Rarefaction is currently the best approach to control for uneven sequencing effort in amplicon sequence analyses

Considering it is common to find as much as 100-fold variation in the number of 16S rRNA gene sequences across samples in a study, researchers need to control for the effect of uneven sequencing effort. How to do this has become a contentious question. Some have argued that rarefying or rarefaction is "inadmissible" because it omits valid data. A number of alternative approaches have been developed to normalize and rescale the data that purport to be invariant to the number of observations. I generated community distributions based on 12 published data sets where I was able to assess the ability of multiple methods to control for uneven sequencing effort. Rarefaction was the only method that could control for variation in uneven sequencing effort when measuring commonly used alpha and beta diversity metrics. Next, I compared the false detection rate and power to detect true differences between simulated communities with a known effect size using various alpha and beta diversity metrics. Although all methods of controlling for uneven sequencing effort had an acceptable false detection rate when samples were randomly assigned to two treatment groups, rarefaction was consistently able to control for differences in sequencing effort when sequencing depth was confounded with treatment group. Finally, the statistical power to detect differences in alpha and beta diversity metrics was consistently the highest when using rarefaction. These simulations underscore the importance of using rarefaction to normalize the number of sequences across samples in amplicon sequencing analyses.

IMPORTANCE

Sequencing 16S rRNA gene fragments has become a fundamental tool for understanding the diversity of microbial communities and the factors that affect their diversity. Due to technical challenges, it is common to observe wide variation in the number of sequences that are collected from different samples within the same study. However, the diversity metrics used by microbial ecologists are sensitive to differences in sequencing effort. Therefore, tools are needed to control for the uneven levels of sequencing. This simulation-based analysis shows that despite a longstanding controversy, rarefaction is the most robust approach to control for uneven sequencing effort. The controversy started because of confusion over the definition of rarefaction and violation of assumptions that are made by methods that have been borrowed from other fields. Microbial ecologists should use rarefaction.

Waste not, want not: revisiting the analysis that called into question the practice of rarefaction

In 2014, McMurdie and Holmes published the provocatively titled "Waste not, want not: why rarefying microbiome data is inadmissible." The claims of their study have significantly altered how microbiome researchers control for the unavoidable uneven sequencing depths that are inherent in modern 16S rRNA gene sequencing. Confusion over the distinction between the definitions of rarefying and rarefaction continues to cloud the interpretation of their results. More importantly, the authors made a variety of problematic choices when designing and analyzing their simulations. I identified 11 factors that could have compromised the results of the original study. I reproduced the original simulation results and assessed the impact of those factors on the underlying conclusion that rarefying data is inadmissible. Throughout, the design of the original study made choices that caused rarefying and rarefaction to appear to perform worse than they truly did. Most important were the approaches used to assess ecological distances, the removal of samples with low sequencing depth, and not accounting for conditions where sequencing effort is confounded with treatment group. Although the original study criticized rarefying for the arbitrary removal of valid data, repeatedly rarefying data many times (i.e., rarefaction) incorporates all the data. In contrast, it is the removal of rare taxa that would appear to remove valid data. Overall, I show that rarefaction is the most robust approach to control for uneven sequencing effort when considered across a variety of alpha and beta diversity metrics.IMPORTANCEOver the past 10 years, the best method for normalizing the sequencing depth of samples characterized by 16S rRNA gene sequencing has been contentious. An often cited article by McMurdie and Holmes forcefully argued that rarefying the number of sequence counts was "inadmissible" and should not be employed. However, I identified a number of problems with the design of their simulations and analysis that compromised their results. In fact, when I reproduced and expanded upon their analysis, it was clear that rarefaction was actually the most robust approach for controlling for uneven sequencing effort across samples. Rarefaction limits the rate of falsely detecting and rejecting differences between treatment groups. Far from being "inadmissible", rarefaction is a valuable tool for analyzing microbiome sequence data.

Machine learning classification by fitting amplicon sequences to existing OTUs

The ability to use 16S rRNA gene sequence data to train machine learning classification models offers the opportunity to diagnose patients based on the composition of their microbiome. In some applications, the taxonomic resolution that provides the best models may require the use of de novo operational taxonomic units (OTUs) whose composition changes when new data are added. We previously developed a new reference-based approach, OptiFit, that fits new sequence data to existing de novo OTUs without changing the composition of the original OTUs. While OptiFit produces OTUs that are as high quality as de novo OTUs, it is unclear whether this method for fitting new sequence data into existing OTUs will impact the performance of classification models relative to models trained and tested only using de novo OTUs. We used OptiFit to cluster sequences into existing OTUs and evaluated model performance in classifying a dataset containing samples from patients with and without colonic screen relevant neoplasia (SRN). We compared the performance of this model to standard methods including de novo and database-reference-based clustering. We found that using OptiFit performed as well or better in classifying SRNs. OptiFit can streamline the process of classifying new samples by avoiding the need to retrain models using reclustered sequences. IMPORTANCE There is great potential for using microbiome data to aid in diagnosis. A challenge with de novo operational taxonomic unit (OTU)-based classification models is that 16S rRNA gene sequences are often assigned to OTUs based on similarity to other sequences in the dataset. If data are generated from new patients, the old and new sequences must be reclustered to OTUs and the classification model retrained. Yet there is a desire to have a single, validated model that can be widely deployed. To overcome this obstacle, we applied the OptiFit clustering algorithm to fit new sequence data to existing OTUs allowing for reuse of the model. A random forest model implemented using OptiFit performed as well as the traditional reassign and retrain approach. This result shows that it is possible to train and apply machine learning models based on OTU relative abundance data that do not require retraining or the use of a reference database.

Gut Microbiome Composition in Lynch Syndrome With and Without History of Colorectal Neoplasia and Non-Lynch Controls

BACKGROUND

While Lynch syndrome (LS) is a highly penetrant colorectal cancer (CRC) syndrome, there is considerable variation in penetrance; few studies have investigated the association between microbiome and CRC risk in LS. We analyzed the microbiome composition among individuals with LS with and without personal history of colorectal neoplasia (CRN) and non-LS controls.

METHODS

We sequenced the V4 region of the 16S rRNA gene from the stool of 46 individuals with LS and 53 individuals without LS. We characterized within community and in between community microbiome variation, compared taxon abundance, and built machine learning models to investigate the differences in microbiome.

RESULTS

There was no difference within or between community variations among LS groups, but there was a statistically significant difference in both within and between community variation comparing LS to non-LS. Streptococcus and Actinomyces were differentially enriched in LS-CRC compared to LS-without CRN. There were numerous differences in taxa abundance comparing LS to non-LS; notably, Veillonella was enriched and Faecalibacterium and Romboutsia were depleted in LS. Finally, machine learning models classifying LS from non-LS controls and LS-CRC from LS-without CRN performed moderately well.

CONCLUSIONS

Differences in microbiome composition between LS and non-LS may suggest a microbiome pattern unique to LS formed by underlying differences in epithelial biology and immunology. We found specific taxa differences among LS groups, which may be due to underlying anatomy. Larger prospective studies following for CRN diagnosis and microbiome composition changes are needed to determine if microbiome composition contributes to CRN development in patients with LS.

Evaluation of software impact designed for biomedical research: Are we measuring what's meaningful?

Software is vital for the advancement of biology and medicine. Through analysis of usage and impact metrics of software, developers can help determine user and community engagement. These metrics can be used to justify additional funding, encourage additional use, and identify unanticipated use cases. Such analyses can help define improvement areas and assist with managing project resources. However, there are challenges associated with assessing usage and impact, many of which vary widely depending on the type of software being evaluated. These challenges involve issues of distorted, exaggerated, understated, or misleading metrics, as well as ethical and security concerns. More attention to the nuances, challenges, and considerations involved in capturing impact across the diverse spectrum of biological software is needed. Furthermore, some tools may be especially beneficial to a small audience, yet may not have comparatively compelling metrics of high usage. Although some principles are generally applicable, there is not a single perfect metric or approach to effectively evaluate a software tool's impact, as this depends on aspects unique to each tool, how it is used, and how one wishes to evaluate engagement. We propose more broadly applicable guidelines (such as infrastructure that supports the usage of software and the collection of metrics about usage), as well as strategies for various types of software and resources. We also highlight outstanding issues in the field regarding how communities measure or evaluate software impact. To gain a deeper understanding of the issues hindering software evaluations, as well as to determine what appears to be helpful, we performed a survey of participants involved with scientific software projects for the Informatics Technology for Cancer Research (ITCR) program funded by the National Cancer Institute (NCI). We also investigated software among this scientific community and others to assess how often infrastructure that supports such evaluations is implemented and how this impacts rates of papers describing usage of the software. We find that although developers recognize the utility of analyzing data related to the impact or usage of their software, they struggle to find the time or funding to support such analyses. We also find that infrastructure such as social media presence, more in-depth documentation, the presence of software health metrics, and clear information on how to contact developers seem to be associated with increased usage rates. Our findings can help scientific software developers make the most out of the evaluations of their software so that they can more fully benefit from such assessments.

Association of bacterial community types, functional microbial processes and lung disease in cystic fibrosis airways

Bacterial infection and inflammation of the airways are the leading causes of morbidity and mortality in persons with cystic fibrosis (CF). The ecology of the bacterial communities inhabiting CF airways is poorly understood, especially with respect to how community structure, dynamics, and microbial metabolic activity relate to clinical outcomes. In this study, the bacterial communities in 818 sputum samples from 109 persons with CF were analyzed by sequencing bacterial 16S rRNA gene amplicons. We identified eight alternative community types (pulmotypes) by using a Dirichlet multinomial mixture model and studied their temporal dynamics in the cohort. Across patients, the pulmotypes displayed chronological patterns in the transition among each other. Furthermore, significant correlations between pulmotypes and patient clinical status were detected by using multinomial mixed effects models, principal components regression, and statistical testing. Constructing pulmotype-specific metabolic activity profiles, we found that pulmotype microbiota drive distinct community functions including mucus degradation or increased acid production. These results indicate that pulmotypes are the result of ordered, underlying drivers such as predominant metabolism, ecological competition, and niche construction and can form the basis for quantitative, predictive models supporting clinical treatment decisions.

Developing and deploying an integrated workshop curriculum teaching computational skills for reproducible research

Inspired by well-established material and pedagogy provided by The Carpentries (Wilson, 2016), we developed a two-day workshop curriculum that teaches introductory R programming for managing, analyzing, plotting and reporting data using packages from the tidyverse (Wickham et al., 2019), the Unix shell, version control with git, and GitHub. While the official Software Carpentry curriculum is comprehensive, we found that it contains too much content for a two-day workshop. We also felt that the independent nature of the lessons left learners confused about how to integrate the newly acquired programming skills in their own work. Thus, we developed a new curriculum that aims to teach novices how to implement reproducible research principles in their own data analysis. The curriculum integrates live coding lessons with individual-level and group-based practice exercises, and also serves as a succinct resource that learners can reference both during and after the workshop. Moreover, it lowers the entry barrier for new instructors as they do not have to develop their own teaching materials or sift through extensive content. We developed this curriculum during a two-day sprint, successfully used it to host a two-day virtual workshop with almost 40 participants, and updated the material based on instructor and learner feedback. We hope that our new curriculum will prove useful to future instructors interested in teaching workshops with similar learning objectives.

mikropml: User-Friendly R Package for Supervised Machine Learning Pipelines

Machine learning (ML) for classification and prediction based on a set of features is used to make decisions in healthcare, economics, criminal justice and more. However, implementing an ML pipeline including preprocessing, model selection, and evaluation can be time-consuming, confusing, and difficult. Here, we present mikropml (prononced "meek-ROPE em el"), an easy-to-use R package that implements ML pipelines using regression, support vector machines, decision trees, random forest, or gradient-boosted trees. The package is available on GitHub, CRAN, and conda.

Women Are Underrepresented and Receive Differential Outcomes at ASM Journals: a Six-Year Retrospective Analysis

Despite 50% of biology Ph.D. graduates being women, the number of women that advance in academia decreases at each level (e.g., from graduate to postdoctorate to tenure track). Recently, scientific societies and publishers have begun examining internal submissions data to evaluate representation and evaluation of women in their peer review processes; however, representation and attitudes differ by scientific field, and to date, no studies have investigated academic publishing in the field of microbiology. Using manuscripts submitted between January 2012 and August 2018 to the 15 journals published by the American Society for Microbiology (ASM), we describe the representation of women at ASM journals and the outcomes of their manuscripts. Senior women authors at ASM journals were underrepresented compared to global and society estimates of microbiology researchers. Additionally, manuscripts submitted by corresponding authors that were women received more negative outcomes than those submitted by men. These negative outcomes were somewhat mediated by whether or not the corresponding author was based in the United States and by the type of institution for United States-based authors. Nonetheless, the pattern for women corresponding authors to receive more negative outcomes on their submitted manuscripts held. We conclude with suggestions to improve the representation of women and decrease structural penalties against women.IMPORTANCE Barriers in science and academia have prevented women from becoming researchers and experts that are viewed as equivalent to their colleagues who are men. We evaluated the participation and success of women researchers at ASM journals to better understand their success in the field of microbiology. We found that women are underrepresented as expert scientists at ASM journals. This is, in part, due to a combination of both low submissions from senior women authors and more negative outcomes on submitted manuscripts for women compared to men.

A Framework for Effective Application of Machine Learning to Microbiome-Based Classification Problems

Machine learning (ML) modeling of the human microbiome has the potential to identify microbial biomarkers and aid in the diagnosis of many diseases such as inflammatory bowel disease, diabetes, and colorectal cancer. Progress has been made toward developing ML models that predict health outcomes using bacterial abundances, but inconsistent adoption of training and evaluation methods call the validity of these models into question. Furthermore, there appears to be a preference by many researchers to favor increased model complexity over interpretability. To overcome these challenges, we trained seven models that used fecal 16S rRNA sequence data to predict the presence of colonic screen relevant neoplasias (SRNs) (n = 490 patients, 261 controls and 229 cases). We developed a reusable open-source pipeline to train, validate, and interpret ML models. To show the effect of model selection, we assessed the predictive performance, interpretability, and training time of L2-regularized logistic regression, L1- and L2-regularized support vector machines (SVM) with linear and radial basis function kernels, a decision tree, random forest, and gradient boosted trees (XGBoost). The random forest model performed best at detecting SRNs with an area under the receiver operating characteristic curve (AUROC) of 0.695 (interquartile range [IQR], 0.651 to 0.739) but was slow to train (83.2 h) and not inherently interpretable. Despite its simplicity, L2-regularized logistic regression followed random forest in predictive performance with an AUROC of 0.680 (IQR, 0.625 to 0.735), trained faster (12 min), and was inherently interpretable. Our analysis highlights the importance of choosing an ML approach based on the goal of the study, as the choice will inform expectations of performance and interpretability.IMPORTANCE Diagnosing diseases using machine learning (ML) is rapidly being adopted in microbiome studies. However, the estimated performance associated with these models is likely overoptimistic. Moreover, there is a trend toward using black box models without a discussion of the difficulty of interpreting such models when trying to identify microbial biomarkers of disease. This work represents a step toward developing more-reproducible ML practices in applying ML to microbiome research. We implement a rigorous pipeline and emphasize the importance of selecting ML models that reflect the goal of the study. These concepts are not particular to the study of human health but can also be applied to environmental microbiology studies.

Gut Microbiota Modulate CD8 T Cell Responses to Influence Colitis-Associated Tumorigenesis

There is increasing evidence that gut microbiome perturbations, also known as dysbiosis, can influence colorectal cancer development. To understand the mechanisms by which the gut microbiome modulates cancer susceptibility, we examine two wild-type mouse colonies with distinct gut microbial communities that develop significantly different tumor numbers using a mouse model of inflammation-associated tumorigenesis. We demonstrate that adaptive immune cells contribute to the different tumor susceptibilities associated with the two microbial communities. Mice that develop more tumors have increased colon lamina propria CD8+ IFNγ+ T cells before tumorigenesis but reduced CD8+ IFNγ+ T cells in tumors and adjacent tissues compared with mice that develop fewer tumors. Notably, intratumoral T cells in mice that develop more tumors exhibit increased exhaustion. Thus, these studies suggest that microbial dysbiosis can contribute to colon tumor susceptibility by hyperstimulating CD8 T cells to promote chronic inflammation and early T cell exhaustion, which can reduce anti-tumor immunity.

Reintroducing mothur: 10 Years Later

More than 10 years ago, we published the paper describing the mothur software package in Applied and Environmental Microbiology Our goal was to create a comprehensive package that allowed users to analyze amplicon sequence data using the most robust methods available. mothur has helped lead the community through the ongoing sequencing revolution and continues to provide this service to the microbial ecology community. Beyond its success and impact on the field, mothur's development exposed a series of observations that are generally translatable across science. Perhaps the observation that stands out the most is that all science is done in the context of prevailing ideas and available technologies. Although it is easy to criticize choices that were made 10 years ago through a modern lens, if we were to wait for all of the possible limitations to be solved before proceeding, science would stall. Even preceding the development of mothur, it was necessary to address the most important problems and work backwards to other problems that limited access to robust sequence analysis tools. At the same time, we strive to expand mothur's capabilities in a data-driven manner to incorporate new ideas and accommodate changes in data and desires of the research community. It has been edifying to see the benefit that a simple set of tools can bring to so many other researchers.

Fecal Short-Chain Fatty Acids Are Not Predictive of Colonic Tumor Status and Cannot Be Predicted Based on Bacterial Community Structure

Colonic bacterial populations are thought to have a role in the development of colorectal cancer with some protecting against inflammation and others exacerbating inflammation. Short-chain fatty acids (SCFAs) have been shown to have anti-inflammatory properties and are produced in large quantities by colonic bacteria that produce SCFAs by fermenting fiber. We assessed whether there was an association between fecal SCFA concentrations and the presence of colonic adenomas or carcinomas in a cohort of individuals using 16S rRNA gene and metagenomic shotgun sequence data. We measured the fecal concentrations of acetate, propionate, and butyrate within the cohort and found that there were no significant associations between SCFA concentration and tumor status. When we incorporated these concentrations into random forest classification models trained to differentiate between people with healthy colons and those with adenomas or carcinomas, we found that they did not significantly improve the ability of 16S rRNA gene or metagenomic gene sequence-based models to classify individuals. Finally, we generated random forest regression models trained to predict the concentration of each SCFA based on 16S rRNA gene or metagenomic gene sequence data from the same samples. These models performed poorly and were able to explain at most 14% of the observed variation in the SCFA concentrations. These results support the broader epidemiological data that questions the value of fiber consumption for reducing the risks of colorectal cancer. Although other bacterial metabolites may serve as biomarkers to detect adenomas or carcinomas, fecal SCFA concentrations have limited predictive power.IMPORTANCE Considering that colorectal cancer is the third leading cancer-related cause of death within the United States, it is important to detect colorectal tumors early and to prevent the formation of tumors. Short-chain fatty acids (SCFAs) are often used as a surrogate for measuring gut health and for being anticarcinogenic because of their anti-inflammatory properties. We evaluated the fecal SCFA concentrations of a cohort of individuals with different colonic tumor burdens who were previously analyzed to identify microbiome-based biomarkers of tumors. We were unable to find an association between SCFA concentration and tumor burden or use SCFAs to improve our microbiome-based models of classifying people based on their tumor status. Furthermore, we were unable to find an association between the fecal community structure and SCFA concentrations. Our results indicate that the association between fecal SCFAs, the gut microbiome, and tumor burden is weak.

The Impact of DNA Polymerase and Number of Rounds of Amplification in PCR on 16S rRNA Gene Sequence Data

PCR amplification of 16S rRNA genes is a critical yet underappreciated step in the generation of sequence data to describe the taxonomic composition of microbial communities. Numerous factors in the design of PCR can impact the sequencing error rate, the abundance of chimeric sequences, and the degree to which the fragments in the product represent their abundance in the original sample (i.e., bias). We compared the performance of high fidelity polymerases and various numbers of rounds of amplification when amplifying a mock community and human stool samples. Although it was impossible to derive specific recommendations, we did observe general trends. Namely, using a polymerase with the highest possible fidelity and minimizing the number of rounds of PCR reduced the sequencing error rate, fraction of chimeric sequences, and bias. Evidence of bias at the sequence level was subtle and could not be ascribed to the fragments' fraction of bases that were guanines or cytosines. When analyzing mock community data, the amount that the community deviated from the expected composition increased with the number of rounds of PCR. This bias was inconsistent for human stool samples. Overall, the results underscore the difficulty of comparing sequence data that are generated by different PCR protocols. However, the results indicate that the variation in human stool samples is generally larger than that introduced by the choice of polymerase or number of rounds of PCR.IMPORTANCE A steep decline in sequencing costs drove an explosion in studies characterizing microbial communities from diverse environments. Although a significant amount of effort has gone into understanding the error profiles of DNA sequencers, little has been done to understand the downstream effects of the PCR amplification protocol. We quantified the effects of the choice of polymerase and number of PCR cycles on the quality of downstream data. We found that these choices can have a profound impact on the way that a microbial community is represented in the sequence data. The effects are relatively small compared to the variation in human stool samples; however, care should be taken to use polymerases with the highest possible fidelity and to minimize the number of rounds of PCR. These results also underscore that it is not possible to directly compare sequence data generated under different PCR conditions.

The gut microbiome can contribute to colon tumor susceptibility via an effect on CD8+ T cell responses

There is increasing evidence that colorectal cancer patients have altered gut microbiomes compared to healthy controls, but the mechanisms by which the microbiota contribute to colon carcinogenesis remain to be fully elucidated. Using the azoxymethane (AOM)/dextran sulfate sodium (DSS) mouse model of inflammation-associated colon tumorigenesis, our lab discovered two colonies of specific pathogen free (SPF) C57BL/6J wild type (WT) mice housed in the same mouse room that develop differential tumor burdens. Mice from the “WT1” colony developed five tumors on average while mice from the “WT2” colony developed 15 tumors on average. The increased tumor susceptibility in WT2 mice can be directly attributed to the gut microbiome as germ-free (GF) mice colonized with WT2 bacteria developed more tumors compared to that of GF mice colonized with WT1 bacteria. Additionally, 16S rRNA gene sequencing of fecal bacteria from WT1 and WT2 mice revealed distinct microbiomes with certain bacteria consistently associated with high or low tumor numbers. Furthermore, naïve and acutely-inflamed (day 10 of AOM/DSS) WT2 mice have increased colon lamina propria CD8+ IFNγ+ T cells compared to WT1 mice as measured by flow cytometry. However, in tumor-bearing WT2 mice, there was decreased tumor-infiltrating CD8+ T cells with reduced IFNγ production, possibly due to T cell exhaustion. GF Rag1−/− mice as well as SPF CD8−/− mice inoculated with WT2 gut microbiota developed fewer tumors than SPF WT2 mice, suggesting that the WT2 gut microbiome increases tumor susceptibility, in part, through an effect on CD8+ T cells. Altogether, our data reveal a potential novel role of microbiota in altering colon CD8+ T cell function that ultimately impacts colon cancer risk.

Toward unrestricted use of public genomic data

Publication interests should not limit access to public data

Nonsteroidal Anti-inflammatory Drugs Alter the Microbiota and Exacerbate Clostridium difficile Colitis while Dysregulating the Inflammatory Response

Clostridium difficile infection (CDI) is a major public health threat worldwide. The use of nonsteroidal anti-inflammatory drugs (NSAIDs) is associated with enhanced susceptibility to and severity of CDI; however, the mechanisms driving this phenomenon have not been elucidated. NSAIDs alter prostaglandin (PG) metabolism by inhibiting cyclooxygenase (COX) enzymes. Here, we found that treatment with the NSAID indomethacin prior to infection altered the microbiota and dramatically increased mortality and the intestinal pathology associated with CDI in mice. We demonstrated that in C. difficile-infected animals, indomethacin treatment led to PG deregulation, an altered proinflammatory transcriptional and protein profile, and perturbed epithelial cell junctions. These effects were paralleled by increased recruitment of intestinal neutrophils and CD4⁺ cells and also by a perturbation of the gut microbiota. Together, these data implicate NSAIDs in the disruption of protective COX-mediated PG production during CDI, resulting in altered epithelial integrity and associated immune responses.IMPORTANCEClostridium difficile infection (CDI) is a spore-forming anaerobic bacterium and leading cause of antibiotic-associated colitis. Epidemiological data suggest that use of nonsteroidal anti-inflammatory drugs (NSAIDs) increases the risk for CDI in humans, a potentially important observation given the widespread use of NSAIDs. Prior studies in rodent models of CDI found that NSAID exposure following infection increases the severity of CDI, but mechanisms to explain this are lacking. Here we present new data from a mouse model of antibiotic-associated CDI suggesting that brief NSAID exposure prior to CDI increases the severity of the infectious colitis. These data shed new light on potential mechanisms linking NSAID use to worsened CDI, including drug-induced disturbances to the gut microbiome and colonic epithelial integrity. Studies were limited to a single NSAID (indomethacin), so future studies are needed to assess the generalizability of our findings and to establish a direct link to the human condition.

Clostridium difficile Alters the Structure and Metabolism of Distinct Cecal Microbiomes during Initial Infection To Promote Sustained Colonization

Susceptibility to Clostridium difficile infection (CDI) is primarily associated with previous exposure to antibiotics, which compromise the structure and function of the gut bacterial community. Specific antibiotic classes correlate more strongly with recurrent or persistent C. difficile infection. As such, we utilized a mouse model of infection to explore the effect of distinct antibiotic classes on the impact that infection has on community-level transcription and metabolic signatures shortly following pathogen colonization and how those changes may associate with persistence of C. difficile Untargeted metabolomic analysis revealed that C. difficile infection had significantly larger impacts on the metabolic environment across cefoperazone- and streptomycin-pretreated mice, which became persistently colonized compared to clindamycin-pretreated mice, where infection quickly became undetectable. Through metagenome-enabled metatranscriptomics, we observed that transcripts for genes associated with carbon and energy acquisition were greatly reduced in infected animals, suggesting that those niches were instead occupied by C. difficile Furthermore, the largest changes in transcription were seen in the least abundant species, indicating that C. difficile may "attack the loser" in gut environments where sustained infection occurs more readily. Overall, our results suggest that C. difficile is able to restructure the nutrient-niche landscape in the gut to promote persistent infection.IMPORTANCEClostridium difficile has become the most common single cause of hospital-acquired infection over the last decade in the United States. Colonization resistance to the nosocomial pathogen is primarily provided by the gut microbiota, which is also involved in clearing the infection as the community recovers from perturbation. As distinct antibiotics are associated with different risk levels for CDI, we utilized a mouse model of infection with 3 separate antibiotic pretreatment regimens to generate alternative gut microbiomes that each allowed for C. difficile colonization but varied in clearance rate. To assess community-level dynamics, we implemented an integrative multi-omics approach that revealed that infection significantly changed many aspects of the gut community. The degree to which the community changed was inversely correlated with clearance during the first 6 days of infection, suggesting that C. difficile differentially modifies the gut environment to promote persistence. This is the first time that metagenome-enabled metatranscriptomics have been employed to study the behavior of a host-associated microbiota in response to an infection. Our results allow for a previously unseen understanding of the ecology associated with C. difficile infection and provide the groundwork for identification of context-specific probiotic therapies.

Identifying and Overcoming Threats to Reproducibility, Replicability, Robustness, and Generalizability in Microbiome Research

The "reproducibility crisis" in science affects microbiology as much as any other area of inquiry, and microbiologists have long struggled to make their research reproducible. We need to respect that ensuring that our methods and results are sufficiently transparent is difficult. This difficulty is compounded in interdisciplinary fields such as microbiome research. There are many reasons why a researcher is unable to reproduce a previous result, and even if a result is reproducible, it may not be correct. Furthermore, failures to reproduce previous results have much to teach us about the scientific process and microbial life itself. This Perspective delineates a framework for identifying and overcoming threats to reproducibility, replicability, robustness, and generalizability of microbiome research. Instead of seeing signs of a crisis in others' work, we need to appreciate the technical and social difficulties that limit reproducibility in the work of others as well as our own.

Leveraging Existing 16S rRNA Gene Surveys To Identify Reproducible Biomarkers in Individuals with Colorectal Tumors

An increasing body of literature suggests that both individual and collections of bacteria are associated with the progression of colorectal cancer. As the number of studies investigating these associations increases and the number of subjects in each study increases, a meta-analysis to identify the associations that are the most predictive of disease progression is warranted. We analyzed previously published 16S rRNA gene sequencing data collected from feces and colon tissue. We quantified the odds ratios (ORs) for individual bacterial taxa that were associated with an individual having tumors relative to a normal colon. Among the fecal samples, there were no taxa that had significant ORs associated with adenoma and there were 8 taxa with significant ORs associated with carcinoma. Similarly, among the tissue samples, there were no taxa that had a significant OR associated with adenoma and there were 3 taxa with significant ORs associated with carcinoma. Among the significant ORs, the association between individual taxa and tumor diagnosis was equal to or below 7.11. Because individual taxa had limited association with tumor diagnosis, we trained Random Forest classification models using only the taxa that had significant ORs, using the entire collection of taxa found in each study, and using operational taxonomic units defined based on a 97% similarity threshold. All training approaches yielded similar classification success as measured using the area under the curve. The ability to correctly classify individuals with adenomas was poor, and the ability to classify individuals with carcinomas was considerably better using sequences from feces or tissue.IMPORTANCE Colorectal cancer is a significant and growing health problem in which animal models and epidemiological data suggest that the colonic microbiota have a role in tumorigenesis. These observations indicate that the colonic microbiota is a reservoir of biomarkers that may improve our ability to detect colonic tumors using noninvasive approaches. This meta-analysis identifies and validates a set of 8 bacterial taxa that can be used within a Random Forest modeling framework to differentiate individuals as having normal colons or carcinomas. When models trained using one data set were tested on other data sets, the models performed well. These results lend support to the use of fecal biomarkers for the detection of tumors. Furthermore, these biomarkers are plausible candidates for further mechanistic studies into the role of the gut microbiota in tumorigenesis.

Biogeography and environmental conditions shape bacteriophage-bacteria networks across the human microbiome

Viruses and bacteria are critical components of the human microbiome and play important roles in health and disease. Most previous work has relied on studying bacteria and viruses independently, thereby reducing them to two separate communities. Such approaches are unable to capture how these microbial communities interact, such as through processes that maintain community robustness or allow phage-host populations to co-evolve. We implemented a network-based analytical approach to describe phage-bacteria network diversity throughout the human body. We built these community networks using a machine learning algorithm to predict which phages could infect which bacteria in a given microbiome. Our algorithm was applied to paired viral and bacterial metagenomic sequence sets from three previously published human cohorts. We organized the predicted interactions into networks that allowed us to evaluate phage-bacteria connectedness across the human body. We observed evidence that gut and skin network structures were person-specific and not conserved among cohabitating family members. High-fat diets appeared to be associated with less connected networks. Network structure differed between skin sites, with those exposed to the external environment being less connected and likely more susceptible to network degradation by microbial extinction events. This study quantified and contrasted the diversity of virome-microbiome networks across the human body and illustrated how environmental factors may influence phage-bacteria interactive dynamics. This work provides a baseline for future studies to better understand system perturbations, such as disease states, through ecological networks.

The human microbiome, the collection of microbial communities that colonize the human body, is a crucial component to health and disease. Two major components of the human microbiome are the bacterial and viral communities. These communities have primarily been studied separately using metrics of community composition and diversity. These approaches have failed to capture the complex dynamics of interacting bacteria and phage communities, which frequently share genetic information and work together to maintain ecosystem homestatsis (e.g. kill-the-winner dynamics). Removal of bacteria or phage can disrupt or even collapse those ecosystems. Relationship-based network approaches allow us to capture this interaction information. Using this network-based approach with three independent human cohorts, we were able to present an initial understanding of how phage-bacteria networks differ throughout the human body, so as to provide a baseline for future studies of how and why microbiome networks differ in disease states.

Spatial Variation of the Native Colon Microbiota in Healthy Adults

The microbiome has been implicated in the development of colorectal cancer and inflammatory bowel diseases. The specific traits of these diseases vary along the axis of the digestive tract. Further, variation in the structure of the gut microbiota has been associated with both diseases. We profiled the microbiota of the healthy proximal and distal mucosa and lumen to better understand how bacterial populations vary along the colon. We used a two-colonoscope approach to sample proximal and distal mucosal and luminal contents from the colons of 20 healthy subjects that had not undergone any bowel preparation procedure. The biopsies and home-collected stool were subjected to 16S rRNA gene sequencing, and random forest classification models were built using taxa abundance and location to identify microbiota specific to each site. The right mucosa and lumen had the most similar community structures of the five sites we considered from each subject. The distal mucosa had higher relative abundance of Finegoldia, Murdochiella, Peptoniphilus, Porphyromonas, and Anaerococcus The proximal mucosa had more of the genera Enterobacteriaceae, Bacteroides, and Pseudomonas The classification model performed well when classifying mucosal samples into proximal or distal sides (AUC = 0.808). Separating proximal and distal luminal samples proved more challenging (AUC = 0.599), and specific microbiota that differentiated the two were hard to identify. By sampling the unprepped colon, we identified distinct bacterial populations native to the proximal and distal sides. Further investigation of these bacteria may elucidate if and how these groups contribute to different disease processes on their respective sides of the colon. Cancer Prev Res; 11(7); 393-402. ©2018 AACR.

Normalization of the microbiota in patients after treatment for colonic lesions

BACKGROUND

Colorectal cancer is a worldwide health problem. Despite growing evidence that members of the gut microbiota can drive tumorigenesis, little is known about what happens to it after treatment for an adenoma or carcinoma. This study tested the hypothesis that treatment for adenoma or carcinoma alters the abundance of bacterial populations associated with disease to those associated with a normal colon. We tested this hypothesis by sequencing the 16S rRNA genes in the feces of 67 individuals before and after treatment for adenoma (N = 22), advanced adenoma (N = 19), and carcinoma (N = 26).

RESULTS

There were small changes to the bacterial community associated with adenoma or advanced adenoma and large changes associated with carcinoma. The communities from patients with carcinomas changed significantly more than those with adenoma following treatment (P value < 0.001). Although treatment was associated with intrapersonal changes, the change in the abundance of individual OTUs in response to treatment was not consistent within diagnosis groups (P value > 0.05). Because the distribution of OTUs across patients and diagnosis groups was irregular, we used the random forest machine learning algorithm to identify groups of OTUs that could be used to classify pre and post-treatment samples for each of the diagnosis groups. Although the adenoma and carcinoma models could reliably differentiate between the pre- and post-treatment samples (P value < 0.001), the advanced-adenoma model could not (P value = 0.61). Furthermore, there was little overlap between the OTUs that were indicative of each treatment. To determine whether individuals who underwent treatment were more likely to have OTUs associated with normal colons we used a larger cohort that contained individuals with normal colons and those with adenomas, advanced adenomas, and carcinomas. We again built random forest models and measured the change in the positive probability of having one of the three diagnoses to assess whether the post-treatment samples received the same classification as the pre-treatment samples. Samples from patients who had carcinomas changed toward a microbial milieu that resembles the normal colon after treatment (P value < 0.001). Finally, we were unable to detect any significant differences in the microbiota of individuals treated with surgery alone and those treated with chemotherapy or chemotherapy and radiation (P value > 0.05).

CONCLUSIONS

By better understanding the response of the microbiota to treatment for adenomas and carcinomas, it is likely that biomarkers will eventually be validated that can be used to quantify the risk of recurrence and the likelihood of survival. Although it was difficult to identify significant differences between pre- and post-treatment samples from patients with adenoma and advanced adenoma, this was not the case for carcinomas. Not only were there large changes in pre- versus post-treatment samples for those with carcinoma, but also these changes were toward a more normal microbiota.

Clostridium difficile Colonizes Alternative Nutrient Niches during Infection across Distinct Murine Gut Microbiomes

Clostridium difficile is the largest single cause of hospital-acquired infection in the United States. A major risk factor for Clostridium difficile infection (CDI) is prior exposure to antibiotics, as they disrupt the gut bacterial community which protects from C. difficile colonization. Multiple antibiotic classes have been associated with CDI susceptibility, many leading to distinct community structures stemming from variation in bacterial targets of action. These community structures present separate metabolic challenges to C. difficile. Therefore, we hypothesized that the pathogen adapts its physiology to the nutrients within different gut environments. Utilizing an in vivo CDI model, we demonstrated that C. difficile highly colonized ceca of mice pretreated with any of three antibiotics from distinct classes. Levels of C. difficile spore formation and toxin activity varied between animals based on the antibiotic pretreatment. These physiologic processes in C. difficile are partially regulated by environmental nutrient concentrations. To investigate metabolic responses of the bacterium in vivo, we performed transcriptomic analysis of C. difficile from ceca of infected mice across pretreatments. This revealed heterogeneous expression in numerous catabolic pathways for diverse growth substrates. To assess which resources C. difficile exploited, we developed a genome-scale metabolic model with a transcriptome-enabled metabolite scoring algorithm integrating network architecture. This platform identified nutrients that C. difficile used preferentially between pretreatments, which were validated through untargeted mass spectrometry of each microbiome. Our results supported the hypothesis that C. difficile inhabits alternative nutrient niches across cecal microbiomes with increased preference for nitrogen-containing carbon sources, particularly Stickland fermentation substrates and host-derived glycans. IMPORTANCE Infection by the bacterium Clostridium difficile causes an inflammatory diarrheal disease which can become life threatening and has grown to be the most prevalent nosocomial infection. Susceptibility to C. difficile infection is strongly associated with previous antibiotic treatment, which disrupts the gut microbiota and reduces its ability to prevent colonization. In this study, we demonstrated that C. difficile altered pathogenesis between hosts pretreated with antibiotics from separate classes and exploited different nutrient sources across these environments. Our metabolite score calculation also provides a platform to study nutrient requirements of pathogens during an infection. Our results suggest that C. difficile colonization resistance is mediated by multiple groups of bacteria competing for several subsets of nutrients and could explain why total reintroduction of competitors through fecal microbial transplant currently is the most effective treatment for recurrent CDI. This work could ultimately contribute to the identification of targeted, context-dependent measures that prevent or reduce C. difficile colonization, including pre- and probiotic therapies.

Preprinting Microbiology

The field of microbiology has experienced significant growth due to transformative advances in technology and the influx of scientists driven by a curiosity to understand how microbes sustain myriad biochemical processes that maintain Earth. With this explosion in scientific output, a significant bottleneck has been the ability to rapidly disseminate new knowledge to peers and the public. Preprints have emerged as a tool that a growing number of microbiologists are using to overcome this bottleneck. Posting preprints can help to transparently recruit a more diverse pool of reviewers prior to submitting to a journal for formal peer review. Although the use of preprints is still limited in the biological sciences, early indications are that preprints are a robust tool that can complement and enhance peer-reviewed publications. As publishing moves to embrace advances in Internet technology, there are many opportunities for preprints and peer-reviewed journals to coexist in the same ecosystem.

DNA from fecal immunochemical test can replace stool for detection of colonic lesions using a microbiota-based model

BACKGROUND

There is a significant demand for colorectal cancer (CRC) screening methods that are noninvasive, inexpensive, and capable of accurately detecting early stage tumors. It has been shown that models based on the gut microbiota can complement the fecal occult blood test and fecal immunochemical test (FIT). However, a barrier to microbiota-based screening is the need to collect and store a patient's stool sample.

RESULTS

Using stool samples collected from 404 patients, we tested whether the residual buffer containing resuspended feces in FIT cartridges could be used in place of intact stool samples. We found that the bacterial DNA isolated from FIT cartridges largely recapitulated the community structure and membership of patients' stool microbiota and that the abundance of bacteria associated with CRC were conserved. We also found that models for detecting CRC that were generated using bacterial abundances from FIT cartridges were equally predictive as models generated using bacterial abundances from stool.

CONCLUSIONS

These findings demonstrate the potential for using residual buffer from FIT cartridges in place of stool for microbiota-based screening for CRC. This may reduce the need to collect and process separate stool samples and may facilitate combining FIT and microbiota-based biomarkers into a single test. Additionally, FIT cartridges could constitute a novel data source for studying the role of the microbiome in cancer and other diseases.

Microbiota-based model improves the sensitivity of fecal immunochemical test for detecting colonic lesions

BACKGROUND

Colorectal cancer (CRC) is the second leading cause of death among cancers in the United States. Although individuals diagnosed early have a greater than 90% chance of survival, more than one-third of individuals do not adhere to screening recommendations partly because the standard diagnostics, colonoscopy and sigmoidoscopy, are expensive and invasive. Thus, there is a great need to improve the sensitivity of non-invasive tests to detect early stage cancers and adenomas. Numerous studies have identified shifts in the composition of the gut microbiota associated with the progression of CRC, suggesting that the gut microbiota may represent a reservoir of biomarkers that would complement existing non-invasive methods such as the widely used fecal immunochemical test (FIT).

METHODS

We sequenced the 16S rRNA genes from the stool samples of 490 patients. We used the relative abundances of the bacterial populations within each sample to develop a random forest classification model that detects colonic lesions using the relative abundance of gut microbiota and the concentration of hemoglobin in stool.

RESULTS

The microbiota-based random forest model detected 91.7% of cancers and 45.5% of adenomas while FIT alone detected 75.0% and 15.7%, respectively. Of the colonic lesions missed by FIT, the model detected 70.0% of cancers and 37.7% of adenomas. We confirmed known associations of Porphyromonas assaccharolytica, Peptostreptococcus stomatis, Parvimonas micra, and Fusobacterium nucleatum with CRC. Yet, we found that the loss of potentially beneficial organisms, such as members of the Lachnospiraceae, was more predictive for identifying patients with adenomas when used in combination with FIT.

CONCLUSIONS

These findings demonstrate the potential for microbiota analysis to complement existing screening methods to improve detection of colonic lesions.

Multicenter Comparison of Lung and Oral Microbiomes of HIV-infected and HIV-uninfected Individuals

RATIONALE

Improved understanding of the lung microbiome in HIV-infected individuals could lead to better strategies for diagnosis, therapy, and prophylaxis of HIV-associated pneumonias. Differences in the oral and lung microbiomes in HIV-infected and HIV-uninfected individuals are not well defined. Whether highly active antiretroviral therapy influences these microbiomes is unclear.

OBJECTIVES

We determined whether oral and lung microbiomes differed in clinically healthy groups of HIV-infected and HIV-uninfected subjects.

METHODS

Participating sites in the Lung HIV Microbiome Project contributed bacterial 16S rRNA sequencing data from oral washes and bronchoalveolar lavages (BALs) obtained from HIV-uninfected individuals (n = 86), HIV-infected individuals who were treatment naive (n = 18), and HIV-infected individuals receiving antiretroviral therapy (n = 38).

MEASUREMENTS AND MAIN RESULTS

Microbial populations differed in the oral washes among the subject groups (Streptococcus, Actinomyces, Rothia, and Atopobium), but there were no individual taxa that differed among the BALs. Comparison of oral washes and BALs demonstrated similar patterns from HIV-uninfected individuals and HIV-infected individuals receiving antiretroviral therapy, with multiple taxa differing in abundance. The pattern observed from HIV-infected individuals who were treatment naive differed from the other two groups, with differences limited to Veillonella, Rothia, and Granulicatella. CD4 cell counts did not influence the oral or BAL microbiome in these relatively healthy, HIV-infected subjects.

CONCLUSIONS

The overall similarity of the microbiomes in participants with and without HIV infection was unexpected, because HIV-infected individuals with relatively preserved CD4 cell counts are at higher risk for lower respiratory tract infections, indicating impaired local immune function.

Sequencing 16S rRNA gene fragments using the PacBio SMRT DNA sequencing system

Over the past 10 years, microbial ecologists have largely abandoned sequencing 16S rRNA genes by the Sanger sequencing method and have instead adopted highly parallelized sequencing platforms. These new platforms, such as 454 and Illumina's MiSeq, have allowed researchers to obtain millions of high quality but short sequences. The result of the added sequencing depth has been significant improvements in experimental design. The tradeoff has been the decline in the number of full-length reference sequences that are deposited into databases. To overcome this problem, we tested the ability of the PacBio Single Molecule, Real-Time (SMRT) DNA sequencing platform to generate sequence reads from the 16S rRNA gene. We generated sequencing data from the V4, V3-V5, V1-V3, V1-V5, V1-V6, and V1-V9 variable regions from within the 16S rRNA gene using DNA from a synthetic mock community and natural samples collected from human feces, mouse feces, and soil. The mock community allowed us to assess the actual sequencing error rate and how that error rate changed when different curation methods were applied. We developed a simple method based on sequence characteristics and quality scores to reduce the observed error rate for the V1-V9 region from 0.69 to 0.027%. This error rate is comparable to what has been observed for the shorter reads generated by 454 and Illumina's MiSeq sequencing platforms. Although the per base sequencing cost is still significantly more than that of MiSeq, the prospect of supplementing reference databases with full-length sequences from organisms below the limit of detection from the Sanger approach is exciting.

De novo clustering methods outperform reference-based methods for assigning 16S rRNA gene sequences to operational taxonomic units

Background. 16S rRNA gene sequences are routinely assigned to operational taxonomic units (OTUs) that are then used to analyze complex microbial communities. A number of methods have been employed to carry out the assignment of 16S rRNA gene sequences to OTUs leading to confusion over which method is optimal. A recent study suggested that a clustering method should be selected based on its ability to generate stable OTU assignments that do not change as additional sequences are added to the dataset. In contrast, we contend that the quality of the OTU assignments, the ability of the method to properly represent the distances between the sequences, is more important. Methods. Our analysis implemented six de novo clustering algorithms including the single linkage, complete linkage, average linkage, abundance-based greedy clustering, distance-based greedy clustering, and Swarm and the open and closed-reference methods. Using two previously published datasets we used the Matthew's Correlation Coefficient (MCC) to assess the stability and quality of OTU assignments. Results. The stability of OTU assignments did not reflect the quality of the assignments. Depending on the dataset being analyzed, the average linkage and the distance and abundance-based greedy clustering methods generated OTUs that were more likely to represent the actual distances between sequences than the open and closed-reference methods. We also demonstrated that for the greedy algorithms VSEARCH produced assignments that were comparable to those produced by USEARCH making VSEARCH a viable free and open source alternative to USEARCH. Further interrogation of the reference-based methods indicated that when USEARCH or VSEARCH were used to identify the closest reference, the OTU assignments were sensitive to the order of the reference sequences because the reference sequences can be identical over the region being considered. More troubling was the observation that while both USEARCH and VSEARCH have a high level of sensitivity to detect reference sequences, the specificity of those matches was poor relative to the true best match. Discussion. Our analysis calls into question the quality and stability of OTU assignments generated by the open and closed-reference methods as implemented in current version of QIIME. This study demonstrates that de novo methods are the optimal method of assigning sequences into OTUs and that the quality of these assignments needs to be assessed for multiple methods to identify the optimal clustering method for a particular dataset.

Intra- and interindividual variations mask interspecies variation in the microbiota of sympatric peromyscus populations

Using populations of two sympatric Peromyscus species, we characterized the importance of the host species, physiology, environment, diet, and other factors in shaping the structure and dynamics of their gut microbiota. We performed a capture-mark-release experiment in which we obtained 16S rRNA gene sequence data from 49 animals at multiple time points. In addition, we performed 18S rRNA gene sequencing of the same samples to characterize the diet of each individual. Our analysis could not distinguish between the two species of Peromyscus on the basis of the structures of their microbiotas. However, we did observe a set of bacterial populations that were found in every animal. Most notable were abundant representatives of the genera Lactobacillus and Helicobacter. When we combined the 16S and 18S rRNA gene sequence analyses, we were unable to distinguish the communities on the basis of the animal's diet. Furthermore, there were no discernible differences in the structure of the gut communities based on the capture site or their developmental or physiological status. Finally, in contrast to humans, where each individual has a unique microbiota when sampled over years, among the animals captured in this study, the uniqueness of each microbiota was lost within a week of the original sampling. Wild populations provide an opportunity to study host-microbiota interactions as they originally evolved, and the ability to perform natural experiments will facilitate a greater understanding of the factors that shape the structure and function of the gut microbiota.

The daily dynamics of cystic fibrosis airway microbiota during clinical stability and at exacerbation

BACKGROUND

Recent work indicates that the airways of persons with cystic fibrosis (CF) typically harbor complex bacterial communities. However, the day-to-day stability of these communities is unknown. Further, airway community dynamics during the days corresponding to the onset of symptoms of respiratory exacerbation have not been studied.

RESULTS

Using 16S rRNA amplicon sequencing of 95 daily sputum specimens collected from four adults with CF, we observed varying degrees of day-to-day stability in airway bacterial community structures during periods of clinical stability. Differences were observed between study subjects with respect to the degree of community changes at the onset of exacerbation. Decreases in the relative abundance of dominant taxa were observed in three subjects at exacerbation. We observed no relationship between total bacterial load and clinical status and detected no viruses by multiplex PCR.

CONCLUSION

CF airway microbial communities are relatively stable during periods of clinical stability. Changes in microbial community structure are associated with some, but not all, pulmonary exacerbations, supporting previous observations suggesting that distinct types of exacerbations occur in CF. Decreased abundance of species that are dominant at baseline suggests a role for less abundant taxa in some exacerbations. Daily sampling revealed patterns of change in microbial community structures that may prove useful in the prediction and management of CF pulmonary exacerbations.

The human gut microbiome as a screening tool for colorectal cancer

Recent studies have suggested that the gut microbiome may be an important factor in the development of colorectal cancer. Abnormalities in the gut microbiome have been reported in patients with colorectal cancer; however, this microbial community has not been explored as a potential screen for early-stage disease. We characterized the gut microbiome in patients from three clinical groups representing the stages of colorectal cancer development: healthy, adenoma, and carcinoma. Analysis of the gut microbiome from stool samples revealed both an enrichment and depletion of several bacterial populations associated with adenomas and carcinomas. Combined with known clinical risk factors of colorectal cancer (e.g., BMI, age, race), data from the gut microbiome significantly improved the ability to differentiate between healthy, adenoma, and carcinoma clinical groups relative to risk factors alone. Using Bayesian methods, we determined that using gut microbiome data as a screening tool improved the pretest to posttest probability of adenoma more than 50-fold. For example, the pretest probability in a 65-year-old was 0.17% and, after using the microbiome data, this increased to 10.67% (1 in 9 chance of having an adenoma). Taken together, the results of our study demonstrate the feasibility of using the composition of the gut microbiome to detect the presence of precancerous and cancerous lesions. Furthermore, these results support the need for more cross-sectional studies with diverse populations and linkage to other stool markers, dietary data, and personal health information.

The dynamics of a family's gut microbiota reveal variations on a theme

BACKGROUND

It is clear that the structure and function of the human microbiota has significant impact on maintenance of health and yet the factors that give rise to an adult microbiota are poorly understood. A combination of genetics, diet, environment, and life history are all thought to impact the development of the gut microbiome. Here we study a chronosequence of the gut microbiota found in eight individuals from a family consisting of two parents and six children ranging in age from two months to ten years old.

RESULTS

Using 16S rRNA gene and metagenomic shotgun sequence data, it was possible to distinguish the family from a cohort of normal individuals living in the same geographic region and to differentiate each family member. Interestingly, there was a significant core membership to the family members' microbiota where the abundance of this core accounted for the differences between individuals. It was clear that the introduction of solids represents a significant transition in the development of a mature microbiota. This transition was associated with increased diversity, decreased stability, and the colonization of significant abundances of Bacteroidetes and Clostridiales. Although the children and mother shared essentially the identical diet and environment, the children's microbiotas were not significantly more similar to their mother than they were to their father.

CONCLUSIONS

This analysis underscores the complex interactions that give rise to a personalized microbiota and suggests the value of studying families as a surrogate for longitudinal studies.

Dynamics and associations of microbial community types across the human body

A primary goal of the Human Microbiome Project (HMP) was to provide a reference collection of 16S ribosomal RNA gene sequences collected from sites across the human body that would allow microbiologists to better associate changes in the microbiome with changes in health. The HMP Consortium has reported the structure and function of the human microbiome in 300 healthy adults at 18 body sites from a single time point. Using additional data collected over the course of 12-18 months, we used Dirichlet multinomial mixture models to partition the data into community types for each body site and made three important observations. First, there were strong associations between whether individuals had been breastfed as an infant, their gender, and their level of education with their community types at several body sites. Second, although the specific taxonomic compositions of the oral and gut microbiomes were different, the community types observed at these sites were predictive of each other. Finally, over the course of the sampling period, the community types from sites within the oral cavity were the least stable, whereas those in the vagina and gut were the most stable. Our results demonstrate that even with the considerable intra- and interpersonal variation in the human microbiome, this variation can be partitioned into community types that are predictive of each other and are probably the result of life-history characteristics. Understanding the diversity of community types and the mechanisms that result in an individual having a particular type or changing types, will allow us to use their community types to assess disease risk and to personalize therapies.

Structure of the gut microbiome following colonization with human feces determines colonic tumor burden

BACKGROUND

A growing body of evidence indicates that the gut microbiome plays a role in the development of colorectal cancer (CRC). Patients with CRC harbor gut microbiomes that are structurally distinct from those of healthy individuals; however, without the ability to track individuals during disease progression, it has not been possible to observe changes in the microbiome over the course of tumorigenesis. Mouse models have demonstrated that these changes can further promote colonic tumorigenesis. However, these models have relied upon mouse-adapted bacterial populations and so it remains unclear which human-adapted bacterial populations are responsible for modulating tumorigenesis.

RESULTS

We transplanted fecal microbiota from three CRC patients and three healthy individuals into germ-free mice, resulting in six structurally distinct microbial communities. Subjecting these mice to a chemically induced model of CRC resulted in different levels of tumorigenesis between mice. Differences in the number of tumors were strongly associated with the baseline microbiome structure in mice, but not with the cancer status of the human donors. Partitioning of baseline communities into enterotypes by Dirichlet multinomial mixture modeling resulted in three enterotypes that corresponded with tumor burden. The taxa most strongly positively correlated with increased tumor burden were members of the Bacteroides, Parabacteroides, Alistipes, and Akkermansia, all of which are Gram-negative. Members of the Gram-positive Clostridiales, including multiple members of Clostridium Group XIVa, were strongly negatively correlated with tumors. Analysis of the inferred metagenome of each community revealed a negative correlation between tumor count and the potential for butyrate production, and a positive correlation between tumor count and the capacity for host glycan degradation. Despite harboring distinct gut communities, all mice underwent conserved structural changes over the course of the model. The extent of these changes was also correlated with tumor incidence.

CONCLUSION

Our results suggest that the initial structure of the microbiome determines susceptibility to colonic tumorigenesis. There appear to be opposing roles for certain Gram-negative (Bacteroidales and Verrucomicrobia) and Gram-positive (Clostridiales) bacteria in tumor susceptibility. Thus, the impact of community structure is potentially mediated by the balance between protective, butyrate-producing populations and inflammatory, mucin-degrading populations.

Comparison of the respiratory microbiome in healthy nonsmokers and smokers

RATIONALE

Results from 16S rDNA-encoding gene sequence-based, culture-independent techniques have led to conflicting conclusions about the composition of the lower respiratory tract microbiome.

OBJECTIVES

To compare the microbiome of the upper and lower respiratory tract in healthy HIV-uninfected nonsmokers and smokers in a multicenter cohort.

METHODS

Participants were nonsmokers and smokers without significant comorbidities. Oral washes and bronchoscopic alveolar lavages were collected in a standardized manner. Sequence analysis of bacterial 16S rRNA-encoding genes was performed, and the neutral model in community ecology was used to identify bacteria that were the most plausible members of a lung microbiome.

MEASUREMENTS AND MAIN RESULTS

Sixty-four participants were enrolled. Most bacteria identified in the lung were also in the mouth, but specific bacteria such as Enterobacteriaceae, Haemophilus, Methylobacterium, and Ralstonia species were disproportionally represented in the lungs compared with values predicted by the neutral model. Tropheryma was also in the lung, but not the mouth. Mouth communities differed between nonsmokers and smokers in species such as Porphyromonas, Neisseria, and Gemella, but lung bacterial populations did not.

CONCLUSIONS

This study is the largest to examine composition of the lower respiratory tract microbiome in healthy individuals and the first to use the neutral model to compare the lung to the mouth. Specific bacteria appear in significantly higher abundance in the lungs than would be expected if they originated from the mouth, demonstrating that the lung microbiome does not derive entirely from the mouth. The mouth microbiome differs in nonsmokers and smokers, but lung communities were not significantly altered by smoking.

Multiphasic analysis of the temporal development of the distal gut microbiota in patients following ileal pouch anal anastomosis

BACKGROUND

The indigenous gut microbiota are thought to play a crucial role in the development and maintenance of the abnormal inflammatory responses that are the hallmark of inflammatory bowel disease. Direct tests of the role of the gut microbiome in these disorders are typically limited by the fact that sampling of the microbiota generally occurs once disease has become manifest. This limitation could potentially be circumvented by studying patients who undergo total proctocolectomy with ileal pouch anal anastomosis (IPAA) for the definitive treatment of ulcerative colitis. A subset of patients who undergo IPAA develops an inflammatory condition known as pouchitis, which is thought to mirror the pathogenesis of ulcerative colitis. Following the development of the microbiome of the pouch would allow characterization of the microbial community that predates the development of overt disease.

RESULTS

We monitored the development of the pouch microbiota in four patients who underwent IPAA. Mucosal and luminal samples were obtained prior to takedown of the diverting ileostomy and compared to samples obtained 2, 4 and 8 weeks after intestinal continuity had been restored. Through the combined analysis of 16S rRNA-encoding gene amplicons, targeted 16S amplification and microbial cultivation, we observed major changes in structure and function of the pouch microbiota following ileostomy. There is a relative increase in anaerobic microorganisms with the capacity for fermentation of complex carbohydrates, which corresponds to the physical stasis of intestinal contents in the ileal pouch. Compared to the microbiome structure encountered in the colonic mucosa of healthy individuals, the pouch microbial community in three of the four individuals was quite distinct. In the fourth patient, a community that was much like that seen in a healthy colon was established, and this patient also had the most benign clinical course of the four patients, without the development of pouchitis 2 years after IPAA.

CONCLUSIONS

The microbiota that inhabit the ileal-anal pouch of patients who undergo IPAA for treatment of ulcerative colitis demonstrate significant structural and functional changes related to the restoration of fecal flow. Our preliminary results suggest once the pouch has assumed the physiologic role previously played by the intact colon, the precise structure and function of the pouch microbiome, relative to a normal colonic microbiota, will determine if there is establishment of a stable, healthy mucosal environment or the reinitiation of the pathogenic cascade that results in intestinal inflammation.

Stabilization of the murine gut microbiome following weaning

Ecologists hypothesize that community structure and stability affect productivity, sensitivity to invasion and extinction, and resilience and resistance to perturbations. Viewed in the context of the gut microbiome, the stability of the gut community is important for understanding the effects of antibiotics, diet change and other perturbations on host health and colonization resistance. Here we describe the dynamics of a self-contained community, the murine gut microbiome. Using 16S rRNA gene sequencing of fecal samples collected daily from individual mice, we characterized the community membership and structure to determine whether there were significant changes in the gut community during the first year of life. Based on analysis of molecular variance, we observed two community states. The first was observed in the 10 days following weaning and the second was observed by 15 days following weaning. Interestingly, these two states had the same bacterial populations, but those populations had different relative abundances in the two states. By calculating the root mean squared distances between samples collected in the early and late states for each mouse, we observed that the late state was more stable than the early state. This increase in stability was not correlated with increased taxonomic richness, taxonomic diversity, or phylogenetic diversity. In the absence of an experimentally induced perturbation, the second community state was relatively constant through 364 days post weaning. These results suggest a high degree of stability in the microbiome once the community reached the second state.

The metatranscriptome of a deep-sea hydrothermal plume is dominated by water column methanotrophs and lithotrophs

Microorganisms mediate geochemical processes in deep-sea hydrothermal vent plumes, which are a conduit for transfer of elements and energy from the subsurface to the oceans. Despite this important microbial influence on marine geochemistry, the ecology and activity of microbial communities in hydrothermal plumes is largely unexplored. Here, we use a coordinated metagenomic and metatranscriptomic approach to compare microbial communities in Guaymas Basin hydrothermal plumes to background waters above the plume and in the adjacent Carmen Basin. Despite marked increases in plume total RNA concentrations (3-4 times) and microbially mediated manganese oxidation rates (15-125 times), plume and background metatranscriptomes were dominated by the same groups of methanotrophs and chemolithoautotrophs. Abundant community members of Guaymas Basin seafloor environments (hydrothermal sediments and chimneys) were not prevalent in the plume metatranscriptome. De novo metagenomic assembly was used to reconstruct genomes of abundant populations, including Marine Group I archaea, Methylococcaceae, SAR324 Deltaproteobacteria and SUP05 Gammaproteobacteria. Mapping transcripts to these genomes revealed abundant expression of genes involved in the chemolithotrophic oxidation of ammonia (amo), methane (pmo) and sulfur (sox). Whereas amo and pmo gene transcripts were abundant in both plume and background, transcripts of sox genes for sulfur oxidation from SUP05 groups displayed a 10-20-fold increase in plumes. We conclude that the biogeochemistry of Guaymas Basin hydrothermal plumes is mediated by microorganisms that are derived from seawater rather than from seafloor hydrothermal environments such as chimneys or sediments, and that hydrothermal inputs serve as important electron donors for primary production in the deep Gulf of California.