Imperatorin Targets MCL-1 to Sensitize CD133⁺ Lung Cancer Cells to γδ-T Cell-Mediated Cytotoxicity

CD133 is a cell surface glycoprotein that has been identified as an important molecular marker of stem-like cells. In some types of cancer, including lung cancer, the CD133⁺ subset is also referred to as cancer stem cells [1-3]. Previous studies have demonstrated that the CD133⁺ subgroup of cancer cells is highly tumorigenic and can undergo self-renewal [4, 5]. Furthermore, CD133⁺ cancer cells exhibit significant resistance to anti-tumor treatment, including chemotherapy [6], radiotherapy [7], and immunotherapy [8]. CD133⁺ cancer cells have become important targets for improving the efficiency of cancer therapy.

Human γδ T lymphocytes are a subgroup of T lymphocytes. They are innate immune cells that exhibit non-major histocompatibility complex restricted cytotoxicity against a broad range of tumor types, including lung cancer [9-11]. Efforts have been made to improve the feasibility of γδ T cell-based treatment in the clinical setting for cancer therapy [12, 13]. However, the mechanisms by which γδ T cells kill cancer cells are still unclear. Despite it having been proved that γδ T cells induce cytotoxicity against tumor cells through direct contact, it is not clear whether γδ T cells also exhibit indirect anti-cancer activity. Moreover, strategies are required to elevate the efficiency of γδ T cell-based cancer therapy.

Imperatorin, extracted from the root of Angelica dahurica, is a linear furanocoumarin compound that is used for analgesia, anti-inflammation, and anticoagulation [14-16]. Recent studies have demonstrated that imperatorin has some anti-tumor activity and it is able to inhibit cell proliferation and promote apoptosis in colon cancer and lung cancer [17, 18]. Furthermore, imperatorin is also used as an adjuvant drug to sensitize cancer cells to chemotherapy [19]. However, the effect of imperatorin on immunotherapy needs to be explored further. The aim of this study was to investigate the potential role of imperatorin in γδ T cell-mediated cytotoxicity toward CD133⁺ lung cancer.

Human NSCLC cell lines A549 and PC9 were obtained from the American Type Culture Collection (Rockville, MD, USA) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco, Invitrogen, Carlsbad, CA, USA) in an incubator with 5% CO₂ at 37 °C. CD133⁺ and CD133^- populations of the A549 and PC9 cell lines were sorted by using an anti-CD133-FITC antibody (Miltenyi Biotec, Bergisch Gladbach, Germany) with flow cytometry (FACSCALIBUR; BD Biosciences, Franklin Lakes, NJ, USA). γδ T cells were expanded ex vivo by using blood samples obtained from healthy donors (n = 10; informed consent was obtained from all participants). In brief, peripheral blood mononuclear cells (PBMCs) were separated by using the Ficoll reagent (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer’s instructions. To induce the generation of γδ T cells, PBMCs were cultured in DMEM supplemented with 10% FBS, 3 μM BrHPP (Innate Pharma, Marseille, France), and 400 IU/mL IL-2 (R&D Systems, Minneapolis, MN, USA) for 2 weeks [20]. To purify the γδ T cells, the above-mentioned cells were negatively selected by using an anti-αβ-TCR antibody (BioLegend, San Diego, CA, USA) and MACS LD depletion columns (Miltenyi Biotec) [21]. γδ T cells were washed 3 times before co-culture with A549 and PC9 cells.

Targeted CD133⁺ and CD133^- A549 and PC9 cells were co-cultured with γδ T cells on Transwell plates (0.4 μm) at different E:T ratio (γδ T effectors : targeted A549 and PC9 cells) in the presence or absence of imperatorin (10 μM). Following incubation for 12 h, the cytotoxicity of γδ T cells and imperatorin to the targeted cells was evaluated by measuring lactate dehydrogenase (LDH) in the supernatant. Lysis of the targeted cells was calculated by using the standard formula: [(LDH_experimental – LDH_spontaneous)/(LDH_maximum – LDH_spontaneous)] × 100%. Blocking assays to determine the involvement of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) in γδ T cell-mediated cytotoxicity were performed by adding TRAIL-neutralizing antibodies (R&D Systems, Minneapolis, MN, USA) into the co-culture system of γδ T cells with targeted A549 and PC9 cells.

Targeted CD133⁺ and CD133^- A549 and PC9 cells were co-cultured with γδ T cells on Transwell plates at an E:T ratio of 10: 1 in the presence or absence of imperatorin (10 μM). After incubation for 12 h, the concentrations of TRAIL in the co-culture system were determined by using a TRAIL enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems) according to the manufacturer’s protocol.

CD133⁺ A549 and PC9 cells were treated with imperatorin (10 μM). After incubation for 12 h, the amount of cell death receptor 4 (DR4 and DR5 on the cell surface was measured by using anti-DR4-PE and anti-DR5-PE antibodies (R&D Systems) with flow cytometry.

For the overexpression of myeloid cell leukemia 1 (MCL-1) in CD133⁺ A549 and PC9 cells, we constructed an MCL-1 eukaryotic expression vector by cloning the open reading frame of MCL-1 into the pcDNA3.1 plasmid. For MCL-1 knockdown, MCL-1 small interfering RNA (siRNA) was purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Subsequently, 2 µg/mL MCL-1 plasmid or 50 pmol/mL MCL-1 siRNA was transfected into the cells by using Lipofectamine 2000 (Invitrogen) before co-culture with imperatorin and γδ T cells. Empty plasmid transfected cells were used as the control.

After treatment with the MCL-1 plasmid, imperatorin, and γδ T cells, a Mitochondria/Cytosol Fraction Kit (BioVision, Milpitas, CA, USA) was used to isolate mitochondria from the cytoplasm of CD133⁺ A549 and PC9 cells. Subsequently, cytochrome c in the mitochondria and cytoplasm was evaluated by western blot analysis.

After treatment with the MCL-1 plasmid, imperatorin, and γδ T cells, CD133⁺ A549 and PC9 cells were collected and proteins were extracted with a RIPA lysis buffer (Cell Signaling Technology, Danvers, MA, USA). Equal amounts of proteins were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis followed by transfer to polyvinylidene fluoride membranes. Subsequently, the membranes were blocked with 5% non-fat milk at room temperature for 1 h and then incubated with antibodies (Cell Signaling Technology) against DR4, DR5, Bax, Bak, Bcl-xl, Bcl-2, MCL-1, cytochrome c, caspase-8, caspase-3, cox IV, and GAPDH overnight at 4 °C. Then, the membranes were washed 3 times before probing the bands with an enhanced chemiluminescence system (Thermo Fisher Scientific, Waltham, MA, USA).

After treatment, targeted CD133⁺ A549 and PC9 cells were collected. Subsequently, cell apoptosis was measured by using an Annexin V-FITC Apoptosis Detection Kit (Sigma-Aldrich), and mitochondrial membrane potential (MMP) was detected by using JC-1 staining (Molecular Probes, Eugene, OR, USA) [22] with flow cytometry.

Thirty-two immunodeficient nude BALB/c mice (4–6 weeks old) were purchased from Shanghai Super-B&K Laboratory Animal Corp., Ltd. (Shanghai, China), and 2.0 × 10⁶ CD133⁺ A549 cells were injected subcutaneously into the right armpit. Subsequently, 6.0 × 10⁷ γδ T cells were injected into tumor tissue in situ when the tumors reached a mean volume of 100 mm3. Imperatorin (50 mg/kg) was administered by intraperitoneal injection twice a week. Tumor size was measured every 3 days. Tumor volume = length × (width²)/2. At 28 days after inoculation, the mice were sacrificed and tumor tissues were harvested. The animal care and experimental protocols were approved by the Animal Care Committee of Ruijin Hospital, Shanghai Jiaotong University School of Medicine.

Data are presented as the mean ± standard deviation. Statistical analysis was performed by using oneway analysis of variance. Results were considered significant for P-values < 0.05.

To evaluate the sensitivity of CD133⁺ lung cancer cells to γδ T cells, we sorted CD133⁺ and CD133^- lung cancer cells from the A549 and PC9 cell lines. We observed that γδ T cells induced significantly more cell lysis in CD133^- A549 and PC9 cells than in CD133⁺ A549 and PC9 cells (Fig. 1A), indicating that CD133⁺ lung cancer cells exhibited resistance to γδ T cell treatment. Previous studies have demonstrated that γδ T cell-induced cytotoxicity to cancer cells is mainly dependent on the secretion of TRAIL [23]. Consistent with this report, we found that the addition of TRAIL-neutralizing antibodies significantly suppressed the cytotoxicity of γδ T cells to both CD133⁺ and CD133^- A549 and PC9 cells (Fig. 1B). These results indicated that one important mechanism by which γδ T cells induced cell lysis of lung cancer cells was dependent on γδ T cell-secreted TRAIL. However, we did not observe a significant difference of TRAIL concentration in the supernatant of the co-culture system with γδ T cells and CD133⁺ or CD133^- lung cancer cells (Fig. 1C). These results indicated that CD133⁺ lung cancer cells exhibited significantly lower sensitivity to γδ T cell-secreted TRAIL compared to their corresponding CD133^- lung cancer cells.

We added imperatorin to the co-culture system of CD133⁺ lung cancer cells and γδ T cells. Interestingly, we observed that γδ T cells induced significantly more cell lysis of CD133⁺ A549 and PC9 cells in the presence of imperatorin (Fig. 2A). The preceding results proved the importance of TRAIL in γδ T cell-induced cytotoxicity to lung cancer cells (Fig. 1B); however, we found that imperatorin did not influence the ability of γδ T cells to secrete TRAIL in the co-culture system (Fig. 2B). This suggested that imperatorin was a potential sensitizer to increase the sensitivity of CD133⁺ lung cancer cells to γδ T cell-secreted TRAIL. Thus, we treated CD133⁺ A549 and PC9 cells with imperatorin and recombinant TRAIL directly and then performed a cell lysis assay. The results showed that imperatorin co-treatment significantly enhanced the cytotoxicity of recombinant TRAIL to CD133⁺ A549 and PC9 cells (Fig. 2C). Taken together, we demonstrated that imperatorin increased the anti-tumor effect of γδ T cells to CD133⁺ lung cancer cells by sensitizing them to γδ T cell-secreted TRAIL.

Since the preceding results demonstrated that imperatorin increased the sensitivity of CD133⁺ lung cancer cells to γδ T cell-secreted TRAIL, we next investigated the underlying mechanisms. TRAIL is a ligand of DR4 and DR5 and delivers the apoptotic signal by binding with them [24]. Therefore, we detected the expression of DR4/DR5 on the cell surface of CD133⁺ A549 and PC9 cells after they were treated with imperatorin. However, we observed that imperatorin did not change the level of DR4/DR5 on the surface of CD133⁺ lung cancer cells (Fig. 3A). As Bcl-2 family proteins control the sensitivity of cancer cells to TRAIL-dependent apoptotic signaling, we next investigated the expression profile of Bcl-2 family proteins in CD133⁺ and CD133^- A549 and PC9 cells [25]. Among these proteins, we observed that the expression of MCL-1, which is an anti-apoptotic Bcl-2 family protein [26], was clearly increased in γδ T cell-resistant CD133⁺ A549 and PC9 cells (Fig. 3B). Furthermore, we found that imperatorin treatment reversed the overexpression of MCL-1 in CD133⁺ A549 and PC9 cells (Fig. 3C). These results suggested that imperatorin treatment may impair γδ T cell-resistance in CD133⁺ lung cancer cells by targeting MCL-1.

To determine whether the mechanism by which imperatorin sensitizes CD133⁺ lung cancer cells to γδ T cell treatment is dependent on the inhibition of MCL-1, we transfected the cells with an MCL-1-overexpressing plasmid before treatment with γδ T cells and imperatorin. As shown in Fig. 4A, transfection with the MCL-1 plasmid induced the overexpression of MCL-1 in CD133⁺ A549 and PC9 cells. An LDH release assay showed that CD133⁺ A549 and PC9 cells transfected with the MCL-1 plasmid exhibited significant resistance to imperatorin and γδ T cell co-treatment. However, CD133⁺ A549 and PC9 cells with MCL-1 knockdown were sensitive to γδ T cell treatment, even in the absence of imperatorin co-treatment (Fig. 4B). Furthermore, under co-treatment with imperatorin and γδ T cells, the apoptotic rate of CD133⁺ A549 and PC9 cells transfected with the MCL-1 plasmid was significantly lower than CD133⁺ A549 and PC9 cells transfected with the empty plasmid (Fig. 4C). These results demonstrated that imperatorin increased the cytotoxicity of γδ T cells to CD133⁺ lung cancer cells by targeting MCL-1.

The results of flow cytometry analysis showed that imperatorin clearly promoted the ability of γδ T cells to decrease the MMP of CD133⁺ A549 and PC9 cells. However, the overexpression of MCL-1 in these cells protected the mitochondria from γδ T cell-induced damage (Fig. 5A). Cytochrome c, an apoptotic inducer located in mitochondria, is the key regulator linking mitochondrial collapse and caspase-dependent apoptosis [27]. Our results showed that γδ T cells clearly induced the release of cytochrome c from mitochondria into the cytoplasm in the presence of imperatorin. However, the overexpression of MCL-1 reduced the release of cytochrome c in CD133⁺ A549 and PC9 cells (Fig. 5B). As for the downstream pathway, although γδ T cell and imperatorin co-treatment induced significant cleavage of caspase-9 and caspase-3, MCL-1 overexpression inhibited their activation. Taken together, these results demonstrated that imperatorin promotes γδ T cells to induce mitochondrial apoptosis in CD133⁺ lung cancer cells by targeting MCL-1.

To investigate the anti-tumor effect of imperatorin and γδ T cells on CD133⁺ lung cancer cells in vivo, we established a mouse model of lung cancer by inoculating mice with CD133⁺ A549 cells. We found that the anti-tumor effect of γδ T cells on this lung cancer model was slight; however, co-treatment with imperatorin and γδ T cells significantly suppressed tumor growth (Fig. 6A). Furthermore, the combination of γδ T cells and imperatorin efficiently limited the weight of xenografts originating from CD133⁺ A549 cells, whereas single treatment with γδ T cells did not achieve a satisfactory curative effect (Fig. 6B). In addition, consistent with the preceding results, we observed that imperatorin, but not γδ T cells, decreased the expression of MCL-1 in the xenografts (Fig. 6C). These results demonstrated that imperatorin inhibited the expression of MCL-1 and enhanced the anti-tumor effect of γδ T cells against CD133⁺ lung cancer cells in vivo.

Lung cancer is the most common malignant tumor worldwide. Mortality in lung cancer patients is high because lung cancer cells are very aggressive and thus the tumors have often metastasized on primary diagnosis [28-31]. Surgery is the most effective approach to treat lung cancer; however, for patients with metastasis, the tumors are unresectable. At this stage, chemotherapy or immunotherapy is indispensable [32, 33]. Recent studies consider that CD133⁺ lung cancer cells are stem-like cancer cells, and this group of lung cancer cells is reported to be resistant to anti-cancer treatment [34, 35].

γδ T cells comprise 1–5% of peripheral blood T cells, and they are capable of killing cancer cells directly. Furthermore, several clinical trials have shown that in vivo as well as adoptive transferred γδ T cells exhibited an anti-tumor effect in lung cancer patients [36-38]. TRAIL belongs to the TNF superfamily, which is mainly secreted by immune cells. TRAIL has been reported to induce mitochondrial apoptosis in cancer cells by binding selectively to DR4 and DR5 [39-41]. However, CD133⁺ cancer cells are resistant to TRAIL treatment, which is responsible for treatment failure [42, 43]. In the present study, our results showed that CD133⁺ lung cancer cells were resistant to γδ T cell treatment. Mechanically, we proved that CD133⁺ lung cancer cells exhibited low sensitivity to γδ T cell-secreted TRAIL. To improve the anti-tumor effect of γδ T cells on CD133⁺ lung cancer cells, adjuvant therapy is required to sensitize these lung cancer cells to γδ T cell-secreted TRAIL.

MCL-1 is a member of the anti-apoptotic Bcl-2 family and is mainly located at the mitochondrial outer membrane. It suppresses mitochondrial apoptosis by inactivating pro-apoptotic proteins such as Bax, Bak, and Noxa in cancer cells [44, 45]. Studies have reported that MCL-1 is usually overexpressed in cancers, and a high level of MCL-1 expression predicts a poor prognosis for cancer patients [46, 47]. Therefore, MCL-1 has become a potential target for improving the efficiency of cancer treatment. In this study, we found that MCL-1 was overexpressed in CD133⁺ lung cancer cells compared to CD133^- cells. Therefore, compared with CD133^- lung cancer cells, CD133⁺ cells exhibited lower sensitivity to γδ T cell-induced mitochondrial apoptosis. Imperatorin, a natural drug with low toxicity, was found to suppress the expression of MCL-1 in CD133⁺ lung cancer cells. We then proved that imperatorin sensitized CD133⁺ lung cancer cells to γδ T cell-secreted TRAIL by targeting MCL-1 Furthermore, our experimental results demonstrated that γδ T cell treatment dramatically induced the apoptosis of CD133⁺ lung cancer cells through the mitochondrial pathway in the presence of imperatorin.

Given the above, we demonstrated the value of adjuvant therapy with imperatorin in γδ T cell treatment. Despite the fact that CD133⁺ cancer cells show obvious resistance to standard treatments against cancer, the combination of γδ T cell treatment with natural drugs such as imperatorin may represent a potential strategy for reducing the occurrence of treatment failure.

This study was supported by the Guangdong Province Science and Technology Project (grant no: 2017B090901067).

The authors declare to have no conflict of interests.

C. You and Y. Yang contributed equally to this work.