31. L. J. Bayne et al., Cancer Cell 21, 822–835 (2012).32. Y. Pylayeva-Gupta, K. E. Lee, C. H. Hajdu, G. Miller, D. Bar-Sagi,

Cancer Cell 21, 836–847 (2012).33. D. C. Strachan et al., OncoImmunology 2, e26968 (2013).34. Y. Zhu et al., Cancer Res. 74, 5057–5069 (2014).35. A. J. Garcia et al., Mol. Cell. Biol. 34, 2017–2028 (2014).36. D. G. DeNardo et al., Cancer Discov. 1, 54–67 (2011).37. B. Molon et al., J. Exp. Med. 208, 1949–1962 (2011).38. G. T. Motz et al., Nat. Med. 20, 607–615 (2014).39. X. Zang et al., Mod. Pathol. 23, 1104–1112 (2010).40. R. J. Buckanovich et al., Nat. Med. 14, 28–36 (2008).41. Y. Huang, S. Goel, D. G. Duda, D. Fukumura, R. K. Jain, Cancer

Res. 73, 2943–2948 (2013).42. G. C. Jayson, D. J. Hicklin, L. M. Ellis, Nat. Rev. Clin. Oncol. 9,

297–303 (2012).43. R. K. Shrimali et al., Cancer Res. 70, 6171–6180 (2010).44. Y. Huang et al., Proc. Natl. Acad. Sci. U.S.A. 109, 17561–17566

(2012).45. D. I. Gabrilovich et al., Nat. Med. 2, 1096–1103 (1996).46. J. Hamzah et al., Nature 453, 410–414 (2008).47. A. Johansson, J. Hamzah, C. J. Payne, R. Ganss, Proc. Natl.

Acad. Sci. U.S.A. 109, 7841–7846 (2012).48. A. Calcinotto et al., J. Immunol. 188, 2687–2694 (2012).49. E. D. Thompson, H. L. Enriquez, Y. X. Fu, V. H. Engelhard, J. Exp.

Med. 207, 1791–1804 (2010).50. Y. Ma et al., Cancer Res. 74, 436–445 (2014).51. M. L. Broz et al., Cancer Cell 26, 638–652 (2014).52. K. Hildner et al., Science 322, 1097–1100 (2008).53. B. Ruffell et al., Cancer Cell 26, 623–637 (2014).54. X. Yang et al., Cancer Cell 25, 37–48 (2014).55. M. Z. Noman et al., J. Exp. Med. 211, 781–790 (2014).56. O. R. Colegio et al., Nature 513, 559–563 (2014).57. M. J. Cavnar et al., J. Exp. Med. 210, 2873–2886 (2013).58. P. J. Murray et al., Immunity 41, 14–20 (2014).59. B. Z. Qian, J. W. Pollard, Cell 141, 39–51 (2010).60. D. H. Munn, A. L. Mellor, Trends Immunol. 34, 137–143

(2013).61. M. Platten, W. Wick, B. J. Van den Eynde, Cancer Res. 72,

5435–5440 (2012).62. C. Smith et al., Cancer Discov. 2, 722–735 (2012).63. R. B. Holmgaard, D. Zamarin, D. H. Munn, J. D. Wolchok,

J. P. Allison, J. Exp. Med. 210, 1389–1402 (2013).64. M. D. Sharma et al., Immunity 38, 998–1012 (2013).65. P. Zhou et al., Nature 506, 52–57 (2014).66. S. M. Pyonteck et al., Nat. Med. 19, 1264–1272 (2013).67. G. L. Beatty et al., Science 331, 1612–1616 (2011).68. N. I. Affara et al., Cancer Cell 25, 809–821 (2014).69. A. Pommier et al., Proc. Natl. Acad. Sci. U.S.A. 110, 13085–13090

(2013).70. Y. Murata, T. Kotani, H. Ohnishi, T. Matozaki, J. Biochem. 155,

335–344 (2014).71. K. Weiskopf et al., Science 341, 88–91 (2013).72. H. Salmon et al., J. Clin. Invest. 122, 899–910 (2012).73. M. Kraman et al., Science 330, 827–830 (2010).74. E. W. Roberts et al., J. Exp. Med. 210, 1137–1151 (2013).75. M. H. Sherman et al., Cell 159, 80–93 (2014).76. C. J. Scotton et al., Cancer Res. 62, 5930–5938 (2002).77. J. J. Liang et al., Cancer Epidemiol. Biomarkers Prev. 19,

2598–2604 (2010).78. Y. Akishima-Fukasawa et al., Am. J. Clin. Pathol. 132, 202–210,

quiz 307 (2009).79. M. C. Poznansky et al., Nat. Med. 6, 543–548 (2000).80. J. Eyles et al., J. Clin. Invest. 120, 2030–2039 (2010).81. J. Predina et al., Proc. Natl. Acad. Sci. U.S.A. 110, E415–E424

(2013).82. T. Chiba et al., Br. J. Cancer 91, 1711–1717 (2004).

ACKNOWLEDGMENTS

J.A.J. is supported by grants from the National Cancer Institute,the American Cancer Society, the Breast Cancer ResearchFoundation, New York State Health Research Science Board, Cyclefor Survival, and the Center for Metastasis Research at MSKCC.D.T.F. is the Walter B. Wriston Professor of Pancreatic CancerResearch and is supported by a Lustgarten Distinguished ScholarAward. D.T.F. is a cofounder of Myosotis, which has licensed thepotential intellectual property that may be associated with thefinding that interfering with the interaction of CXCR4 with CXCL12promotes the antitumor effects of T cell checkpoint antagonists.The authors thank C. Connell for Fig. 1C. We apologize to all theinvestigators whose research could not be appropriately citedbecause of the journal’s space limitations.

10.1126/science.aaa6204

REVIEWS

Cancer and the microbiotaWendy S. Garrett1,2,3,4

A host’s microbiota may increase, diminish, or have no effect at all on cancersusceptibility. Assigning causal roles in cancer to specific microbes and microbiotas,unraveling host-microbiota interactions with environmental factors in carcinogenesis,and exploiting such knowledge for cancer diagnosis and treatment are areas ofintensive interest. This Review considers how microbes and the microbiota mayamplify or mitigate carcinogenesis, responsiveness to cancer therapeutics, andcancer-associated complications.

The relationship between cancer and mi-crobes is complex. Although cancer is gen-erally considered to be a disease of hostgenetics and environmental factors, mi-croorganisms are implicated in ~20% of

human malignancies (1). Microbes present atmucosal sites can become part of the tumormicroenvironment of aerodigestive tract ma-lignancies, and intratumoral microbes can affectcancer growth and spread in many ways (2–6).In counterpoise, the gut microbiota also func-tions in detoxification of dietary components,reducing inflammation, and maintaining a bal-ance in host cell growth and proliferation. Thepossibility of microbe-based cancer therapeuticshas attracted interest for more than 100 years,from Coley’s toxins (one of the earliest formsof cancer bacteriotherapy) to the current era ofsynthetic biology’s designer microbes and micro-biota transplants. Thus, interrogation of the rolesof microbes and the microbiota in cancer re-quires a holistic perspective.The ways in which microbes and the micro-

biota contribute to carcinogenesis, whether byenhancing or diminishing a host’s risk, fall intothree broad categories: (i) altering the balanceof host cell proliferation and death, (ii) guidingimmune system function, and (iii) influencingmetabolism of host-produced factors, ingestedfoodstuffs, and pharmaceuticals (Fig. 1). Assign-ing microbial communities, their members, andaggregate biomolecular activities into these cat-egories will require a substantial research com-mitment. This Review discusses how microbesand the microbiota may contribute to cancerdevelopment and progression, responsivenessto cancer therapeutics, and cancer-associatedcomplications.

Microbial contributions to carcinogenesis

Of the estimated 3.7 × 1030 microbes living onEarth (7), only 10 are designated by the Inter-national Agency for Cancer Research (IACR)

as carcinogenic to humans (1). Although mostof these carcinogenic microbes colonize largepercentages of the human population, only asubset of affected individuals develop cancer,because host and microbial genotypes influ-ence cancer susceptibility.Tumors arising at boundary surfaces, such as

the skin, oropharynx, and respiratory, digestive,and urogenital tracts, harbor amicrobiota, whichcomplicates cancer-microbe causality. Enrich-ment of a microbe at a tumor site does not con-note that a microbe is directly associated, letalone causal, in disease. Rather, microbes mayfind a tumor’s oxygen tension or carbon sourcespermissive and take advantage of an underusednutritional niche. Decreased abundances of spe-cific microbes may also place a host at enhancedrisk for cancer development at sites local ordistant from this microbial shift. Thus, rigorousframeworks for interpreting tumor-associatedmicrobiota data are essential (2).

Oncomicrobes, shifting the balance ofwhen to die and when to grow

Bona fide oncomicrobes—microbes that triggertransformation events in host cells—are rare.Beyond the 10 IACR-designated microbes, thereare a handful of other microorganisms with ro-bust but fewer aggregate data supporting their rolein human carcinogenesis. As many of these andtheir carcinogenic mechanisms have been recent-ly reviewed (2–6, 8), select activities representingcommon pathways by which microbes influencecancer will be highlighted.Human oncoviruses can drive carcinogene-

sis by integrating oncogenes into host genomes.Human papillomaviruses (HPV) express onco-proteins such as E6 and E7. Data from recentgenomic analyses of HPV+ cervical cancers sug-gest that viral integration also selectively triggersamplification of host genes in pathways with es-tablished roles in cancer (9).Microbes also drive transformation by affect-

ing genomic stability, resistance to cell death,and proliferative signaling. Many bacteria haveevolvedmechanisms to damage DNA, so as to killcompetitors and survive in the microbial world.Unfortunately, these bacterial defensive factorscan lead to mutational events that contribute tocarcinogenesis (Fig. 2). Examples include coli-bactin encoded by the pks locus [expressed by B2

80 3 APRIL 2015 • VOL 348 ISSUE 6230 sciencemag.org SCIENCE

1Department of Immunology and Infectious Diseases andDepartment of Genetics and Complex Diseases, HarvardT. H. Chan School of Public Health, Boston, MA 02115, USA.2Department of Medical Oncology, Dana-Farber CancerInstitute, Boston, MA 02115, USA. 3Department of Medicine,Harvard Medical School, Boston, MA 02115, USA. 4BroadInstitute of Harvard and MIT, Cambridge, MA 02142, USA.Corresponding author. E-mail: wendy_garrett@dfci.harvard.edu

CANCER IMMUNOLOGY AND IMMUNOTHERAPY on S

eptember 15, 2020

http://science.sciencem

ag.org/D

ownloaded from

group Escherichia coli (10) as well as by otherEnterobacteriaceae (11)], Bacteroides fragilistoxin (Bft) produced by enterotoxigenicB. fragilis,and cytolethal distending toxin (CDT) producedby several e- and g-proteobacteria. Colibactin hasemerged as a molecule of interest in colorectalcarcinogenesis, given the detection of pks+ E. coliin human colorectal cancers and the ability ofcolibactin-expressingE. coli to potentiate intestinaltumorigenesis inmice (12, 13). Accumulating dataalso support a role for enterotoxigenic B. fragilisin both human and animal models of colon tu-mors (14–17). Both colibactin and CDT can causedouble-stranded DNA damage in mammaliancells (18). In contrast, Bft acts indirectly by elicit-ing high levels of reactive oxygen species (ROS),which in turn damage hostDNA (19). Chronicallyhigh ROS levels can outpace a host’s DNA repairmechanisms, leading to DNA damage andmuta-tions (Fig. 2).Beyond damaging DNA, several microbes

possess proteins that engage host pathwaysinvolved in carcinogenesis. The Wnt/b-cateninsignaling pathway, which regulates cell stemness,polarity, and growth (20), is one example and isaltered in many malignancies. Several cancer-associated bacteria also can influence b-cateninsignaling (Fig. 2). Oncogenic type 1 strains ofHelicobacter pylori express a protein called CagA,which is injected directly into the cytoplasm ofhost cells and aberrantlymodulates b-catenin todrive gastric cancer (8). CagA-mediated b-cateninactivation leads to up-regulation of genes in-volved in cellular proliferation, survival, and mi-gration, as well as angiogenesis—all processescentral to carcinogenesis.Fusobacteriumnucleatumis a member of the oral microbiota and is associ-ated with human colorectal adenomas and adeno-carcinomas and amplified intestinal tumorigenesisin mice (21–24). F. nucleatum expresses FadA, abacterial cell surface adhesion component thatbinds host E-cadherin, leading to b-catenin acti-vation (25). Enterotoxigenic B. fragilis, whichis enriched in somehumancolorectal cancers (14),can stimulate E-cadherin cleavage via Btf, leadingto b-catenin activation (26). Salmonella typhistrains that maintain chronic infections secreteAvrA, which can activate epithelial b-catenin sig-naling (27, 28), and are associated with hepato-biliary cancers (29–31).

This phenomenon of activating b-catenin sig-naling reflects an interesting convergence of evo-lution, as several of these bacteria are normalconstituents of the human microbiota. Althoughmicrobial engagement of b-catenin signaling mayreflect a drive to establish a niche in a new tissuesite, the presence of these cancer-potentiatingmicrobes and their access to E-cadherin in evolv-ing tumors demonstrate that a loss of appropri-ate boundaries and barrier maintenance betweenhost and microbe is a critical step in the develop-ment of some tumors (Figs. 1 and 2).

The immune system, microbes,microbiota, and cancer

Mucosal surface barriers permit host-microbialsymbiosis (32); they are susceptible to constantenvironmental insult and must rapidly repairto reestablishhomeostasis. Compromised resiliencyof the host or microbiota can place tissues on apath to malignancy. Cancer and inflammatorydisorders can arise when barriers break downand microbes and immune systems find them-selves in geographies and assemblages for whichtheyhavenot coevolved.Oncebarriers arebreached,microbes can further influence immune responsesin evolving tumor microenvironments by elicit-ing proinflammatory or immunosuppressiveprograms (Fig. 2).

Proinflammatory responses canbe procarcinogenic

Both the chronic, high-grade inflammation ofinflammatory disorders (e.g., inflammatory boweldisease) and the lower-grade smoldering in-flammation of malignancies and obesity drivea tumor-permissive milieu. Inflammatory factorssuch as reactive oxygen and nitrogen species,cytokines, and chemokines can contribute totumor growth and spread (Fig. 2). Data fromhuman tissues and animal models show thattumors can up-regulate and activate many pat-tern recognition receptors, including Toll-likereceptors (3, 8). Activation of these receptorsresults in feedforward loops of activation ofNF-kΒ, a master regulator of cancer-associatedinflammation (33) (Fig. 2). Numerous cancer-associated microbes appear to activate NF-kΒsignaling within the tumor microenvironment[e.g., the colon cancer–associated F. nucleatum(23)]. The activation of NF-kΒ by F. nucleatummay be the result of pattern recognition recep-tor engagement (10, 34–37) or FadA engagementof E-cadherin (25). Other pattern recognitionreceptors, such as the nucleotide-binding oligo-merization domain–like receptor (NLR) familymembers NOD-2, NLRP3, NLRP6, and NLRP12,may play a role in mediating colorectal cancer;mice deficient in these NLRs display an enhancedsusceptibility to colitis-associated colorectal can-cer (caCRC) (38–44).Engagement of the immune system within

the tumor microenvironment is not restrictedto the innate immune system. Once barriers arebreached and the innate immune system is acti-vated, subsequent adaptive immune responsesensue, often with deleterious consequence for

SCIENCE sciencemag.org 3 APRIL 2015 • VOL 348 ISSUE 6230 81

Altered

balance of

host cell

proliferation

and death

Altered

immune

system

function

Altered

host and

microbial

metabolism

Persistent barrier breach

Healthy tissue

Barrier breach

Infection, trauma, dietary factors,

germline mutations

Failure to re-attain healthy host and

microbial homeostasis

Aerodigestive tract lumen

MU

CO

SA

L B

AR

RIE

R

Microbiota

Impaired host and microbial resiliency

MU

CO

SA

L B

AR

RIE

R

Fig. 1. The path from health to solid tumor ma-lignancies at mucosal sites and the microbiota’scontribution. Human body surfaces are subjectto constant environmental insult and injury. In-fections, trauma, dietary factors, and germlinemutations can contribute to breach of the body’smucosalbarriers. Inmost individuals, barrierbreachesare rapidly repaired and tissue homeostasis is re-stored. Impaired host or microbial resiliency contrib-utes to persistent barrier breach and a failure torestore homeostasis. In these settings, the micro-biota may influence carcinogenesis by (i) alteringhost cell proliferation and death, (ii) perturbing im-mune system function, and (iii) influencing metab-olism within a host.

on Septem

ber 15, 2020

http://science.sciencemag.org/

Dow

nloaded from

tumor progression. The interleukin-23 (IL-23)–IL-17 axis (45), tumor necrosis factor–a (TNF-a)–TNF receptor signaling (3, 5, 6, 46), IL-6–IL-6family member signaling (46, 47), and STAT3activation (48, 49)—an output of these cytokine-mediated signaling pathways—all represent in-nate and adaptive pathways contributing to tumorprogression and growth (Fig. 2).The microbiota is responsive and adapts to

changes in its host, such as inflammation. Ad-aptation to new selective pressures may resultin a microbiota at a tissue site that is not wellsuited for barrier repair, immune homeostasis,or maintenance of traditional host and microbeboundaries. Mouse models of caCRC furnishinsight in this regard. One such model usesazoxymethane, a genotoxin, and dextran sodi-um sulfate, a colon barrier–disrupting agent.Either agent alone results in colon tumors insusceptible mouse strains; using them togetheraccelerates tumorigenesis. Although this modeldoes not recapitulate the molecular and envi-ronmental events that lead to caCRC, it pro-vides an opportunity to study the convergenceof an environmental genotoxin, barrier disrup-tion, and severe chronic inflammation on cancerdevelopment.Microbiota transfer studies in caCRCmodels

support the idea that perturbations to a hostimmune system, either by genetic deletion orgenotoxin coupled with inflammatory stimulus,may select for microbiotas enriched for bacte-rial clades adept at attaching to host surfaces,invading host tissue, or triggering host inflam-matory mediators (21, 22, 40, 50, 51). Fecalmicrobiota fromNod2- orNlrp6-deficient miceacquire features that enhance the susceptibil-ity of wild-type mice to caCRC (40, 44). In mice,the gut microbiota modulate colon tumorigen-esis, independent of genetic deficiencies. Whengerm-free mice were colonized with microbes

82 3 APRIL 2015 • VOL 348 ISSUE 6230 sciencemag.org SCIENCE

DNA damage

Pro-inflammatory programs

CagA AvrA

A

B

C

Toxin-mediated

direct DNA damage

Toxin

Reactive oxygen andnitrogen species–mediated

indirect DNA damage

Altered -catenin signaling

CA

NC

ER

D

EV

EL

OP

ME

NT

&

P

RO

GR

ES

SIO

N

O2-

O2-

O2-

O2-

NO

NO

NO

NO

NF-κB activation

STAT3 activation

β-catenin β-catenin

β-catenin

β-catenin

Cell growthStemnessPolarity

Pattern recognition

receptor engagement

Cytokine signaling

IL-23/IL-17,

IL-6 family, TNF-α

E-cadherin

TCF

Fig. 2. Mechanisms by which microbes influ-ence cancer development and progression. (A)Bacterial toxins can directly damage host DNA.Bacteria also damage DNA indirectly via host-produced reactive oxygen and nitrogen species.When DNA damage exceeds host cell repair ca-pacity, cell death or cancer-enabling mutationsoccur. (B) b-Catenin signaling alterations are afrequent target of cancer-associated microbes.Some microbes bind E-cadherin on colonic epi-thelial cells, with altered polarity or within a dis-rupted barrier, and trigger b-catenin activation.Other microbes inject effectors (e.g., CagA orAvrA) that activate b-catenin signaling, resultingin dysregulated cell growth, acquisition of stemcell–like qualities, and loss of cell polarity. (C) Pro-inflammatory pathways are engaged upon muco-sal barrier breach in an evolving tumor. Loss ofboundaries between host and microbe engagespattern recognition receptors and their signalingcascades. Feedforward loops of chronic inflamma-tion mediated by NF-kB and STAT3 signaling fuelcarcinogenesis within both transforming and non-neoplastic cells within the tumors.

CANCER IMMUNOLOGY AND IMMUNOTHERAPY on S

eptember 15, 2020

http://science.sciencem

ag.org/D

ownloaded from

from donors with or without caCRC, followedby treatments that induced caCRC, those re-cipients that received gut microbiomes fromcaCRC-bearing mice developed more tumors(51). Similar mouse experiments using fecal trans-fers from humans with colon cancer suggestthat there are microbiome structures, both pro-tective and risk-elevating, that influence tumor-igenesis (52).Inflammation also results in the generation

of respiratory electron acceptors such as nitrate,ethanolamine, and tetrathionate, which somebacterial clades can use for their own fitnessadvantage (53–59). Several bacteria (e.g., E. coliand Salmonella spp.) can use these electronacceptors and also possess the key features thatreinforce the chronic inflammatory programsthat can enhance cancer growth and spread.However, it remains to be determined whether

bacterial use of these electron acceptors enhancescancer growth.

Immune-dampening responses canbe cancer-permissive

Microbes not only trigger and reinforce pro-inflammatory immune circuits but also exploitor elicit immunosuppressive responses. Amicrobemay take advantage of preexisting immuno-suppressionor elicit immune-dampening responsesto avoid destruction. Chronic systemic immuno-suppression, as seenwith advancedHIV infection,increases the risk for many cancers, especiallyvirally associated malignancies. Microbial-elicitedimmunosuppression can also contribute to im-paired antitumor immunity.Most current cancer-directed immunotherapies are focused on rousingimmune responsiveness to tumors (60). The coloncancer–associated bacterium F. nucleatum may

directly inhibit antitumor immunity by engagingTIGIT, a receptorwith immunoglobulin and ITIMdomains expressed on some T cells and naturalkiller cells, and blocking its ability to kill tumorcells (61). Whethermicrobes contribute to immu-notherapeutic resistance in other cancers remainsto be investigated.

Interrogating the role of microbesand microbiotas in cancer with newand old technologies

Microbiota studies in cancer remain at an earlystage. Information gathering and descriptivestudies are still necessary, and many criticalquestions remain.What other mechanismsmightmicrobes use to influence tumorigenesis? If sin-gle microbes can compromise antitumor immu-nity or enhance susceptibility to oncomicrobes,are there configurations of the microbiota thatdo this, too (or are protective)? Are there mi-crobes or microbiotas that enhance responsive-ness to immunotherapies or other therapeuticinterventions? To answer these questions, it isimportant to identify the key next steps in un-derstanding how the humanmicrobiota affectstumor growth and spread.Sequencing-based technologies are a boon to

both cancer biology and microbiology. Cancergenomes and their functional analyses have ledto the implementation of precision medicine ap-proaches to cancer care. Efforts to sequence in-dividual microbes and human microbiomes areproviding insight into how they influence humanhealth and disease. Computational tools that iden-tify microbial data within human sequencingdata sets are welcome new additions to the ar-mamentarium of cancermicrobe hunters (62, 63).Despite the affordable price of sequencing,

advances in culture techniques (64–66), andhigh-throughput analysis pipelines, the path ofcancer microbiome discovery is fraught withpitfalls. Cancers may develop over decades, anddifferent microbes and microbiotas may par-ticipate at distinct stages of the neoplastic pro-cess. For many malignancies, by the time acancer is detected, the window of opportunityfor identifying the inciting microbial agent(s)may have passed, allowing these organisms toremain elusive. However, the microbiota shouldremain a focus of study in locally advanced andmetastatic cancer, as microbes may contributeto an established cancer’s continued growthand spread.Beyond sequencing, microscopy and flow

cytometry–based approaches are useful tools todetect and study tumor-associated microbio-tas. Human colon tumors may harbor specificconsortia of bacteria that assemble themselvesinto biofilms (17). These biofilms appear to bespecific to certain biogeographies within thegastrointestinal tract and have members thathave been associated with colorectal adenomasand adenocarcinomas in human and mousestudies (e.g., enterotoxigenic B. fragilis andF. nucleatum). Microbiological studies of the oralcavity have shed light on microbial biofilms andtheir roles in human health and disease (67, 68).

SCIENCE sciencemag.org 3 APRIL 2015 • VOL 348 ISSUE 6230 83

Colonmicrobiota

metabolizes fiberButyric acid

Butyricacid

Increased cellular

proliferation

Enhanced tumorigenesis

Apoptosis

Tumor suppression

Intranuclear butyrate

accumulation

Inhibition of histone

deacetylases

Glycolysis and

lactic acid fermentation

Fatty acid

oxidation

Healthy colonocytes

Study with azoxymethane and dextran sodium sulfate–treated mice with specialized diets and defined microbes

Neoplasticcolonocytes

O

OH

Study with mutations

in Msh2-/-, ApcMin+/-

Fig. 3. Dietary fiber,microbiota, butyrate, and tumorigenesis.Metabolismof fiber by colonicmicrobesresults in generation of butyric acid. When genetic mutations in Msh2 and Apc are present, butyrateincreases cell proliferation and enhances tumorigenesis. Data from another model of colorectal carcino-genesis indicate the opposite outcome: Neoplastic colonocytes engage in glycolysis for cellular energy,unlike healthy colonocytes (which favor fatty acid oxidation). As a result, butyrate accumulates in thenucleus of neoplastic cells, engaging tumor-suppressive pathways and apoptosis.

on Septem

ber 15, 2020

http://science.sciencemag.org/

Dow

nloaded from

Within biofilms, microbial cross-feeding andco-metabolism occur (69). Consortia of tumor-associated microbes have the potential to gener-ate metabolites that require collective microbialmetabolism, and these co-metabolites may con-tribute to or halt carcinogenesis. The role ofmicrobial metabolism in host physiology is anexciting area, with several recent studies re-examining the role of microbial metabolites incancer (4, 70).

Microbes, metabolism, and cancer

In 1956, Warburg put forth the hypothesis thataltered cellular metabolism is the root cause ofcarcinogenesis (71), and cancer cell metabolismis currently a promising therapeutic target (72).Microbes participate in a range of host metabolicactivities.Microbialmetabolites or co-metabolites(generated with contributions from both hostandmicrobe) can contribute to inflammatory toneand can influence the balance of proliferation andcell death in tissues (4). Consideration of the ef-fects of a microbiota’s metabolism, and specif-icallymicrobialmetabolites generatedwithin thetumor microenvironment, on cancer growth andspread adds another therapeutic and diagnosticangle for targeting cancers through metabolicalterations.

A meal fit/unfit for a tumor:Fiber and fats

What defines a microbial oncometabolite (73),and how are such metabolites generated? Boththe host and its microbes affect themetabolismof dietary fiber, fats, ethanol, and phytoestro-gens. As with microbes, metabolites can affectimmune cell function, barrier function, and cellproliferation and death. Metabolites generatedfrom dietary fiber and fats that have an estab-lished effect on cancer are considered below,along with recent insights.Intestinal fermentation of dietary fiber by

members of the colonic microbiota results inthe generation of several short-chain fatty acids(SCFAs) including acetic, propionic, and butyricacids. These SCFAs have a range of effects onmany cell types, including anti-inflammatoryeffects on myeloid cells (74) and colonic regu-latory T cells (75–77), with consequences for in-tratumoral inflammation. SCFA’s effects maybe tuned by the receptors that they bind (e.g.,Niacr1/Gpr109a, Gpr43, Gpr41, or Olfr78). Gpr109ais a receptor for niacin and butyrate. It plays animportant role in mediating the effects of die-tary fiber and the microbiota in the colon, whereit is expressed by both colonic epithelial cellsand intestinalmyeloid cells. Activation ofGpr109aby butyrate results in anti-inflammatory hostresponses inmyeloid cells that lead to regulatoryT cell generation, and loss of Gpr109a increasessusceptibility to caCRC (78).SCFAs also affect host gene expression pat-

terns, cell proliferation, and cell death via bothreceptor-mediated and receptor-independentmechanisms. SCFAs and their activation of Gpr43reduce the proliferation rate of leukemia cells(79). In a study of ~70 human colon adenocarci-

nomas, GPR43 expression was reduced in cancerversus healthy tissue; restoration of GPR43 in ahuman colon cancer line increased apoptoticcell death upon SCFA exposure (80).SCFAs’ effects on host cellular processes vary

according to concentration and host genotype.Two recent mouse studies, which arrived at dif-ferent conclusions regarding the relationship ofdietary fiber, the microbiota, and butyrate tocolorectal tumorigenesis, reflect this heteroge-neous response to SCFAs (Fig. 3). Dietary fiberand butyrate-producing bacteria suppressedtumors in mice that harbored strictly definedmicrobial communities, received specialized diets,and were treated with azoxymethane and dex-tran sodium sulfate (81). This study’s data sup-ported amodelwherein the glycolyticmetabolismof cancer cells resulted in reduced metabolismof butyrate and enhanced butyrate nuclear ac-cumulation. High intranuclear butyrate levelsincreased histone acetylation and led to increasedapoptosis and reduced cellular proliferation. Ina mouse model of intestinal tumorigenesis driv-en by mutations in both the Apc gene and themismatch repair gene Msh2, the microbiotaand butyrate had tumor-promoting effects (82).Butyrate’s principal effect in this model systemwas to drive a hyperproliferative response inMsh2-deficient epithelial cells. Cancer geneticsand butyrate concentrations were critical factorsin SCFAs’ disparate effects on tumorigenesis be-tween these studies. These studies underscore

the challenges of translating microbiome, diet,and cancer basic science data into consensusguidelines for dietary interventions to reducecancer risk. Given that a single microbial me-tabolite can mediate a range of effects in tumormodels, investigators will require additional ex-perimental systems to unravel the effects of thehuman-microbial meta-metabolome for healthand cancer susceptibility.In contrast with the conflicting basic science

and epidemiological data surrounding dietaryfiber (83), there is consensus that high satu-rated fat intake heightens cancer risk. Debatesurrounding a high-fat diet (HFD) focuses onseveral mechanisms that may act alone or incombination, involving obesity, the microbiome,bile acids, and inflammation. There are a myr-iad of studies exploring the interconnectionbetween obesity and malignancy (84–86). Obe-sity is now regarded as an inflammatory state(87), and we are learning more about the gutmicrobiome’s contribution to obese and leanstates (88, 89). Data support the idea that in-flammation, the microbiota, and obesity con-stitute an inseparable trio that fuels cancer.However, a recent study suggests otherwise. In amousemodel of duodenal hyperplasia, adenomas,and invasive cancer driven by k-ras mutation,HFD and microbial dysbiosis amplified tumorgrowth and spread in the absence of obesity orthe development of a robust proinflammatoryresponse (90); mutated k-rasmodulated Paneth

84 3 APRIL 2015 • VOL 348 ISSUE 6230 sciencemag.org SCIENCE

Chemotherapy ImmunotherapyOxaliplatin

ROS

HN

Pt

Cl

NH

N

NHO

O

P

O O

O O

Cyclophosphamide CpG oligonucleotide

Small intestinal

barrier breach

Responsive to

immunotherapy

Impaired

response to

immunotherapy

No oral antibiotics

Subcutaneous

tumor

Oral antibiotics

Gut microbiotadisturbance

Bacterial

translocation

Immune activation

TH1 and TH17

antitumor response

2

2

O2-

Fig. 4. How the microbiota modulate chemotherapy and immunotherapy efficacy in mouse mod-els.The gut microbiota stimulate immune cells to produce reactive oxygen species (ROS). ROS enhanceDNA damage caused by oxaliplatin, blocking DNA replication and transcription and resulting in cell death.Cyclophosphamide can cause small intestinal barrier breach. This barrier disruption results in bacterialtranslocation that potentiates antitumor TH1 and TH17 responses. CpG oligonucleotides are a microbial-associated molecular pattern and are used in immunotherapy. Antibiotic disruption of the gut microbiotain mice compromised the efficacy of CpG in a mouse subcutaneous tumor model.

CANCER IMMUNOLOGY AND IMMUNOTHERAPY on S

eptember 15, 2020

http://science.sciencem

ag.org/D

ownloaded from

cell antimicrobial expression and HFD affectedintestinal mucin expression, thereby altering theintestinal microbiota. The fecal microbiota ofHFD k-rasmutant mice was sufficient to trans-mit the cancer-potentiating effects of the HFDwhen transferred to antibiotic-treated k-rasmu-tant mice.Another mechanism by which HFD influences

cancer risk is via bile acids that are producedto solubilize and digest the consumed fats—specifically, the microbially generated secondarybile acids. The role of secondary bile acids inincreased or decreased cancer risk has beenstudied for decades (2). One recent study pro-vided new insight into deoxycholic acid’s pro-oncogenic mechanisms in liver cancer: HFD orgenetic susceptibility to obesity can increase de-oxycholic acid–mediated activation of a mito-genic and proinflammatory response programin hepatic stellate cells, thereby potentiatingliver cancer in mice (91). These studies reinforcethe importance of gene-environment interac-tions in carcinogenesis and underscore the needto consider how dietary patterns influence thegenomes and genomic outputs of both hostand microbiome in mitigating or amplifyingcancer risk.

Drugs, bugs, and cancer

The gutmicrobiota function in drugmetabolism,influencing toxicity and efficacy (92, 93). Becausechemotherapeutic agents have a narrow thera-peutic window, there is interest in the micro-biota’s modulation of chemotherapy toxicity andefficacy (Fig. 4). Irinotecan is a topoisomerase-1inhibitor that is used in combination with otherchemotherapies to treat several cancers. A com-mon side effect is diarrhea. For some patients,the severity of the diarrhea requires hospitaliza-tion. Microbial-produced b-glucuronidases regu-late levels of irinotecan’s bioactive form withinthe intestinal lumen and thus influence irinote-can’s toxicity (94). Oral bacterial b-glucuronidaseinhibitors blunt the dose-limiting toxicities ofirinotecan in mice and do not harm host cells orkill bacteria, which suggests that microbial me-tabolism is a plausible target in cancer care (95).The gut microbiota also affect the efficacy of

chemotherapy. Oxaliplatin is a platinum-basedchemotherapyused to treat several gastrointestinalmalignancies. Together, the microbiota and im-mune system contribute to oxaliplatin’s efficacy(96). The gut microbiota prime myeloid cells forhigh-level ROS production. The resultant intra-tumoral oxidative stress augments oxaliplatin-associated DNA damage, triggering cancer celldeath (96). Cyclophosphamide, an alkylatingagent used in hematologic malignancies andsolid tumors, can injure the small intestinalepithelium. The ensuing barrier breach results ingut microbiota–dependent, T helper (TH) cell–mediated antitumor responses (97). Delineatingthe roles of gut microbiota in response to che-motherapy in model systems and undertakingepidemiologic studies with microbiome analysisin patients with and at risk for cancer will becritical for realizing the microbiota as an adju-

vant therapy that enhances efficacy or attenuatestoxicity of chemotherapies.

The microbiota and immunotherapy:Friend or foe?

The success of immunotherapy (in the form ofcytokine therapy, targeting immune checkpointblockade, and vaccine therapy) has been one ofthe most exciting developments in cancer careover the past decade (98). Given the intertwinednature of the microbiota and the immune sys-tem, it is plausible that the microbiota influencea host’s responsiveness to immunotherapy. Insupport of this idea, antibiotic-mediated disrup-tion of the microbiota in mice bearing subcu-taneous tumors impaired the effectiveness of CpGoligonucleotide immunotherapy (Fig. 4) (96).Observations that immunotherapies are show-ing efficacy in melanoma and bladder, renal, andlung cancer but not in cancer of the colon (whichis densely populated by bacteria) fuel interest inhow the microbiota contributes to immuno-therapy’s efficacy. Furthermore, given the severecolitis observed in some patients receiving im-munotherapies (99) (e.g., antibodies to CTLA4and PD-L1) and the role of gut microbes in colitis,it is possible that the gut microbiota influencesthis toxicity. As patient populations expand,investigators will hopefully interrogate whetherthere are microbiota that are predictive for coli-tis and other toxicities. Examining the micro-biota and its effects on immunotherapy efficacyand toxicity in preclinical models and patientsis a critical next step.

Hematopoietic transplants,complications, and the microbiota

Allogeneic hematopoietic stem cell transplant(allo-HSCT), a mainstay in hematologic malig-nancy treatment, is a challenge to both hostand microbiota. An individual’s microbiota isconfronted with a new host within its host aswell as chemotherapy, radiation, oral and gastro-intestinal barrier breach, and broad-spectrumantibiotics. Studies have begun to examine per-turbations to the gut microbiota and clinical out-comes during allo-HSCT (100).Bacteremia, Clostridium difficile infection,

and graft-versus-host disease (GVHD) are com-mon events in allo-HSCT patients. Bacteremiaswith vancomycin-resistant Enterococcus (VRE)are a grave concern. Two preclinical studies ex-amining how antibiotics perturb the gut micro-biota to enable VRE displacement of a healthymicrobiota (101) and how the anaerobic bacteriaBarnesiella spp. may confer resistance to VRE(102) have provided mechanistic insight intothese bloodstream infections. These studies setthe stage for a clinical study showing that en-terococcal gut microbiota domination was as-sociated with a factor of 9 higher risk of VREbacteremia in allo-HSCT patients (103). Hospi-talized patients and allo-HSCT patients bothconfront toxigenic C. difficile infection. Usingmouse models, microbiome analysis, and allo-HSCT patient populations, researchers identi-fied amicrobe that can restore bile acid–mediated

resistance to C. difficile (104). The workflows ofthis precision medicine–based study are appli-cable to many diseases associated with alteredmicrobiotas.Allo-HSCT patients can experience gastro-

intestinal, pulmonary, and skin complicationsafter transplant; some of these are idiopathicclinical syndromes while others are GVHDman-ifestations. Using shotgun DNA sequencing ofcolon tissue and the PathSeq pipeline, investi-gators found that Bradyrhizobium enterica wasenriched in affected colonic tissue from patientswith idiopathic colitis after receiving a cord bloodtransplant (105), providing insight and a potentialtreatment. Using samples frommice and humansthat had undergone allogeneic bone marrowtransplants, investigators characterized the gutmicrobiota changes in active intestinal GVHD(106). In mice, depletion of lactobacilli exacer-bated GVHD-associated intestinal inflammationand their reintroduction attenuated inflam-mation (106). The challenge intrinsic to thesestudies, and realized in (104), is to use our evolv-ing knowledge of the microbiome and microbesto identify bacteriotherapy for cancer and itscomplications.

Back to the future:Perspectives and directions forcancer bacteriotherapy

The genesis of immunotherapy came from anappreciation for the co-adaptation between hostand microbe. Exploiting this knowledge andusing bacteria to trigger the immune system toattack and destroy cancers dates back to the1850s, when several German physicians noticedthat some cancer patients with active infectionsshowed signs of tumor regression. This led Coleyto test bacterial extracts in patients with bonecancers around 1900. Heat-killed cultures ofStreptococcus pyogenes and Serratia marcescens,or Coley’s toxins, were one of earliest forms ofimmunotherapy (60). Since this seminal work,one bacterium has entered the mainstream ofcancer treatment. For the past three to fourdecades, Bacillus Calmette-Guerin (BCG) hasbeen used to treat non–muscle-invasive bladdercancer. The live bacteria, which are delivered di-rectly into the bladder, elicit inflammation thattriggers an antitumor immune response (107).Much still remains to be learned about the im-mune response to BCG and antitumor immunity,and why BCG loses efficacy once the cancer ismore invasive (108).Over the past 30 years, several bacterial-based

approaches to cancer therapy have emerged.Bacterial-based vaccines that express tumor an-tigens have shown efficacy in preclinical studies,and recombinant Listeria monocytogenes–basedvaccines showed tremendous promise in mice(109). Interest remains in using bacteria as adelivery vehicle for plant toxins, such as ricinand saporin, or pseudomonal exotoxins that canblock protein synthesis and induce apoptosisin cancer cells (110). Bacteria have evolved ele-gant systems to communicate with each other,to kill one another (111), and to deliver their

SCIENCE sciencemag.org 3 APRIL 2015 • VOL 348 ISSUE 6230 85

on Septem

ber 15, 2020

http://science.sciencemag.org/

Dow

nloaded from

effectors into host cells (112). The extension andapplication of these secretion systems, whichhave been honed bymillennia of evolution, seemslike a therapeutic slam dunk but has been chal-lenging in practice. A recent study in dogs (113)has breathed new life into the concept of bacterio-therapy with Clostridium novyii, which emergedas a promising concept in preclinical models al-most 15 years ago (114); however, balancing tox-icity with efficacy remains difficult.Synthetic biology approaches to cancer care

hold enormous potential, especially those thatmake use of bacteria. These efforts involve thereengineering of bacterial cells for the deliveryof biomolecules under tunable networks andon/off toggle switches triggered by host re-sponses (115). The goals are simple: to targetcancers and minimize damage to healthy tissuesvia genetic network designs informed by engi-neering principles. Proof of concept that de-signer microbes can invade cancer cells (116) totarget and perturb key cancer pathways hasbeen established (117). Evaluation in robust pre-clinical models will be the next step. Applicationand design for cancer care will need to focus onmaximizing anticancer responses while mini-mizing toxicities and infectious complications.Like synthetic biology, microbiome studies

have emerged as a promising area of investiga-tion for cancer care over the past decade. Themicrobiome may afford many answers to sev-eral looming questions in cancer biology: Whatare the critical gene-environmental interactionsin cancer susceptibility? Why do certain foods ordietary patterns confer increased or decreasedrisk in certain populations and individuals? Whydo chemotherapies, immunotherapies, and pre-ventive agents fail or succeed for patients, irre-spective of host germline or cancer genotype?The microbiome seems to provide many poten-tial answers in the forms of select clades, con-sortia, metabolites, and enzymatic activities, butit remains unclear whether and how these willtranslate from preclinical models to humans.One opportunity for the microbiota in the nearterm is as a biomarker for diagnosis (118), prog-nostication, or identifying those most at riskfor treatment-related complications. Althoughthere may be dissent about the best next steps,there is consensus that therapeutic considera-tion of cancer and the microbiota requires amultidisciplinary approach andmore intensiveinvestigation.

REFERENCES AND NOTES

1. C. de Martel et al., Lancet Oncol. 13, 607–615 (2012).2. C. L. Sears, W. S. Garrett, Cell Host Microbe 15, 317–328

(2014).3. R. F. Schwabe, C. Jobin, Nat. Rev. Cancer 13, 800–812

(2013).4. P. Louis, G. L. Hold, H. J. Flint, Nat. Rev. Microbiol. 12,

661–672 (2014).5. E. Elinav et al., Nat. Rev. Cancer 13, 759–771 (2013).6. T. Irrazábal, A. Belcheva, S. E. Girardin, A. Martin,

D. J. Philpott, Mol. Cell 54, 309–320 (2014).7. J. Kallmeyer, R. Pockalny, R. R. Adhikari, D. C. Smith,

S. D’Hondt, Proc. Natl. Acad. Sci. U.S.A. 109, 16213–16216(2012).

8. M. T. Abreu, R. M. Peek Jr., Gastroenterology 146,1534–1546.e3 (2014).

9. A. I. Ojesina et al., Nature 506, 371–375 (2014).10. J.-P. Nougayrède et al., Science 313, 848–851 (2006).11. J. Putze et al., Infect. Immun. 77, 4696–4703 (2009).12. J. C. Arthur et al., Nat. Commun. 5, 4724 (2014).13. J. C. Arthur et al., Science 338, 120–123 (2012).14. A. Boleij et al., Clin. Infect. Dis. 60, 208–215 (2015).15. C. L. Sears et al., Clin. Infect. Dis. 47, 797–803 (2008).16. S. Wu et al., Nat. Med. 15, 1016–1022 (2009).17. C. M. Dejea et al., Proc. Natl. Acad. Sci. U.S.A. 111,

18321–18326 (2014).18. L. Guerra, R. Guidi, T. Frisan, FEBS J. 278, 4577–4588 (2011).19. A. C. Goodwin et al., Proc. Natl. Acad. Sci. U.S.A. 108,

15354–15359 (2011).20. H. Clevers, R. Nusse, Cell 149, 1192–1205 (2012).21. M. Castellarin et al., Genome Res. 22, 299–306 (2012).22. A. D. Kostic et al., Genome Res. 22, 292–298 (2012).23. A. D. Kostic et al., Cell Host Microbe 14, 207–215 (2013).24. A. N. McCoy et al., PLOS ONE 8, e53653 (2013).25. M. R. Rubinstein et al., Cell Host Microbe 14, 195–206

(2013).26. C. L. Sears, Clin. Microbiol. Rev. 22, 349–369 (2009).27. R. Lu et al., Oncogenesis 3, e105 (2014).28. R. Lu et al., Am. J. Physiol. Gastrointest. Liver Physiol. 303,

G1113–G1125 (2012).29. U. Dutta, P. K. Garg, R. Kumar, R. K. Tandon, Am. J.

Gastroenterol. 95, 784–787 (2000).30. E. C. Lazcano-Ponce et al., CA Cancer J. Clin. 51, 349–364

(2001).31. I. I. Wistuba, A. F. Gazdar, Nat. Rev. Cancer 4, 695–706

(2004).32. L. V. Hooper, D. R. Littman, A. J. Macpherson, Science 336,

1268–1273 (2012).33. J. A. DiDonato, F. Mercurio, M. Karin, Immunol. Rev. 246,

379–400 (2012).34. G. Zhang, R. Chen, J. D. Rudney, J. Periodontal Res. 46,

558–567 (2011).35. M. R. Milward et al., Clin. Exp. Immunol. 148, 307–324 (2007).36. E. Allen-Vercoe, J. Strauss, K. Chadee, Gut Microbes 2,

294–298 (2011).37. S.-R. Park et al., Infect. Immun. 82, 1914–1920 (2014).38. I. C. Allen et al., J. Exp. Med. 207, 1045–1056 (2010).39. G. Y. Chen, M. Liu, F. Wang, J. Bertin, G. Núñez, J. Immunol.

186, 7187–7194 (2011).40. B. Hu et al., Proc. Natl. Acad. Sci. U.S.A. 110, 9862–9867 (2013).41. B. Hu et al., Proc. Natl. Acad. Sci. U.S.A. 107, 21635–21640

(2010).42. M. H. Zaki, P. Vogel, M. Body-Malapel, M. Lamkanfi,

T.-D. Kanneganti, J. Immunol. 185, 4912–4920 (2010).43. M. H. Zaki et al., Cancer Cell 20, 649–660 (2011).44. A. Couturier-Maillard et al., J. Clin. Invest. 123, 700–711

(2013).45. S. I. Grivennikov et al., Nature 491, 254–258 (2012).46. S. I. Grivennikov, M. Karin, Ann. Rheum. Dis. 70 (suppl. 1),

i104–i108 (2011).47. S. I. Grivennikov, Cancer Cell 24, 145–147 (2013).48. H. Yu, D. Pardoll, R. Jove, Nat. Rev. Cancer 9, 798–809 (2009).49. N. Li, S. I. Grivennikov, M. Karin, Cancer Cell 19, 429–431 (2011).50. R. L. Warren et al., Microbiome 1, 16 (2013).51. J. P. Zackular et al., MBio 4, e00692-13 (2013).52. N. T. Baxter, J. P. Zackular, G. Y. Chen, P. D. Schloss,

Microbiome 2, 20 (2014).53. F. Faber, A. J. Bäumler, Immunol. Lett. 162, 48–53 (2014).54. F. Rivera-Chávez et al., PLOS Pathog. 9, e1003267 (2013).55. S. E. Winter, A. J. Bäumler, Gut Microbes 5, 71–73 (2014).56. S.-P. Nuccio, A. J. Bäumler, MBio 5, e00929-14 (2014).57. S. E. Winter et al., Science 339, 708–711 (2013).58. P. Thiennimitr et al., Proc. Natl. Acad. Sci. U.S.A. 108,

17480–17485 (2011).59. S. E. Winter et al., Nature 467, 426–429 (2010).60. I. Mellman, G. Coukos, G. Dranoff, Nature 480, 480–489 (2011).61. C. Gur et al., Immunity 42, 344–355 (2015).62. A. D. Kostic et al., Nat. Biotechnol. 29, 393–396 (2011).63. D. R. Riley et al., PLOS Comput. Biol. 9, e1003107 (2013).64. A. L. Goodman et al., Proc. Natl. Acad. Sci. U.S.A. 108,

6252–6257 (2011).65. I. Vanwonterghem, P. D. Jensen, D. P. Ho, D. J. Batstone,

G. W. Tyson, Curr. Opin. Biotechnol. 27, 55–64 (2014).66. M. V. Sizova et al., Appl. Environ. Microbiol. 78, 194–203 (2012).67. C. J. Wright et al., Mol. Oral Microbiol. 28, 83–101 (2013).68. M. A. Curtis, C. Zenobia, R. P. Darveau, Cell Host Microbe 10,

302–306 (2011).69. K. E. Boyle, S. Heilmann, D. van Ditmarsch, J. B. Xavier,

Curr. Opin. Microbiol. 16, 207–212 (2013).

70. R. Medzhitov, D. S. Schneider, M. P. Soares, Science 335,936–941 (2012).

71. O. Warburg, Science 123, 309–314 (1956).72. M. G. Vander Heiden, Nat. Rev. Drug Discov. 10, 671–684

(2011).73. S. J. Bultman, C. Jobin, Cell Host Microbe 16, 143–145

(2014).74. K. M. Maslowski et al., Nature 461, 1282–1286 (2009).75. P. M. Smith et al., Science 341, 569–573 (2013).76. N. Arpaia et al., Nature 504, 451–455 (2013).77. Y. Furusawa et al., Nature 504, 446–450 (2013).78. N. Singh et al., Immunity 40, 128–139 (2014).79. L. B. Bindels et al., Br. J. Cancer 107, 1337–1344 (2012).80. Y. Tang, Y. Chen, H. Jiang, G. T. Robbins, D. Nie, Int. J. Cancer

128, 847–856 (2011).81. D. R. Donohoe et al., Cancer Discov. 10.1158/2159-8290.CD-

14-0501 (2014).82. A. Belcheva et al., Cell 158, 288–299 (2014).83. M. Song, W. S. Garrett, A. T. Chan, Gastroenterology

10.1053/j.gastro.2014.12.035 (2015).84. R. W. O’Rourke, Surg. Obes. Relat. Dis. 10, 1208–1219 (2014).85. J. Park, T. S. Morley, M. Kim, D. J. Clegg, P. E. Scherer,

Nat. Rev. Endocrinol. 10, 455–465 (2014).86. N. A. Berger, Ann. N.Y. Acad. Sci. 1311, 57–76 (2014).87. M. F. Gregor, G. S. Hotamisligil, Annu. Rev. Immunol. 29,

415–445 (2011).88. R. E. Ley, Curr. Opin. Gastroenterol. 26, 5–11 (2010).89. P. J. Turnbaugh et al., Nature 457, 480–484 (2009).90. M. D. Schulz et al., Nature 514, 508–512 (2014).91. S. Yoshimoto et al., Nature 499, 97–101 (2013).92. A. D. Patterson, P. J. Turnbaugh, Cell Metab. 20, 761–768

(2014).93. R. N. Carmody, P. J. Turnbaugh, J. Clin. Invest. 124,

4173–4181 (2014).94. A. B. Roberts, B. D. Wallace, M. K. Venkatesh, S. Mani,

M. R. Redinbo, Mol. Pharmacol. 84, 208–217 (2013).95. B. D. Wallace et al., Science 330, 831–835 (2010).96. N. Iida et al., Science 342, 967–970 (2013).97. S. Viaud et al., Science 342, 971–976 (2013).98. J. Couzin-Frankel, Science 342, 1432–1433 (2013).99. C. Gedye, A. van der Westhuizen, T. John, Intern. Med. J.

10.1111/imj.12653 (2014).100. Y. Taur et al., Blood 124, 1174–1182 (2014).101. C. Ubeda et al., J. Clin. Invest. 120, 4332–4341 (2010).102. C. Ubeda et al., Infect. Immun. 81, 965–973 (2013).103. Y. Taur et al., Clin. Infect. Dis. 55, 905–914 (2012).104. C. G. Buffie et al., Nature 517, 205–208 (2015).105. A. S. Bhatt et al., N. Engl. J. Med. 369, 517–528

(2013).106. R. R. Jenq et al., J. Exp. Med. 209, 903–911 (2012).107. M. A. Ingersoll, M. L. Albert, Mucosal Immunol. 6, 1041–1053

(2013).108. G. Redelman-Sidi, M. S. Glickman, B. H. Bochner, Nat. Rev.

Urol. 11, 153–162 (2014).109. Z. K. Pan, G. Ikonomidis, A. Lazenby, D. Pardoll, Y. Paterson,

Nat. Med. 1, 471–477 (1995).110. N. Nair, T. Kasai, M. Seno, Anticancer Res. 34, 6289–6296

(2014).111. B. T. Ho, T. G. Dong, J. J. Mekalanos, Cell Host Microbe 15,

9–21 (2014).112. J. E. Galán, H. Wolf-Watz, Nature 444, 567–573 (2006).113. N. J. Roberts et al., Sci. Transl. Med. 6, 249ra111 (2014).114. L. H. Dang, C. Bettegowda, D. L. Huso, K. W. Kinzler,

B. Vogelstein, Proc. Natl. Acad. Sci. U.S.A. 98, 15155–15160(2001).

115. W. C. Ruder, T. Lu, J. J. Collins, Science 333, 1248–1252(2011).

116. J. C. Anderson, E. J. Clarke, A. P. Arkin, C. A. Voigt, J. Mol.Biol. 355, 619–627 (2006).

117. S. Xiang, J. Fruehauf, C. J. Li, Nat. Biotechnol. 24, 697–702(2006).

118. J. P. Zackular, M. A. M. Rogers, M. T. Ruffin 4th, P. D. Schloss,Cancer Prev. Res. 7, 1112–1121 (2014).

ACKNOWLEDGMENTS

W.S.G. thanks C. Brennan, C. Gallini, G. Hold, C. Lesser, andM. Howitt for critical reading of the manuscript. Supported by NIHgrant R01CA154426, a Burroughs Wellcome Career in MedicalSciences Award, a Searle Scholars Award, a Cancer ResearchInstitute Investigator Award, and a research grant fromHoffman-LaRoche. W.S.G. is a Synlogic SAB member.

10.1126/science.aaa4972

86 3 APRIL 2015 • VOL 348 ISSUE 6230 sciencemag.org SCIENCE

CANCER IMMUNOLOGY AND IMMUNOTHERAPY on S

eptember 15, 2020

http://science.sciencem

ag.org/D

ownloaded from

Cancer and the microbiotaWendy S. Garrett

DOI: 10.1126/science.aaa4972 (6230), 80-86.348Science 

ARTICLE TOOLS http://science.sciencemag.org/content/348/6230/80

CONTENTRELATED

http://stm.sciencemag.org/content/scitransmed/7/271/271ps1.fullhttp://science.sciencemag.org/content/sci/348/6230/54.full

REFERENCES

http://science.sciencemag.org/content/348/6230/80#BIBLThis article cites 118 articles, 40 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2015, American Association for the Advancement of Science

on Septem

ber 15, 2020

http://science.sciencemag.org/

Dow

nloaded from