Poster Session 2023
Poster Session 2023
| Presenter Name | Poster Title |
| Amanda Adams | Bacteria-virus interactions in the vaginal microbiome reduce herpes virus infectivityAmanda Adams, Smita Gopinath |
| Amrisha Bhosle | Response of the gut microbiome and metabolome to dietary fiber in healthy dogsAmrisha Bhosle, Matthew I. Jackson, Aaron Walsh, Eric A. Franzosa, Curtis Huttenhower, Dayakar V. Badri |
| Tobyn Branck | A comprehensive profile of the companion animal gut microbiome integrating reference-based and reference-free methods Tobyn Branck, Zhiji Hu, William A. Nickols, Aaron M. Walsh, Amrisha Bhosle, Meghan I. Short, Jacob Nearing, Artemis Louyakis, Dayakar V. Badri, Christoph Brockel, Kelsey N. Thompson, Curtis Huttenhower |
| Ivan Duran | Using machine learning and multi-omics longitudinal microbiome data for the early prediction of disease developmentIvan Duran, Arnav Srivastava, Maureen M. Leonard, Alessio Fasano, Ali R. Zomorrodi |
| Haley Gause | Cross-kingdom Interactions between Candida albicans and Enterococcus faecalis in the Gut MicrobiomeHaley Gause, UCSF, Johnson Lab |
| Sarah Gayer | Elements of the lean gut environment protect against severe atopic dermatitis in high fat diet-fed miceSarah Gayer, Vaibhav Upadhyay, Owen Jiang, ARum Yoo, Chihiro Tabuchi, Margaret Alexander, Christine Olson, Kyle Spitler, Connie Ha, Moriah Sandy, Laura Dwyer, Jessie Turnbaugh, Tiffany Scharschmidt, Peter Turnbaugh, Sagar Bapat |
| Andrew Ghazi | Quantifying Microbial Strain-Host Associations with ANPAN Andrew R. Ghazi, Yan Yan, Kelsey N. Thompson, Zhendong Mei, Amrisha Bhosle, Fenglei Wang, Kai Wang, Eric A. Franzosa, Curtis Huttenhower |
| Isabella Goodchild-Michelman | Investigating the Effects of Probiotic Treatment on the Gut Microbiome in Preterm InfantsIsabella (Izzy) Goodchild-Michelman , Shirin Moossavi, Marie-Claire Arrieta, Emily Mercer, Ali Zomorrodi |
| Amanda Graboski | Discovery and Characterization of a Gut Microbial Tryptophanase Inhibitor to Ameliorate Chronic Kidney Disease Amanda L. Graboski1, Mark E. Kowalewski2, Joshua B. Simpson2, Matthew R. Redinbo2 |
| Yordan Hodzhev | What does the ratio between blood bacterial and fungal microbiome abundance tell us about fungal-bacterial interactions? Yordan Hodzhev, Borislava Tsafarova, Vladimir Tolchkov, Reni Kalfin, Stefan Panaiotov |
| Katie Hsia | DIFFERENTIAL ABUNDANCE OF THE FUNGAL MICROBIOME IN PATIENTS WITH ULCERATIVE COLITIS Katie Hsia, Laleh Montaser Kouhsari, Khalid Algarrahi, May Fu, Naisi Zhao, Hannah Chen, Dominique Michaud, Sushrut Jangi |
| Jordan Jensen | Integrating reference- and assembly-based methods for improved viral identification from metagenomes, metatranscriptomes, and viromes Jordan Jensen, Ya Wang, Moreno Zolfo, Philipp C. Münch, Nicola Segata, Eric A. Franzosa, Curtis Huttenhower |
| Shanlin Ke | Dissecting the Role of the Human Microbiome in COVID-19 via Metagenome-assembled Genomes Shanlin Ke, Scott T. Weiss, Yang-Yu Liu |
| Gavin Kuziel | Functional diversification of plant small molecules by the gut microbiomeGavin A Kuziel, Gabriel L Lozano, John Manion, Corina Simian, Emmanuel Stephen-Victor, Talal Chatila, Min Dong, Jing-Ke Weng, Seth Rakoff-Nahoum |
| Abigail Lind | Genetic diversity of commensal Blastocystis gut protists reveals strain-specific changes in host-interfacing pathways Abigail L. Lind, Ami S. Bhatt, Katherine S. Pollard |
| Steven Medina | The Harvard Chan Microbiome Collection Core Steven Medina, Curtis Huttenhower |
| Mahsa Monshizadeh | Incorporating metabolic activity, taxonomy, and community structure to improve microbiome-based predictive models for host phenotype prediction Mahsa Monshizadeh, Yuzhen Ye |
| Xochitl Morgan | The Harvard Chan Microbiome Analysis Core Xochitl C. Morgan, Lauren J. McIver, Thomas Kuntz, Curtis Huttenhower |
| Jacob Nearing | Gut microbiome-metabolome interactions during ketogenic diets of varied composition Jacob T. Nearing, Kelsey N Thompson, Amrisha Bhosle, Veronica Perdomo, William A. Nickols, Tobyn Branck, Christoph Brockel, Dayakar Badri, Curtis Huttenhower, Matthew Jackson |
| Etienne Nzabarushimana | The landscape of novel lateral gene transfer events in the human microbiome Etienne Nzabarushimana, Tiffany Y. Hsu, Dennis Wong, Chengwei Luo, Robert G. Beiko, Morgan Langille, Curtis Huttenhower, Long H. Nguyen, Eric A. Franzosa |
 Diana Proctor | Candida auris and the great ESKAPE: the skin as a reservoir for multi-drug resistance and transmission Diana M Proctor, Thomas Karl Atkins, Sarah E Sansom, Mary K Hayden, Julia A Segre |
| Chatpol (Jamie) Samuthpongtorn | THE ROLE OF THE GUT MICROBIOME IN THE ASSOCIATION BETWEEN CITRUS FRUIT AND RISK OF DEPRESSION Chatpol Samuthpongtorn, Allison Chan, Wenjie Ma, Fenglei Wang, Long H. Nguyen, Dong D. Wang, Olivia I. Okereke, Curtis Huttenhower, Andrew T. Chan, Raaj S. Mehta |
| Prioty Sarwar | HMO-metabolizing bacteria in the gut microbiome of infants with atopic dermatitis/eczemaPrioty Sarwar, Deniz Uzun, Vanja Klepac-Ceraj |
| Zeyang Shen | A genome catalog of the early-life human skin microbiome Zeyang Shen, Pamela A. Frischmeyer-Guerrerio, VITALITY team, Julia A. Segre |
| Kelsey Thompson | Identifying strain-specific associations in colorectal cancer Kelsey N. Thompson, Andrew Ghazi, Gianmarco Piccinno, Yan Yan, Andrew M. Thomas, Long H. Nguyen, Lior Lobel, Paolo Manghi, Lauren J. Mciver, Emma Accorsi, Eric A. Franzosa, Francesco Asnicar, Andrew T. Chan, Wendy S. Garrett, Nicola Segata, Curtis Huttenhower |
| A.Delphine Tripp | C. acnes phage predation in the healthy human skin microbiome A. Delphine Tripp, Tami D. Lieberman |
| Emily Van Syoc | Mining metagenomes reveals gut mycobiome alterations with metformin and type 2 diabetes mellitusEmily Van Syoc, Michelle Nixon, Justin D. Silverman, Connie J. Rogers, Frank Gonzalez, Luo Yuhong, Ilze Elbere, Janis Klovins, Andrew Patterson, and Erika Ganda |
| Ya Lea Wang | Scalable virome enrichment methods for community detection and quantification Ya Wang, Jordan Jensen, Kelsey N. Thompson, Eric Franzosa, Seth Rakoff-Nahoum, Erica M. Hartmann, Curtis Huttenhower |
| An-Ni Zhang | CRISPR spacer acquisition is a rare event in human gut microbiome Anni Zhang, Department of Biological Engineering, MIT |
| Yancong Zhang | Predicting functions of uncharacterized gene products in microbial communities Yancong Zhang, Amrisha Bhosle, Sena Bae, Kelly Eckenrode, Xueying (Sonia) Huang, Jingjing Tang, Danylo Lavrentovich, Lana Awad, Ji Hua, Xochitl C. Morgan, Andy Krueger, Wendy S. Garrett1, Eric A. Franzosa, Curtis Huttenhower |
Thanks to our sponsors:
Bacteria-virus interactions in the vaginal microbiome reduce herpes virus infectivity
Presented by: Amanda Adams
The vaginal microbiome is an important determinant of host health and the first barrier encountered by sexually transmitted pathogens during infection. Among the vaginal microbiome, Lactobacilli are associated with reduced susceptibility to viral infection, but the mechanisms by which various Lactobacilli strains reduce viral infectivity remain poorly understood. Using a collection of human vaginal microbial strains, we show that the prominent vaginal strain, Lactobacillus crispatus reduces infectivity of sexually transmitted pathogen Herpes Simplex Virus (HSV). Reduction of HSV infectivity is species specific, with L. crispatus reducing infection and disease better than gut-associated L. reuteri. Active cell metabolism is not required as UV-killed L. crispatus retain the ability to reduce herpes infection. Since one of the most abundant structures on the outside of the L. crispatus cell is peptidoglycan, we assessed whether peptidoglycan could reduce HSV infection. We found that commercially available purified peptidoglycan from multiple bacterial sources reduced herpes infection in vitro and in vivo in a mouse model of genital herpes infection. Mice were susceptible to reinfection, indicating that immunological memory is not activated. Cleavage of the glycosidic linkages in the peptidoglycan chain with lysozyme restored virus infectivity in vitro and in vivo suggesting that antiviral effects are dependent on longer peptidoglycan chains. Current studies aim to determine how Lactobacilli peptidoglycan contributes to a reduction in HSV infectivity focusing on HSV entry receptors and what species-specific peptidoglycan modifications allow L. crispatus to reduce infectivity better than other Lactobacilli. Such results provide a greater understanding of the ways that the vaginal microbiome serves as a physical barrier to infection and why some vaginal communities promote better antiviral protection than others.
Response of the gut microbiome and metabolome to dietary fiber in healthy dogs
Presented by: Amrisha Bhosle
Dietary fiber has been shown to expand beneficial microbial activity in the gut and is being explored to maximize health-related benefits from the microbiome in both humans and companion animals. However, a clear understanding of the precise interactions among fiber-containing diets, specific fibers, and microbiome composition and function is lacking. To fill this knowledge gap, we analyzed fecal metagenomic and metabolomic profiles from 18 healthy dogs at 13 time points each (totaling 226 samples) that were fed 12 test foods containing different fiber sources and quantities (5-13% fiber on an as-fed basis) belonging to three food groups (high starch/low fiber, medium starch/medium fiber, and low starch/high fiber). Taxonomic and functional profiling identified taxa and functions whose abundances were associated either with overall fiber intake or specific fiber compositions. Short-chain fatty acids (SCFAs), acylglycerols, and intermediate products of fiber degradation including sugars (arabinose, xylulose, fucose) and intercalating polyphenols (genistein, hesperidin, limonin, naringenin, and secoisolariciresinol) were enriched in response to intake of insoluble fiber. Some of these were also associated with affected microbes in particular food groups. Eleven species were significantly enriched in response to only one food group. Accordingly, enrichment of beneficial metabolites such as SCFAs was more pronounced in response to these fiber sources, highlighting the importance of interactions between specific dietary components and individual microbes. That is, the production of beneficial metabolites was dependent on both the presence of the appropriate pre-existing microbiota, and the introduction of an appropriate dietary substrate. Correspondingly, the response of microbes to dietary macronutrients varied across subjects, as did the metabolomic consequences of the foods and their overall influence on the microbiome. These data are useful for directing population-wide dietary modifications and for personalized health targeting, with implications for other animals and humans.
A comprehensive profile of the companion animal gut microbiome integrating reference-based and reference-free methods
Presented by: Tobyn Branck
The gut microbiome of companion animals (cats and dogs) is relatively underexplored, despite its relevance to animal health, pet owner health, and basic microbial community biology. Here, we present a comprehensive analysis of cat and dog gut microbiomes, incorporating 2,423 stool shotgun metagenomes (2,056 dog and 367 cat) spanning 12 publicly available datasets (n=513) and 7 populations newly presented in this study (n=1,910). These are compared with a baseline human population of 238 gut metagenomes from the Human Microbiome Project 1-II processed in an identical manner. All microbiomes were characterized using both reference-based taxonomic and functional profiling, as well as de novo assembly and metagenomic assembled genomes (MAG) clustered into species genome bins (SGBs). The latter identified 563 SGBs from the companion animal microbiomes, 333 novel (“unknown” SGBs or uSGBs) without confident taxonomy and 230 “known” (kSGBs) that include at least one genome with prior taxonomic assignment. Companion animal and human SGBs spanned 14 phyla, with several SGBs unique or shared across host species. For instance, 47 SGBs including Bifidobacterium spp., Lactobacillus spp., Fusobacteriaceae, and novel SGBs spanning Firmicutes, Bacteroidetes, and Actinobacteria were unique to dogs. In cats, an even larger number of unique SGBs were identified (n=65) that spanned Prevotella copri clade C, Olsenella spp., Clostridia, and novel SGBs in the Veillonellaceae, Phyllobacteriaceae, Anaplasmataceae, and Ruminococcaceae families, among others. Companion animals shared several SGBs not found in humans (n=55) that included many Blautia spp., Phocaeicola spp., Sutterella wadsworthensis, and a relatively higher proportion of novel SGBs in Butyricicoccus, Mediterraneibacter, and Fusobacteriaceae. Lastly, we identified a subset of 17 SGBs found in all three hosts, including Ruminococcus gnavus, Prevotella copri clade A, Flavonifractor plautii, Lachnospiraceae, Blautia spp., and Phocaeicola vulgatus. We identified patterns of phylogenetic clustering of genomes within SGBs (i.e., strains that are shared universally vs. host-specific lineages within SGBs). For example, Prevotella copri clade A strains recovered from cat and dog gut metagenomes clustered distinctly from those found in human hosts. These phylogenetic patterns suggest genetic divergence within host species, likely with functional implications (e.g., differences in pangenome and mobile element distributions) important to companion animal health. This study provides the largest one-health microbiome resource to date of companion animal gut metagenomes, greatly improves the ability to profile additional animal gut microbiomes in the future and contributes to our understanding of how microbes are transmitted between companion animals and humans in the contexts of infectious diseases, immune modulation, and specific genetic elements such as antimicrobial resistance genes.
Using machine learning and multi-omics longitudinal microbiome data for the early prediction of disease development
Presented by: Ivan Duran
The gut microbiome is intrinsically dynamic and studies that collect longitudinal microbiome data to assess the dynamics of the gut microbiota during disease development or progression, or after a therapeutic intervention are increasing in frequency. However, efficient computational tools to harness multi-omics longitudinal microbiome data to predict clinical outcomes are underdeveloped. In this project, we aim to develop new machine learning (ML) tools to predict clinical outcomes by making use of time-series microbiome multi-omics data. As a case study, we used longitudinal metagenomic and metabolomic data from a prospective, longitudinal birth cohort study of children at high risk of Celiac Disease (CD) and sought to predict CD development in these subjects using pre-onset data. To this end, we trained Random Forest classifiers combined with an efficient feature selection scheme using several pieces of clinical
metadata along with species, strains, pathways, and metabolites abundance data before disease onset as features (predictors). Our analyses revealed that clinical metadata alone are not accurate predictors of disease development (F1-score = 68.67%, 10-fold C.V.). However, we were able to achieve a high prediction performance of 93% (F1-score, 10-fold C.V.) using the abundance of only one pathway at 9 months of age and 100% (F1-score, 10-fold C.V.) using the abundance of only seven microbial strains at 15 months of age. This pilot study demonstrates the utility of ML for inferring key temporal microbiome signatures that are highly predictive of host clinical status. It also lays the foundation for building early predictive tools that would enable physicians to plan for preventive strategies before the clinical manifestation of disease.
Cross-kingdom Interactions between Candida albicans and Enterococcus faecalis in the Gut Microbiome
Presented by: Haley Gause
Candida albicans and Enterococcus faecalis are both common members of the human gut microbiome. Previous studies have suggested that this yeast and bacteria species interact as commensal members in the gut, however the mechanisms behind these interactions have yet to be elucidated. By measuring gene expression of these species in co-culture, we can start to tease apart the genetic underpinnings of this interaction. Using Dual RNA-seq, we characterized the transcriptional profiles of both C. albicans and E. faecalis during growth together and separately in two conditions – 1) gut-like in-vitro culture and 2) Germ-free (GF) mouse gut. The transcriptional response of C. albicans to E. faecalis is very similar across in vitro and GF gut conditions, with over 300 genes up-regulated > 4-fold in either condition. Up-regulated genes include a group of seven transcription factors which are induced exclusively in the presence of E. faecalis. Many of these transcription factors are currently uncharacterized, indicating a possible function in mediating interkingdom interactions with bacteria in biologically relevant environments, such as the gut. Future work is ongoing to determine the transcriptional networks controlled by these upregulated transcription factors and the role of those networks in the interactions between Candida albicans and Enterococcus faecalis.
Elements of the lean gut environment protect against severe atopic dermatitis in high fat diet-fed mice
Presented by: Sarah Gayer
Diet controls core physiological processes, yet how it regulates immune responses in the context of inflammatory disease is not well understood. Previous work from our lab demonstrated that high fat diet (HFD)-fed mice exhibit a non-canonical T helper type 17 (TH17) dominated response to atopic dermatitis (AD), which classically elicits a TH2 response in control lean fat diet (LFD)-fed mice, rendering HFD-fed mice incapable of benefitting from targeted anti-TH2 therapies for AD (Bapat et al., Nature 2022). Diet robustly remodels the gut microbiome, which in turn has been shown to have long-range effects on tissue physiology and inflammation. We hypothesized that diet-induced alterations in the microbiome were important drivers of the TH2 to TH17 inflammatory switch in HFD-fed mice. To test this hypothesis, we conducted bed-swapping experiments, in which we exposed HFD-fed mice to bedding and fecal pellets from LFD-fed mice and vice-versa. Interestingly, we found that HFD-fed mice exposed to LFD-fed bedding presented with a TH2 dominant response to AD challenge and less severe disease score relative to the HFD-fed control. Additionally, we conducted untargeted serum metabolics and 16S metagenomic sequencing of these mice, data which we will describe here. Surprisingly, depletion of the bacterial microbiome via antibiotic administration had no effect on AD severity of LFD-fed vs HFD-fed mice relative to water only controls. TH17 cells are important for anti-fungal immunity, and so further work will be aimed at understanding the potential role of the fungal microbiome on T cell response and disease severity.
Quantifying Microbial Strain-Host Associations with ANPAN
Presented by: Andrew Ghazi
Strain variation can strongly influence the impact of microbes on their environments, however inferential methods for quantifying these important differences have been lacking. Metagenomic data with strain-level resolution has several features that make traditional statistical methods challenging to use, including high dimensionality, individual-specific strain carriage, and complex phylogenetic relatedness. We present ANPAN, an R package that consolidates methods for strain statistics in three key components. First, adaptive filtering methods specifically designed to interrogate microbial strain profiles are combined with linear models to identify strain-specific genetic elements associated with host health outcomes. Second, phylogenetic generalized linear mixed models are used to characterize the effect of strain-level community structure. Finally, random effects models are used to account for species abundance when assessing the impact of gene pathway abundance on outcome variables. We validated our methods by simulation, showing that we achieve more accurate effect size estimation and a lower false positive rate compared to current methodologies. We then applied our methods to a dataset of 1,262 colorectal cancer patients, identifying functionally adaptive genes and strong phylogenetic effects associated with CRC status. The open source ANPAN repository with help documentation and a tutorial vignette are available at https://github.com/biobakery/anpan.
Investigating the Effects of Probiotic Treatment on the Gut Microbiome in Preterm Infants
Presented by: Isabella Goodchild-Michelman
Infants born prematurely have an abnormal set of birth conditions that lead to a sparse,low-diversity population of microbes initially colonizing their guts. Several studies have shown the promise of probiotic treatments to shift the preterm infants’ microbiomes to resemble those of a healthy term infant; however, the functional mechanisms that underlie probiotic’s therapeutic effects remain unknown. The goal of this study is to better understand the role of probiotics in the maturation of the preterm gut microbiomes by using metagenomic data from a longitudinal study of preterm infant gut development where infants sampled from birth to three months. To this end, we constructed metagenome-scale species-resolved computational models of metabolism for each microbiome sample in the study through integrating GEnome-scale Models (GEMs) of metabolism for individual species present in that sample. We further constructed GEMs for all microbial strains that made up the multi-strain probiotic used in the study. We then computationally simulated each microbiota model in the absence and presence of the probiotic strains. Comparing the metabolite production between community models with and without the presence of the probiotic treatment over the sampling time allowed us to identify metabolites that are differentially produced between these two groups and, more importantly, to trace back microbial species that are responsible for their production. Overall, our study is expected to provide unprecedented insight into species and metabolite-level mechanisms of how probiotic treatment accelerates the maturation of the preterm infants’ gut Microbiomes.
Discovery and Characterization of a Gut Microbial Tryptophanase Inhibitor to Ameliorate Chronic Kidney Disease
Presented by: Amanda Graboski
Chronic kidney disease (CKD) afflicts nearly 500 million people worldwide and is one of the fastest growing causes of mortality. A key consequence of a diseased kidney is the serum retention of toxic compounds that have a broad impact on human physiology. One of the most dangerous uremic toxins is indoxyl sulfate (IS), a metabolite produced solely from the breakdown of tryptophan by gut microbial tryptophanases (TPases). High IS levels in preclinical and clinical models have been correlated with altered mitochondrial oxidative phosphorylation, renal fibrosis, and six different phenotypes of cardiovascular disease. Recent studies showed the genetic elimination of TPase in an artificial microbiome of germfree mice prevented the formation of IS and reduced biomarkers of kidney injury, suggesting that inhibition of TPase could prevent or reduce uremic toxicity. In this study, we define the structural and functional landscape of gut microbial TPases to gauge its drugability and to guide novel inhibitor design. First, we selected a group of diverse TPases from the 184 sequences present in the Integrated Genome Catalog of human fecal metagenomes and examined their activity against tryptophan using substrate-turnover assays. We also elucidated the crystal structures of these TPases, revealing highly conserved tertiary structure and active site architecture. Using this structural information and the well-characterized catalytic mechanism, we designed and synthesized a panel of structure-based and mechanism-inspired inhibitors. A handful of these novel tryptophan-like compounds show inhibitory activity, one displaying a 20-30-fold greater affinity relative to the natural substrate. UV-vis spectral readings suggest activity as a transition state analog, and a cocrystal structure reveals many aromatic contacts with conserved active site residues. Minimal toxicity against mammalian and microbial cells was observed when dosed up to 500mM and target engagement in microbial cells and murine models is currently being evaluated. The discovery of a potent TPase inhibitor will help to unravel the molecular basis of increased serum IS levels and may serve as a therapeutic avenue for mitigating IS-induced toxicity in CKD.
What does the ratio between blood bacterial and fungal microbiome abundance tell us about fungal-bacterial interactions?
Presented by: Yordan Hodzhev
Ecological studies demonstrated a strong interaction between bacterial and fungal microbiomes (fungal-bacterial interaction) in various habitats (Wagg et al., 2019, Krüger et al. 2019). The consequences of bacterial-fungal interactions for the human host are largely unknown. Recently, we characterized the composition of whole-blood bacterial and fungal microbiomes in healthy individuals (Panaiotov et al., 2021). Using metagenomic sequence analysis we were able to identify a total of 24 bacterial orders (40 families and 50 genera) and 44 fungal orders (75 families and 94 genera). The aim of the present meta-analysis was to explore possible interactions between microbial and fungal communities. Blood group and gender data were included to assess the findings’ biological relevance.
Three ml of venous whole blood was collected from 28 subjects (14 females, 7 of each blood group – A, B, AB, O). Blood was lysed in d. water and the human DNA was treated with DNase. Microbial DNA was isolated by applying treatment with 4% SDS for microbial lysis. Isolated DNA was divided into two subsamples and 16S and ITS metagenomic analysis was applied for each subject. Microbial total and relative abundance were calculated. Then the bacterial vs. fungal (B/F) reads ratio was analyzed. Data were subjected to nonparametric statistical evaluation (Kruskal-Wallis) of gender and blood group effects.
The major findings were: (1) The overall fungal sequence number (median=8579) was higher than the bacterial (median = 1062; Related-Samples Wilcoxon Signed Rank Test, Z =383; P<0.001). (2) Individually, the B/F ratio varied significantly spanning from full fungal dominance to an almost complete lack of fungal sequences. (3) The mean B/F ratio was higher for males (mean B/F=0.95) as compared to females (mean BF = 0.18; P<0.001). (4) The blood group had an impact on the B/F ratio. For individuals of blood groups, A and B the ratio were around 1 and 0.2 (P<0.05) for individuals of blood groups AB and O.
In conclusion, despite the overall fungal dominance the B/F ratio showed high individual variability ranging from almost full fugal dominance to negligible fungal presence. The dependence of the B/F by gender and blood group suggests that it reflects the physiological status of the host. It could be hypothesized that B/F could serve as a health diagnostic index. It is worth testing the therapeutic correction of B/F in clinical practice.
DIFFERENTIAL ABUNDANCE OF THE FUNGAL MICROBIOME IN PATIENTS WITH ULCERATIVE COLITIS
Presented by: Katie Hsia
Background
The fungal microbiome has been increasingly implicated in the pathogenesis of ulcerative colitis (UC). Circulating antibodies to Saccharomyces antibodies are detected in 10-15% of patients with UC, and patients with high fecal Candida are more likely to demonstrate a favorable response to fecal microbial transplantation (FMT) in inducing remission. However, the fungal microbiome remains poorly characterized in UC, especially across the spectrum of endo-histologic activity and following exposure to immunosuppressive drugs. Furthermore, unidentified fungal sequences are commonly recovered from UC cohort studies.
Aims
In this study, we aim to characterize the fungal microbiome in UC patients with varying levels of endoscopic activity, endo-histologic activity, and treatment exposure as well as additionally identify unknown fungal sequences.
Methods
We performed a secondary analysis using the data extracted from the Crohn’s and Colitis Foundation’s Study of a Prospective Adult Research Cohort with IBD, which contains clinical, endoscopic, histologic, and metagenomic data. Using Internal Transcribed Spacer based deep sequencing of fungal rDNA from fecal samples, we classified sequences utilizing the UNITE fungal database and a fitted classifier. These sequences were also analyzed with the Basic Local Alignment Search Tool (BLAST). Phyloseq and DESeq2 in R studio were used to assess fungal diversity and differential abundance of fungi between comparator groups.
Results
We identified 500 unique fungal amplicon sequence variants across the cohort of 82 patients, belonging to phylum Ascomycota (71.5%), Basidiomycota (11%), Mucoromycota (0.16%), or unidentified (17.2%). We found no differences in alpha or beta diversity in the fungal microbiome during endoscopic activity vs remission. However, patients with endoscopic activity (n=25) had relative increases in Saccharomyces (log 2-fold change 4.54, p-adj< 5 x 10-5) and Candida (2.56, p-adj< 0.03), compared to during endoscopic remission. Similarly, patients with endo-histologic activity (n=19) were also enriched for Saccharomyces (2.78, p-adj<0.08), compared to endo-histologic remission. Exposure to immunosuppressants was not associated with any significant changes in differential abundance about phylum or genera. After adjusting for age, gender, and biologic exposure among patients with endoscopic activity, Saccharomyces (7.76, p-adj < 1 x 10-15) and Candida (7.28, p-adj < 1 x 10-8) remained enriched. Among 270 unclassifiable fungal ASVs, we identified the majority of these to be low-prevalence ASVs (in <5% of samples), with 265 representing dietary contaminants (i.e. seed plants) and 5 representing either a fungus (Candida mesenterica), bacteria (Phocaeicola vulgatus, Eubacterium rectale, or Bacteroides ovatus), or a nematode (Meloidogyne chitwoodi).
Conclusion
These data demonstrate that endoscopic and histologic inflammation in UC is associated with relative expansion of Saccharomyces and Candida compared to remission, even after controlling for exposure to biologic exposure. Unidentified fungal sequences are usually of low prevalence and represent dietary contaminants. The role of fungal ASVs as potential biomarkers and targets for personalized approaches to therapeutics in UC should be evaluated in future studies.
Integrating reference- and assembly-based methods for improved viral identification from metagenomes, metatranscriptomes, and viromes
Presented by: Jordan Jensen
Capturing an accurate representation of the viral members of a microbial community presents significant experimental and computational challenges. Sample preparation approaches for virus-like particle (VLP) enrichment vary greatly in their efficiency among protocols and environments, and sequences from any technology (metagenomic, metatranscriptomic, or VLP enrichments) can be difficult to identify computationally. Limitations include small viral genome size, and subsequently a small proportion of genetic content in samples; lack of universal marker genes; multiple nucleic acid backbone types; rapid evolution, recombination, and sequence divergence; and most prominently, a lack of well-characterized viral reference databases.
To address these limitations, we developed BAQLaVa (Bioinformatic Application for Quantification and Labeling of Viral taxonomy), which integrates both reference- and assembly-based methods to provide viral profiles from shotgun DNA or RNA sequencing (with or without enrichment). Reads are compared with both nucleotide and protein (translated) databases that are pre-screened for viral identification using a modification of the MetaPhlAn algorithm and reconciled with the most recent International Committee on Taxonomy of Viruses (ICTV) taxonomic rankings. In parallel, assembled contigs are classified using deep learning, and viral identifications from all three approaches are harmonized per sample.
We evaluated BAQLaVa with 1) in silico simulated data representing broadly viral material, 2) more detailed synthetic gut microbiomes, and 3) existing human gut metagenomes, metatranscriptomes, and VLP-enriched viromes. Using only nucleotide and protein references, we found that BAQLaVa achieves both greater sensitivity and specificity than existing tools (including MetaPhlAn). We capture 57-87% of both DNA and RNA viral content even in highly novel communities with a PPV of 73-97% and FPR of 2.4-2.5%. This work is ongoing, including optimization of parameter and contig lengths for assembly classification, as well as error rate calculations for viral quantitative abundance profiling vs. qualitative detection. We hope these methods will unlock as-yet-unaccessed information on viral community members from thousands of existing metagenomes and metatranscriptomes, as well as enabling more accurate characterization of future VLPs from a variety of microbial environments.
Dissecting the Role of the Human Microbiome in COVID-19 via Metagenome-assembled Genomes
Presented by: Shanlin Ke
Coronavirus disease 2019 (COVID-19), primarily a respiratory disease caused by infection with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), is often accompanied by gastrointestinal symptoms. However, little is known about the relation between the human microbiome and COVID-19, largely due to the fact that most previous studies fail to provide high taxonomic resolution to identify microbes that likely interact with the SARS-CoV-2 infection. Here we applied de novo assembly and binning strategies to reconstruct metagenome-assembled genomes (MAGs) from the whole-metagenome shotgun sequencing data of 514 nasopharyngeal and fecal samples of patients with COVID-19 and controls in a total of six discovery cohorts (publicly available). We reconstructed a total of 11,584 medium-and high-quality microbial MAGs and obtained 5,403 non-redundant MAGs (nrMAGs) with strain-level resolution. Thanks to the high taxonomic resolution of nrMAGs, we found that a significant reduction of strain richness for many species in the gut microbiome of COVID-19 patients. The gut microbiome signatures can accurately distinguish COVID-19 cases from healthy controls and the generality of COVID-19 microbiome features in machine learning models can be validated across different cohorts. We then demonstrated the ability of nrMAGs to predict the date of negative RT-qPCR result of patients with COVID-19 (progression of COVID-19) and this prediction linked some opportunistic pathogens to the progression of COVID-19, including nrMAGs from Klebsiella quasivariicola, Klebsiella pneumoniae, and Escherichia coli. To further characterize the relation between the human gut microbiome and COVID-19, we applied the GMPT method to move beyond the standard association analysis. Using GMPT, we identified a set of nrMAGs with a putative causal role in the clinical manifestations of COVID-19 and revealed their functional pathways (i.e., pentose phosphate pathway) that potentially interact with SARS-CoV-2 infection. Moreover, we found that the abundance of pentose phosphate pathway (PENTOSE-P-PWY) in COVID-19 patients was significantly higher than that in Non-COVID-19 controls at the community level. Finally, we demonstrated that the main findings of our study can be largely validated in three independent cohorts (314 fecal samples in total, publicly available). The presented results highlight the importance of incorporating the human gut microbiome in our understanding of SARS-CoV-2 infection and disease progression. The genomic content of nrMAGs presented here has the potential to inform microbiome-based therapeutic developments for COVID-19 progression and post-COVID conditions.
Functional diversification of plant small molecules by the gut microbiome
Presented by: Gavin Kuziel
The interaction between diet and the gut microbiome is instrumental in affecting host health and disease. For example, complex dietary carbohydrates actively shape the microbiome as a principal carbon source and concomitantly affect host physiology via microbiome products of carbohydrate fermentation that drive host homeostasis. In contrast, little is known about the interaction between the microbiome and “dietary dark matter”, the additional millions of chemically-diverse plant secondary metabolites, phytochemicals, that we consume daily.
Here, we asked whether chemically-diverse plant secondary metabolites are transformed by gut bacteria, and if products of phytochemical catabolism affect microbiome composition or host physiology. Across a collection of taxonomically-diverse gut bacteria, we identified broad phytochemical glycoside catabolism across diverse species. Glycosides are small molecules (aglycones) conjugated to simple carbohydrates. The Bacteroidales exhibited an enhanced capacity for phytochemical catabolism. Genetic dissection of this catabolism in two prevalent and abundant Bacteroides species, Bacteroides ovatus (Bo) and Bacteroides uniformis (Bu), identified two novel divergent systems for phytochemical catabolism. Whereas Bo harbored a non-specific, generalist system for glycoside and disaccharide catabolism, Bu harbored a multiple-locus glycoside catabolism system in which some loci were selective for glycosides and not chemically-similar disaccharides. Furthermore, we demonstrated that microbial catabolism of glycosides liberates their aglycones, expanding the chemical diversity of these factors and activating new bioactivities unique from their parent glycosides. In a model of colitis, mice were protected from disease when treated with Bu (but not a mutant Bu unable to catabolize glycosides) and the ancient medicinal glycoside salicin. This suggests saligenin, salicin’s aglycone, is the active anti-inflammatory agent. This protection was unique to salicin bioactivation, as mice treated with Bu and arbutin, a salicin analog, were not protected from disease. Our findings highlight new mechanistic insights into microbiome-dependent transformation of dietary phytochemicals and the effects of these metabolic transformations on host homeostasis.
Genetic diversity of commensal Blastocystis gut protists reveals strain-specific changes in host-interfacing pathways
Presented by: Abigail Lind
The human gut microbiome is a microbial ecosystem containing bacteria, archaea, viruses, and microbial eukaryotes. The most common human gut microbial eukaryote is the commensal protist Blastocystis, with an estimated prevalence in industrialized countries at 25%. While the presence of Blastocystis is associated with a significant reduction in pro-inflammatory bacteria and with reduced gut inflammation, little is understood about its role in the gut microbiome and its fundamental biology. Genetic evidence suggests that Blastocystis comprise a group of genetically diverse subtypes, but we lack high quality genomic data for these subtypes and understanding of their functional similarities and differences. Here, we cultivate 6 Blastocystis strains spanning the genetic diversity of the genus and generate contiguous, annotated genomes using a combination of long-read DNA sequencing, Hi-C, and RNA-seq. These genomes range in size from 14-25 Mb and have protein-coding genes with unusual features, including a frequent lack of canonical stop codons and a regular intron length of exactly 30 base pairs. Through comparison with the genomes of closely related stramenopiles, we find a pattern of genome reduction and gene duplication in Blastocystis, as well as genomic organization patterns that likely arose during the transitions from a free-living lifestyle to an obligate within-host lifestyle and in transitions between host species. We find substantial strain and subtype-specific gene duplications, including those of likely host-interfacing genes such as those involved in cell-cell adhesion and cell surface glycan production. Together, these genomes and our analyses reveal the adaptations Blastocystis has undergone to thrive in the gut microbiome. These results identify substantial biological variability between subtypes of Blastocystis which are likely to drive differences in interactions with other gut microbiota and the host.
The Harvard Chan Microbiome Collection Core
Presented by: Steven Medina
The Microbiome Collection Core at the Harvard T.H. Chan School of Public Health (HCMCC) was established in response to a strong demand among the research community for validated microbiome collection kit configurations and easy usability for in-home sampling. Under the umbrella of the Harvard Chan Microbiome in Public Health Center (HCMPH), HCMCC aims to support population-scale microbiome sample collection and expand our understanding of the microbiome to improve population health. The HCMCC has developed a multi carrier-compatible home stool and oral sample collection kit that permits cost-effect multi’omic microbiome studies, leveraging the intellectual and infrastructure foundation laid by the HMP2 (the 2nd phase of the NIH Human Microbiome Project) and the MLSC (Massachusetts Life Sciences Center)-funded MICRO-N (MICRObiome Among Nurses) collection. By providing this customizable microbiome collection kit, we enable researchers to perform multiple different molecular assays and tailor collection plan to study-specific needs.
Incorporating metabolic activity, taxonomy, and community structure to improve microbiome-based predictive models for host phenotype prediction
Presented by: Mahsa Monshizadeh
The human gut microbiome plays a key role in human health and diseases. We developed MicroKPNN, a prior-knowledge guided interpretable neural network for microbiome-based human host phenotype prediction. The prior-knowledge used in MicroKPNN includes the metabolic activities of different bacterial species, phylogenetic relationships, and bacterial community structure. Application of MicroKPNN to seven gut microbiome datasets (involving five different human diseases including inflammatory bowel disease, type 2 diabetes, liver cirrhosis, colorectal cancer, and obesity) shows that incorporation of the prior knowledge helped improve the microbiome-based host phenotype prediction. MicroKPNN outperformed fully-connected neural network based approaches in all seven cases, with the most improvement of accuracy in the prediction of type 2 diabetes. MicroKPNN outperformed a recently developed deep-learning based approach DeepMicro, which selects the best combination of autoencoder and machine learning approach to make predictions, in six out of the seven cases. More importantly, we showed that MicroKPNN provides a way for the interpretation of the predictive models. Our results suggested that the metabolic potential of the bacterial species contributed more than the two other sources of prior knowledge.
MicroKPNN is publicly available at https://github.com/mgtools/MicroKPNN.
The Harvard Chan Microbiome Analysis Core
Presented by: Xochitl Morgan
The Microbiome Analysis Core at the Harvard T.H. Chan School of Public Health was established in response to the rapidly emerging field of microbiome research and its potential to affect studies across the biomedical sciences. The Core’s goal is to aid researchers with microbiome study design and interpretation, reducing the gap between primary data and translatable biology. The Microbiome Analysis Core provides end-to-end support for microbial community and human microbiome research, from experimental design through data generation, bioinformatics, and statistics. This includes general consulting, power calculations, and selection of data generation options, as well as analysis of data from amplicon (16S/18S/ITS), shotgun metagenomic sequencing, metatranscriptomics, metabolomics, and other molecular assays. The Microbiome Analysis Core has extensive experience with microbiome profiles in diverse populations, including taxonomic and functional profiles from large cohorts, qualitative ecology, multi’omics and meta-analysis, and microbial systems and human epidemiological analysis. By integrating microbial community profiles with host clinical and environmental information, we enable researchers to interpret molecular activities of the microbiota and assess its impact on human health.
Gut microbiome-metabolome interactions during ketogenic diets of varied composition
Presented by: Jacob Nearing
High fat and protein (ketogenic) diets that come close to eliminating all dietary carbohydrates force the body to switch from utilizing carbohydrates to fat as its main energy source. This state known as ketosis is characterized by the production of ketone bodies such as 3-hydroxybutyrate (BHB) and has been associated with weight loss, insulin resistance, mitochondria efficiency, and reduced inflammation. For example, recent evidence has suggested that alterations in the gut microbiome by ketogenic diets may play a role in improving colitis and epileptic symptoms through altered microbial metabolism of various metabolites including short chain fatty acids. However, the extent to which microbes contribute to altered host metabolism remains unclear.
To better understand this association we conducted a randomized cross-over study examining two ketogenic diets (one fat-based and one protein-based) and their impact on the gut microbiome composition and fecal and serum metabolomes of canines. Over the span of 15 weeks, healthy dogs (n=35) were fed a series of three different diets varying in fat and protein sources (5 weeks each). All dogs started on the base feed, which is a high carbohydrate standard dog food (25/37/38% percent protein/fat/carbohydrates). Then they were randomly allocated to either a high protein diet (53/39/8%) or a high-fat diet (27/68/5%) for 5 weeks. Dogs were then switched to the opposite diet for the final 5 weeks of the study. Stool samples collected at the end of each 5-week time point were used for shotgun metagenomic sequencing (MGX) and fecal metabolome profiling. Paired blood samples were also collected and subjected to metabolome profiling.
We observed strong associations between diet and overall fecal and serum chemical composition as well as gut microbiome composition (PERMANOVA p < 0.05). Similar to previous literature, we found several species of bifidobacteria such as B. criceti and B. pseudolongum to be negatively associated with both ketogenic diets (q < 0.05). Although we also noted multiple other species within the Actinobacteria phyla such as C. tanakaei, to be more abundant during ketogenic diets. Firmicutes species overall were also broadly more abundant during ketogenic diets including poorly characterized Peptostreptococcaceae and Lachnospiraceace species, a result that contrasts with previous work in humans.
Interestingly, the serum ketogenic marker 3-hydroxybutyrate (BHB) was elevated after the ketogenic fat diet (p < 0.001) but was unchanged after the ketogenic protein diet (p=0.297). Both ketogenic diets led to reduced BHB fecal concentrations (p < 0.001). In association with serum BHB levels, community abundances shifted for several gut microbes including an unclassified Eggerthellaceae species. These microbial associations reveal interesting candidates that may play roles in ketogenesis or in regulating serum metabolite profiles outside of the colon. Overall, our current results suggest that ketogenic diets alter gut microbiome communities and lead to shifts in metabolite profiles locally in feces and systemically via serum.
The landscape of novel lateral gene transfer events in the human microbiome
Presented by: Etienne Nzabarushimana
Lateral gene transfer (LGT) is an important mechanism for genomic diversification in microbial populations and communities, including the human microbiome. While previous work has surveyed ancient LGT events in human-associated microbial isolate genomes, the scope, and dynamics of novel LGT events in human microbiomes are not well understood. We addressed this by developing and validating a computational method (Workflow to Annotate Assemblies and Find LGT Events or WAAFLE) to profile novel LGT events from assembled metagenomes. We assessed WAAFLE on synthetic contigs containing spiked LGTs and identified intergenus LGTs with >91% sensitivity and >99.9% specificity. For more challenging intragenus LGT (due to congeneric overlap), we report a still-respectable 51% sensitivity. Applying WAAFLE to >2K human metagenomes from diverse body sites, we identified >100K high-confidence putative, novel LGT events. These events were enriched for mobile elements (as expected), as well as restriction-modification and transport functions, both being particularly intriguing areas for further study given their putative role in viral/phage-mediated LGT defense. LGT frequency was quantifiably influenced by biogeography, the phylogenetic similarity of the involved taxa, and the ecological abundance of the involved taxa. Our findings suggest that LGT is an active process in the human microbiome, occurring far more frequently than previously suspected.
Candida auris and the great ESKAPE: the skin as a reservoir for multi-drug resistance and transmission
Presented by: Diana Proctor
Background: Candida auris is an emerging fungal pathogen of urgent concern due to its ability to cause healthcare-associated infections and outbreaks, its resistance to antimicrobials and disinfectants and its persistence on human skin and in the environment. Early genome sequencing of surveillance isolates indicates that C. auris exhibits limited genomic diversity within an outbreak.
Methods: We sought to identify transmission patterns of C. auris in a skilled nursing facility with high prevalence of C. auris asymptomatic colonization. Our clinical study leveraged patient trace data, medication history and clinical microbiology findings for 36 residents who underwent serial sample collection monthly over 3 months. To quantify the genomic diversity of C. auris colonizing a patient, we integrated sequencing of 75 isolates (175 Gb total) with plate metagenomics of original culture plates (N=28; 6.4Gb), permitting the analysis of tens to thousands of C. auris isolates per patient. Complementing isolate analyses, shotgun metagenomic sequencing of 210 skin, nasal, and perianal samples (3.2 Tb total) was used for metagenome assembled genome (MAG) and read-based analyses to assess strain sharing of C. auris and ESKAPE (Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp.) pathogens in this facility. To assess the generalizability of our findings, we then analyzed over 1,300 publicly available shotgun metagenomic sequencing samples (stool, skin) collected from patients in 7 other facilities.
Results: Fewer than 300 single nucleotide polymorphisms (SNPs) separated C. auris isolates from this facility and the first isolate identified in the Chicago region, which was collected 3 years earlier. On average, 5 SNPs separated C. auris isolates collected from the same individual with 7-28 SNPs private to each subject, suggesting limited personalized diversification of clones. Strikingly, from these 210 shotgun metagenomic samples, we recovered >5X coverage of the C. auris genome from 22 samples, enabling independent estimates of genomic diversity. In addition, we recovered metagenome-assembled-genomes (MAGs) for all ESKAPE pathogens (16 E.coli, 1 S. aureus, 21 A. baumannii, 39 K. pneumoniae, 37 P. aeruginosa, 2 E. faecalis MAGs), and species previously shown to correlate with C. auris relative abundance, including Staphylococcus pettenkoferi (N=42), Providencia stuartti (N=62), and Proteus mirabilis (N=60). Importantly, MAGs for these species had depth > 5X and fraction aligned > 75% (median 86.22%), suggesting strain tracking could be deployed. Incorporating Refseq genomes into phylogenetic trees of MAG single copy marker genes (N=169) and subsequent sequence typing (ST) suggested clonal separation of each pathogen within the facility. Moreover, SNP-based analyses implicated either undetected transmission of ESKAPE pathogens within the facility or very recent acquisition prior to facility admission. For example, only 0-46 SNPs separated the 39 K. pneumoniae MAGs (mean=8.4) with 2 clusters each reaching >99.9% average nucleotide identity (ANI). Intriguingly, even the commensal S. pettenkoferi MAGs (N=42) clustered together in one group at >99.9% ANI, separated by just 0-32 SNPs. Analysis of publicly available metagenomes revealed clonal spread on skin of S. pettenkoferi in 4 other facilities and widespread sharing of E. coli ST131 across virtually all nursing homes. Taken together, our data suggest the skin is a reservoir for ESKAPE pathogens and that it may potentially serve as a reservoir for their transmission.
THE ROLE OF THE GUT MICROBIOME IN THE ASSOCIATION BETWEEN CITRUS FRUIT AND RISK OF DEPRESSION
Presented by: Chatpol (Jamie) Samuthpongtorn
Background: Diet is known to alter the risk of depression. Increasing data also demonstrate a causal role of the gut microbiome in mental illness, via the gut-brain axis. However, it remains unclear how diet and the microbiome mechanistically influence depression risk in humans. Leveraging dietary, metabolomics, microbiome, and depression data, we assessed how gut microbial species and their pathways may mediate the association between depression and citrus, a food group that possibly protects against risk of depression.
Methods: We conducted a prospective study in the Nurses’ Health Study II (NHSII) between 2003 and 2017 among 32,427 middle-aged women free of depression at baseline. Citrus intake was determined using validated food frequency questionnaires collected every 4 years. Depression was defined according to physician-diagnosis and antidepressant use. Between 2013-2014, 207 NHSII participants enrolled in a nested substudy, providing up to 4 stool samples (profiled by shotgun metagenomics) and a blood sample (profiled by LC-MS-based metabolomics). Cox proportional hazard models were used to relate citrus intake with depression risk. Linear mixed effects models were used to relate diet with gut microbial features, and microbial features with depression. We also associated microbial features with a depression-risk score, derived according to levels of circulating serotonin and GABA. All models were adjusted for multiple dietary, medication and lifestyle variables including age, BMI, calorie/alcohol intake, and diet quality. We validated our findings in a subcohort of 307 men in the Health Professionals Follow-up Study (HPFS).
Results: Total citrus intake was associated with a lower risk of incident depression (ptrend 0.001), with a multivariable relative risk of 0.80 (95% CI, 0.68-0.93), comparing extreme quintiles. Within the NHSII substudy, greater citrus intake was associated with increased abundance of Faecalibacterium prausnitzii (β 0.026, FDR 0.17). In turn, levels of F. prausnitzii were higher in non-depressed individuals compared to depressed participants (p 0.003). Greater abundance of F. prausnitzii was also associated with our metabolomics-based depression-risk score in the NHSII (p 0.03), and in the HPFS validation study (p 0.02). In an exploratory analysis of gut microbial pathways, S-Adenosyl-L-Methionine (SAM) cycle I, encoded by F. prausnitzii, was reduced in depressed participants.
Conclusion: Greater citrus intake was prospectively associated with lower risk of depression, and with greater abundance of F. prausnitzii. In turn, participants with depression had lower levels of F. prausnitzii and lower abundance of its genes capable of producing SAM, a compound known to have antidepressant properties. These data offer a potential mechanism by which diet influences the gut microbiome to reduce risk of depression.
HMO-metabolizing bacteria in the gut microbiome of infants with atopic dermatitis/eczema
Presented by: Prioty Sarwar
Atopic dermatitis, commonly known as eczema, is an inflammatory skin condition that affects up to 20% of infants in the United States. The development of eczema in early life is predictive of other allergic diseases later in life in a process known as the atopic march. Like many inflammatory diseases, infants with eczema have an altered gut microbiome. A diet of breast milk is often associated with protection against eczema development. Human milk oligosaccharides (HMOs), the third most abundant component of breast milk, are of particular interest as they are indigestible by infants and act as a prebiotic to shape the gut microbiome. To understand the role of the gut microbiota in eczema development, it is important to know the differences in the HMO metabolizing capacity of the gut microbiota both at the species and gene level. Here, I use two geographically distinct infant cohorts from Michigan (MARCH) and Rhode Island (RESONANCE) to characterize some of these changes. I analyzed the diversity of the gut microbial communities from the two cohorts and found them to have similar alpha and beta diversity indices when comparing eczema and breastfeeding groups. Further, I show the strain level differences of target HMO metabolizing bacteria in the two cohorts. HMO metabolism is one of the ways the gut microbiota influences the development of an infant’s immune system. Therefore, understanding the differences in HMO metabolizing capacity in infants with eczema is important in halting the progression of the atopic march.
A genome catalog of the early-life human skin microbiome
Presented by: Zeyang Shen
Metagenome-assembled genomes have greatly expanded the reference genomes for skin microbiome. However, the current reference genomes are largely based on samples from children and adults in North America and lack representation from infants and individuals from other continents. Here we used ultra-deep shotgun metagenomics sequencing to profile the skin microbiota of over 500 skin swabs collected longitudinally at age 2 months and 12 months from 217 infants who were part of a vitamin D supplementation food allergy prevention trial in Australia, Oceania. Each sample yielded about 10 million non-human reads or 1 billion base pairs on average. To build metagenome-assembled genomes, we used MEGAHIT for assembly and a combination of MetaBAT, MaxBin, and CONCOCT for binning. Prokaryotic genomes were further refined with metaWRAP and then checked for chimerism with GUNC, and eukaryotic genomes were checked for quality with EukCC. We present the Early-Life Skin Genomes (ELSG) catalog, comprising 9,096 nonredundant prokaryotic and 227 eukaryotic genomes with more than 50% completeness and less than 10% contamination. This genome catalog substantially expanded the diversity of species known to comprise the skin microbiome and improved the classification rate of sequenced data by 25% over the standard Kraken 2 database. The protein catalog derived from these genomes provided insights into the functional elements that distinguish early-life skin microbiome. For example, the proteins specific to the ELSG were enriched for functions related to defense mechanisms. By comparing microbial profiles and genomes from 67 mother-infant pairs, we discovered mother-infant transmission of skin microbiome for 7 different species, including Cutibacterium acnes, Rothia mucilaginosa, and Micrococcus luteus. We also found temporal persistence of microbiome within the same infant by analyzing longitudinal samples at two different times. Overall, the ELSG uncovers the skin microbiome of a previously underrepresented age group and population and provides a comprehensive view of skin microbiome diversity, function, and transmission in early life.
Identifying strain-specific associations in colorectal cancer
Presented by: Kelsey Thompson
Colorectal cancer (CRC) is the second most commonly diagnosed malignancy in women and the third in men, and accounts for around 10% of all deaths related to cancer. The progression from healthy intestinal cells, to benign tumors (adenomas), and then to more malignant forms has profound impacts on the composition of the intestinal microbiota. Additionally, the factors influencing this progression are idiopathic but likely involve a combination of genetics, local tumor environment, and extrinsic factors such as diet. Here, we focus on further elucidating the role of the gut microbiome, a large component of the tumor microenvironment, in cancer initiation and progression by considerably expanding on the current largest meta-analysis to include a total of 3,558 samples from 17 public and private studies. Through expanded sample size, increased resolution of the computational tools, and bioinformatic advances we have improved the understanding of the gut ecosystem in CRC. We found several taxa from the Solobacterium group and Clostridia groups enriched in CRC, and several unknown taxa enriched in healthy individuals. However, as has recently been observed with pks+ E. coli, while species-level identification does provide valuable insights; strain-level resolution can often provide the most actionable downstream targets and help to further elucidate the basic biology of the system. Thus, we developed ANPAN, a collection of statistical methods for microbial strain analysis that can identify associations between several different types of microbial genetic variation and host health outcomes (particularly CRC). This collection of methods includes gene-level, phylogenetic, and pathway-level models. When applied to our dataset, the gene model identified 26 transposases or transposable elements associated with CRC status across 16 species, potentially indicating a role for these genes in helping the microbes acquire other genetic elements necessary to adapt to the inflammatory microenvironment of the CRC gut. Meanwhile, the phylogenetic model identifies strong species-wide phylogenetic signals in several species, including species’ typically found to be CRC-associated: Ruminococcus gnavus, Clostridium leptum, Bacteroides fragilis, which could indicate these species have clades with variable CRC risk.
C. acnes phage predation in the healthy human skin microbiome
Presented by: A.Delphine Tripp
Poor predictability of stable microbial colonization undermines our ability to harness probiotics and phage therapy in human health applications. Person-specific diversity contributes to this unpredictability; it is unclear how microbes colonize and evolve on individuals to create microbiomes with unique compositions, in which most genetic variation occurs within species boundaries. Bacteria-phage dynamics are implicated in promoting microbiome diversity by generating species- and strain-level population fluctuations. Yet, much remains unknown about the assembly and structure of phage populations in human ecosystems, and consequently their role in stable microbial colonization. What role do existing prophage and/or free phage play in the success of new bacterial colonization? How are phage populations structured, and is prophage carriage a source of acquisition? Sebaceous skin offers a tractable model to unravel phage population structure due to its low species-level complexity, ease of temporospatial sampling, and selective conduciveness to colonization at different human developmental stages. Here, we combine a culture-based whole genome sequencing and shotgun metagenomic approach to examine the coevolutionary dynamics of the highly abundant and ubiquitous skin commensal Cutibacterium acnes and its phage. We begin by screening for prophages in our collection of 3,794 whole genome sequenced C. acnes isolates sampled from human skin. We find that the frequency of prophage carriage is low in C. acnes; in 1.27% of isolates the well-characterized dsDNA pseudolysogen is detected and in 5% of isolates a novel ssDNA lysogen is detected. Interestingly, we find that neither phage co-occur within a single bacterial genome, and that the set of isolates collected from an individual tend to be dominated by a single phage type. When age is considered, we find that adolescent and younger individuals tend to carry the ssDNA lysogen more than adults, who carry the dsDNA pseudolysogen at a higher frequency. This result is recapitulated in our metagenomic samples, where we also observed that the relative abundance of C. acnes bacteria increases with age. The possible relationship between this age-based increase in C. acnes bacterial abundance and differences in prophage carriage is the topic of further investigation. Overall, these findings support future studies into the role of prophages and phage predation in the colonization of the skin microbiome.
Mining metagenomes reveals gut mycobiome alterations with metformin and type 2 diabetes mellitus
Presented by: Emily Van Syoc
Multiple recent studies have implicated the gut mycobiome, the fungal component of the microbiome, in human diseases. Considering that the role of the bacterial microbiome (bacteriome) in metabolic disease is well established, it is plausible that the mycobiome could also play an important role. Yet, recent studies of the human mycobiome have been small with inconsistent findings regarding metabolic diseases, and they did not account for use of oral pharmaceuticals. Oral pharmaceuticals, including the antidiabetic drug metformin, interact with gut bacteria and alter microbial metabolism with subsequent consequences for host glucose regulation. However, the potential interactions of pharmaceuticals with gut fungi remain entirely unknown. In this article we reanalyze published shotgun metagenomics from 9 studies to quantify if and to what degree there is a conserved relationship between the gut mycobiome and metabolic disease. To ensure our inferences are reproducible and statistically rigorous, we use Bayesian multinomial logistic normal models to account for numerous sources of variation and potential confounding, including count variation, compositional constraints, and batch effects induced by differences in study design and sample processing. Using these methods, we analyze data from over 1,000 human metagenomic samples and identify consistent associations with fungal genera, type 2 diabetes mellitus, and metformin treatment. Beyond humans, we perform a novel murine study to show that these relationships are reproducible across species. Overall, we conclude that there is a consistent relationship between the gut mycobiome, T2D, and metformin treatment.
Scalable virome enrichment methods for community detection and quantification
Presented by: Ya Lea Wang
Viruses are important but often overlooked members of most microbial communities, including the human gut, where many remain uncharacterized. This is due to a combination of both computational and experimental limitations: viral nucleotides are difficult to enrich and extract, and once sequenced, their uniqueness and rapid evolutionary divergence can make them difficult to classify. The limitations of high-throughput sequencing approaches to address this have been noted previously, but to our knowledge, no study has evaluated the efficiency of specific protocols for retaining viral nucleotides from a community while depleting non-viral members.
Here, we present our work benchmarking varied experimental protocols to isolate virus-like particles (VLP) from gut microbial communities. Different experimental parameters drawn from multiple previous studies were evaluated to develop an optimized protocol, which was further validated in mock communities (viruses representing common gut viral families) and in spiked stool samples. The optimized VLP isolation protocol efficiently reduced bacterial signals below the limit of detection in mock viral communities. In spiked stool samples, the protocol depleted bacterial signals by approximately 100-fold – although, notably, this still left non-viral nucleotides in the majority in many cases. Different viral clades were also differentially affected by changes in experimental parameters, leading to bias relative to the ground truth. We thus provide a standardized and optimized protocol for gut VLP isolation, with known limits of detection and differential extraction efficiency among potential viral targets.
We are currently carrying out analysis of metagenomic and metatranscriptomic sequencing from VLP-treated preemie stool samples to evaluate the protocol on real-world samples at scale. We are also continuing to improve BAQLaVa (Bioinformatic Application for Quantification and Labeling of Viral taxonomy), a newly developed integrative computational method for virome profiling. Together, we hope these tools will improve experimental and bioinformatic capabilities for gut virome profiling.
CRISPR spacer acquisition is a rare event in human gut microbiome
Presented by: An-Ni Zhang
Host-parasite (host-virus) interactions are important for all cellular life, for example, human immunity and SARS-COV2. In bacteria and archaea, CRISPR systems actively acquire spacers to ensure continued defense against phages, which have been reported to acquire a new spacer within a few hours or days in an experimental setting. However, spacer acquisition in natural environments has often been too slow to observe, and limited literature reports suggest a much slower rate (new spacers only acquired over months or years). This high variance highlights the need to improve our understanding of host-parasite interactions in nature.
By investigating temporal WGS datasets and metagenomes of human gut microbiome in healthy individuals, we found that spacer acquisition is a rare event in human gut microbiome, with an average rate of 1 spacer per 2,142-5,000 cell divisions, i.e. over 7-8 years. This low rate reflects only a small proportion of phage challenged the CRISPR systems in the human gut microbiome.
Bifidobacterium longum shows a significantly higher rate of spacer acquisition than the other gut microbiome species. We found six recently acquired spacers were highly prevalent, consecutive in the same order, in B.longum lineages from 14 different human subjects in both the United States and Europe. Those six spacers locate on different parts of the B.longum genomes, while their neighborhood (50k-135k bp) remains highly similar across B.longum lineages, including a transposase and the whole CRISPR system. This indicates that the high spacer acquisition rate in B.longum is mainly contributed by horizontal gene transfer.
The rare spacer acquisition in CRISPR suggests that CRISPR might not be the primary risk of effective phage therapy for the majority of human microbiome. The results of this study may inform future efforts involving phage therapy and pandemic defense.
Predicting functions of uncharacterized gene products in microbial communities
Presented by: Yancong Zhang
Microbial communities are rich reservoirs for molecular functions that influence environmental and host-associated chemistry, with numerous roles in ecosystem maintenance, health, and disease. However, our knowledge of these molecular mechanisms is limited, due to the massive range of microbial genetic material in comparison to the limited throughput available for experimental characterization. Here, we developed a novel method to systematically predict functional capabilities of uncharacterized proteins in microbial communities by assessing high-dimensional community-wide data. We predicted potential functions for the majority of uncharacterized protein families (~70% of total) in 1,595 gut metagenomes (MGX) and 800 metatranscriptomes (MTX) from the Integrative Human Microbiome Project (HMP2). Using only MTX-based information, our approach achieved an average of approximately 0.7 AUC for Gene Ontology (GO) biological process term prediction. By aggregating predictions from other types of information (e.g. sequence similarity), the AUC was further improved to ~0.88. Further evaluations showed that our method recapitulated comparable, realistic prediction profiles from microbial communities when compared to state-of-the-art tools for function prediction in single-organisms. FUGAsseM predicted high-confidence functional annotations for >283,000 protein families (70% previously uncharacterized), including >20,000 novel protein families (i.e. those without substantial homology to isolate genomes). The functional capacity of microbial proteins was extensively unexplored by leveraging MTX-based coexpression for both common gut taxa and less-studied species in the human gut. Our method is generalizable to any type of microbial community, providing a new approach to predict microbial protein functions. We implemented it as an open-source tool, FUGAsseM (Function predictor of Uncharacterized Gene products by Assessing high-dimensional community data in Microbiomes), along with documentation available at http://huttenhower.sph.harvard.edu/fugassem. This study expands the functional landscape of the human microbiome and allows better exploitation of microbial proteins in any under-characterized communities.