Eric A. Franzosa

The Dynamics of the Human Infant Gut Microbiome in Development and in Progression toward Type 1 Diabetes

Colonization of the fetal and infant gut microbiome results in dynamic changes in diversity, which can impact disease susceptibility. To examine the relationship between human gut microbiome dynamics throughout infancy and type 1 diabetes (T1D), we examined a cohort of 33 infants genetically predisposed to T1D. Modeling trajectories of microbial abundances through infancy revealed a subset of microbial relationships shared across most subjects. Although strain composition of a given species was highly variable between individuals, it was stable within individuals throughout infancy. Metabolic composition and metabolic pathway abundance remained constant across time. A marked drop in alpha-diversity was observed in T1D progressors in the time window between seroconversion and T1D diagnosis, accompanied by spikes in inflammation-favoring organisms, gene functions, and serum and stool metabolites. This work identifies trends in the development of the human infant gut microbiome along with specific alterations that precede T1D onset and distinguish T1D progressors from nonprogressors.

MetaPhlAn2 for enhanced metagenomic taxonomic profiling.

 Profiling of all domains of life. Marker and quasi-marker genes are now identified not only for microbes (Bacteria and Archaea), but also for viruses and Eukaryotic microbes (Fungi, Protozoa) that are crucial components of microbial communities.  A 6-fold increase in the number of considered species. Markers are now identified from >16,000 reference genomes and >7,000 unique species, dramatically expanding the comprehensiveness of the method. The new pipeline for identifying marker genes is also scalable to the quickly increasing number of reference genomes. See Supplementary Tables 1-3.  Introduction of the concept of quasi-markers, allowing more comprehensive and accurate profiling. For species with less than 200 markers, MetaPhlAn2 adopts additional quasi-marker sequences (Supplementary Note 2) that are occasionally present in other genomes (because of vertical conservation or horizontal transfer). At profiling time, if no other markers of the potentially confounding species are detected, the corresponding quasi-local markers are used to improve the quality and accuracy of the profiling.  Addition of strain-specific barcoding for microbial strain tracking. MetaPhlAn2 includes a completely new feature that exploits marker combinations to perform species-specific and genus-specific “barcoding” for strains in metagenomic samples (Supplementary Note 7). This feature can be used for culture-free pathogen tracking in epidemiology studies and strain tracking across microbiome samples. See Supplementary Figs. 12-20.  Strain-level identification for organisms with sequenced genomes. For the case in which a microbiome includes strains that are very close to one of those already sequenced, MetaPhlAn2 is now able to identify such strains and readily reports their abundances. See Supplementary Note 7, Supplementary Table 13, and Supplementary Fig. 21.  Improvement of false positive and false negative rates. Improvements in the underlying pipeline for identifying marker genes (including the increment of the adopted genomes and the use of quasi-markers) and the profiling procedure resulted in much improved quantitative performances (higher correlation with true abundances, lower false positive and false negative rates). See the validation on synthetic metagenomes in Supplementary Note 4.  Estimation of the percentage of reads mapped against known reference genomes. MetaPhlAn2 is now able to estimate the number of reads that would map against genomes of each clade detected as present and for which an estimation of its relative abundance is provided by the default output. See Supplementary Note 3 for details.  Integration of MetaPhlAn with post-processing and visualization tools. The MetaPhlAn2 package now includes a set of post-processing and visualization tools (“utils” subfolder of the MetaPhlAn2 repository). Multiple MetaPhlAn profiles can in fact be merged in an abundance table (“merge_metaphlan_tables.py”), exported as BIOM files, visualized as heatmap (“metaphlan_hclust_heatmap.py” or the integrated “hclust2” package), GraPhlAn plots (“export2graphlan.py” and the GraPhlAn package1), Krona2 plots (“metaphlan2krona.py”), and single microbe barplot across samples and conditions (“plot_bug.py”).

Relating the metatranscriptome and metagenome of the human gut.

Significance Recent years have seen incredible growth in both the scale and specificity of projects analyzing the microbial organisms living in and on the human body (the human microbiome). Such studies typically require subjects to report to clinics for sample collection, a complicated practice that is impractical for large studies. To address these issues, we developed a protocol that allows subjects to collect microbiome samples at home and ship them to laboratories for multiple different types of molecular analysis. Measurements of microbial species, gene, and gene transcript composition within self-collected samples were consistent across sampling methods. In addition, our subsequent analysis of these samples revealed interesting similarities and differences between the measured functional potential and functional activity of the human microbiome. Although the composition of the human microbiome is now well-studied, the microbiota’s >8 million genes and their regulation remain largely uncharacterized. This knowledge gap is in part because of the difficulty of acquiring large numbers of samples amenable to functional studies of the microbiota. We conducted what is, to our knowledge, one of the first human microbiome studies in a well-phenotyped prospective cohort incorporating taxonomic, metagenomic, and metatranscriptomic profiling at multiple body sites using self-collected samples. Stool and saliva were provided by eight healthy subjects, with the former preserved by three different methods (freezing, ethanol, and RNAlater) to validate self-collection. Within-subject microbial species, gene, and transcript abundances were highly concordant across sampling methods, with only a small fraction of transcripts (<5%) displaying between-method variation. Next, we investigated relationships between the oral and gut microbial communities, identifying a subset of abundant oral microbes that routinely survive transit to the gut, but with minimal transcriptional activity there. Finally, systematic comparison of the gut metagenome and metatranscriptome revealed that a substantial fraction (41%) of microbial transcripts were not differentially regulated relative to their genomic abundances. Of the remainder, consistently underexpressed pathways included sporulation and amino acid biosynthesis, whereas up-regulated pathways included ribosome biogenesis and methanogenesis. Across subjects, metatranscriptional profiles were significantly more individualized than DNA-level functional profiles, but less variable than microbial composition, indicative of subject-specific whole-community regulation. The results thus detail relationships between community genomic potential and gene expression in the gut, and establish the feasibility of metatranscriptomic investigations in subject-collected and shipped samples.

Natural history of the infant gut microbiome and impact of antibiotic treatment on bacterial strain diversity and stability

A longitudinal strain-level analysis of the infant gut microbiome after repeated antibiotic treatments reveals decreased diversity and stability, as well as transient increases in antibiotic resistance genes. Elucidating the effects of drugs on bugs Despite widespread use of antibiotics in children, the effects of antibiotic exposure on the developing infant gut microbiome have remained underexplored. Here, Yassour et al. present a longitudinal study capturing how the gut microbiome responds to and recovers from antibiotic perturbations. Antibiotic-treated children had less stable and less diverse bacterial communities. Antibiotic resistance genes within the guts of these children peaked after antibiotic treatment but generally returned rapidly to baseline. Delivery mode (vaginal versus cesarean) also had strong long-term effects on microbial diversity. These data give insights into the consequences of early life factors such as birth mode and antibiotic treatment on the infant gut microbiome. The gut microbial community is dynamic during the first 3 years of life, before stabilizing to an adult-like state. However, little is known about the impact of environmental factors on the developing human gut microbiome. We report a longitudinal study of the gut microbiome based on DNA sequence analysis of monthly stool samples and clinical information from 39 children, about half of whom received multiple courses of antibiotics during the first 3 years of life. Whereas the gut microbiome of most children born by vaginal delivery was dominated by Bacteroides species, the four children born by cesarean section and about 20% of vaginally born children lacked Bacteroides in the first 6 to 18 months of life. Longitudinal sampling, coupled with whole-genome shotgun sequencing, allowed detection of strain-level variation as well as the abundance of antibiotic resistance genes. The microbiota of antibiotic-treated children was less diverse in terms of both bacterial species and strains, with some species often dominated by single strains. In addition, we observed short-term composition changes between consecutive samples from children treated with antibiotics. Antibiotic resistance genes carried on microbial chromosomes showed a peak in abundance after antibiotic treatment followed by a sharp decline, whereas some genes carried on mobile elements persisted longer after antibiotic therapy ended. Our results highlight the value of high-density longitudinal sampling studies with high-resolution strain profiling for studying the establishment and response to perturbation of the infant gut microbiome.

Identifying personal microbiomes using metagenomic codes

Significance Recent surveys of the microbial communities living on and in the human body—the human microbiome—have revealed strong variation in community membership between individuals. Some of this variation is stable over time, leading to speculation that individuals might possess unique microbial “fingerprints” that distinguish them from the population. We rigorously evaluated this idea by combining concepts from microbial ecology and computer science. Our results demonstrated that individuals could be uniquely identified among populations of 100s based on their microbiomes alone. In the case of the gut microbiome, >80% of individuals could still be uniquely identified up to a year later—a result that raises potential privacy concerns for subjects enrolled in human microbiome research projects. Community composition within the human microbiome varies across individuals, but it remains unknown if this variation is sufficient to uniquely identify individuals within large populations or stable enough to identify them over time. We investigated this by developing a hitting set-based coding algorithm and applying it to the Human Microbiome Project population. Our approach defined body site-specific metagenomic codes: sets of microbial taxa or genes prioritized to uniquely and stably identify individuals. Codes capturing strain variation in clade-specific marker genes were able to distinguish among 100s of individuals at an initial sampling time point. In comparisons with follow-up samples collected 30–300 d later, ∼30% of individuals could still be uniquely pinpointed using metagenomic codes from a typical body site; coincidental (false positive) matches were rare. Codes based on the gut microbiome were exceptionally stable and pinpointed >80% of individuals. The failure of a code to match its owner at a later time point was largely explained by the loss of specific microbial strains (at current limits of detection) and was only weakly associated with the length of the sampling interval. In addition to highlighting patterns of temporal variation in the ecology of the human microbiome, this work demonstrates the feasibility of microbiome-based identifiability—a result with important ethical implications for microbiome study design. The datasets and code used in this work are available for download from huttenhower.sph.harvard.edu/idability.

Linking the Human Gut Microbiome to Inflammatory Cytokine Production Capacity

Gut microbial dysbioses are linked to aberrant immune responses, which are often accompanied by abnormal production of inflammatory cytokines. As part of the Human Functional Genomics Project (HFGP), we investigate how differences in composition and function of gut microbial communities may contribute to inter-individual variation in cytokine responses to microbial stimulations in healthy humans. We observe microbiome-cytokine interaction patterns that are stimulus specific, cytokine specific, and cytokine and stimulus specific. Validation of two predicted host-microbial interactions reveal that TNFα and IFNγ production are associated with specific microbial metabolic pathways: palmitoleic acid metabolism and tryptophan degradation to tryptophol. Besides providing a resource of predicted microbially derived mediators that influence immune phenotypes in response to common microorganisms, these data can help to define principles for understanding disease susceptibility. The three HFGP studies presented in this issue lay the groundwork for further studies aimed at understanding the interplay between microbial, genetic, and environmental factors in the regulation of the immune response in humans. PAPERCLIP.

Structural principles within the human-virus protein-protein interaction network

General properties of the antagonistic biomolecular interactions between viruses and their hosts (exogenous interactions) remain poorly understood, and may differ significantly from known principles governing the cooperative interactions within the host (endogenous interactions). Systems biology approaches have been applied to study the combined interaction networks of virus and human proteins, but such efforts have so far revealed only low-resolution patterns of host-virus interaction. Here, we layer curated and predicted 3D structural models of human-virus and human-human protein complexes on top of traditional interaction networks to reconstruct the human-virus structural interaction network. This approach reveals atomic resolution, mechanistic patterns of host-virus interaction, and facilitates systematic comparison with the host’s endogenous interactions. We find that exogenous interfaces tend to overlap with and mimic endogenous interfaces, thereby competing with endogenous binding partners. The endogenous interfaces mimicked by viral proteins tend to participate in multiple endogenous interactions which are transient and regulatory in nature. While interface overlap in the endogenous network results largely from gene duplication followed by divergent evolution, viral proteins frequently achieve interface mimicry without any sequence or structural similarity to an endogenous binding partner. Finally, while endogenous interfaces tend to evolve more slowly than the rest of the protein surface, exogenous interfaces—including many sites of endogenous-exogenous overlap—tend to evolve faster, consistent with an evolutionary “arms race” between host and pathogen. These significant biophysical, functional, and evolutionary differences between host-pathogen and within-host protein-protein interactions highlight the distinct consequences of antagonism versus cooperation in biological networks.

Strains, functions and dynamics in the expanded Human Microbiome Project

The characterization of baseline microbial and functional diversity in the human microbiome has enabled studies of microbiome-related disease, diversity, biogeography, and molecular function. The National Institutes of Health Human Microbiome Project has provided one of the broadest such characterizations so far. Here we introduce a second wave of data from the study, comprising 1,631 new metagenomes (2,355 total) targeting diverse body sites with multiple time points in 265 individuals. We applied updated profiling and assembly methods to provide new characterizations of microbiome personalization. Strain identification revealed subspecies clades specific to body sites; it also quantified species with phylogenetic diversity under-represented in isolate genomes. Body-wide functional profiling classified pathways into universal, human-enriched, and body site-enriched subsets. Finally, temporal analysis decomposed microbial variation into rapidly variable, moderately variable, and stable subsets. This study furthers our knowledge of baseline human microbial diversity and enables an understanding of personalized microbiome function and dynamics.

Biogeography of the Intestinal Mucosal and Lumenal Microbiome in the Rhesus Macaque

The gut microbiome is widely studied by fecal sampling, but the extent to which stool reflects the commensal composition at intestinal sites is poorly understood. We investigated this relationship in rhesus macaques by 16S sequencing feces and paired lumenal and mucosal samples from ten sites distal to the jejunum. Stool composition correlated highly with the colonic lumen and mucosa and moderately with the distal small intestine. The mucosal microbiota varied most based on location and was enriched in oxygen-tolerant taxa (e.g., Helicobacter and Treponema), while the lumenal microbiota showed inter-individual variation and obligate anaerobe enrichment (e.g., Firmicutes). This mucosal and lumenal community variability corresponded to functional differences, such as nutrient availability. Additionally, Helicobacter, Faecalibacterium, and Lactobacillus levels in stool were highly predictive of their abundance at most other gut sites. These results quantify the composition and biogeographic relationships between gut microbial communities in macaques and support fecal sampling for translational studies.

Integrated Assessment of Genomic Correlates of Protein Evolutionary Rate

Rates of evolution differ widely among proteins, but the causes and consequences of such differences remain under debate. With the advent of high-throughput functional genomics, it is now possible to rigorously assess the genomic correlates of protein evolutionary rate. However, dissecting the correlations among evolutionary rate and these genomic features remains a major challenge. Here, we use an integrated probabilistic modeling approach to study genomic correlates of protein evolutionary rate in Saccharomyces cerevisiae. We measure and rank degrees of association between (i) an approximate measure of protein evolutionary rate with high genome coverage, and (ii) a diverse list of protein properties (sequence, structural, functional, network, and phenotypic). We observe, among many statistically significant correlations, that slowly evolving proteins tend to be regulated by more transcription factors, deficient in predicted structural disorder, involved in characteristic biological functions (such as translation), biased in amino acid composition, and are generally more abundant, more essential, and enriched for interaction partners. Many of these results are in agreement with recent studies. In addition, we assess information contribution of different subsets of these protein properties in the task of predicting slowly evolving proteins. We employ a logistic regression model on binned data that is able to account for intercorrelation, non-linearity, and heterogeneity within features. Our model considers features both individually and in natural ensembles (“meta-features”) in order to assess joint information contribution and degree of contribution independence. Meta-features based on protein abundance and amino acid composition make strong, partially independent contributions to the task of predicting slowly evolving proteins; other meta-features make additional minor contributions. The combination of all meta-features yields predictions comparable to those based on paired species comparisons, and approaching the predictive limit of optimal lineage-insensitive features. Our integrated assessment framework can be readily extended to other correlational analyses at the genome scale.