Upload
jonathan-eisen
View
1.130
Download
3
Embed Size (px)
DESCRIPTION
Talk by Jonathan Eisen for meeting on "All creatures great and small" at UC Davis.
Citation preview
Slides by Jonathan Eisen for BIS2C at UC Davis Spring 2014 1
Don’t Neglect Their Microbiomes
Jonathan A. Eisen @phylogenomics
November 17, 2014
Talk for Nonhumans Meeting
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Obsessions …
!2
Me and My Girl Annapurna
!3
The Story of a Bird
!4
5
9
10
11
Robin in London Examples
MICROBES
Microbes vs Nonhumans
!14
Nonhumans Word Cloud 1
!15
Nonhumans Word Cloud 2
!16
Microbes Better?
!17
Microbes Better?
!17
Nonhumans Better?
!18
Microbes AND Nonhumans
!19
Microbes AND Nonhumans
!19
Microbes and Non Humans 1: Bad Germs
!20
• Animals get and transmit many pathogens
• But … can lead to excess germophobia
Microbes and Nonhumans 2: Mutualisms
!21
5
Sharpshooter:Cuerna sayi
bacteriomes
Sharpshooters harbor two obligatesymbionts in their bacteriomes
Moran et al. 2003 Environ. Microbiol.Moran et al. 2005 Appl. Environ. Microbiol.
Candidatus “Baumannia cicadellinicola” (Gammaproteobacteria)
Candidatus “Sulcia muelleri” (Bacteroidetes)
D Takiya
0.1mm
Bacteriome dissected from anterior abdomen of H. vitripennis
Orange-red portion- Baumannia only
Yellow portion- Baumannia and Sulcia
(Moran et al. 2003 Environmental Microbiology)
Copyright © National Academy of Sciences. All rights reserved.
The Social Biology of Microbial Communities: Workshop Summary
WORKSHOP OVERVIEW 9
et al., 2012). This simple model of persistent colonization of animal epithelia by Gram-negative bacteria provides a “valuable complement to studies of both beneficial and pathogenic consortial interactions, such as in the mammalian in-testine, and chronic disease that involve persistent colonization by Gram-negative bacteria, such as cystic fibrosis” (Nyholm and McFall-Ngai, 2004).
Plant roots and their partners Plants establish associations with several micro-organisms in a relationship somewhat analogous to that of mammals with their gastrointestinal microbiota. The roots of most higher plant species form mycor-rhizae, an association with specific fungal species that significantly improves the plant’s ability to acquire phosphorous, nitrogen, and water from the soil.12 A few plant families, including legumes, associate with nitrogen-fixing bacteria. They colonize the plant’s roots and form specialized nodules, where the bacteria
12 See http://agronomy.wisc.edu/symbiosis.
DC
Figure WO-3
A B
FIGURE WO-3 The bacterium and the squid. A persistent, symbiotic association be-tween the squid Euprymna scolopes (A) and its luminous bacterial symbiont Vibrio fischeri (B) forms within the squid’s light organ (C and D). After colonization of the host’s light organ tissue, V. fischeri induces a series of irreversible developmental changes that trans-form these tissues into a mature, functional light organ (Nyholm and McFall-Ngai, 2004). SOURCE: (A) Images taken by C. Frazee, provided by M. McFall-Ngai and E. G. Ruby; (B) Image provided courtesy of Marianne Engel; (C and D). Reprinted by permission from Macmillan Publishers Ltd: Nature, Dusheck (2002), copyright 2002.
Copyright © National Academy of Sciences. All rights reserved.
The Social Biology of Microbial Communities: Workshop Summary
148 THE SOCIAL BIOLOGY OF MICROBIAL COMMUNITIES
Figure A5-3.epsbitmap
FIGURE A4-3 Cooperation and conflict within the fungus-growing ant microbe symbio-sis. A) Fungus-growing ants forage for substrate to nourish their cultivated fungus, which they also groom to help remove garden parasites. B) In return, the fungus serves as the primary food source for the ants; with some species producing nutrient-rich hyphal swell-ings (gongylidia) that the ants preferentially feed on. Cooperation and conflict is inherent to the ant-fungus mutualism (black arrows, head points toward recipient of benefit), with each symbiont receiving a benefit (+), at a cost to the other (-). Natural selection favors symbionts that increase their own fitness selfishly by exploiting their partner, receiving a benefit (+) without paying the cost (-) associated with providing a benefit in return. C) The mutualism is parasitized by specialized fungi in the genus Escovopsis, which acquire nutrients from the fungus garden at a direct and indirect cost to the cultivated fungus and ants, respectively. Cooperation is enforced, and cheaters minimized, because the selfish interests of both ants and cultivated fungus are aligned (orange triangle) against the para-site Escovopsis.
Microbes and Nonhumans 3: The Microbiome
!22
The Rise of the Microbiome
0
1000
2000
3000
4000
00 01 02 03 04 05 06 07 08 09 10 11 12 13
Pubmed “Microbiome” Hits
The Rise of the Microbiome
0
1000
2000
3000
4000
00 01 02 03 04 05 06 07 08 09 10 11 12 13
Pubmed “Microbiome” Hits
The Rise of the Microbiome
0
1000
2000
3000
4000
00 01 02 03 04 05 06 07 08 09 10 11 12 13
Pubmed “Microbiome” Hits
The Rise of the Microbiome
0
1000
2000
3000
4000
00 01 02 03 04 05 06 07 08 09 10 11 12 13
Pubmed “Microbiome” Hits
The Rise of the Microbiome
Not Just About Humans
!25
• Animals are covered in a cloud of microbes
!26
The Rise of the Microbiome
• Animals are covered in a cloud of microbes
• This “microbiome” likely is involved in many important animal phenotypes
!27
The Rise of the Microbiome
• Animals are covered in a cloud of microbes
• This “microbiome” LIKELY is involved in many important animal phenotypes
!28
The Rise of the Microbiome
• Animals are covered in a cloud of microbes
• This “microbiome” LIKELY is INVOLVED in many important animal phenotypes
!29
The Rise of the Microbiome
Why Now?
Why Now I: Growing Appreciation of Microbial Diversity
!31
Why Now I: Growing Appreciation of Microbial Diversity
!31
Why Now I: Growing Appreciation of Microbial Diversity
!31
Diversity of Form
Why Now I: Growing Appreciation of Microbial Diversity
!31
Diversity of Form
Phylogenetic Diversity
Why Now I: Growing Appreciation of Microbial Diversity
!31
Functional Diversity
Diversity of Form
Phylogenetic Diversity
Why Now I: Growing Appreciation of Microbial Diversity
!31
Functional Diversity
Diversity of Form
Phylogenetic Diversity
MICROBES RUN THE PLANET
Why Now II: Post Genome Blues
!32
Why Now II: Post Genome Blues
!32
Overselling the Human Genome?
Why Now II: Post Genome Blues
!32
Transcriptome
Overselling the Human Genome?
Why Now II: Post Genome Blues
!32
Transcriptome
Epigenome
Overselling the Human Genome?
Why Now II: Post Genome Blues
!32
Transcriptome
VariomeEpigenome
Overselling the Human Genome?
Why Now II: Post Genome Blues
!32
The Microbiome
Transcriptome
VariomeEpigenome
Overselling the Human Genome?
!33
Why Now III: Advances in Culture-Independent Work
!33
Why Now III: Advances in Culture-Independent Work
!33
Observation
Why Now III: Advances in Culture-Independent Work
!33
Culturing Observation
Why Now III: Advances in Culture-Independent Work
!33
Culturing Observation
CountCount
Why Now III: Advances in Culture-Independent Work
!33
<<<<
Culturing Observation
CountCount
Why Now III: Advances in Culture-Independent Work
!33
<<<<
Culturing Observation
CountCount
http://www.google.com/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&docid=rLu5sL207WlE1M&tbnid=CRLQYP7d9d_TcM:&ved=0CAUQjRw&url=h
ttp%3A%2F%2Fwww.biol.unt.edu%2F~jajohnson
%2FDNA_sequencing_process&ei=hFu7U_TyCtOqsQSu9YGwBg&psig=AFQjCN
G-8EBdEljE7-yHFG2KPuBZt8kIPw&ust=140487395121
1424
Why Now III: Advances in Culture-Independent Work
!33
<<<<
Culturing Observation
CountCount
http://www.google.com/url?sa=i&rct=j&q=&esrc=s&source=images&cd=&docid=rLu5sL207WlE1M&tbnid=CRLQYP7d9d_TcM:&ved=0CAUQjRw&url=h
ttp%3A%2F%2Fwww.biol.unt.edu%2F~jajohnson
%2FDNA_sequencing_process&ei=hFu7U_TyCtOqsQSu9YGwBg&psig=AFQjCN
G-8EBdEljE7-yHFG2KPuBZt8kIPw&ust=140487395121
1424
DNA
Why Now III: Advances in Culture-Independent Work
!34
Why Now IV: Sequencing Has Gone Crazy
!34
!3535
Approaching to NGS
Discovery of DNA structure(Cold Spring Harb. Symp. Quant. Biol. 1953;18:123-31)
1953
Sanger sequencing method by F. Sanger(PNAS ,1977, 74: 560-564)
1977
PCR by K. Mullis(Cold Spring Harb Symp Quant Biol. 1986;51 Pt 1:263-73)
1983
Development of pyrosequencing(Anal. Biochem., 1993, 208: 171-175; Science ,1998, 281: 363-365)
1993
1980
1990
2000
2010
Single molecule emulsion PCR 1998
Human Genome Project(Nature , 2001, 409: 860–92; Science, 2001, 291: 1304–1351)
Founded 454 Life Science 2000
454 GS20 sequencer(First NGS sequencer) 2005
Founded Solexa 1998
Solexa Genome Analyzer(First short-read NGS sequencer) 2006
GS FLX sequencer(NGS with 400-500 bp read lenght) 2008
Hi-Seq2000(200Gbp per Flow Cell) 2010
Illumina acquires Solexa(Illumina enters the NGS business) 2006
ABI SOLiD(Short-read sequencer based upon ligation) 2007
Roche acquires 454 Life Sciences(Roche enters the NGS business) 2007
NGS Human Genome sequencing(First Human Genome sequencing based upon NGS technology) 2008
From Slideshare presentation of Cosentino Cristian http://www.slideshare.net/cosentia/high-throughput-equencing
Miseq Roche Jr Ion Torrent PacBio Oxford
Sequencing Has Gone Crazy
Sequencing Revolution
!36
•More genes and genomes
•Deeper sequencing • The rare biosphere • Relative abundance estimates
•More samples (with barcoding) • Times series • Spatially diverse sampling • Fine scale sampling
!37
Turnbaugh et al Nature. 2006 444(7122):1027-31.
Why Now V: Microbiome Functions
IBD vs. normal
Almost all (99.96%) of the phylogenetically assigned genes belongedto the Bacteria and Archaea, reflecting their predominance in the gut.Genes that were not mapped to orthologous groups were clusteredinto gene families (see Methods). To investigate the functional con-tent of the prevalent gene set we computed the total number oforthologous groups and/or gene families present in any combinationof n individuals (with n 5 2–124; see Fig. 2c). This rarefaction ana-lysis shows that the ‘known’ functions (annotated in eggNOG orKEGG) quickly saturate (a value of 5,569 groups was observed): whensampling any subset of 50 individuals, most have been detected.However, three-quarters of the prevalent gut functionalities consistsof uncharacterized orthologous groups and/or completely novel genefamilies (Fig. 2c). When including these groups, the rarefaction curveonly starts to plateau at the very end, at a much higher level (19,338groups were detected), confirming that the extensive sampling of alarge number of individuals was necessary to capture this considerableamount of novel/unknown functionality.
Bacterial functions important for life in the gut
The extensive non-redundant catalogue of the bacterial genes fromthe human intestinal tract provides an opportunity to identify bac-terial functions important for life in this environment. There arefunctions necessary for a bacterium to thrive in a gut context (thatis, the ‘minimal gut genome’) and those involved in the homeostasisof the whole ecosystem, encoded across many species (the ‘minimalgut metagenome’). The first set of functions is expected to be presentin most or all gut bacterial species; the second set in most or allindividuals’ gut samples.
To identify the functions encoded by the minimal gut genome weuse the fact that they should be present in most or all gut bacterialspecies and therefore appear in the gene catalogue at a frequencyabove that of the functions present in only some of the gut bacterialspecies. The relative frequency of different functions can be deducedfrom the number of genes recruited to different eggNOG clusters,after normalization for gene length and copy number (Supplemen-tary Fig. 10a, b). We ranked all the clusters by gene frequencies anddetermined the range that included the clusters specifying well-known essential bacterial functions, such as those determined experi-mentally for a well-studied firmicute, Bacillus subtilis27, hypothe-sizing that additional clusters in this range are equally important.As expected, the range that included most of B. subtilis essentialclusters (86%) was at the very top of the ranking order (Fig. 5).Some 76% of the clusters with essential genes of Escherichia coli28
were within this range, confirming the validity of our approach.This suggests that 1,244 metagenomic clusters found within the range(Supplementary Table 10; termed ‘range clusters’ hereafter) specifyfunctions important for life in the gut.
We found two types of functions among the range clusters: thoserequired in all bacteria (housekeeping) and those potentially specificfor the gut. Among many examples of the first category are thefunctions that are part of main metabolic pathways (for example,central carbon metabolism, amino acid synthesis), and importantprotein complexes (RNA and DNA polymerase, ATP synthase, generalsecretory apparatus). Not surprisingly, projection of the range clusterson the KEGG metabolic pathways gives a highly integrated picture ofthe global gut cell metabolism (Fig. 6a).
The putative gut-specific functions include those involved in adhe-sion to the host proteins (collagen, fibrinogen, fibronectin) or inharvesting sugars of the globoseries glycolipids, which are carriedon blood and epithelial cells. Furthermore, 15% of range clustersencode functions that are present in ,10% of the eggNOG genomes(see Supplementary Fig. 11) and are largely (74.3%) not defined(Fig. 6b). Detailed studies of these should lead to a deeper compre-hension of bacterial life in the gut.
To identify the functions encoded by the minimal gut metagenome,we computed the orthologous groups that are shared by individuals ofour cohort. This minimal set, of 6,313 functions, is much larger than theone estimated in a previous study8. There are only 2,069 functionallyannotated orthologous groups, showing that they gravely underesti-mate the true size of the common functional complement among indi-viduals (Fig. 6c). The minimal gut metagenome includes a considerablefraction of functions (,45%) that are present in ,10% of thesequenced bacterial genomes (Fig. 6c, inset). These otherwise rare func-tionalities that are found in each of the 124 individuals may be necessaryfor the gut ecosystem. Eighty per cent of these orthologous groupscontain genes with at best poorly characterized function, underscoringour limited knowledge of gut functioning.
Of the known fraction, about 5% codes for (pro)phage-relatedproteins, implying a universal presence and possible important eco-logical role of bacteriophages in gut homeostasis. The most strikingsecondary metabolism that seems crucial for the minimal metage-nome relates, not unexpectedly, to biodegradation of complex sugarsand glycans harvested from the host diet and/or intestinal lining.Examples include degradation and uptake pathways for pectin(and its monomer, rhamnose) and sorbitol, sugars which are omni-present in fruits and vegetables, but which are not or poorly absorbedby humans. As some gut microorganisms were found to degrade bothof them29,30, this capacity seems to be selected for by the gut ecosystemas a non-competitive source of energy. Besides these, capacity toferment, for example, mannose, fructose, cellulose and sucrose is alsopart of the minimal metagenome. Together, these emphasize the
40
30
20
10
0
Clu
ster
(%)
1 2,001 4,001 6,001 8,001 10,001Cluster rank
Range
Figure 5 | Clusters that contain the B. subtilis essential genes. The clusterswere ranked by the number of genes they contain, normalized by averagelength and copy number (see Supplementary Fig. 10), and the proportion ofclusters with the essential B. subtilis genes was determined for successivegroups of 100 clusters. Range indicates the part of the cluster distributionthat contains 86% of the B. subtilis essential genes.
•
•
•
•
•
• •
• •
•
•
•
••
••
• •
•
•
•
•
•
•
•
• •
••
•
•
••
•
•
•
•
•
•
Healthy
Crohn’s disease
Ulcerative colitis
P value: 0.031
PC2
PC1
Figure 4 | Bacterial species abundance differentiates IBD patients andhealthy individuals. Principal component analysis with health status asinstrumental variables, based on the abundance of 155 species with $1%genome coverage by the Illumina reads in at least 1 individual of the cohort,was carried out with 14 healthy individuals and 25 IBD patients (21 ulcerativecolitis and 4 Crohn’s disease) from Spain (Supplementary Table 1). Two firstcomponents (PC1 and PC2) were plotted and represented 7.3% of wholeinertia. Individuals (represented by points) were clustered and centre ofgravity computed for each class; P-value of the link between health status andspecies abundance was assessed using a Monte-Carlo test (999 replicates).
ARTICLES NATURE | Vol 464 | 4 March 2010
62Macmillan Publishers Limited. All rights reserved©2010
!38
Microbiome Forensics
!39
Microbiomes and Plant Health
!40
Model Animal Microbiomes
!4141
Both natural surveys and laboratory experiments indicate that host diet plays a major role in shaping the Drosophila bacterial microbiome.
Laboratory strains provide only a limited model of natural host–microbe interactions
Asthma, Dust, Dogs and Microbiomes
!42
Nice Counter to Germophobia but …
!43
Public Service Reminder
Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation Correlation ≠ causation
!44
Microbiome 101
!45
Methods
!46
Woese: Classification of Cultured Taxa by rRNA
!47
Woese: Classification of Cultured Taxa by rRNA
!47
Woese: Classification of Cultured Taxa by rRNA
!47
Woese: Classification of Cultured Taxa by rRNA
!47
Woese: Classification of Cultured Taxa by rRNA
!47
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
Eukaryotes
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
EukaryotesBacteria
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
EukaryotesBacteria ?????
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
EukaryotesBacteria ?????Archaebacteria
Woese: Classification of Cultured Taxa by rRNA
!47
rRNA rRNArRNA
ACUGC ACCUAU CGUUCG
ACUCC AGCUAU CGAUCG
ACCCC AGCUCU CGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG R ACUCCACCUAUCGUUCG E ACUCCAGCUAUCGAUCG F ACUCCAGGUAUCGAUCG C ACCCCAGCUCUCGCUCG W ACCCCAGCUCUGGCUCG
Taxa Characters S ACUGCACCUAUCGUUCG
E ACUCCAGCUAUCGAUCG
C ACCCCAGCUCUCGCUCG
EukaryotesBacteria ?????ArchaebacteriaArchaea
Culture Independent rRNA PCR: One Taxon
• v
DNA
ACTGC ACCTAT CGTTCG
ACTGC ACCTAT CGTTCG
ACTGC ACCTAT CGTTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACTGCACCTATCGTTCG
EukaryotesBacteria Archaea
!48
Many sequences from one sample all point to the same branch on the tree
DNA
ACTGC ACCTAT CGTTCG
ACTGC ACCTAT CGTTCG
ACCCC AGCTCT CGCTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 ACTGCACCTATCGTTCG
EukaryotesBacteria Archaea
!49
One can estimate cell counts from the number of times each sequence is seen.
Culture Independent rRNA PCR: Two Taxa
DNA
ACTGC ACCTAT CGTTCG
ACTGC ACCTAT CGTTCG
ACCCC AGCTCT CGCTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 ACTGCACCTATCGTTCG
EukaryotesBacteria Archaea
!49
One can estimate cell counts from the number of times each sequence is seen.
Culture Independent rRNA PCR: Two Taxa
DNA
ACTGC ACCTAT CGTTCG
ACTGC ACCTAT CGTTCG
ACCCC AGCTCT CGCTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 ACTGCACCTATCGTTCG
EukaryotesBacteria Archaea
!49
One can estimate cell counts from the number of times each sequence is seen.
Culture Independent rRNA PCR: Two Taxa
DNA
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 AGGGGAGCTCTGCCTCG New3 ACTCCAGCTATCGATCG New4 ACTGCACCTATCGTTCG
EukaryotesBacteria Archaea
!50
ACTGC ACCTAT CGTTCG
ACTCC AGCTAT CGATCG
ACCCC AGCTCT CGCTCG
AGGGG AGCTCT CGCTCG
AGGGG AGCTCT CGCTCG
ACTGC ACCTAT CGTTCG
Even with more taxa it still works
Culture Independent rRNA PCR: Four Taxa
Culture Independent rRNA PCR: Communities
DNA DNADNA
ACTGC ACCTAT CGTTCG
ACTCC AGCTAT CGATCG
ACCCC AGCTCT CGCTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 ACGGCAGCTCTGCCTCG
EukaryotesBacteria Archaea
!51
Culture Independent rRNA PCR: Communities
DNA DNADNA
ACTGC ACCTAT CGTTCG
ACTCC AGCTAT CGATCG
ACCCC AGCTCT CGCTCG
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 ACGGCAGCTCTGCCTCG
!52
Culture Independent “Metagenomics”
DNA DNADNA
!53
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 AGGGGAGCTCTGCCTCG New3 ACTCCAGCTATCGATCG New4 ACTGCACCTATCGTTCG
RecA RecARecA
http://genomebiology.com/2008/9/10/R151 Genome Biology 2008, Volume 9, Issue 10, Article R151 Wu and Eisen R151.7
Genome Biology 2008, 9:R151
sequences are not conserved at the nucleotide level [29]. As a
result, the nr database does not actually contain many more
protein marker sequences that can be used as references than
those available from complete genome sequences.
Comparison of phylogeny-based and similarity-based phylotypingAlthough our phylogeny-based phylotyping is fully auto-
mated, it still requires many more steps than, and is slower
than, similarity based phylotyping methods such as a
MEGAN [30]. Is it worth the trouble? Similarity based phylo-
typing works by searching a query sequence against a refer-
ence database such as NCBI nr and deriving taxonomic
information from the best matches or 'hits'. When species
that are closely related to the query sequence exist in the ref-
erence database, similarity-based phylotyping can work well.
However, if the reference database is a biased sample or if it
contains no closely related species to the query, then the top
hits returned could be misleading [31]. Furthermore, similar-
ity-based methods require an arbitrary similarity cut-off
value to define the top hits. Because individual bacterial
genomes and proteins can evolve at very different rates, a uni-
versal cut-off that works under all conditions does not exist.
As a result, the final results can be very subjective.
In contrast, our tree-based bracketing algorithm places the
query sequence within the context of a phylogenetic tree and
only assigns it to a taxonomic level if that level has adequate
sampling (see Materials and methods [below] for details of
the algorithm). With the well sampled species Prochlorococ-
cus marinus, for example, our method can distinguish closely
related organisms and make taxonomic identifications at the
species level. Our reanalysis of the Sargasso Sea data placed
672 sequences (3.6% of the total) within a P. marinus clade.
On the other hand, for sparsely sampled clades such as
Aquifex, assignments will be made only at the phylum level.
Thus, our phylogeny-based analysis is less susceptible to data
sampling bias than a similarity based approach, and it makes
Major phylotypes identified in Sargasso Sea metagenomic dataFigure 3Major phylotypes identified in Sargasso Sea metagenomic data. The metagenomic data previously obtained from the Sargasso Sea was reanalyzed using AMPHORA and the 31 protein phylogenetic markers. The microbial diversity profiles obtained from individual markers are remarkably consistent. The breakdown of the phylotyping assignments by markers and major taxonomic groups is listed in Additional data file 5.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Alphap
roteo
bacte
ria
Betapr
oteob
acter
ia
Gammap
roteo
bacte
ria
Deltap
roteo
bacte
ria
Epsilo
npro
teoba
cteria
Unclas
sified
prote
obac
teria
Bacter
oidete
s
Chlamyd
iae
Cyano
bacte
ria
Acidob
acter
ia
Therm
otoga
e
Fusob
acter
ia
Actino
bacte
ria
Aquific
ae
Plancto
mycete
s
Spiroc
haete
s
Firmicu
tes
Chloro
flexi
Chloro
bi
Unclas
sified
bacte
ria
dnaGfrrinfCnusApgkpyrGrplArplBrplCrplDrplErplFrplKrplLrplMrplNrplPrplSrplTrpmArpoBrpsBrpsCrpsErpsIrpsJrpsKrpsMrpsSsmpBtsf
Rel
ativ
e ab
unda
nce
RpoB RpoBRpoB
Rpl4 Rpl4Rpl4 rRNA rRNArRNA
Hsp70 Hsp70Hsp70
EFTu EFTuEFTu
Many other genes better than rRNA
Culture Independent “Metagenomics”
DNA DNADNA
!54
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 AGGGGAGCTCTGCCTCG New3 ACTCCAGCTATCGATCG New4 ACTGCACCTATCGTTCG
inputs of fixed carbon or nitrogen from external sources. As withLeptospirillum group I, both Leptospirillum group II and III have thegenes needed to fix carbon by means of the Calvin–Benson–Bassham cycle (using type II ribulose 1,5-bisphosphate carboxy-lase–oxygenase). All genomes recovered from the AMD system
contain formate hydrogenlyase complexes. These, in combinationwith carbon monoxide dehydrogenase, may be used for carbonfixation via the reductive acetyl coenzyme A (acetyl-CoA) pathwayby some, or all, organisms. Given the large number of ABC-typesugar and amino acid transporters encoded in the Ferroplasma type
Figure 4 Cell metabolic cartoons constructed from the annotation of 2,180 ORFs
identified in the Leptospirillum group II genome (63% with putative assigned function) and
1,931 ORFs in the Ferroplasma type II genome (58% with assigned function). The cell
cartoons are shown within a biofilm that is attached to the surface of an acid mine
drainage stream (viewed in cross-section). Tight coupling between ferrous iron oxidation,
pyrite dissolution and acid generation is indicated. Rubisco, ribulose 1,5-bisphosphate
carboxylase–oxygenase. THF, tetrahydrofolate.
articles
NATURE | doi:10.1038/nature02340 | www.nature.com/nature 5© 2004 Nature Publishing Group
Culture Independent “Metagenomics”
DNA DNADNA
!55
Taxa Characters B1 ACTGCACCTATCGTTCG B2 ACTCCACCTATCGTTCG E1 ACTCCAGCTATCGATCG E2 ACTCCAGGTATCGATCG A1 ACCCCAGCTCTCGCTCG A2 ACCCCAGCTCTGGCTCG New1 ACCCCAGCTCTGCCTCG New2 AGGGGAGCTCTGCCTCG New3 ACTCCAGCTATCGATCG New4 ACTGCACCTATCGTTCG
Animal Microbiomes as Ecosystems
!56
Biogeography
!57
Biogeography
!57
a broader range of Proteobacteria, but yielded similar results(Fig. S1 and Tables S2 and S3).Across all samples, we identified 4,931 quality Nitrosomadales
sequences, which grouped into 176 OTUs (operational taxo-nomic units) using an arbitrary 99% sequence similarity cutoff.This cutoff retained a high amount of sequence diversity, butminimized the chance of including diversity because of se-quencing or PCR errors. Most (95%) of the sequences appearclosely related either to the marine Nitrosospira-like clade,known to be abundant in estuarine sediments (e.g., ref. 19) or tomarine bacterium C-17, classified as Nitrosomonas (20) (Fig. S2).Pairwise community similarity between the samples was calcu-lated based on the presence or absence of each OTU usinga rarefied Sørensen’s index (4). Community similarity using thisincidence index was highly correlated with the abundance-basedSørensen index (Mantel test: ρ = 0.9239; P = 0.0001) (21).A plot of community similarity versus geographic distance for
each pairwise set of samples revealed that the Nitrosomonadalesdisplay a significant, negative distance-decay curve (slope = −0.08,P < 0.0001) (Fig. 2). Furthermore, the slope of this curve variedsignificantly among the three spatial scales. The distance-decayslope within marshes was significantly shallower than the overallslope (slope=−0.04;P< 0.0334) and steeper acrossmarsheswithina region than the overall slope (slope= −0.27, P < 0.0007) (Fig. 2).In contrast, at the continental scale, the distance-decay curve didnot differ from zero (P = 0.0953). Thus, there is no evidence thatsampling across continents contributed Nitromonadales OTU di-versity in addition to what was already observed at the marsh andregional scales. Furthermore, additional analyses suggest that theseresults are not driven by a few outlier samples (Fig. S3).Over all spatial scales, both the environment and dispersal lim-
itation appear to influence Nitrosomonadales β-diversity. Rankedpartial Mantel tests revealed that the similarity in Nitrosomo-nadales community composition between samples was highly cor-related with environmental distance (ρ=−0.5339; P=0.0001) andgeographic distance (ρ = −0.2803; P = 0.0001), but not plantcommunity similarity (P = 0.72) (Table S2).To further identify the relative importance of factors con-
tributing to these correlations, we used a multiple regression onmatrices (MRM). The partial regression coefficients of an MRMmodel give a measure of the rate of change in community sim-ilarity per standardized unit of similarity for the variable of in-terest; all other explanatory variables are held constant (22).Over all scales, the MRMmodel explained a large and significantproportion (R2 = 46%; P < 0.0001) of the variability in Nitro-
somonadales community similarity. Geographic distance con-tributed the largest partial regression coefficient (b = 0.40,P < 0.0001), with sediment moisture, nitrate concentration, plantcover, salinity, and air and water temperature contributing tosmaller, but significant, partial regression coefficients (b = 0.09–0.17, P < 0.05) (Table 1). Because salt marsh bacteria may bedispersing through ocean currents, we also used a global oceancirculation model (23), as applied previously (24), to estimaterelative dispersal times of hypothetical microbial cells betweeneach sampling location. Dispersal times between sampling pointsdid not explain more variability in bacterial community similarity(ln dispersal time: b= 0.06, P= −0.0799; with dispersal R2 = 0.47vs. without 0.46). Therefore, in the remaining analyses we usegeographic distance rather than dispersal time.As hypothesized, the relative importance of environmental
factors versus geographic distance to Nitrosomadales communitysimilarity differed across the three spatial scales. Contrary to ourexpectations, however, geographic distance had a strong effecton community similarity within salt marshes (partial regressioncoefficient b = 0.47) but no effect at larger scales (Table 1).Furthermore, the relative importance of different environmentalvariables varied by scale. Sediment moisture, which is likely re-lated to unmeasured variables, such as oxygen availability, wasthe most important variable explaining community similaritywithin marshes (b = 0.63). In contrast, water temperature (b =0.45) and nitrate concentrations (b = 0.17) were more importantat the regional and continental scales, respectively.The varying importance of the environmental parameters at
different spatial scales likely reflects differences in their un-derlying variability at these scales. For example, the MRMmodeldid exceptionally well in explaining variation in Nitrosomadalescommunity similarity at the regional scale (R2 = 0.61) (Table 1).Notably, this spatial scale captures a latitudinal gradient on theeast and west coasts of North America, which results in highvariability in water temperature. Previous studies in the field andlaboratory support the idea that AOB composition is particularlysensitive to temperature (e.g., refs. 25 and 26). Within marshes,
Fig. 1. The 13 marshes sampled (see Table S1 for details). Marshes com-pared with one another within regions are circled. (Inset) The arrangementof sampling points within marshes. Six points were sampled along a 100-mtransect, and a seventh point was sampled ∼1 km away. Two marshes in theNortheast United States (outlined stars) were sampled more intensively,along four 100-m transects in a grid pattern.
Fig. 2. Distance-decay curves for the Nitrosomadales communities. Thedashed, blue line denotes the least-squares linear regression across all spatialscales. The solid lines denote separate regressions within each of the threespatial scales: within marshes, regional (across marshes within regions circled inFig. 1), and continental (across regions). The slopes of all lines (except the solidlight blue line) are significantly less than zero. The slopes of the solid red linesare significantly different from the slope of the all scale (blue dashed) line.
Martiny et al. PNAS | May 10, 2011 | vol. 108 | no. 19 | 7851
ECOLO
GY
a broader range of Proteobacteria, but yielded similar results(Fig. S1 and Tables S2 and S3).Across all samples, we identified 4,931 quality Nitrosomadales
sequences, which grouped into 176 OTUs (operational taxo-nomic units) using an arbitrary 99% sequence similarity cutoff.This cutoff retained a high amount of sequence diversity, butminimized the chance of including diversity because of se-quencing or PCR errors. Most (95%) of the sequences appearclosely related either to the marine Nitrosospira-like clade,known to be abundant in estuarine sediments (e.g., ref. 19) or tomarine bacterium C-17, classified as Nitrosomonas (20) (Fig. S2).Pairwise community similarity between the samples was calcu-lated based on the presence or absence of each OTU usinga rarefied Sørensen’s index (4). Community similarity using thisincidence index was highly correlated with the abundance-basedSørensen index (Mantel test: ρ = 0.9239; P = 0.0001) (21).A plot of community similarity versus geographic distance for
each pairwise set of samples revealed that the Nitrosomonadalesdisplay a significant, negative distance-decay curve (slope = −0.08,P < 0.0001) (Fig. 2). Furthermore, the slope of this curve variedsignificantly among the three spatial scales. The distance-decayslope within marshes was significantly shallower than the overallslope (slope=−0.04;P< 0.0334) and steeper acrossmarsheswithina region than the overall slope (slope= −0.27, P < 0.0007) (Fig. 2).In contrast, at the continental scale, the distance-decay curve didnot differ from zero (P = 0.0953). Thus, there is no evidence thatsampling across continents contributed Nitromonadales OTU di-versity in addition to what was already observed at the marsh andregional scales. Furthermore, additional analyses suggest that theseresults are not driven by a few outlier samples (Fig. S3).Over all spatial scales, both the environment and dispersal lim-
itation appear to influence Nitrosomonadales β-diversity. Rankedpartial Mantel tests revealed that the similarity in Nitrosomo-nadales community composition between samples was highly cor-related with environmental distance (ρ=−0.5339; P=0.0001) andgeographic distance (ρ = −0.2803; P = 0.0001), but not plantcommunity similarity (P = 0.72) (Table S2).To further identify the relative importance of factors con-
tributing to these correlations, we used a multiple regression onmatrices (MRM). The partial regression coefficients of an MRMmodel give a measure of the rate of change in community sim-ilarity per standardized unit of similarity for the variable of in-terest; all other explanatory variables are held constant (22).Over all scales, the MRMmodel explained a large and significantproportion (R2 = 46%; P < 0.0001) of the variability in Nitro-
somonadales community similarity. Geographic distance con-tributed the largest partial regression coefficient (b = 0.40,P < 0.0001), with sediment moisture, nitrate concentration, plantcover, salinity, and air and water temperature contributing tosmaller, but significant, partial regression coefficients (b = 0.09–0.17, P < 0.05) (Table 1). Because salt marsh bacteria may bedispersing through ocean currents, we also used a global oceancirculation model (23), as applied previously (24), to estimaterelative dispersal times of hypothetical microbial cells betweeneach sampling location. Dispersal times between sampling pointsdid not explain more variability in bacterial community similarity(ln dispersal time: b= 0.06, P= −0.0799; with dispersal R2 = 0.47vs. without 0.46). Therefore, in the remaining analyses we usegeographic distance rather than dispersal time.As hypothesized, the relative importance of environmental
factors versus geographic distance to Nitrosomadales communitysimilarity differed across the three spatial scales. Contrary to ourexpectations, however, geographic distance had a strong effecton community similarity within salt marshes (partial regressioncoefficient b = 0.47) but no effect at larger scales (Table 1).Furthermore, the relative importance of different environmentalvariables varied by scale. Sediment moisture, which is likely re-lated to unmeasured variables, such as oxygen availability, wasthe most important variable explaining community similaritywithin marshes (b = 0.63). In contrast, water temperature (b =0.45) and nitrate concentrations (b = 0.17) were more importantat the regional and continental scales, respectively.The varying importance of the environmental parameters at
different spatial scales likely reflects differences in their un-derlying variability at these scales. For example, the MRMmodeldid exceptionally well in explaining variation in Nitrosomadalescommunity similarity at the regional scale (R2 = 0.61) (Table 1).Notably, this spatial scale captures a latitudinal gradient on theeast and west coasts of North America, which results in highvariability in water temperature. Previous studies in the field andlaboratory support the idea that AOB composition is particularlysensitive to temperature (e.g., refs. 25 and 26). Within marshes,
Fig. 1. The 13 marshes sampled (see Table S1 for details). Marshes com-pared with one another within regions are circled. (Inset) The arrangementof sampling points within marshes. Six points were sampled along a 100-mtransect, and a seventh point was sampled ∼1 km away. Two marshes in theNortheast United States (outlined stars) were sampled more intensively,along four 100-m transects in a grid pattern.
Fig. 2. Distance-decay curves for the Nitrosomadales communities. Thedashed, blue line denotes the least-squares linear regression across all spatialscales. The solid lines denote separate regressions within each of the threespatial scales: within marshes, regional (across marshes within regions circled inFig. 1), and continental (across regions). The slopes of all lines (except the solidlight blue line) are significantly less than zero. The slopes of the solid red linesare significantly different from the slope of the all scale (blue dashed) line.
Martiny et al. PNAS | May 10, 2011 | vol. 108 | no. 19 | 7851
ECOLO
GY
Biogeography
!57
a broader range of Proteobacteria, but yielded similar results(Fig. S1 and Tables S2 and S3).Across all samples, we identified 4,931 quality Nitrosomadales
sequences, which grouped into 176 OTUs (operational taxo-nomic units) using an arbitrary 99% sequence similarity cutoff.This cutoff retained a high amount of sequence diversity, butminimized the chance of including diversity because of se-quencing or PCR errors. Most (95%) of the sequences appearclosely related either to the marine Nitrosospira-like clade,known to be abundant in estuarine sediments (e.g., ref. 19) or tomarine bacterium C-17, classified as Nitrosomonas (20) (Fig. S2).Pairwise community similarity between the samples was calcu-lated based on the presence or absence of each OTU usinga rarefied Sørensen’s index (4). Community similarity using thisincidence index was highly correlated with the abundance-basedSørensen index (Mantel test: ρ = 0.9239; P = 0.0001) (21).A plot of community similarity versus geographic distance for
each pairwise set of samples revealed that the Nitrosomonadalesdisplay a significant, negative distance-decay curve (slope = −0.08,P < 0.0001) (Fig. 2). Furthermore, the slope of this curve variedsignificantly among the three spatial scales. The distance-decayslope within marshes was significantly shallower than the overallslope (slope=−0.04;P< 0.0334) and steeper acrossmarsheswithina region than the overall slope (slope= −0.27, P < 0.0007) (Fig. 2).In contrast, at the continental scale, the distance-decay curve didnot differ from zero (P = 0.0953). Thus, there is no evidence thatsampling across continents contributed Nitromonadales OTU di-versity in addition to what was already observed at the marsh andregional scales. Furthermore, additional analyses suggest that theseresults are not driven by a few outlier samples (Fig. S3).Over all spatial scales, both the environment and dispersal lim-
itation appear to influence Nitrosomonadales β-diversity. Rankedpartial Mantel tests revealed that the similarity in Nitrosomo-nadales community composition between samples was highly cor-related with environmental distance (ρ=−0.5339; P=0.0001) andgeographic distance (ρ = −0.2803; P = 0.0001), but not plantcommunity similarity (P = 0.72) (Table S2).To further identify the relative importance of factors con-
tributing to these correlations, we used a multiple regression onmatrices (MRM). The partial regression coefficients of an MRMmodel give a measure of the rate of change in community sim-ilarity per standardized unit of similarity for the variable of in-terest; all other explanatory variables are held constant (22).Over all scales, the MRMmodel explained a large and significantproportion (R2 = 46%; P < 0.0001) of the variability in Nitro-
somonadales community similarity. Geographic distance con-tributed the largest partial regression coefficient (b = 0.40,P < 0.0001), with sediment moisture, nitrate concentration, plantcover, salinity, and air and water temperature contributing tosmaller, but significant, partial regression coefficients (b = 0.09–0.17, P < 0.05) (Table 1). Because salt marsh bacteria may bedispersing through ocean currents, we also used a global oceancirculation model (23), as applied previously (24), to estimaterelative dispersal times of hypothetical microbial cells betweeneach sampling location. Dispersal times between sampling pointsdid not explain more variability in bacterial community similarity(ln dispersal time: b= 0.06, P= −0.0799; with dispersal R2 = 0.47vs. without 0.46). Therefore, in the remaining analyses we usegeographic distance rather than dispersal time.As hypothesized, the relative importance of environmental
factors versus geographic distance to Nitrosomadales communitysimilarity differed across the three spatial scales. Contrary to ourexpectations, however, geographic distance had a strong effecton community similarity within salt marshes (partial regressioncoefficient b = 0.47) but no effect at larger scales (Table 1).Furthermore, the relative importance of different environmentalvariables varied by scale. Sediment moisture, which is likely re-lated to unmeasured variables, such as oxygen availability, wasthe most important variable explaining community similaritywithin marshes (b = 0.63). In contrast, water temperature (b =0.45) and nitrate concentrations (b = 0.17) were more importantat the regional and continental scales, respectively.The varying importance of the environmental parameters at
different spatial scales likely reflects differences in their un-derlying variability at these scales. For example, the MRMmodeldid exceptionally well in explaining variation in Nitrosomadalescommunity similarity at the regional scale (R2 = 0.61) (Table 1).Notably, this spatial scale captures a latitudinal gradient on theeast and west coasts of North America, which results in highvariability in water temperature. Previous studies in the field andlaboratory support the idea that AOB composition is particularlysensitive to temperature (e.g., refs. 25 and 26). Within marshes,
Fig. 1. The 13 marshes sampled (see Table S1 for details). Marshes com-pared with one another within regions are circled. (Inset) The arrangementof sampling points within marshes. Six points were sampled along a 100-mtransect, and a seventh point was sampled ∼1 km away. Two marshes in theNortheast United States (outlined stars) were sampled more intensively,along four 100-m transects in a grid pattern.
Fig. 2. Distance-decay curves for the Nitrosomadales communities. Thedashed, blue line denotes the least-squares linear regression across all spatialscales. The solid lines denote separate regressions within each of the threespatial scales: within marshes, regional (across marshes within regions circled inFig. 1), and continental (across regions). The slopes of all lines (except the solidlight blue line) are significantly less than zero. The slopes of the solid red linesare significantly different from the slope of the all scale (blue dashed) line.
Martiny et al. PNAS | May 10, 2011 | vol. 108 | no. 19 | 7851
ECOLO
GY
a broader range of Proteobacteria, but yielded similar results(Fig. S1 and Tables S2 and S3).Across all samples, we identified 4,931 quality Nitrosomadales
sequences, which grouped into 176 OTUs (operational taxo-nomic units) using an arbitrary 99% sequence similarity cutoff.This cutoff retained a high amount of sequence diversity, butminimized the chance of including diversity because of se-quencing or PCR errors. Most (95%) of the sequences appearclosely related either to the marine Nitrosospira-like clade,known to be abundant in estuarine sediments (e.g., ref. 19) or tomarine bacterium C-17, classified as Nitrosomonas (20) (Fig. S2).Pairwise community similarity between the samples was calcu-lated based on the presence or absence of each OTU usinga rarefied Sørensen’s index (4). Community similarity using thisincidence index was highly correlated with the abundance-basedSørensen index (Mantel test: ρ = 0.9239; P = 0.0001) (21).A plot of community similarity versus geographic distance for
each pairwise set of samples revealed that the Nitrosomonadalesdisplay a significant, negative distance-decay curve (slope = −0.08,P < 0.0001) (Fig. 2). Furthermore, the slope of this curve variedsignificantly among the three spatial scales. The distance-decayslope within marshes was significantly shallower than the overallslope (slope=−0.04;P< 0.0334) and steeper acrossmarsheswithina region than the overall slope (slope= −0.27, P < 0.0007) (Fig. 2).In contrast, at the continental scale, the distance-decay curve didnot differ from zero (P = 0.0953). Thus, there is no evidence thatsampling across continents contributed Nitromonadales OTU di-versity in addition to what was already observed at the marsh andregional scales. Furthermore, additional analyses suggest that theseresults are not driven by a few outlier samples (Fig. S3).Over all spatial scales, both the environment and dispersal lim-
itation appear to influence Nitrosomonadales β-diversity. Rankedpartial Mantel tests revealed that the similarity in Nitrosomo-nadales community composition between samples was highly cor-related with environmental distance (ρ=−0.5339; P=0.0001) andgeographic distance (ρ = −0.2803; P = 0.0001), but not plantcommunity similarity (P = 0.72) (Table S2).To further identify the relative importance of factors con-
tributing to these correlations, we used a multiple regression onmatrices (MRM). The partial regression coefficients of an MRMmodel give a measure of the rate of change in community sim-ilarity per standardized unit of similarity for the variable of in-terest; all other explanatory variables are held constant (22).Over all scales, the MRMmodel explained a large and significantproportion (R2 = 46%; P < 0.0001) of the variability in Nitro-
somonadales community similarity. Geographic distance con-tributed the largest partial regression coefficient (b = 0.40,P < 0.0001), with sediment moisture, nitrate concentration, plantcover, salinity, and air and water temperature contributing tosmaller, but significant, partial regression coefficients (b = 0.09–0.17, P < 0.05) (Table 1). Because salt marsh bacteria may bedispersing through ocean currents, we also used a global oceancirculation model (23), as applied previously (24), to estimaterelative dispersal times of hypothetical microbial cells betweeneach sampling location. Dispersal times between sampling pointsdid not explain more variability in bacterial community similarity(ln dispersal time: b= 0.06, P= −0.0799; with dispersal R2 = 0.47vs. without 0.46). Therefore, in the remaining analyses we usegeographic distance rather than dispersal time.As hypothesized, the relative importance of environmental
factors versus geographic distance to Nitrosomadales communitysimilarity differed across the three spatial scales. Contrary to ourexpectations, however, geographic distance had a strong effecton community similarity within salt marshes (partial regressioncoefficient b = 0.47) but no effect at larger scales (Table 1).Furthermore, the relative importance of different environmentalvariables varied by scale. Sediment moisture, which is likely re-lated to unmeasured variables, such as oxygen availability, wasthe most important variable explaining community similaritywithin marshes (b = 0.63). In contrast, water temperature (b =0.45) and nitrate concentrations (b = 0.17) were more importantat the regional and continental scales, respectively.The varying importance of the environmental parameters at
different spatial scales likely reflects differences in their un-derlying variability at these scales. For example, the MRMmodeldid exceptionally well in explaining variation in Nitrosomadalescommunity similarity at the regional scale (R2 = 0.61) (Table 1).Notably, this spatial scale captures a latitudinal gradient on theeast and west coasts of North America, which results in highvariability in water temperature. Previous studies in the field andlaboratory support the idea that AOB composition is particularlysensitive to temperature (e.g., refs. 25 and 26). Within marshes,
Fig. 1. The 13 marshes sampled (see Table S1 for details). Marshes com-pared with one another within regions are circled. (Inset) The arrangementof sampling points within marshes. Six points were sampled along a 100-mtransect, and a seventh point was sampled ∼1 km away. Two marshes in theNortheast United States (outlined stars) were sampled more intensively,along four 100-m transects in a grid pattern.
Fig. 2. Distance-decay curves for the Nitrosomadales communities. Thedashed, blue line denotes the least-squares linear regression across all spatialscales. The solid lines denote separate regressions within each of the threespatial scales: within marshes, regional (across marshes within regions circled inFig. 1), and continental (across regions). The slopes of all lines (except the solidlight blue line) are significantly less than zero. The slopes of the solid red linesare significantly different from the slope of the all scale (blue dashed) line.
Martiny et al. PNAS | May 10, 2011 | vol. 108 | no. 19 | 7851
ECOLO
GY
!58Huttenhower et al. 2012.
Population Variability
!58Morgan et al. Genome Biology 2012, 13:R79MJ Blaser et al. ISMEJ 2012
US Amerindian
Actinobacteria (Propionibacteria)
Firmicutes (Staphylococcus)
Rela
tive
abun
danc
e
Actinobacteria dominates in the US
Boulder NY Platanillal A Platanillal B
Proteobacteria
Between Countries Age
Vaginal MicrobiomeCorn
at Different
Locations
Individuals
Community Assembly
Community AssemblyFrom Mom
Community AssemblyFrom Mom
Other People
Community AssemblyFrom Mom
From Pets
Other People
Community AssemblyFrom Mom
From Food
From Pets
Other People
Community AssemblyFrom Mom
From Food
From PetsFrom Built
EnvironmentOther People
Disturbance
!60
Disturbance
!60
Disturbance
!60
Disturbance
!60
Switch to solid foods
Disturbance
!60
Switch to solid foods
Disturbance
!60
Switch to solid foods
Disturbance
!60
Switch to solid foods
Captivity and Conservation
!61
1
Research article
Captivity results in disparate loss of gut microbial diversity in closely related hostsKevin D. Kohl1*, Michele M. Skopec2 and M. Denise Dearing1
2
*Corresponding author: +
The gastrointestinal tracts of animals contain diverse communities of microbes that provide a number of services to their hosts. There is recent concern that these communities may be lost as animals enter captive breeding programmes, due to changes in diet and/or exposure to environmental sources. However, empirical evidence documenting the effects of captivity and captive birth on gut communities is lacking. We conducted three studies to advance our knowledge in this area. First, we compared changes in microbial diversity of the gut communities of two species of woodrats (Neotoma albigula, a dietary gen-eralist, and Neotoma stephensi, which specializes on juniper) before and after 6–9 months in captivity. Second, we investi-gated whether reintroduction of the natural diet of N. stephensi could restore microbial diversity. Third, we compared the microbial communities between offspring born in captivity and their mothers. We found that the dietary specialist, N. ste-phensi, lost a greater proportion of its native gut microbiota and overall diversity in response to captivity compared with N. albigula. Addition of the natural diet increased the proportion of the original microbiota but did not restore overall diversity in N. stephensi. Offspring of N. albigula more closely resembled their mothers compared with offspring–mother pairs of N. stephensi. This research suggests that the microbiota of dietary specialists may be more susceptible to captivity. Furthermore, this work highlights the need for further studies investigating the mechanisms underlying how loss of microbial diversity may vary between hosts and what an acceptable level of diversity loss may be to a host. This knowledge will aid conservation biolo-gists in designing captive breeding programmes effective at maintaining microbial diversity.Sequence Accession Numbers: NCBI’s Sequence Read Archive (SRA) – SRP033616
Key words: Neotoma
Editor:
Cite as: Conserv Physiol
IntroductionThe gut microbial communities of animals are hyperdiverse and influence many aspects of their physiology, such as nutri-tion, immune development and even behaviour (Amato, 2013). The preservation of the microbial diversity present in the gut is thought to be critical to the success of their hosts (Redford et al., 2012). For instance, loss of microbial diversity may underlie increased disease prevalence in humans by
resulting in microbial communities that are more susceptible to invasion or by altering host immune function (Blaser and Falkow, 2009). Additionally, gut microbes serve as sources of novel gene products, such as enzymes for biomass degradation (Hess et al., 2011) or bioremediation (Verma et al., 2006).
There is concern that bringing animals into captivity for breeding programmes may result in a loss of microbial diversity, which may contribute to the failures of reintroduced
by guest on Novem
ber 16, 2014http://conphys.oxfordjournals.org/
Dow
nloaded from
Zoos and Shelters
!62
Antimicrobials are in Everything
Restoration
!63
Restoration
!63
Probiotics
Restoration
!63
Animal TransfaunationProbiotics
Restoration
!63
Animal Transfaunation
IlealTransplant
Probiotics
Restoration
!63
Fecal Transplants
Animal Transfaunation
IlealTransplant
Probiotics
History Important Too
History Important Too
Genera that cross the divide. Another way to visualize the vertebrate gut–environment dichotomy is by using a network diagram that displays, in addition to the clus-tering of hosts with similar microbiotas, the bacterial genera they share. In this representation of the data, the vertebrate gut samples are more connected to one another than to the environmental samples (FIG. 4a,b). As in the UniFrac-based analysis, the non-gut human samples also occupy an intermediate position between the free-living and the gut communities. FIGURE 5 shows the phylogenetic classification of operational taxonomic units (OTUs) that are shared between samples: among humans, an over-whelming number of these are from the Firmicutes, with a smaller number from the Bacteroidetes. By contrast, the free-living communities share OTUs from a wider range of phyla. Samples from the guts of obese humans cluster away from the samples of healthy subjects, and most of their shared OTUs are found in the Firmicutes. This obser-vation is consistent with the finding that samples from obese individuals have a higher number of OTUs from Firmicutes than samples from lean subjects31.
Bacterial genera that inhabit both the vertebrate gut-associated microbiotas and the free-living com-munities can be considered to be cosmopolitan. As the analyses discussed above mainly determine the dominant members of a microbiota, these genera are presumed to grow and subsist in the gut environment (autochthonous members) rather than simply passing through as transient members of the gut microbial community (allochthonous members). Among these cosmopolitan groups was the Pseudomonadaceae
family of the gammaproteobacteria class. This fam-ily contained OTUs from both the vertebrate gut and free-living communities in saline and non-saline habitats. Members of the Enterobacteriales order (also from the gammaproteobacteria) were detected in the vertebrate gut, termite gut and other invertebrates, as well as in a surface soil sample and anoxic saline water. Staphylococcaceae family members (from the phylum Firmicutes and class Bacilli) were common in the ver-tebrate gut samples, but were also detected in soil and cultures derived from freshwater and saline habitats. Finally, members of the Fusobacterium genus were detected in salt-water sediments, in addition to the vertebrate gut. The cosmopolitan distribution of these organisms might have made them particularly impor-tant for introducing novel functions during evolution of the gut microbiota, as they could bring new useful genes from the global microbiome into the gut microbiome through horizontal gene transfer. However, it should be noted that some OTUs that are common in humans
Nature Reviews | Microbiology
16S
ribos
omal
RN
A se
quen
ces
(%)
0
20
40
60
80
100
Bacteroidetes (red)
Firmicutes (blue)
Vertebrate
gut
Termite gut
Salt-wate
r surface
Salt wate
r
Subsurface, anoxic or sediment
Other human
Non-saline cultured
Insects or earth
worms
Soils or fr
eshwater se
diments
Mixed wate
r
Figure 3 | Relative abundance of phyla in samples. Bar graph showing the proportion of sequences from each sample that could be classified at the phylum level. The colour codes for the dominant Firmicutes and Bacteroidetes phyla are shown. For a complete description of the colour codes see Supplementary information S2 (figure). ‘Other humans’ refers to body habitats other than the gut; for example, the mouth, ear, skin, vagina and vulva (see Supplementary information S1 (table)).
Figure 4 | Network analysis of bacterial communities from animal-associated and free-living communities. The panel on the left includes a schematic key that illustrates features of the network analysis and genera keys for panels a and b. Labels are sample nodes. Rounded squares represent operational taxonomic units (OTUs) shared by two or more samples (shown in grey in panels a and b), whereas diamonds represent the set of OTUs that are unique to a sample. Network diagrams are colour coded according to habitat.
▶
ANALYSIS
782 | OCTOBER 2008 | VOLUME 6 www.nature.com/reviews/micro
ANALYSIS
Nat Rev Microbiol. 2008 October ; 6(10): 776–788. doi:10.1038/nrmicro1978.
History Important Too
Genera that cross the divide. Another way to visualize the vertebrate gut–environment dichotomy is by using a network diagram that displays, in addition to the clus-tering of hosts with similar microbiotas, the bacterial genera they share. In this representation of the data, the vertebrate gut samples are more connected to one another than to the environmental samples (FIG. 4a,b). As in the UniFrac-based analysis, the non-gut human samples also occupy an intermediate position between the free-living and the gut communities. FIGURE 5 shows the phylogenetic classification of operational taxonomic units (OTUs) that are shared between samples: among humans, an over-whelming number of these are from the Firmicutes, with a smaller number from the Bacteroidetes. By contrast, the free-living communities share OTUs from a wider range of phyla. Samples from the guts of obese humans cluster away from the samples of healthy subjects, and most of their shared OTUs are found in the Firmicutes. This obser-vation is consistent with the finding that samples from obese individuals have a higher number of OTUs from Firmicutes than samples from lean subjects31.
Bacterial genera that inhabit both the vertebrate gut-associated microbiotas and the free-living com-munities can be considered to be cosmopolitan. As the analyses discussed above mainly determine the dominant members of a microbiota, these genera are presumed to grow and subsist in the gut environment (autochthonous members) rather than simply passing through as transient members of the gut microbial community (allochthonous members). Among these cosmopolitan groups was the Pseudomonadaceae
family of the gammaproteobacteria class. This fam-ily contained OTUs from both the vertebrate gut and free-living communities in saline and non-saline habitats. Members of the Enterobacteriales order (also from the gammaproteobacteria) were detected in the vertebrate gut, termite gut and other invertebrates, as well as in a surface soil sample and anoxic saline water. Staphylococcaceae family members (from the phylum Firmicutes and class Bacilli) were common in the ver-tebrate gut samples, but were also detected in soil and cultures derived from freshwater and saline habitats. Finally, members of the Fusobacterium genus were detected in salt-water sediments, in addition to the vertebrate gut. The cosmopolitan distribution of these organisms might have made them particularly impor-tant for introducing novel functions during evolution of the gut microbiota, as they could bring new useful genes from the global microbiome into the gut microbiome through horizontal gene transfer. However, it should be noted that some OTUs that are common in humans
Nature Reviews | Microbiology
16S
ribos
omal
RN
A se
quen
ces
(%)
0
20
40
60
80
100
Bacteroidetes (red)
Firmicutes (blue)
Vertebrate
gut
Termite gut
Salt-wate
r surface
Salt wate
r
Subsurface, anoxic or sediment
Other human
Non-saline cultured
Insects or earth
worms
Soils or fr
eshwater se
diments
Mixed wate
r
Figure 3 | Relative abundance of phyla in samples. Bar graph showing the proportion of sequences from each sample that could be classified at the phylum level. The colour codes for the dominant Firmicutes and Bacteroidetes phyla are shown. For a complete description of the colour codes see Supplementary information S2 (figure). ‘Other humans’ refers to body habitats other than the gut; for example, the mouth, ear, skin, vagina and vulva (see Supplementary information S1 (table)).
Figure 4 | Network analysis of bacterial communities from animal-associated and free-living communities. The panel on the left includes a schematic key that illustrates features of the network analysis and genera keys for panels a and b. Labels are sample nodes. Rounded squares represent operational taxonomic units (OTUs) shared by two or more samples (shown in grey in panels a and b), whereas diamonds represent the set of OTUs that are unique to a sample. Network diagrams are colour coded according to habitat.
▶
ANALYSIS
782 | OCTOBER 2008 | VOLUME 6 www.nature.com/reviews/micro
ANALYSIS
Nat Rev Microbiol. 2008 October ; 6(10): 776–788. doi:10.1038/nrmicro1978.
History Important Too
Genera that cross the divide. Another way to visualize the vertebrate gut–environment dichotomy is by using a network diagram that displays, in addition to the clus-tering of hosts with similar microbiotas, the bacterial genera they share. In this representation of the data, the vertebrate gut samples are more connected to one another than to the environmental samples (FIG. 4a,b). As in the UniFrac-based analysis, the non-gut human samples also occupy an intermediate position between the free-living and the gut communities. FIGURE 5 shows the phylogenetic classification of operational taxonomic units (OTUs) that are shared between samples: among humans, an over-whelming number of these are from the Firmicutes, with a smaller number from the Bacteroidetes. By contrast, the free-living communities share OTUs from a wider range of phyla. Samples from the guts of obese humans cluster away from the samples of healthy subjects, and most of their shared OTUs are found in the Firmicutes. This obser-vation is consistent with the finding that samples from obese individuals have a higher number of OTUs from Firmicutes than samples from lean subjects31.
Bacterial genera that inhabit both the vertebrate gut-associated microbiotas and the free-living com-munities can be considered to be cosmopolitan. As the analyses discussed above mainly determine the dominant members of a microbiota, these genera are presumed to grow and subsist in the gut environment (autochthonous members) rather than simply passing through as transient members of the gut microbial community (allochthonous members). Among these cosmopolitan groups was the Pseudomonadaceae
family of the gammaproteobacteria class. This fam-ily contained OTUs from both the vertebrate gut and free-living communities in saline and non-saline habitats. Members of the Enterobacteriales order (also from the gammaproteobacteria) were detected in the vertebrate gut, termite gut and other invertebrates, as well as in a surface soil sample and anoxic saline water. Staphylococcaceae family members (from the phylum Firmicutes and class Bacilli) were common in the ver-tebrate gut samples, but were also detected in soil and cultures derived from freshwater and saline habitats. Finally, members of the Fusobacterium genus were detected in salt-water sediments, in addition to the vertebrate gut. The cosmopolitan distribution of these organisms might have made them particularly impor-tant for introducing novel functions during evolution of the gut microbiota, as they could bring new useful genes from the global microbiome into the gut microbiome through horizontal gene transfer. However, it should be noted that some OTUs that are common in humans
Nature Reviews | Microbiology
16S
ribos
omal
RN
A se
quen
ces
(%)
0
20
40
60
80
100
Bacteroidetes (red)
Firmicutes (blue)
Vertebrate
gut
Termite gut
Salt-wate
r surface
Salt wate
r
Subsurface, anoxic or sediment
Other human
Non-saline cultured
Insects or earth
worms
Soils or fr
eshwater se
diments
Mixed wate
r
Figure 3 | Relative abundance of phyla in samples. Bar graph showing the proportion of sequences from each sample that could be classified at the phylum level. The colour codes for the dominant Firmicutes and Bacteroidetes phyla are shown. For a complete description of the colour codes see Supplementary information S2 (figure). ‘Other humans’ refers to body habitats other than the gut; for example, the mouth, ear, skin, vagina and vulva (see Supplementary information S1 (table)).
Figure 4 | Network analysis of bacterial communities from animal-associated and free-living communities. The panel on the left includes a schematic key that illustrates features of the network analysis and genera keys for panels a and b. Labels are sample nodes. Rounded squares represent operational taxonomic units (OTUs) shared by two or more samples (shown in grey in panels a and b), whereas diamonds represent the set of OTUs that are unique to a sample. Network diagrams are colour coded according to habitat.
▶
ANALYSIS
782 | OCTOBER 2008 | VOLUME 6 www.nature.com/reviews/micro
ANALYSIS
Nat Rev Microbiol. 2008 October ; 6(10): 776–788. doi:10.1038/nrmicro1978.
History Important Too
Genera that cross the divide. Another way to visualize the vertebrate gut–environment dichotomy is by using a network diagram that displays, in addition to the clus-tering of hosts with similar microbiotas, the bacterial genera they share. In this representation of the data, the vertebrate gut samples are more connected to one another than to the environmental samples (FIG. 4a,b). As in the UniFrac-based analysis, the non-gut human samples also occupy an intermediate position between the free-living and the gut communities. FIGURE 5 shows the phylogenetic classification of operational taxonomic units (OTUs) that are shared between samples: among humans, an over-whelming number of these are from the Firmicutes, with a smaller number from the Bacteroidetes. By contrast, the free-living communities share OTUs from a wider range of phyla. Samples from the guts of obese humans cluster away from the samples of healthy subjects, and most of their shared OTUs are found in the Firmicutes. This obser-vation is consistent with the finding that samples from obese individuals have a higher number of OTUs from Firmicutes than samples from lean subjects31.
Bacterial genera that inhabit both the vertebrate gut-associated microbiotas and the free-living com-munities can be considered to be cosmopolitan. As the analyses discussed above mainly determine the dominant members of a microbiota, these genera are presumed to grow and subsist in the gut environment (autochthonous members) rather than simply passing through as transient members of the gut microbial community (allochthonous members). Among these cosmopolitan groups was the Pseudomonadaceae
family of the gammaproteobacteria class. This fam-ily contained OTUs from both the vertebrate gut and free-living communities in saline and non-saline habitats. Members of the Enterobacteriales order (also from the gammaproteobacteria) were detected in the vertebrate gut, termite gut and other invertebrates, as well as in a surface soil sample and anoxic saline water. Staphylococcaceae family members (from the phylum Firmicutes and class Bacilli) were common in the ver-tebrate gut samples, but were also detected in soil and cultures derived from freshwater and saline habitats. Finally, members of the Fusobacterium genus were detected in salt-water sediments, in addition to the vertebrate gut. The cosmopolitan distribution of these organisms might have made them particularly impor-tant for introducing novel functions during evolution of the gut microbiota, as they could bring new useful genes from the global microbiome into the gut microbiome through horizontal gene transfer. However, it should be noted that some OTUs that are common in humans
Nature Reviews | Microbiology
16S
ribos
omal
RN
A se
quen
ces
(%)
0
20
40
60
80
100
Bacteroidetes (red)
Firmicutes (blue)
Vertebrate
gut
Termite gut
Salt-wate
r surface
Salt wate
r
Subsurface, anoxic or sediment
Other human
Non-saline cultured
Insects or earth
worms
Soils or fr
eshwater se
diments
Mixed wate
r
Figure 3 | Relative abundance of phyla in samples. Bar graph showing the proportion of sequences from each sample that could be classified at the phylum level. The colour codes for the dominant Firmicutes and Bacteroidetes phyla are shown. For a complete description of the colour codes see Supplementary information S2 (figure). ‘Other humans’ refers to body habitats other than the gut; for example, the mouth, ear, skin, vagina and vulva (see Supplementary information S1 (table)).
Figure 4 | Network analysis of bacterial communities from animal-associated and free-living communities. The panel on the left includes a schematic key that illustrates features of the network analysis and genera keys for panels a and b. Labels are sample nodes. Rounded squares represent operational taxonomic units (OTUs) shared by two or more samples (shown in grey in panels a and b), whereas diamonds represent the set of OTUs that are unique to a sample. Network diagrams are colour coded according to habitat.
▶
ANALYSIS
782 | OCTOBER 2008 | VOLUME 6 www.nature.com/reviews/micro
ANALYSIS
Nat Rev Microbiol. 2008 October ; 6(10): 776–788. doi:10.1038/nrmicro1978.
Example: Behavior
!65
12 OCTOBER 2012 VOL 338 SCIENCE www.sciencemag.org 198
PERSPECTIVES
Human bodies house trillions of sym-biotic microorganisms. The genes in this human microbiome outnum-
ber human genes by 100 to 1, and their study is providing profound insights into human health. But humans are not the only ani-mals with microbiomes, and microbiomes do not just impact health. Recent research is revealing surprising roles for microbiomes in shaping behaviors across many animal taxa—shedding light on how behaviors from diet to social interactions affect the compo-sition of host-associated microbial commu-nities ( 1, 2), and how microbes in turn infl u-ence host behavior in dramatic ways ( 2– 6).
Our understanding of interactions between host behavior and microbes stems largely from studies of pathogens. Animal social and mating activities have profound effects on pathogen transmission, and many animals use behavioral strategies to avoid or remove pathogens ( 7). Pathogens can also manipulate host behavior in overt or covert ways. However, given the diversity of microbes in nature, it is important to expand the view of behavior-microbe interactions to include nonpathogens.
For diverse animals, including iguanas, squids, and many insects, behavior plays a central role in the establishment and regula-tion of microbial associations (see the fi rst fi gure). For example, the Kudzu bug (Mega-
copta cribraria), an agricultural pest, is born without any symbionts. After birth it acquires a specifi c symbiont from bacterial capsules left by its mother. If these capsules are removed, the bugs show dramatic wan-dering behaviors, presumably to search for symbiont capsules left with nearby eggs ( 8).
Social contact is another mechanism that can mediate the acquisition and exchange of microbial symbionts. Indeed, a bene-
fi t of social living in many species may be the transmission of benefi cial microbes ( 9). Koch and Schmid-Hempel have shown that in the case of bumble bees (Bombus terres-
tris), either direct contact with nest mates or feeding on feces of nest mates was neces-sary for establishing the normal gut micro-biota. Bees never exposed to feces had an altered gut microbiota and were more sus-ceptible to the parasite Crithidia bombi ( 1).
Social context also shapes establishment of mammalian microbial associations. For example, chimpanzees from the same com-munity have more similar microbial consor-tia than do chimpanzees from different com-munities ( 10).
Yet, despite these and other examples of behavior facilitating microbial colonization, questions remain. To what extent is juvenile behavior driven by the search for benefi cial microbes? How frequently does host choice infl uence the acquisition of microbial part-ners? And, if there is strong selection for ani-mals to acquire microbes from each other, what role do benefi cial microbes play in the evolution of sociality?
Once host-microbe associations are established, microbes can influence host behavior in ways that have far-reaching implications for host ecology and evolution (see the second fi gure). Sharon et al. recently found that fruit fl ies (Drosophila melano-
gaster) strongly prefer to mate with individ-uals reared on the same diet on which they were reared. Antibiotic treatment abolished the mating preference, and inoculation of treated fl ies with microbes from the dietary media restored the preference, indicating that microbes, and not diet, altered mate choice. Changes in presence of one bacte-rium, Lactobacillus plantarum, were linked to the induction of mating preferences ( 2). Flies reared on different diets showed dif-ferences in major cuticular hydrocarbons, which are known to infl uence mating, sug-gesting that the bacteria alter these crucial chemical signals. Similarly, communities of scent gland–inhabiting odor-producing bacteria vary across hyena clans ( 3), sug-gesting that microbes could fundamentally alter social interactions in these animals via effects on their chemical communication.
Animal Behavior and the Microbiome
MICROBIOLOGY
Vanessa O. Ezenwa 1, Nicole M. Gerardo 2, David W. Inouye 3 ,4, Mónica Medina 5, Joao B. Xavier 6
Feedbacks between microbiomes and their
hosts affect a range of animal behaviors.
Gut microbiota
Behaviors im
pact m
icrobiom
es
Juvenile iguanas eat soil
or feces to tailor the
microbiota to their current
diet
Animals may adjust
the microbiota at
different life-history
stages
Ishikawaella
capsulata
When born, bugs feed on
capsules of symbionts; if no
capsules are present, nymphs
wander in search of microbes
Behaviors shape
symbiont acquisition
Vibrio fischeri
Squids eject bioluminescent
bacteria daily
Suggests animals
can actively control
their symbiont
populations
Green iguana
(Iguana iguana)
Bobtail squid
(Euprymna scolopes)
Kudzu bug
(Megacopta cribraria)
Animal Implication
Microbial species
or consortium
Interaction with
behavior
Behaviors alter microbiomes. In Kudzu bugs ( 8), green iguanas ( 15), and bobtail squid ( 16), host behaviors alter microbial acquisition and maintenance.
1Odum School of Ecology and Department of Infectious Diseases, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA. 2Department of Biology, Emory University, Atlanta, GA 30322, USA. 3Rocky Moun-tain Biological Laboratory, Crested Butte, CO 81224, USA. 4Department of Biology, University of Maryland, College Park, MD 20742, USA. 5School of Natural Sciences, Univer-sity of California Merced, Merced, CA 95343, USA. 6Program in Computational Biology, Memorial Sloan-Kettering Can-cer Center, New York, NY 10021, USA. All authors contrib-uted equally. E-mail: [email protected] P
HO
TO
CR
ED
ITS
FIR
ST
FIG
UR
E: (IG
UA
NA
) B
. W
EH
RLE
/CA
LIF
OR
NIA
STA
TE
UN
IVE
RSIT
Y, N
OR
TH
RID
GE
; (S
QU
ID) W
. O
RM
ER
OD
, C
OU
RT
ESY
OF
M. M
CFA
LL-N
GA
I/U
NIV
ER
SIT
Y O
F W
ISC
ON
SIN
; (B
UG
) N
. G
ER
AR
DO
/EM
OR
Y U
NIV
ER
SIT
Y
Published by AAAS
on
Nov
embe
r 21,
201
2w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from
12 OCTOBER 2012 VOL 338 SCIENCE www.sciencemag.org 198
PERSPECTIVES
Human bodies house trillions of sym-biotic microorganisms. The genes in this human microbiome outnum-
ber human genes by 100 to 1, and their study is providing profound insights into human health. But humans are not the only ani-mals with microbiomes, and microbiomes do not just impact health. Recent research is revealing surprising roles for microbiomes in shaping behaviors across many animal taxa—shedding light on how behaviors from diet to social interactions affect the compo-sition of host-associated microbial commu-nities ( 1, 2), and how microbes in turn infl u-ence host behavior in dramatic ways ( 2– 6).
Our understanding of interactions between host behavior and microbes stems largely from studies of pathogens. Animal social and mating activities have profound effects on pathogen transmission, and many animals use behavioral strategies to avoid or remove pathogens ( 7). Pathogens can also manipulate host behavior in overt or covert ways. However, given the diversity of microbes in nature, it is important to expand the view of behavior-microbe interactions to include nonpathogens.
For diverse animals, including iguanas, squids, and many insects, behavior plays a central role in the establishment and regula-tion of microbial associations (see the fi rst fi gure). For example, the Kudzu bug (Mega-
copta cribraria), an agricultural pest, is born without any symbionts. After birth it acquires a specifi c symbiont from bacterial capsules left by its mother. If these capsules are removed, the bugs show dramatic wan-dering behaviors, presumably to search for symbiont capsules left with nearby eggs ( 8).
Social contact is another mechanism that can mediate the acquisition and exchange of microbial symbionts. Indeed, a bene-
fi t of social living in many species may be the transmission of benefi cial microbes ( 9). Koch and Schmid-Hempel have shown that in the case of bumble bees (Bombus terres-
tris), either direct contact with nest mates or feeding on feces of nest mates was neces-sary for establishing the normal gut micro-biota. Bees never exposed to feces had an altered gut microbiota and were more sus-ceptible to the parasite Crithidia bombi ( 1).
Social context also shapes establishment of mammalian microbial associations. For example, chimpanzees from the same com-munity have more similar microbial consor-tia than do chimpanzees from different com-munities ( 10).
Yet, despite these and other examples of behavior facilitating microbial colonization, questions remain. To what extent is juvenile behavior driven by the search for benefi cial microbes? How frequently does host choice infl uence the acquisition of microbial part-ners? And, if there is strong selection for ani-mals to acquire microbes from each other, what role do benefi cial microbes play in the evolution of sociality?
Once host-microbe associations are established, microbes can influence host behavior in ways that have far-reaching implications for host ecology and evolution (see the second fi gure). Sharon et al. recently found that fruit fl ies (Drosophila melano-
gaster) strongly prefer to mate with individ-uals reared on the same diet on which they were reared. Antibiotic treatment abolished the mating preference, and inoculation of treated fl ies with microbes from the dietary media restored the preference, indicating that microbes, and not diet, altered mate choice. Changes in presence of one bacte-rium, Lactobacillus plantarum, were linked to the induction of mating preferences ( 2). Flies reared on different diets showed dif-ferences in major cuticular hydrocarbons, which are known to infl uence mating, sug-gesting that the bacteria alter these crucial chemical signals. Similarly, communities of scent gland–inhabiting odor-producing bacteria vary across hyena clans ( 3), sug-gesting that microbes could fundamentally alter social interactions in these animals via effects on their chemical communication.
Animal Behavior and the Microbiome
MICROBIOLOGY
Vanessa O. Ezenwa 1, Nicole M. Gerardo 2, David W. Inouye 3 ,4, Mónica Medina 5, Joao B. Xavier 6
Feedbacks between microbiomes and their
hosts affect a range of animal behaviors.
Gut microbiota
Behaviors im
pact m
icrobiom
es
Juvenile iguanas eat soil
or feces to tailor the
microbiota to their current
diet
Animals may adjust
the microbiota at
different life-history
stages
Ishikawaella
capsulata
When born, bugs feed on
capsules of symbionts; if no
capsules are present, nymphs
wander in search of microbes
Behaviors shape
symbiont acquisition
Vibrio fischeri
Squids eject bioluminescent
bacteria daily
Suggests animals
can actively control
their symbiont
populations
Green iguana
(Iguana iguana)
Bobtail squid
(Euprymna scolopes)
Kudzu bug
(Megacopta cribraria)
Animal Implication
Microbial species
or consortium
Interaction with
behavior
Behaviors alter microbiomes. In Kudzu bugs ( 8), green iguanas ( 15), and bobtail squid ( 16), host behaviors alter microbial acquisition and maintenance.
1Odum School of Ecology and Department of Infectious Diseases, College of Veterinary Medicine, University of Georgia, Athens, GA 30602, USA. 2Department of Biology, Emory University, Atlanta, GA 30322, USA. 3Rocky Moun-tain Biological Laboratory, Crested Butte, CO 81224, USA. 4Department of Biology, University of Maryland, College Park, MD 20742, USA. 5School of Natural Sciences, Univer-sity of California Merced, Merced, CA 95343, USA. 6Program in Computational Biology, Memorial Sloan-Kettering Can-cer Center, New York, NY 10021, USA. All authors contrib-uted equally. E-mail: [email protected] P
HO
TO
CR
ED
ITS
FIR
ST
FIG
UR
E: (IG
UA
NA
) B
. W
EH
RLE
/CA
LIF
OR
NIA
STA
TE
UN
IVE
RS
ITY,
NO
RT
HR
IDG
E; (S
QU
ID) W
. O
RM
ER
OD
, C
OU
RT
ES
Y O
F M
. M
CFA
LL-N
GA
I/U
NIV
ER
SIT
Y O
F W
ISC
ON
SIN
; (B
UG
) N
. G
ER
AR
DO
/EM
OR
Y U
NIV
ER
SIT
Y
Published by AAAS
on
Nov
embe
r 21,
201
2w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from
www.sciencemag.org SCIENCE VOL 338 12 OCTOBER 2012 199
PERSPECTIVES
Microbial effects on animal chemistry
also recently have been linked to changes
in predator-prey interactions ( 11) and feed-
ing behavior ( 12). Females of the African
malaria mosquito, Anopheles gambiae, use
chemical cues released from human skin
to locate hosts. By analyzing skin emana-
tions from 48 subjects, Verhulst et al. ( 12)
found that humans with higher microbial
diversity on their skin were less attractive
to these mosquitoes. High abundances of
Pseudomonas spp. and Variovorax spp.
were also associated with poor attractive-
ness to A. gambiae. These bacteria may pro-
duce chemicals that repel mosquitoes or
mask attractive volatiles emanating from
human skin. Given the importance of chem-
ical communication throughout the animal
kingdom, symbiont alteration of host chem-
istry may be a potent force that shapes many
fundamental animal behaviors.
Animal microbiomes often consist of
thousands of species of bacteria, many
of which cannot be cultivated outside the
host. Rapid advances in metagenomics are
allowing characterization of microbiomes
beyond the few cultivable microbes ( 10, 13,
14). However, determining which animal
behaviors infl uence and are infl uenced by
microbial symbionts, and the mechanisms
underlying these interactions, will require
a combination of molecular and experimen-
tal approaches. For example, Huang et al.
have studied the settlement behavior in the
marine tubeworm Hydroides elegans. Bac-
terial biofi lms play a key role in the settle-
ment behavior of many marine inverte-
brates, from corals to sea urchins. To study
the H. elegans system, the authors used
transposon mutagenesis to knock out a num-
ber of genes from the bacterium Pseudoal-
teromonas luteoviolacea, which is required
for larval settlement. Mutagenesis of four
genes related to cell adhesion and secretion
generated bacterial strains that altered worm
settlement behavior and metamorphosis ( 4).
It remains to be shown whether similar bac-
terial phenotypes drive this important life-
history transition across metazoans.
Some animal behaviors will be linked
to single microbial species, but many will
involve communities of multiple micro-
bial species. It is unclear how fl uctuations
in the microbiome throughout the host life
cycle drive behavioral traits and vice versa.
Another challenge is to identify when
behavior shapes the microbiome, when the
microbiome shapes behavior, and when
there is a complex feedback between the
two. This requires manipulative experiments
and will be facilitated by studying the under-
lying mechanisms by which signals are sent
between hosts and microbes.
Recent experiments with mice, showing
that the gut microbiome can infl uence stress,
anxiety, and depression-related behavior via
effects on the host’s neuroendrocrine sys-
tem, provide insight into how information
can be passed between a host and microbe
( 5, 6). Mice fed with the probiotic Lacto-
baccillus rhamnosus fared better in a forced
swim test, an indication of lower anxiety, and
showed higher expression of γ-aminobutyric
acid receptors in the brain. Partial removal
of the vagus, a central communication nerve
between gut and brain, obliterated the probi-
otic effect, suggesting that this nerve trans-
mits information on gut bacteria to the brain
( 5). If benefi cial microbes are also found to
modify neural and endocrine activity in the
brain in other animals, then they have enor-
mous potential to influence how animals
behave toward one another.
Experimental approaches that evaluate
the behavioral consequences of microbiome
manipulation will be key to addressing out-
standing questions. Similarly, studies that
manipulate animal behavior—for example,
by swapping social or mating partners or by
enhancing or blocking neuroendocrine func-
tion—can be used to identify behaviors that
alter microbiomes. Through either approach,
the interface between the fi elds of microbi-
ology and behavior is poised to expand our
understanding of complex microbial com-
munities already known to shape animal
nutrition and health ( 13, 14) and to unveil a
hidden dimension of animal behavior.
References and Notes
1. H. Koch, P. Schmid-Hempel, Proc. Natl. Acad. Sci. U.S.A. 108, 19288 (2011).
2. G. Sharon et al., Proc. Natl. Acad. Sci. U.S.A. 107, 20051 (2010).
3. K. R. Theis, T. M. Schmidt, K. E. Holekamp, Sci. Rep. 2, 615 (2012).
4. Y. Huang, S. Callahan, M. G. Hadfi eld, Sci. Rep. 2, 228 (2012).
5. J. A. Bravo et al., Proc. Natl. Acad. Sci. U.S.A. 108, 16050 (2011).
6. R. Diaz, Heijtz et al., Proc. Natl. Acad. Sci. U.S.A. 108, 3047 (2011).
7. S. Altizer et al., Annu. Rev. Ecol. Evol. Syst. 34, 517 (2003).
8. T. Hosokawa, Y. Kikuchi, M. Shimada, T. Fukatsu, Biol.
Lett. 4, 45 (2008). 9. M. P. Lombardo, Behav. Ecol. Sociobiol. 62, 479 (2008). 10. P. H. Degnan et al., Proc. Natl. Acad. Sci. U.S.A. 109,
13034 (2012). 11. K. M. Oliver et al., BMC Biol. 10, 11 (2012). 12. N. O. Verhulst et al., PLoS ONE 6, e28991 (2011). 13. P. Engel, V. G. Martinson, N. A. Moran, Proc. Natl. Acad.
Sci. U.S.A. 109, 11002 (2012). 14. C. Huttenhower et al., Nature 486, 207 (2012). 15. K. Troyer, Behav. Ecol. Sociobiol. 14, 189 (1984). 16. K. J. Boettcher, E. G. Ruby, M. J. McFall-Ngai, J. Comp.
Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 179, 65 (1996).
Acknowledgments: This perspective was made possible thanks to NSF meeting grant IOS 1229439 “Meeting: The Future of Research in Animal Behavior.” We thank B. Parker, A. Laughton, B. Wehrle, C. Fontaine, and D. Ditmarsch for discussion and comments.
Animal
Microbiom
es im
pact behaviors
Implication
Microbial species
or consortium
Interaction with
behavior
Human skinmicrobiota
Skin microbes of humans influence attraction to mosquitoes
Differential attraction could impact disease spread
Lactobacillus
rhamnosus
The probiotic L. rhamnosus
decreases anxiety in miceSuggests bacteria can alter mood
Gut microbiota Diet-specific microbiota influence mating preferences
Microbes could drive speciation
Mosquito
(Anopheles gambiae)
Mouse
(Mus musculus)
Fruit fly
(Drosophila melanogaster)
Microbiomes alter behaviors. In fruit fl ies ( 2), mosquitoes ( 12), and mice ( 5, 6), microbes alter mating, feeding, and anxiety levels.
PH
OT
O C
RE
DIT
S S
EC
ON
D F
IGU
RE
: (M
OS
QU
ITO
) J.
GA
TH
AN
Y/C
EN
TE
R F
OR
DIS
EA
SE
CO
NT
RO
L; (M
OU
SE
) G
. S
HU
KLIN
/WIK
IME
DIA
CO
MM
ON
S; (F
LIE
S) T.
CH
AP
MA
N/U
NIV
ER
SIT
Y O
F E
AS
T A
NG
LIA
10.1126/science.1227412
Published by AAAS
on
Nov
embe
r 21,
201
2w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from
www.sciencemag.org SCIENCE VOL 338 12 OCTOBER 2012 199
PERSPECTIVES
Microbial effects on animal chemistry
also recently have been linked to changes
in predator-prey interactions ( 11) and feed-
ing behavior ( 12). Females of the African
malaria mosquito, Anopheles gambiae, use
chemical cues released from human skin
to locate hosts. By analyzing skin emana-
tions from 48 subjects, Verhulst et al. ( 12)
found that humans with higher microbial
diversity on their skin were less attractive
to these mosquitoes. High abundances of
Pseudomonas spp. and Variovorax spp.
were also associated with poor attractive-
ness to A. gambiae. These bacteria may pro-
duce chemicals that repel mosquitoes or
mask attractive volatiles emanating from
human skin. Given the importance of chem-
ical communication throughout the animal
kingdom, symbiont alteration of host chem-
istry may be a potent force that shapes many
fundamental animal behaviors.
Animal microbiomes often consist of
thousands of species of bacteria, many
of which cannot be cultivated outside the
host. Rapid advances in metagenomics are
allowing characterization of microbiomes
beyond the few cultivable microbes ( 10, 13,
14). However, determining which animal
behaviors infl uence and are infl uenced by
microbial symbionts, and the mechanisms
underlying these interactions, will require
a combination of molecular and experimen-
tal approaches. For example, Huang et al.
have studied the settlement behavior in the
marine tubeworm Hydroides elegans. Bac-
terial biofi lms play a key role in the settle-
ment behavior of many marine inverte-
brates, from corals to sea urchins. To study
the H. elegans system, the authors used
transposon mutagenesis to knock out a num-
ber of genes from the bacterium Pseudoal-
teromonas luteoviolacea, which is required
for larval settlement. Mutagenesis of four
genes related to cell adhesion and secretion
generated bacterial strains that altered worm
settlement behavior and metamorphosis ( 4).
It remains to be shown whether similar bac-
terial phenotypes drive this important life-
history transition across metazoans.
Some animal behaviors will be linked
to single microbial species, but many will
involve communities of multiple micro-
bial species. It is unclear how fl uctuations
in the microbiome throughout the host life
cycle drive behavioral traits and vice versa.
Another challenge is to identify when
behavior shapes the microbiome, when the
microbiome shapes behavior, and when
there is a complex feedback between the
two. This requires manipulative experiments
and will be facilitated by studying the under-
lying mechanisms by which signals are sent
between hosts and microbes.
Recent experiments with mice, showing
that the gut microbiome can infl uence stress,
anxiety, and depression-related behavior via
effects on the host’s neuroendrocrine sys-
tem, provide insight into how information
can be passed between a host and microbe
( 5, 6). Mice fed with the probiotic Lacto-
baccillus rhamnosus fared better in a forced
swim test, an indication of lower anxiety, and
showed higher expression of γ-aminobutyric
acid receptors in the brain. Partial removal
of the vagus, a central communication nerve
between gut and brain, obliterated the probi-
otic effect, suggesting that this nerve trans-
mits information on gut bacteria to the brain
( 5). If benefi cial microbes are also found to
modify neural and endocrine activity in the
brain in other animals, then they have enor-
mous potential to influence how animals
behave toward one another.
Experimental approaches that evaluate
the behavioral consequences of microbiome
manipulation will be key to addressing out-
standing questions. Similarly, studies that
manipulate animal behavior—for example,
by swapping social or mating partners or by
enhancing or blocking neuroendocrine func-
tion—can be used to identify behaviors that
alter microbiomes. Through either approach,
the interface between the fi elds of microbi-
ology and behavior is poised to expand our
understanding of complex microbial com-
munities already known to shape animal
nutrition and health ( 13, 14) and to unveil a
hidden dimension of animal behavior.
References and Notes
1. H. Koch, P. Schmid-Hempel, Proc. Natl. Acad. Sci. U.S.A. 108, 19288 (2011).
2. G. Sharon et al., Proc. Natl. Acad. Sci. U.S.A. 107, 20051 (2010).
3. K. R. Theis, T. M. Schmidt, K. E. Holekamp, Sci. Rep. 2, 615 (2012).
4. Y. Huang, S. Callahan, M. G. Hadfi eld, Sci. Rep. 2, 228 (2012).
5. J. A. Bravo et al., Proc. Natl. Acad. Sci. U.S.A. 108, 16050 (2011).
6. R. Diaz, Heijtz et al., Proc. Natl. Acad. Sci. U.S.A. 108, 3047 (2011).
7. S. Altizer et al., Annu. Rev. Ecol. Evol. Syst. 34, 517 (2003).
8. T. Hosokawa, Y. Kikuchi, M. Shimada, T. Fukatsu, Biol.
Lett. 4, 45 (2008). 9. M. P. Lombardo, Behav. Ecol. Sociobiol. 62, 479 (2008). 10. P. H. Degnan et al., Proc. Natl. Acad. Sci. U.S.A. 109,
13034 (2012). 11. K. M. Oliver et al., BMC Biol. 10, 11 (2012). 12. N. O. Verhulst et al., PLoS ONE 6, e28991 (2011). 13. P. Engel, V. G. Martinson, N. A. Moran, Proc. Natl. Acad.
Sci. U.S.A. 109, 11002 (2012). 14. C. Huttenhower et al., Nature 486, 207 (2012). 15. K. Troyer, Behav. Ecol. Sociobiol. 14, 189 (1984). 16. K. J. Boettcher, E. G. Ruby, M. J. McFall-Ngai, J. Comp.
Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 179, 65 (1996).
Acknowledgments: This perspective was made possible thanks to NSF meeting grant IOS 1229439 “Meeting: The Future of Research in Animal Behavior.” We thank B. Parker, A. Laughton, B. Wehrle, C. Fontaine, and D. Ditmarsch for discussion and comments.
Animal
Microbiom
es im
pact behaviors
Implication
Microbial species
or consortium
Interaction with
behavior
Human skinmicrobiota
Skin microbes of humans influence attraction to mosquitoes
Differential attraction could impact disease spread
Lactobacillus
rhamnosus
The probiotic L. rhamnosus
decreases anxiety in miceSuggests bacteria can alter mood
Gut microbiota Diet-specific microbiota influence mating preferences
Microbes could drive speciation
Mosquito
(Anopheles gambiae)
Mouse
(Mus musculus)
Fruit fly
(Drosophila melanogaster)
Microbiomes alter behaviors. In fruit fl ies ( 2), mosquitoes ( 12), and mice ( 5, 6), microbes alter mating, feeding, and anxiety levels.
PH
OT
O C
RE
DIT
S S
EC
ON
D F
IGU
RE
: (M
OSQ
UIT
O) J.
GA
TH
AN
Y/C
EN
TE
R F
OR
DIS
EA
SE
CO
NT
RO
L; (M
OU
SE
) G
. SH
UK
LIN
/WIK
IME
DIA
CO
MM
ON
S; (F
LIE
S) T.
CH
AP
MA
N/U
NIV
ER
SIT
Y O
F E
AST
AN
GLIA
10.1126/science.1227412
Published by AAAS
on
Nov
embe
r 21,
201
2w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from
Where You Reside / Spend Time Important
!66
ORIGINAL ARTICLE
Architectural design influences the diversity andstructure of the built environment microbiome
Steven W Kembel1, Evan Jones1, Jeff Kline1,2, Dale Northcutt1,2, Jason Stenson1,2,Ann M Womack1, Brendan JM Bohannan1, G Z Brown1,2 and Jessica L Green1,3
1Biology and the Built Environment Center, Institute of Ecology and Evolution, Department ofBiology, University of Oregon, Eugene, OR, USA; 2Energy Studies in Buildings Laboratory,Department of Architecture, University of Oregon, Eugene, OR, USA and 3Santa Fe Institute,Santa Fe, NM, USA
Buildings are complex ecosystems that house trillions of microorganisms interacting with eachother, with humans and with their environment. Understanding the ecological and evolutionaryprocesses that determine the diversity and composition of the built environment microbiome—thecommunity of microorganisms that live indoors—is important for understanding the relationshipbetween building design, biodiversity and human health. In this study, we used high-throughputsequencing of the bacterial 16S rRNA gene to quantify relationships between building attributes andairborne bacterial communities at a health-care facility. We quantified airborne bacterial communitystructure and environmental conditions in patient rooms exposed to mechanical or windowventilation and in outdoor air. The phylogenetic diversity of airborne bacterial communities waslower indoors than outdoors, and mechanically ventilated rooms contained less diverse microbialcommunities than did window-ventilated rooms. Bacterial communities in indoor environmentscontained many taxa that are absent or rare outdoors, including taxa closely related to potentialhuman pathogens. Building attributes, specifically the source of ventilation air, airflow rates, relativehumidity and temperature, were correlated with the diversity and composition of indoor bacterialcommunities. The relative abundance of bacteria closely related to human pathogens was higherindoors than outdoors, and higher in rooms with lower airflow rates and lower relative humidity.The observed relationship between building design and airborne bacterial diversity suggests thatwe can manage indoor environments, altering through building design and operation the communityof microbial species that potentially colonize the human microbiome during our time indoors.The ISME Journal advance online publication, 26 January 2012; doi:10.1038/ismej.2011.211Subject Category: microbial population and community ecologyKeywords: aeromicrobiology; bacteria; built environment microbiome; community ecology; dispersal;environmental filtering
Introduction
Humans spend up to 90% of their lives indoors(Klepeis et al., 2001). Consequently, the way wedesign and operate the indoor environment has aprofound impact on our health (Guenther andVittori, 2008). One step toward better understandingof how building design impacts human healthis to study buildings as ecosystems. Built envi-ronments are complex ecosystems that containnumerous organisms including trillions of micro-organisms (Rintala et al., 2008; Tringe et al., 2008;Amend et al., 2010). The collection of microbiallife that exists indoors—the built environment
microbiome—includes human pathogens and com-mensals interacting with each other and with theirenvironment (Eames et al., 2009). There have beenfew attempts to comprehensively survey the builtenvironment microbiome (Rintala et al., 2008;Tringe et al., 2008; Amend et al., 2010), with moststudies focused on measures of total bioaerosolconcentrations or the abundance of culturable orpathogenic strains (Berglund et al., 1992; Toivolaet al., 2002; Mentese et al., 2009), rather than a morecomprehensive measure of microbial diversity inindoor spaces. For this reason, the factors thatdetermine the diversity and composition of the builtenvironment microbiome are poorly understood.However, the situation is changing. The develop-ment of culture-independent, high-throughputmolecular sequencing approaches has transformedthe study of microbial diversity in a variety ofenvironments, as demonstrated by the recent explo-sion of research on the microbial ecology of aquaticand terrestrial ecosystems (Nemergut et al., 2011)
Received 23 October 2011; revised 13 December 2011; accepted13 December 2011
Correspondence: SW Kembel, Biology and the Built EnvironmentCenter, Institute of Ecology and Evolution, Department of Biology,University of Oregon, Eugene, OR 97405, USA.E-mail: [email protected]
The ISME Journal (2012), 1–11& 2012 International Society for Microbial Ecology All rights reserved 1751-7362/12
www.nature.com/ismej
Microbial Biogeography of Public Restroom SurfacesGilberto E. Flores1, Scott T. Bates1, Dan Knights2, Christian L. Lauber1, Jesse Stombaugh3, Rob Knight3,4,
Noah Fierer1,5*
1 Cooperative Institute for Research in Environmental Science, University of Colorado, Boulder, Colorado, United States of America, 2 Department of Computer Science,
University of Colorado, Boulder, Colorado, United States of America, 3 Department of Chemistry and Biochemistry, University of Colorado, Boulder, Colorado, United
States of America, 4 Howard Hughes Medical Institute, University of Colorado, Boulder, Colorado, United States of America, 5 Department of Ecology and Evolutionary
Biology, University of Colorado, Boulder, Colorado, United States of America
Abstract
We spend the majority of our lives indoors where we are constantly exposed to bacteria residing on surfaces. However, thediversity of these surface-associated communities is largely unknown. We explored the biogeographical patterns exhibitedby bacteria across ten surfaces within each of twelve public restrooms. Using high-throughput barcoded pyrosequencing ofthe 16 S rRNA gene, we identified 19 bacterial phyla across all surfaces. Most sequences belonged to four phyla:Actinobacteria, Bacteriodetes, Firmicutes and Proteobacteria. The communities clustered into three general categories: thosefound on surfaces associated with toilets, those on the restroom floor, and those found on surfaces routinely touched withhands. On toilet surfaces, gut-associated taxa were more prevalent, suggesting fecal contamination of these surfaces. Floorsurfaces were the most diverse of all communities and contained several taxa commonly found in soils. Skin-associatedbacteria, especially the Propionibacteriaceae, dominated surfaces routinely touched with our hands. Certain taxa were morecommon in female than in male restrooms as vagina-associated Lactobacillaceae were widely distributed in femalerestrooms, likely from urine contamination. Use of the SourceTracker algorithm confirmed many of our taxonomicobservations as human skin was the primary source of bacteria on restroom surfaces. Overall, these results demonstrate thatrestroom surfaces host relatively diverse microbial communities dominated by human-associated bacteria with clearlinkages between communities on or in different body sites and those communities found on restroom surfaces. Moregenerally, this work is relevant to the public health field as we show that human-associated microbes are commonly foundon restroom surfaces suggesting that bacterial pathogens could readily be transmitted between individuals by the touchingof surfaces. Furthermore, we demonstrate that we can use high-throughput analyses of bacterial communities to determinesources of bacteria on indoor surfaces, an approach which could be used to track pathogen transmission and test theefficacy of hygiene practices.
Citation: Flores GE, Bates ST, Knights D, Lauber CL, Stombaugh J, et al. (2011) Microbial Biogeography of Public Restroom Surfaces. PLoS ONE 6(11): e28132.doi:10.1371/journal.pone.0028132
Editor: Mark R. Liles, Auburn University, United States of America
Received September 12, 2011; Accepted November 1, 2011; Published November 23, 2011
Copyright: ! 2011 Flores et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported with funding from the Alfred P. Sloan Foundation and their Indoor Environment program, and in part by the NationalInstitutes of Health and the Howard Hughes Medical Institute. The funders had no role in study design, data collection and analysis, decision to publish, orpreparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
Introduction
More than ever, individuals across the globe spend a largeportion of their lives indoors, yet relatively little is known about themicrobial diversity of indoor environments. Of the studies thathave examined microorganisms associated with indoor environ-ments, most have relied upon cultivation-based techniques todetect organisms residing on a variety of household surfaces [1–5].Not surprisingly, these studies have identified surfaces in kitchensand restrooms as being hot spots of bacterial contamination.Because several pathogenic bacteria are known to survive onsurfaces for extended periods of time [6–8], these studies are ofobvious importance in preventing the spread of human disease.However, it is now widely recognized that the majority ofmicroorganisms cannot be readily cultivated [9] and thus, theoverall diversity of microorganisms associated with indoorenvironments remains largely unknown. Recent use of cultiva-tion-independent techniques based on cloning and sequencing ofthe 16 S rRNA gene have helped to better describe these
communities and revealed a greater diversity of bacteria onindoor surfaces than captured using cultivation-based techniques[10–13]. Most of the organisms identified in these studies arerelated to human commensals suggesting that the organisms arenot actively growing on the surfaces but rather were depositeddirectly (i.e. touching) or indirectly (e.g. shedding of skin cells) byhumans. Despite these efforts, we still have an incompleteunderstanding of bacterial communities associated with indoorenvironments because limitations of traditional 16 S rRNA genecloning and sequencing techniques have made replicate samplingand in-depth characterizations of the communities prohibitive.With the advent of high-throughput sequencing techniques, wecan now investigate indoor microbial communities at anunprecedented depth and begin to understand the relationshipbetween humans, microbes and the built environment.
In order to begin to comprehensively describe the microbialdiversity of indoor environments, we characterized the bacterialcommunities found on ten surfaces in twelve public restrooms(six male and six female) in Colorado, USA using barcoded
PLoS ONE | www.plosone.org 1 November 2011 | Volume 6 | Issue 11 | e28132
the stall in), they were likely dispersed manually after women usedthe toilet. Coupling these observations with those of thedistribution of gut-associated bacteria indicate that routine use oftoilets results in the dispersal of urine- and fecal-associated bacteriathroughout the restroom. While these results are not unexpected,they do highlight the importance of hand-hygiene when usingpublic restrooms since these surfaces could also be potentialvehicles for the transmission of human pathogens. Unfortunately,previous studies have documented that college students (who arelikely the most frequent users of the studied restrooms) are notalways the most diligent of hand-washers [42,43].
Results of SourceTracker analysis support the taxonomicpatterns highlighted above, indicating that human skin was theprimary source of bacteria on all public restroom surfacesexamined, while the human gut was an important source on oraround the toilet, and urine was an important source in women’srestrooms (Figure 4, Table S4). Contrary to expectations (seeabove), soil was not identified by the SourceTracker algorithm asbeing a major source of bacteria on any of the surfaces, includingfloors (Figure 4). Although the floor samples contained family-leveltaxa that are common in soil, the SourceTracker algorithmprobably underestimates the relative importance of sources, like
Figure 3. Cartoon illustrations of the relative abundance of discriminating taxa on public restroom surfaces. Light blue indicates lowabundance while dark blue indicates high abundance of taxa. (A) Although skin-associated taxa (Propionibacteriaceae, Corynebacteriaceae,Staphylococcaceae and Streptococcaceae) were abundant on all surfaces, they were relatively more abundant on surfaces routinely touched withhands. (B) Gut-associated taxa (Clostridiales, Clostridiales group XI, Ruminococcaceae, Lachnospiraceae, Prevotellaceae and Bacteroidaceae) were mostabundant on toilet surfaces. (C) Although soil-associated taxa (Rhodobacteraceae, Rhizobiales, Microbacteriaceae and Nocardioidaceae) were in lowabundance on all restroom surfaces, they were relatively more abundant on the floor of the restrooms we surveyed. Figure not drawn to scale.doi:10.1371/journal.pone.0028132.g003
Figure 4. Results of SourceTracker analysis showing the average contributions of different sources to the surface-associatedbacterial communities in twelve public restrooms. The ‘‘unknown’’ source is not shown but would bring the total of each sample up to 100%.doi:10.1371/journal.pone.0028132.g004
Bacteria of Public Restrooms
PLoS ONE | www.plosone.org 5 November 2011 | Volume 6 | Issue 11 | e28132
high diversity of floor communities is likely due to the frequency ofcontact with the bottom of shoes, which would track in a diversityof microorganisms from a variety of sources including soil, which isknown to be a highly-diverse microbial habitat [27,39]. Indeed,bacteria commonly associated with soil (e.g. Rhodobacteraceae,Rhizobiales, Microbacteriaceae and Nocardioidaceae) were, on average,more abundant on floor surfaces (Figure 3C, Table S2).Interestingly, some of the toilet flush handles harbored bacterialcommunities similar to those found on the floor (Figure 2,Figure 3C), suggesting that some users of these toilets may operatethe handle with a foot (a practice well known to germaphobes andthose who have had the misfortune of using restrooms that are lessthan sanitary).
While the overall community level comparisons between thecommunities found on the surfaces in male and female restroomswere not statistically significant (Table S3), there were gender-
related differences in the relative abundances of specific taxa onsome surfaces (Figure 1B, Table S2). Most notably, Lactobacillaceaewere clearly more abundant on certain surfaces within femalerestrooms than male restrooms (Figure 1B). Some species of thisfamily are the most common, and often most abundant, bacteriafound in the vagina of healthy reproductive age women [40,41]and are relatively less abundant in male urine [28,29]. Ouranalysis of female urine samples collected as part of a previousstudy [26] (Figure 1A), found that Lactobacillaceae were dominant inurine, therefore implying that surfaces in the restrooms whereLactobacillaceae were observed were contaminated with urine. Otherstudies have demonstrated a similar phenomenon, with vagina-associated bacteria having also been observed in airplanerestrooms [11] and a child day care facility [10]. As we foundthat Lactobacillaceae were most abundant on toilet surfaces andthose touched by hands after using the toilet (with the exception of
Figure 2. Relationship between bacterial communities associated with ten public restroom surfaces. Communities were clustered usingPCoA of the unweighted UniFrac distance matrix. Each point represents a single sample. Note that the floor (triangles) and toilet (asterisks) surfacesform clusters distinct from surfaces touched with hands.doi:10.1371/journal.pone.0028132.g002
Table 1. Results of pairwise comparisons for unweighted UniFrac distances of bacterial communities associated with varioussurfaces of public restrooms on the University of Colorado campus using the ANOSIM test in Primer v6.
Door in Door out Stall in Stall outFaucethandle
Soapdispenser
Toilet flushhandle Toilet seat Toilet floor
Door in
Door out 20.139
Stall in 0.149 20.053
Stall out 20.074 20.083 20.037
Faucet handle 20.062 20.011 20.092 20.040
Soap dispenser 20.020 0.014 20.060 20.001 0.070
Toilet flush handle 0.376* 0.405* 0.221 0.350* 0.172* 0.470*
Toilet seat 0.742* 0.672* 0.457* 0.586* 0.401* 0.653* 0.187*
Toilet floor 0.995* 0.988* 0.993* 0.961* 0.758* 0.998* 0.577* 0.950*
Sink floor 1.000* 0.995* 1.000* 0.974* 0.770* 1.000* 0.655* 0.982* 20.033
The R-statistic is shown for each comparison with asterisks denoting comparisons that were statistically significant at P#0.01.doi:10.1371/journal.pone.0028132.t001
Bacteria of Public Restrooms
PLoS ONE | www.plosone.org 4 November 2011 | Volume 6 | Issue 11 | e28132
10 FEBRUARY 2012 VOL 335 SCIENCE www.sciencemag.org 650
NEWSFOCUS
CR
ED
ITS
(T
OP
TO
BO
TT
OM
): (P
HO
TO
) C
OU
RT
ES
Y G
ILB
ER
TO
FLO
RE
S; (C
HA
RT
) G
. E
. F
LO
RE
S E
T A
L.,
PLO
S O
NE
6, 1
1 (2
01
1);
PH
OT
O B
Y S
ISIR
A G
OR
TH
ALA
In just that short time, the microbes had begun to take on a “signature” of outside air (more types from plants and soil), and 2 hours after the windows were shut again, the proportion of microbes from the human body increased back to pre-vious levels.
The s tudy, which appeared online 26 Janu-ary in The ISME Journal, found that mechanically ventilated rooms had lower microbial diversity than ones with open win-dows. The availability of fresh air translated into lower proportions of microbes associ-ated with the human body, and consequently, fewer potential pathogens. Although this result suggests that having natural airfl ow may be healthier, Green says answering that question requires clinical data; she’s hoping to convince a hospital to participate in a study to see if the incidence of hospital-acquired infections is associated with a room’s micro-bial community.
For his part, Peccia, who is also a Sloan grantee, is merging microbiology and the
physics of aerosols to look more closely at how the movement of air affects microbes. Peccia says his group is building on work by air-quality engineers and scientists, but “we want to add biology to the equation.”
Bacteria in air behave like other particles; their size dictates how they disperse or settle. Humans in a room not only shed microbes from their skin and mouths, but they also drum up microbial material from the fl oor as
they move around. But to quantify those con-tributions, Peccia’s team has had to develop new methods to collect airborne bacteria and extract their DNA, as the microbes are much less abundant in air than on surfaces.
In one recent study, they used air fi lters to sample airborne particles and microbes in a classroom during 4 days during which students were present and 4 days during which the room was vacant. They measured the abundance and type of fungal and bac-terial genomes present and estimated the microbes’ concentrations in the entire room. By accounting for bacteria entering and leav-
ing the room through ventilation, they calculated that people shed or resuspended about 35 million bacterial cells per person per hour. That number is much higher than the several-hundred-thousand maximum previously estimated to be present in indoor air, Peccia reported last fall at the American Association for Aerosol Research Conference in Orlando, Florida.
His group’s data also suggest that rooms have “memories” of past human inhabitants. By kick-ing into the air settled microbes from the fl oor, occupants expose themselves not just to the microbes of a person coughing next to them, but also possibly to those from a person who coughed in the room a few hours or even days ago.
Peccia hopes to come up with ways to describe the distribution of bacteria indoors that can be used in conjunction with exist-ing knowledge about particulate matter and chemicals in designing healthier buildings. “My hope is that we can bring this enough to the forefront that people who do aerosol sci-ence will fi nd it as important to know biology as to know physics and chemistry,” he says.
Still, even though he’s a willing partici-
pant in indoor microbial ecology research, Peccia thinks that the field has yet to gel. And the Sloan Foundation’s Olsiewski shares some of his con-cern. “Everybody’s gen-erating vast amounts of
data,” she says, but looking across data sets can be diffi cult because groups choose dif-ferent analytical tools. With Sloan support, though, a data archive and integrated analyt-ical tools are in the works.
To foster collaborations between micro-biologists, architects, and building scientists, the foundation also sponsored a symposium on the microbiome of the built environment at the 2011 Indoor Air conference in Austin, Texas, and launched a Web site, MicroBE.net, that’s a clearinghouse of information on the fi eld. Although Olsiewski won’t say how long the foundation will fund its indoor microbial ecology program, she says Sloan is committed to supporting all of the current projects for the next few years. The program’s ultimate goal, she says, is to create a new fi eld of scientifi c inquiry that eventually will be funded by tradi-tional government funding agencies focused on basic biology and environmental policy.
Matthew Kane, a microbial ecologist and program director at the U.S. National Sci-ence Foundation (NSF), says that although there was interest in these questions prior to the Sloan program, the Sloan Foundation has taken a directed approach to funding the research, and “I have no doubt that their investment is going to reap great returns.” So far, though, NSF has funded only one study on indoor microbes: a study of Pseudomonas bacteria in human households.
As studies like Green’s building ecology analysis progress, they should shed light on how indoor environments differ from those traditionally studied by microbial ecologists. “It’s important to have a quantitative under-standing of how building design impacts microbial communities indoors, and how these communities impact human health,” Green says. But it remains to be seen whether we’ll someday design and maintain our build-ings with microbes in mind.
–COURTNEY HUMPHRIES
Courtney Humphries is a freelance writer in Boston and author of Superdove.
100
80
60
40
20
0
Ave
rag
e c
on
trib
uti
on
(%
)
Door in
Door out
Stall i
n
Stall o
ut
Faucet h
andles
Soap disp
enser
Toile
t seat
Toile
t flu
sh h
andle
Toile
t flo
or
Sink f
loor
SOURCES
Soil
Water
Mouth
Urine
Gut
Skin
Outside infl uence. Students prepare to sample air outside a class-room in China as part of an indoor ecology study.
Bathroom biogeography. By swabbing different surfaces in public restrooms, researchers determined that microbes vary in where they come from depend-ing on the surface (chart).
Published by AAAS
on
Febr
uary
9, 2
012
ww
w.s
cien
cem
ag.o
rgD
ownl
oade
d fro
m
Example: Context
!67
K.R. Amato
12
with reduced resource availability [71]. Such a trend is likely
to lead to increased inter-individual differences in juvenile gut
microbiomes in these habitats. Finally, habitat differences such
as forest fragmentation can alter host population densities
[75-79]. In habitats with increased population densities individuals
are likely to come into contact with one another more often and
should therefore exhibit fewer inter-individual differences in gut
microbial community composition. The opposite should occur in
habitats with decreased population densities.
To date, no study measures differences in host social
interactions or population densities across habitats in conjunction
with analyses of the gut microbiota. However, it is likely that a
relationship between individual host contact and gut microbial
community composition exists. Studies of parasites and disease
in wild animals frequently take these factors into account when
examining patterns of transmission [80-83], and processes that
are relevant for pathogenic microbes are likely to be relevant for
commensal microbes as well.
Because established gut microbial communities have been
shown to resist colonization by certain types of microbes [e.g.
84-87], a framework that does not include competition between
frequencies of social interaction and contact are likely to have
fewer inter-individual differences in gut microbial community
composition than host species that spend more time solitary
and less time engaged in social behavior. Although additional
data are necessary to test this prediction thoroughly, data from
black howler monkeys (Alouatta pigra) and chimpanzees provide
support for it [53,66]. Howler monkeys live in highly cohesive
social groups and show an inter-individual Bray-Curtis similarity
index of 0.51 within social groups [53,67], while chimpanzees
live in less cohesive social groups and show an index of
approximately 0.20 [66,68,69].
Because social interactions within species vary in response
to habitat [70-74], we would also expect different patterns in
gut microbial community composition among individuals of
a given species in distinct habitats. For example, a study of
chacma baboons (Papio ursinus) showed that more grooming
occurs when temperatures are higher [74]. It follows that inter-
individual differences in gut microbial community composition
should be smaller among baboons in warmer habitats compared
to baboons in cooler habitats. Similarly, juvenile gelada baboons
(Theropithecus gelada) spend less time playing in habitats
Figure 1. Basic model of factors influencing host fitness, including predicted interactions between host and gut microbiota. Relationships and factors represented by dashed lines indicate areas that are not well studied in wild animal populations.
UnauthenticatedDownload Date | 11/17/14 4:11 AM
Research Article • DOI: 10.2478/micsm-2013-0002 • MICSM • 2013 • 10-29
MicrobioMe Science and Medicine
10
* E-mail: [email protected]
Introduction
As sequencing technology makes data generation faster,
cheaper, and more comprehensive, studies of gut microbial
communities are multiplying at an astonishing rate. As a
result, our understanding of the host-gut microbe relationship
is constantly improving. Studies to date have demonstrated
that the gut microbiota contributes to host nutrition, health
and behavioral patterns by providing energy and nutrients,
improving immune function, and influencing the production of
neuroactive molecules [1-12]. Changes in the composition of
the gut microbial community are known to lead to changes in
its function, which can alter host nutrition, health and behavior
[6,13-23]. Environmental factors such as diet or social contact
are largely responsible for determining the composition of the
gut microbial community [24-31], but host genotype also affects
the abundances of some microbial genera [28,32,33].
Because host-gut microbe relationships are influenced to
some extent by host genotype, and gut microbial community
composition differs according to host phylogeny [34-36],
discussions of the co-evolution of host and gut microbiota are
common in the current literature [7,34-37]. Some researchers
argue that since microbes are found in animals as simple as
earthworms, the co-evolution of animals and bacteria has been
Co-evolution in context: The importance of studying gut microbiomes in wild animals
1Program in Ecology Evolution and Conservation Biology, University of Illinois at Urbana-Champaign,
Urbana, IL, USA, 61801
2Department of Anthropology, University of Illinois at Urbana-Champaign, Urbana, IL, USA, 61801
Katherine R. Amato1,2*
Received 05 August 2013Accepted 29 September 2013
Abstract Because the gut microbiota contributes to host nutrition, health and behavior, and gut microbial community composition differs according to host phylogeny, co-evolution is believed to have been an important mechanism in the formation of the host-gut microbe relationship. However, current research is not ideal for examining this theme. Most studies of the gut microbiota are performed in controlled settings, but gut microbial community composition is strongly influenced by environmental factors. To truly explore the co-evolution of host and microbe, it is necessary to have data describing host-microbe dynamics in natural environments with variation in factors such as climate, food availability, disease prevalence, and host behavior. In this review, I use current knowledge of host-gut microbe dynamics to explore the potential interactions between host and microbe in natural habitats. These interactions include the influence of host habitat on gut microbial community composition as well as the impacts of the gut microbiota on host fitness in a given habitat. Based on what we currently know, the potential connections between host habitat, the gut microbiota, and host fitness are great. Studies of wild animals will be an essential next step to test these connections and to advance our understanding of host-gut microbe co-evolution.
KeywordsGut microbiota • host-microbe • co-evolution • habitat • ecology • fitness
occurring for more than 800 million years [38,39]. Additionally, the
increased complexity and stability of gut microbial communities
in vertebrates as well as the presence of fewer physical barriers
to bacteria has been used to suggest that the adaptive immune
system evolved in vertebrates to recognize gut bacteria and
improve host-gut microbe interactions [40]. Nevertheless, while
it seems likely that co-evolution is an important mechanism for
understanding host-gut microbe relationships, current research
is not ideal for examining the co-evolution of host and microbe.
Most studies of the gut microbiota are performed in
controlled laboratory settings or are focused solely on human
populations [9,16,25,41-49]. Therefore, despite an understanding
that environmental factors greatly influence the host-gut microbe
relationship [25,27-29,31], the effects of natural environmental
variation in factors such as food availability on the host-gut
microbe relationship have generally not been explored. Because
the host-gut microbe mutualism evolved in a natural environment
with complex patterns of climate, food availability, disease
prevalence, and host behavior, a comprehensive examination
of host-gut microbe dynamics must consider these factors.
Specifically, we must establish the ways in which a host’s habitat
influences the selective environment the host imposes upon its
gut microbiota, and in turn, how the gut microbiota influences
the host’s response to its own selective environment in a
© 2013 Katherine R. Amato, licensee Versita Sp. z o. o.This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs license, which means that the text may be used for non-commercial purposes, provided credit is given to the author
UnauthenticatedDownload Date | 11/17/14 4:11 AM
!68
From Wu et al. 2009 Nature 462, 1056-1060
Challenge 1: Biological Dark Matter
Challenge 2: Function Prediction Difficult
Lateral Gene Transfer
Metagenomic Binning
HypotheticalProteins
Solution: Better Prediction Methods and Data
!70
Characterizing the niche-space distributions of components
Sit
es
N orth American E ast C oast_G S 005_E mbayment
N orth American E ast C oast_G S 002_C oasta l
N orth American E ast C oast_G S 003_C oasta l
N orth American E ast C oast_G S 007_C oasta l
N orth American E ast C oast_G S 004_C oasta l
N orth American E ast C oast_G S 013_C oasta l
N orth American E ast C oast_G S 008_C oasta l
N orth American E ast C oast_G S 011_E stuary
N orth American E ast C oast_G S 009_C oasta l
E astern Tropica l Pacific_G S 021_C oasta l
N orth American E ast C oast_G S 006_E stuary
N orth American E ast C oast_G S 014_C oasta l
Polynesia Archipelagos_G S 051_C ora l R eef Atoll
G alapagos Islands_G S 036_C oasta l
G alapagos Islands_G S 028_C oasta l
Indian O cean_G S 117a_C oasta l sample
G alapagos Islands_G S 031_C oasta l upwelling
G alapagos Islands_G S 029_C oasta l
G alapagos Islands_G S 030_W arm S eep
G alapagos Islands_G S 035_C oasta l
S argasso S ea_G S 001c_O pen O cean
E astern Tropica l Pacific_G S 022_O pen O cean
G alapagos Islands_G S 027_C oasta l
Indian O cean_G S 149_H arbor
Indian O cean_G S 123_O pen O cean
C aribbean S ea_G S 016_C oasta l S ea
Indian O cean_G S 148_Fringing R eef
Indian O cean_G S 113_O pen O cean
Indian O cean_G S 112a_O pen O cean
C aribbean S ea_G S 017_O pen O cean
Indian O cean_G S 121_O pen O cean
Indian O cean_G S 122a_O pen O cean
G alapagos Islands_G S 034_C oasta l
C aribbean S ea_G S 018_O pen O cean
Indian O cean_G S 108a_Lagoon R eef
Indian O cean_G S 110a_O pen O cean
E astern Tropica l Pacific_G S 023_O pen O cean
Indian O cean_G S 114_O pen O cean
C aribbean S ea_G S 019_C oasta l
C aribbean S ea_G S 015_C oasta l
Indian O cean_G S 119_O pen O cean
G alapagos Islands_G S 026_O pen O cean
Polynesia Archipelagos_G S 049_C oasta l
Indian O cean_G S 120_O pen O cean
Polynesia Archipelagos_G S 048a_C ora l R eef
Component 1
Component 2
Component 3
Component 4
Component 5
0 .1 0 .2 0 .3 0 .4 0 .5 0 .6
0 .2 0 .4 0 .6 0 .8 1 .0
Sa
lin
ity
Sa
mp
le D
ep
th
Ch
loro
ph
yll
Te
mp
era
ture
Ins
ola
tio
n
Wa
ter
De
pth
G enera l
H ighM ediumLowN A
H ighM ediumLowN A
W ater depth
>4000m2000!4000m900!2000m100!200m20!100m0!20m
>4000m2000!4000m900!2000m100!200m20!100m0!20m
(a) (b) (c)
Figure 3: a) Niche-space distributions for our five components (HT ); b) the site-similarity matrix (HT H); c) environmental variables for the sites. The matrices arealigned so that the same row corresponds to the same site in each matrix. Sites areordered by applying spectral reordering to the similarity matrix (see Materials andMethods). Rows are aligned across the three matrices.
Figure 3a shows the estimated niche-space distribution for each of the five com-ponents. Components 2 (Photosystem) and 4 (Unidentified) are broadly distributed;Components 1 (Signalling) and 5 (Unidentified) are largely restricted to a handful ofsites; and component 3 shows an intermediate pattern. There is a great deal of overlapbetween niche-space distributions for di�erent components.
Figure 3b shows the pattern of filtered similarity between sites. We see clear pat-terns of grouping, that do not emerge when we calculate functional distances withoutfiltering, or using PCA rather than NMF filtering (Figure 3 in Text S1). As withthe Pfams, we see clusters roughly associated with our components, but there is moreoverlapping than with the Pfam clusters (Figure 2b).
Figure 3c shows the distribution of environmental variables measured at each site.Inspection of Figure 3 reveals qualitative correspondence between environmental factorsand clusters of similar sites in the similarity matrix. For example, the “North AmericanEast Coast” samples are divided into two groups, one in the top left and the other in thebottom right of the similarity matrix. Inspection of the environmental features suggeststhat the split in these samples could be mostly due to the di�erences in insolation andwater depth.
We can also examine patterns of similarity between the components themselves,using niche-site distributions or functional profiles (see Figure 5 in Text S1). All 5
8
Better Prediction Methods
More Reference Data
In Situ Function
AB
C
��
�
�
�
�
�� �
�
�
�
�
�
��
��
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
��
�
�
�
��
��
�
�
��
�
�
� ��
�
�
��
��
�
��
�
�
�
�
�
�
�
�
�
�
�
�
��
��
�
�
�
�
�
� �
�
�
��
�
�
� �
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
� �
��
�
�
�
�
�
�
�
��
�
� �
�
�
�
�
�
�
�
�
�
� �
�
�
�
�
�
�
�
�
� �
�
��
�
�
��
�
�
� ��
�
��
�
��
�
�
��
�
��
�
��
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
��
��
�
�
� �
�
�
�
��
�
�
�
��
�
�
�
�
�
�
� �
�
�
��
�
�
�
��
�
�
�
�
�
�
�
�
�
�
� �
�
�
�
�
��
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
��
�
�
��
��
�
�
��
�
�
� �
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
���
�
�
�
�
�
�
�
�
�
��
�
�
�
��
� �
�
�
��
�
�
��
�
�
�
�
�
�
�
�
�
��
�
� ��
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
� �
�
�
�
�
�
��
�
�
�
�
�
�
��
�
�
�
� �
�
�
� �� �
�
�
�
�
� �
�
��
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
��
���
�
�
�
�
�
�
� �
�
�
��
� �
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
��
�
�
��
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
���
�
�
��
�
�
��
��
�
�
�
�
��
�
�
�
�
�
��
�
� �
�
�
��
�
��
�
��
�
��
�
��
��
�
�
�
�
�
�
�
���
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
��
�
�
�
�
�
�
�
��
�
�
��
�
�
�
�
�
��
� �
��
� �
�
�
�
�
�
�
��
�
�� �
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
��
� �
�
�
�
�
�
�
�
��
�
�
�
�
�
�
��
�
�
�
�
�
�
��
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
��
��
�
��
�
�
�
�
�
�
�
�
�� �
�
�
�
�
�
�
�
�
��
��
�
�
��
�
�
���
�
�
�
�
�
�
�
��
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
��
��
�
��
�
�
�
�
�
�
�
�
�
�
�
�� �
�
�
�
�
�
�
�
�
�
�
�
�
�
�� �
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
��
��
�
�
�
�
�
�
�
�
�
�
��
��
�
�
�
�
�
�
�
�
�
�� �
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
� ��
�
�
�
�
�
�
�
�
� �
��
�
�
�
�
�
�
��
�
�
��
�
�
�
�
�
�
� �
�
��
�
�
�
�
� �
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
� �
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
��
�
�
�
�
�
�
��
�
��
�
�
�
� �
�
�
��
�
�
�
��
��
� �
�
�
�
�
��
�
�
�
�
�
�
�
�
��
��
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
��
�
�
��
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
�
� �
��
�
�
�
�
��
�
��
��
��
�
�
�
�
�
�
� �
��
�
�
�
�
RepresentativeGenomes
ExtractProtein
Annotation
All v. AllBLAST
HomologyClustering
(MCL)
SFams
Align & Build
HMMs
HMMs
Screen forHomologs
NewGenomes
ExtractProtein
Annotation
Figure 1
Challenge 3: Systems are Complex
!71
• How distinguish ! Good vs. Bad ! Correlation
vs. Causation
• Solutions: ! More
controlled ecological studies
! Better analysis tools
Good? Bad?
How define “bad” vs. “good” ecosystems
• Can (try to) define features that indicate whether an ecosystem is good or bad ! Productivity ! Diversity ! Stability ! Resilience
• Key major challenge is predicting future “health” of ecosystem
!72
How define “bad” vs. “good” ecosystems
• Can (try to) define features that indicate whether a microbiome is good or bad ! Productivity ! Diversity ! Stability ! Resilience
• Key major challenge is predicting future “health” of a microbiome
!73
How define “bad” vs. “good” ecosystems
• Can (try to) define features that indicate whether a microbiome is good or bad ! Health of host ! Diversity ! Stability ! Resilience
• Key major challenge is predicting future “health” of a microbiome
!74
How do these relate to health?
• Idea of a healthy community vs. a unhealthy community is very complex
• The enormous variation between people and over time makes it VERY difficult and very risky to try and say what is “normal”
!75
Challenge 4: Need More Reference Data
HistoricalCollections
GlobalAutomatedSampling
Fillingin the
Tree of Life
Acknowledgements• GEBA:
• $$: DOE-JGI, DSMZ • Eddy Rubin, Phil Hugenholtz, Hans-Peter Klenk, Nikos Kyrpides, Tanya Woyke, Dongying Wu, Aaron Darling,
Jenna Lang • GEBA Cyanobacteria
• $$: DOE-JGI • Cheryl Kerfeld, Dongying Wu, Patrick Shih
• Haloarchaea • $$$ NSF • Marc Facciotti, Aaron Darling, Erin Lynch,
• Phylosift • $$$ DHS • Aaron Darling, Erik Matsen, Holly Bik, Guillaume Jospin
• iSEEM: • $$: GBMF • Katie Pollard, Jessica Green, Martin Wu, Steven Kembel, Tom Sharpton, Morgan Langille, Guillaume Jospin,
Dongying Wu, • aTOL
• $$: NSF • Naomi Ward, Jonathan Badger, Frank Robb, Martin Wu, Dongying Wu
• Others (not mentioned in detail) • $$: NSF, NIH, DOE, GBMF, DARPA, Sloan • Frank Robb, Craig Venter, Doug Rusch, Shibu Yooseph, Nancy Moran, Colleen Cavanaugh, Josh Weitz • EisenLab: Srijak Bhatnagar, Russell Neches, Lizzy Wilbanks, Holly Bik