The synergistic antitumor effects of the combination therapy imatinib mesylate (IM) and IL-2 depended upon NK1.1- expressing cells and were associated with the accumulation of CD11cintB220+NK1.1+ IFN-producing killer dendritic cells (IKDC) into tumor beds. In this study, we show that the antitumor efficacy of the combination therapy was compromised in IL-15 and IFN-type 1R loss-of-function mice. IL-15Rα was required for the proliferation of IKDC during IM plus IL-2 therapy. Trans-presentation of IL-15/IL-15Rα activated IKDC to express CCR2 and to respond to type 1 IFN by producing CCL2. Moreover, the antitumor effects of the combination therapy correlated with a CCL2-dependent recruitment of IKDC, but not B220 NK cells, into tumor beds. Altogether, the IL-15-driven peripheral expansion and the CCL-2-dependent intratumoral chemoattraction of IKDC are two critical parameters dictating the antitumor efficacy of IM plus IL-2 in mice.

Accumulating evidence indicates that the innate and adaptive immune systems make a crucial contribution to the antitumor effects of conventional chemotherapy and radiotherapy-based cancer treatments (1, 2). The molecular and cellular bases of the immunogenicity of cell death induced by cytotoxic agents are being progressively unraveled (3, 4, 5, 6). Along these lines, we highlighted the off-target immunological effects of a tyrosine kinase inhibitor imatinib mesylate (IM;4 STI571/Gleevec) (7, 8). The 2-phenylaminopyrimidine compound has been initially designed to specifically block the ATP binding site of BCR/ABL tyrosine kinase and found to inhibit the kinase activity of the three related kinases BCR-ABL, platelet-derived growth factor receptor, and KIT (8, 9). IM is a clinically approved drug commonly used for the treatment of gastrointestinal stromal tumors because it directly targets the pathogenomic c-kit mutation responsible for the deregulated proliferation of tumor precursors (10). However, IM could mediate antitumor effects using an alternate mode of action than the direct effect on tumoral c-kit mutations. Borg et al. (7) selected mouse tumor models that resisted the antiproliferative effects of IM in vitro, yet responded to long-term exposure to IM in vivo. They showed that IM acted on host dendritic cells (DC) to promote NK cell activation and NK cell-dependent antitumor effects in mice and humans (7). c-kit triggering was required for the DC-mediated NK cell activation induced by IM.

To potentiate the NK cell activation induced by c-kit tyrosine kinase inhibition in vivo, IM was combined with IL-2 (11). Flow cytometry studies of single-cell suspensions obtained from regressing lung metastases highlighted that the CD11c+B220+NK1.1+ cells were increased by 4-fold during the combination therapy with IM plus IL-2 compared with each treatment modality alone (IM or IL-2). The CD11c+B220+NK1.1+ cell population expressed other NK cell markers such as the integrin VLA-2 recognized by anti-CD49b/Dx5 mAb, CD122, NKG2D, CD11b, but failed to express CD3, CD4, CD8α, CD25, PDCA-1, and costimulatory molecules (CD40, CD80, CD86). Because up to 50% of the CD11c+B220+NK1.1+ cells coexpressed I-Ab in tumors (11) and because they could stimulate naive OVA -specific TCR-transgenic T cells when isolated from lymph nodes (12) and produce IFN-γ in contact with tumor cells (11), these CD11c+ cells were named IFN-producing killer DC (IKDC).

In an attempt to delineate the mechanisms by which IKDC could proliferate and/or be recruited into tumor beds, we investigated the role of key regulatory cytokines and chemokines in the antitumor effects mediated by IM plus IL-2 and found the critical involvement of the type 1 IFN/IL-15/CCL2 axis in the proliferation and recruitment of IKDC into tumors.

Female C57BL/6 wild-type (WT) mice were obtained from the Centre d’ Elevage Janvier and used at 6–10 wk of age. IFN-type 1R−/− and CCL2/MCP-1−/− mice backcrossed on a C57BL/6 background were provided by Centre d’Elevage d’Orléans. IL-15Rα−/− and IL-15−/− mice were backcrossed on a C57BL/6 6–8 times and maintained at the animal facility of EB (Research Center Borstel, Borstel, Germany). Animals were all maintained according to the Animal Experimental Ethics Committee Guidelines. B16F10 is a melanoma cell line syngeneic of C57BL/6 (provided by M. T. Lotze, University of Pittsburgh, Pittsburgh, PA) that was cultured in RPMI 1640 (Life Technologies 31870) with 10% heat-inactivated FBS enriched with 5% l-glutamine, non essential amino acids, sodium pyruvate, and antibiotics. MS5 feeder cell lines (provided by W. Vainchenker, Institut Gustave Roussy, Villejuif, France) were cultured in IMDM (Sigma 13390) containing 10% heat-inactivated FBS, 5% l-glutamine, sodium pyruvate, and antibiotics.

Lung metastases model.

B16F10 cells (3 × 105) were injected at day 0 into the tail vein and mice were sacrificed at day 11. Two hundred microliters of H2O or 150 mg/kg IM (Gleevec; Novartis) was given orally in mice twice a day (bid) from days 0 to 10 after tumor inoculation alone or combined with other therapies. Control groups (H2O plus PBS named PBS), IM alone (IM plus PBS named IM), IL-2 alone (H2O plus IL-2 named IL-2), cyclophosphamide (CTX, Endoxan; Baxter) alone (H2O plus CTX named CTX), CpG oligodeoxynucleotide (ODN) alone (H2O plus CpG named CpG), and combination therapies IM plus CTX, IM plus CpG, or IM plus IL-2 were preformed as stated. rhIL-2 (1 × 105 IU; Chiron) were injected bid by the i.p. route from days 7 to 10 or 100 mg/kg CTX was administered i.p. at day 0, while 10 μg of CpG ODN (5′-TGACTGTGAACGTTCGAGATGA, given by A. Carpentier (Paris, France), AP-HP Pitié Salpétrière) were injected s.c. from days 7 to 10. PBS was injected i.p. or s.c. in control groups.

Skin model.

B16F10 cells (3 × 105) were injected s.c., and mice were treated with combination therapy IM plus IL-2 as described above, but with two different doses of IL-2 (105 or 3 × 105 bid).

Depletion of plasmacytoid DC (pDC).

Mice were depleted by injection of 300 μg of anti-PDCA1-depleting mAb (clone 120G8; Miltenyi Biotec) at day −3 before tumor inoculation and day 0, day +3, and day +9 after tumor inoculation. Effective depletion was monitored by FACS using the pDC-specific staining (CD3CD19NK1.1CD11cintB220+Gr1+).

FACS analyses were performed using allophycocyanin-conjugated anti-CD11c mAb (HL3), PE-Cy7-conjugated anti-NK1.1 mAb (PK136), allophycocyanin-Cy7-conjugated anti-B220 mAb (RA3-6B2), FITC-conjugated anti-CD3 mAb (17A2), FITC-conjugated CD19 (1D3), and PE-conjugated Gr1. IKDC are defined as CD3CD19NK1.1+B220+CD11cintGR1, NK as CD3CD19NK1.1+B220CD11c+/−GR1, and pDC as CD3CD19NK1.1B220CD11cintGR1+. Abs were purchased from BD Pharmingen or eBioscience. Anti-CCR2 mAb (MC21, rat IgG2a) has been provided by M. Mack (Universitätsklinikum Regensburg, Abteilung für Nephrologie, Regensburg, Germany) (13). Cells were preincubated with Fc block for 20 min (CD16/CD32, 2.4G2; BD Pharmingen, diluted in PBS with 2% mice serum and 2% FBS) and afterward stained for 20 min at 4°C with the different Abs at 1/200 (anti-CCR2 at 1/20). Alexa Fluor 488 goat anti-rat IgG has been used as secondary Ab at 1/400 (Invitrogen). Immediately before FACS analysis, 4′,6-diamidino-2-phenylindole (Sigma-Aldrich) was added. FACS analysis was performed by an LSRII (BD Biosciences) using FACSDiva software (BD Biosciences), CellQuestPro software (BD Biosciences), or Flow Jo (Tree Star).

Sorting of NK and IKDC cells and in vitro expansion of IKDC15 is described in our unpublished data. Sorting of NK and IKDC was realized on a Mo-Flo instrument (DakoCytomation), CD3CD10C11cintB220+NK1.1+ cells defined “IKDC” and CD3CD19CD11c+/−B220NK1.1+ cells defined B220NK cells. In vitro expansion of IKDC15 was realized as follows: 104 freshly cell sorted IKDC were cultured in the presence of murine stromal cells MS-5 in DMEM (GIBCO 41966) culture medium containing 4500 mg/l of glucose, 5% l-glutamin, pyruvate, and enriched with Abs, 10% BGS (Lot ANB 18298, HyClone), and 20 ng/ml rmIL-15/ml (R&D Systems). While B220 NK cells could not proliferate ex vivo in similar conditions as IKDC, we could maintain NK cells at higher concentrations (5 × 105/ml) for 7 days on MS-5 feeders and rIL-15 to allow fair comparisons with IKDC.

In brief, 105 freshly sorted NK cells and IKDC or IKDC15 (obtained at day 7 of ex vivo expansion) or NK and IKDC stimulated with recombinant murine IL-15 (20 ng/ml; R&D Systems) for 24 h in the presence of MS-5 were further incubated with recombinant human (rh) IL-2 (105 IU/ml) or type 1 IFN at 2500 IU/ml (14) (provided by M. Ferrantini, Instituto di Sanita, Rome, Italy). These in vitro cultures were performed in 200 μl of complete medium in 96-round-bottom well plates. After 24–36 h, cell supernatants were collected and commercial LUMINEX kits, (Linco and BioSource International) were used to determine chemokine and cytokine profiles (used according to the manufacturer’s conditions).

WT C57BL/6 and IL-15Rα−/− mice were treated with IL-2 combined with IM according to a protocol already reported (11). Mice received an i.p. injection of 100 μg of BrdU/100 μl PBS) 1 day before sacrifice. Spleen cells were harvested and processed according to the manufacturer’s protocol (BrdU Flow Kit; BD Biosciences). Briefly, cells were stained for surface Ags, fixed, and permeabilized. DNase digestion followed by staining with anti-BrdU mAb was performed before flow cytometry analyses.

Aberrant values were excluded using Dixon’s test. Normality of distributions was assessed using the Shapiro-Wilk test. Normal distributions were compared by the Student t test; non-normal samplings were compared using the Mann-Whitney U test. ANOVAs were performed with the Kruskal-Wallis test. Values of p inferior to 0.05 were considered significant. All tests were done using Prism 5 software (GraphPad).

The natural immunosurveillance against B16F10 melanoma developing into lung metastases is known to rely on innate NK cell effectors (15, 16). We previously showed that IM could induce NK1.1+ cell-dependent tumor regressions (7). To further improve the NK cell triggering effects of IM, we combined IM with various NK adjuvants such as TLR9 ligands (CpG ODN (17, 18)) or metronomic CTX suppressing the regulatory T cell-mediated NK cell inhibitory effects (19, 20, 21) or IL-2 (22, 23). We observed significant synergistic effects with CTX (Fig. 1,A) with CpG ODN (Fig. 1,B) or rhIL-2 (Fig. 1,C) but the most remarkable antitumoral effects were achieved with the combination of IM + IL-2 (Fig. 1,D). Similarly, the growth of B16F10 melanoma inoculated into the skin was also impaired by the combination therapy but higher dosages of IL-2 were required for a significant tumor growth delay (Fig. 1 E). It is of note that all of these agents did not impact on B16F10 proliferation in vitro (data not shown). We already reported that depleting anti-NK1.1 mAb completely curtailed the tumoricidal activity induced by the combination of IM + IL-2 (11).

FIGURE 1.

Combining IM with NK cell adjuvants. Established B16F10 lung metastases model in C57BL/6 mice. B16F10 tumor cells (5 × 105) were inoculated in the tail vein at day 0. A, Mice were injected i.p. with PBS or low-dose CTX (100 mg/kg) and were fed with H2O or IM bid at 150 mg/kg per day from days 0 to 10. B, Mice were fed with H2O or IM as in A, but were treated with s.c. administration of PBS or CpG ODN (10 μg/day) from days 7 to 10. C, Mice were fed with H2O or IM as in A, but were treated bid with i.p. administration of PBS or 105 IU of rhIL-2 from days 7 to 10. D, Mice were fed and treated systemically with the above-mentioned immunotherapies (A–C) in the same experiment. Graphs represent the number of lung metastases at day 11. E, s.c. B16F10 tumor model in C57BL/6 mice. Same as in C but 3 × 105 B16F10 cells were inoculated in the skin of flanks and two doses of IL-2 (105 IU or 3 × 105 IU bid) were used. The tumor size was monitored twice a week. A representative experiment of two to five including five mice per group is depicted. Statistical analyses were performed using the Kruskal-Wallis test. Values of p inferior to 0.05, 0.01, or 0.001 are indicated with ∗, ∗∗, or ∗∗∗, respectively.

FIGURE 1.

Combining IM with NK cell adjuvants. Established B16F10 lung metastases model in C57BL/6 mice. B16F10 tumor cells (5 × 105) were inoculated in the tail vein at day 0. A, Mice were injected i.p. with PBS or low-dose CTX (100 mg/kg) and were fed with H2O or IM bid at 150 mg/kg per day from days 0 to 10. B, Mice were fed with H2O or IM as in A, but were treated with s.c. administration of PBS or CpG ODN (10 μg/day) from days 7 to 10. C, Mice were fed with H2O or IM as in A, but were treated bid with i.p. administration of PBS or 105 IU of rhIL-2 from days 7 to 10. D, Mice were fed and treated systemically with the above-mentioned immunotherapies (A–C) in the same experiment. Graphs represent the number of lung metastases at day 11. E, s.c. B16F10 tumor model in C57BL/6 mice. Same as in C but 3 × 105 B16F10 cells were inoculated in the skin of flanks and two doses of IL-2 (105 IU or 3 × 105 IU bid) were used. The tumor size was monitored twice a week. A representative experiment of two to five including five mice per group is depicted. Statistical analyses were performed using the Kruskal-Wallis test. Values of p inferior to 0.05, 0.01, or 0.001 are indicated with ∗, ∗∗, or ∗∗∗, respectively.

Close modal

Thus, combining IM to IL-2 appeared to be the optimal therapeutic option to induce NK cell-mediated tumor regressions.

IKDC were found in tumor beds and were necessary and sufficient to prevent tumor outgrowth after adoptive cell transfer in Rag−/− × IL-2R−/− hosts (11, 24, 25). Moreover, we just showed that IL-15/IL-15Rα dictates IKDC effector functions (our unpublished data). Therefore, we investigated the role of IL-15 in the efficacy of IM + IL-2 against B16F10 lung metastases by comparing the numbers of tumor foci in lung parenchyma of WT vs IL-15−/− mice (devoid of IKDC). IL-15 was not only involved in the natural immunosurveillance against B16F10 establishment but also in the IM + IL-2-induced antitumor activity (Fig. 2,A). Like recombinant murine IL-15, IM + IL-2 could promote the accumulation of IKDC in the spleen by 4-fold (Fig. 2,B and our unpublished data) while the absolute numbers of B220 NK and B220+ cells remained unchanged (Fig. 2,B). Importantly, IM + IL-2 promoted the entry of cells into cycle but IKDC cell proliferation was more efficient than that of B220NK cells (Fig. 2,C). The IM + IL-2-driven IKDC proliferation was not entirely abolished in IL-15Rα−/− mice (Fig. 2 C), presumably as a result of a direct effect of IL-2 (our unpublished data). Nevertheless, these data outline the involvement of IL-15/IL-15Rα in the proliferation of IKDC during IM + IL-2 treatment.

FIGURE 2.

IL-15 is required for the proliferation of IKDC and the antitumor effects of IM + IL-2. A, Established B16F10 lung metastases model in C57BL/6 mice (WT) and IL-15 knockout (IL-15−/−) mice. Same experimental setting as in Fig. 1 C, but injections were performed in WT vs IL-15−/− mice. Statistical analyses were performed using the Kruskal-Wallis; ∗∗∗, p < 0.001. B, Naive C57BL/6 mice (WT) mice were treated with PBS and combined therapy IM + IL-2. Splenocytes were harvested at day 11 for the enumeration of IKDC (CD3CD19CD11cintB220+NK1.1+), B220 NK cells (CD3CD19B220NK1.1+), and B220+ cells. C, Naive C57BL/6 mice (WT) or IL-15Rα−/− mice were treated with PBS and combined therapy IM + IL-2 and injected with BrdU before flow cytometry analysis of splenocytes according to protocols described in Materials and Methods. Dividing cells that incorporated BrdU are measured as a percentage among the IKDC (upper panel) or B220 NK cell subsets (lower panel). ∗, p < 0.05; ND, Not detected; ns, nonsignificant.

FIGURE 2.

IL-15 is required for the proliferation of IKDC and the antitumor effects of IM + IL-2. A, Established B16F10 lung metastases model in C57BL/6 mice (WT) and IL-15 knockout (IL-15−/−) mice. Same experimental setting as in Fig. 1 C, but injections were performed in WT vs IL-15−/− mice. Statistical analyses were performed using the Kruskal-Wallis; ∗∗∗, p < 0.001. B, Naive C57BL/6 mice (WT) mice were treated with PBS and combined therapy IM + IL-2. Splenocytes were harvested at day 11 for the enumeration of IKDC (CD3CD19CD11cintB220+NK1.1+), B220 NK cells (CD3CD19B220NK1.1+), and B220+ cells. C, Naive C57BL/6 mice (WT) or IL-15Rα−/− mice were treated with PBS and combined therapy IM + IL-2 and injected with BrdU before flow cytometry analysis of splenocytes according to protocols described in Materials and Methods. Dividing cells that incorporated BrdU are measured as a percentage among the IKDC (upper panel) or B220 NK cell subsets (lower panel). ∗, p < 0.05; ND, Not detected; ns, nonsignificant.

Close modal

In conclusion, the combination of IM + IL-2 promoted the regression of established tumors in an IL-15-dependent manner.

Trans-presentation of IL-15 is a biological attribute of conventional DC (cDC) stimulated through IFN-type 1R (26, 27) and is essential for the cDC-mediated NK cell triggering during TLR activation in vivo. We addressed whether the combination of IM + IL-2 would also implicate IFN-type 1R signaling. IFN-type 1R played a major role in the tumor growth inhibition provoked through therapy combining IM + IL-2 (Fig. 3,A). Since the main source of IFN-α remains pDC (28), we addressed the role of pDC in the antitumor effects of the combination therapy. Depletion of pDC using the 120G8 mAb before tumor inoculation significantly curtailed the therapeutic efficacy of IM + IL-2 (Fig. 3 B).

FIGURE 3.

Role of type 1 IFN in IM + IL-2-mediated antitumor effects. A, Established B16F10 lung metastases model in C57BL/6 mice (WT) and type I IFN receptor-deficient mice. Same experimental setting as in Fig. 1,C. B, Depletion of C57BL/6 mice using the anti-PDCA1 Ab. Same experimental setting as in Fig. 1 C. A representative experiment of two to three including five mice per group is depicted. Statistical analyses were performed using the Kruskal-Wallis test. Values of p inferior to 0.05 or 0.01 confidence interval are indicated with ∗ or ∗∗, respectively. ns, nonsignificant.

FIGURE 3.

Role of type 1 IFN in IM + IL-2-mediated antitumor effects. A, Established B16F10 lung metastases model in C57BL/6 mice (WT) and type I IFN receptor-deficient mice. Same experimental setting as in Fig. 1,C. B, Depletion of C57BL/6 mice using the anti-PDCA1 Ab. Same experimental setting as in Fig. 1 C. A representative experiment of two to three including five mice per group is depicted. Statistical analyses were performed using the Kruskal-Wallis test. Values of p inferior to 0.05 or 0.01 confidence interval are indicated with ∗ or ∗∗, respectively. ns, nonsignificant.

Close modal

Because pDC were not found in tumor infiltrates (11), it remains to be investigated at which sites pDC may play a role and whether they could represent the source of IFN-α.

In summary, IL-15 and IFN-type 1 signaling are markedly involved in the therapeutic effectiveness of the c-kit tyrosine kinase inhibitor combined with IL-2.

Because IL-15 and type 1 IFN are major players in IM + IL-2-mediated immunosurveillance against B16F10, we investigated their role in modulating the chemokine/chemokine receptor expression profile in freshly sorted IKDC before and after stimulation by IL-15 trans-presentation (so-called “IKDC15” (our unpublished data)). In resting conditions, IKDC cells responded weakly to type 1 IFN and IL-2 for CCL2 (MCP-1), CCL3 (MIP-1α), CCL4 (MIP-1β), CCL5 (RANTES), CXCL10 (IFN-γ-inducible protein 10), and vesicular endothelial growth factor (Fig. 4,A). Following trans-stimulation with IL-15, IKDC strongly produced CCL2 (MCP-1) and factors involved in tissue remodeling and/or angiogenesis such as vesicular endothelial growth factor and CXCL1 (KC/GROα) in response to type 1 IFN or IL-2 (Fig. 4,A, right panel). Moreover, following trans-stimulation with IL-15, IKDC enhanced their transcription of the inflammatory chemokine receptors CCR2 and CCR5 (Fig. 4,B). Flow cytometry analyses confirmed the expression of CCR2 and CCR5 on the IKDC15 cell surface (Fig. 4,C and data not shown). Importantly, IL-15/IL-15Rα-driven B220 NK cells failed to express CCR2 (Fig. 4 C).

FIGURE 4.

IL-15/IL-15Rα stimulates IKDC to respond to type 1 IFN and IL-2 for CCL2 and CCR2 expression. A, Cytokine and chemokine release. Resting IKDC or IKDC15 (IL-15Rα/IL-15 expanded IKDC, as described in our unpublished data) were stimulated for 24 h with medium, type 1 IFN, or IL-2. Multiplex analyses of cytokine and chemokine release were performed in the culture supernatants. B, Quantitative RT-PCR were performed on freshly sorted IKDC and IKDC15 (at day 7 of expansion). C, FACS analyses of CCR2 expression were performed on freshly sorted IKDC or B220 NK cells or on IL-15/IL15-Rα-stimulated IKDC or NK (for 7 days). Dotted line, Isotype control; black line, anti-CCR2 Ab staining.

FIGURE 4.

IL-15/IL-15Rα stimulates IKDC to respond to type 1 IFN and IL-2 for CCL2 and CCR2 expression. A, Cytokine and chemokine release. Resting IKDC or IKDC15 (IL-15Rα/IL-15 expanded IKDC, as described in our unpublished data) were stimulated for 24 h with medium, type 1 IFN, or IL-2. Multiplex analyses of cytokine and chemokine release were performed in the culture supernatants. B, Quantitative RT-PCR were performed on freshly sorted IKDC and IKDC15 (at day 7 of expansion). C, FACS analyses of CCR2 expression were performed on freshly sorted IKDC or B220 NK cells or on IL-15/IL15-Rα-stimulated IKDC or NK (for 7 days). Dotted line, Isotype control; black line, anti-CCR2 Ab staining.

Close modal

In conclusion, IL-15 trans-presentation licensed IKDC to respond to type 1 IFN and IL-2 and greatly synergized with these stimuli (Fig. 4 A) for the production of CCL2 by IKDC. In addition, IL-15/IL-15Rα induced the up-regulation of CCR2 receptors on the surface of IKDC as a prerequisite for an autocrine-positive feedback loop for IKDC recruitment in tumor beds.

We next addressed the role of CCL2 in the accumulation of IKDC into B16F10 lung metastases. Because our in vitro data support the view that CCL2 production by IKDC is stimulated by the synergistic effects between IL-15/IL-15Rα and type 1 IFN (Fig. 4), we determined the impact of IM + IL-2 therapy on the proportion of NK1.1+ innate effectors (B220NK vs IKDC) accumulating in tumor beds in WT, IL-15−/−, and CCL2−/−. IM + IL-2 significantly increased (up to 10-fold) the numbers of IKDC in WT mice in a CCL2 and IL-15-dependent manner (Fig. 5,A, left panel). In contrast, B220NK cell numbers barely increased in a CCL2-independent fashion (Fig. 5,A, right panel). Accumulation of IKDC into lung metastases could result from IKDC recruitment and/or local proliferation. The ratio between spleen and tumor-infiltrating IKDC augmented (by up to 3-fold) during therapy in WT but not CCL2-deficient hosts, supporting the contention that both recruitment and proliferation occurred during IM + IL-2 (Fig. 5 B).

FIGURE 5.

CCL2-dependent IKDC homing into tumor tissues. B16F10 tumor cells (5 × 105) were inoculated in the tail vein at day 0 and mice were sacrificed at day 11. A, Enumeration of tumor-infiltrating IKDC and B220 NK cells during PBS or combined therapy with IM + IL-2 in C57BL/6 (WT), CCL2-deficient (CCL2−/−), and IL-15-deficient (IL-15−/−) mice after lung dissociations. B, As in A, but calculation of an IKDC ratio defined as: number of IKDC in the lung divided by number of IKDC in the spleen (no. of lung IKDC ÷ no. of spleen IKDC). These data have been obtained from seven animals per group. Statistical analyses were performed using Student’s t test, p inferior to 0.05 are indicated with ∗. ns, nonsignificant.

FIGURE 5.

CCL2-dependent IKDC homing into tumor tissues. B16F10 tumor cells (5 × 105) were inoculated in the tail vein at day 0 and mice were sacrificed at day 11. A, Enumeration of tumor-infiltrating IKDC and B220 NK cells during PBS or combined therapy with IM + IL-2 in C57BL/6 (WT), CCL2-deficient (CCL2−/−), and IL-15-deficient (IL-15−/−) mice after lung dissociations. B, As in A, but calculation of an IKDC ratio defined as: number of IKDC in the lung divided by number of IKDC in the spleen (no. of lung IKDC ÷ no. of spleen IKDC). These data have been obtained from seven animals per group. Statistical analyses were performed using Student’s t test, p inferior to 0.05 are indicated with ∗. ns, nonsignificant.

Close modal

Therefore, IL-15 and CCL2 are mandatory for proliferation and recruitment of IKDC into tumor beds during therapy with IM + IL-2.

These findings underscore the antitumoral efficacy of associating two NK cell adjuvants, one blocking the c-kit tyrosine kinase of cDC (7) and the IL-2, a major lymphokine-inducing proliferation and activation of T and NK cells (22). This study addresses the mechanisms by which IKDC, recently described as novel innate effectors involved in antitumor immunosurveillance, could accumulate in tumor beds and be associated with tumor regression (11). In this study, we unravel the key role of the IFN type 1R/IL-15/CCL2 axis in the antitumor effects mediated by IM + IL-2 and underscore the potential of IL-15 to license IKDC to proliferate, kill (our unpublished data) and respond to IFN type 1 (and IL-2) by producing CCL2 and expressing CCR2. Finally, we highlight the relevance of the IL-15-driven CCL2/CCR2 selective expression on IKDC for their preferential homing to tumor tissues.

Before the description of IKDC cells, the functional interaction between conventional DC and bona fide NK cells (29, 30) was described to play a key role in the interplay between innate and cognate immunity (31, 32). IL-15/IL-15Rα is crucial for the DC-mediated NK cell priming in vitro (26, 33) and in vivo (26). TLR ligands or agonistic anti-CD40 mAb could both induce the expression of IL-15Rα on DC in vivo (26). Moreover, IFN-α has been identified as a key inducer of IL-15/IL-15Rα (26, 27, 34). The precise mechanism by which IM + IL-2 could induce IL-15/IL-15Rα in our model system remains unclear. Our unpublished data indicated that IM alone could not mediate antitumor efficacy in TRIF−/− mice, indicating that IFN type 1 could be necessary in the immunostimulatory effects of IM. Supporting this contention, we now show that the combination of IM + IL-2 required IFN type IR signaling (Fig. 3,A). A previous work reported that IKDC derived from BALB/c mice and stimulated through TLR9 ligation could secrete IFN-α (12), but these data remain controversial (35). Since pDC constitute a main source of IFN-α (28), at least during viral infection or TLR7 and TLR9-dependent insults, we could hypothesize that pDC may also be implicated in the efficacy of IM + IL-2. Depleting pDC resulted in an abrogation of the IM +IL-2-mediated antitumor effects in vivo (Fig. 3 B).

We could postulate that IM + IL-2 might stimulate IFN-α production by pDC in vivo, leading to the up-regulation of IL-15/IL-15Rα on cDC, thereby promoting the DC/NK cell cross-talk. Alternatively, a direct pDC-NK cell interaction could occur whereby IL-2 and/or IM could play a regulatory role. Indeed, we and others reported that IL-2 could trigger TRAIL expression on both pDC and NK cells during a pDC/NK cell cross-talk in vitro (36) and that IL-2-activated NK cells could stimulate IFN-α and IL-6 production by pDC (37). Interestingly, IM could synergize with a concomitant viral stimulation to enhance IFN-α production by pDC in leukemic patients (38). It remains to be determined whether IM could increase IFN-α secretion by pDC when associated with IL-2 in our model. In these two above-mentioned hypotheses (cDC/NK or pDC/NK dialogs), bona fide NK cells could be activated either through trans-presentation of IL-15 and/or by IFN-α, accounting for the efficacy of IM + IL-2. However, several lines of evidence argue against those hypotheses. First, a significant proliferation of NK cells could not be documented during IM + IL-2 therapy (Fig. 2), neither in a tumor-free (Fig. 2) nor in tumor-bearing mice (Ref. 11 and not shown). Second, IM + IL-2 could not only prevent the establishment of B16F10 lung metastases (that are notoriously NK cell dependent) but also the s.c. expansion of the same tumor (Fig. 1), supporting the notion that effectors other than NK cells might be involved. Third, in synergy with IL-15/IL-15Rα, IFN type 1 and IL-2 could directly stimulate B220 NK cells for CCL3 (MIP-1α) secretion (data not shown). However, CCR5 was not critical for the efficacy of IM + IL-2 (our unpublished data).

To date, it remains quite complex to directly ascribe the tumoricidal activity of IM + IL-2 to IKDC. Since adaptive transfer of IM + IL-2-activated IKDC into Rag−/− × IL-2Rγ−/− mice could abrogate B16F10 growth while NK cells failed to do so (11), we proposed that IKDC could represent the most potent final effector leading to an efficient antitumor activity of IM + IL-2. Supporting this view, we demonstrate here that IKDC are significantly enriched compared with B220 NK cells in spleens of mice treated with IM + IL-2 (Fig. 2,B). Indeed, the IKDC:B220 NK cell ratio was increased by 4-fold during therapy (0.2 vs 0.05, in IM + IL-2 and PBS-treated mice, respectively, p < 0.001). Moreover, we showed the crucial role of IL-15/IL-15 Rα in the proliferation and in the CCR2/CCL2-dependent trafficking of IKDC cells into tumor beds during the therapy (Figs. 2, B and C and 5 A). Finally, in our unpublished data, we showed that IL-15Ra/IL-15 licensed IKDC (and not bona fide NK cells) with TRAIL-dependent killing, which was critical in the efficacy of IM + IL-2 (11). However, the final demonstration of the key role of IKDC in tumor immunosurveillance, as an effector or conceivably as a professional APC (12), will await further investigations and the delineation of specific markers assigned to IKDC and not to other subsets of NK cells (39).

Because our data suggested that the combination therapy with IM + IL-2 could be useful to boost the natural immunosurveillance against tumors sensitive to TRAIL-dependent apoptosis (11) and was more potent than other combination therapies (Fig. 1), we launched a phase I trial “IMAIL-2” at the Institut Gustave Roussy aimed at targeting IM-resistant or stabilized gastrointestinal sarcomas and TRAIL-sensitive cancers. It is conceivable that the monitoring of innate effectors in patients treated with high dosages of IL-2 combined with Gleevec might allow the identification of the human counterpart of mouse IKDC.

We thank W. Vainchenker, A. Mackensen, A. Caignard, and B. Azzarone for discussing our data and H. Yagita, E. Tomasello, E. Vivier, and T. Walzer for providing reagents and mice. We thank Y. Lecluse, P. Rameau, and D. Métivier for cell sorting and G. Elain, N. Brunel, and B. Besson-Lescure for technical assistance and the animal facility of the Institut Gustave Roussy under the direction of P. Gonin.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by European Union grants (ALLOSTEM, DC THERA), Association pour la Recherche sur le Cancer, and Ligue Nationale Contre le Cancer (équipes labelisées de G. Kroemer and L.Z.). G.M. was supported by the Association pour la Recherche sur le Cancer, E.U. received a fellowship from the Deutsche Forschungsgemeinschaft and from the Fondation pour la Recherche Médicale, and M.B. was supported by the Poste d’Accueil Institut National de la Santé et de la Recherche Médicale.

4

Abbreviations used in this paper: IM, imatinib mesylate; DC, dendritic cell; IKDC, IFN-producing killer DC; pDC, plasmacytoid DC; rh, recombinant human; bid, twice a day; CTX, cyclophosphamide; ODN, oligodeoxynucleotide; WT, wild type; cDC, conventional DC.

1
Lake, R. A., B. W. Robinson.
2005
. Immunotherapy and chemotherapy: a practical partnership.
Nat. Rev. Cancer
5
:
397
-405.
2
Apetoh, L., M. Obeid, A. Tesniere, F. Ghiringhelli, G. M. Fimia, M. Piacentini, G. Kroemer, L. Zitvogel.
2007
. Immunogenic chemotherapy: discovery of a critical protein through proteomic analyses of tumor cells.
Cancer Genomics Proteomics
4
:
65
-70.
3
Apetoh, L., F. Ghiringhelli, A. Tesniere, M. Obeid, C. Ortiz, A. Criollo, G. Mignot, M. C. Maiuri, E. Ullrich, P. Saulnier, et al
2007
. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy.
Nat. Med.
13
:
1050
-1059.
4
Casares, N., M. O. Pequignot, A. Tesniere, F. Ghiringhelli, S. Roux, N. Chaput, E. Schmitt, A. Hamai, S. Hervas-Stubbs, M. Obeid, et al
2005
. Caspase-dependent immunogenicity of doxorubicin-induced tumor cell death.
J. Exp. Med.
202
:
1691
-1701.
5
Nowak, A. K., R. A. Lake, B. W. Robinson.
2006
. Combined chemoimmunotherapy of solid tumours: improving vaccines?.
Adv. Drug Delivery Rev.
58
:
975
-990.
6
Obeid, M., A. Tesniere, F. Ghiringhelli, G. M. Fimia, L. Apetoh, J. L. Perfettini, M. Castedo, G. Mignot, T. Panaretakis, N. Casares, et al
2007
. Calreticulin exposure dictates the immunogenicity of cancer cell death.
Nat. Med.
13
:
54
-61.
7
Borg, C., M. Terme, J. Taieb, C. Menard, C. Flament, C. Robert, K. Maruyama, H. Wakasugi, E. Angevin, K. Thielemans, et al
2004
. Novel mode of action of c-kit tyrosine kinase inhibitors leading to NK cell-dependent antitumor effects.
J. Clin. Invest.
114
:
379
-388.
8
Smyth, M. J..
2006
. Imatinib mesylate: uncovering a fast track to adaptive immunity.
N. Engl. J. Med.
354
:
2282
-2284.
9
Mauro, M. J., M. O'Dwyer, M. C. Heinrich, B. J. Druker.
2002
. STI571: a paradigm of new agents for cancer therapeutics.
J. Clin. Oncol.
20
:
325
-334.
10
Demetri, G. D., M. von Mehren, C. D. Blanke, A. D. Van den Abbeele, B. Eisenberg, P. J. Roberts, M. C. Heinrich, D. A. Tuveson, S. Singer, M. Janicek, et al
2002
. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors.
N. Engl. J. Med.
347
:
472
-480.
11
Taieb, J., N. Chaput, C. Menard, L. Apetoh, E. Ullrich, M. Bonmort, M. Pequignot, N. Casares, M. Terme, C. Flament, et al
2006
. A novel dendritic cell subset involved in tumor immunosurveillance.
Nat. Med.
12
:
214
-219.
12
Chan, C. W., E. Crafton, H. N. Fan, J. Flook, K. Yoshimura, M. Skarica, D. Brockstedt, T. W. Dubensky, M. F. Stins, L. L. Lanier, et al
2006
. Interferon-producing killer dendritic cells provide a link between innate and adaptive immunity.
Nat. Med.
12
:
207
-213.
13
Bruhl, H., J. Cihak, J. Plachy, L. Kunz-Schughart, M. Niedermeier, A. Denzel, M. Rodriguez Gomez, Y. Talke, B. Luckow, M. Stangassinger, M. Mack.
2007
. Targeting of Gr-1+, CCR2+ monocytes in collagen-induced arthritis.
Arthritis Rheum.
56
:
2975
-2985.
14
Proietti, E., L. Bracci, S. Puzelli, T. Di Pucchio, P. Sestili, E. De Vincenzi, M. Venditti, I. Capone, I. Seif, E. De Maeyer, et al
2002
. Type I IFN as a natural adjuvant for a protective immune response: lessons from the influenza vaccine model.
J. Immunol.
169
:
375
-383.
15
Harning, R., G. C. Koo, J. Szalay.
1989
. Regulation of the metastasis of murine ocular melanoma by natural killer cells.
Invest. Ophthalmol. Visual Sci.
30
:
1909
-1915.
16
Parhar, R. S., P. K. Lala.
1987
. Amelioration of B16F10 melanoma lung metastasis in mice by a combination therapy with indomethacin and interleukin 2.
J. Exp. Med.
165
:
14
-28.
17
Krieg, A. M..
2002
. A role for Toll in autoimmunity.
Nat. Immunol.
3
:
423
-424.
18
Krieg, A. M..
2002
. CpG motifs in bacterial DNA and their immune effects.
Annu. Rev. Immunol.
20
:
709
-760.
19
Ghiringhelli, F., C. Menard, F. Martin, L. Zitvogel.
2006
. The role of regulatory T cells in the control of natural killer cells: relevance during tumor progression.
Immunol. Rev.
214
:
229
-238.
20
Ghiringhelli, F., C. Menard, M. Terme, C. Flament, J. Taieb, N. Chaput, P. E. Puig, S. Novault, B. Escudier, E. Vivier, et al
2005
. CD4+CD25+ regulatory T cells inhibit natural killer cell functions in a transforming growth factor-β-dependent manner.
J. Exp. Med.
202
:
1075
-1085.
21
Smyth, M. J., M. W. Teng, J. Swann, K. Kyparissoudis, D. I. Godfrey, Y. Hayakawa.
2006
. CD4+CD25+ T regulatory cells suppress NK cell-mediated immunotherapy of cancer.
J. Immunol.
176
:
1582
-1587.
22
Grimm, E. A., A. Mazumder, H. Z. Zhang, S. A. Rosenberg.
1982
. Lymphokine-activated killer cell phenomenon: lysis of natural killer-resistant fresh solid tumor cells by interleukin 2-activated autologous human peripheral blood lymphocytes.
J. Exp. Med.
155
:
1823
-1841.
23
Mazumder, A., S. A. Rosenberg.
1984
. Successful immunotherapy of natural killer-resistant established pulmonary melanoma metastases by the intravenous adoptive transfer of syngeneic lymphocytes activated in vitro by interleukin 2.
J. Exp. Med.
159
:
495
-507.
24
Bonmort, M., E. Ullrich, G. Mignot, B. Jacobs, N. Chaput, L. Zitvogel.
2007
. Interferon-γ is produced by another player of innate immune responses: the interferon-producing killer dendritic cell (IKDC).
Biochimie
89
:
872
-877.
25
Ullrich, E., M. Bonmort, G. Mignot, N. Chaput, J. Taieb, C. Menard, S. Viaud, T. Tursz, G. Kroemer, L. Zitvogel.
2007
. Therapy-induced tumor immunosurveillance involves IFN-producing killer dendritic cells.
Cancer Res.
67
:
851
-853.
26
Lucas, M., W. Schachterle, K. Oberle, P. Aichele, A. Diefenbach.
2007
. Dendritic cells prime natural killer cells by trans-presenting interleukin 15.
Immunity
26
:
503
-517.
27
Mattei, F., G. Schiavoni, F. Belardelli, D. F. Tough.
2001
. IL-15 is expressed by dendritic cells in response to type I IFN, double-stranded RNA, or lipopolysaccharide and promotes dendritic cell activation.
J. Immunol.
167
:
1179
-1187.
28
Colonna, M., G. Trinchieri, Y. J. Liu.
2004
. Plasmacytoid dendritic cells in immunity.
Nat. Immunol.
5
:
1219
-1226.
29
Fernandez, N. C., A. Lozier, C. Flament, P. Ricciardi-Castagnoli, D. Bellet, M. Suter, M. Perricaudet, T. Tursz, E. Maraskovsky, L. Zitvogel.
1999
. Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo.
Nat. Med.
5
:
405
-411.
30
Terme, M., E. Tomasello, K. Maruyama, F. Crepineau, N. Chaput, C. Flament, J. P. Marolleau, E. Angevin, E. F. Wagner, B. Salomon, et al
2004
. IL-4 confers NK stimulatory capacity to murine dendritic cells: a signaling pathway involving KARAP/DAP12-triggering receptor expressed on myeloid cell 2 molecules.
J. Immunol.
172
:
5957
-5966.
31
Adam, C., S. King, T. Allgeier, H. Braumuller, C. Luking, J. Mysliwietz, A. Kriegeskorte, D. H. Busch, M. Rocken, R. Mocikat.
2005
. DC-NK cell cross talk as a novel CD4+ T-cell-independent pathway for antitumor CTL induction.
Blood
106
:
338
-344.
32
Mocikat, R., H. Braumuller, A. Gumy, O. Egeter, H. Ziegler, U. Reusch, A. Bubeck, J. Louis, R. Mailhammer, G. Riethmuller, et al
2003
. Natural killer cells activated by MHC class Ilow targets prime dendritic cells to induce protective CD8 T cell responses.
Immunity
19
:
561
-569.
33
Koka, R., P. Burkett, M. Chien, S. Chai, D. L. Boone, A. Ma.
2004
. Cutting edge: murine dendritic cells require IL-15Rα to prime NK cells.
J. Immunol.
173
:
3594
-3598.
34
Jinushi, M., T. Takehara, T. Tatsumi, T. Kanto, V. Groh, T. Spies, T. Suzuki, T. Miyagi, N. Hayashi.
2003
. Autocrine/paracrine IL-15 that is required for type I IFN-mediated dendritic cell expression of MHC class I-related chain A and B is impaired in hepatitis C virus infection.
J. Immunol.
171
:
5423
-5429.
35
Vremec, D., M. O'Keeffe, H. Hochrein, M. Fuchsberger, I. Caminschi, M. Lahoud, K. Shortman.
2007
. Production of interferons by dendritic cells, plasmacytoid cells, natural killer cells, and interferon-producing killer dendritic cells.
Blood
109
:
1165
-1173.
36
Ullrich, E., C. Menard, C. Flament, M. Terme, G. Mignot, M. Bonmort, J. Plumas, L. Chaperot, N. Chaput, L. Zitvogel.
2008
. Dendritic cells and innate defense against tumor cells.
Cytokine Growth Factor Rev.
19
:
79
-92.
37
Della Chiesa, M., C. Romagnani, A. Thiel, L. Moretta, A. Moretta.
2006
. Multidirectional interactions are bridging human NK cells with plasmacytoid and monocyte-derived dendritic cells during innate immune responses.
Blood
108
:
3851
-3858.
38
Mohty, M., E. Jourdan, N. B. Mami, N. Vey, G. Damaj, D. Blaise, D. Isnardon, D. Olive, B. Gaugler.
2004
. Imatinib and plasmacytoid dendritic cell function in patients with chronic myeloid leukemia.
Blood
103
:
4666
-4668.
39
Hayakawa, Y., M. J. Smyth.
2006
. CD27 dissects mature NK cells into two subsets with distinct responsiveness and migratory capacity.
J. Immunol.
176
:
1517
-1524.