The State of Microbiome Science at the Intersection of Infectious Diseases and Antimicrobial Resistance

Abstract

Along with the rise in modern chronic diseases, ranging from diabetes to asthma, there are challenges posed by increasing antibiotic resistance, which results in difficult-to-treat infections, as well as sepsis. An emerging and unifying theme in the pathogenesis of these diverse public health threats is changes in the microbial communities that inhabit multiple body sites. Although there is great promise in exploring the role of these microbial communities in chronic disease pathogenesis, the shorter timeframe of most infectious disease pathogenesis may allow early translation of our basic scientific understanding of microbial ecology and host-microbiota-pathogen interactions. Likely translation avenues include development of preventive strategies, diagnostics, and therapeutics. For example, as basic research related to microbial pathogenesis continues to progress, Clostridioides difficile infection is already being addressed clinically through at least 2 of these 3 avenues: targeted antibiotic stewardship and treatment of recurrent disease through fecal microbiota transplantation.

This special issue highlights recent work exploring the indigenous microbiota, demonstrated through the direct and indirect effects it has on the pathogenesis of infectious diseases and antibiotic resistance. We discuss translation of basic scientific understanding that has advanced the diagnosis, prevention, and treatment of disease. The articles are thematically organized into chapters of related science, based on the convergence of common principles and the state of scientific knowledge. We hope this supplement provides clinical context for researchers studying the microbiome and provide insight into potential novel strategies for clinical care based on understanding our indigenous microbiota.

THE CENTRAL ROLE OF THE MICROBIOTA IN PATHOGEN RESISTANCE AND HOST TISSUE HOMEOSTASIS AND IMMUNITY

In the 150 years since the formulation of Koch’s postulates, microbes were primarily recognized as the cause of significant illnesses. Efforts in understanding the pathogenesis of infectious diseases centered on virulence characteristics intrinsic to microbial pathogens isolated in pure culture and their respective individual interactions with host defenses. Although it was recognized that the human body is also normally inhabited by large and complex microbial communities, these organisms were considered to be “commensals,” ie, organisms that only take advantage of food sources provided by the host without potential benefit or harm to the host. However, we are increasingly aware that the indigenous microbiota are not simply inert freeloaders but instead form a complex host-microbe system benefiting both partners (ie, mutualism). As such, infectious disease pathogenesis reflects intertwined interactions between the pathogen, the host, and the indigenous microbiota (Figure 1). As this complexity is more appreciated, it has become clear that studying this entire system and its contextual relationships can improve our understanding of infectious disease pathogenesis.

Figure 1.

The complex interactions between pathogen, host, and indigenous microbiota. There are multiple bidirectional interactions that influence the outcome of infection and other host inflammatory or immune-related processes. The microbiota appears to play a central role in influencing both host and pathogen behavior thus necessitating a system-level approach that improves our understanding of disease pathogenesis.

The concept of colonization resistance is a prime example of how understanding such relationships between host, indigenous microbiota, and pathogens benefit human health. The idea that indigenous microbiota could provide a barrier to colonization by pathogenic microbes (ie, colonization resistance) originated in the 1950s and was further advanced by van der Waaij et al in the 1970s [1]. When initially proposed, the focus was on the interaction between indigenous microbes and bacterial pathogens. However, our modern concept of colonization resistance extends to other microbial pathogens and also recognizes the cooperative nature of the host and indigenous microbes in defending against pathogens [2, 3]. This contemporary view of colonization resistance forms the basis for investigating and understanding the pathogenesis of several bacterial and viral pathogens (eg, Salmonella, Escherichia coli, Vibrio cholera, vancomycin-resistant enterococci, Clostridioides difficile, norovirus, poliovirus and reovirus). This new vantage point places the human microbiota in a critical role, one that mediates host-pathogen interactions through mechanisms that involve direct inhibition, metabolic competition, and maintenance of host physiology including barrier function and immunomodulation. The first 2 articles in this supplement explore the central topic of colonization resistance and transkingdom interactions in the context of bacterial and viral pathogens.

In the article by Pike and Theriot [4], the authors describe a subset of the microbiota-mediated mechanisms underlying susceptibility to C difficile infection (CDI). They describe how antibiotic use and other factors contribute to the loss of colonization resistance, C difficile growth, and the expression of virulence factors that lead to the development of disease. The authors suggest that advances in our mechanistic understanding of colonization resistance are key to developing new strategies to decrease CDI-associated morbidity and mortality. The research described in this article is at the forefront of the field and has produced clear targets and novel approaches for future prophylactic and therapeutic development.

Much of what we know about the composition of the indigenous microbiota comes from gene marker studies involving the 16S rRNA gene, which is highly conserved across bacteria species. However, the increasing use of metagenomics to more fully characterize microbial communities (and their collective genome) has revealed associations among bacteria, fungi, parasites, and viruses that appear important in the context of infectious disease. In the article by Nishimoto et al [5], the authors review the landscape of transkingdom interactions important for viral pathogenesis and host immunity to viral pathogens. Studies of these interactions, in the context of animal models, has provided proof-of-concept that modulating indigenous gut bacteria is a viable strategy for reducing the pathogenesis of enteric viruses. Even more interesting is that certain bacteria-virus interactions in the lung appear to be synergistic, in which the presence of one pathogen can directly influence the other. The fundamental knowledge and mechanistic insight from these cross-cutting studies is growing and may provide new targets (viral, bacterial, or host) for therapeutic intervention. However, the more immediate impact will likely be in the realm of treatment choices (ie, antibiotic selection) based on a more complete understanding of the microbial community interactions that can influence viral pathogenesis and host immune responses.

THE CONVERGENCE OF ANTIMICROBIAL RESISTANCE AND THE MICROBIOME

Antibiotic resistance is a leading threat to public health [6]; the US National Action Plan for Combating Antibiotic-Resistant Bacteria includes goals and priorities that embrace the microbiome’s role in resistance development [7]. Furthermore, Anthony et al [8] discuss the gut microbiome as an important reservoir for antimicrobial resistance genes (ARGs). The authors argue that antimicrobials can disrupt colonization resistance, which leads to colonization by pathogens, increased load, dissemination of ARGs (ie, resistome), and, eventually, clinically significant infections. However, many of the studies correlating increased ARG acquisition with the risk of infection fail to mechanistically link the 2 outcomes, and this represents an important knowledge gap in the field. Part of the difficulty in making mechanistic inferences is technical in nature (ie, short read sequencing technology is insufficient), but it is also true that human microbiome data can be extremely complex, highly variable over time, and almost exclusively derived from stool as part of ongoing observational studies. Golob et al [9] tackled this topic and posit that data based on 16S gene sequence retrieval, although not optimal, can be applied in a complementary fashion along with new analytic approaches and a conceptual framework of causal inference. They suggest that a holistic “systems” approach, involving specific analytic strategies (eg, dimensionality reduction and machine learning), new in vitro models, and proper study design, should be used to generate testable hypotheses that can be applied to future human cohort studies.

The combined picture that emerges from these first 4 articles is that therapeutically restoring colonization resistance in the context of recurrent CDI can provide a new, microbiome-based, treatment option that does not contribute to the problem of increasing antibiotic resistance. However, in the context of difficult-to-treat antibiotic-resistant infections, we are still making correlations, lacking mechanistic insight, and learning how to demonstrate patient benefit. This knowledge gap can be addressed through multidisciplinary, team science, where microbiologists and infectious disease physicians work with computational biologists to study disease pathogenesis on a “systems” level. This need to embrace complexity becomes more apparent in the study of pneumonia and other acute diseases in which host-microbiota interactions play a more central role and the inciting pathogen (or pathogens) is not always the primary focus.

THE CENTRAL ROLE THE MICROBIOTA PLAYS IN ACUTE DISEASES AT MULTIPLE BODY SITES

Microbiota harbored at different anatomic sites across the human body differ both in bacterial composition and function. The composition and function can change significantly over time depending on host physiology, environmental exposures, etc, all of which can have a significant impact on the host-microbe interactions. In the article by Tuddenham et al [10], the authors discuss the cervicovaginal microbiota, which is less complex than the gut microbiota and typically dominated by 1 or more species of Lactobacillus. They further describe how the vaginal microbiota has been both mechanistically and epidemiologically associated with protection against, and risk for, sexually transmitted infections ([STIs] eg, human immunodeficiency virus, gonorrhea, chlamydia, trichomoniasis). It is interesting to note that a vaginal microbiota comprising more phylogenetically diverse species of anaerobic bacteria is often associated with clinically diagnosed bacterial vaginosis (BV), a condition that can be associated with significant morbidity and increased risk of STIs. Tuddenham et al [10] discuss, in detail, the unique compositions of the male and female urogenital microbiota, the associations with STIs, as well as a form of colonization resistance that appears to mediate host protection and susceptibility to STIs. The authors also discuss important knowledge gaps in the field and considerations for developing microbiome-based STI prevention strategies, including an exciting new study involving a live biotherapeutic product ([LBP] LACTIN-V); work that is further highlighted later in the supplement in a piece by Lagenaur et al [11].

Studies of the gut and vagina have prompted researchers to investigate the potential role of microbial communities in the pathogenesis of other conditions at varied anatomic sites. This includes infectious diseases where potential pathogenic agents are variable or remain elusive (eg, pneumonia, cystic fibrosis [CF], necrotizing enterocolitis [NEC], and sepsis). Unlike the urogenital or gastrointestinal systems, the pulmonary system, including specifically the lungs below the glottis, was historically thought to be largely sterile and harboring only transient microbial residents that were present as part of the respiratory process. The article by Ethridge et al [12] articulates the more contemporary view of the upper and lower airways and how the entire airway system is home to stable microbial communities. These microbes function to maintain lung homeostasis through important transkingdom communication mediated, in part, by microbially derived metabolites. Ethridge et al provide a detailed review of the myriad of microbiota-derived metabolites. They describe the importance of the host-microbe dialog in maintaining mucosal immunity with a particular focus on immunomodulatory molecules that impact epithelial cell physiology (eg, barrier function), antigen presentation, and cellular immunity. This fine tuning, the authors argue, is mediated through ligand-receptor interactions that are critical for rapidly mobilizing a response to an invading pathogen as well as for maintaining mucosal tolerance to allergens and the variations in resident microbes.

The new ecological view of the lungs is gaining traction, particularly in the context of pneumonia where a single inciting pathogen is not routinely cultured or identified. In this context, the use of empiric broad-spectrum antibiotics will continue to drive increasing rates of antimicrobial resistance and eventually lead to high rates of treatment failure. The piece by Pettigrew et al [13] tackles this topic and summarizes a new conceptual framework for pneumonia pathogenesis in which the lung microbiota plays a critical role. The authors explore whether culture-independent research on these hard-to-reach microbial populations can be used to improve the diagnosis, prevention, and treatment of pneumonia. Knowing the identity and source of pneumonia will help guide new prevention and treatment strategies, including the value of microbiome-targeted interventions for pneumonia. Pettigrew et al also touch on the relationship between respiratory viruses, the lung microbiota, and pneumonia, an area of great interest, given the coronavirus disease 2019 pandemic.

Culture-independent sequencing has also been applied to assess the microbiome of sputum from patients with CF. Patients with CF have decreased lung mucociliary clearance and often suffer from chronic polymicrobial infections. Expectorated sputum, whereas not ideal, can serve as a proxy for the lung microbiota and is a noninvasive way to collectively assess resident microbes. In a new primary research article by Zhao et al [14], the investigative team leverage this concept and hypothesize that the inclusion of nonpathogen data can improve the prediction of lung function in CF patients. The authors generated quantitative 16S rRNA data from a cohort of 77 CF patients and applied it, along with data from patient electronic medical health records, to train machine learning models that predicted lung function. It is not surprising that models trained on pathogen data were consistent with the traditional view of CF microbiology, that CF pathogens (ie, Achromobacter, Pseudomonas) are the major drivers of patient health. However, the inclusion of quantitative 16S community data improved the accuracy of the models and found that certain taxa (ie, Rothia and Fusobacterium) were positive predictors of lung function. This new research, taken together, supports the concept of an ecological framework for studying infections of the lungs and highlights the challenges and opportunities to augment traditional microbiology and advance the field.

Necrotizing enterocolitis and sepsis are 2 conditions characterized by the rapid onset of systemic inflammation and subsequent dysregulation of the host response. Although not related, recent evidence suggests that both conditions can be driven by aberrant host-microbial community interactions in which an altered microbiota contributes to a feedback loop that exacerbates the condition and results in tissue injury.

Preterm birth is the primary known risk factor for NEC, but the underlying cause appears to be multifactorial and includes an immature immune system, overall gut function, and abnormal microbial colonization. In the review by Thanert et al [15], they discuss how next-generation sequencing studies from high-risk cohorts find “compositionally distinct and less diverse microbiota” before NEC onset. As noted by the authors, this concept is not new, but recent data from various cohorts support the hypothesis that overrepresentation of Gammaproteobacterial/Proteobacteria (ie, Enterobacteriaceae) and underrepresentation of certain anaerobes is a common microbiome signature of NEC. Thanert et al suggest that future studies should focus on developing robust classifiers from multiomic profiling of both host and microbiota in high-risk cohorts.

Sepsis, initiated as a result of infection and driven by an aberrant host response, is a leading cause of mortality worldwide [16]. Until recently, an immune-centered approach has been applied in the study of sepsis; however, in the review by Miller et at [17], a new paradigm for understanding the septic response is proposed, and the microbiota appears to play a key role. The authors cite evidence which demonstrates that major disruptions to the microbiota occur before the onset of sepsis, and that treatment for sepsis itself (eg, antibiotics and artificial nutrition) can further drive this disruption. The resulting gut “pathobiome” can lead to impaired host immune function, polymicrobial dissemination via compromised tissues, and the development of organ system dysfunction. The authors describe this as a vicious cycle that begins with infection as the primary “insult” and is further amplified through sepsis treatment. They argue this new paradigm can inform novel approaches to sepsis care, one that drives recovery through interventions that restore a healthy microbiota.

An interesting concept put forward in the Thanert et al and Miller et al articles, is that emergence of a pathogenic microbiota, is key to systemic disease. This is in stark contrast to the beneficial role that indigenous microbes play during health. It also reinforces the host and microbiota relationship that can lead to new insights regarding disease pathogenesis. It remains to be seen whether the microbiota can be used as a therapeutic, given the lack of clear targets and the acute nature of the disease. Perhaps advances in the area of microbiome-based clinical predictors and diagnostics will enable progress with these devastating conditions.

MICROBIOME-BASED CLINICAL PREDICTORS AND DIAGNOSTICS

Over the past decade, there have been significant advances in our ability to characterize, quantify, and infer the functional capacity of the human microbiota. Futhermore, with these advances comes the desire to leverage this information to detect or predict disease. More importantly, the ability to diagnose or predict disease risk can occur in the absence of a complete understanding of causal mechanisms. Damhorst et al [18] provide a comprehensive and insightful view of the various methods used to analyze the microbiome and how this can have diagnostic value. The authors describe each method of microbiome characterization, from traditional culture to quantitative microbiome profiling, and provide concrete examples of how this information may be used to inform clinical decisions, including in the context of infectious diseases and antimicrobial resistance. The outlook from Damhorst et al suggests that traditional microbiome-based diagnostics (ie, culture and quantitative polymerase chain reaction) will continue to have a place in the clinical microbiology setting and that untargeted characterization of the microbiome, when coupled with advanced analytic tools such as metabolomic profiling approaches, will lead to more precise and reliable diagnostic information. Moreover, although the use of microbiome or metabolome data to predict or diagnose disease has been advanced with technological innovations, many significant hurdles remain, including the following: (1) the need for validated benchmarks indicative of a healthy, fully functional microbiota; (2) the need for application-specific wet laboratory and bioinformatic standards; (3) an enhanced understanding of the microbiome-based causality for a particular disease state. Only when these issues are addressed can we engage in prospective validation of signatures for microbiome-mediated disease states and realize the full potential of microbiome-based diagnostics.

LIVE BIOTHERAPEUTIC PRODUCTS AND MICROBIOME-BASED THERAPEUTICS

Intentional manipulation of the gut microbiome, for the purpose of curing disease, is likely one of the most ancient forms of therapy. The process of purposely transferring gut microbes through the consumption of fecal matter can be traced back approximately 3000 years and was used to cure stomach-related disorders [19]. The contemporary view of this process, as seen through the lens of modern-day drug development, constitutes the emerging field of LBPs and microbiome-based therapeutics. Both fecal microbiota transplantation (FMT) and LBPs are highlighted as potential novel strategies to address antimicrobial resistance in the recent US Centers for Disease Control and Prevent Antibiotic Resistance Threats in the United States Report, 2019 [6].

Gerardin et al [20] take a bullish stance on LBPs as they reflect on the current literature regarding the clinical benefits of FMT for recurrent CDI. In light of consistently positive data, the authors argue that biotechnology companies have enough evidence to justify a full-throated effort towards advancing a new drug development paradigm, one based on the naturally occurring bacteria that inhabit the human body, for example, FMT. However, for these benefits to be realized, drug developers need to overcome the inherent challenges of producing a complex drug product that meets regulatory standards for safety, identity, purity, and potency. For biologically sourced microbiome-based therapeutics, the authors point to persisting issues with safety and infection risk, particularly in the context of establishing rigorous, yet adaptable processes to screen for known pathogens. Beyond this immediate issue lies the challenge of establishing the long-term safety of LBPs. Manufacturability, establishing product uniformity through rigorous characterization, new formulations, and optimal delivery modalities are all still nascent topics that need to be addressed in this emerging field. Gerardin et al also discuss how LBPs can be considered fit-for-purpose, where full-spectrum microbiota (eg, FMT) may still be necessary to correct complex microbiota impairments and at the same time be used as a powerful tool for establishing proof of concept in new therapeutic areas. In contrast, limited spectrum, rationally designed microbial consortia of selected strains, can be used with precision and under circumstances in which the microbiota deficiency is defined.

These 2 categories (ie, full-spectrum microbiota and microbial consortia) of LBPs are further discussed in the piece by Ducarmon et al [21] who suggest that the inherent biological complexity of LBPs represents the single greatest challenge to the field. Because most LBPs are being studied and designed based on concepts derived from microbial ecology, the influence of LBPs on host physiology can be pleiotropic, and the rules that govern this very complex host-microbe system are still largely unknown. Biotechnology companies, the authors argue, lack the deep fundamental knowledge and model systems needed to accurately predict the potential preclinical activity of their drug candidates. The lack of precedent in this space has necessitated the use of healthy human cohort studies and dose-ranging studies in patients to provide representative data on LBP pharmacokinetic and pharmacodynamic relationships. Additional challenges discussed by Ducarmon et al [21] include the establishment of intellectual property and the lack of sufficient economic or “pull incentives” in the anti-infective marketplace.

Innovations in live microbiome-based drug development have not been exclusive to the gastrointestinal compartment. As mentioned above, a new investigational product (LACTIN-V) constitutes the first vaginal microbiome-based LBP and was designed to restore a Lactobacillus-dominated vaginal microbiota in women with BV. Unlike most LBPs designed for the gut microbiota, LACTIN-V consists of a single, rationally selected, strain of Lactobacillus crispatus and is being developed as adjunctive therapy after antimicrobial treatment for BV. The unique formulations and extensive clinical development of LACTIN-V is reviewed in the article by Lagenaur et al [11] including a recent proof-of-concept Phase 2 clinical trial demonstrating significant efficacy in preventing BV recurrence. However, most importantly, there is the mechanistic evidence that LACTIN-V colonization is closely correlated with the prevention of recurrent BV. These data have inspired the prospect of conducting a larger, pivotal, Phase 3 trial to support an application for US Food and Drug Administration approval. If successful, the positive impact of having an efficacious drug to prevent recurrent BV will likely extend to improvements in women’s health, including a reduced risk of STIs.

The 3 articles summarized above provide a comprehensive overview of the field of LBPs and saliently capture the numerous opportunities and challenges that lie ahead. This new framework for drug discovery and development has the potential to be a disruptive force for anti-infective drug development, especially in the context of increasing antimicrobial resistance.

PROTECTING THE MICROBIOTA AND EXTREMELY NARROW-SPECTRUM ANTIMICROBIALS

The growing associations and mechanistic insights between an impaired microbiota and susceptibility to infectious disease has motivated research into new approaches that protect the microbiota while specifically eliminating pathogens. Rooney et al [22] present a compelling case for this specific line of drug development and discuss evidence from recent clinical trials on the effectiveness of some of these products. One example of an extremely narrow spectrum antibiotic is the advanced clinical candidate ridinilazole for the treatment of CDI. Ridinilazole appears to specifically target Clostridial spp and has the unique characteristics of extreme low bioavailability after oral administration and a limited impact on the members of the gut microbiota. This drug, the authors note, appears to avoid the C difficile “treatment paradox,” where the treatment can act as both a cure and a potential cause of recurrent infection. Another promising approach to limiting the impact of broad-spectrum antibiotics is to irreversibly destroy or capture residual antibiotic before it accumulates in the gut. The clinical data for these products show they can be effective; however, demonstrating a clear reduction in infection risk presents a challenging clinical development pathway.

Leveraging the complex principles and mechanisms of microbial ecology to discover and develop highly targeted antimicrobials is also another broad and active area of research. Components of the gut microbiota have evolved to establish their nutritional niche and limit interspecies competition. Several microbial strategies have been developed to overcome so-called “nutritional immunity” where the host limits trace elements to protect against colonization by pathogens [23]. One such strategy is reviewed in the article by Sargun et al [24] in which the authors discuss sophisticated strategies used by pathogenic bacteria to acquire a limited supply of free iron (via siderophores) and other essential nutrients. The authors discuss recent advances in understanding siderophore-mediated iron-acquisition systems and how they can be leveraged to develop “trojan-horse” strategies, able to specifically deliver a toxic payload that selectively targets pathogenic members of the gut microbiota. The existence of natural trojan-horse antimicrobials, in the form of siderophore-antibiotic conjugates (SACs), can be used to understand their mechanisms of antibacterial activity and also provide the chemical framework to rationally designed synthetic SACs. As noted by Sargun et al, the first synthetic siderophore-cephalosporin conjugate (ie, cefiderocol) was approved in 2019 for the treatment of complicated urinary tract infections. This appears to have revitalized interest in using siderophore modifications as a means to specifically target the antibacterial activity of SACs against drug-resistant pathogens. Another interesting translational application discussed in this article involves using iron acquisition system components as antigens for immunization against Gram-negative pathogens (eg, E coli). Early studies using this strategy have shown some promise, yet more work is needed to fully appreciate the potential of this approach.

CONCLUSIONS AND FUTURE PERSPECTIVES

This special issue explores the state of microbiome science at the intersection of infectious diseases and antimicrobial resistance, an undertaking that would not have been possible a decade ago. Through each article we learn the different ways in which the human microbiome can play a central role in the protection from infectious diseases and how host-microbiota interactions forms a complex system of interactions that are critical to maintain tissue homeostasis and immunity. Systems-level approaches complemented with mechanistic studies (1) can be used to reveal the forces that govern the structure and function of each community, (2) can lead to the discovery of new therapeutics, and (3) can be used to blunt the impact of increasing antimicrobial resistance. Scientific advances in the field have already produced new possibilities and treatment modalities (ie, FMT and LBPs) for CDI and BV, in which wholesale replacement of the microbiota can be used as a highly efficacious treatment. However, this new knowledge may not always translate directly into novel therapeutics, instead it may result in enhanced diagnostics or adjunctive therapies that seek to limit the impact to host-associated microbial communities with the goal of reducing risk of infection.

Notes

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention (CDC) and the National Institute of Allergy and Infectious Diseases (NIAID). This article was written by the National Institutes of Health (NIH) and CDC employees in the course of their usual duties without additional funding support.

Financial support. This article was written by the National Institutes of Health and Centers for Disease Control and Prevention employees in the course of their usual duties without additional funding support.

Supplement sponsorship. This work is part of a supplement sponsored by the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health (NIH) and the Centers for Disease Control and Prevention (CDC).

Potential conflicts of interest. V. B. Y. reports nonfinancial support from Vedanta Bioscience, nonfinancial support from Bio-K + International, and nonfinancial support from Pantheryx, outside the submitted work. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.