mycobacterium bovis bacillus calmette-guérin vaccination ... ·...

of September 23, 2012.This information is current as Dependent Nitric Oxide Production−

IL-1RRestraining In Vivo T Cell Priming via Innate Myeloid-Derived Suppressor CellsCalmette-Guérin Vaccination Mobilizes

BacillusMycobacterium bovis

Combadière, Christophe Combadière and Nathalie WinterBalloy, Michel Chignard, Michel-Yves Mistou, Béhazine Angelo Martino, Edgar Badell, Valérie Abadie, Viviane

http://www.jimmunol.org/content/184/4/2038doi: 10.4049/jimmunol.0903348January 2010;

2010; 184:2038-2047; Prepublished online 18J Immunol

MaterialSupplementary

8.DC1.htmlhttp://www.jimmunol.org/content/suppl/2010/01/13/jimmunol.090334

Referenceshttp://www.jimmunol.org/content/184/4/2038.full#ref-list-1

, 23 of which you can access for free at: cites 45 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2010 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on September 23, 2012

http://jimm

unol.org/D

ownloaded from

http://jimmunol.org/cgi/adclick/?ad=33331&adclick=true&url=http%3A%2F%2Fwww.biolegend.com%2Fbrilliantviolet

http://jimmunol.org/

The Journal of Immunology

Mycobacterium bovis Bacillus Calmette-Guerin VaccinationMobilizes Innate Myeloid-Derived Suppressor CellsRestraining In Vivo T Cell Priming via IL-1R–DependentNitric Oxide Production

Angelo Martino,* Edgar Badell,*,1 Valerie Abadie,*,1 Viviane Balloy,† Michel Chignard,†

Michel-Yves Mistou,‡ Behazine Combadiere,x Christophe Combadiere,x and Nathalie Winter*

Early immune response to the largely usedMycobacterium bovis bacillus Calmette-Guerin (BCG) intradermal vaccine remains ill defined.

Three days after BCG inoculation into the mouse ear, in addition to neutrophils infiltrating skin, we observed CD11b+Ly-6CintLy-6G2

myeloid cells.Neutrophil depletionmarkedly enhanced their recruitment.These cells differed from inflammatorymonocytes andrequired

MyD88-dependent BCG-specific signals to invade skin, whereas neutrophil influx was MyD88 independent. Upon BCG phagocytosis,

CD11b+Ly-6CintLy-6G2 cells producedNO,which required the IL-1 receptor.DespiteNOproduction, theywere unable to kill BCGor the

nonpathogenicMycobacterium smegmatis. However, theymarkedly impairedT cell priming in the draining lymphnode.Their elimination

by all-trans retinoid acid treatment increased the number of IFN-g–producing CD4 T cells. Thus, BCG vaccination recruits innate

myeloid-derived suppressor cells, akin to mouse tumor-infiltrating cells. These propathogenic cells dampen the early T cell response

and might facilitate BCG persistence. The Journal of Immunology, 2010, 184: 2038–2047.

Promptly after pathogen entry or danger signal encounter,early tissue inflammatory infiltrate includes neutrophils,the master cell type involved in controlling invaders, and

monocytes (MOs), which mature in dendritic cells (DCs) or tissuemacrophages. MOs circulating in the blood are heterogeneous inphenotype and function (1). In the mouse, the different levels ofCX3CR1 expression, the fractalkine receptor, and Ly-6C, a gly-cosylphosphatidylinositol-anchored protein, at the cell surfacedefine two different subsets (2): CX3CR1loLy-6Chi MOs migrateto inflamed tissues upon CCR2 activation of the host responseagainst infection (3), and CX3CR1hiLy-6CloCCR22 MOs re-plenish the resident tissue macrophage pool under steady-stateconditions (2). Blood MOs with intermediate levels of Ly-6C mayrepresent different differentiation phases of a continuum (4–6) orbona fide MO subsets (7). Bone marrow-derived precursors for

neutrophils and MOs (8), sometimes referred to as band-cells orring-shaped nucleus cells, also circulate under steady-state con-ditions (9).Inflammation, intertwined with the innate immune response,

becomes involved in adaptive immunity whenDCs prime T cells forcontrol of primary infection or for protection against a secondpathogen encounter. In response to moderate inflammatory signals,CX3CR1loLy-6Chi MOs migrate to the draining lymph node, wherethey differentiate in DCs and prime the T cell response (10). Uponvaccination, the consequences of these early immune events onprotective immunity remain ill defined. Notably, understandinginflammatory/innate response to the tuberculosis vaccine Myco-bacterium bovis bacillus Calmette-Guerin (BCG) might give someclues as to why BCG does not consistently protect adults againstpulmonary disease (11).We showed that during BCG vaccination inmouse models, neutrophils shuttle BCG from the skin to thedraining lymph node (12) and establish cross-talk with DCs to fa-cilitate the presentation of Ag to T cells (13).Inflammation is a double-edged sword that must be tightly con-

trolled. The first checkpoint occurs during acute inflammation andentails the active coordination of suppression of neutrophil influxand local release of molecular host response machinery. The secondcheckpoint involvesbackupprocesses thatdampen the inflammatoryT cell responses to avoid chronic inflammation and tissue injury.These include the production of anti-inflammatory cytokines byinnate cells and generation of regulatory T cells. Myeloid-derivedsuppressor cells (MDSCs) suppress T cells in pathological situationssuch as cancer (14) or sepsis (15). MDSCs constitute a heteroge-neous population of CD11b+Gr1 (Ly-6C/Ly-6G)+ cells, comprisingMOs, granulocytes, and DCs at different stages of differentiation(16). In tumor-bearing mice, MDSCs expand in the blood, spleen,and tumors, where they contribute to tumor-associated Ag-specificT cell suppression and tolerance via NO production (17, 18).We characterized innate/inflammatory cells infiltrating the skin

early after BCG injection and their effects on T cell priming in thedraining lymph node in a mouse model of intradermal BCG

*Institut Pasteur Unite Genetique Mycobacterienne; †Institut Pasteur and InstitutNational de la Sante et de la Recherche Medicale, Unite Defense Innee etInflammation; ‡Institut Pasteur Unite Biologie des Bacteries pathogenes a Gram-positif; and xInstitut National de la Sante et de la Recherche Medicale, Laboratoired’Immunologie Cellulaire, Universite Pierre et Marie Curie-Paris 6, AP-HP GroupeHospitalier Pitie-Salpetriere Service d’Immunologie, Paris, France

1E.B. and V.A. contributed equally to this work.

Received for publication October 13, 2009. Accepted for publication December 10,2009.

This work was supported by the European Union Contract LSHP-CT-2003-503240MUVAPRED and Agence Nationale de la Recherche sur le SIDA. A.M. receives anInstitut National du Cancer fellowship.

Address correspondence and reprint requests to Dr. Nathalie Winter, INRA, United’Infectiologie Animale et Sante Publique, Centre de Recherches de Tours, 37380Nouzilly, France. E-mail address: [email protected]

The online version of this paper contains supplemental material.

Abbreviations used in this paper: ATRA, all-trans retinoic acid; BCG, bacillus Calmette-Guerin; DC, dendritic cell; EGFP, enhanced GFP; inMDSC, innate myeloid-derivedsuppressor cell; iNOS, inducible NO synthase; L-NMMA, L-NG-monomethyl argininecitrate; Ly-6Cint, intermediate level of Ly-6C; MDSC, myeloid-derived suppressor cell;MO, monocyte; NOS2, NO synthase 2; SFC, spot forming cell; SkEx, skin explant.

Copyright� 2010 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/10/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.0903348


http://jimm

unol.org/D

ownloaded from

mailto:[email protected]


vaccination. During the first 3 d, after BCG injection, we observedthe recruitment of a subset of CD11b+ myeloid-derived cells inaddition to neutrophils in the skin. These CD11b+ cells producedinducible NO synthase (iNOS) but were unable to kill mycobac-teria. However, they blocked T cell proliferation and dampenedAg-specific priming in the draining lymph node through NOproduction in vivo. Our findings thus showed that BCG vaccina-tion recruits innate MDSCs, similar to those infiltrating tumors inmice, that impair the in vivo T cell response.

Materials and MethodsInjection into ear skin in mice and mycobacterial strains

Female C57BL/6 or BALB/c (Centre d’Elevage Janvier, Le Genest-Saint-Isle,France), IL-1R2/2 and F1 C57BL/6 3 CX3CR1GFP/GFP (Centre de Cry-oconservation, Distribution, Typage et Archivage Animal, Orleans, France)and in house-bredMyD882/2, TLR23 42/2, TLR92/2, and OT-II micewerehoused in our animal care facility under specific pathogen-free conditions.After anesthesia with 5 ml xylamine 2% (v/v) and 1 mg ketamine, each mousereceived 10 ml culture containing 106 CFUs of mycobacteria per ear dorsum.Strains used were BCG Pasteur 1173P2, wild-type or recombinant producingthe enhanced GFP (EGFP) (strain Myc409) (12), or the fast-growing Myco-bacterium smegmatis strain mc2155 (19). Mock injections were performedwith PBS. All experiments were approved by the ethical committee of theInstitut Pasteur. They were conducted in full compliance with French andEuropean regulations and guidelines on experiments with live animals.

Measurement of NO, blocking of MDSC differentiation, andall-trans retinoic acid treatment

NO produced in cell supernatants from 23 105 cells cultured for 72 h in 2 mlcomplete RPMI 1640 medium (RPMI supplemented with 10% [v/v] FCS[Dominique Dutscher, Brumath, France], 100 mM HEPES, and 100 mg/mlampicillin) was measured with the Griess reagent (Molecular Probes, Eugene,OR). For inducible NO synthase inhibition in vitro, 20 mg/ml L-NG-mono-methyl arginine citrate (L-NMMA; Sigma-Aldrich, St. Louis, MO) was addedduring T cell proliferation tests. For MDSC differentiation blocking in vivo,all-trans retinoic acid (ATRA) or placebo pellets (5mg, 21-d sustained release;Innovative Research of America, Sarasota, FL)were implanted s.c. under neckskin, 4 d before BCG inoculation, using methods recommended by the man-ufacturer. Lymph node cells were harvested 3 d after BCG injection.

Neutrophil depletion

Cells were depleted in vivo by one 200-mg i.p. NIMP-R14 (Rat IgG2b) (20)or 1A8 (21)mAb injection 15 h before BCG injection in skin and two 145-mgi.p. injections 24 and 48 h after the first mAb injection. NIMP-R14 and 1A8gave similar results (Supplemental Fig. 1), therefore we used NIMP-R14throughout the study. The rat IgG2b isotype mAb HB152 (kindly providedby Genvieve Milon) was used in control mice. Neutrophils were depletedin vitro from skin explant (SkEx) cells after anti–Ly-6G labeling and usingnegative magnetic bead selection according to the manufacturer’s in-structions (Miltenyi Biotec, Bergisch Gladbach, Germany).

SkEx, blood, and bone marrow cell preparation

Dorsal and ventral halves of ears from five mice were split and layered oncomplete RPMI 1640 medium for 24 h at 37˚C to recover SkEx cells asdescribed (12). Blood cells from the retro-orbital sinuswere collected in PBScontaining 10 U/ml heparin (Sanofi Aventis, Bridgewater, NJ). Blood (3 ml)was mixed with 17–18 ml Dextran 500 (Fluka, Sigma-Aldrich) 1.25% (w/v)in PBS. Erythrocytes were isolated by sedimentation for 20 min at 20˚C.Leukocyte-rich suspension was collected and centrifuged for 5 min at 1300rpm, and the remaining erythrocyteswere lysed using 1mlRBC lysing buffer(Sigma-Aldrich) before staining for flow cytometry. Bone marrow wasflushed from the femurs and tibias of C57BL/6 mice. Total cells were platedin DMEM (Life Technologies, Rockville, MD) medium, containing 10%FCS and 10% L929 supernatant for macrophages, or IMDM medium(Cambrex, East Rutherford, NJ) containing 10% FCS and 1% J558 cell su-pernatant for DCs, as previously described (22). DCs were 95% CD11c+,CD11b+, I-Adint. After overnight in vitro infection with BCG (10 bacteria/cell), DCs shifted to the I-Adhi, CD80hi, CD86hi mature phenotype.

T cell assays

OT-IImicewere injected in ear skinwith 106BCGCFUswith orwithout 50mgOVA (Sigma-Aldrich). Total lymph node cells were prepared from three in-dividual mice, and 2 3 105 cells were cultured in 200 ml HL-1 medium

(Cambrex) with 10 U/ml penicillin and streptomycin and 1 mM glutamine.OVA protein (5mg/ml; Sigma-Aldrich) was added to stimulated cells for 36 hand IFN-g spot forming cells (SFCs) were analyzed with the IFN-g ELISPOTassay (BD Biosciences, San Jose, CA). For allogenic MLR, BALB/c spleenT cells were positively selected using anti-CD3 (17A2, BD Pharmingen, SanDiego, CA) and magnetic bead selection (Miltenyi Biotec). BALB/c T cells(105) were mixed with 2.5 3 104 to 105 C57BL/6 SkEx cells. Alternatively,a fixed number of 43 104 C57BL/6 SkEx cells was added to 2.53 104 to 105

C57BL/6 bone marrow-derived BCG-activated DCs fixed with 2% (v/v)paraformaldehyde and 105 BALB/c T cells. After 72 h culture, 0.5 mCi[3H]thymidine was added in each well for the last 8 h. Under similar condi-tions, IFN-g present in the 36-h cell supernatant was measured by ELISA(R&D Systems, Minneapolis, MN).

Flow cytometry

Single suspensions were stained for surface markers for 30 min at 4˚C afterblocking Fc receptors with anti-CD16/CD32 (2.4G2), in PBS containing 0.1%(w/v) sodium azide and 10% (v/v) FCS. Cells were stained with Abs againstCD11b (M1/70), Ly-6C (AL-21), Ly-6G (1A8), CD11c (HL3), IAd-IEd (2G9)and CD31 (MEC 13.3) fromBDPharmingen or against F4/80 (A3-1, Serotec,Oxford, U.K.), CD115 (AFS98, eBioscience, San Diego, CA).

For intracellular staining, cellswerepermeabilizedwithCytofix-Cytopermkit (554714, BD Pharmingen) and incubated with rabbit polyclonal anti-NOsynthase 2 (NOS2) (SA-200, Biomol International, Plymouth Meeting, PA)followed with Alexa Fluor 594-conjugated goat anti-rabbit IgG (MolecularProbes) or with anti-TNF (BD Bioscience). Data were analyzed usingFACSCalibur; 3 3 104 events were collected and analyzed with FlowJosoftware (BD Bioscience).

Immunohistochemistry

In situ expression of Ly-6C+ was assessed in ear skin snap-frozen andembedded in 1.5 ml optimum cutting temperature compound (Tissue Tek,Sakura, Torrance, CA). Cryostat sections (5–8 mm) were fixed in cold 80%(v/v) acetone and 20% (v/v) ethanol, air dried, and blocked for 30 min inBSA containing medium (Vector Laboratories, Burlingame, CA). Sectionswere incubated with anti–Ly-6C and Alexa Fluor 633- or 594-conjugatedgoat anti-rat IgG (Molecular Probes). Rabbit polyclonal anti-NOS2 andAlexa Fluor 594- or 488-conjugated goat anti-rabbit IgG were used forNOS2 detection.

Microscope and image acquisition

For immunohistochemistry on tissues, images were acquired on microscope(Axioskop, Zeiss, Oberkochen, Germany) with objective (Achroplan 203/0.45, Zeiss) using a camera (Leica Camera AG, Solms, Germany) andLeica Qwin software (Leica) or on a confocal microscope equipped witha blue laser diode (405 nm), an Argon laser (488 nm), helium-neon laser(543 nm) and objective (Plan Apochromat 633/1.4 NA, Zeiss) with LSM510 V3.2 software (Zeiss). For cytocentrifuged cells, images were acquiredon a light microscope (Axioskop, Zeiss) equipped with an oil iris objective(Achroplan 1003/1.25 NA, Zeiss), with a camera (Leica 300F) and Qwinsoftware (Leica). For fluorescence microscopy on SkEx cells, images wereacquired on a microscope (eclipse E600; Nikon, Melville, NY) equippedwith an objective (Plan Apochromat 603/1.40, Nikon) with camera (DXM1200F, Nikon) using imaging software (Nikon, NIS document BR). Allimages were transferred to Adobe Photoshop (Adobe Systems, San Jose,CA) for editing figures.

Statistical analysis

Values are expressed as mean 6 SD unless otherwise indicated. TheKruskal-Wallis test was used first to ascertain that significant varianceexisted among the groups studied. The Mann-Whitney U test was thenused to test statistical significance of the differences between two groups,and p values (indicated on each figure) of less than 0.05 were consideredsignificant. All calculations were made using Prism 5.0 software (Graph-Pad, San Diego, CA).

ResultsCD11b+Ly-6CintLy-6G2 cells accumulate early in skin andblood after intradermal BCG

C57/BL6 mice were inoculated in each ear dorsum with 106 BCGCFUs. After 24 and 72 h, cells migrating from ear SkExs (23)were analyzed ex vivo. SkEx cells were almost exclusivelyCD11b+ myeloid cells, among which we distinguished CD11b+

Ly-6C+Ly-6G+ neutrophils (2, 5) from CD11b+Ly-6C+Ly-6G2

The Journal of Immunology 2039


http://jimm

unol.org/D

ownloaded from


cells. Neutrophils were the predominant SkEx cell type after 24 h(Fig. 1A), as described previously (12). However, a significantproportion of CD11b+Ly-6G2 cells uniformly expressed in-termediate levels of Ly-6C at the surface (Ly-6Cint) and hada heterogeneous nuclear morphology (data not shown). Whereasthe absolute numbers of neutrophils remained constant, CD11b+

Ly-6CintLy-6G2 cell numbers increased (Fig. 1B). Few CD11b+

cells were recruited after injection of PBS into the skin (data notshown). CD11b+Ly-6CintLy-6G2 cells expressed surface MHCclass II and F4/80 and low levels of the myeloid marker CD31.They were negative for CD115 and CD11c (Fig. 1C) and thuswere not DCs or MOs. Given that a major inflammatory MOsubset has been defined on the basis of CX3CR1 and Ly-6C levelsat the cell surface (7), we used CX3CR1+/GFP knockin mice (24) todecipher inflammatory MO recruitment in skin following BCG

injection. No CX3CR1 expressing cells were observed amongCD11b+Ly-6CintLy-6G2 SkEx cells 24 (not shown) or 72 h afterBCG injection (Fig. 1D). The number of CD11b+Ly-6CintLy-6G2

SkEx cells observed in the ear dermis after BCG inoculation didnot differ between CX3CR1GFP/GFP knockout and CX3CR1+/GFP

knockin mice (Fig. 1E). This suggested that CX3CR1 did notcontrol their recruitment. Thus, myeloid cells, distinct from neu-trophils, DCs, and MOs, the major actors during the inflammatoryresponse, were recruited early after BCG vaccination in skin.In order to determine the origin of CD11b+Ly-6CintLy-6G2 skin

cells, CX3CR1+/GFP mouse peripheral blood CD11b+ cells wereanalyzed 24 and 72 h after BCG vaccination. After gating outneutrophils, we found blood CD11b+Ly-6CintLy-6G2CX3CR12

cells in addition to the two expected CX3CR1+ MO subsets. Thesecells were detected also in mock-injected animals albeit to smaller

FIGURE 1. BCG injection induces early accumulation of CD11b+Ly-6CintLy-6G2 cells, which do not belong to CX3CR1 monocyte subsets, in skin and

blood.A, Cellsmigrating fromSkEx 24 h and 72 h after BCG injectionwere analyzed byflowcytometry. CD11b+ cells contained Ly-6C+Ly-6G+ neutrophils and

Ly-6CintLy-6G2 subsets. B, CD11b+Ly-6G+, Ly-6C+ neutrophils, and CD11b+Ly-6C+Ly-6G2 cells from 24 and 72 h SkEx cells were counted. The number of

neutrophils remained constant, whereas CD11b+Ly-6C+Ly-6G2 cells increased in number (mean6 SD cell number per mouse; n = 5 experiments; p = 0.0079,

Mann-WhitneyU test).C, Surface expression of CD11c,MHCclass II, F4/80, CD31, andCD115 (filled histograms) onCD11b+Ly-6CintLy-6G2 cells 72 h after

BCG injection in the ear. Open histograms represent isotype controls.D, Characterization of CD11b+Ly-6CintLy-6G2 cells pooled fromfiveCX3CR1+/GFPmice

72 h after BCG injection in the ear (gate R1). CD11b+Ly-6CintLy-6G2 cells were negative forCX3CR1 (GFP), and the number ofmigrating CD11b+Ly-6CintLy-

6G2 cells was identical in CX3CR1+/GFP knockin and CX3CR1GFP/GFP knockout mice (E). F, Two CX3CR1+ MO populations were distinguished from pe-

ripheral blood CD11b+Ly-6G2 gated cells. The Ly-6Cint population that increased in number upon BCG injection did not express CX3CR1. Cytocentrifuged

cells were observed under a light microscope harbored ring-shaped nuclei (R1, May-Grunwald-Giemsa stain). Bar, 10 mM; original magnification3100.

2040 BCG RECRUITS INNATE MYELOID-DERIVED SUPPRESSOR CELLS


http://jimm

unol.org/D

ownloaded from


levels. They displayed ring-shaped nuclei, as previously describedfor immature myeloid cells (9) (Fig. 1F). Thus, CD11b+Ly-6Cint

Ly-6G2 cells, that increased in blood and migrated to skin early inresponse to BCG, were distinct from the CX3CR1hiLy-6Clo andCX3CR1loLy-6Chi MO subsets circulating in mouse blood (2).

Neutrophil depletion increases influx ofCD11b+Ly-6CintLy-6G2

cells to skin

Upon acute inflammation, an initial massive influx of neutrophils isfollowed by the recruitment of MOs, which help to resolve in-flammation. The nature of extravasation kinetics (25) and interplaybetween the different subsets remains unclear (26, 27).We examinedthe effect of neutrophils on the recruitment of CD11b+Ly-6CintLy-6G2 cells to the dermis in CX3CR1+/GFP mice. Neutrophil-depletingAbs NIMP-R14 (rat IgG2b) (21) and 1A8 (20) were injected i.p.during the first 72 h after BCG injection. Both treatments inducedneutrophil depletion amongSkExcells and stimulated the recruitmentofCD11b+Ly-6CintLy-6G2 cells (Supplemental Fig. 1). In SkEx cellsfrom neutrophil-depleted CX3CR1+/GFP mice, we only observedCD11b+Ly-6CintLy-6G2CX3CR12 cells 24 and 72 h after BCG in-jection (Fig. 2A). Their numbers were 2.1 times and 2.9 times higherat 24 and 72 h after neutrophil depletion than in isotype-injectedcontrolmice (Fig. 2B).Wedidnot detectCX3CR1+(GFP+)SkExcellsat these early time points (data not shown), although we observedCX3CR1+ SkEx cells by day 25 following BCG injection in the ear(Supplemental Fig. 2). Neutrophil-depleted C57/BL6 mice were in-jectedwith the BCG strainMyc409 producing EGFP (12).We did notdetectLy-6C+ cells in ear cryosections 4hafterBCG injection in thesemice.Bycontrast, large numbers ofLy-6C+ cells infiltrated the dermis

between 24 and 72 h (Fig. 2C), indicating that these cells were re-cruited from the periphery. These cells mostly colocalized with BCG(Fig. 2C). Therefore, CD11b+Ly-6CintLy-6G2 myeloid cells re-cruited to skin early after BCG injectionwere increased by neutrophildepletion.

BCG-recruited CD11b+Ly-6CintLy-6G2 myeloid cells arespecialized for iNOS production

Seventy-two hours after BCG (Myc409) inoculation in C57BL/6ear skin, both SkEx neutrophils and CD11b+Ly-6CintLy-6G2 cellsshowed similar levels for EGFP staining, suggesting that they havesimilar capacities to phagocytose BCG (Fig. 3A). Mouse neu-trophils produce TNF in response to BCG infection (13). Similarlevels of intracellular TNF staining were observed for neutrophilsand CD11b+Ly-6CintLy-6G2 cells after BCG infection (Fig. 3A).Only CD11b+Ly-6CintLy-6G2 cells showed intense staining foriNOS (Fig. 3B, in red), an enzyme playing a major role in my-cobacterial control. Whereas few iNOS+ cells were detected innondepleted mice, almost all SkEx cells displayed high iNOSstaining in neutrophil-depleted mice. Accordingly, NO productionby SkEx cells was eight times higher in neutrophil-depleted thanin control mice (Fig. 3C). In situ analysis revealed colocalizationof Ly-6C+ cells (blue) with iNOS (red) and BCG (Myc409, green)(Fig. 3D). Intracellular iNOS was analyzed in blood cells fromCX3CR1+/GFP mice 72 h after BCG injection. No iNOS+ cellswere detected in mouse blood when PBS was injected in ear (datanot shown). The two CX3CR1+ blood MO subsets were negativefor iNOS. Intracellular iNOS was detected in CD11b+Ly-6CintLy-6G2CX3CR12 cells only in the presence and absence of

FIGURE 2. Neutrophil depletion increased the accumulation of CD11b+Ly-6CintLy-6G2 cells in skin and blood in response to BCG. A, Five mice were

depleted for neutrophils by NIMP-R14 injection before and after BCG inoculation. Control mice were treated with HB152 control isotype. Ly-6C and Ly-

6G surface expression on SkEx cells was analyzed 24 and 72 h after BCG injection and counted from five pooled mice (mean6 SD cell number per mouse;

n = 3 experiments) (B). C, Ear cryosections from neutrophil-depleted mice CD11b+Ly-6CintLy-6G2 cells 4, 24, and 72 h after green fluorescent BCG

(Myc409) injection were immunostained with anti–Ly-6C (red), and Ly-6C+ cells were absent in skin 4 h after BCG (Myc409) injection. Large numbers of

these cells were then recruited by 72 h and colocalized with BCG (Myc409, yellow). Bars, 50 mM; original magnification 320.



http://jimm

unol.org/D

ownloaded from


FIGURE 3. CD11b+Ly-6CintLy-6G2 cells recruited to skin, in addition to neutrophils, phagocytose BCG, secrete TNF, and express high levels of iNOS. A, BCG

(Myc409) producingEGFPwas injected into themouse ear dorsum, andSkEx cellswere analyzed 72 h later. OfmigratingCD11b+Ly-6CintLy-6G2cells and neutrophils,

62% were infected with BCG, suggesting that both subsets have a similar phagocytic capacity. Similar levels of intracellular TNF were detected for the two subsets. B,

Intracellular cytosolic iNOS (red) was detected in SkEx cells under a fluorescence microscope with DAPI-stained nuclei (blue) from neutrophil-depleted (NIMP-R14-

treated) andmice treatedwithHB152 control isotype 72 h after BCG injection. Bars, 20mM; originalmagnification350.C, NO secreted by SkEx cells pooled fromfive

mice injectedwith control isotypeHB152or neutrophil-depletingNIMP-R14wasdetected byGriess reagent (mean6SD;n=3 experiments;p=0.031, Student t test).D,

Cells infiltrating the ear tissuewere analyzed in cryosections fromneutrophil-depletedmice 72 h after BCG (Myc409, green) injection. Ly-6C+ cells (blue) infiltrating the

skin show strong staining for iNOS (red). Onmerged images, iNOS colocalized with Ly-6C and BCG (Myc409). Bars, 50mM; original maginification320. E, Among

blood CD11b+Ly-6G2 gated cells from CX3CR1+/GFP mice, intracellular iNOS was only detected in CD11b+Ly-6CintLy-6G2 cells (R1 gate). The two CX3CR1+ MO

subsets (R2andR3gates)werenegative for iNOS.CD11b+Ly-6CintLy-6G2cellsproducing iNOSweredetected inblood in thepresence (HB152control isotype injection)

and absence (NIMP-R14 injection) of neutrophils, although a higher number of these cells were found in the neutrophil-depleted mice.



http://jimm

unol.org/D

ownloaded from


neutrophils (Fig. 3E) upon BCG injection, illustrating the speci-ficity of these myeloid cells to produce iNOS both in blood andtissues.

Despite high NO production, innate CD11b+Ly-6CintLy-6G2

myeloid cells do not kill mycobacteria

NO is a major antimycobacterial molecule in the mouse. Wetherefore studied BCG killing by CD11b+Ly-6CintLy-6G2 cells.Despite their high numbers in skin from neutrophil-depleted ani-mals, the BCG load 4 h after injection did not decrease for thenext 3 (Fig. 4A) or 4 d (data not shown). Both BCG and M.smegmatis belong to the mycobacterial genus. However, whereasthe slow-growing BCG is resistant to the hostile macrophage NO-rich environment, the fast-growing M. smegmatis is rapidly killed.CD11b+Ly-6CintLy-6G2iNOS+ cells were purified from SkExcells from neutrophil-depleted BCG injected mice and incubatedin vitro with M. smegmatis for 36 h. Similar M. smegmatis CFUswere recovered from infected CD11b+Ly-6CintLy-6G2iNOS+

cells and from the medium (Fig. 4B). Therefore, despite highlevels of NO production, CD11b+Ly-6CintLy-6G2 cells did notkill pathogenic or nonpathogenic mycobacteria, whereas classi-cally activated macrophages did (Supplemental Fig. 3).

CD11b+Ly-6CintLy-6G2 cells respond to BCG and MyD88/IL-1R signals to invade skin and express maximum iNOS

To determine which signals, on the host and microbial side, mightregulate the recruitment of CD11b+Ly-6CintLy-6G2 cells and iNOSexpression, we analyzed the role ofMyD88, amajor TLR adaptor formycobacterial control (28, 29). In MyD88+/+ mice, CD11b+Ly-6Cint

Ly-6G2 cells represented one-third of the total SkEx cells 72 h afterBCG injection (Fig. 5A). By contrast, very few CD11b+Ly-6CintLy-

6G2 cells were present in MyD882/2 mice, in which neutrophillevels were six times higher. In neutrophil-depleted mice, CD11b+

Ly-6CintLy-6G2 cells were not detected among MyD882/2 SkExcells 24 h after BCG inoculation (data not shown) and were detectedat levels four times lower than in MyD88+/+ mice 72 h after BCGinjection (Fig. 5A). Major innate receptors, upstream from MyD88,instrumental inmycobacterial control include TLR2, TLR9 (30), andIL-1R (31). None of these receptors were involved in the recruitmentof CD11b+Ly-6CintLy-6G2 cells to the skin in the absence of neu-trophils (Fig. 5B). We also found that CCR2, a major receptor forCX3CR1hi MO recruitment, was not involved in this process (datanot shown).The recruitment and full activation of CD11b+Ly-6CintLy-6G2

cells may require different signals. We did not find any differences inthe levels of iNOS expression in CD11b+Ly-6CintLy-6G2 cells be-tween TLR+/+ and TLR23 42/2 or TLR92/2 SkEx cells after BCGinjection (data not shown). By contrast, despite similar numbers ofCD11b+Ly-6CintLy-6G2 SkEx recruited cells, iNOS levels weresubstantially lower in IL-1R2/2mice than in IL-1R+/+mice (Fig. 5C).Accordingly, NO levels in IL-1R2/2 SkEx cells were eight timeslower than in IL-1R+/+ SkEx cells (Fig. 5D). Therefore, IL-1R con-trolled intrinsic capacity of CD11b+Ly-6CintLy-6G2cells to be acti-vated. On the microbial side, to determine whether CD11b+Ly-6Cint

Ly-6G2 cell activation was a general response to mycobacteria, wecompared the skin response to BCG and to M. smegmatis. WhereasBCG derives from the slow-growing pathogenic speciesM. bovis,M.smegmatis is an opportunistic, fast-growing species. Both speciesshare several pathogen-associated molecular patterns, yet only BCGis able to persist in the host. Similar numbers of CD11b+Ly-6CintLy-6G2 SkEx cells were observed in CX3CR1+/GFP mice injected withM. smegmatis or BCG (data not shown). However, intracellular iNOSwas barely detected in SkEx cells from M. smegmatis-injected ears(Fig. 5E). Hence, MyD88-mediated signaling appeared to be neces-sary for CD11b+Ly-6CintLy-6G2 cell recruitment. The production ofiNOS required a second signal dependent on molecular patternspresent in BCG but not inM. smegmatis and involving IL-1R but notTLR2, -4, or -9.

CD11b+Ly-6CintLy-6G2 cells suppress T cells in vitro anddampen T cell priming in the lymph node draining BCGvaccination site

To gain further insight into CD11b+Ly-6CintLy-6G2 cell function,C57BL/6 SkEx cells were prepared from neutrophil-depleted orcontrol mice 72 h after BCG injection andmixedwith BALB/c CD3+

T cells (Fig. 6A). SkEx cells from controlmice, including neutrophils,and possibly some professional APCs such as DCs, were able tostimulate allogenic T cells. However, in vitro neutrophil depletionabolished SkEx cell stimulatory activity. Similarly, iNOS-rich SkExcells from in vivo neutrophil-depleted animals were unable to stim-ulateT cells. The iNOS inhibitor L-NMMAwas added to cell culturesand inhibited SkEx cell-mediated T cell suppression. Fixed numbersof C57BL/6 SkEx cells from BCG-vaccinated micewere then mixedwith increasing numbers of bone marrow-derived BCG-activatedC57BL/6 DCs to improve stimulatory capacity of BALB/c allogenicCD3 T cells (Fig. 6B). SkEx cells were either obtained in vivo fromneutrophil-depletedmice or invitro fromneutrophil-depleted cells. Inboth cases, SkEx cells completely suppressed allogenic T cell pro-liferation (Fig. 6B) and IFN-g secretion (Supplemental Fig. 4) despitethe presence of strongly stimulating BCG-activated DCs. This effectwas observed even at the highest DC/T cell ratio and was reversed byL-NMMA addition. iNOS+ cells from isotype-treated mice sup-pressed T cell proliferation to a lesser extent than iNOS+ cells fromneutrophil-depleted animals but still prevented IFN-g secretion(Supplemental Fig. 4). Hence, BCG vaccination in skin recruited

FIGURE 4. CD11b+Ly-6CintLy-6G2 cells did not kill mycobacteria

despite high iNOS levels. A, Fifteen C57BL/6 mice, treated with HB152

control isotype or neutrophil-depleting NIMP-R14, received 106 BCG

CFUs in each ear dorsum. After 4, 24, and 72 h, ears from five animals

were homogenized separately, and BCG load was determined for each ear

(mean 6 SD CFUs per ear; n = 10 ears). B, SkEx cells from five mice

(NIMP-R14 treated) harvested 72 h after BCG ear injection, were infected

in vitro with fast-growingM. smegmatis (multiplicity of infection: 1) in the

presence or absence of the iNOS inhibitor L-NMMA. Control wells with

medium alone were inoculated with free M. smegmatis. The CFUs de-

termined after 36 h inM. smegmatis grown intracellularly were higher than

the CFUs for cells recovered 1 h postinfection, demonstrating that the cells

were not bactericidal (mean 6 SD CFUs per well; n = 3 experiments).



http://jimm

unol.org/D

ownloaded from


CD11b+Ly-6CintLy-6G2iNOS+ innate cells, capable of marked NO-mediated T cell suppression. The phenotype and suppressive functionof skin CD11b+Ly-6CintLy-6G2 cells were reminiscent of MDSCspreviously described under pathological conditions such as in sepsisand cancer (14). We therefore called these cells innate MDSCs(inMDSCs). We also found that jointly recruited neutrophils coun-terbalanced inMDSC-mediated T cell suppression in vitro.We then investigated a potential effect of inMDSCs on T cell

priming in the draining lymph node. iNOS-producing CD11c2 cellswere detected in the auricular lymph node by immunohistochem-istry 4 and 72 h after BCG injection in the ear skin. iNOS stainingunder the lymph node capsule was stronger in neutrophil-depletedthan in isotype-treated mice (Fig. 6C). This suggested that, as pre-viously shown for neutrophils (12), inMDSCswere able to reach thelymph node. No iNOS+ cells were detected in mock-injected mice.Increased recruitment of iNOS+ cells in the draining lymph node byneutrophil depletion had no impact on BCG transport from the skin(Fig. 6D), suggesting that similar amounts of BCG-derived Agswere available for early T cell priming in both neutrophil-proficientand neutrophil-depleted mice.Neutrophil-depleted or isotype-treated TCR-Ova-specific OT-II

mice (32) were coinjected with BCG and Ova in the ear skin.Draining lymph node cells were harvested after 72 h, and IFN-gSFCs were counted after in vitro restimulation with Ova protein.In correlation with increased iNOS+ cells in draining lymph node,the number of IFN-g SFCs in neutrophil-depleted treated micewas three times lower than in controls (Fig. 6E). Recent studiesshowed that ATRA injection depletes MDSCs in mice tumors(33). We observed that Ova-specific IFN-g SFC levels followingATRA treatment in OT-II mice injected with BCG and Ova in earskin were five times higher than those in vehicle-treated mice,

suggesting that inMDSC elimination in vivo strongly increased Tcell priming following BCG vaccination (Fig. 6F). These datashow that the coordinated early recruitment of neutrophils andinMDSCs regulated early T cell priming in response to BCG-derived Ags after vaccination in skin.

DiscussionInnate MDSCs, phenotypically characterized by Ly-6C, CD11b, andCD31 cell-surface expression and functionally defined by strongNO-mediated in vitro T cell suppression, were recruited to the dermisduring the very early stages following BCG vaccination. These cellsdidnotexpressCD115orLy-6Gatthecellsurface,suggestingthattheywerenotMOsorneutrophils.MDSCswere recentlydefined indiseasestates including overwhelming infection or cancer (14). MDSCsoriginate in the bone marrow, circulate in the blood, and accumulatein lymphoid tissue after the onset of inflammation. They act asa regulatory feedback mechanism to suppress T cells (16). MouseMDSCs are phenotypically characterized by a combination ofmarkers, including CD11b and Gr1 (Ly6-C/G), and are functionallydefined by their potent NO-mediated suppression of T cells (16).Recently, two subsets have been described in mouse tumors: granu-locytic MDSCs that express both Ly-6C and Ly-6G and monocyticMDSCs that only express Ly-6C (34, 35). Both subsets are able toblockT cell proliferation bydifferentmechanisms associatedwith themetabolism of L-arginine (14). iNOS is the main effector for T cellsuppression by monocytic MDSCs (34). Whereas MDSCs accumu-late in lymphoid tissues, inMDSCs appeared early after BCG in-jection, first in the blood and skin, where they peaked by day 3. Inmice, iNOS and NO production induced by classically activatedmacrophages is essential for the destruction of mycobacteria (36).However, pathogenicmycobacteria escape this harsh environment by

FIGURE 5. The recruitment of CD11b+Ly-6CintLy-

6G2 cells to skin and iNOS production in response to

BCG-derived ligandsarecontrolledbyMyD88and IL-1R

signaling. MyD88+/+ and MyD882/2 or TLR2 3 42/2,

TLR92/2, and IL-1R2/2mice (A), treatedwithHB152 or

NIMP-R14,were injectedwithBCG (B). SkEx cellswere

harvested after 72 h, and neutrophils andCD11b+Ly-6Cint

Ly-6G2cellswere counted (mean6SDcell numbers per

mouse; n = 3 experiments). Neutrophils but not CD11b+

Ly-6CintLy-6G2 cells were detected among MyD882/2

SkEx cells. Even in the absence of neutrophils, re-

cruitment of CD11b+Ly-6CintLy-6G2 cells to the dermis

of MyD882/2 mice was impaired. C, SkEx cells were

harvested from neutrophil-depleted (NIMP-R14-in-

jected) IL-1R+/+ and IL-1R2/2 mice 72 h after BCG in-

jection,fixed, and stained for intracellular iNOS (red); cell

nuclei were stained with DAPI (blue). D, NO secreted

from SkEx IL-1R+/+ and IL-1R2/2 cells 72 h after BCG

injection was detected using Griess reagent (mean6 SD

perwell; n= 3 experiments).E, Fluorescencemicroscopy

in fixed SkEx cells from NIMP-R14-treated C57/Bl6

mice injected withM. smegmatis revealed no iNOS (red)

in cells with DAPI-stained nuclei (blue). Bars, 20 mM;

original magnification350.



http://jimm

unol.org/D

ownloaded from


adapting their genetic program for protection against NO (37), iNOSexclusion from the phagosome (38), and macrophage arginase 1 in-duction (39). Despite phagocytosis and high iNOS levels, skininMDSCswere unable to kill BCGor the fast-growingM. smegmatis,showing that mycobacteria were also able to escape the hostile NOenvironment in inMDSCs.We found that skin CD11b+Gr1+ cells, constituting the early

BCG-induced inflammatory infiltrate, were composed of Ly-6Cint

Ly-6G2 cells and Ly-6C+Ly-6G+ neutrophils. Despite expressing

iNOS and TNF, these cells differed from recently characterizedTNF and iNOS-producing DCs (3) by a number of aspects: theydid not express CD115 or CD11c, their migration from blood toskin was CCR2 independent, and they exerted suppression insteadof activation on T cells. However, in respect to their immaturephenotype, we do not exclude that inMDSCs might convert toTNF and iNOS-producing DCs under some circumstances. Un-expectedly, we did not detect the recently described inflammatoryCX3CR1lo MO population between 2 (data not shown) and 72 h

FIGURE 6. CD11b+Ly-6CintLy-6G2 cells accumulating early after BCG injection displayed characteristics reminiscent of MDSCs and dampen T cell

priming. A, Total SkEx cells were harvested 72 h after BCG injection in HB152 control isotype (filled triangles) or neutrophil-depleting NIMP-R14-treated

(open diamonds) mice and mixed at various ratios with 105 BALB/c CD3+ T cells. Proliferation was measured by [3H]thymidine incorporation. Neutrophils

were also depleted in vitro using anti–Ly-6G+ magnetic beads (open circles) in SkEx cells from HB152 control isotype-injected mice. To investigate the role

of NO secreted by CD11b+Ly-6CintLy-6G2 SkEx in T cell suppression, L-NMMAwas added to cells from neutrophil-depleted (NIMP-R14–treated) mice

(filled diamonds). B, The proliferation of BALB/c CD3+ T cells (105/well) proliferation was measured in response to 105 to 2.5 3 104 allogeneic C57/BL6

BCG-activated DCs (filled squares). Addition of 4 3 103 CD11b+Ly-6CintLy-6G2 SkEx cells, obtained from C57/BL6 mice depleted for neutrophils

in vivo (NIMP-R14 treatment, open diamonds), completely suppressed T cell proliferation in response to activated DCs, even at the highest DC/T cell ratio.

In vitro magnetic bead depletion of Ly-6G+ neutrophils in SkEx cells from HB152-treated mice (open circles) also suppressed T cells, although not

completely. The addition of L-NMMA to CD11b+Ly-6CintLy-6G2 cells from SkEx cells depleted for neutrophils in vivo (filled diamonds) or in vitro (filled

circles) reversed their T cell suppressive activity. C, Immunohistochemistry was performed on auricular lymph node cryosections 4 and 72 h after BCG ear

injection. CD11c+ DCs (red) and iNOS+ cells (green) were observed, but these cells did not colocalize. Higher amounts of subcapsular iNOS+ cells were

detected in the absence of neutrophils (NIMP-R14 treatment); this effect was particularly marked at 72 h after BCG injection. No iNOS+ cells were

observed in mock-treated mice. Bars = 50 mM; original magnification 320. D, NIMP-R14– and HB152-treated mice were injected in ear skin with BCG.

CFU values were determined in the draining lymph node 4 and 72 h later (mean CFU 6 SD; n = 5 mice). TCR Ova transgenic OT-II mice treated with

NIMP-R14 to induce the recruitment of CD11b+Ly-6CintLy-6G2 MDSC-like cells to the draining lymph node (E) or with ATRA to eliminate CD11b+Ly-

6CintLy-6G2 MDSC-like cells (F) were injected in the ear with BCG (106 CFU) with or without 50 mg Ova. After 72 h, IFN-g SFCs were analyzed in

draining lymph node after restimulation in vitro with Ova protein (mean 6 SD; n = 3 mice; p = 0.0133 in E; p = 0.0004 in F, unpaired t test).



http://jimm

unol.org/D

ownloaded from


after BCG inoculation in the skin. However, migrating CX3CR1+

SkEx cells were observed by day 25 after BCG inoculation(Supplemental Fig. 2), demonstrating that these cells were re-cruited to skin at later time points. Although we cannot excludethat minute amounts of CX3CR1+Ly-6C+ MOs were presentduring the early stages of infection in skin, the BCG-inducedenvironment may delay the recruitment of inflammatory MOs.Consistent with this, Rotta et al. (40) described that strong in-flammatory signals, delivered through TLR4 from endotoxin orSalmonella typhimurium injection in skin, blocked the migrationof inflammatory MOs migration and differentiation of DCs in thedraining lymph node, occurring under mild inflammatory con-ditions (10). Upon disruption of homeostasis, the content of theinflammatory cell infiltrate must be tightly regulated. It is widelyaccepted that neutrophils infiltrate the lesion first and are swiftlyfollowed by MOs, although the role played by neutrophils in thecontrol of MO recruitment is unclear (26, 27, 41). The poorspecificity of anti-Gr1 (Ly-6C/Ly-6G) Ab has contributed toconfusion in the field. Upon BCG injection, we observed co-ordinated fluctuations in neutrophil and inMDSC levels in skininfiltrates over time; although neutrophils remained the pre-dominant cell type, inMDSCs numbers increased at later timepoints. Moreover, the specific depletion of neutrophils markedlyincreased the recruitment of inMDSCs. These data suggested thatthe recruitment of inMDSCs was suppressed by neutrophils.The recruitment of neutrophils and inMDSCs to skin was regulated

by different signaling pathways. Whereas neutrophils did not needMyD88to reach the skin, inMDSCswerestrictlydependentonMyD88signaling, highlighting differences between the host mechanismscontrolling these two cell populations. MyD88-deficient mice cannotsuppress lung Mycobacterium tuberculosis (28, 29) or BCG (42)multiplication. Tissue infiltrates in these mice also display higherlevels of Ly-6G+ granulocytic cells than those inMyD88+/+mice (28),consistent with theMyD88-independent recruitment of neutrophils inresponse to mycobacteria. Given the suppressive function ofinMDSCs, the extensive lung inflammation observed inmycobacteria-infected MyD882/2mice (28, 42) could be partly due to unrestrictedneutrophils influx and/or the loss of inMDSCs. MyD88 signaling is atthe crossroads of TLR-, IL-1R–, and IL-18R–dependent pathways.We did not observe the impaired influx of inMDSCs in skin in TLR23 42/2, TLR 92/2, or IL-1R2/2mice. The receptor upstream fromMyD88 involved in the recruitment of these cells remains unknown.However, levels of inMDSC-induced iNOS and NO production in IL-1R2/2mice, which, similarly toMyD882/2mice (31), are susceptibleto tuberculosis infection, were greatly reduced. The b form of pro-IL-1–induced gastric inflammation and cancer have recently been linkedto the recruitment ofmouseMDSCs (43).Hence, theb formof pro-IL-1 and IL-1R appear to be master regulators of inMDSC activation.NO produced by inMDSCs was unable to inhibit mycobacterial

growth, but completely suppressed T cell proliferation in vitro andcurtailed T cell priming in vivo. In BCG-injected mice, increasedrecruitment of iNOS+ cells to the lymph node after neutrophildepletion correlated with decreased T cell priming. Conversely,in vivo treatment with ATRA, which eliminates MDSCs, increasedT cell priming (33). In mice, treatment with ATRA significantlyenhanced cancer vaccine efficacy (44). In mycobacteria-inducedexperimental autoimmune encephalomyelitis models, CD11b+

Gr1+Ly-6G2 cells are compartmentalized in spleen and preventT cell expansion by secreting NO upon contact with Ag-activatedT cells. NO-mediated suppression is counterbalanced by O2

2

produced by neutrophils (45). These feedback mechanisms havemajor effects on experimental autoimmune encephalomyelitispathogenesis. Our data are consistent with the hypothesis that thecoordinated innate influx of neutrophils and inMDSCs, and the

associated NO/O22 balance, regulates BCG Ag-specific T cell

priming during vaccination. This mechanism may control theBCG-induced T cell response level and avoid excessive tissueinflammation. Alternatively, the recruitment of propathogeniccells, such as inMDSCs, which are unable to kill BCG and delay Tcell priming, could be an additional mechanism employed byslow-growing mycobacteria to parasitize their host.

AcknowledgmentsWe thank Dmitry Gabrilovich for key discussions and advice. We thank

Dominique Buzoni-Gatel, Genevieve Milon, and Jean-Marc Cavaillon

for constructive discussions and Pieter Leenen for the critical reading of

our manuscript. CX3CR1GFP/GFP mice were originally from Dan Littman’s

laboratory. We thank Genevieve Milon, Dominique Buzoni-Gatel, and

Steffen Jung for making them available to us. We warmly thank Bernhard

Ryffel for IL-1R2/2 mice.

DisclosuresThe authors have no financial conflicts of interest.

References1. Hume, D. A. 2006. The mononuclear phagocyte system. Curr. Opin. Immunol.

18: 49–53.2. Geissmann, F., S. Jung, and D. R. Littman. 2003. Blood monocytes consist of

two principal subsets with distinct migratory properties. Immunity 19: 71–82.3. Serbina, N. V., T. Jia, T. M. Hohl, and E. G. Pamer. 2008. Monocyte-mediated

defense against microbial pathogens. Annu. Rev. Immunol. 26: 421–452.4. Qu, C., E. W. Edwards, F. Tacke, V. Angeli, J. Llodra, G. Sanchez-Schmitz,

A. Garin, N. S. Haque, W. Peters, N. van Rooijen, et al. 2004. Role of CCR8 andother chemokine pathways in the migration of monocyte-derived dendritic cellsto lymph nodes. J. Exp. Med. 200: 1231–1241.

5. Sunderkotter, C., T. Nikolic, M. J. Dillon, N. Van Rooijen, M. Stehling,D. A. Drevets, and P. J. Leenen. 2004. Subpopulations of mouse blood mono-cytes differ in maturation stage and inflammatory response. J. Immunol. 172:4410–4417.

6. Tacke, F., F. Ginhoux, C. Jakubzick, N. van Rooijen, M. Merad, andG. J. Randolph. 2006. Immature monocytes acquire antigens from other cells inthe bone marrow and present them to T cells after maturing in the periphery.J. Exp. Med. 203: 583–597.

7. Geissmann, F., C. Auffray, R. Palframan, C. Wirrig, A. Ciocca, L. Campisi,E. Narni-Mancinelli, and G. Lauvau. 2008. Blood monocytes: distinct subsets,how they relate to dendritic cells, and their possible roles in the regulation ofT-cell responses. Immunol. Cell Biol. 86: 398–408.

8. Gordon, S., and P. R. Taylor. 2005. Monocyte and macrophage heterogeneity.Nat. Rev. Immunol. 5: 953–964.

9. Biermann, H., B. Pietz, R. Dreier, K. W. Schmid, C. Sorg, and C. Sunderkotter.1999. Murine leukocytes with ring-shaped nuclei include granulocytes, mono-cytes, and their precursors. J. Leukoc. Biol. 65: 217–231.

10. Randolph, G. J., K. Inaba, D. F. Robbiani, R. M. Steinman, and W. A. Muller.1999. Differentiation of phagocytic monocytes into lymph node dendritic cellsin vivo. Immunity 11: 753–761.

11. Colditz, G. A., T. F. Brewer, C. S. Berkey, M. E. Wilson, E. B. Burdick,H. V. Fineberg, and F. Mosteller. 1994. Efficacy of BCG vaccine in the pre-vention of tuberculosis. Meta-analysis of the published literature. JAMA 271:698–702.

12. Abadie, V., E. Badell, P. Douillard, D. Ensergueix, P. J. Leenen, M. Tanguy,L. Fiette, S. Saeland, B. Gicquel, and N. Winter. 2005. Neutrophils rapidlymigrate via lymphatics after Mycobacterium bovis BCG intradermal vaccinationand shuttle live bacilli to the draining lymph nodes. Blood 106: 1843–1850.

13. Morel, C., E. Badell, V. Abadie, M. Robledo, N. Setterblad, J.-C. Gluckman,B. Gicquel, S. Boudaly, and N. Winter. 2008. Mycobacterium bovis BCG-infected neutrophils and dendritic cells cooperate to induce specific T cell re-sponses in humans and mice. Eur. J. Immunol. 38: 437–447.

14. Gabrilovich, D. I., and S. Nagaraj. 2009. Myeloid-derived suppressor cells asregulators of the immune system. Nat. Rev. Immunol. 9: 162–174.

15. Delano, M. J., P. O. Scumpia, J. S. Weinstein, D. Coco, S. Nagaraj, K. M. Kelly-Scumpia, K. A. O’Malley, J. L. Wynn, S. Antonenko, S. Z. Al-Quran, et al. 2007.MyD88-dependent expansion of an immature GR-1(+)CD11b(+) populationinduces T cell suppression and Th2 polarization in sepsis. J. Exp. Med. 204:1463–1474.

16. Sica, A., and V. Bronte. 2007. Altered macrophage differentiation and immunedysfunction in tumor development. J. Clin. Invest. 117: 1155–1166.

17. Nagaraj, S., and D. I. Gabrilovich. 2008. Tumor escape mechanism governed bymyeloid-derived suppressor cells. Cancer Res. 68: 2561–2563.

18. Serafini, P., I. Borrello, and V. Bronte. 2006. Myeloid suppressor cells in cancer:recruitment, phenotype, properties, and mechanisms of immune suppression.Semin. Cancer Biol. 16: 53–65.



http://jimm

unol.org/D

ownloaded from


19. Snapper, S. B., R. E. Melton, S. Mustafa, T. Kieser, and W. R. Jacobs, Jr. 1990.Isolation and characterization of efficient plasmid transformation mutants ofMycobacterium smegmatis. Mol. Microbiol. 4: 1911–1919.

20. Fleming, T. J., M. L. Fleming, and T. R. Malek. 1993. Selective expression ofLy-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb togranulocyte-differentiation antigen (Gr-1) detects members of the Ly-6 family.J. Immunol. 151: 2399–2408.

21. Lopez, A. F., M. Strath, and C. J. Sanderson. 1984. Differentiation antigens onmouse eosinophils and neutrophils identified by monoclonal antibodies. Br.J. Haematol. 57: 489–494.

22. Rousseau, C., N. Winter, E. Pivert, Y. Bordat, O. Neyrolles, P. Ave, M. Huerre,B. Gicquel, and M. Jackson. 2004. Production of phthiocerol dimycocerosatesprotects Mycobacterium tuberculosis from the cidal activity of reactive nitrogenintermediates produced by macrophages and modulates the early immune re-sponse to infection. Cell. Microbiol. 6: 277–287.

23. Romani, N., G. Ratzinger, K. Pfaller, W. Salvenmoser, H. Stossel, F. Koch, andP. Stoitzner. 2001. Migration of dendritic cells into lymphatics-the Langerhanscell example: routes, regulation, and relevance. Int. Rev. Cytol. 207: 237–270.

24. Jung, S., J. Aliberti, P. Graemmel, M. J. Sunshine, G. W. Kreutzberg, A. Sher,and D. R. Littman. 2000. Analysis of fractalkine receptor CX(3)CR1 function bytargeted deletion and green fluorescent protein reporter gene insertion. Mol. Cell.Biol. 20: 4106–4114.

25. Auffray, C., D. Fogg, M. Garfa, G. Elain, O. Join-Lambert, S. Kayal, S. Sarnacki,A. Cumano, G. Lauvau, and F. Geissmann. 2007. Monitoring of blood vesselsand tissues by a population of monocytes with patrolling behavior. Science 317:666–670.

26. Garcıa-Ramallo, E., T. Marques, N. Prats, J. Beleta, S. L. Kunkel, andN. Godessart. 2002. Resident cell chemokine expression serves as the majormechanism for leukocyte recruitment during local inflammation. J. Immunol.169: 6467–6473.

27. Henderson, R. B., J. A. R. Hobbs, M. Mathies, and N. Hogg. 2003. Rapid re-cruitment of inflammatory monocytes is independent of neutrophil migration.Blood 102: 328–335.

28. Fremond, C. M., V. Yeremeev, D. M. Nicolle, M. Jacobs, V. F. Quesniaux, andB. Ryffel. 2004. Fatal Mycobacterium tuberculosis infection despite adaptiveimmune response in the absence of MyD88. J. Clin. Invest. 114: 1790–1799.

29. Scanga, C. A., A. Bafica, C. G. Feng, A. W. Cheever, S. Hieny, and A. Sher.2004. MyD88-deficient mice display a profound loss in resistance to Myco-bacterium tuberculosis associated with partially impaired Th1 cytokine and ni-tric oxide synthase 2 expression. Infect. Immun. 72: 2400–2404.

30. Bafica, A., C. A. Scanga, C. G. Feng, C. Leifer, A. Cheever, and A. Sher. 2005.TLR9 regulates Th1 responses and cooperates with TLR2 in mediating optimalresistance to Mycobacterium tuberculosis. J. Exp. Med. 202: 1715–1724.

31. Fremond, C. M., D. Togbe, E. Doz, S. Rose, V. Vasseur, I. Maillet, M. Jacobs,B. Ryffel, and V. F. J. Quesniaux. 2007. IL-1 receptor-mediated signal is anessential component of MyD88-dependent innate response to Mycobacteriumtuberculosis infection. J. Immunol. 179: 1178–1189.

32. Barnden, M. J., J. Allison, W. R. Heath, and F. R. Carbone. 1998. Defective TCRexpression in transgenic mice constructed using cDNA-based alpha- and beta-

chain genes under the control of heterologous regulatory elements. Immunol.Cell Biol. 76: 34–40.

33. Nefedova, Y., M. Fishman, S. Sherman, X. Wang, A. A. Beg, and D. I. Gabrilovich.2007. Mechanism of all-trans retinoic acid effect on tumor-associated myeloid-derived suppressor cells. Cancer Res. 67: 11021–11028.

34. Movahedi, K., M. Guilliams, J. Van den Bossche, R. Van den Bergh, C. Gysemans,A. Beschin, P. De Baetselier, and J. A. Van Ginderachter. 2008. Identification ofdiscrete tumor-induced myeloid-derived suppressor cell subpopulations with dis-tinct T cell-suppressive activity. Blood 111: 4233–4244.

35. Youn, J.-I., S. Nagaraj, M. Collazo, and D. I. Gabrilovich. 2008. Subsets ofmyeloid-derived suppressor cells in tumor-bearing mice. J. Immunol. 181: 5791–5802.

36. MacMicking, J. D., R. J. North, R. LaCourse, J. S. Mudgett, S. K. Shah, andC. F. Nathan. 1997. Identification of nitric oxide synthase as a protective locusagainst tuberculosis. Proc. Natl. Acad. Sci. USA 94: 5243–5248.

37. Schnappinger, D., S. Ehrt, M. I. Voskuil, Y. Liu, J. A. Mangan, I. M. Monahan,G. Dolganov, B. Efron, P. D. Butcher, C. Nathan, and G. K. Schoolnik. 2003.Transcriptional adaptation of Mycobacterium tuberculosis within macrophages:insights into the phagosomal environment. J. Exp. Med. 198: 693–704.

38. Davis, A. S., I. Vergne, S. S. Master, G. B. Kyei, J. Chua, and V. Deretic. 2007.Mechanism of inducible nitric oxide synthase exclusion from mycobacterialphagosomes. PLoS Pathog. 3: e186.

39. El Kasmi, K. C., J. E. Qualls, J. T. Pesce, A. M. Smith, R. W. Thompson,M. Henao-Tamayo, R. J. Basaraba, T. Konig, U. Schleicher, M.-S. Koo, et al.2008. Toll-like receptor-induced arginase 1 in macrophages thwarts effectiveimmunity against intracellular pathogens. Nat. Immunol. 9: 1399–1406.

40. Rotta, G., E. W. Edwards, S. Sangaletti, C. Bennett, S. Ronzoni, M. P. Colombo,R. M. Steinman, G. J. Randolph, and M. Rescigno. 2003. Lipopolysaccharide orwhole bacteria block the conversion of inflammatory monocytes into dendriticcells in vivo. J. Exp. Med. 198: 1253–1263.

41. Doherty, D. E., G. P. Downey, G. S. Worthen, C. Haslett, and P. M. Henson.1988. Monocyte retention and migration in pulmonary inflammation. Re-quirement for neutrophils. Lab. Invest. 59: 200–213.

42. Nicolle, D. M., X. Pichon, A. Bouchot, I. Maillet, F. Erard, S. Akira, B. Ryffel,and V. F. J. Quesniaux. 2004. Chronic pneumonia despite adaptive immune re-sponse to Mycobacterium bovis BCG in MyD88-deficient mice. Lab. Invest. 84:1305–1321.

43. Tu, S., G. Bhagat, G. Cui, S. Takaishi, E. A. Kurt-Jones, B. Rickman, K. S. Betz,M. Penz-Oesterreicher, O. Bjorkdahl, J. G. Fox, and T. C. Wang. 2008. Over-expression of interleukin-1beta induces gastric inflammation and cancer andmobilizes myeloid-derived suppressor cells in mice. Cancer Cell 14: 408–419.

44. Kusmartsev, S., F. Cheng, B. Yu, Y. Nefedova, E. Sotomayor, R. Lush, andD. Gabrilovich. 2003. All-trans-retinoic acid eliminates immature myeloid cellsfrom tumor-bearing mice and improves the effect of vaccination. Cancer Res.63: 4441–4449.

45. Dietlin, T. A., F. M. Hofman, B. T. Lund, W. Gilmore, S. A. Stohlman, andR. C. van der Veen. 2007. Mycobacteria-induced Gr-1+ subsets from distinctmyeloid lineages have opposite effects on T cell expansion. J. Leukoc. Biol. 81:1205–1212.



http://jimm

unol.org/D

ownloaded from


by guest on September 23, 2012 http://jimmunol.org/ Downloaded from



http://jimm

unol.org/D

ownloaded from


by guest on September 23, 2012 http://jimmunol.org/ Downloaded from


mycobacterium bovis bacillus calmette-guérin vaccination ... ·...

Documents