Microbes are the oldest and most diverse organisms on Earth. They thrive in virtually every environment imaginable, including within the bodies of other organisms. The community of microbes in a given habitat is called a microbiome.

Microbiomes are comprised of diverse taxa engaged in a variety of metabolic and biochemical activities beneficial to ecosystem health [Ferreiro, Crook, Gasparrini et al. Cell 2018]. In the Dantas Lab, we study the ecology and functional dynamics of microbiomes from diverse host- and environment-associated habitats, and in particular, their connections to human health and disease.

To understand the dynamics of healthy microbiomes, we investigate how microbiomes respond to disruptions. Biotic or abiotic perturbations (e.g., antibiotics and bacterial pathogens) can acutely and persistently disrupt the taxonomic composition and metabolic and biochemical functions encoded by the microbiome [Fishbein, Mahmud et al. Nat Rev Microbiol 2023]. Depending on the severity of the perturbation and the resilience of the microbiome, this may lead to ‘dysbiotic’ or unhealthy states associated with numerous human pathologies.

Antibiotics use is among the most impactful microbiome perturbations. Though these “wonder drugs” cure once-fatal infectious diseases, they have two major drawbacks: 1) they kill beneficial commensal bacteria along with disease-causing pathogens, and 2) they enrich for antibiotic resistance in bacteria that undermines their future utility [Crofts, Gasparrini et al. Nat Rev Microbiol 2017]. In 2019, an estimated 4.95 million deaths were associated with antibiotic resistance globally, making it an urgent global health threat. Knowledge of the impact of antibiotics and antibiotic-resistant pathogens on the gut microbiome is key to maintaining the health of individuals’ microbiomes and minimizing the spread of resistance.

In the Dantas Lab, we aim to systematically 1) UNDERSTAND the functional effects of microbiome disruptions, then use that information to 2) PREDICT features indicative of microbiome-specific responses to disruption and validate these using in vivo animal models, and 3) REMEDIATE the negative effects of disruptions with engineered probiotic platforms. These three goals are accomplished leveraging extensive clinical collaborations alongside cutting-edge molecular, multi-omic, and engineering approaches [Boolchandani, D’Souza et al. Nat Rev Genetics 2019].

RESEARCH PROJECTS

UNDERSTAND

By characterizing the impacts of perturbations at the level of microbial communities, species/strains, and gene products, we can identify specific microbes and their functions as biomarkers for predicting health outcomes, and potentially as actionable targets for remediation. This information generates hypotheses that can quantitatively explain microbiome responses to disruption and enables development of strategies to minimize or remediate their negative impacts.

A few of our recent and ongoing efforts to UNDERSTAND how microbiomes respond to disruptions are described below:

Community-level multi-omic analyses

• Resistomes of soil bacteria: Soil bacteria are a significant reservoir of antibiotic resistance genes, fostering the conditions for high-level resistance to evolve then transfer into human pathogens. We provided the first genetic evidence for exchange of antibiotic resistance genes between benign soil multidrug-resistant (MDR) Proteobacteria [Dantas, Sommer et al. Science 2008] [Press] and MDR human pathogens [Forsberg, Reyes et al. Science 2012] [Press]. In contrast, we demonstrated that a majority of the uncultured soil resistome is not poised for horizontal gene exchange with human pathogens and is instead structured by phylogeny and habitat [Forsberg, Patel et al. Nature 2014][Press].

• Pediatric microbiome development: Early life is a critical period for microbiome development, and perturbations during this process can permanently “scar” the microbiome [Thänert, Sawhney, Schwartz et al. Cell Host & Microbe 2022]. We demonstrated that differences during the first years of life such as breastfeeding, formula composition, and maternal weight correlate with distinct microbiome profiles [Baumann-Dudenhoeffer et al. Nature Medicine 2018]. Conversely, we demonstrated that antibiotic use in premature infants acutely perturbs the infant gut microbiome [Gibson et al. Nature Microbiology 2016] [Press], and leads to persistent enrichment of antibiotic resistance genes and MDR organisms >1 year after discharge [Gasparrini et al. Nature Microbiology 2019]. We developed machine learning models that can distinguish between the microbiomes of preterm post-NICU infants and age-matched antibiotic-naïve near-term infants, based on antibiotic resistance genes and bacterial taxa.

Species-level genomic & functional profiling

• Pathogenomics: Infections with multidrug-resistant pathogens are an urgent global health problem. The antibiotic-resistant species of the greatest threat to human health are termed ESKAPEE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter species, and Escherichia coli). Comparative whole-genome sequencing of clinical isolates can generate a wealth of data on the genetic and functional diversity of these species [Sukhum, Diorio-Toth et al. Clin Pharm Therap 2019]. Leveraging extensive collaborations with clinical researchers, we have investigated the genomics of a variety of antibiotic-resistant bacterial pathogens—collectively referred to as “pathogenomics.” These include studies of Gardnerella genomospecies during bacterial vaginosis [Potter et al. Clinical Chemistry 2019], in-host adaptation of Staphylococcus pseudintermedius [Sawhney, Vargas et al. Nature Commun 2023] and Mycobacterium abscessus [Choi et al. J Infect Dis 2023], and plasmid epidemiology in clinical Escherichia coli [Mahmud et al. mSystems 2d022].

• Hospital surfaces: Bacterial pathogens that contaminate the hospital built environment are frequently linked to outbreaks in human patients [Blake, Choi et al. Cell Mol Life Sci 2021]. We demonstrated that common nosocomial pathogens such as Pseudomonas aeruginosa can move between intensive care room surfaces and patients [Sukhum, Newcomer et al. Commun Med 2022]. Further, we identified synergistic interactions promoting the persistence of opportunistic MDR pathogens, Acinetobacter baumannii and Enterococcus faecium, on surfaces [D’Souza, Potter et al. Nature Commun 2019], and that reservoirs of MDR pathogens can persist in sink drains for years [Diorio-Toth et al. mSystems2023].

Sequence-structure-function characterization of gene products

• Tetracycline destructases: Third-generation tetracycline antibiotics are reserved as drugs of “last resort,” for use only with multidrug-resistant pathogens. Using functional metagenomic selection, we discovered a novel family of flavoenzymes which confer high-level resistance to all tetracycline antibiotics, which we termed the ‘tetracycline destructases’ [Forsberg et al. Chemistry & Biology 2015][Gasparrini et al. Commun Biol 2020]. To counter resistance by these enzymes, we developed small-molecule inhibitors that prevent the degradation of tetracyclines by these tetracycline destructases [Markley et al. ACS Infect Dis 2019], established the biochemical and structural basis for inhibition [Park, Gasparrini et al. Nature Chemical Biology 2017] [Press], and identified the sequence determinants of inactivating function [Blake, Kumar, Loganathan, Williford et al. Commun Biol 2024].

Additional projects

• Antibiotic resistance transmission on farms [Sukhum, Vargas et al. mBio 2021] [Sun et al. Nature Communications 2020] [Mahmud, Vargas et al. Nature Microbiology 2024]

• Role of pant-associated microbes using the model organism Arabidopsis thaliana

• Human skin microbiome dynamics during health and disease

• Genomics of gut microbiome-associated fungi and bacteriophages [Zhang, Potter et al. mSystems 2023]

• Role of gut microbiota in urinary tract infections [Adu-Oppong, Thänert et al. Microbiome 2022] [Thänertet al. mBio 2019]

• Mobile genetic elements and their dynamics across human and built-environment microbiomes [Diorio-Toth et al. mSystems 2023] [Mahmud et al. mSystems 2022] [Thänert, Choi. Cell Host & Microbe 2022]

PREDICT

To bridge the gap between UNDERSTANDING microbiome disruptions and REMEDIATING disrupted microbiome states, we develop predictive computational and pre-clinical animal models of microbiome dynamics. Using microbiota-humanized gnotobiotic mouse models to recapitulate human microbiome-host interactions, we can validate computational predictions of pathogen- or antibiotic-induced perturbations.

A few of our recent and ongoing efforts to PREDICT microbiome dynamics are described below:

Models of microbiome dynamics

• Gut-brain axis and Alzheimer’s disease: Alzheimer’s disease is the most common type of dementia, progressing from minor memory impairment to major disruptions to daily life and death. Several lines of evidence suggest that dysbiosis in the gut microbiome is associated with Alzheimer’s disease, raising the potential for its use as a non-invasive early screening tool. We demonstrated that the gut microbiomes of people with preclinical Alzheimer’s disease are distinct from healthy people, and are correlated with β-amyloid (Aβ) and tau pathological biomarkers [Ferreiro et al. Sci Trans Med 2023]. Including these microbiome features improves machine learning models’ ability to predict preclinical Alzheimer’s disease status. Additionally, we’ve shown using mouse models that bacterial sepsis exacerbates amyloid plaque deposition, promoting the progression of Alzheimer’s disease [Basak, Ferreiro et al. Neurobiol Dis 2021]. We are currently investigating whether gut microbiome differences are a result or cause of early Alzheimer’s disease, applying multi-omics approaches to longitudinal samples from healthy, preclinical, and symptomatic individuals.

• Pathogen-microbiome dynamics in high infectious disease-burden areas: Travel to high infectious disease-burden regions, often associated with elevated risk of travelers’ diarrhea, exacerbates the risk of individuals’ acquiring MDR bacteria and contributes to the global spread of antibiotic resistance. We demonstrated that low-income, high infectious disease-burdened regions have higher rates of resistance than in developed countries, work which was featured on the cover of Nature [Pehrsson, Tsukayama et al. Nature 2016] [Press]. In subsequent work, we found that travelers’ diarrhea disrupts microbiome stability and increases the abundance of antibiotic resistance genes for weeks [Boolchandani et al. Nature Commun 2022]. Using differentially-enriched taxa, we developed a machine learning model that can discriminate between diarrheal vs. healthy samples based on microbiome composition. Encouragingly, we demonstrated that single-dose antibiotic treatment for diarrhea does not significantly worsen gut microbiome dysbiosis [Blake, Schwartz, Paruthiyil et al. mBio2023].

• Clostridioides difficile infection: C. difficile infection is a major cause of healthcare-associated diarrhea. We have demonstrated that oral vancomycin fails to permanently clear C. difficile and perturbs the gut microbiome [Fishbein et al. mSphere 2021]. Additionally, we have shown that FMT success is correlated with donor microbiome configurations, and that FMTs can introduce new resistant organisms into recipients [Kwak et al. Microbiome 2020] [Langdon, Schwartz et al. Genome Med 2021]. We have identified key gut microbial taxa, functions, and metabolites which distinguish patients asymptomatically colonized with toxigenic C. difficile from those with symptomatic infection [Fishbein et al. eLife 2022], and tracked instances of C. difficile transmission and persistence between hospital surfaces and patients [Newcomer, Fishbein et al. mBio 2024]. We are currently utilizing innovative microbiota-humanized gnotobiotic mouse models to capture the influence of the human gut microbiome on C. difficile colonization and disease.

Additional projects

• Synergistic antibiotic combinations to kill methicillin-resistant Staphylococcus aureus (MRSA) [Gonzaleset al. Nature Chemical Biology 2016] [Press]

• “Antibiotic scarring” in healthy volunteers [Anthony et al. Cell Reports 2022]

REMEDIATE

Probiotic microorganisms have the exciting potential to ameliorate perturbed microbiome states and treat gastrointestinal diseases at the molecular level. We can design and engineer probiotics that manipulate specific species or functions identified as contributing to microbiome disruptions or disease and return these to healthy states.

A few of our recent and ongoing efforts to REMEDIATE microbiome dynamics and gut diseases are described below:

• Detection of microbiome disruptions: Engineered probiotic microbes can be used to identify microbiome disruptions, such as pathogens, and then deliver biologics to remediate the disruptions. We have developed a large array of tools for the engineering of the commensal yeast S. boulardii as a safe probiotic [Kwak, Mahmud et al. ACS Synth Biol 2021]. Ongoing work includes probiotic display or secretion of epitope-binding proteins for the detection or clearance of bacterial pathogens, and delivery of biologics in response to gut-relevant chemical signals including pH, or concentration of oxygen, bile salts, or short-chain fatty acids.

• Provide missing metabolic functions: Many human metabolic disorders stem from loss-of-function mutations in key metabolic enzymes. Probiotic microbes that encode enzymes which mimic or complement these missing functions are promising therapeutics for these disorders. For example, the rare inherited disorder phenylketonuria is caused by a mutation in the human enzyme needed to break down phenylalanine. We have developed the probiotic E. coli Nissle as a robust probiotic chassis to provide the missing phenylalanine degradation function [Crook, Ferreiro et al. Cell Host and Microbe2019]. In complementary work to improve the predictability and robustness of the E. coli Nissle chassis, we have developed a multiplexed transcript barcoding system to tune gene expression and probiotic performance [Crook et al. ACS Synth Biol 2020], and CRISPR-based kill switches for biocontainment [Rottinghaus, Ferreiro et al. Nature Commun 2022].

• Biofuels: Microbes have the capacity to produce high-value industrial compounds, such as bioplastics and biofuels. We developed high-throughput functional screening systems to identify novel tolerance and synthase genes from soil microbiomes [Forsberg et al. Appl Environ Microbiol 2016] and human gut microbiomes [Kwak, Crook et al. Biotechnol Biofuels Bioprod 2022]. Further, we have developed Rhodococcus opacus as a chassis for production of high-value compounds from diverse aromatic substrates [Henson, Campbell et al. Metabolic Engineering 2018] [Anthony, Carr et al. Biotechnology for Biofuels 2019].

Additional projects

• Treatment of colorectal cancer by in situ delivery of immune checkpoint inhibitors and other immunomodulatory proteins via engineered probiotics

• Treating severe acute malnutrition with antibiotics [Schwartz, Langdon et al. Lancet Microbe 2023]

• Microbiota restoration via fecal microbiota transplantation [Kwak, Choi et al. Microbiome 2020][Langdon, Schwartz et al. Genome Med 2021]