Microbial Community Assembly and Dynamics:.from AMD biofilms to colonization of the premature infant gut

Transcription

1 Microbial Community Assembly and Dynamics:.from AMD biofilms to colonization of the premature infant gut Jill Banfield

2 Talk Overview i) AMD microbial biofilms: an example of reproducible community assembly in a natural system - ecological succession -structuring (by O 2 ) i) Transitioning approaches established in AMD (integrated omics ) and new technologies (high throughput sequencing) to time series analyses of other systems - > colonization of the premature infant gut iii) Community Proteomics: cultivation-independent functional study 2

3 Microbial metabolism impacts environmental chemistry slow FeS O 2 +H 2 O Fe SO H + Microbial biofilms in acid 511 g C m -2 y -1 (~Mediteranianshrubland) Increase in soluble metal Acidification FeS Fe + 8H 2 O 15Fe + 2SO H ph -3.5 to ~ 1.2 O 2 microbial up to ~10 6 times faster MICROBES Chemoautotrophs O 2 Lab experiments

4 A B m u c k T h in B io film 1, J u n e F e (II) (m M )

5 Chemical gradients may explain biological and functional differentiation that emerges as biofilms thicken and communities diversify Fluorescent in situ hybridization of organism-specific probes Archaea Bacteria Antibody (cyt579) Wilmeset al. 2009

6 Community (meta)genomics - Access to genomes of uncultivated organisms - Assays of population level diversity (isolates vs. populations) - Enable functional (e.g., proteomic) studies in community context Insert Read Read

7 16S rrna gene sequence tree: ARMAN-5 ARMAN-1 Fungi Reconstruction of genomes ( ) from AMD biofilms Tyson et al Goltsman et al Baker et al Denefet al Yeltonet al Dick et al. in prep. Miller et al.in prep. Sharon et al. in prep. Tyson et al IV Modified from G. Dick et al. 2009

8 Complex interactions within microbial communities ice Support grid ice Single membrane Cell wall vacuole Thermoplasmatales lineage cell ARMAN cell virus 3D tomographic reconstruction L. Comolli et al Baker et al. 2009

9 Microbial community proteomics: functional assays in situ LC-MS/MS DNA proteins peptides Genomic data from a microbial community cells MS spectra predicted protein sequence predicted peptides protein identification 9

10 Proteomic analysis of the dominant bacterium Early in colonization: Mobile DNA elements Ribosome assembly Cell division Stress defense Unknown function Transcription Mueller et al Late in colonization: General biosynthesis Transport & secretion Motility & signaling Translation DNA replication The main factor determining how an organism uses its genome in these biofilms is the species membership of the microbial community

11 Where we started (2001): AMD Communities Abundance Proteomics ~ 45 % Sequencing Rank Where we are going.. Abundance Sediment, Soil, Marine, Human Tyson et al., Nature, 2004 Rank 11

12 Microbial communities associated with the human body, in particular the gastrointestinal tract, play crucial roles in health and disease QUESTION: How do specific patterns in gut microbial succession relate to health and disease, and specifically to neonatal necrotizing enterocolitis(nec)? METHOD: Time-series analyses of the gut colonization in premature newborns over the first 3 weeks to identify potential connections between microbial colonization, intestinal inflammation, and disease BROADER RELEVANCE: The newborn intestinal tract is a model ecosystem suitable for development of methods for human microbiome studies 12

13 Community genomic analyses of preterm infant gut colonization First study subject -Preterm female infant (28 weeks) - Delivered by caesarean section - Day 1-7: broad-spectrum antibiotics - Three feeding regime phases Metagenomicanalysis of the third colonization phase using 454 sequencing. Objective: whole genome-based analysis for dominant organisms and phage. Morowitz et al. PNAS, 2010

14 454 16S rrna tag sequencing and community genomic analysis Community genomics ~ 60 Mb / sample, co-assembled Dominance by: Serratia Citrobacter Enterococcus faecalis minor populations of Pseudomonas, Klebsiella, Enterobacter, multiple plasmids, phage Morowitz et al. PNAS 2010 (16S rrna tag sequencing by Costello and Relman)

15 UC1 Serratia Curated genome has 9 gaps, 7 of these are the rrnaregions Genome is syntenous with S. marcescens and S. proteamaculans -97.3% aaidentity (AAI) over 4,089 genes with S. marcescens (courtesy, Sanger Inst.) % AAI over 3,672 genes with S. proteamaculans Rearranged or in/del regions show elevated sequence divergence relative to syntenousorthologs(77 % and 58% AAI, respectively), encode: e.g., protocatechuateutilization, fimbrialbiosynthesis and export, nitrate reduction, general secretion, siderophore(enterobactin) synthesis and transport, tetrathionate reduction and regulation, osmoprotectanttransport. NOVEL genes include: transport (e.g., iron uptake), regulation, EPS biosynthesis, adhesion, antibiotic biosynthesis, virulence, quorum sensing, pgaoperon genes involved in polysaccharide synthesis for biofilm adhesion, propanoate metabolism, a regulon for allantoin utilization (virulence-associated), yjf-sgaoperon genes to use vitamin Cas an energy source, a large nonribosomal peptide biosynthesis protein. 15

16 Citrobacter : metagenomicdata confuses assemblers Strain sequence variation regions and insertions /deletions fragment assemblies Morowitz et al. PNAS, 2010

17 UC1 Citrobacter UC1CIT 2 co-existing strain populations: highly fragmented assembly (~1,400 contigs) Manual curation-> 10 contigs(largest is 2.55 Mb), 8 gaps are rrnaregions Genome(s) syntenouswith Citrobacter30_2 * draft (Broad Institute) and C. koseri * isolated from a patient with Crohn s disease Both UC1CIT strains and C. koseri have lateral flagella, 30_2 does not TWO UC1 STRAINS: Significant differences in strain abundance over time: -> differences are ecologically significant Strain genomes differ in: - gene sequence - gene content (blocks) - intergenic regions Cit-ii Cit-i Day 16 Day 18 Day 21

18 Link population abundance shifts to genetic traits that underlie microbial physiology and impact the host Genome-wide analysis Highly divergent intergenicregions Highly divergent proteins Unique gene content UC1CIT-ii UC1CIT-ii strain orthologs UC1CIT-i strain genome Other Citrobacter species orthologs (HMP isolate genomes) 18

19 Hotspots of intergenic variation mapped, genome-wide by comparing minor vs. dominant strain (99% nt ID) 131 bpintergenicregion vs. 57 bpintergenicregion with no secondary structure Differential expression of shared genes may be as important in adaptation as strainspecific genes