preexisting commensal dysbiosis is a host- intrinsic ... · tumor biology and immunology...

Tumor Biology and Immunology

Preexisting Commensal Dysbiosis Is a Host-Intrinsic Regulator of Tissue Inflammation andTumor Cell Dissemination in Hormone Receptor–Positive Breast CancerClaire Buchta Rosean1, Raegan R. Bostic2, Joshua C.M. Ferey3, Tzu-Yu Feng1,Francesca N. Azar1, Kenneth S. Tung4, Mikhail G. Dozmorov5, Ekaterina Smirnova5,Paula D. Bos6, and Melanie R. Rutkowski1

Abstract

It is unknown why some patients with hormone receptor–positive (HRþ) breast cancer present with more aggressive andinvasive disease.Metastatic dissemination occurs early in diseaseand is facilitated by cross-talk between the tumor and tissueenvironment, suggesting that undefined host-intrinsic factorsenhance early dissemination and the probability of developingmetastatic disease. Here, we have identified commensal dysbio-sis as a host-intrinsic factor associated with metastatic dissem-ination. Using a mouse model of HRþ mammary cancer, wedemonstrate that a preestablished disruption of commensalhomeostasis results in enhanced circulating tumor cells andsubsequent dissemination to the tumor-draining lymph nodesand lungs. Commensal dysbiosis promoted early inflammationwithin the mammary gland that was sustained during HRþ

mammary tumor progression. Furthermore, dysbiosis enhancedfibrosis and collagen deposition both systemically and locally

within the tumor microenvironment and induced significantmyeloid infiltration into the mammary gland and breast tumor.These effects were recapitulated both by directly targeting gutmicrobes using nonabsorbable antibiotics and by fecal micro-biota transplantation of dysbiotic cecal contents, demonstratingthe direct impact of gut dysbiosis on mammary tumor dissem-ination. This study identifies dysbiosis as a preexisting, host-intrinsic regulator of tissue inflammation, myeloid recruitment,fibrosis, and dissemination of tumor cells in HRþ breast cancer.

Significance: Identification of commensal dysbiosis as ahost-intrinsic factor mediating evolution of metastatic breastcancer allows for development of interventions or diagnostictools for patients at highest risk for developing metastaticdisease.

See related commentary by Ingman, p. 3539

IntroductionBreast cancer is a major global health issue, with 1 in 8 women

expected to be diagnosed with breast cancer in their lifetime. Themajority of breast cancers (�65%) are diagnosed as hormonereceptor–positive (HRþ; estrogen and progesterone receptorþ,HER2neg/low; ref. 1). Because of the effectiveness of adjuvant hor-monal therapies,HRþbreast cancer has amore favorable short-termprognosis compared with more aggressive subtypes such as triple-negative breast cancer (TNBC; ref. 2). However, patients diagnosed

with HRþ breast cancer have heterogeneity in tumor aggressivenessand overall survival, with a subset of patients remaining at signif-icant risk for developing recurrent disease after 5 years of standardendocrine therapy (3). This heterogeneity in outcome results in asmall but sustained risk for developing recurrent metastatic disease5 years or more after initial remission (4).

Metastatic dissemination to distal lymph nodes and the lungsoccurs early in HRþ breast cancer progression (5). Although itremains unknown whether host-intrinsic differences in immunefunction contribute to increased early dissemination of tumorcells, lymph node involvement is one of the most reliable pre-dictors of recurrent disease (6, 7) and identifies patients thatwouldbenefit from extended adjuvant hormonal therapy. This therapyreduces late recurrence for some high-risk patients but also com-monlyresults inundesirableandpotentially serious sideeffects (8).Identification of early or preexisting host-associated factors thatinfluence the evolution of disseminated breast cancer could there-fore aid in thedevelopmentof alternative interventions forpatientsthat are at highest risk for recurrent metastatic disease.

Factors such as obesity, genetic polymorphisms, race, diet, andchronic use of antibiotics have been associated with adverseoutcomes in breast cancer (9–12). All of these factors are alsolinked to a disruption in the homeostasis of the commensalmicrobiome. The commensal microbiota is a host-intrinsic reg-ulator of systemic innate and adaptive immune function (13–15).Altered diversity in the microbiota, known as dysbiosis, can

1Department of Microbiology, Immunology, and Cancer Biology, University ofVirginia, Charlottesville, Virginia. 2Department of Cell Biology, University ofVirginia, Charlottesville, Virginia. 3University of Virginia School of Medicine,Charlottesville, Virginia. 4Department of Pathology, University of Virginia,Charlottesville, Virginia. 5Department of Biostatistics, Virginia CommonwealthUniversity, Richmond, Virginia. 6Department of Pathology, Massey CancerCenter, Virginia Commonwealth University, Richmond, Virginia.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Melanie R. Rutkowski, University of Virginia MedicalCenter, MR6 G526, 345 Crispell Drive, Charlottesville, VA 22908. Phone: 434-297-8103; Fax: 434-982-1071; E-mail: [email protected]

Cancer Res 2019;79:3662–75

doi: 10.1158/0008-5472.CAN-18-3464

�2019 American Association for Cancer Research.

CancerResearch

Cancer Res; 79(14) July 15, 20193662

on January 25, 2021. © 2019 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 7, 2019; DOI: 10.1158/0008-5472.CAN-18-3464

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-18-3464&domain=pdf&date_stamp=2019-6-29

http://cancerres.aacrjournals.org/

influence the systemic immune environment (16) and lead toincreased production of inflammatorymediators that associatewitha poor outcome for multiple diseases, including cancer (17–19).Indeed, infectionofAPCMin/þmicewithH. hepaticus led to increasedinflammation and significant incidence of breast cancer in femalemice (20), supporting the idea that perturbation of commensalorganisms can influence breast cancer risk (21). However, it iscurrently unknown whether preexisting commensal dysbiosis leadsto changes within the tissue and/or tumor microenvironment thatpromote tumor aggressiveness and metastatic dissemination.

We set out to determine whether commensal microbes influ-ence breast cancer progression: specifically, if preexisting dysbio-sis in the commensal microbiome associates with poor outcomesin HRþ breast cancer. Here, we demonstrate that a disruption incommensal homeostasis promotes tumor cell dissemination todistal sites, enhances fibrosis, and collagen deposition bothsystemically and locally within the tissue and tumor microenvi-ronment, and results in significant early inflammation and mye-loid infiltration into themammary tissue and tumor. These effectsare initiated by signals from the gut, as commensal dysbiosisinducedbothby treatmentwithnonabsorbable antibiotics andbyfecal microbiota transplantation (FMT) of dysbiotic cecal con-tents recapitulates these outcomes. This study identifies commen-sal dysbiosis as a host-intrinsic regulator of tissue inflammation,myeloid recruitment, fibrosis, and dissemination of tumor cells–all of which contribute to reduced survival in HRþ breast cancer.

Materials and MethodsMice

Five- to 8-week-old female C57BL/6mice were purchased fromCharles River Laboratories. L-Stop-L-KRasG12Dp53flx/flxL-Stop-L-Myristoylated p110a�GFPþ mice on a C57BL/6 background (22)were bred andmaintained in house. All animals weremaintainedin pathogen-free barrier facilities at the University of Virginia(Charlottesville, VA). All experiments in this study were approvedby the University of Virginia Institutional Animal Care and UseCommittee.

Genetic tumor model and cell linesThe poorly metastatic HRþ mouse mammary cancer cell line

BRPKp110 has been described previously (23). 5E5 BRPKp110cells were injected orthotopically into the abdominal mammaryfat pad. The highly metastatic mouse mammary cancer cell linePyMT, which expresses luciferase, was cloned from a metastatictumor derived from the HRþ MMTV-PyMT model as describedpreviously (24). In experiments using this cell line, 1E5 PyMT cellswere injected orthotopically into the abdominal mammary fatpad. Cell lines were authenticated bymaintaining at less than fourpassages, monitoring of morphology, and testing forMycoplasma.Autochthonous mammary tumors were induced in L-Stop-L-KRasG12Dp53flx/flxL-Stop-L-Myristoylated p110a�GFPþ mice on aC57BL/6backgroundas describedpreviously (22). Briefly, tumorswere initiated via intraductal injection of 2.5E7 PFU of adenovi-rus-Cre through the nipple in the fourth mammary fat pad.

Antibiotic treatmentFor most experiments, mice were orally gavaged for 14 days

with antibiotic cocktail or water as a vehicle control as describedpreviously (25). After 14 days of gavage, mice were left untreatedfor four days before tumor initiation to allow for the establish-

ment of commensal dysbiosis. Mice treated with an absorbable(systemic) antibiotic cocktail received 100 mL of a cocktail con-taining vancomycin (0.5 mg/mL), ampicillin (1 mg/mL), metro-nidazole (1 mg/mL), neomycin (1 mg/mL), and gentamicin(1 mg/mL). Vehicle-treated mice received 100 mL of water as acontrol. Mice treated with nonabsorbable (nonsystemic) antibio-tics received 200 mL of a cocktail containing neomycin (2mg/mL)andbacitracin (2mg/mL). Vehicle-treatedmice received 200mL ofwater as a control. All antibiotics were purchased from Sigma-Aldrich with the exception of vancomycin (Gold Biotechnology).For FMT experiments, dysbiosis was induced in donor mice asdescribed above. Cecal contents from dysbiotic and nondysbioticanimals were collected, homogenized, and frozen at �80�C insterile 1:1 glycerol/PBS. Recipient mice were orally gavaged withabsorbable antibiotics for 7 days, followed by 3 consecutive daysof oral gavage with either dysbiotic or nondysbiotic cecal con-tents. After the final day of gavage, recipientmice were rested for 7days before tumor initiation to allow for microbial engraftment.

Flow cytometryTumors, mammary glands, and lungs were removed, weighed,

homogenized (gentleMACS Dissociator, Miltenyi Biotec), anddigested with collagenase D (Sigma-Aldrich) for 30 minutes at37�Cprior to 70-mmfiltration. Tumor-draining lymphnodeswerehomogenized using glass slides and processed as the other tissues.Flow cytometrywas performed by staining single-cell suspensionswith Zombie Aqua viability dye (BioLegend), blocking Fc recep-tors with anti-CD16/32 (93, purified), and staining with anti-mouse antibodies at the manufacturer's recommended dilution.Myeloid cell infiltration was quantitated in tumors and normal-adjacent mammary glands by surface staining with anti-mouseCD45 (30-F11, Pacific Blue), CD11b (M1/70, PE/Cy7), F4/80(BM8,PerCP/Cy5.5), CD86(GL-1, BV650),CD206 (C068C2,PE/Dazzle-594), Ly6C (HK1.4, APC/Cy7), and Ly6G (1A8, FITC) andintracellular staining using anti-mouse arginase-1 (IC5868P, PE)and IL6 (MP5-20F3, APC). Tumor cell dissemination was quan-titated using single-cell suspensions from lungs and tumor-drain-ing lymph nodes through surface staining with anti-mouse CD45(30-F11, PE) and intracellular staining with anti-GFP (FM264G,APC). Circulating tumor cells from the blood were quantitated asdescribed previously (26). Briefly, equal volumes of EDTA-treatedbloodunderwent redblood cell lysis,wereplated into6-well culturedishes, and incubated for 7 days. Cells were detached and stained asdescribed for quantitation of disseminated tumor cells. All anti-bodies were purchased from BioLegend with the exception ofanti-arginase-1 (R&D Systems). Counting beads (AccuCount,Spherotech) were added to samples at the manufacturer's recom-mended concentration, and the sampleswere subsequently runonaBeckman Coulter CytoFLEX cytometer and analyzed using FlowJo.

Cytokine analysisCytokines and chemokines were analyzed in serum and mam-

mary glands using a custom multiplex U-PLEX assay from MesoScale Diagnostics. Blood was collected by cardiac puncture andallowed to clot at room temperature for 30 minutes beforecentrifugation and collection of serum. Mammary glands wereflash frozen in liquid nitrogen, ground in amortar and pestle withlysis buffer (150mmol/L NaCl, 20mmol/L Tris, 1mmol/L EDTA,1 mmol/L EGTA, 1% Triton X-100, pH 7.5), sonicated, andcentrifuged. The supernatants were then collected, protein con-centrations were quantified by BCA assay (Pierce, Thermo Fisher

Dysbiosis Enhances Dissemination of HRþ Breast Cancer

www.aacrjournals.org Cancer Res; 79(14) July 15, 2019 3663




Scientific), and supernatants were used for U-PLEX analysis.Mammary gland readouts were normalized to total protein input.

Histology and microscopyTo quantitate fibrosis, tissues were fixed in neutral-buffered

formalin, paraffin-embedded, and cut in 5-mm sections (ResearchHistology Core, University of Virginia, Charlottesville, VA). Slideswere stained using PicroSirius Red (0.1% Direct Red 80 in satu-rated aqueous picric acid, Sigma-Aldrich). Fibrosis was quantifiedusing ImageJ software by calculating the area of tissue and theintensity of red staining. To examine mammary gland pathology,hematoxylin and eosin–stained slides were scored by a patholo-gist in a blinded fashion. To determine estrus cycling, vaginallavages were collected daily after the 14-day antibiotic gavageperiod and examined for vaginal cytology. Mice were consideredto be cycling when they progressed from proestrus to estrus.

Bioluminescent imagingMice were injected intravenously with 3 mg of D-luciferin

potassium salt (Gold Biotechnology) and allowed to saturatesystemically for 2 minutes. Mice were then euthanized, and lungswere removed for ex vivo imaging to detect bioluminescent lightemission using a Caliper IVIS Spectrum bioluminescence instru-ment (PerkinElmer; Molecular Imaging Core, University of Virgi-nia, Charlottesville, VA). Bioluminescence from ex vivo lungs wascalculated as photons per second within a standardized region ofinterest and normalized to final tumor burden.

16S microbiome sequencingSequencing was performed by the University of Maryland

Institute for Genome Science. Sequences were demultiplexedusing the mapping file split_libraries_fastq.py, a QIIME-depen-dent script. Fastq files were split by using seqtk (https://github.com/lh3/seqtk), primer sequences removed using TagCleaner(0.16), followed by downstream processing using DADA2Work-flow for Big Data and dada2 (v. 1.5.2; https://benjjneb.github.io/dada2/bigdata.html). Forward and reverse reads were trimmedusing lengths of 255 and 225 bp, respectively, to contain noambiguous bases, have aminimumquality score of 2, and containless than two expected errors based on quality score. Reads wereassembled and chimeras removed per dada2 protocol.

Taxonomic assignments. Taxonomywas assigned to each ampliconsequence variant (ASV) generated by dada2 using a combinationof the SILVA v128 database and the RDP na€�ve Bayesian classifieras implemented in the dada2Rpackage species-level assignments.Read counts for ASVs assigned to the same taxonomy weresummed for each sample.

Statistical analysis. Alpha diversity of microbiome samples wasmeasured using Shannon alpha diversity measure and a Kruskal–Wallis test to measure significance. Beta diversity of microbiomesequences was assessed using Bray–Curtis dissimilarity measuresbased upon relative abundance data and a Permanova test wasperformed to measure statistical significance.

Metabolomic analysisMetabolomic analysis was performed by Metabolon, Inc. Prin-

cipal component analysis was performed using the "prcomp()" Rfunction (https://www.R-project.org/). Samples were coloredaccording to tumor status, dysbiosis, or both (sample annota-

tions). The association between the sample annotations and thefirst two principal components was tested by one-way ANOVA. Asignificant association suggests that the variability explained by aprincipal component is driven by a corresponding sample anno-tation. Calculations and visualizations were performed in the R/Bioconductor environment v.3.5.1.

Statistical analysisP values were generated using Student t tests (unpaired, two-

tailed, at 95% confidence interval) or ANOVA when appropriate.Statistical significance is designated when P < 0.05.

ResultsPreestablished commensal dysbiosis results in enhanced tumorcell dissemination

To evaluate the impact of commensal dysbiosis on metastaticHRþ breast cancer, we used the poorly metastatic GFPþ

BRPKp110 syngeneic model (23) of HRþ mammary cancer. Ourgoal was to determine whether preexisting commensal dysbiosisaffects mammary tumor metastasis and progression. To that end,mice were orally gavaged with a broad-spectrum cocktail ofantibiotics (25) for 14 days, a sufficient duration to enableoutgrowth of antibiotic-resistant commensal species (27). Micewere then rested for four days following cessation of antibiotics toallow for dysbiotic reacquisition of commensal species (28), andmammary tumors were initiated by injecting tumor cells into thefourth abdominal mammary fat pad (Fig. 1A).

Bacterial composition was evaluated in the feces of mice on theday of tumor initiation using 16S rDNA sequencing. Comparedwith vehicle-treated mice, those receiving antibiotics showedshifts in bacterial communities at both the phylum and genuslevels (Supplementary Fig. S1A).Using aBray–Curtis dissimilaritymeasure, beta diversity between the two treatment groups differedsignificantly (Supplementary Fig. S1B). Antibiotic-treated micehad a significant reduction in community richness within the gut,as indicated by calculation of Shannon alpha diversity index(Supplementary Fig. S1C), as well as enlarged ceca (Supplemen-tary Fig. S1D), demonstrating that antibiotic-treated mice hadcommensal dysbiosis prior to tumor initiation.

To evaluate the impact of antibiotic-driven gut dysbiosis in thebreast tissue, we first performed histologic examination of themammary gland at the time of tumor implantation. Blindedhistopathologic scoring of the mammary tissues indicated thatthe antibiotic treatment resulted in minimal changes to tissuemorphology (Supplementary Fig. S2A and S2B).

Metabolic output often reflects the status of the commensalecosystem. To indirectly evaluate whether functional changes tothe commensal ecosystem within the mammary tissue microen-vironment occur in response to antibiotic treatment or tumorstatus, we compared metabolites in normal-adjacent mammarytissue from antibiotic- or nonantibiotic-treated mice, with orwithout a tumor 6 days after tumor cell inoculation. As a positivecontrol, we collected feces from treated and nontreated mice.Similar to changes observed with 16S rDNA sequencing at day 0,global fecal metabolites significantly differed when compared bydysbiotic status, but were independent of tumor status (Supple-mentary Fig. S2C). On the contrary, metabolites within themammary tissue remained unchanged despite dysbiosis or tumorstatus (Supplementary Fig. S2D). Thus, at the time of tumorimplantation, histologic and metabolic changes within the

Buchta Rosean et al.

Cancer Res; 79(14) July 15, 2019 Cancer Research3664



https://github.com/lh3/seqtk

https://github.com/lh3/seqtk

https://benjjneb.github.io/dada2/bigdata.html

https://benjjneb.github.io/dada2/bigdata.html

https://www.R-project.org/


mammary tissue were similar between treatment groups. Togeth-er, these data suggest that the mammary tissue is minimallyaffected by antibiotic treatment.

Using the poorly metastatic syngeneic BRPKp110 model, weevaluated tumor cell dissemination to lungs and axillary lymphnodes by flow cytometric quantitation of GFPþCD45� mam-mary tumor cells (Supplementary Fig. S3A). Significantly moredisseminated tumor cells were detected within the lungs(Fig. 1B and C), peripheral blood (Fig. 1D; SupplementaryFig. S3B), and tumor-draining axillary lymph nodes (Fig. 1E;Supplementary Fig. S3C) of dysbiotic mice. Importantly,

although commensal bacteria have been shown to modulateestrogen levels, which could impact the estrus cycle and tumorgrowth kinetics of HRþ mammary tumors, we did not observedifferences in primary tumor growth kinetics in mice with orwithout commensal dysbiosis (Fig. 1F). Estrus cycle monitoringbetween the last day of antibiotic treatment until the day oftumor implantation using vaginal lavage confirmed no sub-stantial differences in cycle patterns (Supplementary Fig. S4A)or in the proportions of animals that were cycling, as defined byprogressing from proestrus to estrus (Supplementary Fig. S4B).These data suggest that commensal dysbiosis influences tumor

SSC

Anti-GFP

1.59

A

Tumorinitiation

-17 -4 0Days tumor End

Gavageantibiotics or

vehicle

Elimination of floraoutgrowth antibioticresistant strains

Dysbioticcommensalrepopulation

NO ABX

Begin

Tumor growth

F

Non-dysbiosis

Dysbiosis

0.71

8.68 3.88

C

Non-dysbiosis

Dysbiosis

G H

Non-dysbiosis Dysbiosis

Pho

tons

/sec

ond

per

mg

of

PyM

T t

um

or

I

Days post PyMT

DysbiosisNon-dysbiosis

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

10 15 20 250

500

1,000

1,500

2,000

CTC

s/m

L bl

ood

***

0

2,000

4,000

6,000 **E

Non-dysbiosis Dysbiosis Non-dysbiosis Dysbiosis 0 10 20 300

500

1,000 Non-dysbiosisDysbiosis

Days post BRPKp110

Tota

l CD

45-

lung

GFP

+ tu

mor

cel

ls

Tumorspiked

+ control

0

20

40

60

80

100 **

Non-dysbiosis DysbiosisTo

tal C

D45

-ly

mph

nod

e G

FP+ tu

mor

cel

ls0

2,000

4,000

6,000

8,000

10,000

1.62

3.17

TumorspikedFMO

SSC

Anti-GFP

37 0.07

B

D

0

1,000

2,000

3,000

4,000 *

Figure 1.

Commensal dysbiosis enhances mammary tumor cell dissemination.A, Experimental design for antibiotic-induced dysbiosis and tumor initiation. C57BL/6 micewere orally gavaged for 14 dayswith a broad-spectrum cocktail of antibiotics or an equal volume of water as a vehicle control. Gavage was ceased 4 days prior totumor initiation in the fourth mammary fat pad. Tumor size was measured by calipers every 2–3 days after reaching a palpable size. B, Representative plotsdemonstrating GFPþ tumor cell quantitation in the lungs. The anti-GFP gate was chosen based upon fluorescence minus one (FMO) controls and a stained lungsample spiked with BRPKp110 tumor cells. Numbers represent percent cells within the anti-GFP gate of total live cells. C–F, C57BL/6 mice were treated asdescribed inA. Twenty-seven days after BRPKp110 tumor initiation, GFPþ tumor cell dissemination was quantified in lung tissue (C), peripheral blood (D), andtumor-draining axillary lymph nodes (E) by flow cytometry. Data are represented as absolute number of GFPþCD45� cells of live, singlet cells. F,Growth kineticsof BRPKp110mammary tumors. Representative of at least three independent experiments with 5mice/group. G–I, C57BL/6 mice were treated as described in A.Twenty-five days after PyMT-luciferase tumor initiation, tumor cell disseminationwas quantified in lung tissue by bioluminescence. G, Representative images ofbioluminescence in lungs from advanced tumor-bearing mice bearing PyMT-luciferase tumors. H,Quantitation of luminescence represented as photons/secondnormalized to final tumor burden for mice with PyMT tumors. I, Growth kinetics of PyMTmammary tumors. Representative of two independent experiments with5 mice/group. �, P¼ 0.01–0.05; �� , P¼ 0.001–0.01; �� , P < 0.001.






cell seeding to distal sites, independent of changes to the tissuemicroenvironment or estrus cycle.

We next evaluated whether preestablished commensal dys-biosis would impact tumor growth or metastatic disseminationin a more aggressive and metastatic tumor model. For theseexperiments, we used a tumor cell line that was derived from anadvanced primary mammary tumor arising in the autochtho-nous MMTV-PyMT mouse model (24). Similar to what weobserved using the BRPKp110 model, we found significantlygreater dissemination of tumors into the lungs, independent oftumor volume (Fig. 1G and H), and tumors progressed withequal kinetics regardless of the dysbiotic status of the animals(Fig. 1I). Together, these results suggest that commensal dys-biosis has a sustained impact on the dissemination of HRþ

mammary cancer, and that increased dissemination in micewith commensal dysbiosis is occurring independently of tumorgrowth kinetics.

We also tested the effects of commensal dysbiosis using aGFPþ-inducible autochthonous model (22) that is established by tar-geting latent mutations with intraductal delivery of adenovirus-Cre. Using a similar experimental approach as depicted in Fig. 1A,despite variable tumor kinetics, no significant differences inprimary tumor growth between dysbiotic and nondysbioticgroups were observed (Supplementary Fig. S5A). However, thisvariability makes it challenging to evaluate effects on tumorcell dissemination. To confirm whether dysbiosis-driven, host-intrinsic differences induce dissemination in the autochthonousmodel, we performed paired tumor transfers into dysbiotic ornondysbiotic recipient mice. Autochthonous mammary tumorswere induced in wild-type mice without commensal dysbiosis,sterilely excised when they reached approximately 1 cm in diam-eter, and equal-weight fragments from the same tumor sectionwere surgically implanted in the mammary tissue of recipientanimals pretreated with antibiotics or vehicle gavage as describedin Fig. 1A. After the implanted tumors grew to an advanced size,lungs and axillary lymph nodes from the recipient mice wereexamined for the presence of disseminated tumor cells. Similar towhat we observed using the syngeneic models, lungs from dys-biotic mice showed significantly greater frequencies of dissemi-nated tumor cells (Supplementary Fig. S5B and S5C), whereasthere was a nonsignificant increase in GFPþ tumor cells in distallymph nodes (Supplementary Fig. S5D). Together, these resultsconfirm that commensal dysbiosis enhances tumor cell dissem-ination independently of primary tumor growth, and suggest thathost-associated changes in response to commensal dysbiosis, nottumor-intrinsic differences in invasive capacity, enhance tumordissemination.

Commensal dysbiosis leads to enhanced inflammation andmyeloid cell infiltration within the mammary gland

To identify possible microenvironmental factors associatedwith increased dissemination in tumor-bearing mice with com-mensal dysbiosis, we first defined the immunologic changesoccurring within the primary tumor microenvironment, specifi-cally within the mammary tissue. Previous studies have demon-strated that inflammation within the mammary tissue associateswith an increased risk for breast cancer (29). However, it is notwell understood how this inflammation arises. The commensalmicrobiome is important for the maintenance of systemic andlocal immune homeostasis, giving rise to the possibility thatcommensal dysbiosis may result in a disruption of mammary

tissue homeostasis and increased inflammation within the tissue.To examine the effects of commensal dysbiosis on the mammarygland during tumor progression, mammary tissues were har-vested during early (day 12) and advanced (day 27) stages oftumor progression (Fig. 2A). Mammary glands from nontumor-bearing mice with or without dysbiosis were also evaluated inparallel. To normalize the systemic effects of hormones on mam-mary gland homeostasis and immune function, estrus cycles weresynchronized in all animals prior to analysis using a modifiedWhitten effect (30).

Macrophages within the mammary gland associate with anincreased risk of developing metastatic breast cancer (31). Wetherefore determined whether commensal dysbiosis influencedthe frequencies and numbers of macrophages during early oradvanced stages of mammary tumor progression. During earlystages of tumor progression, the presence of a tumorwas sufficientto significantly enhance myeloid accumulation into the normal-adjacent mammary gland (Fig. 2B). This effect persisted intoadvanced stages of tumor progression, resulting in a substantialincrease inmyeloid frequencieswithin thenormal-adjacentmam-mary gland (Fig. 2C). However, commensal dysbiosis lead to adramatic and significant increase in total accumulated myeloidcells within the normal-adjacent mammary glands at both early(Fig. 2B) and advanced (Fig. 2C) stages of tumor progression.Because tumors progressed with similar kinetics, these resultssuggest that commensal dysbiosis synergizes with a developingtumor to enhance myeloid cell infiltration within mammarytissue. Importantly, these data indicate that the effects of pre-established commensal dysbiosis on myeloid-driven inflamma-tion within the mammary gland are sustained throughout tumorprogression. Similar to published studies in humans (31), mye-loid cells infiltrating into normal-adjacent mammary tissue indysbiotic mice expressed high levels of the inflammatory media-tors arginase-1 and IL6, specifically in the context of an earlydeveloping mammary tumor (Fig. 2D). However, duringadvanced stages of tumor progression, myeloid production ofarginase-1 and IL6 was unchanged (Fig. 2E). These results dem-onstrate that commensal dysbiosis promotes the early recruit-ment of inflammatory myeloid cells into the mammary tissuemicroenvironment.

We next wanted to define signals that arise in response tocommensal dysbiosis and could enhance the recruitment ofinflammatory myeloid cells into the breast tissue. Therefore,mammary tissues were harvested prior to tumor initiation (pre-tumor; day 0) and during early (early tumor; day 12) stages oftumor progression. Similar to the experiment depicted in Fig. 2A,estrus cycles were synchronized prior to each experimental end-point (Fig. 2F). We focused our analysis on three myeloid che-moattractants that also associate with adverse outcomes in breastcancer: CXCL10 (32, 33), CCL2 (34–36), and CXLC2 (37).

Compared with nondysbiotic mice, commensal dysbiosisresulted in a significant upregulation of both CXCL10 and CCL2within mammary tissue prior to tumor initiation (pretumormammary tissue; day 0), while CXCL2 was elevated but notsignificantly different (Fig. 2G). These data demonstrate thatthe systemic and/or local response to commensal dysbiosisculminates, at the very least, in low-level nonpathologicinflammation within the mammary gland. Although the mam-mary gland inflammation eventually subsided in nontumor-bearing dysbiotic mice, CXCL10, CCL2, and CXCL2 remainedsignificantly elevatedwithinnormal-adjacentmammary glands in






tumor-bearing mice with preestablished dysbiosis when com-pared with all other experimental groups (Fig. 2G). Together,these data demonstrate that preexisting commensal dysbiosisresults in inflammation within the mammary gland that persistsin the presence of a tumor during early stages of mammary tumorprogression. In addition, these data suggest that cross-talk occur-ring between an inflamed tissue and a developing tumor amplifiesthe accumulation of myeloid cells into the microenvironment,and that the effects of this cross-talk are likely sustained through-out early and advanced stages of tumor progression.

Dysbiosis induces early and sustained systemic inflammationand enhances myeloid infiltration into tumors

In addition to promoting local inflammation within breasttissue, commensal dysbiosis also enhanced systemic inflam-

mation. To examine the systemic impact of commensal dys-biosis, serum was collected prior to tumor initiation (pretumorserum; day 0) and at early (early tumor serum; day 12) andadvanced (advanced tumor serum; day 27) stages of tumorprogression. Serum was analyzed for the presence of GM-CSF,CCL2, and CXLC2, chemokines involved in myeloid cellgeneration and migration; and IL22 and IL23, cytokinesthat mediate inflammatory responses and that are associatedwith gut barrier function. We focused upon these twocytokines as expression of both IL22 and IL23 are enhancedin the serum of patients with breast cancer and have beenassociated with tumor progression and decreased overallsurvival (38, 39).

Interestingly, in response to commensal dysbiosis, all cytokinesand chemokines tested were significantly increased in the serum

Arginase 1IL6

Non-dysbiosis

Dysbiosis

Non-dysbiosis +Tumor

Dysbiosis +Tumor

B

Non-dysbiosis

Dysbiosis

Non-dysbiosis + Tumor

Dysbiosis + Tumor

Non-dysbiosis

Dysbiosis

Non-dysbiosis + Tumor

Dysbiosis + Tumor

pmn-MDSC

mono-MDSCM2

M1

M0

*

C

Non-dysbiosis

Dysbiosis

Non-dysbiosis +Tumor

Dysbiosis +Tumor

Arginase 1IL6

A

D

E

*

**

** *

Early mammary tissue

Advanced mammary tissue

+/-

Tum

or

-17 -4 0Days tumor

+27

ezylanA

End


vehicle

+23

SynchronizeEstrus

Begin+8 +12

ezylanA

SynchronizeEstrus

Early tumor Advanced tumor

pg C

CL2

/mg

mam

mar

y tis

sue

pg C

XC

L2/m

g m

amm

ary

tissu

e

Non-dy

sbios

is

Dysbio

sis

Non-dy

sbios

is + T

umor

Dysbio

sis +

Tumor

pg C

XC

L10/

mg

mam

mar

y tis

sue

Non-dy

sbios

is

Dysbio

sis

Pretumormammary tissue

Early tumor mammary tissue

*

*

*

*

*0

20

40

60

80

0

10

20

30

40

0

5

10

15

20

0

2

4

6

8

100

20

40

60

80

100

0

5

10

15

20

25

+/-

Tum

or



vehicle

Begin+8 +12ezylanA

SynchronizeEstrus

Early tumorSynchronize

Estrus

Pretumor

ezylanA

F

GEarly mammary tissue


Tota

l CD

11b+

/mg

mam

mar

y tis

sue

Tota

l CD

11b+ /

mg

mam

mar

y tis

sue

Tota

l CD1

1b+ /m

g m

amm

ary

tissu

e To

tal C

D11

b+ /m

g m

amm

ary

tissu

e

0

20

40

60

80

050

100150200250300350400450500

0

20

40

60

80

0

20

40

60

pmn-MDSC

mono-MDSCM2

M1

M0

Figure 2.

Dysbiosis enhances mammary gland inflammation andmyeloid cell infiltration.A, Experimental design for B–E. C57BL/6 mice were orally gavaged for 14 dayswith a broad-spectrum cocktail of antibiotics or an equal volume of water as a vehicle control. Gavage was ceased 4 days prior to BRPKp110 tumor initiation intothe abdominal mammary fat pad. Mice were euthanized at early (day 12) and advanced (day 27) tumor timepoints. Four days prior to each timepoint, a modifiedWhitten effect was used to synchronize estrus in these animals. Normal tumor-adjacent mammary glands were harvested, and infiltrating myeloid cellpopulations were quantitated by flow cytometry at early (B) and advanced (C) timepoints after tumor initiation. All populations were gated on live, singlet,CD45þCD11bþ cells. Numbers represent absolute numbers of cells quantitated using counting beads. M0macrophages, F4/80þCD86�CD206�; M1 macrophages,F4/80þCD86þCD206�; M2 macrophages, F4/80þCD86�CD206þ; monocytic MDSC, Ly6ChiLy6G�; polymorphonuclear MDSC, Ly6CmidLy6Gþ. Arginase-1 andIL6 expression was quantitated by intracellular staining in bulk CD45þCD11bþ cells frommammary glands at early (D) and advanced (E) timepoints after tumorinitiation. Numbers represent absolute numbers of cells quantitated using counting beads. F, Experimental design forG. Similar to A, C57BL/6 mice were orallygavaged for 14 dayswith a broad-spectrum cocktail of antibiotics or an equal volume of water as a vehicle control. Gavage was ceased 4 days prior to BRPKp110tumor initiation into the abdominal mammary fat pad. Mice were euthanized prior to tumor initiation (pretumor; day 0) and at an early point (early tumor; day 12)during tumor progression. Four days prior to each timepoint, a modifiedWhitten effect was used to synchronize estrus in these animals. Normal tumor-adjacentmammary glands were harvested, and protein levels of CXCL10, CCL2, and CXCL2 were quantitated bymultiplex cytokine analysis (G). � , P¼ 0.01–0.05;�� , P¼ 0.001–0.01.






of mice with established commensal dysbiosis prior to tumorinitiation.While no significant differenceswere observed in serumfrom dysbiotic or nondysbiotic mice with or without early-stagemammary tumors, dysbiotic mice bearing advanced mammarytumors showed significantly increased serumCCL2 and a low butsignificant increase of IL23, while levels of GM-CSF, CXCL2, andIL22were elevated but not significantly different (Fig. 3A). Impor-tantly, because tumors progress at similar rates regardless of

dysbiotic status, enhanced serum cytokines were not due todifferences in primary tumor burden.

Macrophages are one of themost abundant cell typeswithin thebreast tumor microenvironment (40) and are a significant prog-nostic indicator of reduced survival for patients diagnosed withHRþ breast cancer (41). Confirming this, the majority of myeloidinfiltrates within themammary tumormicroenvironment at earlyand advanced stages of tumor progression were M2-like

pg/m

L S

erum

CX

CL2

Non-dy

sbios

is

Dysbio

sis

Non-dy

sbios

is + T

umor

Dysbio

sis +

Tumor

Non-dy

sbios

is

Dysbio

sis

Pretumor serum Early tumor serum

pg/m

L S

erum

CC

L2pg

/mL

Ser

um

GM

-CS

Fpg

/mL

Ser

umIL

23pg

/mL

Ser

umIL

22

Non-dy

sbios

is + T

umor

Dysbio

sis +

Tumor

Advanced tumor serumA

**

Non-dysbiosis Dysbiosis Non-dysbiosis Dysbiosis

**

Tota

l int

ratu

mor

al

CD

11b+

cel

ls

BEarly tumor Advanced tumor

NS

0

20

40

60

80

100

*

0.0

0.5

1.0

1.5

2.0

NS

0

10

20

30

***

0

200

400

600

800

NS

0.0

0.5

1.0

1.5

0.0

0.2

0.4

0.6

0.8

0

20

40

60

80

0

10

20

30

40

0

2

4

6

8

0

10

20

30

40

50

*

0.0

0.2

0.4

0.6

0.8

*

0

10

20

30

40

50

*

0

20

40

60

*

0.0

0.5

1.0

1.5

2.0

2.5

*

0

1

2

3

4

5

Tota

l int

ratu

mor

al

CD

11b+

cel

ls

0

500,000

1,000,000

1,500,000

pmn-MDSC

mono-MDSCM2

M1

M0

0

20,000

40,000

60,000

Figure 3.

Dysbiosis induces early andsustained systemic inflammationand enhances myeloid infiltrationinto mammary tumors. C57BL/6mice were treated as describedin Fig. 2A.A, Serumwas collectedfrom BRPKp110 tumor- andnontumor-bearing mice, with orwithout established dysbiosis, atvarious timepoints: prior to tumorinitiation (pretumor serum; day 0),and at early (early tumor serum; day12) and advanced (advanced tumorserum; day 27) timepoints aftertumor initiation. Serum GM-CSF,CCL2, CXCL2, IL23, and IL22 werequantitated by multiplex cytokineanalysis. B, BRPKp110mammarytumors were harvested, andinfiltrating myeloid cell populationswere quantitated by flow cytometryat early and advanced timepointsafter tumor initiation. Allpopulations were gated on live,singlet, CD45þCD11bþ cells.Numbers represent absolutenumbers of cells quantitated usingcounting beads. M0macrophages,F4/80þCD86�CD206�;M1 macrophages, F4/80þCD86þCD206�;M2macrophages,F4/80þCD86�CD206þ; monocyticMDSC, Ly6ChiLy6G�;polymorphonuclear MDSC,Ly6CmidLy6Gþ. � , P¼ 0.01–0.05;�� , P¼ 0.001–0.01; NS, notsignificant.






macrophages, based upon CD206 expression (Fig. 3B). Similar towhat we observed in normal-adjacent mammary glands, prees-tablished commensal dysbiosis resulted in enhanced total mye-loid cell accumulation within both early and advancedmammarytumors. Importantly, advanced tumors from dysbiotic mice hadsignificantly increased numbers of infiltrating tumor-promotingM2-like macrophages as compared with nondysbiotic controlswith equal tumor burden. Together, these data demonstrate thatthe systemic expression of inflammatorymediators is increased indysbiotic tumor–bearing mice, enhancing myeloid recruitmentinto mammary tumors.

Dysbiosis enhances both local and systemic fibrosis inmammary tumor-bearing animals

Enhanced stromal density, or dense breasts, is a well-estab-lished risk factor for developing metastatic breast cancer (42) andtumor-promoting inflammation in the breast (43). To determinewhether commensal dysbiosis enhanced fibrosis within the tissueor tumor microenvironment or at systemic sites of metastaticdissemination, we stained tissues from dysbiotic and nondysbio-tic mice with PicroSirius Red to visualize collagen deposition. Wefound that preestablished commensal dysbiosis resulted in sig-nificant enhancement of collagen deposition within the normal-


Early mammary tissue Advanced mammary tissue

Non-dysbiosis

Dysbiosis

Non-dysbiosis+ Tumor

Dysbiosis+ Tumor

Non-dysbiosis

Dysbiosis

Non-dysbiosis

+ Tumor

Dysbiosis

+ Tumor

0

20

40

60

80

% A

rea

stai

ned

Non-dysbiosis

Dysbiosis

Non-dysbiosis

+ Tumor

Dysbiosis

+ Tumor

0

20

40

60

80

% A

rea

stai

ned

A

B C

Non-dysbiosis Dysbiosis0

5

10

15

20

% A

rea

stai

ned *


*

Advanced tumors Advanced lungs

Figure 4.

Dysbiosis enhances both local andsystemic fibrosis within normaltumor-adjacent mammary tissue inadvanced tumor-bearing animals.C57BL/6 mice were treated asdescribed in Fig. 2A. A, Normal-adjacent mammary glands wereharvested from BRPKp110 tumor-and nontumor-bearing mice, with orwithout established dysbiosis, atearly (early mammary tissue; day12) and advanced (advancedmammary tissue; day 27)timepoints after tumor initiation.BRPKp110 mammary tumors (B)and lungs (C) were harvested fromadvanced tumor-bearing animalswith or without establisheddysbiosis. All tissues were formalin-fixed and paraffin-embedded, andsections were stained withPicroSirius Red. Quantification ofstaining intensity was calculatedusing ImageJ software.� , P¼ 0.01–0.05.






adjacent mammary gland (Fig. 4A) and tumors (Fig. 4B) ofadvanced tumor-bearing mice. We also observed a slight, butsignificant, increase in collagen accumulation within the lungs ofadvanced tumor-bearing dysbiotic mice (Fig. 4C). Together, thesedata demonstrate that enhanced fibrosis at both local and distalsites are a long-term consequence of commensal dysbiosis duringmammary cancer.

Dysbiotic phenotype is recapitulated using nonabsorbableantibiotics

The mammary gland is considered a mucosal surface withmammary-specific commensal microorganisms. Although signif-icant metabolic changes within the mammary tissue of dysbioticor nondysbiotic mice were not observed (Supplementary Fig.S2C), suggestive of low to minimal impact on the commensalpopulation residing within the mammary gland, the systemiceffects of the antibiotic cocktail could theoretically contribute toenhanced tumor dissemination. Metronidazole is highlyabsorbed from the gut, whereas ampicillin hasmoderate systemicbioavailability after oral administration. To determine whethertumor dissemination and tissue inflammation arise due to adirecteffect of antibiotics upon the tissue microflora or due to distal

changes within the gut, commensal dysbiosis was induced usingantibiotics that have no to minimal systemic bioavailabili-ty (44, 45). This nonabsorbable antibiotic cocktail consisting ofbacitracin and neomycin is well-established to specifically depletebacteria within the gut (46).

Commensal dysbiosis initiated using nonabsorbable antibio-tics recapitulated the changes within the tissue and tumor micro-environments similar to what was observed using absorbableantibiotics. Myeloid cells infiltrating into normal-adjacent mam-mary glands of dysbiotic mice produced IL6 or arginase-1 whencommensal dysbiosis was established using absorbable or non-absorbable antibiotics (Fig. 5A and B). Early myeloid cell recruit-ment into bothmammary glands (Fig. 5C) andmammary tumors(Fig. 5D) was also enhanced in dysbiotic mice using both absorb-able and nonabsorbable antibiotics. During advanced tumorprogression, fibrosis was enhanced both in mammary tumors(Fig. 6A and B) and normal-adjacent mammary glands (Fig. 6Cand D) from dysbiotic mice using both absorbable and nonab-sorbable antibiotics.

Importantly, preestablished dysbiosis using nonabsorbableantibiotics resulted in significant dissemination of tumor cellsboth into the peripheral blood (Fig. 6E; Supplementary Fig. S6A)

A

C


Tota

l CD

11b+ /m

g m

amm

ary

tissu

e

Absorbable Nonabsorbable

**

**


Tota

l IL6

+ CD

11b+ c

ells

/m

g m

amm

ary

tissu

e ** *

Tota

lArg

1+ CD

11b+ ce

lls/

mg

mam

mar

y tis

sue

Non-dysbiosisDysbiosis

**

**



Tota

l int

ratu

mor

al C

D11

b+ cel

ls


*

*

B

Early mammary tissue Early tumors


D

SS

C

Arginase-1

SS

C

Arginase-1

SS

C

Arginase-1

SS

C

Arginase-1

SS

C

IL6

SS

C

IL6

SS

C

IL6

SS

C

IL6

Non-dysbiosis

Dysbiosis


1.94 1.94 2.4 1.44

3.82 9.15 7.06 5.72

0

20,000

40,000

60,000

80,000

0

20

40

60

80

0

5

10

15

0

2

4

6

8

10

pmn-MDSC

mono-MDSCM2

M1

M0

pmn-MDSC

mono-MDSCM2

M1

M0

Figure 5.

Direct targeting of gut commensals using nonabsorbable antibiotics results in enhanced inflammation andmyeloid cell infiltration in dysbiotic mice. C57BL/6mice were treated as described in Fig. 2A. Half of the antibiotic-treated animals were gavaged with the previously described antibiotic cocktail, while the otherhalf were gavaged with a cocktail of nonabsorbable antibiotics that haveminimal absorption from the gut. Nondysbiotic animals from each group were gavagedwith similar volumes of water per the administered antibiotic cocktail. Normal tumor-adjacent mammary glands (A–C) and BRPKp110 tumors (D) were harvestedat an early timepoint (12 days) after tumor initiation, andmyeloid cell populations were quantitated by flow cytometry. A, Representative plots gating onarginase-1þ and IL6þmyeloid cells. Numbers represent the frequency of CD11bþ cells positive for each factor. B,Arginase-1 and IL6 expression quantitated byintracellular staining in bulk CD45þCD11bþ cells frommammary glands. Numbers represent absolute numbers of cells quantitated using counting beads.Characterization of myeloid populations frommammary glands (C) or tumors (D). All populations were gated on live, singlet, CD45þCD11bþ cells. Numbersrepresent absolute numbers of cells quantitated using counting beads. M0macrophages, F4/80þCD86�CD206�; M1 macrophages, F4/80þCD86þCD206�; M2macrophages, F4/80þCD86�CD206þ; monocytic MDSC, Ly6ChiLy6G�; polymorphonuclear MDSC, Ly6CmidLy6Gþ. Representative of two independentexperiments with 5 mice/group. �, P¼ 0.01–0.05; �� , P¼ 0.001–0.01.






and to the lungs (Fig. 6F; Supplementary Fig. S6B), similar towhatwas observed in animals treated with absorbable antibiotics.Although it cannot be ruled out that nonabsorbable antibioticsindirectly affect the tissue commensalmicroorganismsor enhancebacterial translocation from the gut to distal sites such as themammary tissue, these data demonstrate that targeted disruptionof the gutmicrobiome is sufficient to initiate enhancedmetastaticdissemination during breast cancer.

Fecal transfer of a dysbiotic microbiome enhancesinflammation,myeloid cell infiltration, fibrosis, and tumor celldissemination

To determine whether a dysbiotic microbiome is sufficient toenhance mammary tumor cell dissemination, an FMT approachwas used. Cecal contents from dysbiotic and nondysbiotic donormice were collected after 14 days of antibiotic gavage followedby a 4-day rest, to recapitulate a dysbiotic microbiome at the

time of tumor initiation. Cecal contents were transferred over 3consecutive days into antibiotic-sterilized recipient SPF mice.BRPKp110 mammary tumors were initiated 7 days followingFMT gavage (Fig. 7A).

FMT of a dysbiotic or nondysbioticmicrobiome did not impacttumor kinetics (Supplementary Fig. S7A). Mice receiving FMT ofdysbiotic cecal contentsmirrored the phenotype of dysbioticmicethat had received extended gavage of absorbable or nonabsorb-able antibiotics. Specifically, mice receiving dysbiotic FMTshowed enhanced infiltration of inflammatory myeloid cells intoearly mammary tissue (Fig. 7B–D) and increased myeloid cellaccumulation into tumors (Fig. 7E). Similar effects were observedduring advanced stages of tumor progression, both within themammary tissue (Fig. 7F) and advanced tumors (Fig. 7G). Mam-mary tissue (Fig. 7H) and tumors (Fig. 7I) from mice receivingdysbiotic FMT also showed enhanced tissue fibrosis. Importantly,mice receiving dysbiotic FMT, but not "normal" FMT, had



10

20

30


20

40

60


* **

Day 12 Day 27Absorbable Nonabsorbable

0

10

20

30

40

50

60

% A

rea

stai

ned

*

Day 12 Day 27

% A

rea

stai

ned

% A

rea

stai

ned

% A

rea

stai

ned

***

Non-dysbiosis

Dysbiosis

Dysbiosis

Non-dysbiosis

Day 12 Day 27

Non-dysbiosis

Dysbiosis

Dysbiosis

Non-dysbiosis


0

10

20

30

40

50

60Day 12 Day 27

B

C D

Non-dysbiosis

Dysbiosis

Non-dysbiosis

Dysbiosis

A

Tota

l CD4

5-

lung

GFP

+ tu

mor

cel

ls



0

2,000

4,000

6,000

8,000

***

0

1,000

2,000

3,000

4,000

5,000 ****

CTCs

/mL

Bloo

d



E F

Figure 6.

Direct targeting of gut commensals using nonabsorbable antibiotics results in enhanced fibrosis and tumor cell dissemination in advanced tumor-bearingdysbiotic mice. C57BL/6 mice were treated as described in Fig. 2A. Half of the antibiotic-treated animals were gavaged with the previously described antibioticcocktail, while the other half were gavaged with a cocktail of nonabsorbable antibiotics that haveminimal absorption from the gut. Nondysbiotic animals fromeach group were gavaged with similar volumes of water per the administered antibiotic cocktail. Advanced BRPKp110 mammary tumors (A) and normal tumor-adjacent mammary glands at early (day 12) and advanced (day 27) timepoints after tumor initiation were harvested from tumor-bearing mice, with or withoutestablished dysbiosis (C). Tissues were formalin-fixed and paraffin-embedded, and sections were stained with PicroSirius Red. Quantification of staining intensitywas calculated using ImageJ software for both tumors (B) and mammary glands (D). E and F, GFPþ tumor cell disseminationwas quantified in peripheral blood(E) and lung tissue (F) by flow cytometry. Data are represented as absolute number of GFPþCD45� cells of live, singlet cells. The anti-GFP gate was chosenbased upon FMOs and a stained lung sample spiked with GFPþ BRPKp110 tumor cells. Representative of two independent experiments with 5 mice/group.� , P¼ 0.01–0.05; �� , P¼ 0.001–0.01; �� , P < 0.001.






0

20

40

60

80

0

10

20

30

40

0

500

1,000

Arginase-1 IL6

Tota

l CD

11b+

/mg

mam

mar

y tis

sue

Non-dysbiosis FMTDysbiosis FMTNon-dysbiosis FMT

Dysbiosis FMT

A

+/-

Tum

or


Gavageantibiotics

FMT withnon-dysbiosisor dysbioticcecal slurry

Begin

-9 -8 -7

B

SS

C

Arginase-1

SS

C

IL6

24.8

11.7

30.4

21 C

Non-dysbiosis FMT

Dysbiosis FMT

*

*

Tota

l CD

11b+

/mg

mam

mar

y tis

sue

D E

SynchronizeEstrus

Early andadvancedreadouts


0

20

40

60

% A

rea

stai

ned

mam

mar

y gl

and

Non-dysbiosis FMT

Dysbiosis FMT

Non-dysbiosis FMT

Dysbiosis FMT

H I

0

20

40

60

80

100

% A

rea

stai

ned

tum

or

Non-dysbiosis FMT

Dysbiosis FMT

Non-dysbiosis FMT

Dysbiosis FMT

0

10,000

20,000

30,000

40,000 **

CTC

/mL

Blo

od

Non-dysbiosis FMT

Dysbiosis FMT0

10,000

20,000

30,000

40,000

Tota

l CD

45−

lung

GFP

+ tu

mor

cel

ls

*KJ

**

***

Non-dysbiosis FMT

Dysbiosis FMT0

200

400

600

Tota

l CD

45-

lum

ph n

ode

GFP

+ tu

mor

cel

ls

*

Non-dysbiosis FMT

Dysbiosis FMT

L

Non-dysbiosis FMT

Dysbiosis FMT

Tota

l CD

11b+

/mg

mam

mar

y tis

sue

pmn-MDSCmono-MDSCM2M1M0


*

F*

G

0

50

100

150

200

250

300

350

Advanced tumor tissue

Tota

l CD

11b+

/mg

tum

or ti

ssue

Non-dysbiosis FMT

Dysbiosis FMT

**

**

*

0

500

1,000

1,500

2,000

2,500

**

**

Tota

l CD

11b+

/mg

tum

or ti

ssue

Non-dysbiosis FMT

Dysbiosis FMT

Early tumor tissue

Figure 7.

Fecal microbiota transplantation of dysbiotic cecal contents is sufficient to enhance inflammation, myeloid cell infiltration, fibrosis, and tumor cell dissemination.A, Experimental design for B–L. C57BL/6 mice were orally gavaged for 7 days with a broad-spectrum cocktail of antibiotics to create a niche for fecaltransplantation. Immediately following antibiotic cessation, mice received 3 consecutive days of oral gavage of cecal slurries collected from dysbiotic ornondysbiotic mice from day 0, as depicted in Fig. 2A. Mice were then rested for 7 days to allow for bacterial engraftment and growth prior to BRPKp110 tumorinitiation into the abdominal mammary fat pad. Mice were euthanized at early (day 12) and advanced (day 27) tumor timepoints. Four days prior to eachtimepoint, a modifiedWhitten effect was used to synchronize estrus in these animals. B and C, Arginase-1 and IL6 expression were quantitated by intracellularstaining in bulk CD45þCD11bþ cells frommammary glands. B, Representative intracellular staining of arginase-1 and IL6 frommyeloid cells within normal-adjacent mammary glands. Numbers represent percent of total CD11bþmyeloid cells. C,Quantitation of CD11bþ cells producing arginase-1 or IL6. Numbersrepresent absolute numbers quantitated using counting beads. (Continued on the following page.)






significantly enhanced tumor cell dissemination into the periph-eral blood (Fig. 7J; Supplementary Fig. S7B), within the lungs(Fig. 7K; Supplementary Fig. S7C), and to the distal axillary lymphnodes (Fig. 7L; Supplementary Fig. S7D). These data demonstratethat a dysbiotic microbiome, but not a normal microbiome, issufficient to enhance metastatic dissemination during mammarycancer.

DiscussionOur study demonstrates that commensal dysbiosis occurring

prior to mammary cancer initiation is a host-intrinsic factor thatcontributes to increased tumor cell dissemination and metastaticseeding, leading to adverse outcomes. Our data support the ideathat inflammation within the normal tissue microenvironment,in this case driven by a systemic response to commensal dysbiosis,precedes the development of aggressive breast cancer. In ourmodel of HRþmammary cancer, although primary tumor growthremained unchanged, mice with preestablished commensal dys-biosis showed significantly enhanced tumor cell dissemination tothe peripheral blood, lungs, and distal lymph nodes. Theseseemingly paradoxical data suggest that host factors associatedwith the growth of HRþ tumors are distinct from those thatcontribute to metastatic dissemination. More investigation intomyeloid function and tumor immune surveillance are necessaryto understand these differences.

To understand the role of dysbiosis in driving metastaticdissemination, several key factors known to be involved in pro-moting metastasis and adverse outcome in HRþ breast cancerwere evaluated. The syngeneic model of commensal dysbiosisused in this study recapitulates features of breast cancer that havebeen independently associated with adverse outcome: enhancedinflammation both systemically and within the tissue environ-ment; increased macrophage infiltration within tumors and nor-mal-adjacent mammary tissue; and enhanced fibrosis withinmammary tissue, tumors, and lungs. Importantly, enhanceddissemination as a result of commensal dysbiosis was observedregardless of themetastatic potential of theHRþmammary tumormodels used in this study. Therefore, our data demonstrate that adisruption in commensal homeostasis results in the establish-ment of a tissue and/or tumor microenvironment that favorsenhanced metastatic dissemination.

During commensal dysbiosis, chemokines involved inmyeloidcell chemotaxis and recruitment were enhanced both systemicallyin serum and within the mammary tissue microenvironment ofboth tumor- and nontumor-bearing dysbiotic mice. Whereasinflammation eventually resolved in dysbiotic nontumor-bearinganimals, the presence of a tumor in dysbiotic mice resulted in apersistent elevation ofmyeloid chemoattractants within the tissuemicroenvironment. Thus, it is likely that cross-talk occurs between

the tumor and tissue microenvironments, leading to sustainedinflammation andpossiblemyeloid recruitment into these tissuesin mice with commensal dysbiosis. Indeed, significantly moremyeloid cells–particularly M2-like macrophages–infiltrated intomammary tumors and normal-adjacent mammary glands duringearly and advanced stages of tumor progression. Macrophages areone of the most abundant cell types within the breast tumormicroenvironment. In patients, increased frequencies of macro-phages within stroma and tumors associate with significantlyreduced survival in HRþ breast cancer (47).Macrophages increaseangiogenesis, evasion of antitumor immunity, and promotemetastatic dissemination and seeding of cells into distal sites.Thus, the accumulation of M2 macrophages likely plays a directrole in dysbiosis-dependent enhancement of tumor celldissemination.

Inflammation within the mammary gland also increases sus-ceptibility to developing breast cancer, as women with benignbreast disease who later develop breast cancer have significantlygreater proportions of inflammatorymacrophages present withinthe tissue compared with women who do not go on to developcancer (21, 31). Similar to these human studies, in our model ofHRþ mammary cancer, significantly higher numbers of inflam-matory macrophages producing IL6 or arginase-1 were observedwithin normal-adjacent mammary tissue of mice with commen-sal dysbiosis during early stages of tumor development. Nonab-sorbable antibiotics that specifically target gut commensals andFMT of dysbiotic cecal contents also increased macrophage infil-tration and inflammation within the mammary gland. Together,our studies demonstrate that commensal dysbiosis leads to adisruption in immune homeostasis and increased inflammationwithin both tumor- and nontumor-associated tissue.

Supporting the idea that commensal dysbiosis contributes tothe evolution of a more aggressive and high-grade disease, theinduction of commensal dysbiosis using either absorbable ornonabsorbable antibiotics or FMT of dysbiotic cecal contentsresulted in significant fibrosis within both tumors and the tissuemicroenvironment. Collagen accumulation and stiffening withinthe extracellular matrix of tumors and within the tumor micro-environment is associated with increased dissemination of tumorcells into distal lymph nodes and lungs (48). These results suggestthat a disruption of the gut microflora, not mammary tissue flora,prior to tumor initiation is sufficient to facilitate collagen accu-mulation within the tissue and tumor microenvironments. Fur-thermore, these data suggest that in HRþ breast cancer, commen-sal dysbiosis leads to systemic changes that correspondwithmoreinvasive disease.

Factors that influence the development of dysbiosis are increas-ing worldwide, including rates of obesity and antibiotic use,highlighting the importance of elucidating the mechanisms bywhich commensal dysbiosis may promote cancer development

(Continued.) D–G,Mammary glands and tumors were harvested at early or advanced tumor timepoints and myeloid cell populations were quantitated by flowcytometry. All populations were gated on live, singlet, CD45þCD11bþ cells. Numbers represent absolute numbers of cells quantitated using counting beads.D, Absolute number of myeloid populations in early mammary glands. E, Absolute number of myeloid populations in early tumors. F, Absolute number ofmyeloid populations in advanced mammary glands. G, Absolute number of myeloid populations in advanced tumors. M0macrophages, F4/80þCD86�CD206�;M1 macrophages, F4/80þCD86þCD206�; M2 macrophages, F4/80þCD86�CD206þ; monocytic MDSC, Ly6ChiLy6G�; polymorphonuclear MDSC, Ly6CmidLy6Gþ.Normal-adjacent mammary glands (H) and tumors (I) were harvested from advanced tumor-bearing animals that received either dysbiotic or nondysbiotic FMT.Tissues were formalin-fixed and paraffin-embedded, and sections were stained with PicroSirius Red. Quantification of staining intensity was calculated usingImageJ software. Representative of two independent experiments with 5mice/group. J–L, GFPþ tumor cell dissemination was quantified in peripheral blood (J),lung tissue (K), and distal axillary lymph nodes (L) by flow cytometry. Data are represented as absolute number of GFPþCD45� cells of live, singlet cells. Theanti-GFP gate was chosen based upon FMOs and a stained lung sample spiked with GFPþ BRPKp110 tumor cells. Representative of two independent experimentswith 5 mice/group. �, P¼ 0.01–0.05; �� , P¼ 0.001–0.01.






andprogression.While other studies have examined the impact ofcommensal microorganisms and/or dysbiosis in response tocancer treatments (49, 50), this study is thefirst, to our knowledge,to directly assess whether commensal dysbiosis that is present atthe time of tumor initiation impacts tumor aggressiveness anddissemination. Future studies will investigate the relationshipbetween commensal microbe composition and fibrosis andmye-loid cell invasion within premalignant mammary tissue. In addi-tion, it will be important to evaluate whether commensal dys-biosis during triple-negative or HER2þ breast cancer results insimilar outcomes. Overall, our study identifies commensal dys-biosis as a preexisting, host-intrinsic factor that promotes adverseoutcomes in HRþ breast cancer. Based upon this study, wespeculate that commensal dysbiosis could thus serve as apotentialbiomarker or as a therapeutic target to reduce tumor-promotinginflammation within the tissue microenvironment.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: C. Buchta Rosean, M.R. RutkowskiDevelopment of methodology: C. Buchta Rosean, R.R. Bostic, J.C.M. Ferey,M.R. RutkowskiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Buchta Rosean, R.R. Bostic, J.C.M. Ferey,T.-Y. Feng, F.N. Azar, P.D. Bos, M.R. Rutkowski

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Buchta Rosean, R.R. Bostic, T.-Y. Feng, F.N. Azar,K.S. Tung, M.G. Dozmorov, E. Smirnova, P.D. Bos, M.R. RutkowskiWriting, review, and/or revision of the manuscript: C. Buchta Rosean,R.R. Bostic, J.C.M. Ferey, T.-Y. Feng, K.S. Tung, M.G. Dozmorov,E. Smirnova, P.D. Bos, M.R. RutkowskiStudy supervision: M.R. Rutkowski

AcknowledgmentsThis study was supported by a Susan G. Komen Career Catalyst award

CCR17483602 (to M.R. Rutkowski), IRG-17-097-31 (ACS; to M.R.Rutkowski), the University of Virginia Cancer Center and support fromNCI Cancer Center Support grant P30CA44570 as startup funds (to M.R.Rutkowski), and training grant 5T32AI007496-24 (to C. Buchta Rosean).The authors would like to acknowledge the University of Virginia ResearchHistology, Advanced Microscopy, Biorepository and Tissue Research Facility,Molecular Imaging, and Biomolecular Analysis Core Facilities, as well asthe Carter Immunology Center. In addition, we would like to thank JoseR. Conejo-Garcia for the BRPKp110 HRþ breast tumor cell line andSteven N. Fiering for the L-Stop-L-KRasG12Dp53flx/flxL-Stop-L-Myristoylatedp110a�GFPþ mice.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received November 1, 2018; revised March 12, 2019; accepted May 3, 2019;published first May 7, 2019.

References1. Burstein HJ, Cirrincione CT, Barry WT, Chew HK, Tolaney SM, Lake DE,

et al. Endocrine therapy with or without inhibition of epidermal growthfactor receptor and human epidermal growth factor receptor 2: a random-ized, double-blind, placebo-controlled phase III trial of fulvestrant with orwithout lapatinib for postmenopausal women with hormone receptor-positive advanced breast cancer-CALGB 40302 (Alliance). J Clin Oncol2014;32:3959–66.

2. DeSantis CE, Ma J, Goding Sauer A, Newman LA, Jemal A. Breast cancerstatistics, 2017, racial disparity inmortality by state. CACancer J Clin 2017;67:439–48.

3. Song F, Zhang J, Li S, Wu J, Jin T, Qin J, et al. ER-positive breast cancerpatients with more than three positive nodes or grade 3 tumors are at highrisk of late recurrence after 5-year adjuvant endocrine therapy.Onco TargetsTher 2017;10:4859–67.

4. Colzani E, Johansson AL, Liljegren A, Foukakis T, Clements M, Adolfsson J,et al. Time-dependent risk of developing distantmetastasis in breast cancerpatients according to treatment, age and tumour characteristics. Br J Cancer2014;110:1378–84.

5. Hosseini H, Obradovic MM, Hoffmann M, Harper KL, Sosa MS, Werner-KleinM, et al. Early dissemination seedsmetastasis in breast cancer. Nature2016;540:552–8.

6. Lee ES, Han W, Kim MK, Kim J, Yoo TK, Lee MH, et al. Factors associatedwith late recurrence after completion of 5-year adjuvant tamoxifen inestrogen receptor positive breast cancer. BMC Cancer 2016;16:430.

7. Yamashita H, Ogiya A, Shien T, Horimoto Y, Masuda N, Inao T, et al.Clinicopathological factors predicting early and late distant recurrence inestrogen receptor-positive, HER2-negative breast cancer. Breast Cancer2016;23:830–43.

8. Davies C, Pan H, Godwin J, Gray R, Arriagada R, Raina V, et al. Long-termeffects of continuing adjuvant tamoxifen to 10 years versus stopping at 5years after diagnosis of oestrogen receptor-positive breast cancer: ATLAS, arandomised trial. Lancet 2013;381:805–16.

9. Martin AM, Weber BL. Genetic and hormonal risk factors in breast cancer.J Natl Cancer Inst 2000;92:1126–35.

10. Rutkowski MR, Stephen TL, Svoronos N, Allegrezza MJ, Tesone AJ, Perales-Puchalt A, et al. Microbially driven TLR5-dependent signaling governs

distal malignant progression through tumor-promoting inflammation.Cancer Cell 2015;27:27–40.

11. Nechuta S, Chen WY, Cai H, Poole EM, Kwan ML, Flatt SW, et al.A pooled analysis of post-diagnosis lifestyle factors in association withlate estrogen-receptor-positive breast cancer prognosis. Int J Cancer2016;138:2088–97.

12. Ewertz M, JensenMB, Gunnarsdottir KA, Hojris I, Jakobsen EH, Nielsen D,et al. Effect of obesity on prognosis after early-stage breast cancer. J ClinOncol 2011;29:25–31.

13. Abt MC, Osborne LC, Monticelli LA, Doering TA, Alenghat T, SonnenbergGF, et al. Commensal bacteria calibrate the activation threshold of innateantiviral immunity. Immunity 2012;37:158–70.

14. Clarke T, Davis K, Lysenko E, Zhou A, Yu Y, Weiser J. Recognition ofpeptidoglycan from the microbiota by Nod1 enhances systemic innateimmunity. Nat Med 2010;16:228–31.

15. Schirmer M, Smeekens SP, Vlamakis H, Jaeger M, Oosting M, Franzosa EA,et al. Linking the human gut microbiome to inflammatory cytokineproduction capacity. Cell 2016;167:1125–36.e8.

16. Buchta Rosean CM, Rutkowski MR. The influence of the commensalmicrobiota on distal tumor-promoting inflammation. Semin Immunol2017;32:62–73.

17. Willing BP, Russell SL, Finlay BB. Shifting the balance: antibiotic effects onhost–microbiota mutualism. Nat Rev Microbiol 2011;9:233–43.

18. Kim YG, Udayanga KG, Totsuka N, Weinberg JB, N�u~nez G, Shibuya A. Gutdysbiosis promotes M2 macrophage polarization and allergic airwayinflammation via fungi-induced PGE2. Cell Host Microbe 2014;15:95–102.

19. Arendt LM, McCready J, Keller PJ, Baker DD, Naber SP, Seewaldt V, et al.Obesity promotes breast cancer by CCL2-mediated macrophage recruit-ment and angiogenesis. Cancer Res 2013;73:6080–93.

20. Rao VP, Poutahidis T, Ge Z, Nambiar PR, Boussahmain C, Wang YY, et al.Innate immune inflammatory response against enteric bacteria Helicobac-ter hepaticus induces mammary adenocarcinoma in mice. Cancer Res2006;66:7395–400.

21. Rao VP, Poutahidis T, Fox JG, Erdman SE. Breast cancer: should gastroin-testinal bacteria be on our radar screen? Cancer Res 2007;67:847–50.






22. Sheen MR, Marotti JD, Allegrezza MJ, Rutkowski M, Conejo-Garcia JR,Fiering S. Constitutively activated PI3K accelerates tumor initiation andmodifies histopathology of breast cancer. Oncogenesis 2016;5:e267.

23. Allegrezza MJ, Rutkowski MR, Stephen TL, Svoronos N, Perales-Puchalt A,Nguyen JM, et al. Trametinib drives T-cell-dependent control of KRAS--mutated tumors by inhibiting pathological myelopoiesis. Cancer Res2016;76:6253–65.

24. Bos PD, Plitas G, Rudra D, Lee SY, Rudensky AY. Transient regulatory T cellablation deters oncogene-driven breast cancer and enhances radiotherapy.J Exp Med 2013;210:2435–66.

25. Hill DA, Hoffmann C, Abt MC, Du Y, Kobuley D, Kirn TJ, et al. Metage-nomic analyses reveal antibiotic-induced temporal and spatial changes inintestinal microbiota with associated alterations in immune cell homeo-stasis. Mucosal Immunol 2010;3:148–58.

26. Harney AS, Arwert EN, Entenberg D, Wang Y, Guo P, Qian BZ, et al. Real-time imaging reveals local, transient vascular permeability, and tumor cellintravasation stimulated by TIE2hi macrophage–derived VEGFA.Cancer Discov 2015;5:932–43.

27. Becattini S, Taur Y, Pamer EG. Antibiotic-induced changes in the intestinalmicrobiota and disease. Trends Mol Med 2016;22:458–78.

28. Croswell A, Amir E, Teggatz P, BarmanM, Salzman NH. Prolonged impactof antibiotics on intestinal microbial ecology and susceptibility to entericSalmonella infection. Infect Immun 2009;77:2741–53.

29. Iyengar NM, Zhou XK, Gucalp A, Morris PG, Howe LR, Giri DD, et al.Systemic correlates of white adipose tissue inflammation in early-stagebreast cancer. Clin Cancer Res 2016;22:2283–9.

30. Dalal SJ, Estep JS, Valentin-Bon IE, Jerse AE. Standardization of theWhittenEffect to induce susceptibility to Neisseria gonorrhoeae in female mice.Contemp Top Lab Anim Sci 2001;40:13–7.

31. Degnim AC, Hoskin TL, Arshad M, Frost MH, Winham SJ, Brahmbhatt R,et al. Alterations in the immune cell composition in premalignant breasttissue that precede breast cancer development. Clin Cancer Res 2017;23:3945–52.

32. Jafarzadeh A, Fooladseresht H, Nemati M, Assadollahi Z, Sheikhi A,Ghaderi A. Higher circulating levels of chemokine CXCL10 in patientswith breast cancer: evaluation of the influences of tumor stage andchemokine gene polymorphism. Cancer Biomark 2016;16:545–54.

33. Ejaeidi AA, Craft BS, Puneky LV, Lewis RE, Cruse JM. Hormone receptor-independent CXCL10 production is associated with the regulation ofcellular factors linked to breast cancer progression and metastasis.Exp Mol Pathol 2015;99:163–72.

34. Kitamura T, Qian BZ, Soong D, Cassetta L, Noy R, Sugano G, et al. CCL2-induced chemokine cascade promotes breast cancer metastasis by enhanc-ing retention of metastasis-associated macrophages. J Exp Med 2015;212:1043–59.

35. Ueno T, Toi M, Saji H, Muta M, Bando H, Kuroi K, et al. Significance ofmacrophage chemoattractant protein-1 inmacrophage recruitment, angio-

genesis, and survival in human breast cancer. Clin Cancer Res 2000;6:3282–9.

36. Qian BZ, Li J, Zhang H, Kitamura T, Zhang J, Campion LR, et al. CCL2recruits inflammatory monocytes to facilitate breast-tumour metastasis.Nature 2011;475:222–5.

37. Sharma B, Nawandar DM, Nannuru KC, Varney ML, Singh RK. TargetingCXCR2 enhances chemotherapeutic response, inhibits mammary tumorgrowth, angiogenesis, and lung metastasis. Mol Cancer Ther 2013;12:799–808.

38. Rui J, Chunming Z, Binbin G, Na S, Shengxi W, Wei S. IL-22 promotes theprogression of breast cancer through regulating HOXB-AS5. Oncotarget2017;8:103601–12.

39. Sheng S, Zhang J, Ai J, Hao X, Luan R. Aberrant expression of IL-23/IL-23Rin patients with breast cancer and its clinical significance. Mol Med Rep2018;17:4639–44.

40. WilliamsCB, Yeh ES, Soloff AC. Tumor-associatedmacrophages: unwittingaccomplices in breast cancer malignancy. NPJ Breast Cancer 2016;2. doi:10.1038/npjbcancer.2015.25.

41. Gwak JM, Jang MH, Kim DI, Seo AN, Park SY. Prognostic value of tumor-associated macrophages according to histologic locations and hormonereceptor status in breast cancer. PLoS One 2015;10:e0125728.

42. Boyd NF, Guo H, Martin LJ, Sun L, Stone J, Fishell E, et al. Mammographicdensity and the risk and detection of breast cancer. N Engl JMed 2007;356:227–36.

43. Huo CW, Hill P, Chew G, Neeson PJ, Halse H, Williams ED, et al. Highmammographic density in women is associated with protumor inflam-mation. Breast Cancer Res 2018;20:92.

44. Donoso J, Craig GO, Baldwin RS. The distribution and excretion of zincbacitracin-14C in rats and swine. Toxicol Appl Pharmacol 1970;17:366–74.

45. Pedersoli WM, Ravis WR, Jackson J, Shaikh B. Disposition and bio-availability of neomycin in Holstein calves. J Vet Pharmacol Ther 1994;17:5–11.

46. Aguilera M, Cerd�a-Cu�ellar M, Martínez V. Antibiotic-induced dysbiosisalters host-bacterial interactions and leads to colonic sensory and motorchanges in mice. Gut Microbes 2015;6:10–23.

47. Gwak JM, Jang MH, Kim DI, Seo AN, Park SY. Prognostic value of tumor-associated macrophages according to histologic locations and hormonereceptor status in breast cancer. PLoS One 2015;10:e0125728.

48. Cox TR, Erler JT. Molecular pathways: connecting fibrosis and solid tumormetastasis. Clin Cancer Res 2014;20:3637–43.

49. Vetizou M, Pitt JM, Daillere R, Lepage P, Waldschmitt N, Flament C, et al.Anticancer immunotherapy by CTLA-4 blockade relies on the gut micro-biota. Science 2015;350:1079–84.

50. Iida N, Dzutsev A, Stewart CA, Smith L, BouladouxN,Weingarten RA, et al.Commensal bacteria control cancer response to therapy bymodulating thetumor microenvironment. Science 2013;342:967–70.






2019;79:3662-3675. Published OnlineFirst May 7, 2019.Cancer Res Claire Buchta Rosean, Raegan R. Bostic, Joshua C.M. Ferey, et al.

Positive Breast Cancer−Receptor Tissue Inflammation and Tumor Cell Dissemination in Hormone

Preexisting Commensal Dysbiosis Is a Host-Intrinsic Regulator of

Updated version

10.1158/0008-5472.CAN-18-3464doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2019/05/07/0008-5472.CAN-18-3464.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/79/14/3662.full#ref-list-1

This article cites 49 articles, 17 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/79/14/3662.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/79/14/3662To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-18-3464

http://cancerres.aacrjournals.org/content/suppl/2019/05/07/0008-5472.CAN-18-3464.DC1

http://cancerres.aacrjournals.org/content/79/14/3662.full#ref-list-1

http://cancerres.aacrjournals.org/content/79/14/3662.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/79/14/3662


preexisting commensal dysbiosis is a host- intrinsic ... · tumor biology and immunology...

Documents