Immunomodulatory drugs Revlimid^® (lenalidomide) and CC-4047 induce apoptosis of both hematological and solid tumor cells through NK cell activation

Reagents

CC-4047 and CC-5013 (Celgene Corporation, Summit, NJ, USA) were freshly dissolved in dimethyl sulfoxide (DMSO). Human recombinant Interleukin-2 protein (IL-2) was purchased from BD Biosciences (BD Biosciences, San Jose, CA, USA).

Cells

Tumor-derived cell lines K562, Raji, and PC-3 were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA), and maintained in RPMI 1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Hyclone, San Diego, CA, USA), 50 unit/ml penicillin, and 50 μg/ml streptomycin (Invitrogen). Peripheral blood was obtained from healthy donors (San Diego Blood Bank, San Diego, CA, USA). PBMC were isolated by Ficoll-Paque PLUS (GE Healthcare Bioscience, Piscataway, NJ, USA) gradient centrifugation and maintained in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 100 unit/ml of penicillin, 100 μg/ml streptomycin, and 0.25 μg/ml amphotericin (Cascade Biologics, Portland, OR, USA).

NK cell depletion and isolation

PBMC were isolated as described above and cultured in tissue culture plates for 2 h to deplete monocytes. For NK cell depletion, non-adherent PBMC were first incubated with CD56 MicroBeads (Miltenyi Biotec, Auburn, CA, USA). Then CD56-positive NK cells were depleted through magnetic separation according to the manufacturer’s protocol. For NK cell isolation, non-adherent PBMC were used to isolate NK cells using NK cell Isolation Kit II (Miltenyi Biotec).

BrdU incorporation assay

Raji cells were plated at 10,000 cells/well in 96-well plates in triplicate. Cells were then treated with CC-4047, CC-5013 and Rituximab in the presence of various doses of complements. Plates were incubated in a 37°C humidified CO₂ incubator for 72 h. Six hours before the end of the incubation period, Bromodeoxyuridine (BrdU) (Roche, Indianapolis, IN, USA) was added to the cells. Cells were then dried at 60°C for 1 h in an oven and anti-BrdU antibody was added followed by substrate and chemiluminescence intensity of the substrate was read with Wallac Victor 2 plate reader (Perkin Elmer, Atlanta, GA, USA) according to the manufacturer’s instructions.

Pre-label of tumor cells with PKH26

Tumor cells were labeled with fluorescence membrane labeling dye, PKH26, by adapting a protocol described previously [12]. Briefly, K562, Raji or PC-3 cells were washed twice with serum-free RPMI 1640 medium, and then re-suspended in 1 ml Diluent C of Red Fluorescent Cell Linker Kit (Sigma, St Louis, MO, USA). PKH26 dye stock (1 mM) was diluted to 4 μM with Diluent C and added into the cell suspension. The cell suspension was shaken gently at room temperature for 40 min. The reaction was stopped by adding one volume of heat inactivated FBS and incubating for 1 min at room temperature. One volume of complete medium was added and cells were washed three times with complete medium.

Treatment of PBMC, NK-depleted PBMC, and NK cells

PBMC, NK-depleted PBMC, or NK cells were treated either in flasks or in 96-well plates for different applications. In some studies, each group of effector cells was divided into two groups. Twenty units/ml of human recombinant IL-2 were added to one group, while the other group served as no IL-2 control. Both groups were incubated in a 37°C humidified CO₂ incubator for 72 h. PBMC were then spun down and resuspended in fresh RPMI complete medium to 2 × 10⁶ cells/ml. Each of two groups (+IL-2 and −IL-2) was further divided into several subgroups to be treated with vehicle control (DMSO) or various doses of CC-4047 and CC-5013 for 72 h prior to co-culture with tumor cells. The final concentration of DMSO in all experiments was 0.2%.

Co-culture model and FACS analysis

Co-culture models were set up in two formats to accommodate analysis by FACSAria flow cytometer (BD Biosciences) or by FACSArray bioanalyzer (BD Biosciences). For FACSAria analysis, PBMC treated in flasks were counted, mixed with PKH26 labeled tumor cells at various ratios in 1.5 ml microcentrifuge tubes, and incubated at 37°C for 3–4 h for K562 and Raji or 18 h for PC-3. The cells were subsequently labeled with Annexin V-FITC (BD Biosciences) and subject to FACSAria scan. For FACSArray bioanalyzer analysis, PBMC were treated with CC-4047 or CC-5013 in 96-well plate, then transferred to another 96-well plate where pre-labeled tumor cells were seeded. The PBMC and tumor cell ratios and incubation time for the mixture were the same as stated in previous format. After incubation, cells were labeled with Annexin V-APC (BD Biosciences) and subject to FACSArray scan. Non-labeled tumor cells, PKH26 labeled tumor cells and Annexin V-FITC/APC labeled apoptotic tumor cells (pretreated by cytotoxic agent) were included in each assay as compensation controls. Spontaneous apoptosis of tumor cells in the absence of PBMC served as background and was subtracted from all treated samples. Co-culture of tumor cells with NK-depleted PBMC or NK cells was set up following the same procedures as stated above.

Combination studies with chemodrugs using the co-culture model

PKH26-labeled Raji cells were resuspended to 1 × 10⁶ cells/ml in fresh complete medium containing 10 μg/ml Rituximab (Genentech, South San Francisco, CA, USA) and incubated for 20 min at room temperature. Cells were then washed once with complete medium and proceeded to co-culture. For combination studies with solid tumor cells, PKH26-labeled PC-3 cells were plated at 20,000 cells/well in low adherence 96-well plates (Costar, Lowell, MA, USA) and treated with Docetaxel (Save-Mart Pharmaceuticals, San Diego, CA, USA) for 6 h prior to co-culture with PBMC.

Data analysis

To determine whether the combination treatment with CC-4047 or CC-5013 and other chemodrugs showed additive, synergistic or antagonistic effects, we used the fractional product method. In the fractional product method, the effect of two independently acting agents is defined as the product of the unaffected fractions after treatment with either drug alone: fu(1, 2) = fu(1) × fu(2). This formula allows calculation of the predicted effect of co-treatment, based on the assumption that two agents do not interact or cooperate in inducing their effects. If the observed unaffected fraction, that is, the remaining live cells, is below, equal or above the calculated product fu(1, 2) after co-treatment, the two agents show synergy, additivity or antagonism, respectively [7].

CC-4047 enhances PBMC-mediated apoptosis of human leukemia cell K562 in co-culture model

We initially evaluated the direct effects of CC-4047 on human leukemia cell K562 apoptosis in a simple mono-cell culture model. No apoptosis of K562 cells was induced by CC-4047 at concentrations up to 100 μM (data not shown). These data suggest that CC-4047 may have minimal direct effects on leukemia cells or that one of the drug metabolites is the active form. Considering the potent properties of certain IMiDs^® immunomodulatory drugs shown in various immune cell models [8], a co-culture model using PBMC and K562 cells was used to study anti-cancer activities of those drugs. In this model, freshly isolated human PBMC were treated with or without IL-2 for 72 h followed by CC-4047 for 72 h and then co-cultured with K562 cells for 3–4 h. K562 tumor cells were prelabeled with a lipophilic membrane dye PKH26, using an optimal protocol that minimizes the impact of staining on tumor cell viability [12]. After co-culture, the mixture of K562 cells and PBMC was labeled with Annexin V-FITC and analyzed by FACSAria flow cytometer. Prelabeling of K562 cells with PKH26 allowed us to distinguish them from PBMC. Annexin V-FITC and PKH26 double positive cells represent tumor cells in early apoptotic state. Percentage of apoptotic K562 cells was calculated for every experiment. First, we detected that in this co-culture model, PBMC increased apoptosis of K562 cells from 4.4% basal level (no PBMC) to 22.1% (DMSO no IL-2) (Fig. 1a). PBMC pretreated with IL-2 further increased K562 apoptosis to 59.5% (DMSO + IL-2) (Fig. 1a). Next, we determined the effect of CC-4047 in this co-culture model in the presence or absence of IL-2. Importantly, treatment of PBMC with CC-4047 further induced K562 apoptosis from 22.1 to 61.9% in the no IL-2 group, and from 59.5 to 80.1% in the IL-2 group (Fig. 1a). Therefore, we conclude that this co-culture model is able to detect anti-tumor properties of CC-4047, which were not observed in the mono-cell culture model. In addition, these data show that the addition of IL-2 is not required to reveal the proapoptotic effects of CC-4047. We also analyzed the effect of CC-4047 on PBMC-mediated K562 apoptosis at three different effector-to-_target cell ratios. CC-4047 was able to enhance K562 apoptosis at all three ratios tested with or without IL-2 co-treatment (Fig. 1b). The percentage of apoptotic K562 cells correlated well with effector-to-_target cell ratio (Fig. 1b). This result suggests that the apoptotic effects observed are likely directly dependent on the amount of effector cells present.

CC-4047 and CC-5013 enhance PBMC-induced K562 apoptosis in the absence of IL-2 co-stimulation

Since CC-4047 showed profound activity in inducing K562 apoptosis, even in the absence of IL-2, we next assessed CC-5013 in this co-culture model without IL-2 co-stimulation. PBMC pretreated with various concentrations of CC-4047 and CC-5013 for 72 h were incubated with K562 at a 15:1 effector-to-_target ratio for 3 h. When compared to control PBMC cultures in DMSO, twofold increase in K562 apoptosis were detected in CC-4047 treated K562/PBMC co-culture (Fig. 2). CC-5013 appeared to enhance PBMC-mediated K562 apoptosis to a similar degree as CC-4047 (Fig. 2). Together with the minimal impact of either CC-4047 or CC-5013 on K562 apoptosis observed in mono-cell culture model, these data indicate that the K562/PBMC co-culture model provides a suitable in vitro model to reveal anti-tumor activity of CC-4047 and CC-5013.

NK cells are required but not sufficient for CC-4047 to enhance PBMC cytotoxicity against K562

Cell-mediated cytotoxicity is a major effector pathway of immune protection against intracellular pathogens or tumors, and a fundamental mechanism to maintain homeostasis of the immune system. It has been demonstrated previously that selected IMiDs^® immunomodulatory drugs seem to augment the cytotoxicity of natural killer (NK) cells against multiple myeloma cells [15]. We first evaluated the effects of CC-4047 and CC-5013 on NK cell surface markers. Treatment of PBMC with CC-4047 and CC-5013 caused a marked upregulation of the NK cell marker and adhesion molecule CD56 (NCAM1) in the CD3–CD56+ NK cell population (Fig. 3a). We also observed a twofold increase in the numbers of NKT (CD3+CD56+) cells in the samples treated with CC-4047 and CC-5013 (Fig. 3a).

Next, we determined whether NK cells are required in mediating CC-4047 activity in this K562/PBMC co-culture model. NK cells were depleted from PBMC as described in Materials and methods. Typically, 94–97% NK depletion was achieved in those experiments. NK-depleted PBMC and non-depleted, regular PBMC from the same donor were treated with IL-2 followed by incubation with CC-4047. These PBMC were tested in the co-culture model with K562 cells. While non-depleted PBMC induced K562 cell apoptosis at the similar level as shown previously, NK-depleted PBMC lost this ability (Fig. 3b). Only basal level of K562 apoptosis was detected in the NK-depleted PBMC group (Fig. 3b). Furthermore, CC-4047 was not able to rescue or augment the ability of NK-depleted PBMC to induce K562 apoptosis (Fig. 3b). These data demonstrate that the NK cell subpopulation of PBMC is required for CC-4047 activity in K562/PBMC co-culture model.

To further investigate the role of NK cells in CC-4047 function, NK cells were purified from PBMC and treated with IL-2 followed by CC-4047 as described under Materials and methods. As controls, non-depleted regular PBMC and NK-depleted PBMC from the same donor were also treated with IL-2 and CC-4047. In this co-culture assay using isolated NK cells with 88% purity, the total NK cell number was adjusted to match the total NK cell number of non-depleted PBMC samples. One micromolar of CC-4047 increased non-depleted PBMC activity against K562 more than twofold over DMSO treated PBMC (Fig. 3c). DMSO-treated purified NK cells were able to induce K562 apoptosis to a similar degree as non-depleted PBMC (Fig. 3c). CC-4047 marginally enhanced purified NK cell activity against K562 (Fig. 3c), even though CC-4047 treatment did not stimulate proliferation or apoptosis of purified NK cells (data not shown). Additionally, in all three effector-to-_target cell ratios tested, only a slight increase of NK-mediated K562 apoptosis was induced by CC-4047 treatment (Fig. 3d). Together, these data indicate that CC-4047 enhancement of NK cell cytotoxicity requires the presence of other cell population(s) within PBMC.

CC-4047 and CC-5013 show modest effect on Raji cell proliferation and no additive effect with Rituximab in Raji mono-cell culture model

CC-5013 has been shown to have anti-tumor activities in multiple hematological tumors, including NHL [16, 30, 37]. Here, we first tested the direct anti-proliferative activities of CC-4047 and CC-5013 on NHL cells by treating Raji cells with a standard of care for NHL, Rituximab, in the absence or presence of CC-4047 and CC-5013. As single agent, Rituximab significantly inhibited Raji cell proliferation in the presence of complement (Fig. 4a). CC-4047 and CC-5013 alone had a modest effect on Raji cell proliferation (Fig. 4a). In combination with Rituximab and complement, neither CC-4047 nor CC-5013 showed additive effect in inhibiting Raji cell proliferation in the mono-cell culture model (Fig. 4a).

CC-4047 and CC-5013 enhance Raji cell apoptosis mediated by PBMC, as single agent or in combination with Rituximab

Since no additive effect between CC-4047 or CC-5013 and Rituximab was observed in the Raji mono-cell culture model, we next evaluated the co-culture model as a more appropriate cellular model to assess CC-4047 and CC-5013 anti-tumor activities in NHL. PBMC and prelabeled Raji cells were co-cultured, and apoptotic cells were labeled by Annexin V and quantified by FACS analysis. We validated a 96-well plate format FACS analysis using the BD FACSArray Bioanalyzer to perform dose response studies of CC-4047 and CC-5013. CC-4047 or CC-5013 alone did not induce Raji apoptosis in the mono-cell culture model. In the co-culture model, both CC-4047 and CC-5013 enhanced human PBMC activity in inducing Raji cell apoptosis in a dose dependent manner (Fig. 4b). We further evaluated CC-4047 and CC-5013 in combination with Rituximab. Raji cells were incubated with Rituximab for 30 min prior to co-culture for 4 h with PBMC that had been treated with CC-4047 and CC-5013. As shown in Fig. 4c, Rituximab was able to trigger PBMC-mediated Raji apoptosis in the absence of CC-4047 or CC-5013. Very importantly, synergistic effects between Rituximab and CC-4047 or CC-5013 to induce Raji cell apoptosis were detected in the Raji/PBMC co-culture model (Fig. 4c). These data provide the preclinical basis for clinical application of CC-4047 and CC-5013 in combination with Rituximab to improve the outcome of NHL patients.

CC-4047 and CC-5013 enhance PBMC-induced apoptosis of PC-3 with or without Docetaxel pretreatment

There have been numerous reports on certain IMiDs^® immunomodulatory drugs anti-tumor activities against hematological tumors [1, 37]. However, very few studies have been done to investigate the effect of these drugs on solid tumor cells. Minimum impact of CC-4047 and CC-5013 on PC-3 proliferation or apoptosis was detected in mono-cell culture model (data not shown). Thus, we expanded the utility of our co-culture model into prostate cancer. In this model, PBMC and PKH26 labeled PC-3 cells were co-cultured in low-adherence 96-well plates for 18 h, and apoptotic cells were labeled by Annexin V-APC and quantified by FACS analysis using BD FACSArray Bioanalyzer. CC-4047 and CC-5013 dramatically enhanced human PBMC activity in inducing PC-3 apoptosis, in a dose dependent manner (Fig. 5a). Using this co-culture model, we also evaluated CC-4047 and CC-5013 anti-tumor activities in combination with Docetaxel, the standard of care for prostate cancer. PC-3 cells were exposed to the indicated suboptimal concentrations of Docetaxel for 6 h prior to co-culture with PBMC that had been treated with CC-4047 and CC-5013. Various doses of Docetaxel alone exerted minimum level of PC-3 apoptosis (10% at 100 nM), as shown in Fig. 5b. In the absence of PBMC, the combination of various doses of Docetaxel and CC-4047 or CC-5013 induced the same low level of PC-3 apoptosis as Docetaxel alone. In contrast, the combination of Docetaxel with CC-4047 or CC-5013-treated PBMC showed a dramatic increase in PC-3 cell apoptosis (Fig. 5b, c). In co-culture model, the combination treatment with Docetaxel and all doses of CC-4047 resulted in additive or synergistic apoptosis as determined by the fractional product method of drug interactions as described in Materials and methods (Fig. 5b). Co-administration of 10 or 100 nM Docetaxel with either 30 or 300 nM CC-5013-treated PBMC also resulted in a significant increase of apoptosis in PC-3 cells (Fig. 5c). Collectively these findings suggest, for the first time, that these two compounds, particularly CC-4047, may provide potential clinical benefits to prostate cancer patients as single agent or in combination with standard chemotherapy.