RESEARCH I REPORTS Nationale de la Recherche. AS was supported by BMSI YIG 2014. E.G. study have been submitted to the NCBI under theBicproject Figs. SI to s22 Reed Archive ables s1 to s5 sociation pour la Recherche contre le References(19-35) 200851) F.C. was supported by INCA-DGOS (GOLD sRR28006.SRR2758031,SRR275878.sRR275879.sRR2758180 2012-1-RT-14-GR-01). L'Oreal awarded a prize to M V. We are grateful SUPPLEMENTARY MATERIALS 21 October 2015 gelique, N. Chanthapathet, www.sciencemag.org/content/350/6264/1079/suppl/dcl ublished online 5 November 2015 and S. Zuberagoitia for technical help. DNA sequence reads from this Materials and Methods 10.1126/science aad1329 CANCER IMMUNOTHERAPY and JAX mice appeared to acquire the jAX pheno- type, which suggested that JAX mice may be col- Commensal bifidobacterium onized by commensal microbes that dominantly facilitate antitumor immunity To directly test the role of commensal bacteria promotes antitumor immunity and in regulating antitumor immunity, we transferred JAX or TAC fecal suspensions into TAC and JAX facilitates anti-PD-Ll efficacy recipients by oral gavage before tumor implan- tation (fig. SIA). We found that prophylactic trans- fer of JAX fecal material, but not saline or TAc Ayelet Sivan, Leticia Corrales, Nathaniel Hubert, Jason B. williams, fecal material, into TAC recipients was sufficient Keston Aquino-Michaels, " Zachary M. Earley, Franco w Benyamin, 'Yuk Man Lei, to delay tumor growth(Fig 2A)and to enhance Bana Jabri, Maria-Luisa Alegre, Eugene B Chang, Thomas F Gajewski,t T cells(Fig. 2, B and C, and fig SIB), which sup- s T cell infiltration of solid tumors is associated with favorable patient outcomes, yet the ported a microbe-derived effect. Reciprocal trans. mechanisms underlying variable immune responses between individuals are not well fer of TAC fecal material into JAX recipients had understood. One possible modulator could be the intestinal microbiota. We compared a minimal effect on tumor growth rate and anti- sD melanoma growth in mice harboring distinct commensal microbiota and observed differences in spontaneous antitumor immunity, which were eliminated upon cohousing SIB), consistent with the JAX-dominant effects g observed upon cohousing or after fecal transfer Sequencing of the 16S ribosomal RNA identified Bifidobacterium asTo test whether manipulation of the microbial associated with the antitumor effects Oral administration of Bifidobacterium alone improved tumor control to the same degree as programmed cell death protein l ligand community could be effective as a therapy, we ad- 1(PD-L1)-specific antibody therapy(checkpoint blockade), and combination treatment ministered jAX fecal material alone or in combi nearly abolished tumor outgrowth. Augmented dendritic cell function leading to enhanced nation with antibodies targeting PD-Ll(aPD-Ll) 35 CD8" T cell priming and accumulation in the tumor microenvironment mediated the effect. to TAc mice bearing established tumors.Trans- Our data suggest that manipulating the microbiota may modulate cancer immunotherapy. icantly slower tumor growth (Fig 2D, ccompanied a by increased tumor-specific T cell responses E armessing the host immune system consti- mediating immune activation in response to chemo(Fig. 2E)and infiltration of antigen-specific Tcells tutes a promising cancer therapeutic be- therapeutic agents has been demonstrated (10, Ir). into the tumor(Fig. 2F), to the same degree as auseofitspotentialtospecificallytargetHowever,itisnotknownwhethercommensaltreatmentwithsystemicapd-llmab.combina- tumor cells although limiting harm to nor. microbiota influence spontaneous immune re- tion treatment with both JAX fecal transfer and 3 mal tissue. Enthusiasm has been fueled sponses against tumors and thereby affect the aPD-LImAb improved tumor control (Fig 2D)and a specifically CTLA-4 and the axis between pro- antibodies(mAbs). on accumulation of activated t cells within the grammed cell death protein 1(PD-1) and its To address this question, we compared sub- tumor microenvironment (Fig. 2F). Consistent with igand1(PD-LD(1, 2).Clinical responses to these cutaneous B16 STY melanoma growthin genetically these results, aPD-LI therapy alone was signifi- are more frequent in patients similar C57BL/6 mice derived from two different cantly more efficacious in JAX mice compared who show evidence of an endogenous T cell re- mouse facilities, Jackson Laboratory (AX) and with TAC mice(Fig. 2G), which paralleled improved sponse ongoing in the tumor microenvironment Taconic Farms(TAC), which have been shown to antitumor T cell responses(fig. SIC). These data before therapy(3-6). However, the mechanisms differ in their commensal microbes(12). We found indicate that the commensal microbial compo- hat govern the presence or absence of this phe- that JAX and TAC mice exhibited significant sition can influence spontaneous antitumor im- notype are not well understood. Theoretical sources differences in B16 STY melanoma growth rate, munity, as well as a response to immunotherapy of interpatient heterogeneity include host germ- with tumors growing more aggressively in TAc with aPD-LI mAb ine genetic differences, variability in patterns of mice(Fig. 1A). This difference was immune- To identify specific bacteria associated with im- somatic alterations in tumor cells, and environ- mediated: Tumor-specific T cell responses(Fig. 1, proved antitumor immune responses, we moni mental differences B and C) and intratumoral CDS*T cell accumu- tored the fecal bacterial content over time of mice he gut microbiota plays an important role in lation(Fig. ID)were significantly higher in JAX that were subjected to administration of fecal shaping systemicimmune responses(7-9). In the than in TAC mice. To begin to address whether permutations, using the 16S ribosomal RNA(rRNA) cancer context, a role for intestinal microbiota in this difference could be mediated by commensal miSeq Illumina platform. Principal coordinate microbiota, we choused JAX and TAC mice be nalysis revealed that fecal samples analyzed from fore tumor implantation. We found that cohous- TAC mice that received JAX fecal material grad ent of Medicine, University of Chicago, ing ablated the differences in tumor growth(Fig. ually separated from samples obtained from sham- Chicago, IL 60637, USA 3Section of Genetic Medicine 1E) and immune responses(Fig. 1, F to h)be- d TAC feces-inoculated TAC "These authors contributed equally to this work. cOrresponding tween the two lations,which sug.(P=0.001 and P=0.003, respect NOSIM author. E-mail: tgajewsk@medicine bsd uchicago. ed gested an enviro nce Choused TAc I multivariate data analysis )and 1084 27 NOVEMBER 2015. VOL 350 ISSUE 6264 sciencemag. org SCIENCENationale de la Recherche. A.S. was supported by BMSI YIG 2014. F.G. is supported by SIgN core funding. L.Z., M.C., and I.B.G. are all sponsored by Association pour la Recherche contre le Cancer (PGA120140200851). F.C. was supported by INCA-DGOS (GOLD H78008). N.C. was supported by INCA-DGOS (GOLD study; 2012-1-RT-14-IGR-01). L’Oreal awarded a prize to M.V. We are grateful to the staff of the animal facility of Gustave Roussy and Institut Pasteur. We thank P. Gonin, B. Ryffel, T. Angelique, N. Chanthapathet, H. Li, and S. Zuberogoitia for technical help. DNA sequence reads from this study have been submitted to the NCBI under the Bioproject IDPRJNA299112 and are available from the Sequence Read Archive (SRP Study accession SRP065109; run accession numbers SRR2758006, SRR2758031, SRR2758178, SRR2758179, SRR2758180, SRR2758181, SRR2768454, and SRR2768457. SUPPLEMENTARY MATERIALS www.sciencemag.org/content/350/6264/1079/suppl/DC1 Materials and Methods Figs. S1 to S22 Tables S1 to S5 References (19–35) 3 April 2015; accepted 21 October 2015 Published online 5 November 2015; 10.1126/science.aad1329 CANCER IMMUNOTHERAPY Commensal Bifidobacterium promotes antitumor immunity and facilitates anti–PD-L1 efficacy Ayelet Sivan,1 * Leticia Corrales,1 * Nathaniel Hubert,2 Jason B. Williams,1 Keston Aquino-Michaels,3 Zachary M. Earley,2 Franco W. Benyamin,1 Yuk Man Lei,2 Bana Jabri,2 Maria-Luisa Alegre,2 Eugene B. Chang,2 Thomas F. Gajewski1,2† T cell infiltration of solid tumors is associated with favorable patient outcomes, yet the mechanisms underlying variable immune responses between individuals are not well understood. One possible modulator could be the intestinal microbiota. We compared melanoma growth in mice harboring distinct commensal microbiota and observed differences in spontaneous antitumor immunity, which were eliminated upon cohousing or after fecal transfer. Sequencing of the 16S ribosomal RNA identified Bifidobacterium as associated with the antitumor effects. Oral administration of Bifidobacterium alone improved tumor control to the same degree as programmed cell death protein 1 ligand 1 (PD-L1)–specific antibody therapy (checkpoint blockade), and combination treatment nearly abolished tumor outgrowth. Augmented dendritic cell function leading to enhanced CD8+ T cell priming and accumulation in the tumor microenvironment mediated the effect. Our data suggest that manipulating the microbiota may modulate cancer immunotherapy. H arnessing the host immune system consti￾tutes a promising cancer therapeutic be￾cause of its potential to specifically target tumor cells although limiting harm to nor￾mal tissue. Enthusiasm has been fueled by recent clinical success, particularly with anti￾bodies that block immune inhibitory pathways, specifically CTLA-4 and the axis between pro￾grammed cell death protein 1 (PD-1) and its ligand 1 (PD-L1) (1, 2). Clinical responses to these immunotherapies are more frequent in patients who show evidence of an endogenous T cell re￾sponse ongoing in the tumor microenvironment before therapy (3–6). However, the mechanisms that govern the presence or absence of this phe￾notype are not well understood. Theoretical sources of interpatient heterogeneity include host germ￾line genetic differences, variability in patterns of somatic alterations in tumor cells, and environ￾mental differences. The gut microbiota plays an important role in shaping systemic immune responses (7–9). In the cancer context, a role for intestinal microbiota in mediatingimmune activationin response to chemo￾therapeutic agents has been demonstrated (10,11). However, it is not known whether commensal microbiota influence spontaneous immune re￾sponses against tumors and thereby affect the therapeutic activity of immunotherapeutic inter￾ventions, such as anti–PD-1/PD-L1 monoclonal antibodies (mAbs). To address this question, we compared sub￾cutaneous B16.SIYmelanoma growthin genetically similar C57BL/6 mice derived from two different mouse facilities, Jackson Laboratory (JAX) and Taconic Farms (TAC), which have been shown to differ in their commensal microbes (12). We found that JAX and TAC mice exhibited significant differences in B16.SIY melanoma growth rate, with tumors growing more aggressively in TAC mice (Fig. 1A). This difference was immune￾mediated: Tumor-specific T cell responses (Fig. 1, B and C) and intratumoral CD8+ T cell accumu￾lation (Fig. 1D) were significantly higher in JAX than in TAC mice. To begin to address whether this difference could be mediated by commensal microbiota, we cohoused JAX and TAC mice be￾fore tumor implantation. We found that cohous￾ing ablated the differences in tumor growth (Fig. 1E) and immune responses (Fig. 1, F to H) be￾tween the two mouse populations, which sug￾gested an environmental influence. Cohoused TAC and JAX mice appeared to acquire the JAX pheno￾type, which suggested that JAX mice may be col￾onized by commensal microbes that dominantly facilitate antitumor immunity. To directly test the role of commensal bacteria in regulating antitumor immunity, we transferred JAX or TAC fecal suspensions into TAC and JAX recipients by oral gavage before tumor implan￾tation (fig. S1A). We found that prophylactic trans￾fer of JAX fecal material, but not saline or TAC fecal material, into TAC recipients was sufficient to delay tumor growth (Fig. 2A) and to enhance induction and infiltration of tumor-specific CD8+ T cells (Fig. 2, B and C, and fig. S1B), which sup￾ported a microbe-derived effect. Reciprocal trans￾fer of TAC fecal material into JAX recipients had a minimal effect on tumor growth rate and anti￾tumor T cell responses (Fig. 2, A to C, and fig. S1B), consistent with the JAX-dominant effects observed upon cohousing. To test whether manipulation of the microbial community could be effective as a therapy, we ad￾ministered JAX fecal material alone or in combi￾nation with antibodies targeting PD-L1 (aPD-L1) to TAC mice bearing established tumors. Trans￾fer of JAX fecal material alone resulted in signif￾icantly slower tumor growth (Fig. 2D), accompanied by increased tumor-specific T cell responses (Fig. 2E) and infiltration of antigen-specific T cells into the tumor (Fig. 2F), to the same degree as treatment with systemic aPD-L1 mAb. Combina￾tion treatment with both JAX fecal transfer and aPD-L1 mAb improved tumor control (Fig. 2D) and circulating tumor antigen–specific T cell responses (Fig. 2E), although there was little additive effect on accumulation of activated T cells within the tumor microenvironment (Fig. 2F). Consistent with these results, aPD-L1 therapy alone was signifi￾cantly more efficacious in JAX mice compared with TACmice (Fig. 2G), which paralleled improved antitumor T cell responses (fig. S1C). These data indicate that the commensal microbial compo￾sition can influence spontaneous antitumor im￾munity, as well as a response to immunotherapy with aPD-L1 mAb. To identify specific bacteria associated with im￾proved antitumor immune responses, we moni￾tored the fecal bacterial content over time of mice that were subjected to administration of fecal permutations, using the 16S ribosomal RNA (rRNA) miSeq Illumina platform. Principal coordinate analysis revealed that fecal samples analyzed from TAC mice that received JAX fecal material grad￾ually separated from samples obtained from sham￾and TAC feces–inoculated TAC mice over time (P = 0.001 and P = 0.003, respectively, ANOSIM multivariate data analysis) and became similar 1084 27 NOVEMBER 2015 • VOL 350 ISSUE 6264 sciencemag.org SCIENCE 1 Department of Pathology, University of Chicago, Chicago, IL 60637, USA. 2 Department of Medicine, University of Chicago, Chicago, IL 60637, USA. 3 Section of Genetic Medicine, University of Chicago, Chicago, IL 60637, USA. *These authors contributed equally to this work. †Corresponding author. E-mail: tgajewsk@medicine.bsd.uchicago.edu RESEARCH | REPORTS on June 24, 2016 http://science.sciencemag.org/ Downloaded from