACT

There are a plethora of probiotic formulae that supposedly benefit human health on the market. However, the scientific underpinnings of the claimed benefits have remained poorly established. Scientific evidence is now increasingly being provided that explains those benefits, for example, by immune-stimulatory effects or inter-bacterial competition between beneficial and pathogenic bacteria. In our recent study (Piewngam et al. Nature 2018), we show that Bacillus colonization of the human intestine is negatively correlated with that of the human pathogen, Staphylococcus aureus. This type of colonization resistance is achieved by secretion of a class of lipopeptides by Bacillus species that inhibits S. aureus quorum-sensing signaling, which we found is crucial for S. aureus intestinal colonization. Here, we discuss what these findings imply for the general role of S. aureus intestinal colonization, the role of quorum-sensing in that process, and potential alternative ways to control S. aureus infection.

Introduction

It is now widely acknowledged that the ensemble of microorganisms that colonize human epithelial surfaces, the human microbiota, has a key role in determining human health. The number of microorganisms in the human gut exceeds by far that on any other body part, and intestinal colonization has, therefore, received the most attention.¹ Some of the bacteria that colonize the human gastrointestinal tract are of considerable benefit to the host, while others may cause severe infections.² A major reason for the expansion of pathogenic populations or establishment of foreign introduced pathogens, which ultimately can lead to infection, is the eradication of the healthy human microbiota by antibiotic consumption.³ Infection with Clostridium difficile (CDI), a leading cause of healthcare-associated morbidity and mortality worldwide, is a posterchild example of disease that is caused by the depletion of intestinal microbiota by antibiotics.⁴

The contribution of the microbiota to host defense against pathogen colonization and overgrowth is called “colonization resistance” and it has been suggested to exploit mechanisms of colonization resistance by next-generation probiotics consisting of beneficial commensal bacteria.⁵,⁶ However, there are no FDA-approved colonization resistance-based therapies as of yet, which is mostly due to the fact that the mechanisms underlying bacterial interactions that lead to such resistance have remained poorly understood. Generally, the marketing and use of probiotics have far outpaced the scientific investigation of their alleged effects.

Colonization resistance exerted by Bacillus against Staphylococcus aureus

Staphylococcus aureus is a dangerous human pathogen that can cause a variety of infections ranging from minor skin infection to life-threatening disease and fatal pneumonia and sepsis.⁷ Emergence and spread of methicillin-resistant S. aureus (MRSA) is particularly worrisome given the high rate and morbidity of MRSA infections in hospitals and the community.⁸ This situation is exacerbated by the lack of a working S. aureus vaccine, calling for the investigation of alternative treatment methods for this pathogen.

The predominant source for infection by S. aureus is previous asymptomatic colonization.⁹,¹⁰ Traditionally, the human nasal cavity has been considered the primary habitat for S. aureus.¹¹ Approximately 20–30% of the healthy population are persistently colonized with S. aureus in the nose¹²,¹³ and S. aureus nasal colonization is a well-established risk factor for infection.⁹ However, S. aureus is also commonly found on other body sites such as the skin, throat, axillae, vagina, and especially the intestine.¹¹ More recent studies have shown that intestinal colonization with S. aureus can persist following cessation of antibiotic treatment and forms an important reservoir for outbreaks of infectious S. aureus disease,¹⁴,¹⁵ which can perpetuate the spread of S. aureus in hospital wards and the community. Furthermore, S. aureus is a leading factor of food poisoning¹⁶ and has been discussed as a potential source of antibiotic-associated diarrhea.¹¹ Nevertheless, as compared to the many studies that have addressed the molecular underpinnings of S. aureus nasal colonization, intestinal S. aureus colonization has remained severely understudied.

To investigate whether there are mechanisms of colonization resistance that can potentially be exploited to limit S. aureus intestinal colonization as a source for S. aureus infection, we collected fecal samples from 200 healthy individuals from rural Thai populations and performed 16S rRNA metagenomic sequence analysis to detect potential differences in the intestinal microbiota between S. aureus carriers and non-carriers.¹⁷ Nasal swabs were also taken. Of note, the observed rates of S. aureus nasal (13%) and intestinal (12.5%) colonization were considerably lower than those reported in the many previous studies that investigated primarily urbanized Western populations.¹⁸ To our initial disappointment, we detected no significant differences in the microbiome composition between S. aureus carriers and non-carriers using this 16S rRNA sequencing approach. However, using culture-based analysis we found a strong correlation between the presence of Bacillus spp. in the gut and the absence of both intestinal and nasal colonization by S. aureus: Fecal samples that grew Bacillus never grew S. aureus, and vice versa. That we could not detect this correlation using 16S rRNA sequencing-based analysis is not surprising given that such sequencing-based analyses are set up to detect high-order taxonomic shifts rather than specific differences on the species or genus level. This emphasizes the value of traditional, culture-based analyses to detect bacterial interactions even in the era of sequencing-based microbiome research.

In intensive follow-up analyses, we deciphered the mechanism underlying the observed colonization resistance (Figure 1). We showed that Bacillus species produce a family of lipopeptides, the fengycins, that have a strong capacity to inhibit the S. aureus quorum-sensing system Agr,¹⁹ which we revealed is indispensable for S. aureus colonization of the gut. Fengycins are cyclic lipo-decapeptides containing a frequently β-hydroxylated fatty acid with a sidechain length of 16–19 carbon atoms. They are known to specifically inhibit the growth of filamentous fungi while generally showing no antimicrobial activity against yeast and bacteria.²⁰ The cyclic fengycin structure is reminiscent of the thiolactone ring structure of Agr extracellular signals (called AIPs for autoinducing peptides);^21,22 and in fact, we showed that they work as competitors of AIP binding to their membrane receptor, AgrC, in all S. aureus Agr allelic groups (I–IV). Notably, feeding fengycin-producing (but not fengycin-deficient) Bacillus spores to mice completely eradicated S. aureus intestinal colonization. These results indicate a potential value of probiotic Bacillus to eradicate S. aureus colonization and of fengycins as quorum-sensing blockers for S. aureus infection. They also give new insight into the role of staphylococcal quorum-sensing for intestinal colonization and the relative role of different S. aureus colonization sites. We will discuss these implications in the following.

Figure 1.

Inhibition of Agr quorum-sensing in S. aureus by Bacillus. Bacillus ssp. produce fengycin lipopeptides, which resemble the Agr auto-inducing peptide (made from the AgrD precursor by AgrB) and compete with the AIP for binding to the AgrC receptor, part of the AgrC-AgrA two-component system. This blocks signal transduction, which is dependent on AgrA phosphorylation and binding of AgrA to specific promoters, which include that driving agrABCD expression in an auto-regulatory fashion and toxin gene promoters in a direct or indirect manner. As a consequence, there is a diminished capacity to produce toxins and colonize the intestinal tract.

Relationship between S. aureus nasal and intestinal colonization

Despite several studies that have emphasized the widespread and considerable extent of S. aureus intestinal colonization,¹⁸ it has often been regarded as secondary to nasal colonization and possibly initiated by the nasopharyngeal-to-intestinal passage. It has also been speculated that the presence of S. aureus in the feces, which is commonly used to determine the intestinal presence, is caused by colonization of the anorectal skin area rather than the intestine. However, adhesion of S. aureus to intestinal mucus has been shown directly in human tissue²³ and results from mouse studies by us and others are in support of the notion that there is genuine intestinal rather than only anorectal S. aureus colonization.^17,24

Our results showing that an S. aureus colonization resistance mechanism that is only present in the intestine also strongly impacts nasal colonization calls for a critical re-evaluation of the notion that the nose is pivotal to S. aureus colonization. In fact, one possible explanation for our results is that nasal S. aureus colonization is secondary and transient and the seemingly permanent nasal colonization is a consequence of frequent re-inoculation from a permanently S. aureus-colonized intestine by anorectal-nasopharyngeal transmission. A more central than previously assumed role of the intestine for general human colonization by S. aureus is not without further support. In addition to a series of reports suggesting that intestinal carriage is an important risk factor for S. aureus infections,^14,15,18 it has been shown that mice challenged intravenously with S. aureus develop intestinal staphylococcal colonization, and fecal shedding results in S. aureus transmission to cohoused naïve mice. Interestingly, patients who were both S. aureus nasal and intestinal carriers were significantly more likely to develop S. aureus infection than were those with nasal carriage only.¹⁵ Furthermore, intestinal S. aureus colonization has been suggested to explain the failure of topical decolonization measures aimed solely at the nose.^13,18 Moreover, staphylococcal skin and soft tissue infections are linked to rectal, but not nasal, colonization by MRSA strains in children.²⁵ Finally, the sheer size of the intestinal as compared to nasal colonization site indicates that there are many more S. aureus bacteria living in the intestine than the nose in a given individual, suggesting a much more extended reservoir for further spread. Altogether, these findings suggest that the intestine rather than the nose may be the dominating habitat of S. aureus in humans from which re-colonization of other sites initiates. Ultimately, this hypothesis needs to be verified directly in humans.

Role of the S. aureus agr quorum-sensing system in intestinal colonization

The molecular factors underlying asymptomatic S. aureus colonization are much less well studied than those causing infection. This is in part due to the lack of animal colonization models that closely reflect the human situation. The mouse may be suited for the investigation of intestinal colonization at least to some extent, but the difference between mouse and human skin and also the nasal cavities are quite pronounced. To study nasal colonization, cotton rats have been used with somewhat more success. Overall, however, only a very limited list of molecular factors that impact S. aureus colonization of the nose and intestine have been identified, and the skin – while frequently investigated as a site of S. aureus infection – has remained virtually unexplored in that regard. These factors include teichoic acids²⁶ and a series of cell surface-exposed binding proteins such as ClfB.^27–29 The latter are believed to connect the bacteria to the nasal epithelium. Of note, mutants lacking clfB showed no defect in intestinal colonization,³⁰ suggesting that the mechanisms underlying nasal and intestinal colonization by S. aureus differ significantly.

The Agr quorum-sensing system, while clearly regulating infectivity,^19,31,32 had not been implicated in asymptomatic colonization before our study. In competitive experiments with equal amounts of S. aureus wild-type and isogenic agr mutant strain, we observed a dramatic impact of agr on intestinal colonization, with only wild-type S. aureus detected in the feces and the intestine. Furthermore, in a non-competitive experimental set-up, only those bacteria expressing the intracellular Agr effector RNAIII achieved colonization; agr-negative control strains never did. This extremely strong impact of Agr on intestinal colonization was quite unexpected. Commonly regarded as a negative regulator of surface binding proteins,³² Agr had not been among the likely candidates for colonization factors. However, we showed recently that in widespread clinical strains such as USA300, the notion that Agr reduces expression of surface-binding proteins does not universally hold true, and in particular the fibrinogen protein ClfA, which has been implicated in intestinal colonization,³⁰ is in fact positively regulated by Agr in that strain.³¹ Then again, it is quite unlikely that fibrinogen binding plays a role for colonization of at least the healthy intestine and ClfA regulation by itself can explain the dramatic impact of Agr on intestinal colonization that we observed. Furthermore, we found that effect in three unrelated strains, suggesting that it is a more general feature of S. aureus. Altogether, the Agr-controlled factors that drive the impact of Agr quorum-sensing on intestinal colonization yet remain unidentified.

Potential application of Bacillus probiotics and fengycins to control S. aureus infection

There are at least two possible routes for the translational use of our findings. First, live Bacillus may be used as a probiotic regimen to eradicate S. aureus colonization as a source for infection. Second, the fengycin quorum-sensing inhibitors may be used in synthetic or purified form as a quorum-sensing blocking drug directly against S. aureus infections (Figure 2).

Figure 2.

Potential therapeutic applications of Bacillus quorum-sensing blocking activity. Left, live Bacillus (subtilis) probiotics given orally in spore form germinate in the intestine to produce metabolically active cells that inhibit S. aureus quorum-sensing and thereby S. aureus intestinal colonization. Given the key role of intestinal colonization that our study implied, this may also lead to general decolonization including of the nose. Right, pure fengycin preparations may be given to control infection (e.g., lung, blood, or skin infection) in an anti-virulence quorum-sensing blocking approach.

Bacillus subtilis spores are a frequent component of probiotic formulae and also available as a stand-alone probiotic.³³ A probiotic approach with B. subtilis would offer several advantages over other methods of decolonization. First, it is not antibiotic-based as the other methods that have been used to decolonize S. aureus, and thus does not come with the disadvantage of the extensive antibiotic use that would be necessary to decolonize a large subset of individuals or patients. In particular, for the decolonization of the intestine, antibiotic use is severely problematic due to the risk of infection with C. difficile and other pathogens that benefit from a broad eradication of the intestinal microbiota. Second, the Agr system is generally absent from Gram-negatives and not present in beneficial members of the human gut microbiota, while its presence in some other pathogenic bacteria³⁴ may suggest even broader applicability of the Bacillus probiotic-based approach or the quorum-sensing blocker approach discussed below. Thus, the probiotic approach offers pathogen-targeted eradication that is not possible with antibiotics. Third, live Bacillus constantly produces the active substance, providing for a maintained presence that cannot be achieved even with multiple applications of drugs. Still, Bacillus is not a permanent colonizer of the intestine and our mouse data suggest that the probiotic would have to be given daily. Heterologous production in a permanent beneficial colonizer is an alternative approach to consider achieving more permanent production of the inhibitor without the need for repeated oral application; yet, the permanent establishment of a genetically modified organism in the gut is likely not well received by many people. Fourth, the specific use of Bacillus spores as a highly resistant form of a probiotic organism also has the advantage of improved stomach passage – a considerable problem with many other probiotic supplements. After passage through the stomach, the spores germinate to produce metabolically active cells in the intestine.³⁵ Lastly, Bacillus subtilis is generally considered safe and also reported to have additional probiotic benefits, which however except for a specific mechanism of immune stimulation in the intestinal epithelium³⁶ remain poorly established scientifically.

While our results indicate that intestinal S. aureus eradication would have as a consequence a general decolonization of the human body, in principle a probiotic approach of interference could also directly be applied to the nose, if two prerequisites are met: (i) a beneficial colonizer can be found for heterologous expression of fengycins (or other S. aureus quorum-sensing blockers), and (ii) Agr proves to be as crucial for colonization in that body site as it is in the intestine. We believe such an approach to be superior to antibiotic- and also bacteriocin-based approaches that have been proposed recently,³⁷ as a much more specific eradication of the pathogen can be achieved. This is hardly possible with antibiotics or bacteriocins that commonly have a broad target spectrum and also kill beneficial colonizers. We need to add that our current data indicate that other, potentially beneficial staphylococci are also targeted by fengycins, which is not a likely problem for the Bacillus-mediated intestinal application, but would have to be considered in approaches that target body sites such as the skin or nose where S. epidermidis and other staphylococcal species are abundant and supposedly play an important role in the stability of the healthy microbiota.

The other possible route of translational use comprises the application of fengycins as quorum-sensing blockers in an anti-virulence-based approach to control S. aureus infection. This is based on the fact that many S. aureus toxins, including phenol-soluble modulins (PSMs), Panton-Valentine leucocidin (PVL) and other leukocidins, as well as α-toxin, are Agr-controlled and significantly impact S. aureus infection.³⁸ This approach is often described as being less prone to the development of resistance as compared to traditional antibiotics, because genuine anti-virulence drugs do not kill the bacteria.³⁹ However, the respective studies on resistance development are commonly only performed in vitro, and whether this holds true in an in-vivo setting with the selective pressure from efficient host defenses is debatable. Furthermore, many studies on quorum-sensing blocking drugs have not adequately ruled out growth effects that overlie the anti-virulence effect. Moreover, blocking quorum-sensing may not be advantageous in all infection types. Quorum-sensing blockers may be counterproductive in particular in biofilm infections.⁴⁰ As for blood infections, the role of Agr is not entirely clear. While Agr-negative strains can be isolated at an increased percentage from patients with persistent bacteremia,⁴¹ strong contributions of Agr and Agr-regulated toxins to mortality in experimental sepsis argue in favor of a significant role of Agr in that disease type.^42,43 Generally, quorum-sensing blockers are commonly seen as a valid alternative to conventional antibiotics that is worth considering at least in acute types of infection and in combination therapy with antibiotics. For S. aureus, this would include lung infections, and likely sepsis, as the most common causes of mortality due to S. aureus, as well as severe skin infections.

While still only in pre-clinical development, several quorum-sensing blockers have been described for S. aureus.⁴⁴ Most of those, such as savirin, target the intracellular response regulator AgrA, an essential component of Agr.⁴⁵ Fengycins, in contrast, work as competitive inhibitors of the extracellular Agr signal, and thus are active in the extracellular space. There are pros and cons associated with either type. Extracellularly active substances do not need transport into the cell and can thus be less hydrophobic, a considerable pharmacokinetic advantage. On the other hand, competitive inhibitors usually need to be applied in higher amounts. In that regard, the purified main fengycin species β-hydroxy-fengycin B had considerable activity in our study. However, a future detailed structure-function analysis of fengycins as Agr inhibitors and medicinal chemistry may lead to further improvement of activity.

Concluding remarks

While rigorous scientific evidence underlying the health benefits provided by probiotics is generally rare, those have been explained by several mechanisms, including modulation of the immune system, enhancement of the intestinal epithelial barrier, competition of bacteria for nutrients, and bacteriocin-mediated interference.⁴⁶ Our study added direct interaction via a signaling system to this list and provided evidence obtained in humans indicating that a healthy diet that includes unsterilized vegetables is associated with the absence of a major pathogen. These findings suggest several translational approaches to limit S. aureus infection, among which the most elegant and promising appears the administration of a probiotic to limit intestinal and possibly general colonization with S. aureus. Whether this works and also can be extended to limit infection in hospital settings will be the subject of future clinical trials. Finally, whether fengycin-based therapy can be used for other pathogens that have Agr-like systems, such as C. difficile,⁴⁷ is the subject of current investigation.

Funding Statement

This work was supported by the Division of Intramural Research, National Institute of Allergy and Infectious Diseases [ZIA AI000904-16].

Acknowledgments

This work was supported by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases (NIAID), U.S. National Institutes of Health (NIH).