In the current study, the authors sought to identify the molecular mechanisms underlying the chemoresistance of lung cancer stem or initiation cells (cancer stem cells).
In the current study, the authors sought to identify the molecular mechanisms underlying the chemoresistance of lung cancer stem or initiation cells (cancer stem cells).
A549 lung cancer cells before and after selective enrichment of a subpopulation of cancer stem cells were treated with superoxide and traditional chemotherapeutics to determine their sensitivity or resistance to these cytotoxic agents. Apoptotic activity was measured using a variety of fluorescence-based and biochemical techniques. Specific pathways involved in the chemoresistance of cancer stem cell-enriched lung cancer cells were analyzed with Western blotting and pharmacologic targeting therapy in a xenograft model.
Lung cancer stem cells exhibited significantly decreased apoptotic response to treatment with superoxide, cisplatin, gemcitabine, or a combination of cisplatin and gemcitabine compared with control A549 cells. Apoptotic resistance was mediated through the inactivation of caspase-9 and caspase-3. Increased activation of p38MAPK, MAPKAPK2, and Hsp27 was observed in lung cancer stem cells compared with control A549 cells both before and after exposure to superoxide and chemotoxic agents. In a mouse model of lung cancer, chemotherapy-induced cells increased in the antiapoptosis pathway, and quercetin, an inhibitor of Hsp27, combined with traditional chemotherapy was effective in blocking the pathway and in the treatment of lung tumors in vivo.
The authors' data demonstrate that lung cancer stem cells have elevated levels of activated Hsp27 upon treatment with superoxide and traditional chemotherapy. When combined with chemotoxic agents, blockage of Hsp27 decreased the survival of lung cancer stem cells, which otherwise were resistant to traditional chemotherapy. Cancer 2011. © 2010 American Cancer Society.
Lung cancer has become the most frequent cause of cancer death in the world.1 The poor prognosis of lung cancer patients is due mainly to poor response, early relapse, and metastasis after treatment with chemotherapy and radiotherapy. The discovery of cancer stem or initiation cells (cancer stem cells) as cancer-initiating components in leukemia and solid tumors opens an attractive approach in treating cancer.2 Cancer stem cells represent a minor population of cancer cells and possess stem cell self-renewing and multipotent properties.3 Dysregulation of cancer stem cell self-renewal is a likely requirement for the development of cancer.4 In addition, unresponsiveness or survival of cancer stem cells after treatment may be responsible for the recurrence and resistance of cancer to modern therapies.5 Therefore, current cancer therapies may be improved by targeting the self-renewal and survival capacities of cancer stem cells.6
The underlying mechanisms of resistance of lung cancer stem cells to chemotherapy are unclear. Heat shock proteins (Hsps) comprise several different families of proteins that are induced in response to a wide variety of physiological and environmental insults.7 Hsp27 is a 27 kDa protein that regulates apoptosis through interaction with key components of apoptotic signaling pathways, in particular, those involved in caspase activation.8 Several lines of evidence support a role for Hsp27 in cancer. In general, high basal levels of Hsp27 are found in a wide range of tumor cells and tissues, including colorectal and breast cancers.9, 10 In lung cancer, Hsp27 was reported to be associated with the poor prognosis of patients.11 It is well documented that culture conditions with reduced serum in the presence of epidermal growth factor (EGF) and fibroblast growth factor 2 (FGF2) can enrich cancer stem cells.12 In the current study, we enriched a cancer stem cell population in a lung cancer cell line using this method and demonstrated that cancer stem cell-enriched lung cancer cells are more resistant to traditional chemotherapy. Increased resistance is because of constitutive activation of Hsp27 and its upstream signaling pathways, p38MAPK-MAPKAPK2. Moreover, we showed that pharmacological inactivation of the Hsp27 pathways sensitized cancer stem cells to undergo apoptosis and improved the effects of traditional chemotherapy.
The lung cancer cell line A549 was obtained from the American Type Culture Collection. Cells were grown in Dulbecco modified Eagle medium (DMEM) (Gibco, Grand Island, NY) supplemented with 10 U/mL penicillin, 10 μg/mL streptomycin, 2 mM glutamine, and 10% fetal bovine serum (Gibco), in a 37°C humidified atmosphere with 5% CO2. For enrichment, cancer stem cells were cultured in tumor sphere medium consisting of serum-free DMEM/F12 medium, N2 supplement, 5 ng/mL human recombinant basic fibroblast growth factor (bFGF), and 10 ng/mL EGF.12
The cell suspensions were labeled with Hoechst 33342 dye (Molecular Probes, Eugene, Ore; Invitrogen, Carlsbad, Calif) using the methods described by Goodell et al13 with modifications. Briefly, cells were resuspended at 1 × 106/mL in prewarmed DMEM (Invitrogen; Life Technologies, Rockville, Md) with 2% fetal calf serum (FCS) (Invitrogen, Life Technologies) and 10 mmol/L HEPES buffer (Invitrogen, Life Technologies). Hoechst 33342 dye was added at a final concentration of 5 μg/mL in the presence or absence of verapamil (50 μmol/L; Sigma, St Louis, Mo), and the cells were incubated at 37°C for 90 minutes with intermittent shaking, then centrifuged down at 4°C, and resuspended in ice-cold Hanks balanced salt solution containing 2% FCS and 10 mmol/L HEPES. Propidium iodide (Molecular Probes, Invitrogen) at a final concentration of 2 μg/mL was added to the cells to gate viable cells. The cells were filtered through a 40 μm cell strainer to obtain single cell suspension before sorting. Analyses and sorting were done on a FACSVantage SE (Becton Dickinson, Bedford, Mass). The Hoechst 33342 dye was excited at 357 nm, and its fluorescence was dual-wavelength analyzed (blue, 402-446 nm; red, 650-670 nm).
Cells were harvested and total RNA was extracted using the RNeasy Micro kit (Qiagen, Valencia, Calif). Total RNA was treated with DNase I (Invitrogen) and subsequently reverse transcribed using random hexamers and SuperScript II reverse transcriptase enzyme (Invitrogen) according to the manufacturer's instructions. Real-time polymerase chain reaction (PCR) was done with SYBR Green Real-Time Core Reagents (Applied Biosystems, Foster City, Calif) according to the manufacturer's instructions on the ABI Prism 7900 Sequence Detection System (Applied Biosystems). Primers were designed to generate a PCR product of <200 bp. Each 15 μL PCR contained 1.5 μL diluted cDNA (24 ng starting total RNA). Thermal cycling conditions were 50°C for 2 minutes and 95°C for 5 minutes followed by 40 cycles of 15 seconds at 95°C, 30 seconds at 58°C, and 30 seconds at 72°C. Levels of expression were normalized to the GAPDH (glyceraldehyde-3-phosphate dehydrogenase) housekeeping gene.
Cell extracts were prepared with M-PER (Pierce, Rockford, Ill) plus a protease inhibitor cocktail (Halt; Pierce), and protein concentrations were determined using the BCA assay (Pierce). Aliquots of protein lysates were separated on sodium dodecyl sulfate–10% polyacrylamide gels and transferred to polyvinylidene difluoride membrane filters, which were blocked with 5% blotting grade milk (Bio-Rad, Hercules, Calif) in TBST (20 mM Tris-HCl [pH 7.6], 137 mM NaCl, 1% Tween 20). Membranes were then probed with the indicated primary antibodies (Hsp27, Nanog, Oct4 [H-134], Sox2, Santa Cruz, Heidelberg, Germany; pHSP27, R&D Systems, Minneapolis, Minn; MAPKAPK2, pMAPKAPK2, p38MAPK, pp38MAPK, caspase3, caspase9, Cell Signaling Technologies, Beverly, Mass), reacted with corresponding secondary antibodies, and detected using a chemiluminescence assay (Millipore, Billerica, Mass). Membranes were exposed to x-ray film to observe the bands (Amersham Pharmacia Biotech, Piscataway, NJ).
Cell numbers were determined using the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay (Sigma). Tumor cells and enriched cancer stem cells (5 × 103 cells per well) were grown in 96-well plates, and treated with H2O2, cisplatin, gemcitabine, and quercetin at different dosages for 48 hours. After incubation, the media was replaced with 50 μL of MTT reagent (2 mg/mL) and incubated in 5% CO2 at 37°C for 2 hours. After incubation, the media were aspirated and dimethylsulfoxide (50 μL) was added to each well. The optical density of each well was measured using a microplate reader (560 nm).
Apoptosis was assayed with a TUNEL assay that detected and stained fragmented DNA (In Situ Cell Death Detection Kit, Roche, Indianapolis, Ind). In brief, cells were harvested by trypsin treatment, stained with TUNEL, and observed using a fluorescence microscope. For the detection of cleaved caspase-3 and caspase-9, protein lysates were prepared and detected by Western blot with primary antibodies purchased from Cell Signaling Technologies.
Study protocols involving mice were approved by the Institutional Animal Committee of Taipei Veterans General Hospital. Nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice were obtained from Jackson Laboratory (Bar Harbor, Me) and maintained as a colony at the National Taiwan University Animal Facility (Taipei, Taiwan) in specific pathogen-free conditions. The mice were used for experiments at 6 to 8 weeks of age. Lung cancer cells and enriched cancer stem cells were injected subcutaneously at a dose of 104, 105, and 106, respectively, per injection site. Tumor nodules usually became palpable within 1 week and reached about 2 to 7 mm in diameter 2 weeks after cell injection. The length and width of the tumors were measured, and tumor volume was calculated (tumor volume = length × width2/2). Each mouse was injected intraperitoneally with 80 mg/kg gemcitabine (Genentech, South San Francisco, Calif), 5 mg/kg cisplatin (Bristol-Myers Squibb, Seattle, Wash), or 50 mg/kg quercetin (Sigma) every 3 days for 3×.
Tumor cell xenografts comprised of A549 tumor cells and cancer stem cells were removed and submitted for hematoxylin & eosin, TTF-1, CK7, CK20, pp38MAPK, pMAPKAPK2, and pHsp27 immunohistochemical staining and Western blotting analysis. Immunohistochemical staining was performed on 4 μm-thick sections of formalin fixed, paraffin-embedded tumor tissue. After deparaffinization and rehydration, all sections were treated with microwave irradiation in 10 mM citrate buffer (pH 6.0) for 10 minutes at 121°C for antigen retrieval. To block endogenous peroxidase activity, the sections were immersed in serum for 10 minutes. Then all sections were incubated at 4°C overnight with the antibodies. Color reaction was performed using the labeled streptavidin biotin technique with 3,3′-diaminobenzidine as a chromogen (LSAB+ kit, HRP; DakoCytomation; Glostrup, Denmark). Tissues were counterstained with hematoxylin and then dehydrated.
All values are expressed as mean ± standard deviation. Independent t test was performed for comparison of data of independent samples. A P value <.05 was considered significant.
A549 tumor cells cultured in serum-free medium supplemented with EGF and FGF2 (tumor sphere medium) showed sphere formation approximately 14 days after subculture from parental A549 cells (Fig. 1A) and showed increased expression of the stem cell markers Nanog, Sox2, and Oct4 (Fig. 1B). These cells also demonstrated increased capacity to form tumors in immunodeficient mice compared with the bulk control cells (Fig. 1C). Tumors formed by these cells had the same histomorphology and expressed the same markers as primary lung tumors or tumors formed by bulk A549 cells (Fig. 1D), indicating that tumors were indeed of lung cancer origin. These data suggest cells cultured in tumor sphere medium became enriched in properties of cancer stem cells. In this study, these enriched cancer stem cells are referred to as tumor stemlike cells.
When treated with superoxide, tumor stemlike cells exhibited resistance to superoxide-induced cell death at superoxide concentrations up to 1000 mM. At the same concentration of superoxide, A549 cells showed significant cell death (Fig. 2A). Tumor stemlike cells also exhibited resistance to cell death induced by cisplatin, gemcitabine, and a combination of cisplatin and gemcitabine at concentrations that caused significant death in A549 cells (Fig. 2B). To investigate if the resistance of tumor stemlike cells to these cytotoxic agents was because of inhibition of apoptosis, various measurements for apoptosis were performed under these conditions. DNA fragmentation assayed by the TUNEL technique revealed that apoptosis induced by superoxide, or cisplatin, gemcitabine, and a combination of cisplatin and gemcitabine, treatment was significantly reduced in tumor stemlike cells compared with control A549 cells (Fig. 3A). Parallel Western blot analysis of the proapoptotic proteins caspase-9 and caspase-3 showed decreased amounts of the activated forms in tumor stemlike cells, providing confirmation of decreased apoptotic activity in these cells under cytotoxic conditions (Fig. 3B). These data suggest that the survival of tumor stemlike cells after superoxide exposure and treatment with cytotoxic agents may be mediated by inhibition of apoptosis through the inactivation of caspase-9 and caspase-3.
Evidence has shown that Hsp27 can inhibit apoptosis through inhibition of caspase-9 and caspase-3 cleavage by directly interacting with procaspase-9 and procaspase-3 molecules.14, 15 To investigate whether the activation of Hsp27 underlies the antiapoptotic effects exhibited by lung tumor stemlike cells upon treatment with superoxide and cytotoxic agents, activation levels of Hsp27 were compared between tumor stemlike cells and A549 cells. Interestingly, increased phosphorylation of Hsp27 at residue Ser78 was observed in tumor stemlike cells compared with control A549 cells both before and after exposure to superoxide and cytotoxic agents (Fig. 4A). Upon stimulation by stress, p38MAPK is phosphorylated, which then phosphorylates MAPKAPK2 to phosphorylate Hsp27.16 To investigate whether the same molecular mechanisms are involved in activation of Hsp27 in lung cancer tumor stemlike cells, activation levels of these signaling molecules were compared between tumor stemlike cells and A549 cells. Similar to Hsp27 activation, increased activation of p38MAPK and MAPKAPK2 was observed in tumor stemlike cells compared with control A549 cells both before and after exposure to superoxide and cytotoxic agents (Fig. 4B). These data suggest that the p38MAPK-MAPKAPK2-Hsp27 stress pathways are constitutively active in tumor stemlike cells.
To examine the involvement of these pathways in the antiapoptotic effects of tumor stemlike cells upon stimulation with superoxide or cytotoxic agents, quercetin, a specific inhibitor of Hsp27, KRIBB3, an inhibitor of Hsp27 phosphorylation, and SB203580, a selective inhibitor of p38MAPK were coincubated with superoxide or cytotoxic agents. In addition, knockdown experiments to silence the expression of Hsp27 and p38MAPK using siRNAs were also performed. Interestingly, quercetin (Fig. 5A), KRIBB3 (Fig. 5B), siRNA against Hsp27 (Fig. 5C), SB203580 (Fig. 5D), and siRNA against p38MAPK (Fig. 5E) all increased cleavage of caspase-9 and caspase-3 and subsequent apoptosis upon treatment with the proapoptotic agents. Together, these data suggest that constitutive activation of p38MAPK-MAPKAPK2-Hsp27 plays an antiapoptotic role in lung tumor stemlike cells. Because side population cells in different human lung cancer cell lines have been shown to have the properties of cancer stem cells,17 we examined whether these cells are resistant to chemotherapy via the Hsp27-associated pathway. However, we found that A549 side population cells did not show heightened resistance to traditional chemotherapy, although side population cells did express more ABCG2 transporter than nonside population cells as both mRNA and protein. Moreover, the addition of quercetin did not change the response of A549 to traditional chemotherapy (Fig. 6).
Because the Hsp27 inhibitor quercetin has been studied in clinical trials,18 we decided to test the effects of this compound on an in vivo system. First, the MTT assay demonstrated that treatment with quercetin combined with cisplatin and gemcitabine is more efficacious than cisplatin, gemcitabine, and a combination of cisplatin and gemcitabine in killing tumor stemlike cells (Fig. 7A). To transfer these findings into an in vivo system, we developed a mouse xenograft model of human lung cancer by subcutaneously transplanting A549 lung cancer cells in NOD-SCID mice. Once tumors became palpable (around 0.2 to 0.7 cm in diameter), cytotoxic therapy with cisplatin, gemcitabine, and quercetin was administered. The volume of xenograft was measured and compared between different groups up to 28 days. The cytotoxic treatment of cisplatin or gemcitabine marginally inhibited tumor progression, and the combination of cisplatin and gemcitabine further inhibited tumor growth. Moreover, the tumor volume was significantly smaller in the group treated with quercetin combined with traditional chemotherapy, compared with those groups treated with traditional chemotherapy only (P < .05, Mann-Whitney test, Fig. 7B). Tumor blocks retrieved from the animals before and after treatment were submitted for immunohistochemical analysis for the expression of embryonic markers and p38MAPK-MAPKAPK2-Hsp27 pathway proteins. The expression of embryonic markers such as Oct 4, Nanog, and Sox2 increased after chemotherapy treatment. After combined treatment with quercetin, the expression of embryonic markers decreased (Fig. 8). Moreover, the p38MAPK-MAPKAPK2-Hsp27 pathway proteins increased after chemotherapy, but decreased again when quercetin was added in combination (Fig. 9A and B). Similarly, Western blotting analysis also demonstrated that p38MAPK, MAPKAPK2, and Hsp27 protein expression increased in tumors treated with cytotoxic agents, and decreased in tumors treated with cytotoxic agents and quercetin (Fig. 9C). These results suggest that quercetin is effective in the treatment of lung tumor in vivo when combined with traditional chemotherapy for lung cancer.
Cancer stem cells are thought to play multiple roles in tumorigenicity. Previous studies have demonstrated that the side population cells in human lung cancer cell lines are an enriched source of lung tumor-initiating cells with stem cell properties.17 Recently, we demonstrated that Oct4 expression plays a critical role in maintaining the self-renewal, cancer stem cell-like, and chemo- and radioresistance properties of lung cancer stem cells.19 Hilbe et al showed that lung cancer patients had a significant increase in CD133-positive endothelial progenitor cells, and the authors speculated that these cells may be involved in tumor vasculogenesis and tumor growth.20 On the basis of these studies, lung cancer cell lines either with CD133-positive cells or other cells with stem cell-like properties including Oct4 expression may be good models for investigation of lung cancer treatment. The current surface markers, such as CD133, used for cancer stem cell identification still lack reliability, because only 1 of 262 CD133+ cells is a true cancer stem cell,21 and CD133− cells also possess some cancer stem cell properties to initiate tumor in cancer such as colorectal cancer and lung cancer.22, 23 In the current study, we therefore chose tumor stemlike cells enriched with serum-free medium in the presence of EGF and FGF2 for investigating the antiapoptosis pathways of lung cancer stem cells. Reynolds and Weiss described the method of isolation using serum-free medium to acquire spheres in which stem cells such as neural stem cells can obtain their proliferative benefit and induce renewal capacity.24 Since the initial description of this method, it has been successfully applied in the isolation of cancer stem cells from other kinds of malignant tumors including brain tumor and lung cancer.12, 25 In this study, the A549 subpopulation used for test was derived from spheres formed 2 weeks after culture. Cells derived from these spheres increased in the expression of embryonic markers including Oct4 and Nanog compared with control A549 parental cells both in vitro and in vivo. We also showed that these sphere-derived cells can form secondary spheres. These cells, after in vivo implantation, increased in the ability to form lung tumor compared with control A549 parental cells and expressed characteristic markers of differentiated lung cancer such as CK7 and CK20, as shown in the Figure 1. In our laboratory, we had tested the subpopulation 45 days after enrichment and found that the cells maintained the features of cancer stem cells, including the expression of embryonic markers and increase in tumorigenicity. Similar to this report, Eramo et al also demonstrated that cells enriched in these conditions are tumorigenic in vivo, reproduce human tumor, and lose self-renewal and tumorigenic potential upon differentiation.25
In addition to the capacity of cancer stem cells to initiate tumor growth after in vivo transplantation, cancer stem cells also are thought to be responsible for tumor recurrence and progression after cancer treatment. This is thought to occur by the increased activation of specific molecules and signaling pathways in cancer stem cells. For example, the chemoresistant property of colorectal cancer stem cells is mediated by activation of interleukin-4, and radioresistance of brain tumors is mediated by the DNA repair mechanism in glioma cancer stem cells and by the PTEN/stat3 pathway in medulloblastoma cancer stem cells.4, 26, 27 The recurrence and progression of cancer with continuous treatment remains a major problem, and the underlying mechanisms are still unclear. As shown here, the lung cancer tumor stemlike cells used in these studies exhibit properties of cancer stem cells. Tumor stemlike cells were resistant to superoxide treatment and chemotherapy in vitro through constitutive activation of Hsp27. Superoxide treatment was used to demonstrate the resistance of tumor stemlike cells to stress such as oxidative stress compared with the parental lung cancer cells. The effect of radiation was not determined in this study. However, we had previously investigated the effect of radiation on lung tumor stemlike cells and found that knockdown of Oct4 expression in lung cancer cells with positive CD133 surface marker can effectively enhance their chemoradiosensitivities and apoptotic activities in response to irradiation and chemotherapy. On the basis of the multiple roles of Hsp27 in response to different stresses, the role of Hsp27 in the antiapoptosis effects of cancer stem cells necessitates more investigation.
In the study, we also performed side population cell analysis of A549 cells and found that drug transporter-ABCG2 expression was >2× higher in side population cells than nonside population cells, not as high as was reported in Ho et al,17 but similar to the report of Seo et al.28 However, we could not find the property of chemoresistance in A549 side population cells, and inhibition of the Hsp27 pathway did not have any effects on chemotherapy-induced apoptosis in these cells. These data suggest that tumor stemlike cells enriched by serum-free culture and side population cells are different subpopulations in A549 cells.
Heat shock proteins are overexpressed in a wide range of malignancies and are involved in tumor cell proliferation, differentiation, invasion, metastasis, and death. Hsp27 was also reported to be associated with poor prognosis in gastric, liver, and prostate cancer.29 Oba et al recently demonstrated that interferon-γ can down-regulate Hsp27 expression and enhance tumor suppression in A549 lung cancer cells.30 The studies regarding the relationship between the Hsp27-p38MAPK-MAPKAPK2 signal pathway and chemoresistance of lung cancer are limited. We showed in this study that this signal pathway of Hsp27 plays a role in chemoresistance of lung cancer. Other signal pathways such as the PTEN/PI3K/AKT pathway have been reported to be critical for prostate cancer stemlike cell maintenance.31 Bleau et al also showed that PTEN/PI3K/AKT pathway regulates the side population phenotype and ABCG2 activity in glioma tumor stemlike cells.32
Quercetin (3,3′,4”,5,7-pentahydroxy flavone), 1 of the most widely distributed bioflavonoids in the plant kingdom, is known to have antitumor effects.33, 34 In 1990, Hosokawa demonstrated that quercetin can inhibit the synthesis of Hsps in HeLa cells.33 In 1994, Wei et al reported that quercetin displayed antitumor activity by triggering apoptosis in vitro.34 Furthermore, Jakubowicz-Gil and coworkers suggested that reduction of Hsp27 expression in HeLa cells promotes quercetin-induced apoptosis.35 Until now, the anticancer effects of quercetin combined with traditional chemotherapy including cisplatin and gemcitabine in lung cancer cells was undetermined. Here we demonstrate for the first time the sustained activation of Hsp27 in lung cancer stem cells when treated with chemotherapy, suggesting that Hsp27 may serve as a potential target to eliminate lung cancer resistance to anticancer therapies. We also show for the first time that quercetin, when combined with traditional chemotherapy for lung cancer, is effective to reduce lung tumor growth in a xenograft model.
In conclusion, our data demonstrate that lung cancer stem cells with elevated Hsp27 activation are resistant to insult by superoxide and traditional chemotherapy for lung cancer. When combined with traditional chemotherapy, quercetin, an inhibitor of Hsp27, effectively decreased the survival of lung cancer stem cells, which otherwise were resistant to traditional chemotherapy. Targeting the antiapoptotic pathway of cancer stem cells may represent a new strategy for the treatment of lung cancer.
This work was assisted in part by the Division of Experimental Surgery of the Department of Surgery, Taipei Veterans General Hospital, and supported by grants V97E2-007 and V98E2-004 from Taipei Veterans General Hospital to H.-S.H.
The authors declare no conflict of interest.