Chemotherapeutic Agents in Noncytotoxic Concentrations Increase Antigen Presentation by Dendritic Cells via an IL-12-Dependent Mechanism¹

Animals

Six- to 8-wk-old male C57BL/6 (Taconic), IL-12 knockout, and control WT mice (The Jackson Laboratory) were housed in a pathogen-free facility under controlled temperature, humidity, and 12-h light/dark cycle with food and water available ad libitum. Mice were acclimatized for at least 2 wk before the experiments.

Chemotherapeutic agents, Abs, and cytokines

Chemotherapeutic agents vinblastine, vincristine, paclitaxel, methotrexate, mitomycin C, doxorubicin, and bleomycin were from Calbiochem. FITC-labeled anti-mouse I-A^b, CD86, CD80, CD40, PE-labeled anti-mouse CD11c, FITC- or PE-labeled isotype-matched controls, and PE-labeled anti-mouse IL-12 (clone C15.6, p40/p70) mAbs were from BD Pharmingen. Monensin, an inhibitor of intracellular protein transport, was from eBioscience. Abs to APM components including anti-low molecular mass polypeptide (LMP)7, -LMP10, -delta, -MB1, -tapasin, -TAP, and -calnexin were generated and characterized as described previously (37) and were provided by Dr. S. Ferrone (University of Pittsburgh Cancer Institute, Pittsburgh, PA). The recombinant mouse IL-4 and GM-CSF were from PeproTech.

DC cultures and cell analysis

Bone marrow cells were isolated from the femur and tibia, depleted of erythrocytes, CD4⁺, CD8⁺, and B220⁺ cells, and cultured overnight in a complete RPMI 1640 medium supplemented with 10% FBS, 100 U/100 µg/ml penicillin/streptomycin, 1 mM sodium pyruvate, 2 mM l-glutamine, and 0.1 mM nonessential amino acids (Invitrogen) as described earlier (37). The nonadherent cells were collected and seeded in complete medium containing 1000 U/ml GM-CSF and IL-4 for 6 days. The chemotherapeutics were added to DC cultures on day 1 or 5 and on day 6 DCs were harvested and analyzed as stated below. The low noncytotoxic concentrations of cytotoxic agents that do not induce apoptosis of tumor cell lines and bone marrow-derived DCs were selected as described earlier (32) and are shown in Tables I and II.

Ag presentation assay

Presentation of OVA Ags by DCs was assessed as described (38). Briefly, control DCs and DCs treated with chemotherapeutic agents (0.5 × 10⁶ cells/ml) were loaded with dialyzed OVA (1 mg/ml; Sigma-Aldrich) overnight. Then, DCs were washed and mixed (1 × 10⁴ cells) with syngeneic B3Z86/90.14 (B3Z) OVA-specific T cells (2 × 10⁴) in triplicate in 96-well plates. After 48 h, cell-free culture supernatants were assessed for IL-2 production by ELISA (Endogen) according to the manufacturer’s instructions. B3Z cells obtained from B3Z86/90.14 transgenic mice (H-2K^b, I-A^b) were provided by Dr. B. A. Osborne (University of Massachusetts, Amherst, MA).

Flow cytometry

For the phenotypic analysis, DCs were incubated with Abs for 30 min and washed in PBS containing 0.1% BSA and 0.1% sodium azide. For the intracellular staining of APM protein expression, DCs were washed in PBS with 0.5% BSA, fixed in 2% paraformaldehyde, and permeabilized in PBS containing 1% BSA and 0.1% saponin (37). Then, DCs were stained with Abs to individual APM components followed by incubation with FITC-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories). Murine IgG was used as a control. To examine intracellular IL-12p70 expression, DCs were treated with 2 µM monensin for 5 h, washed, fixed in 2% paraformaldehyde, and permeabilized with 0.1% saponin (Sigma-Aldrich) in PBS with 0.5% BSA. Cells were then stained with 0.06 µg of PE-conjugated rat anti-mouse IL-12 Ab. Cells were analyzed on FACScan flow cytometer using the CellQuest software (BD Biosciences). The results are expressed as the percentage of positive cells or the mean fluorescence intensity (MFI).

Data analysis

Statistical analysis of experimental data was performed with software package SigmaStat (STSS). For all analyses, the level of significance was set at probability of p < 0.05. ANOVA was used for comparison of multiple groups. For single comparison of two groups, the Student t test was used after evaluation of normality. Data are presented as the means ± SEM.

Regulation of Ag-presenting function of DCs by low concentrations of chemotherapeutic agents

To determine how chemotherapeutic agents alter Ag processing and presentation by DCs, we have utilized OVA as a model Ag given that antigenic epitopes of this Ag have been well characterized. Concentrations of tested drugs were selected based on our previous data and were 10- to 15-fold lower than the lowest concentrations that induced apoptosis of DCs and tumor cells in vitro (32). From 13 initially selected antineoplastic chemotherapeutic agents, only 7 were chosen for these studies as demonstrating a well-displayed dose-dependent inhibition of DC viability in the annexin V/propidium iodide assay (32). DC cultures were treated with antineoplastic drugs on day 1, harvested on day 6, loaded with OVA, and mixed with B3Z CD8⁺ T cell clone specific for the H-2K^b-restricted SIINFEKL peptide of OVA. Production of IL-2, which reflects T cell activation upon recognition of the OVA epitope 257–264 in the context of the H-2K^b molecule, is shown in Fig. 1. As this figure depicts, vinblastine (0.2 nM), vincristine (1 nM), and methotrexate (5 nM) up-regulated OVA Ag presentation by DCs to Ag-specific T cells by up to 150% (p < 0.01, ANOVA). Paclitaxel (5 nM), mitomycin C (50 nM), and doxorubicin (10 nM) had even stronger effects and increased OVA presentation from 648.0 ± 45.5 pg/ml in control DCs to 1198.4 ± 123.0, 1184.6 ± 56.6, and 1698.4 ± 32.8 pg/ml in treated DCs, respectively (p < 0.01). Bleomycin had no effect on DCs in this assay.

Treatment of DCs with chemotherapeutic drugs in low, noncytotoxic concentrations added to DC cultures on day 5 resulted in similar alterations of Ag-presenting function of DCs, although the magnitude of the effect was slightly smaller (data not shown). Taken together, these data demonstrate that selected chemotherapeutic drugs in low, noncytotoxic concentrations are able to increase Ag presentation by DCs. Because up-regulation of APC function of DCs is usually associated with cell maturation, the results in Fig. 1 raise the question whether the up-regulation of APC function of DCs by chemotherapeutic drugs correlates with the developmentally regulated expression of MHC class I Ag-processing machinery components.

Effect of chemotherapeutic agents on Ag-processing machinery in DCs

Next, we determined if chemotherapeutics alter Ag processing in DCs. Using Abs specific for APM components that have been tested in our laboratory (39, 40), we demonstrated that chemotherapeutic agents could up-regulate expression of key APM proteins in DCs in vitro (Table I), which has been shown to correlate with peptide-MHC class I complex expression and Ag presentation to T cells (37). For instance, methotrexate (5 nM) was a strong up-regulator of immunoproteasome, constitutive proteasome, and chaperone proteins in DCs (p < 0.05), while bleomycin (5 nM) had no significant effect on APM in DCs with only a slight down-regulation of calnexin expression. Mitomycin C (50 nM), doxorubicin (10 nM), paclitaxel (5 nM), and vinblastine (0.2 nM) increased expression of several APM proteins in DCs, while vincristine (1 nM) elevated only LMP10. The results in Table I were calculated on the basis of MFI values and are expressed as the percentage of MFI alterations in comparison to MFI in untreated DCs (n = 3). Similar results were obtained when the percentage of positive cells was assessed by flow cytometry. For instance, Fig. 2 shows the data from a representative experiment focusing on TAP binding protein (tapasin), which mediates the interaction between newly assembled MHC class I molecules and TAP. As shown, methotrexate and doxorubicin increased MFI values of tapasin expression in DCs by 3- and 2-fold, respectively. The percentage of positive cells also increased (Fig. 2). In contrast, bleomycin displayed a slight down-regulation of APM protein expression in DCs in vitro. Thus, chemotherapeutic agents, including methotrexate, paclitaxel, and doxorubicin, in low, noncytotoxic concentrations directly up-regulate expression of Ag-processing machinery components in DCs, which may represent the first mechanism (i.e., signal 1) of stimulation of APC function of DCs by these agents.

Regulation of DC maturation/activation by chemotherapeutic agents

Next, we tested how chemotherapeutic drugs in low, noncytotoxic concentrations affect DC maturation and expression of costimulatory molecules, that is, signal 2 for T cell activation. The results were expressed as the percentage of CD80⁺, CD86⁺, CD40⁺, and MHC class II⁺ cells within the CD11c⁺ DC population. Table II summarizes the results of three independent experiments and shows that paclitaxel (5 and 1 nM), methotrexate (5 and 1 nM), and doxorubicin (10 and 2 nM) significantly up-regulated expression of CD80, CD86, CD40, and MHC class II molecules on immature DCs, while mitomycin C (50 nM) significantly increased CD80 and CD40 expression (p < 0.05 vs control nontreated DCs). Vinblastine, vincristine, and bleomycin did not alter expression of CD80, CD86, CD40, and MHC class II molecules in tested concentrations (Table II). A representative example of actual flow cytometry data is shown in Fig. 3 and demonstrates that methotrexate (5 nM) shifted phenotypically immature DCs to a more mature phenotype; paclitaxel (5 nM) doubled the percentage of CD11c⁺CD86⁺ cells and increased the percentage of CD11c⁺MHC class II⁺ cells from 16% in control to 24% in treated group. Vincristine (1 nM) had no significant effect on DC phenotype. Analysis of MFI data confirmed these results and also demonstrated that paclitaxel, methotrexate, and doxorubicin increased expression of MHC II and costimulatory molecules on DCs. For instance, in the experiment shown in Fig. 3, MFI values for MHC class II expression on nontreated DCs was 219.8; on methotrexate-, paclitaxel-, and vincristine-treated DCs MFI values were 276.9, 320.3, and 224.3, respectively. These data were reproduced in three independent studies and shown in Table II. The data in Table II also demonstrate that paclitaxel, methotrexate, and doxorubicin shifted DCs to more mature phenotypes. For example, analysis of expression of costimulatory molecules on the total CD11c⁺ population assessed that in the presence of methotrexate, the expression of CD40 and MHC class II on DCs increased from 8.6% to 19.9% and from 47.2% to 78.8%, respectively (p < 0.05, Table II). Importantly, if chemotherapeutic agents were added to DC cultures on day 5, the results were similar, although to a lesser extent. Therefore, these data demonstrate that paclitaxel, methotrexate, and doxorubicin in low concentrations up-regulate the expression of costimulatory molecules on DCs and support DC maturation. Increase of costimulatory molecules on DCs (i.e., signal 2 for T cell activation) might represent the second mechanism of augmentation of Ag-presenting capacity of DCs by chemotherapeutic agents used in low, nontoxic concentrations.

Chemotherapeutic agents in low concentrations increase IL-12 production by DCs

The strength of T cell activation is determined by the level of MHC-peptide complexes on APCs (signal 1), the expression of costimulatory molecules (signal 2), and a third signal delivered by IL-12, which is essential for the clonal expansion, effector function, and memory phenotype of T cells (34, 41, 42). Consequently, to explain the up-regulation of APC function of DCs by chemotherapeutics, we examined whether IL-12 expression could be modified in DCs by tested antineoplastic drugs. The results revealed that paclitaxel (5 nM) was the strongest inducer of IL-12 expression in DCs (p < 0.001, ANOVA). Vinblastine (0.2 nM), methotrexate (5 nM), mitomycin C (50 nM), and doxorubicin (10 nM) increased IL-12 expression by up to 75% (p < 0.05, ANOVA). For example, spontaneous intracellular IL-12 levels in DCs corresponded to 4.2 ± 0.3, 13.9 ± 0.8, 7.4 ± 0.7, and 5.7 ± 0.5 MFI in control, paclitaxel-, methotrexate-, and doxorubicin-treated DCs, respectively (p < 0.05 for all treated cells vs control). Vincristine (1 nM) and bleomycin (1 nM) had no effect on spontaneous IL-12 production (Fig. 4). Thus, these results demonstrate that paclitaxel, vinblastine, methotrexate, mitomycin C, and doxorubicin are able to induce IL-12 expression in nonactivated DCs and raise the next question whether the increased expression of IL-12 in DCs treated with chemotherapeutics is involved in upregulation of their Ag-presenting function.

IL-12^−/− DCs treated with low concentrations of paclitaxel, methotrexate, and doxorubicin lack their ability to up-regulate OVA peptide presentation to Ag-specific T cells

To reveal the role of IL-12 in augmenting the Ag-presenting function of DCs by low concentrations of cytotoxic drugs, DC cultures were initiated from the bone marrow precursors obtained from IL-12 knockout and control wild-type mice. DCs were then treated with chemotherapeutic agents and tested for their potential to present OVA Ag as above. Fig. 5 presents the results of these studies shown as the percentage of activation or inhibition of OVA presentation by drug-treated DCs relative to the control nontreated DCs. As expected, Ag presentation in wild-type DCs was up-regulated by all tested drugs by up to ~150–250% with the exception of bleomycin, confirming the data presented in Fig. 1. For example, Ag presentation by DCs treated with 5 nM paclitaxel and 10 nM doxorubicin was increased to 184.9 ± 12.4% and 186.3 ± 15.7%, respectively (p < 0.05, ANOVA, n = 3) when compared with the control nontreated DCs from wild-type mice. However, the ability of DCs treated with paclitaxel, methotrexate, doxorubicin, and vinblastine to increase Ag presentation to Ag-specific T cells was abolished in DCs generated from IL-12 knockout mice (Fig. 5). For instance, OVA Ag presentation by paclitaxel- and doxorubicin-treated DCs was 82.5 ± 7.8% and 70.1 ± 5.6% of control IL-12^−/− DCs, respectively. This indicates that up-regulation of Ag presentation by DCs induced by paclitaxel, methotrexate, doxorubicin, and vinblastine is IL-12-dependent and most likely mediated by autocrine or paracrine mechanisms. At the same time, IL-12 knockout and wild-type DCs demonstrated similar capacity to up-regulate OVA presentation after their pretreatment with low concentrations of mitomycin C (50 nM) and vincristine (1 nM) (p < 0.05) (Fig. 5), suggesting that these agents do not utilize IL-12-mediated pathways in DCs for stimulating their APC function.

To understand how IL-12 regulates Ag presentation in DCs after chemomodulation, we also analyzed the expression of MHC and costimulatory molecules on wild-type and IL-12^−/− DCs pretreated with chemotherapeutic drugs in low concentrations. The capacity of paclitaxel and doxorubicin to up-regulate MHC, CD80, CD86, and CD40 molecules on DCs were completely or significantly abrogated in the absence of IL-12 production by DCs (data not shown). This suggests that induced expression of IL-12 in paclitaxel- and doxorubicin-treated DCs plays a principal role in stimulating DC maturation and Ag processing and, in turn, in Ag presentation.

In summary, our results revealed that several chemotherapeutic agents are able to up-regulate DC maturation and Ag presentation when used in low nontoxic concentrations. These effects were mediated in the autocrine or paracrine manner by DC-derived IL-12, which was induced by pretreatment with paclitaxel, methotrexate, doxorubicin, and vinblastine, but not mytomycin C and vincristine.