Inflammation enhances myeloid-derived suppressor cell ... - CiteSeerX

Uncorrected Version. Published on March 4, 2009 as DOI:10.1189/jlb.0708446

Inflammation enhances myeloid-derived suppressor cell crosstalk by signaling through Toll-like receptor 4 Stephanie K. Bunt,1 Virginia K. Clements,1 Erica M. Hanson, Pratima Sinha, and Suzanne Ostrand-Rosenberg2 Department of Biological Sciences, University of Maryland Baltimore County, Baltimore, Maryland, USA

Abstract: Myeloid-derived suppressor cells (MDSC) are potent inhibitors of anti-tumor immunity that facilitate tumor progression by blocking the activation of CD4ⴙ and CD8ⴙ T cells and by promoting a type 2 immune response through their production of IL-10 and down-regulation of macrophage production of IL-12. MDSC accumulate in many cancer patients and are a significant impediment to active cancer immunotherapies. Chronic inflammation has been shown recently to enhance the accumulation of MDSC and to increase their suppression of T cells. These findings led us to hypothesize that inflammation contributes to tumor progression through the induction of MDSC, which create a favorable environment for tumor growth. As chronic inflammation also drives type 2 immune responses, which favor tumor growth, we asked if inflammation mediates this effect through MDSC. We find that IL-1␤-induced inflammation increased IL-10 production by MDSC and induces MDSC, which are more effective at down-regulating macrophage production of IL-12 as compared with MDSC isolated from less-inflammatory tumor microenvironments, thereby skewing tumor immunity toward a type 2 response. Inflammation heightens MDSC phenotype by signaling through the TLR4 pathway and involves up-regulation of CD14. Although this pathway is well-recognized in other myeloid cells, it has not been implicated previously in MDSC function. These studies demonstrate that MDSC are an intermediary through which inflammation promotes type 2 immune responses, and they identify the TLR4 pathway in MDSC as a potential target for down-regulating immune suppression and promoting anti-tumor immunity. J. Leukoc. Biol. 85: 000 – 000; 2009. Key Words: tumor-induced immune suppression 䡠 T cell activation

䡠

inflammation

INTRODUCTION

responses [1]. Immune cells, primarily macrophages, dendritic cells (DC), and immature bone marrow-derived cells respond to inflammatory signals, such as LPS, from gram-negative bacteria through activation of the TLR4 and CD14 pathway [2]. These acute inflammatory responses typically induce type 1 immunity, which eliminates the initiating stimulus and dampens the inflammatory response. However, persistence of the pathogen, dys-regulation of the inflammatory response, or the inhibition of anti-inflammatory mechanisms may transform acute responses into chronic inflammation and lead to the development of chronic inflammatory diseases [1, 3]. In addition, failure to dampen the inflammatory response is frequently associated with the onset of cancers and is characterized by a predominantly type 2 immune response, which favors tumor progression [4]. Chronic inflammation is also associated with the induction and promotion of immune-suppressive mechanisms, such as the accumulation of myeloid-derived suppressor cells (MDSC), and we have proposed that inflammation may contribute to malignancy by the induction of MDSC [5, 6], which are a heterogeneous population of immature myeloid cells that are present in low levels in healthy individuals and are elevated in patients and experimental animals with cancers [5, 7–16]. They function as potent inhibitors of anti-tumor immunity by blocking the activation of CD4⫹ and CD8⫹ T cells [7, 9 –12], NK cytotoxicity [17, 18], skewing macrophage activity [19], and maturation of DC [8]. In contrast to most immature myeloid cells that are matured by inflammatory stimuli to support tumor-rejecting type 1 immune responses [2], MDSC maintain an immature phenotype when exposed to proinflammatory signals and instead, contribute to a tumor-promoting type 2 phenotype by their production of IL-10 and their blocking of macrophage production of IL-12 [19]. If inflammation facilitates tumor progression through the induction of more suppressive MDSC, then an increasingly proinflammatory environment may enhance the potency of MDSC. Previous studies have confirmed this hypothesis with respect to T cell activation and have shown that increasing inflammation yields more suppres-

1

These authors contributed equally to this work. Correspondence: Department of Biological Sciences, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA. E-mail: [email protected] Received July 25, 2008; revised January 26, 2009; accepted January 30, 2009. doi: 10.1189/jlb.0708446 2

Acute inflammation is a self-limiting, localized response to tissue injury or trauma caused by wounding or infection. This response is aimed at healing the affected tissue by the induction of signals, which trigger innate and adaptive immune 0741-5400/09/0085-0001 © Society for Leukocyte Biology

Journal of Leukocyte Biology Volume 85, June 2009

Copyright 2009 by The Society for Leukocyte Biology.

1

sive MDSC [5, 6]. We now demonstrate that inflammation inhibits anti-tumor immunity further by enhancing the crosstalk between MDSC and macrophages by signaling through the LPS-TLR4 pathway.

for 24 h. Culture supernatants were collected and stored at – 80°C until analyzed for cytokines. Cells were detached using a cell scraper, washed with cold, sterile PBS, stained with antibodies, and analyzed by flow cytometry. For experiments with soluble IL-10, macrophages were prepared and cultured as above, except soluble IL-10 (1.2 ng/ml) was added instead of MDSC.

IL-10 and IL-12 ELISAs MATERIALS AND METHODS Mice Mating pairs of BALB/c, BALB/c TLR4⫺/⫺, and IL-10⫺/⫺ mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). Mating pairs of IL-1R antagonist-deficient (IL-1Ra⫺/⫺) [20] mice, backcrossed to BALB/c mice for 10 –12 generations, were provided by Dr. J. Stuart (University of Tennessee Health Science Center, Memphis, TN, USA). Mice were maintained and bred in the University of Maryland Baltimore County (UMBC; Baltimore, MD, USA) animal facility, and female mice less than 6 months of age were used for all experiments. The UMBC Institutional Animal Care and Use Committee approved all animal procedures.

Thawed supernatants from macrophage/MDSC cocultures were assayed by ELISA for IL-10 and IL-12 using duo set kits (R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s protocols. Plates were read at 420 nm using a Bio-Tek 311 microplate reader and quantified using a standard curve. Data are the mean ⫾ SD of triplicate wells.

Flow cytometry

mAb Gr1-PE, Gr1-FITC, CD11b-FITC, CD11b-PerCP, CD14-FITC, CD80FITC, CD86-PE, CD11c-PE, IL-10-PE, IL-12-PE, rat IgG2␣-PE isotype control, and rat IgG2␣-FITC isotype control were from BD Pharmingen (San Diego, CA, USA); F4/80-FITC and DEC-205 were from Caltag (now Invitrogen, Carlsbad, CA, USA); and TLR4-PE was from Biolegend (San Diego, CA, USA). LPS was from Difco (Detroit, MI, USA), and IFN-␥ was from Pierce (Rockford, IL, USA).

Cells were stained for cell-surface antigens as described [21]. Intracellular staining for IL-10 and IL-12 was performed according to the manufacturer’s directions (BD Pharmingen). Briefly, cells were cultured for 12 h at 3 ⫻ 106/ml with 1 ␮g/ml Brefeldin A. Harvested cells were transferred to 96-well plates (2⫻105 cells/well), incubated for 15 min with FcR-blocking reagent (1 ␮g/100 ␮l PBS/well), labeled externally with antibodies to Gr1 and CD11b for 30 min, washed 2⫻, treated with fix/perm solution for 20 min in the dark, washed 2⫻ with fix/perm solution, stained internally for IL-10 or IL-12, and washed 2⫻. All staining procedures were done on ice. Gating for internal fluorescence was done based on internal isotype control staining. CD11b⫹ macrophages were considered IL-12low or IL-12high if their fluorescence intensity was 26 –95 or ⬎95, respectively (see Fig. 1B). Labeled cells were run on an Epics XL flow cytometer and analyzed using Expo32 Advanced Digital Compensation (Beckman Coulter, Fullerton, CA, USA) or FCS-Express software (DeNovo Software, Los Angeles, CA, USA).

Cell lines

Statistical analyses

The 4T1 mammary carcinoma and 4T1/IL-1␤ cell lines were maintained as described [5, 21].

Student’s two-tailed t-test for unequal variance was performed using Microsoft Excel 2003. Wilcoxon paired-sample rank test was performed as described [22].

Reagents and antibodies

Tumor inoculations and measurements Tumor inoculations and tumor measurements were as described previously [5, 21]. Briefly, female 6- to 10-week-old BALB/c, TLR4⫺/⫺, or IL-1Ra⫺/⫺ mice were inoculated in the mammary fat pad with 7 ⫻ 103 tumor cells in 50 ␮l PBS. Mice were euthanized when tumor diameters reached 10 –12 mm or when mice were moribund.

Blood MDSC Blood Gr1⫹CD11b⫹ MDSC were obtained as described previously [19]. Briefly, mice with ⬎7– 8 mm diameter 4T1 primary mammary tumors have ⬎80% Gr1⫹CD11b⫹ cells in their blood. Tumor-bearing mice were tail bled in heparinized tubes, and the RBC were removed by Gey’s treatment. The remaining white blood cells (WBC) were stained for Gr1 and CD11b and analyzed by flow cytometry, and those preparations that consisted of ⬎90% Gr1⫹CD11b⫹ cells were used in experiments. In some cases, mice were bled the day before an experiment to ascertain that they contained ⬎90% Gr1⫹CD11b⫹ WBC.

Peritoneal macrophages Mice were inoculated i.p. with 1 ml 3% thioglycolate (Difco), and peritoneal exudate cells were collected 5 days later. The peritoneal exudates cells (⬎90%) were F4/80⫹ and CD11b⫹ as measured by flow cytometry.

Macrophage and MDSC coculture experiments Peritoneal macrophages from BALB/c or TLR4⫺/⫺ mice were cocultured with blood-derived MDSC as described previously [19]. Briefly, macrophages were plated at 7.5 ⫻ 105 cells/well in 500 ␮l macrophage medium (DMEM supplemented with 10% FBS) in 24-well plates and incubated at 37°C for 3 h. Nonadherent cells were removed, and 500 ␮l 5% FBS in DMEM was added to the adherent cells. MDSC (1.5⫻106) in 500 ␮l 5% FBS in DMEM were added to some wells. Macrophages and MDSC cocultures were then stimulated with LPS (100 ng/ml) and/or IFN-␥ (2 ng/ml), and cultures were incubated at 37°C

2


RESULTS Inflammation enhances MDSC-mediated suppression of macrophage IL-12 production Macrophages can facilitate tumor progression or tumor rejection, depending on how they have been activated. In most tumor-bearing individuals, tumor-associated macrophages (TAMs; or M2 macrophages) promote tumor progression by their elevated production of IL-10 and minimal production of IL-12 and NO. In vitro activation with IL-4 and IL-13 through the alternative pathway gives a similar profile. In contrast, macrophages activated via the classical pathway with LPS and IFN-␥ or macrophages from IL-1Ra⫺/⫺ mice promote tumor rejection through their production of high levels of IL-12 and NO, and minimal amounts of IL-10 [23–25]. One of the mechanisms by which MDSC facilitate tumor progression is through their cross-talk with macrophages, resulting in a decrease in macrophage production of IL-12 [19]. To determine if inflammation is involved in the MDSC-mediated down-regulation of IL-12, IL-12 production was measured in cultures of peritoneal macrophages from BALB/c mice classically activated with LPS and IFN-␥ in the presence of MDSC, which were obtained from the blood of BALB/c mice with 4T1 mammary carcinoma or from BALB/c mice with 4T1/IL-1␤ tumor cells. We have demonstrated previously that 4T1/IL-1␤ tumors produce a heightened, proinflammatory tumor microenvironment [5]. http://www.jleukbio.org

MDSC from 4T1-bearing mice decreased macrophage production of IL-12 by 33%, and MDSC from mice with 4T1/IL-1␤ tumors decreased macrophage IL-12 production by 84% (Fig. 1A). To confirm that the IL-12 is produced by the macrophages, cells from macrophage alone and macrophage plus MDSC cocultures were surface-stained for CD11b and Gr1 and intracellularly stained for IL-12. Gr1–CD11b⫹ cells (macrophages) were gated and analyzed for IL-12 expression. In the absence of MDSC, ⬎12% of macrophages contained IL-12, and coculture with 4T1 or 4T1/IL-1␤ MDSC reduced this number to ⬍3% (Fig. 1B). Of the CD11b⫹IL-12⫹ macrophages cultured with 4T1 MDSC, 33% had an IL-12high phenotype, and 28% of CD11b⫹IL-12⫹ macrophages cultured with 4T1/IL-1␤ MDSC had an IL-12high phenotype. Therefore, at the cellular level, 4T1/IL-1␤ MDSC are more efficient than 4T1 MDSC at downregulating the quantity of IL-12 made by individual macrophages. Gr1⫹CD11b⫹ MDSC do not stain for IL-12 (data not shown). Experiments using MDSC from IL-1Ra⫺/⫺ mice, which have heightened inflammation as a result of their inability to down-regulate IL-1␤ responses, showed similar downregulation of IL-12 as measured by ELISA (data not shown). Collectively, the intracellular staining and ELISA results demonstrate that IL-1␤ MDSC are more efficient at down-regulating macrophage production of IL-12 made by individual macrophages as compared with the 4T1 MDSC.

Inflammation-induced MDSC produce elevated levels of IL-10 We have shown previously that MDSC-induced down-regulation of macrophage production of IL-12 is dependent on MDSC

synthesis of IL-10 [19]. Therefore, inflammation may decrease macrophage IL-12 production by increasing MDSC production of the type 2 cytokine IL-10. This hypothesis was tested by measuring IL-10 in the supernatants of classically activated macrophages cocultured with MDSC from the blood of BALB/c mice with 4T1 or 4T1/IL-1␤ tumors (Fig. 2A). As observed previously [19], cocultures of MDSC from 4T1 tumor-bearing mice with classically activated macrophages produced significant amounts of IL-10. However, MDSC from tumor-bearing mice with heightened inflammation (i.e., BALB/c mice with 4T1/IL-1␤ tumors) produced significantly more IL-10. Additionally, MDSC from heightened, inflammatory environments and activated with LPS and IFN-␥ in the absence of macrophages produced significantly more IL-10 than MDSC from less inflammatory environments (i.e., MDSC from BALB/c mice with 4T1 tumors). MDSC were also obtained from 4T1 tumor-bearing IL1Ra⫺/⫺ mice and cocultured with BALB/c macrophages. IL1Ra⫺/⫺ mice are unable to attenuate IL-1␤ signaling and therefore have heightened inflammation. MDSC from 4T1 tumor-bearing IL-1Ra⫺/⫺ mice similarly showed increased IL-10 production relative to MDSC from less inflammatory environments (Fig. 2B). Therefore, MDSC, induced by heightened inflammation, by increasing IL-1␤ or inhibiting receptor attenuation by eliminating the receptor antagonist, produce elevated levels of the type 2 cytokine IL-10. To confirm that the IL-10 is produced by the MDSC and not by the macrophages, IL-10⫺/⫺ mice were inoculated with 4T1 or 4T1/IL-1␤ cells, and the resulting MDSC were cocultured with BALB/c macrophages. Cocultures of BALB/c macrophages with 4T1 or 4T1/IL-1␤ MDSC from IL-10⫺/⫺ mice

Fig. 1. Inflammation enhances the ability of MDSC to down-regulate macrophage production of IL-12. BALB/c mice were inoculated with 7000 4T1 or 4T1/IL-1␤ mammary carcinoma cells and bled when their primary tumors were 8 –10 mm in diameter. The resulting nucleated cells, labeled as 4T1 or IL-1␤, respectively, were ⬎88% Gr1⫹CD11b⫹ and were cocultured with peritoneal macrophages (M␾; ⬎97% CD11b⫹) from tumor-free BALB/c mice for 24 h in the presence or absence of LPS and IFN-␥. (〈) Supernatants were tested for IL-12 by ELISA. Data are from one of five experiments. Wilcoxon paired-sample rank test for all experiments shows significance at P ⬍ 0.05 for the indicated comparisons. (B) Cells were stained for CD11b, Gr1, and IL-12, and the gated single CD11b⫹ population was analyzed by flow cytometry for intracellular IL-12. Percent IL-12hi cells is the percent of CD11b⫹IL-12⫹ macrophages that have a fluorescence intensity of ⬎95. Data are from one of three experiments.

Bunt et al. Inflammation facilitates MDSC cross-talk through TLR4

3

IL-10 produced by MDSC down-regulates IL-12 production by macrophages As seen in Figures 1 and 2, inflammation-induced MDSC produced elevated levels of IL-10 and down-regulate macrophage production of IL-12. To determine if the MDSC-produced IL-10 contributes to the decrease in IL-12, MDSC were prepared from BALB/c and IL-10-deficient mice carrying IL-1␤ tumors and cocultured at varying ratios with LPS and IFN-␥-treated BALB/c macrophages, and IL-12 production was measured by ELISA (Fig. 3A). MDSC from wild-type and IL-10⫺/⫺ mice suppress IL-12 production; however, IL-10⫺/⫺ MDSC are significantly less suppressive than wild-type MDSC. To further confirm that IL-10 is responsible for the downregulation, BALB/c and IL-10⫺/⫺ macrophages were cultured with soluble IL-10 (Fig. 3B). Inclusion of soluble IL-10 reduced the level of IL-12 significantly. These results confirm that heightened inflammation induced by IL-1␤ enhances cross-talk between MDSC and macrophages and polarizes immunity toward a tumor-promoting type 2 response of elevated IL-10 and reduced IL-12.

CD14 is up-regulated on inflammation-induced MDSC Figures 1–3 demonstrate that inflammation-primed MDSC are activated further by LPS and IFN-␥ to produce increased levels

Fig. 2. Inflammation increases IL-10 production by MDSC. BALB/c and IL-1Ra⫺/⫺ mice were inoculated with 7000 4T1 or 4T1/IL-1␤ mammary carcinoma cells and bled when their primary tumors were 8 –10 mm in diameter. The resulting nucleated cells were ⬎85% Gr1⫹CD11b⫹ MDSC and were cocultured with peritoneal macrophages from tumor-free BALB/c mice for 24 h in the presence or absence of LPS and IFN-␥. Supernatants were tested for IL-10 by ELISA. (A) MDSC from BALB/c mice with 4T1 or 4T1/IL-1␤ tumors (cocultures treated with LPS and IFN-␥). (B) MDSC from BALB/c or IL-1Ra⫺/⫺ mice with 4T1 tumors (cocultures treated with LPS and IFN-␥). (C) MDSC from BALB/c or IL-10⫺/⫺ mice with 4T1 or 4T1/IL-1␤ tumors (all cocultures include LPS). MDSC and/or macrophages without LPS gave the same results as in A and B. Data for each panel are from one of four to five experiments. Wilcoxon paired-sample rank test for all experiments per panel shows significance at P ⬍ 0.05 for the indicated comparisons.

contained only minimal IL-10, indicating that the MDSC are the source of IL-10 (Fig. 2C). This experiment was performed with LPS and without IFN-␥ to eliminate the possibility that IFN-␥ contributes to the down-regulation of IL-10 [26]. Consistent with the results of Figure 2A, IL-1␤ MDSC produce more IL-10 than MDSC from a less-inflammatory environment. The IL-10 is made by the MDSC and not by the macrophages, as cocultures with IL-10-deficient MDSC make negligible amounts of IL-10. 4


Fig. 3. IL-1␤-enhanced inflammation increases MDSC production of IL-10, which down-regulates macrophage production of IL-12. (A) BALB/c macrophages (⬎90% F4/80⫹ cells) were cocultured with LPS and IFN-␥ at varying ratios with MDSC from BALB/c (⬎86% Gr1⫹CD11b⫹ cells) or IL-10-deficient (⬎90% Gr1⫹CD11b⫹ cells) mice with 4T1/IL-1␤ tumors. IL-12 levels were measured by ELISA. (B) BALB/c or IL-10-deficient macrophages were cocultured with LPS and IFN-␥ and soluble IL-10. Data are from one of two independent experiments for A and B and are from one of five experiments per panel. Wilcoxon paired-sample rank test for all experiments per panel shows significance at P ⬍ 0.05 for the indicated comparisons.

http://www.jleukbio.org

of IL-10 and to down-regulate macrophage production of IL-12. LPS activation occurs through the binding of LPS to the LPSbinding protein, which transfers LPS to CD14 (reviewed in ref. [2]). CD14 then associates with TLR4 and other coreceptors to mediate LPS signaling [27–29], which through this pathway, leads to the induction of several inflammatory mediators, including IL-1, IL-6, cyclooxygenase 2, and inducible NO synthase (reviewed in ref. [30]). The LPS/TLR4 signaling pathway is well-documented in macrophages and DC; however, it has not been described previously in MDSC. To determine whether inflammation up-regulates IL-10 in MDSC through this pathway, MDSC were obtained from wild-type BALB/c or TLR4⫺/⫺ mice with 4T1 or 4T1/IL-1␤ tumors and treated with LPS and IFN-␥ prior to staining with antibodies to Gr1, CD11b, CD14, and TLR4. Gr1⫹CD11b⫹ MDSC were then gated (Fig. 4A) and analyzed by flow cytometry. TLR4 expression is equivalent on MDSC induced by 4T1 and 4T1/IL-1␤ tumor cells and does not change with LPS and IFN-␥ treatment (data not shown). In contrast, CD14 expression is elevated by LPS and IFN-␥ treatment and is highest on Gr1⫹CD11b⫹ MDSC induced by 4T1/IL-1␤ tumor cells (Fig. 4B). MDSC from TLR4-deficient mice, which are unresponsive to LPS [31], display minimal up-regulation of CD14. Therefore, inflammation increases MDSC expression of CD14 through a TLR4-dependent mechanism. As CD11b is also known to interact with TLR4 [32] and to enhance responsiveness to LPS [33], MDSC, induced by 4T1 or 4T1/IL-1␤ tumor cells in wild-type BALB/c or TLR4⫺/⫺ mice, were also analyzed for expression of CD11b (Fig. 4C). CD11b levels are elevated by LPS and IFN-␥ treatment; however, there is no difference in expression of CD11b based on the type of tumor or host, suggesting that heightened inflammation does not mediate its effects on MDSC through differential regulation of CD11b expression. As LPS treatment of immature myeloid cells can induce their differentiation to DC or macrophages, LPS and IFN-␥treated MDSC from BALB/c mice with 4T1 tumors were tested for their expression of the macrophage and DC markers F4/80, CD11c, and DEC205. The cells were also tested for the costimulatory molecules CD80 and CD86, which are elevated on mature DC and activated macrophages, as well as on MDSC [5, 7, 34]. F4/80, CD11c, and DEC205 levels on MDSC from 4T1 tumor-bearing mice remained unchanged following a 24-h incubation with LPS and IFN-␥, whereas CD80 and CD86 levels increased (Fig. 4D). Similar results were seen for MDSC from BALB/c mice with 4T1/IL-1␤ tumors and for cells treated for 48 h (data not shown). These results demonstrate that LPS plus IFN-␥ treatment of MDSC does not induce MDSC differentiation to macrophages or DC. Therefore, LPS plus IFN-␥ treatment increases CD14 expression on MDSC, suggesting that MDSC may be activated through the CD14/TLR4 pathway.

Cross-talk between inflammation-induced MDSC and LPS-activated macrophages is TLR4dependent If inflammation is increasing MDSC production of IL-10 through the CD14/TLR4 signaling pathway, then MDSC from TLR4-deficient, tumor-bearing mice with heightened inflammation should not have elevated levels of IL-10. Likewise, if

MDSC reduce macrophage production of IL-12 through a CD14/TLR4 signaling pathway, then MDSC from TLR4-deficient, tumor-bearing mice should not decrease macrophage production of IL-12. This hypothesis was tested by comparing IL-10 and IL-12 levels in LPS and IFN-␥-treated cocultures of BALB/c macrophages with MDSC from BALB/c or TLR4⫺/⫺ mice with 4T1 or 4T1/IL-1␤ tumors. As observed previously in Figure 2, LPS and IFN-␥-activated, inflammation-induced MDSC (4T1/IL-1␤ vs. 4T1 MDSC) produced elevated levels of IL-10 in the presence and absence of macrophages. In contrast, MDSC from TLR4⫺/⫺ mice produce no IL-10, regardless of the presence of LPS and IFN-␥ or inflammation (Fig. 5A). Similarly, inflammation-induced MDSC from BALB/c mice are potent inhibitors of macrophage production of IL-12 (4T1/ IL-1␤ vs. 4T1 MDSC), as seen previously in Figure 1, whereas MDSC from TLR4⫺/⫺ mice are significantly less effective in reducing macrophage production of IL-12 (Fig. 5B). Therefore, inflammation enhances the ability of MDSC to secrete IL-10 and to down-regulate macrophage production of IL-12 through a LPS/TLR4-dependent pathway in MDSC.

Macrophage induction of IL-10 by MDSC is TLR4-dependent To determine if the effects of macrophages on MDSC are also regulated through the LPS/TLR4 pathway, IL-10 and IL-12 levels were measured in cocultures of TLR4⫺/⫺macrophages and MDSC from BALB/c mice with 4T1 tumors (Fig. 5C). As expected, LPS-treated TLR4⫺/⫺ macrophages do not produce significant levels of IL-12, as they are unable to respond to LPS. In contrast to earlier findings with BALB/c macrophages (Fig. 2), TLR4⫺/⫺ macrophages do not increase MDSC production of IL-10. Therefore, the cross-talk between MDSC and macrophages that promotes a tumor-promoting cytokine phenotype of high IL-10 and low IL-12 is regulated by the TLR4 signaling pathway in MDSC and macrophages.

DISCUSSION The LPS/TLR4 signaling pathway has been well-documented in mature myeloid cells, such as macrophages, but has not been identified previously in myeloid-derived suppressor cells. We now report that IL-1␤-induced inflammation activates MDSC through the TLR4/CD14 pathway and that activation of this pathway in MDSC enhances the cross-talk between MDSC and macrophages in which macrophages up-regulate MDSC production of IL-10, and MDSC down-regulate macrophage production of IL-12. As increases in IL-10 and decreases in IL-12 direct tumor immunity away from a tumor-rejecting type 1 response and toward a tumor-promoting type 2 response, these findings identify another mechanism by which inflammation through TLR4 promotes tumor growth. NF-␬B is the downstream transcription factor that transmits signals following engagement of CD14 and TLR4. It is widely recognized as a key regulator of tumor progression; however, whether it promotes or delays tumor growth depends in which cell type it is expressed and when it is expressed. NF-␬B activation in macrophages results in the production of proin-


5

flammatory mediators, which produce a tumor-rejecting type 1 phenotype [30, 35] that contributes to immune surveillance and tumor rejection [36]. However, increasing tumor burden inhibits the activation of NF-␬B in TAMs, thereby polarizing TAMs towards a tumor-promoting M2 phenotype [37, 38]. In contrast, inactivation of NF-␬B in tumor cells blocks tumor development in models of liver, colon, and mammary cancer, demonstrating that tumor cell expression of NF-␬B enhances tumor progression [39 – 41]. Inactivation of NF-␬B in tumor cells may delay tumor promotion and progression by enhancing apoptosis [40] and/or by shifting the balance from the production of tumor-promoting, proinflammatory TNF-␣ toward TRAIL-mediated apoptosis [41]. Activation through the TLR4 pathway in MDSC also contributes to tumor progression; however, it does so by stimulating immune suppression that blocks tumor immunity. It is intriguing that although macrophages and MDSC share a common hemopoietic lineage, the two populations respond differently to LPS and IFN-␥. As demonstrated in this report, LPS and IFN-␥ treatment of MDSC supports a tumor-promoting type 2 response with decreases in IL-12 and increases in IL-10. As a result, tumor growth is enhanced through the induction of immune suppression. In contrast, treatment of macrophages and DC with LPS and IFN-␥ upregulates MHC II and costimulatory molecule expression and increases macrophage production of IL-12 [42], resulting in a tumor-rejecting type 1 response [2]. Although LPS is a differentiation agent for macrophages [2], it is unlikely that the different responses of macrophages and MDSC are a result of the lack of plasticity of MDSC, as MDSC differentiate to mature DC following treatment with all-trans retinoic acid [10]. The MDSC response to LPS is also unlikely to be a result of LPS-triggered desensitization, a phenomenon that has been reported in macrophages [43]. Macrophage tolerance to LPS is accompanied by increases in CD14 expression [44] and IL-10 production [43], similar to what occurs for MDSC. However, NF-␬B is down-regulated in macrophages that are tolerant to LPS, and NF␬B activation is increased in LPS-treated MDSC (S. K. Bunt and S. Ostrand-Rosenberg, unpublished). Therefore, the differential responses of macrophages and MDSC to

Fig. 4. LPS up-regulates CD14 expression by MDSC. Mice were inoculated in the mammary fat pad with 7000 4T1 or 4T1/IL-1␤ mammary carcinoma cells and bled when their primary tumors were 8 –10 mm in diameter. The resulting nucleated cells were cultured with or without LPS and/or IFN-␥, fluorescently labeled for Gr1, CD11b, and other cell surface markers, and analyzed by flow cytometry. (A) Gated Gr1⫹CD11b⫹ MDSC analyzed in B–D. (B) Nucleated cells were incubated for 24 h without LPS and IFN-␥, with LPS and IFN-␥, or with IFN-␥ alone and stained for Gr1, CD11b, and TLR4, and the gated

6


4 Gr1⫹CD11b⫹ MDSC were analyzed for TLR4 expression. (C) Nucleated cells were incubated for 4 or 24 h without LPS and IFN-␥, with LPS and IFN-␥, or with IFN-␥ alone and stained for Gr1, CD11b, and CD14, and the gated Gr1⫹CD11b⫹ MDSC were analyzed for CD14 expression. Isotype control and CD14 staining were equivalent for MDSC, incubated in the absence of LPS and IFN-␥ so that only one trace is shown. In the TLR4⫺/⫺ MDSC panels, isotype (dotted line) and CD14 (black line) without LPS or IFN-␥ are shown as separate traces, and CD14 with LPS and IFN-␥ (shaded area) is shown. (D) Nucleated cells were incubated for 1 or 24 h without LPS and IFN-␥, with LPS and IFN-␥, or with IFN-␥ alone and stained for Gr1 and CD11b, and the Gr1-gated cells were analyzed for CD11b expression. Isotype control staining was equivalent for cells incubated with or without LPS and IFN-␥ so that only one trace is shown. (E) Nucleated cells from 4T1 tumor-bearing BALB/c mice were incubated for 24 h with or without LPS and IFN-␥ and stained for Gr1, CD11b, and the indicated surface markers, and the gated Gr1⫹CD11b⫹ cells were analyzed. Analyses were done at 24 h, as MDSC viability decreased significantly thereafter. (B–D) Staining patterns for some mAb gave the same pattern so only one tracing is presented (e.g., Panel B, staining with CD14 or isotype mAb gave the same histogram). Data are from one of five, three, three, and three independent experiments for A–D, respectively.


Fig. 5. Cross-talk between MDSC and macrophages requires TLR4. BALB/c or TLR4⫺/⫺ mice were inoculated with 7000 4T1 or 4T1/IL-1␤ mammary carcinoma cells and bled when their primary tumors were 8 –10 mm in diameter. The resulting cells were ⬎85% Gr1⫹CD11b⫹ and were cocultured for 24 h with peritoneal macrophages from BALB/c or TLR4⫺/⫺ mice in the presence or absence of LPS and/or IFN-␥. (A) Gr1⫹CD11b⫹ MDSC from BALB/c mice with 4T1 or 4T1/IL-1␤ tumors and MDSC from TLR4⫺/⫺ mice with 4T1 tumors were cocultured with BALB/c macrophages, and culture supernatants were assayed for IL-10 by ELISA. (B) Gr1⫹CD11b⫹ MDSC from BALB/c mice with 4T1 or 4T1-IL-1␤ tumors and MDSC from TLR4⫺/⫺ mice with 4T1 tumors were cocultured with BALB/c macrophages, and culture supernatants were assayed for IL-12 by ELISA. (C) Gr1⫹CD11b⫹ MDSC from BALB/c mice with 4T1 tumors were cocultured with TLR4⫺/⫺ macrophages, and supernatants were assayed for IL-10 and IL-12 by ELISA. Data are from one of five experiments per panel. Wilcoxon paired-sample rank test for all experiments per panel shows significance at P ⬍ 0.05 for the indicated comparisons.

LPS are not a result of different maturation states or developmental plasticity or to LPS desensitization. The studies reported here used IL-1␤ to activate MDSC in vivo and LPS and IFN-␥ to boost MDSC and activate macrophages in vitro. In addition to IL-1␤, MDSC are activated in vivo by other molecules, including PGE2, IL-6, vascular endothelial growth factor, and the proinflammatory proteins S100A8/A9 [6, 10, 45– 47]. As these molecules have specific cell surface receptors on MDSC, TLR4 may not be involved in their signaling pathways. However, mobilization of NF-␬B is involved in S100A8/A9 activation of MDSC [46], suggesting that activation by the different MDSC inducers may converge on a common pathway requiring NF-␬B. MDSC treated with LPS plus IFN-␥ have increased levels of CD14 and unchanged levels of TLR4. LPS and IFN-␥ are reported to up-regulate CD14 on macrophages [48], and LPS down-regulates TLR4, and IFN-␥ up-regulates TLR4 [49]. As the combined treatment of MDSC with LPS and IFN-␥ gives the maximal response, it is likely that the LPS-mediated decrease of TLR4 is prevented by IFN-␥. By acting synergistically, LPS and IFN-␥ increase CD14 expression, which promotes survival by protecting against apoptosis [50] and enhances LPS responsiveness by reducing the concentration of LPS required for TLR4 signaling [43]. The experiments described in this report used LPS and IFN-␥ to activate inflammation-induced MDSC in vitro, raising the question of which molecules activate MDSC through the CD14/TLR4 pathway in vivo. Chronic inflammatory conditions are triggered frequently and/or maintained by infectious pathogens [51], which have the potential to produce molecules such as LPS. However, the etiology of many inflammatory conditions is unknown. In these cases, stimulation of TLR4 on MDSC may be a result of endogenous ligands released by injured or necrotic tissues, such as high-mobility group box 1 protein, a proinflammatory cytokine released by necrotic cells [52], heat shock proteins 60, 70, or 90, heparin sulfate, or soluble hyaluronan [53]. These molecules have been implicated as TLR4 ligands, so it is not unreasonable that they would also activate MDSC through TLR4. Interestingly, paclitaxel is also a TLR4 ligand, raising the possibility that certain cancer chemotherapeutic drugs may actually promote tumor progression by activating MDSC [54]. Tumor progression is a balance among many factors, some of which are regulated by inflammation. The connection among inflammation, MDSC-suppressive activity, and TLR4 demonstrated in this report identifies TLR4 and/or its downstream signaling factors as potential targets for reducing and/or inactivating this suppressor cell population.

ACKNOWLEDGMENTS This work was supported by National Institutes of Health Grants R01 CA115880 and R01 CA84232 and Susan G. Komen Grant for the Cure BCTR0503885. S. K. B. is supported by a predoctoral fellowship from the Department of Defense Breast Cancer Research Program (W81XWH-05-10276). The authors thank Dr. Jeff Leips for his invaluable help with statistical analyses, Dr. J. Stuart for providing the IL-


7

1Ra⫺/⫺ mice, and Ms. Sandy Mason and Mr. Terry King for their excellent care of our mice.

REFERENCES 1. Coussens, L. M., Werb, Z. (2002) Inflammation and cancer. Nature 420, 860 – 867. 2. Medzhitov, R. (2001) Toll-like receptors and innate immunity. Nat. Rev. Immunol. 1, 135–145. 3. Baniyash, M. (2006) Chronic inflammation, immunosuppression and cancer: new insights and outlook. Semin. Cancer Biol. 16, 80 – 88. 4. Becker, Y. (2006) Molecular immunological approaches to biotherapy of human cancers—a review, hypothesis and implications. Anticancer Res. 26, 1113–1134. 5. Bunt, S. K., Sinha, P., Clements, V. K., Leips, J., Ostrand-Rosenberg, S. (2006) Inflammation induces myeloid-derived suppressor cells that facilitate tumor progression. J. Immunol. 176, 284 –290. 6. Sinha, P., Clements, V. K., Fulton, A. M., Ostrand-Rosenberg, S. (2007) Prostaglandin E2 promotes tumor progression by inducing myeloid-derived suppressor cells. Cancer Res. 67, 4507– 4513. 7. Sinha, P., Clements, V. K., Ostrand-Rosenberg, S. (2005) Reduction of myeloid-derived suppressor cells and induction of M1 macrophages facilitate the rejection of established metastatic disease. J. Immunol. 174, 636 – 645. 8. Almand, B., Clark, J. I., Nikitina, E., van Beynen, J., English, N. R., Knight, S. C., Carbone, D. P., Gabrilovich, D. I. (2001) Increased production of immature myeloid cells in cancer patients: a mechanism of immunosuppression in cancer. J. Immunol. 166, 678 – 689. 9. Bronte, V., Chappell, D. B., Apolloni, E., Cabrelle, A., Wang, M., Hwu, P., Restifo, N. P. (1999) Unopposed production of granulocyte-macrophage colony-stimulating factor by tumors inhibits CD8⫹ T cell responses by dysregulating antigen-presenting cell maturation. J. Immunol. 162, 5728 –5737. 10. Gabrilovich, D. I., Velders, M. P., Sotomayor, E. M., Kast, W. M. (2001) Mechanism of immune dysfunction in cancer mediated by immature Gr-1⫹ myeloid cells. J. Immunol. 166, 5398 –5406. 11. Kusmartsev, S., Nefedova, Y., Yoder, D., Gabrilovich, D. I. (2004) Antigen-specific inhibition of CD8⫹ T cell response by immature myeloid cells in cancer is mediated by reactive oxygen species. J. Immunol. 172, 989 –999. 12. Serafini, P., De Santo, C., Marigo, I., Cingarlini, S., Dolcetti, L., Gallina, G., Zanovello, P., Bronte, V. (2004) Derangement of immune responses by myeloid suppressor cells. Cancer Immunol. Immunother. 53, 64 –72. 13. Schmielau, J., Finn, O. J. (2001) Activated granulocytes and granulocytederived hydrogen peroxide are the underlying mechanism of suppression of T-cell function in advanced cancer patients. Cancer Res. 61, 4756 – 4760. 14. Young, M. R., Petruzzelli, G. J., Kolesiak, K., Achille, N., Lathers, D. M., Gabrilovich, D. I. (2001) Human squamous cell carcinomas of the head and neck chemoattract immune suppressive CD34(⫹) progenitor cells. Hum. Immunol. 62, 332–341. 15. Mirza, N., Fishman, M., Fricke, I., Dunn, M., Neuger, A. M., Frost, T. J., Lush, R. M., Antonia, S., Gabrilovich, D. I. (2006) All-trans-retinoic acid improves differentiation of myeloid cells and immune response in cancer patients. Cancer Res. 66, 9299 –9307. 16. Diaz-Montero, C. M., Salem, M. L., Nishimura, M. I., Garrett-Mayer, E., Cole, D. J., Montero, A. J. (2009) Increased circulating myeloid-derived suppressor cells correlate with clinical cancer stage, metastatic tumor burden, and doxorubicin-cyclophosphamide chemotherapy. Cancer Immunol. Immunother. 58, 49 –59. 17. Suzuki, E., Kapoor, V., Jassar, A. S., Kaiser, L. R., Albelda, S. M. (2005) Gemcitabine selectively eliminates splenic Gr-1⫹/CD11b⫹ myeloid suppressor cells in tumor-bearing animals and enhances antitumor immune activity. Clin. Cancer Res. 11, 6713– 6721. 18. Liu, C., Yu, S., Kappes, J., Wang, J., Grizzle, W. E., Zinn, K. R., Zhang, H. G. (2007) Expansion of spleen myeloid suppressor cells represses NK cell cytotoxicity in tumor-bearing host. Blood 109, 4336 – 4342. 19. Sinha, P., Clements, V. K., Bunt, S. K., Albelda, S. M., Ostrand-Rosenberg, S. (2007) Cross-talk between myeloid-derived suppressor cells and macrophages subverts tumor immunity toward a type 2 response. J. Immunol. 179, 977–983. 20. Hirsch, E., Irikura, V. M., Paul, S. M., Hirsh, D. (1996) Functions of interleukin 1 receptor antagonist in gene knockout and overproducing mice. Proc. Natl. Acad. Sci. USA 93, 11008 –11013.

8


21. Pulaski, B. A., Ostrand-Rosenberg, S. (1998) Reduction of established spontaneous mammary carcinoma metastases following immunotherapy with major histocompatibility complex class II and B7.1 cell-based tumor vaccines. Cancer Res. 58, 1486 –1493. 22. Zar, J. H. (1999) Biostatistical Analysis, Upper Saddle River, NJ, USA, Prentice Hall. 23. Mantovani, A., Sozzani, S., Locati, M., Allavena, P., Sica, A. (2002) Macrophage polarization: tumor-associated macrophages as a paradigm for polarized M2 mononuclear phagocytes. Trends Immunol. 23, 549 –555. 24. Gordon, S. (2003) Alternative activation of macrophages. Nat. Rev. Immunol. 3, 23–35. 25. Sinha, P., Clements, V. K., Ostrand-Rosenberg, S. (2005) Interleukin-13regulated M2 macrophages in combination with myeloid suppressor cells block immune surveillance against metastasis. Cancer Res. 65, 11743– 11751. 26. Trinchieri, G. (1997) Cytokines acting on or secreted by macrophages during intracellular infection (IL-10, IL-12, IFN-␥). Curr. Opin. Immunol. 9, 17–23. 27. Jiang, Q., Akashi, S., Miyake, K., Petty, H. R. (2000) Lipopolysaccharide induces physical proximity between CD14 and Toll-like receptor 4 (TLR4) prior to nuclear translocation of NF-␬ B. J. Immunol. 165, 3541–3544. 28. Zarewych, D. M., Kindzelskii, A. L., Todd III, R. F., Petty, H. R. (1996) LPS induces CD14 association with complement receptor type 3, which is reversed by neutrophil adhesion. J. Immunol. 156, 430 – 433. 29. Pfeiffer, A., Bo¨ttcher, A., Orso´, E., Kapinsky, M., Nagy, P., Bodnar, A., Spreitzer, I., Liebisch, G., Drobnik, W., Gempel, K., Horn, M., Holmer, S., Hartung, T., Multhoff, G., Schu¨tz, G., Schindler, H., Ulmer, A. J., Heine, H., Stelter, F., Schu¨tt, C., Rothe, G., Szo¨lloˆsi, J., Damjanovich, S., Schmitz, G. (2001) Lipopolysaccharide and ceramide docking to CD14 provokes ligand-specific receptor clustering in rafts. Eur. J. Immunol. 31, 3153–3164. 30. Karin, M., Greten, F. R. (2005) NF-␬B: linking inflammation and immunity to cancer development and progression. Nat. Rev. Immunol. 5, 749 –759. 31. Vogel, S. N., Wax, J. S., Perera, P. Y., Padlan, C., Potter, M., Mock, B. A. (1994) Construction of a BALB/c congenic mouse, C.C3H-Lpsd, that expresses the Lpsd allele: analysis of chromosome 4 markers surrounding the Lps gene. Infect. Immun. 62, 4454 – 4459. 32. Zhou, X., Gao, X. P., Fan, J., Liu, Q., Anwar, K. N., Frey, R. S., Malik, A. B. (2005) LPS activation of Toll-like receptor 4 signals CD11b/CD18 expression in neutrophils. Am. J. Physiol. Lung Cell. Mol. Physiol. 288, L655–L662. 33. Perera, P. Y., Mayadas, T. N., Takeuchi, O., Akira, S., Zaks-Zilberman, M., Goyert, S. M., Vogel, S. N. (2001) CD11b/CD18 acts in concert with CD14 and Toll-like receptor (TLR) 4 to elicit full lipopolysaccharide and taxol-inducible gene expression. J. Immunol. 166, 574 –581. 34. Yang, R., Cai, Z., Zhang, Y., Yutzy IV, W. H., Roby, K. F., Roden, R. B. (2006) CD80 in immune suppression by mouse ovarian carcinoma-associated Gr-1⫹CD11b⫹ myeloid cells. Cancer Res. 66, 6807– 6815. 35. Lin, W. W., Karin, M. (2007) A cytokine-mediated link between innate immunity, inflammation, and cancer. J. Clin. Invest. 117, 1175–1183. 36. Dunn, G. P., Old, L. J., Schreiber, R. D. (2004) The immunobiology of cancer immunosurveillance and immunoediting. Immunity 21, 137–148. 37. Biswas, S. K., Gangi, L., Paul, S., Schioppa, T., Saccani, A., Sironi, M., Bottazzi, B., Doni, A., Vincenzo, B., Pasqualini, F., Vago, L., Nebuloni, M., Mantovani, A., Sica, A. (2006) A distinct and unique transcriptional program expressed by tumor-associated macrophages (defective NF-␬B and enhanced IRF-3/STAT1 activation). Blood 107, 2112–2122. 38. Saccani, A., Schioppa, T., Porta, C., Biswas, S. K., Nebuloni, M., Vago, L., Bottazzi, B., Colombo, M. P., Mantovani, A., Sica, A. (2006) p50 nuclear factor-␬B overexpression in tumor-associated macrophages inhibits M1 inflammatory responses and antitumor resistance. Cancer Res. 66, 11432– 11440. 39. Pikarsky, E., Porat, R. M., Stein, I., Abramovitch, R., Amit, S., Kasem, S., Gutkovich-Pyest, E., Urieli-Shoval, S., Galun, E., Ben-Neriah, Y. (2004) NF-␬B functions as a tumor promoter in inflammation-associated cancer. Nature 431, 461– 466. 40. Greten, F. R., Eckmann, L., Greten, T. F., Park, J. M., Li, Z. W., Egan, L. J., Kagnoff, M. F., Karin, M. (2004) IKK␤ links inflammation and tumorigenesis in a mouse model of colitis-associated cancer. Cell 118, 285–296. 41. Luo, J. L., Maeda, S., Hsu, L. C., Yagita, H., Karin, M. (2004) Inhibition of NF-␬B in cancer cells converts inflammation-induced tumor growth mediated by TNF␣ to TRAIL-mediated tumor regression. Cancer Cell 6, 297–305. 42. Takeuchi, O., Akira, S. (2001) Toll-like receptors; their physiological role and signal transduction system. Int. Immunopharmacol. 1, 625– 635.


43. Fujihara, M., Muroi, M., Tanamoto, K., Suzuki, T., Azuma, H., Ikeda, H. (2003) Molecular mechanisms of macrophage activation and deactivation by lipopolysaccharide: roles of the receptor complex. Pharmacol. Ther. 100, 171–194. 44. Ziegler-Heitbrock, H. W., Wedel, A., Schraut, W., Strobel, M., Wendelgass, P., Sternsdorf, T., Bauerle, P. A., Haas, J. G., Riethmuller, G. (1994) Tolerance to lipopolysaccharide involves mobilization of nuclear factor ␬ B with predominance of p50 homodimers. J. Biol. Chem. 269, 17001– 17004. 45. Bunt, S. K., Yang, L., Sinha, P., Clements, V. K., Leips, J., OstrandRosenberg, S. (2007) Reduced inflammation in the tumor microenvironment delays the accumulation of myeloid-derived suppressor cells and limits tumor progression. Cancer Res. 67, 10019 –10026. 46. Sinha, P., Okoro, C., Foell, D., Freeze, H. H., Ostrand-Rosenberg, S., Srikrishna, G. (2008) Proinflammatory S100 proteins regulate the accumulation of myeloid-derived suppressor cells. J. Immunol. 181, 4666 – 4675. 47. Cheng, P., Corzo, C. A., Luetteke, N., Yu, B., Nagaraj, S., Bui, M. M., Ortiz, M., Nacken, W., Sorg, C., Vogl, T., Roth, J., Gabrilovich, G. I. (2008) Inhibition of dendritic cell differentiation and accumulation of myeloid-derived suppressor cells in cancer is regulated by S100A9 protein. J. Exp. Med. 205, 2235–2249. 48. Takeshita, S., Nakatani, K., Takata, Y., Kawase, H., Sekine, I., Yoshioka,

49.

50.

51. 52.

53.

54.

S. (1998) Interferon-␥ (IFN-␥) and tumor necrosis factor-␣ (TNF-␣) enhance lipopolysaccharide binding to neutrophils via CD14. Inflamm. Res. 47, 101–103. Bosisio, D., Polentarutti, N., Sironi, M., Bernasconi, S., Miyake, K., Webb, G. R., Martin, M. U., Mantovani, A., Muzio, M. (2002) Stimulation of Toll-like receptor 4 expression in human mononuclear phagocytes by interferon-␥: a molecular basis for priming and synergism with bacterial lipopolysaccharide. Blood 99, 3427–3431. Frey, E. A., Finlay, B. B. (1998) Lipopolysaccharide induces apoptosis in a bovine endothelial cell line via a soluble CD14 dependent pathway. Microb. Pathog. 24, 101–109. Balkwill, F., Mantovani, A. (2001) Inflammation and cancer: back to Virchow? Lancet 357, 539 –545. Lotze, M. T., Tracey, K. J. (2005) High-mobility group box 1 protein (HMGB1): nuclear weapon in the immune arsenal. Nat. Rev. Immunol. 5, 331–342. Chen, R., Alvero, A. B., Silasi, D. A., Mor, G. (2007) Inflammation, cancer and chemoresistance: taking advantage of the Toll-like receptor signaling pathway. Am. J. Reprod. Immunol. 57, 93–107. Kelly, M. G., Alvero, A. B., Chen, R., Silasi, D. A., Abrahams, V. M., Chan, S., Visintin, I., Rutherford, T., Mor, G. (2006) TLR-4 signaling promotes tumor growth and paclitaxel chemoresistance in ovarian cancer. Cancer Res. 66, 3859 –3868.


9