Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

Sociomicrobiology andPathogenic Bacteria

JOAO B. XAVIER1

1Program for Computational Biology, Memorial Sloan Kettering Cancer Center,New York, NY 10065

ABSTRACT The study of microbial pathogenesis has beenprimarily a reductionist science since Koch’s principles.Reductionist approaches are essential to identify the causalagents of infectious disease, their molecular mechanisms ofaction, and potential drug targets, and much of medicine’ssuccess in the treatment of infectious disease stems fromthat approach. But many bacteria-caused diseases cannot beexplained by a single bacterium. Several aspects of bacterialpathogenesis will benefit from a more holistic approach thattakes into account social interaction among bacteria of the samespecies and between species in consortia such as the humanmicrobiome. The emerging discipline of sociomicrobiologyprovides a framework to dissect microbial interactions insingle and multi-species communities without compromisingmechanistic detail. The study of bacterial pathogenesis canbenefit greatly from incorporating concepts from otherdisciplines such as social evolution theory and microbialecology, where communities, their interactions with hosts,and with the environment play key roles.

INTRODUCTIONMicrobiology has gathered much attention in recentyears in part thanks to major scientific advancements inthe microbiome field. Large-scale projects such as theNIH-funded Human Microbiome Project (1–3) provideextensive catalogues of the microbes that live in andon the human body. Statements like “the human body ishome to bacteria that outnumber human cells by morethan 10:1” or “the genetic content of these bacteriacan be 100x that of the human genome” are often usedby mainstream media and are known to the generalpublic. Vast explorations of the human and nonhumanmicrobiomes are to a large extent boosted by recentbreakthroughs in DNA sequencing and communitymetagenomics (4–6), and the many studies that have

emerged reveal an expanding role of multispecies host-associated microbial communities in several host func-tions (7, 8). Arguably, one of the most notable functionsof commensal microbiota, i.e., nonpathogenic microbes,is protecting the host from colonization by other mi-crobes (9). This is an exciting area of research that aimsto address open questions in pathogenesis such as whyindividuals exposed to the same pathogen can differ intheir levels of infection. It can also explain why patientscan have increased risk of infections after antibiotictherapy destroys the commensal microbiota that wouldnaturally protect against pathogen invasion.

Understanding the ability of microbiomes to protectagainst colonization by pathogens and other relatedaspects of microbial pathogenesis requires a new setof experimental and theoretical tools. The focus mustbroaden beyond the single pathogen as the cause ofdisease and start to include the host-resident microbiota.Understanding how microbial communities function,how they are assembled, and how they change in timeafter perturbations such as antibiotics or diet changes,is a complex problem that is best suited to an integra-tive approach. Fortunately, there is an extensive body

Received: 12 May 2015, Accepted: 14 October 2015,Published: 13 May 2016

Editors: Indira T. Kudva, National Animal Disease Center, AgriculturalResearch Service, U.S. Department of Agriculture, Ames, IA; andPaul J. Plummer, Department of Veterinary Diagnostic andProduction Animal Medicine, College of Veterinary Medicine,Iowa State University, Ames, IA

Citation: Xavier JB. 2016. Sociomicrobiology and pathogenicbacteria. Microbiol Spectrum 4(3):VMBF-0019-2015. doi:10.1128/microbiolspec.VMBF-0019-2015.

Correspondence: Joao B. Xavier, xavierj@mskcc.org

© 2016 American Society for Microbiology. All rights reserved.

ASMscience.org/MicrobiolSpectrum 1

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

of knowledge on the functioning of complex biologicalconsortia in the fields of ecology and evolution fromwhich we can learn.

I start by reviewing the findings of sociomicrobiology,a discipline that aims to address how bacteria functionin communities (10). I then analyze how seemingly co-operative microbes may actually be driven by selfishmotives even within communities where every microbe isof the same species. I move on to multispecies commu-nities, a more complex scenario where both conflict andcooperation occur and may both be essential compo-nents of the robust behaviors that microecosystemsoften have. I end with an ecologist’s view of the humanmicrobiome and a discussion of how resistance againstpathogen colonization can be interpreted as a problem intheoretical ecology.

BIOFILMS, QUORUM SENSING, AND THEDAWN OF SOCIOMICROBIOLOGYMost species of bacteria are social. Biofilms, dense com-munities of bacteria, are a common cause of persistentinfections, and the list of biofilm-forming pathogensincludes common threats such as Pseudomonas aeru-ginosa (11), Escherichia coli (12), Salmonella enterica(13), Klebsiella pneumoniae (14), Vibrio cholerae (15,16), and Clostridium difficile (17). Clinical microbiol-ogists came to realize the importance of biofilm forma-tion in pathogenesis in part because bacteria in biofilms

have much higher tolerance to antibiotics, and the mech-anism of this tolerance appears to be distinct from con-ventional antibiotic resistance (18, 19).

Biofilms saw a surge in interest among microbiolo-gists in the late 1990s. Even though it was well knownthat microbes form dense surface-attached films and thatthese films have medical implications, the topic of howbacteria form biofilms seemed to get more interest fromengineers who were interested their detrimental rolesin industrial biofouling but also beneficial applicationssuch as biofilm reactors for wastewater treatment (20,21). When experimentalists first showed that quorumsensing played a role in regulating biofilm formation(22, 23), the search for genetic mechanisms of biofilmformation became a very hot topic. The excitement inthe field quickly grew as new molecular mechanisms ofbiofilm formation came to light (24, 25). The growingfield generated a new model (Fig. 1), primarily inspiredby experiments in P. aeruginosa but later supported byother species, in which biofilm formation follows a de-velopmental program with different stages and phases,each one potentially driving expression of a distinct setof genes, much a like the developmental programs ofmulticellular eukaryotic organisms (26).

The excitement experienced at the time was under-standable. If a genetic program similar to developmentalpathways in multicellular organisms controls biofilmformation, then this would open the way to new thera-pies. Antibiofilm drugs such as quorum sensing inhibi-

FIGURE 1 A model of biofilm development and life cycle proposed in reference 18.Planktonic bacteria attach to surfaces, initiate expression of biofilm genes such as syn-thesis of extracellular polymeric matrices, and grow a biofilm. A cell can detach from amature biofilm and go back to the planktonic state, closing the biofilm life cycle.

2 ASMscience.org/MicrobiolSpectrum

Xavier

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

tors (27) would be a huge new opportunity for medicinewhen resistance to traditional antibiotics is a growingproblem at the global scale and pharmaceutical com-panies invest less in new antibiotic discovery (28). Couldwe find ways to fight bacteria by jamming their cell-to-cell communication channels and preventing them fromorganizing themselves in communities that make themharder to treat?

The years that followed the onset of sociomicro-biology were a boom for biofilm research, leading toimportant findings. A notable example is the role of thesecondary messenger cyclic-di-GMP in regulating thebacterial transition from motile to biofilm modes (29).This small molecule, which can regulate many otherfunctions in multiple species of bacteria, plays a key rolein biofilm formation by informing when the cell shoulddownregulate genes for motility and upregulate biofilmgenes. In P. aeruginosa, there is an emerging picturein which the bacterium mechanically senses a surfaceusing the transmembrane Wsp system (30). This systemis a multiprotein complex composed of WspA, WspB,WspC, WspD, WspE, and WspF. The transmembraneWspA protein possibly changes conformation whencells contact an attachment surface. This triggers phos-phorylation of a response regulator called WspR thatthen leads to cyclic-diGMP production. Downstream oftheWsp complex, the transcriptional activator FleQ reg-ulates expression of flagella genes when cyclic-diGMPlevels are low but switches to expression of biofilmmatrix genes (the pel operon) when cyclic-diGMP levelsare high. The ability of FleQ to bind to c-diGMP andregulate the motility-to-biofilm transition depends on aprotein-protein interaction between FleQ and the anti-activator FleN (31, 32). Knocking out FleN producesP. aeruginosa cells that are multiflagellated but immotile(33), whereas point mutations in that protein can pro-duce multiflagellated mutants that are hypermotile,the so-called hyperswarmers (34). Hyperswarmers arelocked in a perpetual motility mode and cannot makeproper biofilms. Similarly, a number of mutants inc-diGMP-related genes have been found to be lockedin either motility or biofilm modes (35). c-diGMP istherefore key to the molecular decision making processthat dictates how cells transition from the planktonicmode to the surface-attached mode of bacterial living(Fig. 1). Work in this area may reveal new molecularmechanisms that can become targets to prevent patho-gens from forming biofilms (36).

Molecular biology often seems to take for grantedthe view of biofilms as highly organized communities.It is not uncommon to find descriptions of a “city of

microbes” (37) where bacteria live together in synergy,communicate via cell-cell signaling, and share secretedresources. But how realistic is that view? Natural se-lection is a selfish process whereby the fittest surviveand leave more offspring. Could we expect biofilmswith millions of individual bacteria to be immune to theevolution of exploitative mutants that benefit from thecooperation of others? In the next section I discussthe arrival of social evolutionary theory to the field ofmicrobiology and how the view of natural selectionacting primarily at the level of the gene can help clar-ify some of these issues and shed light on microbialpathogenesis.

BACTERIAL SOCIAL BEHAVIOR:COOPERATION OR CONFLICT?Social evolution theory is a field that aims to dissectthe evolutionary mechanisms of social behavior. Theevolution of cooperative behavior, in particular, is anold problem that puzzled Darwin, yet even 10 years agothis problem was recognized as one of the top “125unknowns” by the journal Science (38). Around thattime, the field of social evolutionary theory started itsforay into bacterial pathogenesis (39, 40).

A landmark paper at the time looked at the produc-tion of iron-scavenging siderophores under the lens ofsocial evolution (39). Siderophores are compounds se-creted by bacteria that have high affinity to iron. Once inthe extracellular space, siderophores scavenge iron thatwould otherwise be inaccessible to the bacteria and freeit up to be taken up by bacterial cells (Fig. 2). Sidero-phore scavenging allows bacteria such as P. aeruginosato grow in iron-limited environments such as hosttissues, where extracellular iron is normally maintainedat very low concentrations exactly to prevent the growthof pathogens. The problem from an evolutionary per-spective is that siderophores are what is called a “publicgood” in a bacterial society. A public good is a concepttaken from economics that means a resource that isavailable to all individuals within a population, irres-pective of who’s producing it. When the production ofa public good is costly, there is a strong incentive forcheating, meaning that individuals would not producethe public good but just exploit the public goods pro-duced by others. In these situations, what preventscheaters from taking over and causing the collapse of thepopulation?

The study by Griffin and colleagues (39) first showedexperimentally that siderophores of P. aeruginosa arecostly public goods. They compared the growth of

ASMscience.org/MicrobiolSpectrum 3

Sociomicrobiology and Pathogenic Bacteria

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

a siderophore-producing strain and a nonproducingstrain in iron-limited conditions where siderophores arekey to bacterial growth. As expected, they saw that thesiderophore-producing strain grows much better thanthe nonproducing strain when the two are comparedin monocultures. However, when mixed together in acoculture, the non-siderophore-producing strain grewbetter than the producing strain because it could use thesiderophores without paying a metabolic cost of theirproduction. Importantly, the final numbers in the pop-ulation were lower for the coculture than for the mono-culture of siderophore producers. The nonproducingstrain benefited from being mixed with the producer,but the whole population suffered as a consequence(Fig. 3A).

This dramatic outcome is a hallmark of cheating andcaptures the essence of the problem of the evolution ofcooperation (41). Cheaters gain a selfish benefit frombeing in a mixed population with cooperators, but thewhole population suffers from cheating compared to apopulation that is composed 100% of cooperators. Sohow can we observe so many cooperative traits in na-ture, where organisms seem to altruistically sacrificetheir own fitness for the benefit of others? This questionhas been on the minds of evolutionary biologists for along time. One answer is kin selection. J. B. S. Haldane,one of the architects of modern synthesis, reportedly

joked that he would altruistically give his life for twobrothers or eight cousins, reflecting the Mendelian in-heritance probability of one-third of sharing a gene witha brother and one-eighth of sharing a gene with a first-degree cousin. Kin selection explains that a cooperativestrategy can evolve if the fitness costs that the coopera-tive behavior has to the actor are less than the fitnessbenefit to the recipient multiplied by a relatedness coef-ficient between actor and recipient. This relationship, r ×b > c, is known as Hamilton’s rule, in honor of WilliamHamilton, who’s seminal work laid the foundations forinvestigating the genetic basis of social behavior (42,43). The insight behind this elegant rule is that selectionacts at the level of the genes, and a gene encoding for acooperative behavior will increase in frequency within apopulation if its function is to make the organism thatcarries it help other carriers of that gene. Since fitnessconcerns the increase of gene frequency in a population,a gene may be fit if it increases other copies of itself,and social evolutionary biologists often use the term“inclusive fitness” to account for social effects. The con-cepts of social evolution theory were popularized inlarge part thanks to the book The Selfish Gene byRichard Dawkins (44).

Why is this relevant to microbial pathogens? Spon-taneous mutants that lose function in a gene occurcommonly in bacterial populations. If a mutant has a

FIGURE 2 Siderophore production as a cooperative trait (74). Some bacterial pathogenslike Pseudomonas aeruginosa secrete siderophores to scavenge iron the in iron-limitedenvironments of host tissues (panel 1). Siderophores have high affinity to iron and can betaken up by bacteria including non–siderophore producers that still have the siderophorereceptors (panel 2). Non–siderophore producers exploit wild-type producers by notpaying the cost of siderophore production, but this cheating behavior can lead to theextinction of siderophore production in the population (panel 3).

4 ASMscience.org/MicrobiolSpectrum

Xavier

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

loss of function in a cooperative gene, for example, inthe gene pvdA that catalyzes a key step in the synthesisof the siderophore pyoverdin (45), then this mutantcould become a cheater. A way for cooperation to bemaintained in the face of cheaters is if the remainingcooperators cooperate only with individuals that stillcarry a functional copy, but not with mutants that lackthe gene. This would be the case where r would have ahigh value.

Griffin and colleagues (39) tested this prediction bymixing bacteria in different ways to manipulate relat-edness experimentally. Their experiments revealed thatconditions of high relatedness favored siderophoreproducers (cooperators), whereas conditions of low re-latedness, where strains mixed more frequently withother clones, favored nonproducing strains (cheaters),which is in accordance with social evolution theory.

Mechanisms Stabilizing Cooperationin Bacterial PathogensMixing reduces the likelihood that cooperative benefitswill be received by related individuals. In nature andin the clinic we expect that bacterial strains and species

will often be in mixed communities. Are there othermechanisms stabilizing cooperation in communitieswhere relatedness is low?

The extracellular polymeric substances of biofilmscould naïvely be viewed as a public good. These sub-stances, which make up the gooey matrix that sticksbacteria to each other and to the solid substratum in abiofilm (Fig. 1), require significant metabolic resourcesto be produced. Mutants that do not produce matrixcould still benefit from the matrix produced by others.However, a series of studies, first with computer simu-lations (46) and later with experiments with V. cholerae(47) and Pseudomonas fluorescens (48) showed this notto be the case and provided a new mechanism to explainthe evolutionary stability of biofilm communities. Thereason is that bacterial biofilms have very steep gradientsof nutrients and other solute substances. For example, inbiofilms of aerobic bacteria and in colonies growing onagar plates, diffusional gradients are often so steep thatbacteria on the inside of the biofilm cannot grow becauseall oxygen is consumed before it reaches the interior. In amixed biofilm of polymer producers and non–polymerproducers the producers gain an advantage because the

FIGURE 3 Laboratory experiments reveal the hallmarks of cheating. (A) Siderophore-producing Pseudomonas aeruginosa grow reasonably well in iron-depleted media byincreasing iron uptake thanks to siderophore scavenging (Fig. 2). Non–siderophoreproducers (cheaters) grow poorly in the same environment when alone but do betterwhen mixed with producers by not paying the cost of siderophore production. The ad-vantage of nonproducers comes at the expense of the whole population (74). (B) Thecompetitive advantage of cheaters decreases as their frequency increases because thereare fewer cooperators to exploit in the population. This example is taken from a study oftype III secretion systems in P. aeruginosawhere exsA− mutants lacking the type III systemcould cheat over wild-type bacteria (WT), but their measured competitive index decreasedas cheater numbers increased in the population (53).

ASMscience.org/MicrobiolSpectrum 5

Sociomicrobiology and Pathogenic Bacteria

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

polymers allow them to be pushed to the top and reachhigher concentrations of nutrients. By secreting poly-mers, a polymer-producing bacterium benefits itself andits lineage and literally suffocates non–polymer pro-ducers in the inner layers of the biofilm. This mechanismof “competitive smothering” proposes that polymerproduction is a competitive strategy, not a cooperativebehavior. What could at first glance be perceived as acooperation reveals a selfish motive.

Bacteria have alternative ways to tilt Hamilton’sequation in their favor, even when a trait is clearlycooperative. P. aeruginosa colonies are capable of a re-markable collective motility behavior called swarmingbehavior (49, 50). Swarming allows the colony to spreadacross large surfaces in a way that single cells cannot,and this benefits the population. However, swarmingrequires that cells produce and secrete large amountsof biosurfactants, called rhamnolipids, that lubricatethe surface and allow the bacteria to slide on top of it(50). Rhamnolipids are a public good, like siderophores.Strains that do not produce surfactants, such as an rhlA−

mutant, cannot swarm on their own but will do sowhen mixed with surfactant producers (51). However,unlike with siderophores, there is no public good di-lemma. With a 1:1 mixture of wild-type bacteria, rhlA−

remains roughly at 1:1 even though the wild type isproducing copious amounts of surfactants and the rhlA−

strain is benefiting from them. How do we explain thisconundrum?

The rhlA gene is an operon (rhlAB) that is tightlyregulated by a combination of quorum sensing andmetabolic sensing. The regulatory circuit ensures thatP. aeruginosa wild-type cells express rhamnolipids notonly when they have reached a quorum but also whenthey have a carbon source in excess of that neededto grow. This regulation, called metabolic prudence,ensures that P. aeruginosa delays expression of cooper-ation to times when it becomes affordable because it hasan abundance of carbon. By doing this, P. aeruginosauses transcriptional regulation to decrease the cost ofcooperation, the c in Hamilton’s equation (51). Meta-bolic prudence allows cooperation even in situations oflow relatedness where constitutive cooperation is notpossible (52).

Cheating Can Explain ClonalDiversity in InfectionsIn the absence of a mechanism to reduce costs or increaserelatedness, a social trait that provides a benefit to othersmay be doomed. This is the case in opportunistic in-fections by P. aeruginosa where virulence mediated by a

type III secretion system can be seen as an altruistic trait.Type III systems are important factors in pathogenesisthat consist of huge needle-like transmembrane proteincomplexes that inject toxins into host cells. The systemis essential for pathogenesis, but the protein complex iscostly. Experiments using a mouse model of lung infec-tion by P. aeruginosa showed that cheating by mutantslacking type III secretion can happen in vivo (53). Al-though these mutants fail to infect mice on their own,they do well when coinfected with type III–positive iso-genic strains at a 1:1 ratio. When coinfected at differentratios, the type III advantage was high when they werethe minority in the mix, but this advantage decreasedwhen they were in the majority. This is called frequency-dependent selection in evolutionary theory (54), and thefinding that fitness decreases with increasing frequency isa hallmark of cheating (Fig. 3B). The public goods in thiscase are likely the metabolic products released by thekilling of eukaryotic host cells, which should benefitall individuals in a bacterial population irrespective ofwhich ones have a type III system. Type III secretionmutants are often found in patients with P. aeruginosa,and their rise may be due to cheating (53). In the absenceof a mechanism to protect against cheating, cheatersare fated to dominate, and cooperation is doomed toextinction. Perhaps this type of phenomenon could beexploited in the development of “Trojan horse” ap-proaches (55), where engineered cheaters exploit wild-type pathogens.

MULTISPECIES COMMUNITIESAND THE MICROBIOMEIn the previous section I gave several examples of co-operation and conflict between bacteria of the samespecies. However, many bacterial communities are mul-tispecies, and the number and richness of social behav-iors can grow exponentially, making them harder todissect experimentally. Computer models of multispeciesbiofilms suggest that the presence of competing strainscan have a strong influence on within-species coopera-tion and that both within-species and between-speciesinteractions are highly influenced by environmental con-ditions (56). In spite of this complexity, understandinghow bacteria interact within multispecies communitiescan be an essential step toward a mechanistic basis ofhost-associated microbiomes relevant for many aspectsof host health (57–60) and even host behavior (61).

Low biodiversity in the gut microbiota seems to in-crease the risk of enteric infection (9). We can learn a lotfrom extreme examples, and here the infectious diseases

6 ASMscience.org/MicrobiolSpectrum

Xavier

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

of bone marrow transplantation patients is providing aninsightful model. Patients receiving bone marrow trans-plants typically have a blood or bone marrow cancersuch as leukemia or lymphoma, and they are hospital-ized during the procedure. During this time the patientsbecome immunocompromised and may have to receivelarge doses of antibiotics in different combinations toprevent opportunistic infections. In many cases the pa-tients are diagnosed with infection by pathogens suchas C. difficile and vancomycin-resistant Enterococcus(VRE) following administration of antibiotics.

Resistance Against Pathogen ColonizationA recent large-scale study analyzed the gut microbiotaof a cohort of 94 allogeneic hematopoietic stem celltransplantation patients at Memorial Sloan Kettering(62). Metagenomics from fecal samples taken at sev-eral time points relative to the day of the transplantshowed many cases in which the biodiversity of themicrobiota fell sharply during antibiotic treatment. Thisdrop in biodiversity is typically due to the expansionof a single member of the gut microbiota. When a singlemember dominates the microbiome this boosts the riskof infection.

The observation that intestinal domination in bonemarrow transplants increases the risk of infection sug-gests an ecological interpretation in which the com-

mensal gut microbiota is a biodiverse ecosystem thatnaturally resists invasion by a foreign species. When thehost takes antibiotics, the composition of the gut micro-biota is perturbed which can cause a cascading loss ofspecies that interact with each other and lead to a dropin biodiversity that opens the way to invasions (Fig. 4).This ecological interpretation is supported by mathe-matical models that take into account the dynamic so-cial interactions between species. Computer simulationsshow that sudden shifts in microbiota composition canlead the system to a state of dysbiosis that is difficult torecover from (63). The same effect was replicated inmouse models in which antibiotics were given to mice toperturb their gut microbiota before they received a doseof pathogens. The procedure has been tested in a rangeof antibiotics and at least two pathogens, C. difficileand vancomycin-resistantEnterococcus (64, 65). In thesemouse models pretreatment with antibiotics increasesinfection rates dramatically.

Understanding resistance to invasion in the gut mi-crobiota is a problem in ecology, and it makes senseto apply the tools of mathematical ecology to dissect itsmechanisms. A recent approach used the classical modelof predator-prey dynamics, called the generalized Lotka-Volterra equation, to describe the interactions betweenmicrobes in the gut (66–68). In its most detailed form,the model includes three terms to describe (i) the specific

FIGURE 4 Colonization resistance by the gut microbiota can be harmed by antibiotictherapy. (Panel 1) The gut microbiota can resist colonization by pathogens such asClostridium difficile. (Panel 2) Antibiotics disrupt the ecology of the commensal micro-biota. (Panel 3) Antibiotic-challenged microbiota open the way to colonization.

ASMscience.org/MicrobiolSpectrum 7

Sociomicrobiology and Pathogenic Bacteria

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

growth rate of each microbe, (ii) the pairwise inter-actions between all microbial species, and (iii) the effectsof external factors such as antibiotics on each species.These models have a large lumber of parameters, anddetermining their values can be challenging. It is none-theless possible by using large enough data sets andcomputational techniques to avoid overfitting. Oncea model is correctly parameterized it reveals the net-work of interactions occurring between members ofthe microbiota, which is valuable information for theinvestigation of mechanisms of resilience to antibioticperturbations but also to identify microbes that protectagainst pathogen invasion (67).

The Lotka-Volterra approach was applied recentlyto model the microbiota of human patients and mousemodels following antibiotic treatment and C. difficileinfection. The model revealed that a single commensalmicrobe, Clostridium scindens, explained a significantpart of the protection in both mice and humans. How-ever, a community was always better than a singlemicrobe. Experimental follow-up studies unveiled themechanism by showing that the capability of C. scindensto metabolize secondary bile acids is key to hinderingC. difficile colonization (69).

CONCLUSIONMicrobes have rich and diverse social lives (70), andpathogenic bacteria are not an exception. In many cases,pathogens invading a human host encounter a com-mensal community that may be viewed as the first lineof defense against pathogen invasion. Understandinghow these communities function requires an investiga-tion of its social behaviors, both cooperative and com-petitive, and how they produce a biodiverse and robustmicrobiota.

These are exciting times for the field of the humanmicrobiome. Although the millions of bacteria that livein and on our bodies have been long recognized to playimportant roles in health and disease, their study hastraditionally been hampered because most microbes aredifficult to cultivate in laboratory conditions. Recentadvancements in DNA sequencing, metagenomic anal-ysis, and culturing of microbial communities (71, 72)enable a direct mechanistic analysis of microbiome bi-ology. The booming field of metagenomics-based anal-ysis of the microbiota is opening a new perspective thatpresents new challenges and many opportunities. Wecan anticipate that microbiome analysis of patients maybecome routine in the future, enabling the use of com-puter models based on ecological theory to assist medi-

cine, for example, in the rational design of antibiotictherapies (73).

Before this is possible we must gain a better under-standing of the ecology and evolution of social inter-action in microbial communities. As we discussed here,even monospecies communities have conflict and coop-eration. Analyzing the features of microbial communi-ties requires new frameworks that expand the field ofsociomicrobiology to include concepts from evolution.There is a tremendous potential for the field of microbialpathogenesis, because the social behaviors of microbesmay reveal new therapeutic targets.

ACKNOWLEDGMENTSI thank Joana S. Torres for making the figures.

This work was supported by the Office of the Director, Na-tional Institutes of Health under award number DP2OD008440to J.B.X. and a seed grant from the Lucille Castori Center forMicrobes, Inflammation, and Cancer.

REFERENCES1. Peterson J, Garges S, Giovanni M, McInnes P, Wang L, Schloss JA,Bonazzi V, McEwen JE,Wetterstrand KA, Deal C. 2009. The NIH humanmicrobiome project. Genome Res 19:2317–2323.2. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R,Gordon JI. 2007. The human microbiome project. Nature 449:804–810.3. The Human Microbiome Project Consortium. 2012. A framework forhuman microbiome research. Nature 486:215–221.4. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, SargentM, Gill SR, Nelson KE, Relman DA. 2005. Diversity of the human in-testinal microbial flora. Science 308:1635–1638.5. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A,Ley RE, Sogin ML, Jones WJ, Roe BA, Affourtit JP. 2009. A core gutmicrobiome in obese and lean twins. Nature 457:480–484.6. Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, GordonJI. 2005. Obesity alters gut microbial ecology. Proc Natl Acad Sci USA102:11070–11075.7. Cho I, Blaser MJ. 2012. The human microbiome: at the interface ofhealth and disease. Nat Rev Genet 13:260–270.8. Morgan XC, Tickle TL, Sokol H, Gevers D, Devaney KL, Ward DV,Reyes JA, Shah SA, LeLeiko N, Snapper SB. 2012. Dysfunction ofthe intestinal microbiome in inflammatory bowel disease and treatment.Genome Biol 13:R79.9. Taur Y, Pamer EG. 2013. The intestinal microbiota and susceptibility toinfection in immunocompromised patients. Curr Opin Infect Dis 26:332–337.10. Parsek MR, Greenberg EP. 2005. Sociomicrobiology: the connectionsbetween quorum sensing and biofilms. Trends Microbiol 13:27–33.11. Costerton J, Stewart PS, Greenberg E. 1999. Bacterial biofilms: acommon cause of persistent infections. Science 284:1318–1322.

12. Ashby M, Neale J, Knott S, Critchley I. 1994. Effect of antibiotics onnon-growing planktonic cells and biofilms of Escherichia coli. J Anti-microb Chemother 33:443–452.

13. Ledeboer NA, Jones BD. 2005. Exopolysaccharide sugars contributeto biofilm formation by Salmonella enterica serovar Typhimuriumon HEp-2 cells and chicken intestinal epithelium. J Bacteriol 187:3214–3226.

14. Anderl JN, Franklin MJ, Stewart PS. 2000. Role of antibiotic pene-tration limitation in Klebsiella pneumoniae biofilm resistance to ampicillinand ciprofloxacin. Antimicrob Agents Chemother 44:1818–1824.

8 ASMscience.org/MicrobiolSpectrum

Xavier

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

15. Drescher K, Nadell CD, Stone HA, Wingreen NS, Bassler BL. 2014.Solutions to the public goods dilemma in bacterial biofilms. Curr Biol24:50–55.16. Nadell CD, Drescher, Wingreen NS, Bassler BL. 2015. Extracellularmatrix structure governs invasion resistance in bacterial biofilms. ISME J9:1700–1709.17. Dapa T, Unnikrishnan M. 2013. Biofilm formation by Clostridiumdifficile. Gut Microbes 4:397–402.18. O’Toole G, Kaplan HB, Kolter R. 2000. Biofilm formation as mi-crobial development. Annu Rev Microbiol 54:49–79.19. Stewart PS, Costerton JW. 2001. Antibiotic resistance of bacteria inbiofilms. Lancet 358:135–138.20. Characklis W, Cooksey K. 1983. Biofilms and microbial fouling. AdvAppl Microbiol 29:93–138.21. Henze M, Harremoës P. 1983. Anaerobic treatment of wastewater infixed film reactors: a literature review. Water Sci Technol 15:1–101.22. Davies DG, Parsek MR, Pearson JP, Iglewski BH, Costerton J,Greenberg E. 1998. The involvement of cell-to-cell signals in the devel-opment of a bacterial biofilm. Science 280:295–298.23. Parsek MR, Greenberg E. 2005. Sociomicrobiology: the connectionsbetween quorum sensing and biofilms. Trends Microbiol 13:27–33.24. O’Toole GA, Kolter R. 1998. Flagellar and twitching motility arenecessary for Pseudomonas aeruginosa biofilm development. Mol Micro-biol 30:295–304.25. O’Toole GA, Pratt LA, Watnick PI, Newman DK, Weaver VB, KolterR. 1999. Genetic approaches to study of biofilms. Methods Enzymol310:91–109.26. O’Toole G, Kaplan HB, Kolter R. 2000. Biofilm formation as mi-crobial development. Annu Rev Microbiol 54:49–79.27. Hentzer M, Riedel K, Rasmussen TB, Heydorn A, Andersen JB,ParsekMR, Rice SA, Eberl L, Molin S, Høiby N, Kjelleberg S, GivskovM.2002. Inhibition of quorum sensing in Pseudomonas aeruginosa biofilmbacteria by a halogenated furanone compound. Microbiology 148:87–102.28. Boyle KE, Heilmann S, van Ditmarsch D, Xavier JB. 2013. Exploitingsocial evolution in biofilms. Curr Opin Microbiol 16:207–212.29. Jenal U, Malone J. 2006. Mechanisms of cyclic-di-GMP signaling inbacteria. Annu Rev Genet 40:385–407.30. Güvener ZT, Harwood CS. 2007. Subcellular location characteristicsof the Pseudomonas aeruginosa GGDEF protein, WspR, indicate that itproduces cyclic-di-GMP in response to growth on surfaces.Mol Microbiol66:1459–1473.31. Hickman JW, Harwood CS. 2008. Identification of FleQ from Pseu-domonas aeruginosa as a c-di-GMP-responsive transcription factor. MolMicrobiol 69:376–389.32. Baraquet C, Murakami K, Parsek MR, Harwood CS. 2012. The FleQprotein from Pseudomonas aeruginosa functions as both a repressor andan activator to control gene expression from the pel operon promoter inresponse to c-di-GMP. Nucleic Acids Res 40:7207–7218.33. Dasgupta N, Arora SK, Ramphal R. 2000. fleN, a gene that regulatesflagellar number in Pseudomonas aeruginosa. J Bacteriol 182:357–364.34. van Ditmarsch D, Boyle KE, Sakhtah H, Oyler JE, Nadell CD, DezielE, Dietrich LE, Xavier JB. 2013. Convergent evolution of hyperswarmingleads to impaired biofilm formation in pathogenic bacteria. Cell Rep4:697–708.35. Kuchma SL, Brothers KM, Merritt JH, Liberati NT, Ausubel FM,O’Toole GA. 2007. BifA, a cyclic-di-GMP phosphodiesterase, inverselyregulates biofilm formation and swarming motility by Pseudomonasaeruginosa PA14. J Bacteriol 189:8165–8178.36. Kearns DB. 2013. You get what you select for: better swarmingthrough more flagella. Trends Microbiol 21:508–509.37. Watnick P, Kolter R. 2000. Biofilm, city of microbes. J Bacteriol182:2675–2679.

38. Pennisi E. 2005. How did cooperative behavior evolve? Science309:93.39. Griffin AS, West SA, Buckling A. 2004. Cooperation and competitionin pathogenic bacteria. Nature 430:1024–1027.40. Keller L, Surette MG. 2006. Communication in bacteria: an ecologicaland evolutionary perspective. Nat Rev Microbiol 4:249–258.41. West SA, Griffin AS, Gardner A, Diggle SP. 2006. Social evolutiontheory for microorganisms. Nat Rev Microbiol 4:597–607.42. Hamilton WD. 1964. Genetical evolution of social behaviour. I.J Theor Biol 7:1–16.43. Hamilton WD. 1964. The genetical evolution of social behaviour. II.J Theor Biol 7:17–52.44. Dawkins R. 2006. The Selfish Gene. Oxford University Press, Oxford,UK.45. Visca P, Ciervo A, Orsi N. 1994. Cloning and nucleotide sequence ofthe pvdA gene encoding the pyoverdin biosynthetic enzyme L-ornithineN5-oxygenase in Pseudomonas aeruginosa. J Bacteriol 176:1128–1140.46. Xavier JB, Foster KR. 2007. Cooperation and conflict in microbialbiofilms. Proc Natl Acad Sci USA 104:876–881.47. Nadell CD, Bassler BL. 2011. A fitness trade-off between local com-petition and dispersal in Vibrio cholerae biofilms. Proc Natl Acad Sci USA108:14181–14185.48. Kim W, Racimo F, Schluter J, Levy SB, Foster KR. 2014. Importanceof positioning for microbial evolution. Proc Natl Acad Sci USA 111:E1639–E1647.49. Köhler T, Curty LK, Barja F, Van Delden C, Pechère JC. 2000.Swarming of Pseudomonas aeruginosa is dependent on cell-to-cell sig-naling and requires flagella and pili. J Bacteriol 182:5990–5996.50. Rashid MH, Kornberg A. 2000. Inorganic polyphosphate is neededfor swimming, swarming, and twitching motilities of Pseudomonasaeruginosa. Proc Natl Acad Sci USA 97:4885–4890.51. Xavier JB, Kim W, Foster KR. 2011. A molecular mechanism thatstabilizes cooperative secretions in Pseudomonas aeruginosa. Mol Micro-biol 79:166–179.52. de Vargas Roditi L, Boyle KE, Xavier JB. 2013. Multilevel selectionanalysis of a microbial social trait. Mol Syst Biol 9:684.53. Czechowska K, McKeithen-Mead S, Al Moussawi K, Kazmierczak BI.2014. Cheating by type 3 secretion system-negative Pseudomonas aeru-ginosa during pulmonary infection. Proc Natl Acad Sci USA 111:7801–7806.54. van Ditmarsch D, Xavier JB. 2014. Seeing is believing: what ex-periments with microbes reveal about evolution. Trends Microbiol 22:2–4.55. Brown SP, West SA, Diggle SP, Griffin AS. 2009. Social evolution inmicro-organisms and a Trojan horse approach to medical interventionstrategies. Philos Trans R Soc Lond B Biol Sci 364:3157–3168.56. Mitri S, Xavier JB, Foster KR. 2011. Social evolution in multispeciesbiofilms. Proc Natl Acad Sci USA 108(Suppl 2):10839–10846.57. Grice EA, Kong HH, Conlan S, Deming CB, Davis J, Young AC,Bouffard GG, Blakesley RW, Murray PR, Green ED, Turner ML, SegreJA, Progra NCS. 2009. Topographical and temporal diversity of the hu-man skin microbiome. Science 324:1190–1192.58. Dewhirst FE, Chen T, Izard J, Paster BJ, Tanner ACR, Yu W-H,Lakshmanan A, Wade WG. 2010. The human oral microbiome. J Bac-teriol 192:5002–5017.59. Kau AL, Ahern PP, Griffin NW, Goodman AL, Gordon JI. 2011.Human nutrition, the gut microbiome and the immune system. Nature474:327–336.60. Wu GD, Chen J, Hoffmann C, Bittinger K, Chen Y-Y, Keilbaugh SA,BewtraM, Knights D,WaltersWA, Knight R, Sinha R, Gilroy E, Gupta K,Baldassano R, Nessel L, Li H, BushmanFD Lewis JD. 2011. Linking long-term dietary patterns with gut microbial enterotypes. Science 334:105–108.

ASMscience.org/MicrobiolSpectrum 9

Sociomicrobiology and Pathogenic Bacteria

Downloaded from www.asmscience.org by

IP: 140.163.254.150

On: Tue, 17 May 2016 01:08:55

61. Ezenwa VO, Gerardo NM, Inouye DW, Medina M, Xavier JB. 2012.Microbiology. Animal behavior and the microbiome. Science 338:198–199.62. Taur Y, Xavier JB, Lipuma L, Ubeda C, Goldberg J, Gobourne A,Lee YJ, Dubin KA, Socci ND, Viale A, Perales MA, Jenq RR, van denBrink MR, Pamer EG. 2012. Intestinal domination and the risk of bac-teremia in patients undergoing allogeneic hematopoietic stem cell trans-plantation. Clin Infect Dis 55:905–914.63. Bucci V, Bradde S, Biroli G, Xavier JB. 2012. Social interaction,noise and antibiotic-mediated switches in the intestinal microbiota. PLoSComput Biol 8:e1002497. doi:10.1371/journal.pcbi.1002497.

64. Ubeda C, Taur Y, Jenq RR, Equinda MJ, Son T, Samstein M, Viale A,Socci ND, van den BrinkMR, Kamboj M, Pamer EG. 2010. Vancomycin-resistant Enterococcus domination of intestinal microbiota is enabledby antibiotic treatment in mice and precedes bloodstream invasion inhumans. J Clin Invest 120:4332–4341.

65. Buffie CG, Jarchum I, Equinda M, Lipuma L, Gobourne A, Viale A,Ubeda C, Xavier J, Pamer EG. 2012. Profound alterations of intestinalmicrobiota following a single dose of clindamycin results in sustained sus-ceptibility to Clostridium difficile-induced colitis. Infect Immun 80:62–73.

66. Marino S, Baxter NT, Huffnagle GB, Petrosino JF, Schloss PD. 2014.Mathematical modeling of primary succession of murine intestinalmicrobiota. Proc Natl Acad Sci USA 111:439–444.

67. Stein RR, Bucci V, Toussaint NC, Buffie CG, Ratsch G, PamerEG, Sander C, Xavier JB. 2013. Ecological modeling from time-series

inference: insight into dynamics and stability of intestinal microbiota.PLoS Comput Biol 9:e1003388. doi:10.1371/journal.pcbi.1003388.68. Fisher CK, Mehta P. 2014. Identifying keystone species in the humangut microbiome from metagenomic timeseries using sparse linear regres-sion. PloS One 9:e102451. doi:10.1371/journal.pone.0102451.69. Buffie CG, Bucci V, Stein RR, McKenney PT, Ling L, Gobourne A,No D, Liu H, Kinnebrew M, Viale A, Littmann E, van den Brink MR,Jenq RR, Taur Y, Sander C, Cross J, Toussaint NC, Xavier JB, Pamer EG.2014. Precision microbiome reconstitution restores bile acid mediatedresistance to Clostridium difficile. Nature. [Epub ahead of print.]70. Nadell CD, Xavier JB, Foster KR. 2009. The sociobiology of biofilms.FEMS Microbiol Rev 33:206–224.71. Goodman AL, Kallstrom G, Faith JJ, Reyes A, Moore A, Dantas G,Gordon JI. 2011. Extensive personal human gut microbiota culturecollections characterized and manipulated in gnotobiotic mice. Proc NatlAcad Sci USA 108:6252–6257.72. Cullen TW, Schofield WB, Barry NA, Putnam EE, Rundell EA, TrentMS, Degnan PH, Booth CJ, Yu H, Goodman AL. 2015. Gut microbiota.Antimicrobial peptide resistance mediates resilience of prominent gutcommensals during inflammation. Science 347:170–175.73. Bucci V, Xavier JB. 2014. Towards predictive models of the humangut microbiome. J Mol Biol. [Epub ahead of print.]74. Griffin AS, West SA, Buckling A. 2004. Cooperation and competitionin pathogenic bacteria. Nature 430:1024–1027.

10 ASMscience.org/MicrobiolSpectrum

Xavier