Manipulating the microbiome: evolution of a strategy to prevent S. aureus disease in children.

Hospitalized infants have the highest rates of invasive Staphylococcus aureus disease of any population and infection control strategies such as decolonization have been insufficient. For decades, researchers began studying the microbiome in search of new prevention strategies. The resident microbiota was found to be closely associated with susceptibility and at times, resistance to S. aureus colonization. The evolution of nucleic acid based techniques has enhanced our understanding of the complex relationship between the nasal microbiota and S. aureus colonization. We review what is known about bacterial communities in the nasal cavity of infants and discuss how future microbiome studies may help identify novel interventions to protect high-risk infants from S. aureus disease.

I. Introduction

Staphylococcus aureus is a leading cause of infectious morbidity and mortality in hospitalized infants and children[1]. Despite decades of research, S. aureus remains the second most common cause of nosocomial infections in neonates and children, including central-line associated bloodstream infection, surgical site infection, and ventilator-associated pneumonia [1, 2, 3, 4]. S. aureus can colonize the nasal cavity, and colonization is associated with an increased risk of subsequent infection[5, 6, 7]. Intranasal antibiotics and antiseptic agents have been used to eradicate colonization in an attempt to prevent S. aureus disease[8, 9]. The agents used for decolonization do not solely target S. aureus, but eradicate many bacterial communities with unknown long-term consequences[10]. Unfortunately, comprehensive infection control strategies including decolonization have been insufficient at preventing S. aureus disease in hospitalized patients, especially in neonates, who have the highest rates of invasive S. aureus disease of any age group [11, 12, 13, 14, 15].

Over the last 10 years, the field of microbiome science has exploded using nucleic acid based tools to characterize organisms, including S. aureus, in their ecologic niche. A more comprehensive understanding of the nasal microbiota may lead to the identification of bacterial communities, interactions, and functional characteristics that are associated with S. aureus colonization. As clinicians and researchers, we are hopeful that these findings will uncover new strategies for the prevention of S. aureus disease. Below we review 1) early studies manipulating the nasal microbiome to eradicate existing S. aureus colonization or prevent colonization following exposure, 2) how nucleic acid based techniques have enhanced our understanding of the nasal microbiota and S. aureus colonization, 3) major differences between the nasal microbiome of adults and children, and 4) how future microbiome studies may identify new strategies for the prevention of S. aureus disease.

II. Manipulating the microbiome: evolution of a strategy to prevent S. aureus disease

Some bacterial communities may produce an environment in the nasal cavity that is inhospitable to S. aureus colonization. For decades, researchers have been investigating the role that these commensal bacterial species play on a human host’s susceptibility to S. aureus colonization and disease. As a result, bacterial interference emerged as a strategy to use commensal bacterial species to eliminate or prevent the colonization of an anatomic niche by a pathogenic species [16].

IIa. Early trials of bacterial interference

In the 1960s, the clinical application of bacterial interference was spurred by several nursery outbreaks of highly virulent S. aureus strain 80/81 [9, 17, 18, 19, 20]. Light et al. found that topical decolonization with intranasal neomycin and hexachlorophene bathing failed to effectively eradicate S. aureus 80/81 carriage in these nurseries. They subsequently began inoculating all neonates admitted to the nursery with an attenuated S. aureus strain, hypothesizing that the attenuated strain 502A would occupy the nasal cavity and inhibit invasion by competing S. aureus strains [20, 21]. After two inoculation cycles with the attenuated S. aureus strain, only 1 out of 202 cultured infants tested positive for strain 80/81 [18]. This method, though successful in reducing the rates of colonization by a highly virulent strain of S. aureus, was discontinued after several neonates developed invasive infections with what was perceived as the non-virulent strain 502A [9, 17, 18, 19, 20]. Since the 1960s, there has been continued interest in the concept of nasal microbiome manipulation to eliminate existing S. aureus colonization or prevent S. aureus colonization following exposure.

IIb. Studies of bacterial interference to eliminate existing S. aureus colonization

Until recently, studies of bacterial interference to eliminate existing S. aureus colonization relied on culture-dependent techniques to characterize organism interactions. Researchers identified cultivable bacterial species that were associated with S. aureus in nasal cultures[22, 23, 24]. Species cultured from people without S. aureus colonization were then studied for their potential to eliminate existing S. aureus colonization or prevent S. aureus colonization following exposure. For example, Corynebacteria are some of the most frequently identified organisms in the nares of individuals who are not colonized by S. aureus [25]. As common commensals of the skin and mucosal surfaces, Corynebacteria have various characteristics that confer ecological fitness, which may contribute to the prevention or elimination of colonization by competing species. Uehara et al. found that Corynebacterium strain Co304 had a higher affinity to nasal mucin than S. aureus for example, which they suggest may contribute to its ability to outcompete S. aureus and eliminate carriage [25]. Some species of Corynebacteria have also been shown to decrease the gene expression of factors that are important for S. aureus colonization and virulence like surface protein A, which allows S. aureus to evade detection and elimination by a host’s immune system[26]. In one study, Corynebacterium pseudodiphtheriticum was inoculated in six MRSA colonized men and led to a decrease in abundance of S. aureus. However, as the abundance of C. pseudodiphtheriticum decreased, MRSA abundance returned to baseline level [27]. Similarly, when Corynebacterium strain Co304 was inoculated in the nasal cavities of over a dozen adults, S. aureus was eradicated in 71% of subjects and maintained in follow up periods of up to 3 years [25].

Staphylococcus epidermidis is another common commensal species found in the human nasal cavity[23]. A recent study in HIV-positive adults found that those colonized with S. epidermidis in the nose and throat were 80% less likely to be colonized with S. aureus at the same site [28]. Iwase et al. also observed that individuals who were not colonized with S. aureus were twice as likely to be colonized by S. epidermidis. This group inoculated adult S. aureus carriers with a serine protease-producing S. epidermidis strain and reported significantly reduced rates of S. aureus colonization[23]. These studies postulate that S. epidermidis may have a functional impact on the environment that makes the nasal cavity inhospitable to S. aureus. The exact mechanisms by which these common commensal organisms eliminate S. aureus colonization remain uncertain. Bacterial interference trials in humans, though few, have suggested that the bacterial species commonly found in the nasal cavity may help eliminate S. aureus colonization.

IIc. Studies of bacterial interference to prevent S. aureus colonization following exposure

The potential to prevent S. aureus colonization after a known exposure using certain bacterial strains, however, has only been studied in mouse models. Researchers have deliberately exposed naïve mice to S. aureus to assess the protective effects of other organisms. Park et al. assessed the potential of S. epidermidis to prevent acquisition of S. aureus colonization. They first treated mice with antibiotics to clear the resident microbiota, hoping that the absence of competing species would improve their success of inoculating the mice with S. epidermidis. After antibiotic pretreatment, mice were inoculated with S. epidermidis. Mice whose nasal cavities were inoculated with S. epidermidis prior to being exposed to MRSA showed reduced rates of S. aureus colonization compared to controls that were not initially inoculated with S. epidermidis, suggesting that S. epidermidis may alter its environment in the nasal cavity and make it unfavorable for S. aureus acquisition [29]. Similarly, Barbagelata et al. inoculated mice with an attenuated mutant strain of S. aureus (NK41) before exposing them to several virulent S. aureus strains isolated from humans[22]. Prior exposure to NK41 significantly reduced the abundance of virulent strains, suggesting that the mutant strain could exclude competing strains from the nasal cavity.

While the results of bacterial interference studies conducted to date are promising, most of the microorganisms that colonize humans cannot be cultivated, so many organisms in the nasal microbiota may be overlooked[30]. Although these prior studies to eradicate existing colonization or prevent colonization following exposure are limited by their reliance on conventional cultures, they suggest that the resident microbiota may play a significant role in regulating S. aureus colonization (fig 1). Nucleic acid based approaches will help to determine 1) the role of non-cultivatable organisms in resistance to S. aureus colonization and 2) whether a community of organisms, rather than a single species, is necessary to resist S. aureus colonization.

Figure 1

III. Nucleic acid based approaches provide insight into the nasal microbiota and S. aureus colonization

Until recently, researchers relied on culture-dependent techniques to identify organisms in people that did not have S. aureus nasal colonization, then tested those organisms for their potential to eliminate existing S. aureus colonization. Nucleic acid based approaches, such as 16S rRNA gene amplification and sequencing techniques, can provide a more complete and detailed characterization of the nasal microbiota and its potential for colonization resistance against S. aureus.

High-throughput sequencing studies in adults suggest that the nasal microbiota plays an important but complex role in S. aureus colonization. The abundance of certain bacterial species in the nasal microbiota is strongly associated with S. aureus carriage. A study conducted by Liu et al. demonstrated that the absolute abundance of Corynebacterium species is a predictor of S. aureus carriage in a threshold-dependent manner [31]. As the abundance of Corynebacterium species increases, the likelihood of S. aureus colonization decreases. In fact, multiple studies have found a higher relative abundance of Corynebacterium species in people not colonized with S. aureus[32, 33, 34, 35]. Other organisms such as the gram-positive facultative anaerobic cocci of genus Dolosigranulum, have been shown to negatively correlate with S. aureus carriage. The same group also demonstrated that there was a 16% rate of S. aureus colonization among individuals with a certain threshold abundance of Dolosigranulum species. Corynebacterium and Dolosigranulum species often co-colonize non-carriers, suggesting that the presence or absence of a single organism alone may not determine susceptibility or resistance to S. aureus colonization.

In addition to correlating S. aureus with other organisms in the nasal microbiota, culture-independent sequencing methods have evaluated organisms’ distribution and diversity metrics to suggest that resistance to S. aureus colonization may be multifactorial. Frank et al. showed that the nasal microbiota of healthy adults was more than twice as diverse as those of hospitalized patients and more evenly distributed[33]. They also found that the microbiota of hospitalized adults was enriched with S. epidermidis or S. aureus but lacking in Actinobacteria, such as Corynebacterium and Propionibacterium[33]. In addition, the microbiota of S. aureus carriers, has a lower biodiversity and more unevenly distributed bacterial communities than that of non-carriers; however, it remains unclear whether certain commensal species foster the dynamics often found in healthy adults or whether these dynamics favor colonization by commensal species over pathogenic bacteria [32, 36]. Microbiome studies can further inform the relative importance of biodiversity and organism composition in contributing to S. aureus colonization.

IV. Nucleic acid based approaches reveal important differences in the nasal microbiota of children and adults

Studies conducted in adults have revealed that a diverse nasal microbiota composed of certain beneficial bacterial communities appears to provide colonization resistance against S. aureus in exposed individuals[36]. However, there is a notable lack of similar data on the microbiota of neonates and children. Findings in adults cannot be extrapolated to a neonatal population for several reasons. First, the mature microbiome is more diverse than the naïve microbiome, and this diversity contributes significantly to its stability and resistance to pathogen invasion [37].

Although the mature microbiota is highly variable between individuals, it is more resistant to environmental exposures than that of infants and young children [38]. Second, the nasal microbiota of adults is composed of different organisms compared to newborns and children [39]. Both community-dwelling and hospitalized adults host a nasal microbiota dominated by organisms of the phyla Actinobacteria and Firmicutes and rarely by gram negative organisms of the phylum Proteobacteria found commonly in infants [33, 34, 35, 36, 40]. Third, studies in adults have revealed the importance of certain commensal species in the protection against colonization by pathogens like S. aureus [28, 31, 32, 33, 34, 36, 41, 42]. Infants, however, lack these commensals initially, and during the first years of life the microbiome is still dynamic and susceptible to extrinsic exposures [43]. Further studies are needed to characterize potentially beneficial communities and whether they can foster the same protective effects in neonates.

IVa. Emergence of a personalized microbiota in neonates

From birth, neonates are exposed to and subsequently colonized with a wide array of microbes. These organisms play a pivotal role in determining the succession of bacterial communities into mature stable configurations[44]. Neonates initially harbor a microbiota that is homogenous across different anatomic sites and significantly impacted by whether they are born by vaginal delivery or Caesarian-section[45]. Infants delivered by Caesarian-section are initially colonized by their mother’s skin microbiota and therefore are more likely to possess a higher abundance of Corynebacterium, Propionibacterium, and Staphylococcus species[45]. These initial colonizers interact with a neonate’s tolerant immune system, leading to the activation and suppression of different inflammatory responses[46]. The long-term impact of the resident microbiota on the immature immune system may also play a part in dictating when an infant will acquire important commensals and if they will acquire pathogens such as S. aureus [46, 47].

By 6 weeks of age, infants have developed a site-differentiated and personalized microbiota[44, 48]. At this stage, the resident microbiota varies more between different individuals than in the same individual over time[44]. Infants share many of the same organisms in their nasal microbiomes, but in variable proportions. For this reason, the nasal microbiome is often classified by the organism with the highest relative abundance, whether Streptococcus, Moraxella, Staphylococcus, Corynebacterium and/or Dolosigranulum [49]. The most abundant bacterial species indicate the overall stability and diversity of other bacterial communities in the nasal cavity[50]. A high relative abundance of certain commensal organisms, such as Corynebacterium, is associated with more diverse communities and lower overall bacterial density [47]. These characteristics promote the stability of the microbiota over time and help maintain homeostasis in the face of external disruptions[37, 47].

As the nasal microbiome develops, commensal species may play a direct role in host protection by eliminating or excluding pathogens from their niche [16]. Children whose microbiota contains commensals Corynebacterium or Dolosigranulum, for example, are less likely to harbor potential pathogens of genera Streptococcus, Staphylococcus, and Moraxella[44]. Although the specific mechanisms by which commensals exclude other organisms from the nasal cavity remain unclear, studies conducted on the immature nasal microbiota suggest that early acquisition of certain commensal species may impart long-term resistance to colonization by pathogens in newborns[16, 51].

IVb. Plasticity of the neonatal microbiome

Lower biodiversity and higher bacterial burden characterize the immature microbiota and predispose it to disruption, including dysbiosis [42, 52]. Dysbiosis, or disequilibrium, has been shown to increase the risk of colonization and potential domination of a niche by pathogenic species [53]. These changes in the microbiota composition may increase an infant’s risk of infection by a colonizing organism. A study by Hilty et al. explored the association between the nasal microbiota and acute otitis media (AOM) in infants. They show that the majority of infants with AOM caused by S. pneumoniae also had a nasal microbiota dominated by S. pneumoniae. These infants had a significantly higher bacterial burden overall and lower bacterial diversity in their nasal cavities than infants colonized by only commensal streptococcal species[47, 54]. In fact, the dominance of any potential respiratory pathogen, whether Moraxella Catarrhalis, Haemophilus influenzae, or S. aureus, in the nasal microbiome, significantly increased the likelihood of AOM with the same organism [47].

The same qualities that contribute to a vulnerability to pathogen invasion also represent an opportunity for successful colonization by protective commensal species. The nasal microbiome of infants is dynamic and receptive to new organisms [55]. A high relative abundance of certain commensal species is associated with greater biodiversity and lower bacterial burden [47]. Both are characteristics shown to promote stability and subsequently, colonization resistance.

A newborn’s naïve nasal microbiome is highly sensitive to environmental exposures [55]. Breastfeeding, for example, is associated with a nasal microbiome dominated by commensal Corynebacterium and Dolosigranulum species. Formula-fed infants are significantly more likely to harbor profiles dominated by potential pathogens such as S. aureus [48]. Adopting habits that increase rates of colonization by potentially protective bacterial species is not always feasible, however. Antibiotics, for instance, are among the many clinical exposures related to a significant reduction in colonizing commensal species in hospitalized infants and children [47]. The identification of species and the mechanisms by which they exclude S. aureus from the immature nasal microbiome could potentially make other interventions available for high-risk infants.

V. Microbiome studies may identify new strategies to prevent S. aureus disease

To date, scientists’ understanding of the association between the nasal microbiome and S. aureus colonization is just the tip of the iceberg. Nucleic acid based studies are revealing that the relationship between the microbiota and S. aureus colonization is more complex than once thought. The microbiota seems to play an important role in colonization resistance, excluding S. aureus from colonizing the nasal cavity. Microbiome studies using nucleic acid based approaches extend the range of detectable organisms and facilitate genus and even species-level identification. Improving the ability to detect and distinguish bacterial species is essential for studying the association between S. aureus colonization and specific commensal species. The use of nucleic acid sequencing also allows for a more precise measure of organism composition and abundance. This has led to the development of new ways to measure biodiversity and population distribution that may help show the association between S. aureus colonization and overall dynamics of the nasal microbiota [56]. In fact, some studies have suggested that microbiota can be categorized in community types based on the organisms with the highest relative abundance[31]. These community types may potentially identify people at a higher risk for S. aureus colonization and infection despite vast individual variations in the nasal microbiota. In addition to community structure, sequencing enables the detection of genes encoding certain functional attributes of bacterial communities such as the ability to synthesize anti-microbial substances, possibly elucidating the mechanisms behind colonization resistance. Future functional analyses may also reveal how the interdependence of gene expression in a microbial community may be employed to attenuate S. aureus virulence. Burnham et al. suggest that interventions aimed at producing colonization resistance to S. aureus in humans may need to involve more than just one bacterial interference mechanism[57]. Introducing a community of organisms found in healthy non-carriers to the nares of high-risk infants could potentially help establish a temporally stable microbiota less susceptible to colonization by pathobionts like S. aureus. While hospitalized neonates and infants are especially at risk for S. aureus disease, there is a lack of studies on the relationship between the immature nasal microbiota and S. aureus colonization. The increasing accessibility of nucleic acid sequencing technologies may help to inform such studies and to make safer and more targeted prevention strategies available for high-risk infants.

VI. Conclusion

S. aureus remains a leading cause of infectious morbidity in infants and children. Prior studies suggest that the nasal microbiota is closely associated with S. aureus colonization, and that certain resident commensal bacteria may be capable of excluding S. aureus from the nasal cavity. Sequencing technologies may help identify bacterial communities and the mechanisms by which they exclude S. aureus from the immature nasal microbiome and lead to novel targeted interventions to protect high-risk infants from S. aureus disease.