As Myeloid-Derived Suppressor Cells to the Accumulation of Splenocytes That Act Mice Display Aberrant Myelopoiesis Leading

The Journal of Immunology

Mast Cell–deﬁcient KitW-sh ‘‘Sash’’ Mutant Mice Display Aberrant Myelopoiesis Leading to the Accumulation of Splenocytes That Act as Myeloid-Derived Suppressor Cells

Anastasija Michel,* Andrea Schu¨ler,† Pamela Friedrich,* Fatma Do¨ner,* Tobias Bopp,* Markus Radsak,† Markus Hoffmann,* Manfred Relle,‡ Ute Distler,* Jo¨rg Kuharev,* Stefan Tenzer,* Thorsten B. Feyerabend,x Hans-Reimer Rodewald,x Hansjo¨rg Schild,* Edgar Schmitt,* Marc Becker,‡ and Michael Stassen*

Mast cell-deﬁcient KitW-sh ‘‘sash’’ mice are widely used to investigate mast cell functions. However, mutations of c-Kit also affect additional cells of hematopoietic and nonimmune origin. In this study, we demonstrate that KitW-sh causes aberrant extramedullary myelopoiesis characterized by the expansion of immature lineage-negative cells, common myeloid progenitors, and granulocyte/macrophage progenitors in the spleen. A consistent feature shared by these cell types is the reduced expression of c-Kit. Populations expressing intermediate and high levels of Ly6G, a component of the myeloid differentiation Ag Gr-1, are also highly expanded in the spleen of sash mice. These cells are able to suppress T cell responses in vitro and phenotypically and functionally resemble myeloid-derived suppressor cells (MDSC). MDSC typically accumulate in tumor-bearing hosts and are able to dampen immune responses. Consequently, transfer of MDSC from naive sash mice into line 1 alveolar cell carcinoma tumor- bearing wild-type littermates leads to enhanced tumor progression. However, although it can also be observed in sash mice, accelerated growth of transplanted line 1 alveolar cell carcinoma tumors is a mast cell–independent phenomenon. Thus, the KitW-sh mutation broadly affects key steps in myelopoiesis that may have an impact on mast cell research. The Journal of Immunology, 2013, 190: 5534–5544.

he receptor tyrosine kinase c-Kit (CD117) and its ligand rocytic anemia, sterility, pigmentation defects, and intestinal dis- stem cell factor (SCF) have been intensively studied owing orders (7). T to their multifaceted role in development and hemato- The c-Kit/SCF axis is also crucial for the development of mast poiesis (1, 2). These molecules were mapped to the White spotting cells, a finding that led to the widespread use of several strains of c- and Steel loci, respectively, and a variety of mutant alleles have Kit/SCF mutant mice in mast cell research (8). The composite been described (3–6). genotype KitW/KitW-v causes severe mast cell deficiency (9), yet Mice carrying spontaneous loss-of-function mutations at either these animals are also anemic and infertile. More recently, mast Steel or White spotting loci generally show pleiotropic yet similar cell–deficient KitW-sh/KitW-sh ‘‘sash’’ mice serve as a common phenotypes. Pathophysiological manifestations can include mac- tool, as these animals do not suffer from some disadvantages that characterize the KitW/KitW-v strain (10). Although sash mice also lack mast cells, melanocytes, and interstitial cells of Cajal, they *Institute for Immunology, University Medical Center of the Johannes Gutenberg are fertile, not anemic, and show normal numbers of gd T cells University, 55131 Mainz, Germany; †III Medical Department, University Medical ‡ (11–13). Center of the Johannes Gutenberg University, 55131 Mainz, Germany; I Medical W-sh Department, University Medical Center of the Johannes Gutenberg University, The Kit mutation is an inversion of ∼3.1 Mbp encompass- x 55131 Mainz, Germany; and Division for Cellular Immunology, German Cancer ing 27 known genes. However, in contrast to the KitW and KitW-v Research Center, D-69120 Heidelberg, Germany mutations, KitW-sh does not alter the coding region of White spotting Received for publication December 6, 2012. Accepted for publication March 29, locus itself. The 59 breakpoint of this inversion disrupts the Corin 2013. gene, leading to cardiac hypertrophy, and the 39 breakpoint is lo- This work was supported by Deutsche Forschungsgemeinschaft Grants STA 984/4-1 (to M.S. and M. Radsak), STA 984/1-2 (to M.S.), and Ra 988/4-2 (to M. Radsak and cated72kbpupstreamofthec-Kit transcriptional start site (14, H.S.) the European Research Council (Advanced Grant 233074, to H.-R.R.), and by 15). Regulatory elements driving the expression of c-Kit in mast the Immunology Center of Excellence Mainz (to M.S.). cells were mapped within the affected region (16). Thus, it can be Address correspondence and reprint requests to Dr. Michael Stassen, Institute for assumed that the inversion or additional deletions of cis regulatory Immunology, University Medical Center of the Johannes Gutenberg University, Langenbeckstrasse 1, 55131 Mainz, Germany. E-mail address: [email protected] elements prevent mast cell–specific c-Kit expression, which ulti- The online version of this article contains supplemental material. mately leads to their irreversible demise within a few weeks after birth of sash mice (17, 18). Abbreviations used in this article: CMP, common myeloid progenitor; G-MDSC, granulocytic myeloid-derived suppressor cell; HSC, hematopoietic stem cell; L1C2, The use of Kit mutant mice has been of central importance to G line 1 alveolar cell carcinoma; Lin, lineage; L-NMMA, N -monomethyl-L-arginine; unravel mast cell functions in the immune system and beyond (19, LSK, Lin2Sca-1+c-Kit+; LT-HSC, long-term hematopoietic stem cell; MDSC, myeloid-derived suppressor cell; MEP, megakaryotic/erythroid progenitor; M-MDSC, 20). However, novel mast cell–deficient strains with unperturbed monocytic myeloid-derived suppressor cell; MPP, multipotent progenitor; MS, mass c-Kit/SCF function challenge some findings obtained in Kit mu- v spectrometry; nor-NOHA, N -hydroxy-nor-L-arginine; ROS, reactive oxygen spe- tant mice (21, 22). These inconsistent observations may indicate cies; SCF, stem cell factor; ST-HSC, short-term hematopoietic stem cell. limitations of mast cell–dependent mouse strains with mutant Kit Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 alleles (23–25). www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203355 The Journal of Immunology 5535

It this study, we demonstrate that the KitW-sh defect causes ex- and cells were stained for CD11b, Ly6G, Ly6C, or only with propidium io- tramedullary hematopoiesis leading to the accumulation of mye- dide and phenotype was analyzed via flow cytometry. + + loid progenitor cells in the spleen of naive sash mice. CD11b Ly6C Histology cells that express intermediate and high levels of Ly6G are also highly expanded in sash spleen. Morphologically and functionally, Cytospin preparations were stained with a microscopy Hemacolor set (Merck) according to the manufacturer’s recommendations. Slides were these Ly6G-expressing cells resemble granulocytic myeloid-derived analyzed by bright-field microscopy on a Keyence BZ-8000 fluorescence suppressor cells (G-MDSC), which typically expand in tumor- microscope. bearing hosts. Thus, besides its well-documented roles on mast To assess mast cell numbers, ears were removed, fixed in Roti-Histofix cell development and survival, the KitW-sh allele affects myelopoi- (Roth), and embedded in paraffin. Sections were deparaffinized, rehydrated, and stained with avidin-Alexa Fluor 488 (Invitrogen). Slides were analyzed esis, which may have adverse effects regarding the interpretation of in GFP channel on a Keyence BZ-8000 fluorescence microscope. mast cell–dependent phenomena. Bone marrow chimeras Materials and Methods BALB/c mice were lethally irradiated with 8.5 Gy from a [137Cs] source Mice (OB58-BA; Buchler Braunschweig). Next day, 5 3 106 bone marrow donor cells from C.B6-KitW-sh or from BALB/c mice were transferred by W-sh d C.B6-Kit mice (H-2 ) were generated as previously described and i.v. injection. Then, mice were housed under specific pathogen-free con- backcrossed at least 12 generations (26). BALB/c wild-type littermates ditions for a time period of 8 wk before use. were obtained from crosses between heterozygous C.B6-KitW-sh/+ mice. Cpa3Cre/+ mice (22) on a BALB/c background and their wild-type litter- Genotyping mates were obtained from Thorsten Feyerabend (German Cancer Research Center). Cells sorted by flow cytometry cells were genotyped according to a pub- All mice were used in accordance with the guidelines of the Central lished procedure (15). Animal Facility of the University of Mainz. Generation of BMDC Abs and reagents Single bone marrow cell suspensions were cultured in IMDM supple- CD11b-FITC (M1/70), Ly6G-PE (1A8), CD11b-biotin (M1/70), Ly6C- mented with 5% FCS, 1 mM L-glutamine, 1 mM NaPyr, and 50 ng/ml GM- CSF. Medium was changed on days 2 and 4. BMDC were used at day 6 and PerCP-Cy5.5 (HK1.4), CD135-PE (A2F10), CD117-PE-Cy7 (ACK2), + CD16/32 (93) , CD34-FITC (RAM34), and CD150-FITC (9D1) were ob- were at least 85% CD11c . tained from eBioscience. Ly6G-FITC (1A8), CD117-allophycocyanin Allogeneic MLR (ACK2), Sca-1-allophycocyanin (D7), CD8a-biotin (53-6.7), CD5-biotin (53-7.3), CD45R (B220)-biotin (RA3-6B2), CD127-biotin (A7R34), TER- To explore the suppressive capacity of CD11b+ subpopulations on T cells, 119-biotin, Gr-1-biotin (RB6-8C5), and streptavidin-PerCP were obtained sorted cells were given to lymphocytes in a ratio of 1:1, 1:3, or 1:9. from BioLegend. CD16/32-PE (2.4G2) and CD3ε-biotin (500A2) were In general, the allogeneic MLR consisted of 5 3 104 lymphocytes from obtained from BD Biosciences. Propidium iodide was from Sigma- BALB/c mice as responders and 1.67 3 103 BMDC from C57BL/6 mice Aldrich, and CD4-biotin (H129.19) was by our own production and bio- as activators. The ratio of BMDC to lymphocytes (1:30) was constant in all tinylation. experiments. On day 4, [3H]thymidine (0.5 mCi/well) was added and the incorporation was determined in a liquid scintillation counter (LKB Flow cytometry analyses and cell sorting Wallac) 20 h later. The role of reactive oxygen species (ROS), inducible NO synthase, or For staining of lineage (Lin) markers, single-cell suspensions from spleen arginase in the suppression of T cell proliferation was examined by addition or bone marrow were treated with a mixture of biotinylated Lin marker– of 1000 U/ml catalase (Sigma-Aldrich), 500 mM L-NMMA (NG-mono- ε v specific Abs (to CD3 , CD4, CD5, CD8a, CD11b, CD45R, CD127, Gr-1, methyl-L-arginine; Calbiochem), or 500 mM nor-NOHA (N -hydroxy-nor- TER-119) and then with streptavidin-PerCP. L-arginine; Calbiochem), respectively. Inhibitors were added to the cocul- For the detection of long-term hematopoietic stem cells (LT-HSC), short- tures on day 4; 8 h later [3H]thymidine was added and the incorporation was term HSC (ST-HSC), and multipotent progenitors (MPP) (27, 28), spleno- determined 20 h later. cytes or bone marrow cells were incubated with biotinylated Lin marker– specific Abs (except anti-CD127), then with streptavidin-PerCP, and finally Line 1 alveolar cell carcinoma–induced tumors with fluorescently labeled Abs to CD117, Sca-1, CD150, and CD135. 5 For the detection of common myeloid progenitors (CMP), megakaryotic/ Mice were injected with 10 line 1 alveolar cell carcinoma (L1C2) cells d erythroid progenitors (MEP), and granulo-myelomonocytic progenitors (murine bronchoalveolar carcinoma cell line; H-2 ) s.c., two spots in each (GMP) (29), splenocytes or bone marrow cells were incubated with bio- flank. tinylated Lin marker–specific Abs (except anti-CD11b), then with streptavidin- The tumor size was determined 2–3 wk later as average of four tumors per PerCP, and finally with fluorescently labeled Abs to CD117, Sca-1, CD16/32, mouse by cross-sectional measurement and multiplication of the two values and CD34. (square millimeters). Bone marrow chimeras showed moderate progress of For analyses of MDSC–like cells, Fc receptors on splenocytes were tumor development, and therefore their tumor size was determined 4 wk blocked with anti-CD16/32 and stained for CD11b, Ly6G, and Ly6C. after the L1C2 injection. For cell sorting, CD4+, CD8+, and B220+ splenocytes were depleted with DynaBeads (Invitrogen), and then CD11b+ splenocytes were enriched Enrichment of G-MDSC–like cells with CD11b MACS (Miltenyi Biotec). Fc receptors were blocked using Single-cell suspensions of spleen from naive C.B6-KitW-sh mice were pre- anti-CD16/32 and cells were finally stained with CD11b, Ly6G, and Ly6C. pared, and T and B cells were depleted with DynaBeads (CD4, CD8 and Analyses were performed using a FACSCanto or LSR II flow cytometer B220; Invitrogen). Ly6G+ cells were enriched using the FcR blocking re- and FACSDiva software (BD Biosciences). Cell sorting was performed agent, anti–Ly6G-biotin, and anti-biotin microbeads (MACS MDSC isola- on a FACSAria II with FACSDiva software. tion kit; Miltenyi Biotec). The purity of these cells was typically ∼90% Ly6G+ as determined by flow cytometry. These cells (5 3 106) were injected Colony forming and differentiation assays i.v. in the tail vein or 2.5 3 106 s.c. between the four tumors on each flank. ColonyGEL 1202 mouse complete medium (Cell Systems) was used ac- Proteolytic digestion for mass spectrometry cording to the manufacturer’s instructions. In brief, 105 splenocytes were incubated for 7 d at 37°C in duplicates. Colonies consisting of at least 50 Cells were lysed by the addition of lysis buffer (7 M urea, 2 M thiourea, 5 cells were counted. mM DTT, 2% CHAPS), followed by sonication. After cell lysis, protein For differentiation assays, ColonyGEL 1202 mouse complete medium or amounts were determined using the Pierce 660 nm protein assay (Thermo ColonyGEL 1201 mouse base medium (Cell Systems) were used. Sorted Scientific, Rockford, IL). Proteins (20 mg/sample) were digested using CD11b+Ly6GintLy6C+ cells (105) were cultured in duplicates either in a modified filter-aided sample preparation method (30) using trypsin gold complete medium or in base medium supplemented with 20 ng/ml M-CSF (Promega). Resulting tryptic digest solutions were diluted with aqueous (PeproTech). On day 7, cells were washed in PBS, Fc receptors were blocked, 0.1% (v/v) formic acid to a concentration of 200 ng/ml and spiked with 20 5536 Myeloid-derived suppressor cells in KitW-sh mutant mice fmol/ml enolase 1 (Saccharomyces cerevisiae) tryptic digest standard replicate, and score filtering. Only proteins with a minimum Global (Waters, Manchester, U.K.). SERVER identification score of 250.0 and quantified by two or more uniquely assigned peptides, with each detected in at least two liquid chromatography/ Mass spectrometric analyses MS runs with a minimum Global SERVER identification score of 5.0 and minimum length of 6 aa, were considered. Nanoscale LC separation of tryptic peptides was performed with a nano- Acquity system (Waters) equipped with an HSS-T3 C18 1.8 mm, 75 mm 3 Statistical analyses 250 mm analytical reversed-phase column (Waters) in direct injection mode as described (31). Statistical differences were determined using the Student t test. A p Mass spectrometric analysis of tryptic peptides was performed using a value ,0.05 was considered to be statistically significant. Values for all Synapt G2-S mass spectrometer (Waters) operated in positive mode elec- groups are expressed as means 6 SD. trospray ionization with a typical resolution of at least 25,000 full width at half maximum using data-independent modes of analysis (32, 33) in combination Results with on-line ion mobility separations (34). The data were postacquisition The KitW-sh mutation causes extramedullary hematopoiesis in lockmass corrected as described (31). In low-energy liquid chromatography/ mass spectrometry (MS) mode, data were collected at constant collision en- the spleen ergy of 4 eV. In elevated energy MS mode, the collision energy was ramped Myeloid hyperplasia in the spleen of sash mice due to expansion of from 25 to 55 eV. One cycle of low and elevated energy data was acquired + every 1.5 s. All samples were analyzed in four replicates. Gr-1 cells was previously reported (15, 35) and was correlated with neutrophilia in spleen, bone marrow, and blood in one of Data processing and protein identification these studies (15). Following the identification of a traceable polymorphism linked to Continuum liquid chromatography/MS data were processed and searched W-sh using ProteinLynx Global SERVER version 2.5.2 (Waters) using a custom- the Kit allele, we generated congenic mast cell–deficient mice on compiled database containing UniProt reference proteome mouse entries a BALB/c background, referred to as C.B6-KitW-sh (26). Our com- supplemented with sequence information of enolase-1 (S. cerevisiae), bo- parative analyses of C57BL/6-KitW-sh and C.B6-KitW-sh confirmed vine trypsin, and human keratins. The initial false-positive rate for protein 3 the increase of CD11b/Gr-1 double-positive cells in the spleen of identification was set to 4% based on the search of a 3 randomized data- W-sh base (31). The experimental data were typically searched with a 3 ppm both strains, which was even more pronounced in C.B6-Kit mice. precursor and 10 ppm product ion tolerance with one missed cleavage However, on either genetic background, the numbers of these bona allowed and fixed carbamidomethyl cysteine and variable methionine oxi- fide neutrophils in bone marrow and blood were unaffected by the dation set as the modifications. Postidentification analyses, including re- KitW-sh mutation (26). This prompted us to investigate the impact of tention time alignment, exact mass retention time clustering, normalization, W-sh and label-free quantification (32), were done using an in-house–developed the Kit allele on peripheral myelopoiesis in detail. software. The false-positive rate of protein identification was reduced to In 10-wk-old sash mice, spleens were enlarged and, on average, ,0.3% by additional data processing steps including isoform/homology, the spleen weight was .3-fold increased compared with wild-type

FIGURE 1. Mast cell–deficient sash mice display splenomegaly and extramedullary hematopoiesis. (A) Spleen weights of 10-wk-old BALB/c and C.B6-KitW-sh mice (n = 10-11). (B) Splenocytes (105 each, in duplicates) derived from BALB/c (n = 3) and C.B6- KitW-sh (n = 2) mice were cultured in methylcellulose- based medium for colony forming cell assays with SCF, IL-3, IL-6, and erythropoietin. At day 7, colonies consisting of at least 50 cells were counted. (C) Flow cytometric analyses (left) of splenocytes and bone marrow cell suspensions stained with Abs specific for lineage markers to exclude lineage-committed cells (CD3ε, CD4, CD5, CD8a, CD11b, CD45R, CD127, Gr-1, TER119). Percentages of Lin2 cells are depicted (right; n = 4–6). The Journal of Immunology 5537 littermates (Fig. 1A). Colony-formation assays revealed a strong population that encompasses HSC. Compared with C57BL/6 mice increase in CFUs, indicative of extramedullary hematopoiesis (Fig. (Supplemental Fig. 1A), the expression of Sca-1 is reduced but not 1B). To dissect this phenomenon, we determined the numbers of absent on LSK cells from BALB/c mice. immature Lin2 cells that do not express cell-surface markers HSC can be divided into LT-HSC and ST-HSC (Fig. 2B). MPP present on Lin-committed cells in spleen and bone marrow derived reflect the branch point to both common lymphoid progenitors and from sash mice and wild-type littermates. As depicted in Fig. 1C, CMP, with the latter being able to yield MEP and GMP. GMP frequencies of Lin2 cells in bone marrow of both strains are finally differentiate into monocytes and granulocytes (36, 37). comparable, yet there is a relative expansion of these cells in the Flow cytometric analyses revealed that in the spleen of sash mice, spleen of sash mice. As shown in Fig. 2A, Lin2 cells contain both frequencies of LT-HSC, ST-HSC, MPP, CMP, and GMP are in- c-Kit+Sca-12 myeloid progenitors and the Lin2Sca-1+c-Kit+ (LSK) creased (Fig. 2C). In contrast, numbers of MEP are strongly de-

FIGURE 2. Sash mice develop abberant myelopoiesis characterized by the expansion of MPP, CMP, and GMP in the spleen. (A) Analytical setup. Lin2 splenocytes were stained for c-Kit and Sca-1 to identify myeloid progenitors (Lin2Sca-12c-Kit+) and LSK cells. Myeloid progenitors can be subdivided into MEP, GMP, and CMP, whereas LSK cells contain LT-HSC, ST-HSC, and MPP. (B) Simplified scheme of hierarchy in myelopoietic progenitor development. (C) Lin2 splenocytes were stained according to (A) and percentages of distinct progenitor populations are plotted (n = 4–6). (D) Progenitor populations shown in (C) plotted according to their geometric mean fluorescence intensity (gMFI) of c-Kit expression. 5538 Myeloid-derived suppressor cells in KitW-sh mutant mice creased. This is most likely due to the preferred development of These findings imply that the intrinsic sash defect in bone CMP to GMP. Supplemental Fig. 1B depicts equivalent data from marrow cells is both necessary and sufficient to drive aberrant C57BL/6-KitW-sh mice and their wild-type littermates. Regarding myelopoiesis in mice with otherwise physiological expression of the expression levels of c-Kit, both populations of HSC and MPP c-Kit. in sash mice are phenotypically inconspicuous (Fig. 2D). How- + + + ever, CMP, GMP, and MEP from the spleen of these animals show CD11b Ly6G Ly6C splenocytes from sash mice resemble reduced expression of c-Kit, indicating deregulation of c-Kit ex- MDSC pression during myelopoiesis. Morphologically, populations gated in P5, P6, and P7 display distinct characteristic features (Fig. 4A). Cells in P7 appear In naive mice, the KitW-sh mutation leads to the intrinsic + + + monocyte blast-like, whereas in P5 a polymorphnuclear phenotype development and accumulation of CD11b Ly6G Ly6C prevails. However, cytospin preparations of splenocytes gated in splenocytes P6 reveal the abundance of cells with ring-shaped nuclei, which Extramedullary myelopoiesis likely explains the abundance of Gr- were reported to encompass MDSC, although this morphological 1+CD11b+ cells in the spleen of sash mice, as introduced above criterion is not unique to these cells (39, 40). and shown in Fig. 3A. The term MDSC is an operational definition and refers to a het- Gr-1 is a myeloid differentiation marker for granulocytes and erogeneous myeloid population that basically also includes neu- belongs to the Ly6 family (38). Staining of CD11b+ splenocytes for trophils and precursors for monocytes and dendritic cells. Phys- Ly6C and Ly6G reveals distinct populations (Fig. 3A). Strikingly, iologically, these cells accumulate under conditions of systemic populations characterized as Ly6GhighLy6C+ and Ly6GintLy6C+ are insults such as tumors and sepsis and are able to dampen immune enlarged in CD11b+ splenocyte suspensions derived from sash mice responses by inhibiting T cell activation (41, 42). Using flow (Fig. 3A, 3B). The largest difference can be found for Ly6GintLy6C+ cytometry, we isolated Ly6GhighLy6C+ (gate P5) and Ly6GintLy6C+ cells, which are virtually absent in wild-type mice (Fig. 3B and gate (gate P6) cells and demonstrated their ability to potently inhibit P6 in the flow cytometry plot of Ly6C versus Ly6G in Fig. 3A). allogeneic MLRs in vitro (Fig. 4B). According to the expression Supplemental Fig. 1C shows equivalent data from C57BL/6-KitW-sh of CD11b, Ly6G, and Ly6C, murine MDSC have been subdivided mice and their wild-type littermates. Within the bone marrow, all into two subpopulations with either typical monocytic (M-MDSC) these populations gated in P5, P6, and P7 can also be clearly iden- or granulocytic (G-MDSC) morphology (40, 43). M-MDSC are tified within the CD11b+ population. However, no significant dif- CD11b+Ly6ChighLy6G2 and produce NO and arginase to inhibit ferences regarding the frequencies of these populations in bone T cell responses, whereas G-MDSC display the phenotype CD11b+ marrow from wild-type and sash mice are detectable (Fig. 3A). Ly6ClowLy6G+ and release ROS (44, 45). Thus, we next evaluated This prompted us to investigate whether the accumulation of the effects of inhibitors of NO synthase (L-NMMA), arginase (nor- CD11b+Ly6GintLy6C+ cells in the spleen of sash mice (P6 in Fig. NOHA), and ROS (catalase) on their suppressive activity in allo- 3A, 3B) originates from bone marrow emigration. To this end, we geneic MLRs. The presence of either L-NMMA or nor-NOHA only performed quadrupole time-of-flight mass spectrometry following marginally affected the suppressive activities of Ly6GhighLy6C+ ultraperformance liquid chromatography separation of tryptic (gate P5) and Ly6GintLy6C+ (gate P6) cells, whereas catalase sig- peptides derived from CD11b+Ly6GintLy6C+ cells previously nificantly restored proliferation, indicating that production of ROS isolated in high purity by FACS from bone marrow and spleen of is of vital importance for suppression to occur (Fig. 4C). sash mice (Fig. 3C). Interestingly, only 62% of the identified pro- Taken together, based on the expression pattern of Ly6C and teins are expressed at comparable levels, whereas the remaining Ly6G, the appearance of ring-shaped nuclei, and their ability to proteins are either significantly up- or downregulated. This finding suppress T cell proliferation, the expanded myeloid cell popu- indicates major differences between CD11b+Ly6GintLy6C+ cells lations in the spleen of naive sash mice resemble G-MDSC. in bone marrow and spleen from sash mice. This technique also The nature of G-MDSC, their ability to differentiate, and their allowed us to estimate the influence of the sash mutation on relationship to neutrophils are still obscure (45). To investigate protein expression in CD11b+Ly6GintLy6C+ cells derived from the potential of the G-MDSC–like populations Ly6GhighLy6C+ either wild-type or sash bone marrow (Fig. 3C). Our results imply (gate P5) and Ly6GintLy6C+ (gate P6) to differentiate, we used a drastic influence of the sash mutation, as only ∼70% of the methylcellulose-based media to follow their fate in the presence identified proteins are equally expressed in cells derived from of growth factors in vitro. A combination of IL-3, IL-6, SCF, and both strains. erythropoietin causes a shift to higher expression of Ly6G; the To investigate whether the expansion of the above-described phenotype almost completely changes from Ly6Gint to Ly6Ghigh cell populations P5 and P6 in sash mice is due to an intrinsic defect, (Fig. 5A). Essentially the same happens with P6 cells isolated we generated chimeras by transferring bone marrow from sash mice from wild-type spleen. However, in the latter case it is obvious into irradiated wild-type recipients. Shown in Fig. 3D is the strong that a significant proportion of cells reduces expression of CD11b expansion of Ly6GhighLy6C+ (gate P5) and Ly6GintLy6C+ (gate but remains negative for propidium iodide (Fig. 5A). We further P6) cells from sash origin in the spleen of wild-type mice. In investigated whether the presence of M-CSF enhances the ex- contrast, the cells designated as population P7 (Ly6ChighLy6G2) pression of Ly6C, which might indicate a shift from a G-MDSC– develop in comparable numbers from wild-type and sash donors to a M-MDSC–like phenotype. However, even in the presence and thus were excluded from further analyses. The descent of all of M-CSF, only a significant trend toward higher Ly6G expres- these populations from donor sash bone marrow in wild-type sion is observable for P6 cells derived from C.B6-KitW-sh and recipients was verified by genotyping cells isolated by FACS. BALB/c mice (Fig. 5B). Note that under these experimental in The wild-type c-Kit allele is absent in the cell populations gated in vitro conditions, G-MDSC–like cells appeared rather long-lived P5–7 in mice that had been rescued with sash bone marrow fol- and postmitotic; on day 7, at least 84% of the initially seeded cells lowing irradiation, as only the sash mutation is detectable by PCR are negative in propidium iodide stainings (Fig. 5A, 5B). Addi- (Fig. 3E). Furthermore, in wild-type mice, the presence of radio- tionally, we also followed the fate of these G-MDSC–like pop- resistant mast cells does not influence the expansion of myeloid ulations P5 and P6 in allogeneic MLRs on day 4. As depicted in cells from sash bone marrow (Fig. 3F). Fig. 5C, the vast majority of Ly6GhighLy6C+ cells maintain a sta- The Journal of Immunology 5539

FIGURE 3. Spleen of naive sash mice display elevated numbers of Ly6G+ cells. (A) CD11b+ splenocytes and bone marrow cells from BALB/c and C.B6- KitW-sh mice were analyzed by flow cytometry for the expression of Ly6G and Ly6C. Representative histograms of CD11b expression and dot plots of Ly6G/Ly6C expression in both strains are shown. (B) Total percentages of CD11b+ splenocytes, CD11b+Ly6GhighLy6C+ (P5), CD11b+Ly6GintLy6C+ (P6), and CD11b+Ly6GlowLy6Chigh (P7) cells are indicated; n =5.(C) Comparative protein expression pattern of CD11b+Ly6GintLy6C+ (P6) cells isolated by FACS (purity $ 98%, not shown) from bone marrow (BM) and spleen (SP) of the indicated mouse strains. Tryptic digests where separated by ultraperformance liquid chromatography and analyzed by quadrupole time-of-flight mass spectrometry. Identified peptides (n = 1062) were used to compare BALB/c bone marrow–derived CD11b+Ly6GintLy6C+ cells with the corresponding population isolated from C.B6-KitW-sh bone marrow. For comparing CD11b+Ly6GintLy6C+ cells isolated from C.B6-KitW-sh spleen and bone marrow, 1362 peptides were used. Each measurement was performed in quad- ruplicates. Representatives of two equivalent biological sample sets are shown. (D) Bone marrow chimeras were generated by irradiation of BALB/c mice on day 0 and reconstitution with bone marrow cells from congenic C.B6-KitW-sh mice on day 1. Controls received BALB/c bone marrow. Eight weeks later, CD11b+ splenocytes were analyzed and percentages of CD11b+Ly6GintLy6C+ cells (P6) and CD11b+Ly6GhighLy6C+ cells (P5) are shown (n = 7). (E) Genotyping of sorted CD11b+ splenocytes derived from chimeric mice was performed using primers specific for c-kit wild-type and sash alleles. (F) Ear skin sections from bone marrow chimeras were analyzed for the presence of mast cells with fluorescence-conjugated avidin. Some mast cells are indicated by arrows. 5540 Myeloid-derived suppressor cells in KitW-sh mutant mice

compared BALB/c and congenic C.B6-KitW-sh mice in their ability to develop tumors following s.c. inoculation of L1C2. As shown in Fig. 6A, BALB/c mice develop visible and distinct tumors until day 18 following inoculation. However, tumor sizes in C.B6-KitW-sh mice are significantly larger. Subcutaneous tumor development in BALB/c mice is accompanied by slight increases of spleen weights in the absence of overt metastases. Spleens from naive sash mice are variably enlarged but there is no additional enlargement in tumor-bearing mice (Fig. 6B). Splenomegaly in tumor-bearing BALB/c mice is paralleled by the accumulation of CD11b+ cells (Fig. 6C), which are either Ly6GhighLy6C+ (P5) or Ly6GintLy6C+ (P6). This observation is in accordance with the previously reported accumulation of MDSC in spleen of tumor- bearing mice. However, depending on the tumor model, MDSC numbers are highly variable (44). However, there is no further accumulation of cells with MDSC-like phenotype in C.B6-KitW-sh mice (Fig. 6C). To investigate whether increased tumor burden in sash mice is due to the absence of mast cells, we generated chimeras following ablation of bone marrow by irradiation. Displayed in Fig. 6D are tumor sizes of irradiated BALB/c mice that either received wild- type or C.B6-KitW-sh bone marrow before inoculation with L1C2. Obviously, the ability to develop larger tumors can be transferred by sash bone marrow, despite the presence of radio-resistant mast cells in wild-type mice (see also Fig. 3F). Additionally, we used mast cell–deficient Cpa3Cre/+ mice on BALB/c background. In these animals, mast cell deficiency is due to genotoxic effects of Cre expressed under the control of car- boxypeptidase A3 promoter, whereas the c-Kit/SCF axis is unimpaired (22). According to the expectations, tumor sizes in Cpa3Cre/+ mice and their wild-type littermates are comparable (Fig. 6E). Thus, the enhanced growth of L1C2 tumors in sash mice cannot be ascribed to the absence of mast cells. Adoptive transfer of G-MDSC–like splenocytes from sash mice to wild-type recipients enhances L1C2 tumor growth One inherent problem of the operational MDSC definition is the impossibility to selectively ablate these cells without the risk of FIGURE 4. CD11b+Ly6GintLy6C+ and CD11b+Ly6GhighLy6C+ spleno- affecting additional cell populations (47, 48). cytes from sash mice resemble MDSC. (A) HE-stained cytospin preparations of the indicated cell populations sorted by flow cytometry. (B) Furthermore, it is impossible to predict whether G-MDSC–like Allogeneic MLR using 5 3 104 lymph node cells from BALB/c mice as cells from sash mutant mice are in every aspect identical to their responders and 1.33 3 103 bone marrow–derived dendritic cells from wild-type counterparts. Deregulation of c-Kit expression (Fig. 2D) C57BL/6 mice as activators (fixed ratio responders/activators, 30:1). Sor- might significantly alter signaling events during the development ted P5 or P6 cells were added in different ratios on day 0 (1:1 [5 3 104], of MDSC. This assumption is also supported by our mass spec- 1:3 [1.67 3 104], and 1:9 [0.55 3 104]). On day 4, [3H]thymidine was trometry analyses shown in Fig. 3C. added and the incorporation measured 20 h later. Percentages of inhibition Thus, as an equivalent alternative approach to depletion, we C of proliferation compared with controls without sorted cells are shown. ( ) assessed tumor growth in BALB/c mice following adoptive transfer Experiment was performed as described above with the modification that of G-MDSC–like cells from sash donors. To this end, we applied on day 4 either catalase, L-NMMA, or nor-NOHA was added. Shown are MACS technology to isolate Ly6G+ cells representing the Ly6Gint representatives of two experiments each performed in triplicates (B, C). and Ly6Ghigh populations, which we collectively and operation- ally termed G-MDSC–like cells (Fig. 7A). As depicted in Fig. 7B, ble phenotype, whereas Ly6GintLy6C+ cells almost completely transfer of G-MDSC–like cells in BALB/c mice inoculated with shift toward higher expression of Ly6G, irrespective of whether L1C2 cells significantly enhanced tumor growth, irrespective of they are derived from C.B6-KitW-sh or BALB/c mice. whether the cells were given i.v. or s.c. This experiment un- Collectively, these results indicate that both populations of G- equivocally supports our notion that aberrant myelopoiesis in sash MDSC–like cells isolated from the spleen of sash mice have a mice leads to the development of splenocytes with immunosup- limited developmental capacity and mainly differentiate toward a pressive function, closely resembling MDSC. Ly6Ghigh phenotype. Discussion Enhanced growth of transplanted L1C2 tumor cells in W-sh Notwithstanding their application in mast cell research, it has to C.B6-Kit mice is not due to the absence of mast cells be considered that c-Kit mutant strains suffer from additional L1C2 is a cell line derived from a tumor that arose spontaneously defects that may even falsify the results of experiments supposedly in a female BALB/c mouse (46). In a first set of experiments, we addressing the role of mast cells. The Journal of Immunology 5541

FIGURE 5. In vitro, CD11b+Ly6GintLy6C+ cells develop a Ly6Ghigh phenotype. (A) CD11b+Ly6Gint Ly6C+ cells (gate P6) were sorted via flow cytometry from the spleen of wild-type and sash mice and 105 cells were cultured in methylcellulose-based medium in the presence of growth factors as indicated. On day 7, cells were analyzed by flow cytometry for the expression of CD11b. Furthermore, CD11b+ cells were stained for Ly6G and Ly6C. Viability was determined by propidium iodide (PI) staining. (B) Experiment was performed as described above but cells were cultured in the presence of M-CSF. (C) CD11b+Ly6GhighLy6C+ (gate P5)orCD11b+Ly6GintLy6C+ (gate P6) cells were isolated using flow cytometry from sash or wild-type spleen and added to an allogeneic MLR in a ratio of 1:1. On day 4 of coculture, CD11b+ cells were analyzed for the expression of Ly6C and Ly6G and for viability. Shown are representatives from two independent experiments each performed in triplicates.

KitW-sh is a mutation known to block c-Kit expression in some cell In this context, we demonstrate that sash mutant mice develop types and to enhance the expression of c-Kit in others. Importantly, extramedullary myelopoiesis characterized by the accumulation of in sash mice c-Kit expression is shut off in mast cells, causing mast HSC, MPP, CMP, and GMP in the spleen. In contrast, frequencies of cell deficiency, whereas the absence of melanocytes might be due to MEP are decreased, yet this might be a consequence of higher GMP enhanced c-Kit expression at sites of early melanogenesis (10, 11, numbers, as both cells derive from the same precursor. Interestingly, 14, 17, 49). Besides multiple functions in development, the c-Kit/ the expression of c-Kit, measured by flow cytometry, is unimpaired SCF system has essential roles in hematopoiesis during fetal de- in LT-HSC, ST-HSC, and MPP, but decreased in CMP, GMP, and velopment and in the adult (1, 50). MEP derived from sash spleen. These findings demonstrate that 5542 Myeloid-derived suppressor cells in KitW-sh mutant mice

FIGURE 7. Following transfer of G-MDSC–like cells from sash mice into wild-type recipients, L1C2 tumor growth is enhanced. (A) Splenocytes from naive C.B6-KitW-sh mice were used for positive selection of Ly6G+ cells as described in the experimental procedures section. Purity of ∼90% was determined by ﬂow cytometry. (B) BALB/c mice were inoculated with L1C2 cells on day 0. On days 1, 4, 7, and 10 Ly6G+ splenocytes were isolated using MACS and injected either i.v. into the tail vein (5 3 106)or s.c. 2.5 3 106 cells into each ﬂank. On day 14, tumor sizes were determined; n = 5 mice/group.

Physiologically, MDSC accumulate in lymphoid organs under chronic inflammatory conditions or in tumor-bearing hosts. In the latter, expansion of MDSC is variable and strongly depends on the FIGURE 6. Enhanced growth of transplanted L1C2 tumor cells in C.B6- tumor model investigated (44). Tumor-derived SCF was reported W-sh Kit mice is not due to the absence of mast cells. (A) BALB/c and C.B6- to promote the expansion of MDSC in BALB/c mice. This indicates W-sh Kit mice were inoculated s.c. with L1C2 cells on two spots on each that SCF/c-Kit signals may be critical in the generation of these flank using 105 cells per spot. On day 18, the cross-sectional size (mm2)of cells in mice with uncompromised c-Kit expression (52). These each tumor was determined and expressed as mean size per mouse. Arrows point to representative tumors; n =5.(B) On day 18, the spleen weights of results apparently contradict our own observations of expanded the animals shown in (A) were determined. (C) CD11b+ splenocytes were MDSC-like cells in naive sash mice in which c-Kit expression is analyzed for the expression of Ly6G and Ly6C by flow cytometry. Per- reduced in MDSC precursors, CMP, and GMP. However, blockade W-sh centages of P5 and P6 populations of L1C2 tumor-bearing or untreated of SCF production and the effects of the Kit mutation may mice on day 18 are depicted. (D) Chimeras were generated by irradiation have different impacts on c-Kit signaling intensities and cell fate. of BALB/c mice on day 0 and reconstitution with bone marrow cells from Currently, it is a paradigm that MDSC can exert their suppres- congenic C.B6-KitW-sh mice on day 1. Controls received BALB/c bone sive activity on T cell function upon exposure to inflammatory marrow. Eight weeks later, mice were inoculated with L1C2 cells and mediators. Factors that induce activation of MDSC include IFN-g, E Cre/+ tumor sizes were determined on day 28; n =6.( ) BALB/c and Cpa3 IL-3, IL-6, and TGF-b (41). Despite the fact that both tumor- mice were treated as described in (A); n =8. promoting (53) and antitumor activities (54) of mast cells were reported, our results allow us to conclude that the growth of L1C2 the sash mutation broadly affects the expression of c-Kit in precursor tumor cells is unimpaired by the presence or absence of mast cells. cells of the myeloid lineage. Consequently, deregulation of c-Kit However, the presence of large numbers of G-MDSC–like cells expression causes the accumulation of CD11b+Ly6Gint–highLy6C+ in naive sash mice might allow their instant activation following cells in naive mice. These cells phenotypically and functionally tumor inoculation leading to the suppression of specific T cell closely resemble G-MDSC. responses, which eventually promotes tumor growth. So far, we can only speculate on the impact of the sash mutation Mounting evidence suggests that the tumor microenvironment is regarding the generation of these G-MDSC–like cells. The c-Kit/ capable of expanding and activating MDSC by delivering a host of SCF axis was shown to regulate adhesion of HSC and to confer immune mediators. Alternatively, activated T cells are regarded as stable interactions between HSC and stroma cells in fetal liver and source for mediators able to activate MDSC. However, activation in bone marrow (51). Thus, reduced expression of c-Kit on mye- of MDSC might be critical for tumor progression, as it dampens loid progenitor cells might promote their exit from the bone immune responses against the tumor (55). marrow and accumulation in the spleen, where they, at least partly, Collectively, our data demonstrate that the KitW-sh mutation has further develop into G-MDSC–like cells. a deep impact on myelopoiesis and is thus not as mast cell–spe- The Journal of Immunology 5543 cific as it was previously thought. This underpins the need for 24. Brown, M. A., J. K. Hatfield, M. E. Walker, and B. A. Sayed. 2012. A game of kit and mouse: the Kit is still in the bag. Immunity 36: 891–892, author reply additional mast cell–specific and Kit-independent murine models 893–894. in mast cell research. 25. Rodewald, H.-R., and T. B. Feyerabend. 2012. Widespread immunological functions of mast cells: fact or fiction? Immunity 37: 13–24. 26. Becker, M., S. Reuter, P. Friedrich, F. Doener, A. Michel, T. Bopp, M. Klein, Disclosures E. Schmitt, H. Schild, M. P. Radsak, et al. 2011. Genetic variation determines The authors have no financial conflicts of interest. mast cell functions in experimental asthma. J. Immunol. 186: 7225–7231. 27. Aguilo, F., S. Avagyan, A. Labar, A. Sevilla, D.-F. Lee, P. Kumar, I. R. Lemischka, B. Y. Zhou, and H.-W. Snoeck. 2011. Prdm16 is a physiologic regulator of hematopoietic stem cells. Blood 117: 5057–5066. References 28. Pronk, C. J. H., O. P. Veiby, D. Bryder, and S. E. W. Jacobsen. 2011. Tumor 1. Broudy, V. C. 1997. Stem cell factor and hematopoiesis. Blood 90: 1345– necrosis factor restricts hematopoietic stem cell activity in mice: involvement of 1364. two distinct receptors. J. Exp. Med. 208: 1563–1570. 2. Ashman, L. K. 1999. The biology of stem cell factor and its receptor C-kit. Int. J. 29. Akashi, K., D. Traver, T. Miyamoto, and I. L. Weissman. 2000. A clonogenic Biochem. Cell Biol. 31: 1037–1051. common myeloid progenitor that gives rise to all myeloid lineages. Nature 404: 3. Chabot, B., D. A. Stephenson, V. M. Chapman, P. Besmer, and A. Bernstein. 193–197. 1988. The proto-oncogene c-kit encoding a transmembrane tyrosine kinase re- 30. Wisniewski, J. R., A. Zougman, N. Nagaraj, and M. Mann. 2009. Universal ceptor maps to the mouse W locus. Nature 335: 88–89. sample preparation method for proteome analysis. Nat. Methods 6: 359–362. 4. Geissler, E. N., M. A. Ryan, and D. E. Housman. 1988. The dominant-white 31. Tenzer, S., D. Docter, S. Rosfa, A. Wlodarski, J. Kuharev, A. Rekik, spotting (W) locus of the mouse encodes the c-kit proto-oncogene. Cell 55: S. K. Knauer, C. Bantz, T. Nawroth, C. Bier, et al. 2011. Nanoparticle size 185–192. is a critical physicochemical determinant of the human blood plasma co- 5. Nocka, K., J. C. Tan, E. Chiu, T. Y. Chu, P. Ray, P. Traktman, and P. Besmer. rona: a comprehensive quantitative proteomic analysis. ACS Nano 5: 7155– 1990. Molecular bases of dominant negative and loss of function mutations at 7167. the murine c-kit/white spotting locus: W37, Wv, W41 and W. EMBO J. 9: 1805– 32. Silva, J. C., R. Denny, C. A. Dorschel, M. Gorenstein, I. J. Kass, G.-Z. Li, 1813. T. McKenna, M. J. Nold, K. Richardson, P. Young, and S. Geromanos. 2005. 6. Zsebo, K. M., D. A. Williams, E. N. Geissler, V. C. Broudy, F. H. Martin, Quantitative proteomic analysis by accurate mass retention time pairs. Anal. H. L. Atkins, R.-Y. Hsu, N. C. Birkett, K. H. Okino, D. C. Murdock, et al. 1990. Chem. 77: 2187–2200. Stem cell factor is encoded at the Sl locus of the mouse and is the ligand for the 33. Geromanos, S. J., J. P. C. Vissers, J. C. Silva, C. A. Dorschel, G.-Z. Li, c-kit tyrosine kinase receptor. Cell 63: 213–224. M. V. Gorenstein, R. H. Bateman, and J. I. Langridge. 2009. The detection, 7. Pittoni, P., S. Piconese, C. Tripodo, and M. P. Colombo. 2011. Tumor-intrinsic correlation, and comparison of peptide precursor and product ions from data and -extrinsic roles of c-Kit: mast cells as the primary off-target of tyrosine independent LC-MS with data dependant LC-MS/MS. Proteomics 9: 1683– kinase inhibitors. Oncogene 30: 757–769. 1695. 8. Galli, S. J., and Y. Kitamura. 1987. Genetically mast-cell-deficient W/Wv and Sl/ 34. Giles, K., S. D. Pringle, K. R. Worthington, D. Little, J. L. Wildgoose, and Sld mice. Their value for the analysis of the roles of mast cells in biologic R. H. Bateman. 2004. Applications of a travelling wave-based radio- responses in vivo. Am. J. Pathol. 127: 191–198. frequency-only stacked ring ion guide. Rapid Commun. Mass Spectrom. 18: 9. Kitamura, Y., S. Go, and K. Hatanaka. 1978. Decrease of mast cells in W/Wv 2401–2414. mice and their increase by bone marrow transplantation. Blood 52: 447– 35. Zhou, J. S., W. Xing, D. S. Friend, K. F. Austen, and H. R. Katz. 2007. Mast cell 452. deficiency in KitW-sh mice does not impair antibody-mediated arthritis. J. Exp. 10. Lyon, M. F., and P. H. Glenister. 1982. A new allele sash (Wsh) at the W-locus and Med. 204: 2797–2802. a spontaneous recessive lethal in mice. Genet. Res. 39: 315–322. 36. Passegue´, E., C. H. Jamieson, L. E. Ailles, and I. L. Weissman. 2003. Normal 11. Duttlinger, R., K. Manova, T. Y. Chu, C. Gyssler, A. D. Zelenetz, and leukemic hematopoiesis: are leukemias a stem cell disorder or a reac- R. F. Bachvarova, and P. Besmer. 1993. W-sash affects positive and negative quisition of stem cell characteristics? Proc. Natl. Acad. Sci. USA 100(Suppl 1): elements controlling c-kit expression: ectopic c-kit expression at sites of kit- 11842–11849. ligand expression affects melanogenesis. Development 118: 705–717. 37. Wilson, A., G. M. Oser, M. Jaworski, W. E. Blanco-Bose, E. Laurenti, 12. Grimbaldeston, M. A., C.-C. Chen, A. M. Piliponsky, M. Tsai, S.-Y. Tam, and C. Adolphe, M. A. Essers, H. R. Macdonald, and A. Trumpp. 2007. Dormant and S. J. Galli. 2005. Mast cell-deficient W-sash c-kit mutant KitW-sh/W-sh mice as self-renewing hematopoietic stem cells and their niches. Ann. N. Y. Acad. Sci. a model for investigating mast cell biology in vivo. Am. J. Pathol. 167: 835– 1106: 64–75. 848. 38. Bao, Y., and X. Cao. 2011. Revisiting the protective and pathogenic roles of 13. Wolters, P. J., J. Mallen-St Clair, C. C. Lewis, S. A. Villalta, P. Baluk, D. J. Erle, neutrophils: Ly-6G is key! Eur. J. Immunol. 41: 2535–2538. and G. H. Caughey. 2005. Tissue-selective mast cell reconstitution and differ- 39. Biermann, H., B. Pietz, R. Dreier, K. W. Schmid, C. Sorg, and C. Sunderko¨tter. ential lung gene expression in mast cell-deficient KitW-sh/KitW-sh sash mice. Clin. 1999. Murine leukocytes with ring-shaped nuclei include granulocytes, mono- Exp. Allergy 35: 82–88. cytes, and their precursors. J. Leukoc. Biol. 65: 217–231. 14. Duttlinger, R., K. Manova, G. Berrozpe, T. Y. Chu, V. DeLeon, I. Timokhina, 40. Greifenberg, V., E. Ribechini, S. Ro¨ssner, and M. B. Lutz. 2009. Myeloid- R. S. Chaganti, A. D. Zelenetz, R. F. Bachvarova, and P. Besmer. 1995. The Wsh derived suppressor cell activation by combined LPS and IFN-g treatment and Ph mutations affect the c-kit expression profile: c-kit misexpression in impairs DC development. Eur. J. Immunol. 39: 2865–2876. embryogenesis impairs melanogenesis in Wsh and Ph mutant mice. Proc. Natl. 41. Gabrilovich, D. I., and S. Nagaraj. 2009. Myeloid-derived suppressor cells as Acad. Sci. USA 92: 3754–3758. regulators of the immune system. Nat. Rev. Immunol. 9: 162–174. 15. Nigrovic, P. A., D. H. D. Gray, T. Jones, J. Hallgren, F. C. Kuo, B. Chaletzky, 42. Cuenca, A. G., M. J. Delano, K. M. Kelly-Scumpia, C. Moreno, P. O. Scumpia, M. Gurish, D. Mathis, C. Benoist, and D. M. Lee. 2008. Genetic inversion in D. M. Laface, P. G. Heyworth, P. A. Efron, and L. L. Moldawer. 2011. A par- mast cell-deficient Wsh mice interrupts corin and manifests as hematopoietic and adoxical role for myeloid-derived suppressor cells in sepsis and trauma. Mol. cardiac aberrancy. Am. J. Pathol. 173: 1693–1701. Med. 17: 281–292. 16. Berrozpe, G., V. Agosti, C. Tucker, C. Blanpain, K. Manova, and P. Besmer. 43. Movahedi, K., M. Guilliams, J. Van den Bossche, R. Van den Bergh, 2006. A distant upstream locus control region is critical for expression of the Kit C.Gysemans,A.Beschin,P.DeBaetselier,andJ.A.VanGinderachter. receptor gene in mast cells. Mol. Cell. Biol. 26: 5850–5860. 2008. Identification of discrete tumor-induced myeloid-derived suppressor 17. Yamazaki, M., T. Tsujimura, E. Morii, K. Isozaki, H. Onoue, S. Nomura, and cell subpopulations with distinct T cell-suppressive activity. Blood 111: 4233– Y. Kitamura. 1994. C-kit gene is expressed by skin mast cells in embryos but not 4244. in puppies of Wsh/Wsh mice: age-dependent abolishment of c-kit gene ex- 44. Youn, J.-I., S. Nagaraj, M. Collazo, and D. I. Gabrilovich. 2008. Subsets of pression. Blood 83: 3509–3516. myeloid-derived suppressor cells in tumor-bearing mice. J. Immunol. 181: 5791– 18. Berrozpe, G., I. Timokhina, S. Yukl, Y. Tajima, M. Ono, A. D. Zelenetz, and 5802. P. Besmer. 1999. The Wsh, W57, and Ph Kit expression mutations define tissue- 45. Youn, J. I., M. Collazo, I. N. Shalova, S. K. Biswas, and D. I. Gabrilovich. 2012. specific control elements located between 223 and 2154 kb upstream of Kit. Characterization of the nature of granulocytic myeloid-derived suppressor cells Blood 94: 2658–2666. in tumor-bearing mice. J. Leukoc. Biol. 91: 167–181. 19. Heib, V., M. Becker, C. Taube, and M. Stassen. 2008. Advances in the under- 46. Yuhas, J. M., R. E. Toya, and E. Wagner. 1975. Specific and nonspecific stimu- standing of mast cell function. Br. J. Haematol. 142: 683–694. lation of resistance to the growth and metastasis of the line 1 lung carcinoma. 20. Beaven, M. A. 2009. Our perception of the mast cell from Paul Ehrlich to now. Cancer Res. 35: 242–244. Eur. J. Immunol. 39: 11–25. 47. Ugel, S., F. Delpozzo, G. Desantis, F. Papalini, F. Simonato, N. Sonda, S. Zilio, 21. Dudeck, A., J. Dudeck, J. Scholten, A. Petzold, S. Surianarayanan, A. Ko¨hler, and V. Bronte. 2009. Therapeutic targeting of myeloid-derived suppressor cells. K. Peschke, D. Vo¨hringer, C. Waskow, T. Krieg, et al. 2011. Mast cells are key Curr. Opin. Pharmacol. 9: 470–481. promoters of contact allergy that mediate the adjuvant effects of haptens. Im- 48. Kao, J., E. C. Ko, S. Eisenstein, A. G. Sikora, S. Fu, and S.-H. Chen. 2011. munity 34: 973–984. Targeting immune suppressing myeloid-derived suppressor cells in oncology. 22. Feyerabend, T. B., A. Weiser, A. Tietz, M. Stassen, N. Harris, M. Kopf, Crit. Rev. Oncol. Hematol. 77: 12–19. P. Radermacher, P. Mo¨ller, C. Benoist, D. Mathis, et al. 2011. Cre-mediated cell 49. Tono, T., T. Tsujimura, U. Koshimizu, T. Kasugai, S. Adachi, K. Isozaki, ablation contests mast cell contribution in models of antibody- and T cell- S. Nishikawa, M. Morimoto, Y. Nishimune, S. Nomura, et al. 1992. c-kit Gene mediated autoimmunity. Immunity 35: 832–844. was not transcribed in cultured mast cells of mast cell-deficient Wsh/Wsh mice 23. Katz, H. R., and K. F. Austen. 2011. Mast cell deficiency, a game of kit and that have a normal number of erythrocytes and a normal c-kit coding region. mouse. Immunity 35: 668–670. Blood 80: 1448–1453. 5544 Myeloid-derived suppressor cells in KitW-sh mutant mice

50. Kimura, Y., B. Ding, N. Imai, D. J. Nolan, J. M. Butler, and S. Raﬁi. 2011. myeloid-derived suppressor cell development by blockade of stem-cell factor c-Kit-mediated functional positioning of stem cells to their niches is es- function. Blood 111: 219–228. sential for maintenance and regeneration of adult hematopoiesis. PLoS ONE 6: 53. Ribatti, D., and E. Crivellato. 2011. Mast cells, angiogenesis, and tumour e26918. growth. Biochim. Biophys. Acta 1822: 2–8. 51. Mazo, I. B., S. Massberg, and U. H. von Andrian. 2011. Hematopoietic stem and 54. Dalton, D. K., and R. J. Noelle. 2012. The roles of mast cells in anticancer progenitor cell trafﬁcking. Trends Immunol. 32: 493–503. immunity. Cancer Immunol. Immunother. 61: 1511–1520. 52. Pan, P.-Y., G. X. Wang, B. Yin, J. Ozao, T. Ku, C. M. Divino, and S.-H. Chen. 55. Gabrilovich, D. I., S. Ostrand-Rosenberg, and V. Bronte. 2012. Coordinated 2008. Reversion of immune tolerance in advanced malignancy: modulation of regulation of myeloid cells by tumours. Nat. Rev. Immunol. 12: 253–268.