Resilience & Response to...

Resilience & Response to Antimicrobials

Les DethlefsenRelman Lab

[email protected]

Interplay of the Microbiome, Environmental Stressors & Human Health

April 27, 2011

mailto:[email protected]

mailto:[email protected]

Start at the beginning...

Complex histories of colonization...

Palmer et al., PLoS Biology, 2007important consequences for human health and physiology.These interactions can have beneficial nutritional, immuno-logical, and developmental effects, or pathogenic effects forthe host [2,5,7,18,45].

This study began with the development of a DNA micro-array with nearly comprehensive coverage of the bacterialtaxa represented in the available database of SSU rRNA gene

sequences. Our microarray design and experimental methodswere based on lessons learned in the validation of a lesscomprehensive SSU rDNA microarray [46]. These previousexperiments enabled us to optimize our methods forcomputational prediction of SSU rDNA hybridization behav-iors, and to develop an experimental protocol that maxi-mized hybridization specificity. The excellent concordance in

Figure 7. Temporal Profiles of the Most Abundant Level 3 Taxonomic Groups

Level 3 taxonomic groups were selected for display if their mean (normalized) relative abundance across all baby samples was greater than 1%. The x-axis indicates days since birth and is shown on a log scale, and the y-axis shows estimated (normalized) relative abundance. For some babies, no valuesare plotted for the first few days because the total amount of bacteria in the stool samples collected on those days was insufficient for microarray-basedanalysis.doi:10.1371/journal.pbio.0050177.g007

PLoS Biology | www.plosbiology.org July 2007 | Volume 5 | Issue 7 | e1771566

Microbiota of the Infant Intestine

sequencing of the species-specific PCR product confirmed itspresence. In seven of 12 cases, none of the array positive (ornegative) samples yielded an amplified product in the PCRanalysis. For four remaining cases, the ostensibly species-specific PCR assay yielded an amplified product of theexpected size, but the clones sequenced from this productdid not correspond to the expected species. We furtherinvestigated these four cases by sequencing a clone libraryobtained by amplification with the same broad-range primersthat were used in preparation for microarray analysis.Although the sequencing did not confirm the presence ofany of the four questionable species/taxa, it provided strong

evidence for a major source of false-positive hybridizationsignals. Specifically, in three of the four cases, we identified arelatively abundant species whose rDNA sequence wassufficiently similar to the probe sequence that it was likelyto account for the observed signal. In one case (Legionellapneumophila), which was predicted to be present at approx-imately 1%, we were unable to identify any candidate speciesthat could account for the hybridization signal (i.e., none withbest BLAST matches scores !30), among our set of 192sequences. Since our power to detect a species present at apartial abundance of 1% was only 85%, it remains possiblethat this species, or another species with a similar SSU rDNAsequence, could have been present at a low abundance in thesuspect samples.

Detection and Quantification of Fungi and ArchaeaBoth our DNA extraction and rDNA amplification methods

were optimized for bacteria and suboptimal for eukaryotesand archaea, thus we separately tested for the presence andabundance of fungi or archaea by means of qPCR assays withbroad specificity for the respective taxonomic groups. Basedon our qPCR analysis, fungi were intermittently detectable instool samples at relatively low abundance (104–106 rRNAgenes/g fecal wet weight), persisting for varying durations inindividual babies, through the first year of life. One of thebabies in this study (baby 10) was noted to have a diaper rash,as well as oral thrush, both of which are commonly caused bya fungus (Candida), and which were treated with an antifungalagent (nystatin). The qPCR analysis detected especially highlevels of fungal rDNA in stool samples from this baby,particularly during the period in which these findings weredescribed. This baby’s mother also had notably high levels offungal SSU rDNA sequences in her prenatal vaginal swabsample, but not in her ‘‘day 0’’ stool sample.The prevalence of archaea was considerably lower and

more variable than that of fungi or bacteria; qPCR analysisdetected archaeal rRNA genes (in the range of 103–106 rRNAgenes/g) in only seven babies during their first year of life, andin four of these babies, they were detected in only a singlesample. In these babies, archaea appeared only transiently,and almost exclusively in the first few weeks of life; they weredetected in only one infant after the fifth week of life. Limitedanalysis of archaeal sequences amplified from the threematernal stool samples that tested positive for archaea(mothers 4, 9, and 12) revealed a predominance of Methano-brevibacter smithii (7/8 archaeal clones identified, including atleast one clone from each mother), with one additional(uncultured) archaeal phylotype. Results of qPCR analysis offungi and archaea are included in Dataset S5 and showngraphically with bacterial qPCR results in Figure S2.

Discussion

The microbial colonization of the infant GI tract is aremarkable episode in the human lifecycle. Every time ahuman baby is born, a rich and dynamic ecosystem developsfrom a sterile environment. Within days, the microbialimmigrants establish a thriving community whose populationsoon outnumbers that of the baby’s own cells. The evolution-arily ancient symbiosis between the human GI tract and itsresident microbiota undoubtedly involves diverse reciprocalinteractions between the microbiota and the host, with

Figure 6. Temporal Patterns in Average Pairwise Similarity of Infant StoolMicrobiota Profiles

(A) Similarity between infants over time. For each time point for which atleast six babies were profiled, we calculated the mean pairwise Pearsoncorrelation between the level 4 taxonomic profiles of all babies at thattime point. The mean pairwise Pearson correlation between theseprofiles in 18 adult participants in this study (nine males and ninefemales) is also shown (open circle indicated by the arrow).(B) Progression towards adult-like flora over time. For each time point forwhich at least four babies were profiled, we calculated the mean Pearsoncorrelation between the level 4 taxonomic profiles of all babies at thattime point and a ‘‘generic adult’’ profile. The generic adult profile is thecentroid of 18 (nine male and nine female) adults (parents in this study).doi:10.1371/journal.pbio.0050177.g006

PLoS Biology | www.plosbiology.org July 2007 | Volume 5 | Issue 7 | e1771565

Microbiota of the Infant Intestine

Phylogenetically diverse, individualized communities...

Eckburg et al., Science, 2005

continued sequencing from our samples(Q130, Q300, and Q200 phylotypes in subjectsA, B, and C) (Fig. 2 and figs. S2 and S3).These estimates must be considered as lowerbounds, because both the observed and theestimated richness have increased in parallelwith additional sampling effort (Fig. 2 and fig.S3). Coverage was 99.0% over all bacterialclone libraries combined, meaning that onenew unique phylotype would be expected forevery 100 additional sequenced clones (18).

The microbial community appeared morediverse in subject B than in A or C, based oninspection of the richness and evenness of theclone distribution across the phylogenetic tree

(Fig. 1). The Rao diversity coefficient (19),which accounts for both phylotype abundanceand dissimilarity, was indeed higher for Bthan for the other subjects (fig. S7). Thispattern was not found with traditional, that is,Shannon and Simpson, diversity indices,which assess only relative phylotype abun-dance (20). Within each subject, the mucosalsamples demonstrated similar diversity pro-files, regardless of the index used (fig. S7).

Previous investigations have not rigorouslyaddressed possible differences in the intestinalmicroflora between subjects, between anatom-ical sites, or between stool and mucosal com-munities. We applied techniques that are

based on the relative abundance of sequenceswithin communities and the extent of geneticdivergence between sequences. We first com-pared inter- and intrasubject variability usingdouble principal coordinate analysis (DPCoA)(19). The greatest amount of variability wasexplained by intersubject differences; stool-mucosa differences explained most of thevariability remaining in the data (Fig. 3).The relative lack of variation among mucosalsites was further examined. The FST statisticof population genetics (21) was used to com-pare genetic diversity within each subject; thisrevealed that the mucosal populations of sub-jects A and B were significantly distinct com-

Fig. 1. Number of sequences per phy-lotype for each sample. The y axis isa neighbor-joining phylogenetic treecontaining one representative ofeach of the 395 phylotypes from thisstudy; each row is a different phylo-type. The phyla (Bacteroidetes, non-Alphaproteobacteria, unclassified nearCyanobacteria, Actinobacteria, Firmi-cutes, Fusobacteria, and Alphaproteino-bacteria, ordered top to bottom) arecolor coded as in Fig. 3 and fig. S1.Each column is labeled by subject(A, B, C) and anatomical site. Foreach phylotype, the clone abundanceis indicated by a grayscale value.

10 JUNE 2005 VOL 308 SCIENCE www.sciencemag.org1636

R E P O R T S

Ecologically diverse, interdependent communities...

Flint et al., Adv. Appl. Microbiol., 2004of insoluble substrates by obligately anaerobic gut microorganisms.Motility and chemotaxis may assist some bacterial species, fungalzoospores, and protozoa in reaching their substrates. Possible quo-rum-sensing molecules have been detected in culture fluid fromsome rumen bacteria (Mitsumori et al., 2003). Initial attachment ofcellulolytic bacterial cells is often likely to involve the extensiveglycocalyx (Latham et al., 1978; Miron et al., 2001; Roger et al., 1990).In addition, substrate-binding modules in microbial enzymes and

FIG. 2. Nutritional interactions between polysaccharide degrading microorganisms inthe rumen. This figure has been modified from Flint (1997), with permission.

POLYSACCHARIDE BREAKDOWN BY GUT ANAEROBES 95

How about the real beginning...

National Geographic

Migrations recorded by Helicobacter pylori...

Falush et al., Science, 2003

Similarly, AE2 nucleotides are most frequent inSpain, Sudan, and Israel, but the isolates fromSudan and Israel possess lower levels of AE1than do European isolates. Thus, AE1 and AE2probably reached Europe from different sourc-es, AE1 primarily from the direction of centralAsia and AE2 primarily from the Near East andNorth Africa.

Further reconstruction of the history ofH. pylori is best done in the context ofcurrent knowledge about human migration.As with a human population tree (21),hpEurope derives from a short centralbranch between hpEastAsia and hpAfrica1(Fig. 1A), hinting at a parallel history ofintercontinental gene flow to Europe forhumans and bacteria. Furthermore, the rel-ative contribution of AE2 versus AE1 cor-relates significantly with the first principlecomponent of European human variation(table S1), which is thought to reflect theentry of neolithic farmers into Europe fromthe Near East (20). The second principlecomponent has been tentatively attributedto the migratory fluxes that brought Uraliclanguages to Europe, and indeed correlatedweakly with AE1 versus AE2 (r ! 0.6, P !.13) (table S1). It seems that neither AE1nor AE2 was harbored by the original Pa-leolithic hunter-gatherers in Europe, be-cause considerable AE1 or AE2 ancestry isfound outside Europe, whereas paleolithicY-chromosome haplotypes are largely re-stricted to Europe (18).

Known human migrations can also explainthe spread of hpEastAsia and hpAfrica1 popu-lations (Fig. 3B). Current models (22, 23) agreethat speakers of Austronesian languages (Mao-ris and other Polynesians) arrived in New Zea-land after sequential island-hopping that is like-ly to have resulted in repeated human popula-

tion bottlenecks. Indeed, consistent with popu-lation bottlenecks, the genetic diversity withinthe hspMaori sample is extremely low (Fig. 1),and the pattern of nucleotide polymorphismswithin subpopulations implies that there hasbeen strong drift in the evolution of the hsp-Maori population (15) (fig. S3). The isolation ofhpEastAsia from Native Americans (7, 8) canbe similarly explained by hpEastAsia’s beingcarried during the colonization of the Americasthat began at least 12,000 years ago. UnlikehspMaori, hspAmerind did not show signs ofstrong drift, implying that H. pylori accompa-nied the ancestors of modern Amerinds andInuits in large numbers of individuals and/orwas introduced on multiple occasions.

The high degree of similarity betweenhspWAfrica and hspSAfrica (Fig. 1B, fig.S3) is concordant with the low genetic dis-tances (20) observed between speakers of theNiger-Congo family of languages and is con-sistent with hspSAfrica’s being carried toSouthern Africa during the rapid expansionof Bantu farmers from central West Africa(24). Given this scenario, one possibility toaccount for the extremely distinct hpAfrica2

population is that they colonized the Khoisanhunter-gatherer inhabitants of Southern Afri-ca, who fall on one of the deepest branches ofan African human population tree (20) andare very distinct from Bantu.

Modern migrations of slaves from WestAfrica to the Americas and of Europeans toSouth Africa, the Americas, and Australasiaare probably responsible for the current exis-tence of hspWAfrica and hpEurope in theseand other locations (Table 1). According tothis interpretation, the past few centuriessince modern human migrations were tooshort for the distinctions between multiplebacterial populations to become blurred.

The assignments of particular human mi-grations to migrations of H. pylori popula-tions can allow dating of the bacterial pop-ulation tree by archaeological events. Thefive ancestral populations existed beforethe separation of hspAmerind from the oth-er hpEastAsia populations (Fig. 1, B andC), which is estimated to have occurred atleast 12,000 years ago. Accordingly, H.pylori has probably accompanied anatomi-cally modern humans since their origins.

Fig. 2. Ancestral sources of individual nucleo-tides in eight selected isolates. The origin ofeach polymorphic nucleotide (colors as in Fig.1C) is shown for each of the eight gene frag-ments. The geographical sources of each isolateare shown above each graph.

Fig. 3. Putative modern and ancient migrations of H. pylori. (A) Average proportion of ancestralnucleotides by source. Numbers correspond to the codes in Table 1 and colors are as in Fig. 1C. (B)Interpretation. Arrows indicate specific migrations of humans and H. pylori populations. BP, yearsbefore present.

R E P O R T S

7 MARCH 2003 VOL 299 SCIENCE www.sciencemag.org1584

on A

pril 25, 2011

ww

w.s

cie

ncem

ag.o

rgD

ow

nlo

aded fro

m

Phylogenetic diversity but functional similarity...

Turnbaugh et al., Nature, 2008; Arumugam et al., Nature, 2011but see also De Filippo et al., PNAS, 2010

and taxonomic similarity (see Supplementary Methods) disclosed asignificant association: individuals with similar taxonomic profilesalso share similar metabolic profiles (P, 0.001; Mantel test).

Functional clustering of phylum-wide sequence bins representingmicrobiome reads assigned to 23 human gut Firmicutes and 14Bacteroidetes reference genomes showed discrete clustering byphylum (Supplementary Figs 14a and 15). Bootstrap analyses of therelative abundance ofmetabolic pathways in themicrobiome-derivedFirmicutes and Bacteroidetes sequence bins disclosed 26 pathwayswith a significantly different relative abundance (SupplementaryFig. 14a). The Bacteroidetes bins were enriched for several carbohyd-rate metabolism pathways, whereas the Firmicutes bins were enrichedfor transport systems. This finding is consistent with our CAZymeanalysis, which revealed a significantly higher relative abundance ofglycoside hydrolases, carbohydrate-binding modules, glycosyltrans-ferases, polysaccharide lyases and carbohydrate esterases in theBacteroidetes sequence bins (Supplementary Fig. 14b).

One of the major goals of the International Human MicrobiomeProject(s) is to determine whether there is an identifiable ‘coremicrobiome’ of shared organisms, genes or functional capabilitiesfound in a given body habitat of all or the vast majority of humans1.Although all of the 18 gut microbiomes surveyed showed a high level

ofb-diversity with respect to the relative abundance of bacterial phyla(Fig. 3a), analysis of the relative abundance of broad functional cat-egories of genes andmetabolic pathways (KEGG) revealed a generallyconsistent pattern regardless of the sample surveyed (Fig. 3b andSupplementary Table 11): the pattern is also consistent with resultswe obtained from ameta-analysis of previously published gut micro-biome data sets from nine adults20,21 (Supplementary Fig. 16). Thisconsistency is not simply due to the broad level of these annotations,as a similar analysis of Bacteroidetes and Firmicutes reference gen-omes revealed substantial variation in the relative abundance of eachcategory (see Supplementary Fig. 17). Furthermore, pairwise com-parisons of metabolic profiles obtained from the 18 microbiomes inthis study revealed an average value of R2 of 0.976 0.002 (Fig. 2a),indicating a high level of functional similarity.

Overall functional diversity was compared using the Shannonindex22, a measurement that combines diversity (the number of dif-ferent metabolic pathways) and evenness (the relative abundance ofeach pathway). The human gut microbiomes surveyed had a stableand high Shannon index value (4.636 0.01), close to the maximumpossible level of functional diversity (5.54; see SupplementaryMethods). Despite the presence of a small number of abundant meta-bolic pathways (listed in Supplementary Table 11), the overall func-tional profile of each gutmicrobiome is quite even (Shannon evennessof 0.846 0.001 on a scale of 0–1), demonstrating that most metabolicpathways are found at a similar level of abundance. Interestingly, thelevel of functional diversity in each microbiome was significantlylinked to the relative abundance of the Bacteroidetes (R25 0.81,P, 1026); microbiomes enriched for Firmicutes/Actinobacteria hada lower level of functional diversity. This observation is consistentwithan analysis of simulatedmetagenomic reads generated from each of 36Bacteroidetes and Firmicutes genomes (Supplementary Fig. 18): onaverage, the Bacteroidetes genomes have a significantly higher level ofboth functional diversity and evenness (Mann–Whitney U-test,P, 0.01).

At a finer level, 26–53% of ‘enzyme’-level functional groups(KEGG/CAZy/STRING) were shared across all 18 microbiomes,whereas 8–22% of the groups were unique to a single microbiome(Supplementary Fig. 19a–c). The ‘core’ functional groups present inall microbiomes were also highly abundant, representing 93–98% ofthe total sequences. Given the higher relative abundance of these ‘core’groups, more than 95% were found after 26.116 2.02 megabases ofsequence were collected from a given microbiome, whereas the ‘vari-able’ groups continued to increase substantially with each additionalmegabase of sequence. Of course, any estimate of the total size of thecore microbiome will depend on sequencing effort, especially for

0.990.980.97R2 value:

PC1 (20%)

Bac

tero

idet

es (%

)

a

b c

High Firmicutes/ActinobacteriaHigh Bacteroidetes

Twin versustwin

Twin versusmother

Unrelated pairs

Func

tiona

l sim

ilarit

y (R

2 )

*

0.94

0.95

0.96

0.97

0.98

0.99

1

R2 = 0.96

0

20

40

60

80

100

–0.6 –0.4 –0.2 0 0.2 0.4 0.6

!"## #"$% #"$% #"$$ #"$$ #"$$ #"$% #"$% #"$$ #"$%

#"$% !"## #"$$ #"$$ #"$& #"$& #"$% #"$% #"$% #"$& #"$% #"$$

#"$% #"$$ !"## !"## #"$% #"$% #"$% #"$% #"$$ #"$$ #"$$ #"$% #"$% #"$$

#"$$ #"$$ !"## !"## #"$$ #"$$ #"$$ #"$& #"$$ !"## !"## #"$$ #"$% #"$% #"$$

#"$$ #"$% #"$$ !"## #"$$ #"$% #"$& #"$$ #"$$ #"$$ #"$$

#"$$ #"$& #"$% #"$$ #"$$ !"## #"$& #"$% #"$$ #"$$ #"$$

#"$& #"$% #"$$ #"$% #"$& !"## #"$& #"$% #"$$ #"$$ #"$% #"$% #"$& #"$%

#"$& !"## #"$$ #"$%

#"$% #"$& #"$$ #"$% #"$& #"$$ !"## #"$$ #"$$ #"$%

#"$% #"$% #"$$ #"$$ #"$$ #"$% #"$% #"$$ !"## #"$$ #"$$

#"$$ #"$% #"$$ !"## #"$$ #"$$ #"$$ #"$$ #"$$ !"## #"$$ #"$% #"$& #"$& #"$%

#"$% #"$% #"$$ !"## #"$$ #"$$ #"$$ #"$% #"$$ #"$$ !"## #"$% #"$% #"$& #"$%

#"$% #"$$ #"$$ #"$% #"$% #"$% !"## #"$$ #"$& #"$% #"$$ #"$$

#"$& #"$% #"$% #"$% #"$& #"$% #"$$ !"## #"$% #"$$ #"$$ #"$$

#"$& #"$% !"## !"## #"$$ #"$%

#"$% #"$$ !"## !"## #"$$ #"$%

#"$% #"$% #"$% #"$& #"$& #"$& #"$$ #"$$ #"$$ #"$$ !"## #"$$

#"$$ #"$$ #"$$ #"$% #"$% #"$% #"$$ #"$$ #"$% #"$% #"$$ !"##

F1T1LeF1T2Le

F1MOvF2T1Le

F2T2Le

F2MObF3T1LeF3T2Le

F3MOv

F4T1Ob

F4T2Ob

F4MOb

F5T1Ob

F5T2Ob

F5MOv

F6T1ObF6T2Ob

F6MOb

F1T1

Le

F1T2

Le

F1M

Ov

F2T1

Le

F2T2

Le

F2M

Ob

F3T1

Le

F3T2

Le

F3M

Ov

F4T1

Ob

F4T2

Ob

F4M

Ob

F5T1

Ob

F5T2

Ob

F5M

Ov

F6T1

Ob

F6T2

Ob

F6M

Ob

<0.97

Figure 2 | Metabolic-pathway-based clustering and analysis of the humangut microbiome of monozygotic twins. a, Clustering of functional profilesbased on the relative abundance of KEGG metabolic pathways. All pairwisecomparisons were made of the profiles by calculating each R2 value. Sampleidentifier nomenclature: family number, twin number or mother, and BMIcategory (Le, lean; Ov, overweight; Ob, obese; for example, F1T1Le standsfor family 1, twin 1, lean). b, The relative abundance of Bacteroidetes as afunction of the first principal component derived from an analysis of KEGGmetabolic profiles. c, Comparisons of functional similarity between twinpairs, between twins and their mother, and between unrelated individuals.Asterisk indicates significant differences (Student’s t-test with Monte Carlo;P, 0.01; mean6 s.e.m.).

[Q][P][I]

[H][F][E]

[G][C][S]

[R][O][U]

[W][Z][N]

[M][T][V]

[Y][D][B]

[L][K][A]

[J]

0

20

40

60

80

100

Rel

ativ

e ab

unda

nce

(%)

OtherProteobacteria

F1T1

LeF1

T2Le

F1M

Ov

F2T1

LeF2

T2Le

F2M

Ob

F3T1

LeF3

T2Le

F3M

Ov

F4T1

Ob

F4T2

Ob

F4M

Ob

F5T1

Ob

F5T2

Ob

F5M

Ov

F6T1

Ob

F6T2

Ob

F6M

Ob

F1T1

LeF1

T2Le

F1M

Ov

F2T1

LeF2

T2Le

F2M

Ob

F3T1

LeF3

T2Le

F3M

Ov

F4T1

Ob

F4T2

Ob

F4M

Ob

F5T1

Ob

F5T2

Ob

F5M

Ov

F6T1

Ob

F6T2

Ob

F6M

Ob

a COG categoriesBacterial phylum b

ActinobacteriaBacteroidetesFirmicutes

Figure 3 | Comparison of taxonomic and functional variations in the humangut microbiome. a, Relative abundance of major phyla across 18 faecalmicrobiomes frommonozygotic twins and theirmothers, based on BLASTXcomparisons of microbiomes and the National Center for BiotechnologyInformation non-redundant database. b, Relative abundance of categories ofgenes across each sampled gut microbiome (letters correspond to categoriesin the COG database).

NATURE LETTERS

3 ©2008 Macmillan Publishers Limited. All rights reserved

species, which could shed light on their survival strategies in thehumangut. In the samples analysedhere, themost abundantmolecularfunctions generally trace back to themost dominant species. However,we identified some abundant orthologous groups that are contributedto primarily by low-abundance genera (see Supplementary Fig. 2, Sup-plementary Table 6 and Supplementary Notes section 3). For example,low-abundance Escherichia contribute over 90% of two abundantproteins associated with bacterial pilus assembly, FimA (COG3539)

and PapC (COG3188), found in one individual (IT-AD-5). Pili enablethe microbes to colonize the epithelium of specific host organs; theyhelpmicrobes to stay longer in the human intestinal tract by binding tohuman mucus or mannose sugars present on intestinal surface struc-tures18. They are also key components in the transfer of plasmidsbetween bacteria through conjugation, often leading to exchange ofprotective functions such as antibiotic resistance18. Pili can thus pro-videmultiple benefits to these low-abundancemicrobes in their efforts

a

d

e

RhodospirillumRhodospirillum

DesulfovibrioDesulfovibrio

VeillonellaVeillonella

PrevotellaPrevotellaStaphylococcusStaphylococcus

HelicobacterHelicobacter

Escherichia/ShigellaEscherichia/Shigella

LeuconostocLeuconostoc

AkkermansiaAkkermansia

RuminococcaceaeRuminococcaceae

HoldemaniaHoldemania

EggerthellaEggerthella

RhodospirillumRhodospirillum

VeillonellaVeillonella

PrevotellaPrevotellaStaphylococcusStaphylococcus

HelicobacterHelicobacter

Escherichia/ShigellaEscherichia/Shigella

LeuconostocLeuconostoc


RuminococcaceaeRuminococcaceae

*Peptostrepto*Peptostrepto coccaceae coccaceae

HoldemaniaHoldemania

EggerthellaEggerthella

Rhodospirillum

Desulfovibrio

Veillonella

PrevotellaStaphylococcus

Helicobacter

Escherichia/Shigella

Leuconostoc

Akkermansia

*Ruminococcaceae

*Peptostrepto coccaceae

Holdemania

EggerthellaClostridialesClostridiales

BacteroidesBacteroidesParabacteroidesParabacteroides

GeobacterGeobacter

LactobacillusLactobacillus

MethanobrevibacterMethanobrevibacter

AlkaliphilusAlkaliphilus

SlackiaSlackia

Clostridiales

BacteroidesParabacteroides

Geobacter

Lactobacillus

Methanobrevibacter

Alkaliphilus

Catenibacterium

Slackia

StaphylococcusStaphylococcus


DialisterDialister

RuminoRuminococcaceaecoccaceae

RuminococcusRuminococcus

SymbiobacteriumSymbiobacteriumMarvinbryantiaMarvinbryantia

GordonibacterGordonibacterSphingoSphingobacteriumbacterium

Staphylococcus

Akkermansia

Dialister

*Rumino coccaceae

Ruminococcus

SymbiobacteriumMarvinbryantia

GordonibacterSphingobacterium

1 2 3

Prevotella

0.1

0.3

0.2

1 2 3

Ruminococcus

0.01

0.04

0.02

0.06

1 2 3

0.1

0.3

0.5

Bacteroides

Enterotype Enterotype

Main contributors

Genera co-occurring with main contributors

Positive correlation (>0.4)

Negative correlation (<–0.4)

Enterotype

PC1

PC2

DA.AD.1

DA.AD.4

ES.AD.1

ES.AD.2

ES.AD.3

FR.AD.3

FR.AD.6

IT.AD.4

JP.AD.1

JP.AD.4

JP.AD.6

JP.AD.7

JP.AD.8

JP.AD.9

AM.F10.T2

DA.AD.2DA.AD.3

ES.AD.4

FR.AD.1

FR.AD.2

FR.AD.4

FR.AD.5

FR.AD.7

FR.AD.8

IT.AD.2

IT.AD.3

IT.AD.5

IT.AD.6

JP.AD.2

JP.AD.3JP.AD.5

IT.AD.1AM.F10.T1

RoseburiaRoseburia

BacteroidesBacteroides

PrevotellaPrevotellaAkkermansiaAkkermansia

RuminococcusRuminococcus

AlistipesAlistipes

Roseburia

Bacteroides

PrevotellaAkkermansia

Ruminococcus

Alistipes

Abu

ndan

ce

!

BlautiaBlautia


PrevotellaPrevotella

Blautia

Bacteroides

Prevotella

*Peptostreptococcaceae*PeptostreptococcaceaeSubdoligranulumSubdoligranulum

PrevotellaPrevotella

MitsuokellaMitsuokella


CatenibacteriumCatenibacterium

*Lachnospiraceae*Lachnospiraceae

*PeptostreptococcaceaeSubdoligranulum

Prevotella

Mitsuokella

Bacteroides

Catenibacterium

*Lachnospiraceae

Obese

IBD

b

c

Figure 2 | Phylogenetic differences between enterotypes. a–c, Between-classanalysis, which visualizes results from PCA and clustering, of the genuscompositions of 33 Sanger metagenomes estimated by mapping themetagenome reads to 1,511 reference genome sequences using an 85%similarity threshold (a), Danish subset containing 85 metagenomes from apublished Illumina data set8 (b) and 154 pyrosequencing-based 16S sequences5

(c) reveal three robust clusters that we call enterotypes. IBD, inflammatorybowel disease. Two principal components are plotted using the ade4 package in

R with each sample represented by a filled circle. The centre of gravity for eachcluster is marked by a rectangle and the coloured ellipse covers 67% of thesamples belonging to the cluster. IBD, inflammatory bowel disease.d, Abundances of the main contributors of each enterotype from the Sangermetagenomes. See Fig. 1 for definition of box plot. e, Co-occurrence networksof the three enterotypes from the Sanger metagenomes. Unclassified generaunder a higher rank are marked by asterisks in b and e.

ARTICLE RESEARCH

0 0 M O N T H 2 0 1 1 | V O L 0 0 0 | N A T U R E | 3

Macmillan Publishers Limited. All rights reserved©2011

Human-microbe mutualism

Humans get

• Resources

• Pathogen resistance

• Developmental signals

• Immune regulation

• Metabolic regulation

• Cometabolism

Microbes get

• Resources

• Habitat

• Dispersal

Rhinos and oxpeckers, a no-cost mutualism...

Some human-microbe mutualisms are costly...

Hooper et al., Ann. Rev. Nutrition 2002; Lux et al., J. Bact., 2007

8 May 2002 12:8 AR AR161-13.tex AR161-13.sgm LaTeX2e(2002/01/18) P1: GJB

294 HOOPER ! MIDTVEDT ! GORDON

Figure 4 Model of howBacteroides thetaiotaomicron induces production of Fuc!1,2Gal"-

containing glycans in the distal intestinal epithelium. L-fucose induces the bacterial fucose

utilization operon that encodes FucR plus the enzymes involved in fucose metabolism (note

that fucose permease is constitutively expressed and is not a member of the operon). In this

model FucR also functions as a corepressor at another locus, control of signal production

(csp), that regulates signaling to the intestinal epithelium. Signaling induces Fuc!1,2Gal"-

glycans in absorptive enterocytes. These glycans are postulated to function as a nutrient

source for B. thetaiotaomicron.

This second locus, designated control of signal production (csp), has not yet been

directly identified, although its existence is inferred from the genetic experiment.

The ability of FucR to coordinate fucose consumption by the microbe and

fucosylated glycan production by the host appears to occur through a distinctive

regulatory mechanism. Figure 4 presents a model consistent with the observed

fucose utilization and host signaling phenotypes of the various isogenic mutant

B. thetaiotaomicron strains described above. In the model, fucose acts through

FucR to induce transcription of the fucose utilization operon and to repress csp.

This explains the nonsignaling phenotype of fucI mutants. Because this enzyme

catalyzes the first enzymatic step in the breakdown of fucose, its absence leads to an

accumulation of fucose in the bacterium. Fucose accumulation, in turn, increases

the proportion of fucose-bound FucR and thereby silences csp. The model predicts

that lowering intracellular fucose levels will shift FucR to its fucose-unbound

form and release csp from inhibition so signaling can occur. In fact, disruption of

the constitutively expressed fucose permease gene reduces import of fucose into

B. thetaiotaomicron and promotes signaling (43).

Annu. R

ev. N

utr

. 2002.2

2:2

83-3

07. D

ow

nlo

aded

fro

m w

ww

.annual

revie

ws.

org

by S

tanfo

rd U

niv

ersi

ty -

Mai

n C

ampus

- R

ober

t C

row

n L

aw L

ibra

ry o

n 0

4/2

5/1

1. F

or

per

sonal

use

only

.

pncO genes were in the same arrangement as they are in theTIGR4 genome (Fig. 1).

Four loss-of-function 2306 mutants were generated by inser-tion duplication mutagenesis in the following single genes: (i)the response regulator gene spiR2 of the TCS13 two-compo-nent system involved in activation of the bacteriocin genes (13,38); (ii) spiB encoding part of the ABC transporter believed torepresent the dedicated transporter for the signaling peptideSpiP; and (iii) pncO and pncP encoding proteins with homol-

ogy to the CAAX family of proteases suggested to be involvedin processing and activation of bacteriocins and/or immunityproteins (35). Finally, the region between spiA and pncOincluding all four putative bacteriocin genes (pncR, pncT, pncI,and pncJ) was replaced with the spectinomycin resistancecassette, resulting in the construct 2306!pncR-K, leavingthe genes encoding the regulatory system as well as pncO,pncQ, and pncP intact. Replacement of the entire peptidecluster was also done in strain TIGR4, resulting in mutantTIGR4!pncA-N.

When tested for apparent bacteriocin production, the spiR2,spiB, and pncO single-gene mutants displayed a bacteriocin-

TABLE 5. Bacteriocin activities of S. pneumoniae 2306 derivativesand sensitivity to the parental strain S. pneumoniae 2306

as the test strain

S. pneumoniaestrain

Activity withL. lactis

Sensitivity to2306

2306 " #2306spiB::pJDB2 # "2306spiR::pJDR1 # "2306pncO::pJDO3 # "2306!pncR-K " #2306pncP::pJDP4 " #

FIG. 3. Bacteriocin-like activity in S. pneumoniae. Different S. pneumoniae strains were tested for bacteriocin production with L. lactis (A, C,E, and G) or M. luteus (B, D, F, and H) as the indicator strain, using the double-layer technique as described in Materials and Methods. The teststrains included R6 (A and B), TIGR4 (C and D), 2306 (E and F), and 632 (G and H). Strain 632 also showed activity against S. mitis NCTC10712(I) and S. pneumoniae R6 (K) as indicator strains. The bacteriocin activities of different S. pneumoniae mutants against L. lactis were alsodetermined; the mutants used were S. pneumoniae 2306spiR::pJDR1 (L), S. pneumoniae 2306spiB::pJDB2 (M), S. pneumoniae 2306pncO::pJDO3(N), S. pneumoniae 2306!pncR-K (O), S. pneumoniae 2306pncP::pJDP4 (P), and S. pneumoniae TIGR4!pncA-N (Q). The mutants with mutationsin spiR2, spiB, and pncO are bacteriocin negative, and the pncP mutant is bacteriocin positive. The 2306!pncR-K mutant is bacteriocin positive,but the zones of inhibition are slightly smaller than those of the wild type. In contrast, the TIGR4!pncA-N mutant is bacteriocin negative. Bar $1 cm.

TABLE 4. Bacteriocin activities of S. pneumoniae strains againstdifferent indicator strains

Indicator strainActivity against the following S. pneumoniae

test strains:

2306 632 TIGR4 628 R6

M. luteus DSM2786 " " " " #L. lactis M61363 " " " " #S. pneumoniae R6 # " # # #S. pneumoniae 2306 # " # # #S. oralis 510 # " # # #S. pyogenes 15 # " # # #S. mitis NCTC10712 " " # # #S. salivarius 674 " # # # #S. sanguis DSM 20567 # # # # #

VOL. 189, 2007 BACTERIOCINS IN S. PNEUMONIAE 7747

Bacteroicin activity

...driven by a shared evolutionary fate.

• Transmission to next generation easy

- unused resources open niches

• Transmission within generations hard

- numerical disadvantage

- retention mechanisms

Gut microbes share our fate because...

• Exploited by cheaters

- loss of mutualism gives higher microbial fitness within a host

• Supported by community selection on fitter hosts with mutualists

- acts across host generations

Costly mutualism...

...equilibrium between cheater and mutualist strains.

• Potential loss of mutualist functions

- immune modulation

- specific pathogen interference

• vertical, horizontal transmission

- weaker community selection

Context of perturbation study...

Antibiotics

• 235 million doses annually in U.S.

- perhaps half unnecessary

• Acute effects

- antibiotic associated diarrhea

- pseudomembranous colitis

• Chronic effects

- allergies, atopic disease

- ??

Study design

Amplify V1-V3 region16S rDNA

454 Titanium Pyrosequencing

3 subjects

Study design

Amplify V1-V3 region16S rDNA


Randomly sheared genomic DNA


2 subjects

Diversity over time

Dethlefsen & Relman, PNAS, 2010

16S rDNA

Heat map of taxon

abundance


16S rDNA

Heat map of taxon abundanceC

los.

XIV

a C

los.

IV

Blautia

MarvinbrantiaMoreyella

LachnospiraCoprococcus II

Roseburia

FaecalibacteriumSubdoligranulumRuminococcus IRuminococcus II

OscillobacterOscillospira

Coprococcus I

D E F


16S rDNA

dbRDA of Bray-Curtis distances


16S rDNA


16S rDNA dbRDA of Bray-Curtis distances

Sampling day

Bray-Curtis distance relative to last day pre-Cp

16S rDNA

Sampling day

16S rDNABray-Curtis distance

relative to last day pre-Cp

MetagenomicSpecies-level taxonomic assignments

pre-Cp

PCO1 - 39.0%

PCO

2 -

22.6

%

DD-CpEE-Cp


1st Cp

PCO1 - 39.0%

PCO

2 -

22.6

%

DD-CpEE-Cp


Interim

PCO1 - 39.0%

PCO

2 -

22.6

%

DD-CpEE-Cp


2nd Cp

PCO1 - 39.0%

PCO

2 -

22.6

%

DD-CpEE-Cp


Post

PCO1 - 39.0%

PCO

2 -

22.6

%

DD-CpEE-Cp

Metagenomic

Text

Metagenomic

Metagenomic

Sampling day

-100 0 100 200 300

Distances relative to last daypre-Cp for phylogenetic and functional

classification of metagenomic reads

Distance relative to last day pre-Cp for 16S rDNA

D

E

Conclusions• Gut microbiota fluctuates around stable long term averages in

both composition and function

• Cp is dramatic perturbation to community composition

- responds and rebounds quickly, perhaps not completely

- idiosyncratic

• Cp is less dramatic perturbation to community function

- unclear whether any persistent effects

- idiosyncratic

• Perturbation is without symptoms (for these subjects)

- nature of restoring force on community isn’t clear

- probably not based on selection for mutualist functions

THANKS!• Doris Duke Foundation

• Tanya Yatsunyenko & Jeff Gordon (Wash. U.)

- Metagenomic sequencing

• Mani Arumugam & Peer Bork (EMBL)

- SmashCommunity bioinformatics

• Liz Costello, Eoghan Harrington & David Relman (Stanford)

• NAS Committee on Use of Emerging Science for Environmental Health Decisions

Abundant functional categories of genesaffected by ciprofloxacin

Subject DCOG 60 - Isoleucyl-tRNA synthetase

COG 513 - Superfamily II DNA and RNA helicasesCOG 768 - Cell division protein FtsI/penicillin-binding protein 2COG 3292 - Predicted periplasmic ligand-binding sensor domain

COG 3507 - Beta-xylosidaseCOG 4646 - DNA methylase

Subject ECOG 1506 - Dipeptidyl aminopeptidases/acylaminoacyl-peptidases

COG 4799 - Acetyl-CoA carboxylase, carboxyltransferase componentNOG 44491 – Oxidoreductase

Different communities in Burkina Faso & Italy

0.8%, respectively). The differential distribution of Firmicutesand Bacteroidetes delineates profound differences between thetwo groups (Fig. S1).Statistical analysis using a parametric test (ANOVA) indicates

that Firmicutes (P = 7.89 ! 10!5) and Bacteroidetes (P = 1.19 !10!6) signi!cantly differentiate the BF from the EU children.This result is strengthened by the nonparametric Kruskal–Wallistest, which again indicated signi!cant discriminating factors inFirmicutes (P = 3.38 ! 10!5), Bacteroidetes (P = 4.80 ! 10!4),Actinobacteria (P = 8.82 ! 10!3), and Spirochaetes (P = 1.11 !10!5) phyla. Firmicutes are twice as abundant in the EU childrenas evidenced by the different ratio between Firmicutes andBacteroidetes (F/B ratio ± SD, 2.8 ± 0.06 in EU and 0.47 ± 0.05in BF), suggesting a dramatically different bacterial colonizationof the human gut in the two populations. Interestingly, Prevotella,Xylanibacter (Bacteroidetes) and Treponema (Spirochaetes) arepresent exclusively in BF children microbiota (Figs. 2 A and B,Fig. S2, and Table S5). We can hypothesize that among theenvironmental factors separating the two populations (diet,sanitation, hygiene, geography, and climate) the presence of

these three genera could be a consequence of high !ber intake,maximizing metabolic energy extraction from ingested plantpolysaccharides.Diet plays a central role in shaping the microbiota, as dem-

onstrated by the fact that bacterial species associated with a high-fat, high-sugar diet promote obesity in gnotobiotic mice (12). Insuch a model, indigenous bacteria maintain energy homeostasisby in"uencing metabolic processes. The ratio of Firmicutes toBacteroidetes differs in obese and lean humans, and this pro-portion decreases with weight loss on low-calorie diet (9). It istherefore reasonable to surmise that the increase in the F/B ratioin EU children, probably driven by their high-calorie diet, mightpredispose them to future obesity. This F/B ratio may also beconsidered a useful obesity biomarker.

16S rRNA Gene Surveys Reveal Hierarchical Separation of the TwoPediatric Populations. We further assessed differences in the totalbacterial community at the single sample level by clustering theEU and BF samples according to their bacterial genera as foundby the RDP classi!er (Ribosomal Database Project v. 2.1).

Fig. 2. 16S rRNA gene surveys reveal a clear separation of two children populations investigated. (A and B) Pie charts of median values of bacterial generapresent in fecal samples of BF and EU children (>3%) found by RDP classi!er v. 2.1. Rings represent corresponding phylum (Bacteroidetes in green andFirmicutes in red) for each of the most frequently represented genera. (C) Dendrogram obtained with complete linkage hierarchical clustering of the samplesfrom BF and EU populations based on their genera. The subcluster located in the middle of the tree contains samples taken from the three youngest (1–2 yold) children of the BF group (16BF, 3BF, and 4BF) and two 1-y-old children of the EU group (2EU and 3EU). (D) Relative abundances (percentage of sequences)of the four most abundant bacterial phyla in each individual among the BF and EU children. Blue area in middle shows abundance of Actinobacteria, mainlyrepresented by Bi!dobacterium genus, in the !ve youngest EU and BF children. (E) Relative abundance (percentage of sequences) of Gram-negative andGram-positive bacteria in each individual. Different distributions of Gram-negative and Gram-positive in the BF and EU populations re"ect differences in thetwo most represented phyla, Bacteroidetes and Firmicutes.

De Filippo et al. PNAS | August 17, 2010 | vol. 107 | no. 33 | 14693

EVOLU

TION

De Filippo et al., PNAS, 2010

• 3 Subjects

• 52-56 samples/subject

• 1-2 DNA extractions/sample

• 1-3 PCR amplifications/extraction

• 4 pyrosequencing runs

• Quality filtering & trimming

• 1.76 million reads analyzed

• 2,582 refOTUs

Operational Taxonomic Unit:

• OTUs by sequence similarity potentially problematic w/pyrosequencing errors

• Clustered Silva database 97% similarity

• Reference database = ‘seeds’ for pyro reads

• Reads clustering w/same reference seq at 95% similarity = refOTU

• Reads not clustering were discarded

Sampling day

Jens

en-S

hann

on d

ista

nce

PhylogeneticFunctional

Metagenomic reads classified according to phylogenetic or functional categoriesJensen-Shannon distance to last pre-Cp sample

Resilience & Response to...

Documents

Transcript of Resilience & Response to...