Radiotherapy Targeting Cancer Stem Cells: Current Views and Future Perspectives

Mechanisms of Radiation Resistance in CSCs

CSCs are regarded as virtually resistant to radiation. The emerging role of CSCs in tumor response to radiotherapy urges investigation of the molecular mechanisms underlying radioresistance of these cells. The precise mechanisms of radioresistance of CSCs remain to be elucidated. However, in some studies, increased radioresistance may be mechanically associated with increased self-renewal activity, increased DNA repair capacity and reduced induction of DNA damage, such as the one by reduced reactive oxygen species (ROS). Table I shows the possible mechanisms of radioresistance of CSCs (4-20).

One of the defining characteristics of CSCs is their ability to undergo self-renewal. During the past decade, a number of developmental pathways that regulate the self-renewal of normal stem cells have been elucidated. These pathways include WNT, OCT4, B-lymphoma Mo-MLV insertion region 1 homolog (BMI1), NOTCH, SNAILl/SlUG and Sonic Hedgehog (SHH) (21). These pathways have recently been implicated in radiation resistance of CSCs cells. Woodward et al. found that WNT/beta-catenin pathways mediate radiation resistance of mouse mammary CSCs (10). Wang et al. indicated that the NOTCH pathway is critically implicated in stem cell fate determination and cancer (6). They also found that inhibition of the Notch pathway with gamma-secretase inhibitors renders glioma CSCs more sensitive to radiation at clinically relevant doses. Chen et al. irradiated several lung cancer cell lines and found that after radiation treatment, the survival rate and number of cluster of differentiation (CD)133⁺ cells were significantly higher than those of CD133⁻ cells (p<0.01). In addition, the treatment effect of chemoradiotherapy for the CD133⁺ group is significantly improved with OCT4 siRNA compared with non-treated CD133⁺ cells or CD133⁺ cells treated with scrambled siRNA (5). Kurray et al. indicated that Snail/Slug is critical for cancer cells to acquire stem cell characteristics and for resisting radiotherapy-mediated cellular stress, and this may be a determining aspect of aggressive cancer metastases (8). These reports indicate that the self-renewal capacity is one of the main factors of radioresistance of CSCs.

The DNA repair capacity is considered to be responsible for the radioresistance of several tumors. In 2006, Bao et al. at first demonstrated the existence of CSCs in solid tumors. They showed that glioblastoma cells expressing CD133 are resistant to ionizing radiation because they are more efficient at inducing repair of damaged DNA than the bulk of the tumor cells (13). They also found that the mechanism of radioresistance involves the cell-cycle regulating proteins Chk1/Chk2. Yin et al. showed that ataxia telangiectasia mutated (ATM) signaling contributes to radioresistance in CSCs (14). D'Andrea et al. demonstrated that radioresistance in mesenchymal CSCs is likely due to N-methyltransferase (NNMT) overexpression, which may affect the DNA repair mechanisms (15). These reports demonstrate that the DNA repair capacity contributes to radioresistance in cancer CSCs.

Recent reports have shown that similarly to normal tissue stem cells, subsets of CSCs in some tumors have lower ROS levels and enhanced ROS defenses compared to their non-tumorigenic progeny. These factors may contribute to tumor radioresistance. Diehn et al. demonstrated that subsets of CSCs in some human and murine breast tumors have lower ROS levels than corresponding non-tumorigenic cells, and CSCs in these tumors develop less DNA damage and are preferentially spared after irradiation compared to non-tumorigenic cells (16). Blazek et al. showed that Daoy medulloblastoma cells that express CD133 are more radioresistant than CD133- cells, and that the CD133⁺ subset is enlarged by hypoxia, suggesting a low level of ROS in CSCs (22). Kim et al. indicated that increased CD13 expression reduces ROS, and promotes survival of liver CSCs via an epithelial-mesenchymal transition (EMT)-like phenomenon, which is an important step in tumor invasion and metastasis (23). These results show that reduced ROS plays a critical role in the radioresistance of CSCs.

Other factors, such as the microenvironment, autophagy and protein kinase C-delta have been shown to confer radioresistance in CSCs (18-20). Jamal et al. found that the brain microenvironment protects glioblastoma cells from radiation-induced double-strand breaks (DSBs) and facilitates their repair, a situation indicative of a relative decrease in intrinsic radiosensitivity (18). Lemonaco et al. indicated that gamma-radiation activates the process of autophagy, and the autophagy inhibitor bafilomycin A1 and silencing of ATG5 and Beclin1 sensitize CD133⁺ cells to gamma-radiation and significantly reduce the viability of the irradiated cells and their ability to form neurospheres (19). These findings imply that other factors, which are currently unknown, may underlie the potential mechanisms of radioresistance in CSCs.

Potential Biomarkers

Although the importance of the number of CSCs for local tumor control is obvious, until recently, no marker was available that could measure the stem cell density in different tumors in the clinic. Recently, several reports have indentified several cell surface markers of CSCs. These markers include CD133, CD44 and CD44⁺CD24^−/low (17, 24-28). Studies testing cellular radiosensitivity in vitro have revealed differences between marker-positive and -negative cells, with marker-positive cells generally being more radioresistant (7, 9, 10, 13, 22, 29, 30). Moreover, such biomarkers may be important as potential predictors of clinical outcomes in patients treated with radiotherapy (Table II). Currently, CD133 represents a major marker of CSCs in various types of cancer.

Tamura et al. showed that the number of CD133⁺ glioma cells is dramatically increased in recurrent high-grade glioma after high-dose irradiation with gamma knife surgery compared to those before gamma knife surgery (25). They concluded that glioma CD133⁺ glioma cells can survive, leading to recurrence, despite prolonged damage to tumor blood vessels. Murat et al. analyzed gene expression profiles of 80 patients with glioblastoma that showed resistance to concurrent chemoradiotherapy (24). They found that the HOX signature, which includes CD133, and epidermal growth factor receptor (EGFR) expression, were independent prognostic factors for survival in multivariate analysis. Saigusa et al. analyzed 50 patients with rectal cancer who had undergone preoperative chemoradiotherapy followed by surgery. They found that patients with tumor cells with CD133 expression on both the luminal surface and in the cytoplasm showed a poorer response than patients without CD133 expression (26). Shien et al. also showed that CD133 is a significant prognostic factor for survival in patients with lung cancer treated with radiotherapy (27). These results indicate the prognostic importance of CD133 in the clinical setting.

Concerning other biomarkers, De Jong et al. analyzed 52 patients with early laryngeal cancer who were treated with radiotherapy-alone and found that CD44 expression significantly correlated with local tumor control (28). Svendsen et al. found that the expression of the stem cell marker neuron-glial-2 (NG2) in glioblastoma predicts poor survival and resistance to ionizing radiation (17). These results indicate that other markers, as well as CD133, appear to be potential candidate biomarkers of CSCs.

Targeting CSCs

The relative resistance of CSCs to radiation and the increase in the number of CSCs after sub-lethal doses of radiation highlight the need to develop new treatment strategies to target this cell population (7, 31). Concerning therapeutic targets for CSCs, several reports have indicated that pathways of self-renewal, the CSC niche and several signal transduction pathways may be attractive targets for eradicating CSCs. Table III indicates candidates and new innovative therapies that target CSCs (4-6, 12, 14, 19, 32-51).

Many studies have implicated several developmental signaling pathways as positive regulators of stem cell self-renewal and proliferation. These pathways are therefore potentially important targets in CSCs (52). SHH, NOTCH and OCT4 pathways have received the most attention. BMI1 has also been shown to play a role in stem cell self-renewal and also is a target of SHH signaling (53). Wang et al. demonstrated that inhibition of the Notch pathway with gamma-secretase inhibitors renders the glioma cells more sensitive to radiation at clinically relevant doses, suggesting that inhibition of Notch signaling promises to improve the efficacy of current radiotherapy for the treatment of gliomas (6). Zeng et al. indicated that inhibition of the SHH pathway increases in vivo radiation efficacy in pre-clinical lung cancer models (32). Ponnurangam et al. found that Honoliol in combination with radiation targets NOTCH signaling pathways and inhibits colon CSCs (33). These results indicate the self-renewal pathway is an attractive target in CSCs.

Concerning DNA-repair pathways, Facchino et al. showed that BMI1 copurifies with the DNA DSB response and nonhomologous end-joining (NHEJ) repair proteins in glioblastoma cells, and BMI1 deficiency in glioblastoma cells severely impairs the DNA DSB response, resulting in increased sensitivity to radiation (12). They concluded that pharmacological inhibition of BMI1 combined with radiotherapy may provide an effective means of targeting glioblastoma stem cells. Zhang et al. indicated that pharmacological inhibition of the AKT pathway in both mammospheres and syngeneic mice bearing tumors, blocks WNT signaling and the repair of DNA damage selectively in CSCs (39). They speculated that pre-treatment with AKT inhibitors before ionizing radiation treatment may have a potential therapeutic benefit to patients. These results suggest that DNA repair pathways are attractive targets for targeting CSCs.

Several reports have indicated the efficacy of targeting other signal transduction pathways in CSCs when treated with radiation. Hsu et al. indicated that targeting STAT3 signaling in CD133⁺ non-small cell lung cancer cells with cucurbitacin I suppresses CSCs and enhances the response to chemoradiotherapy (38). Chen et al. also indicated that cucurbitacin I reduces CSC-like radioresistant properties of head and neck squamous cell carcinoma-derived CD44⁺/aldehyde dehydrogenase (ALDH)⁺ cells (40). Song et al. indicated that metformin kills and radiosensitizes cancer cells and eradicates radioresistant CSCs by activating 5’ AMP-activated protein kinase (AMPK) and suppressing mammalian target of rapamycin (mTOR) (41).

Another strategy for treatment is to target the cells that provide the supportive environment for CSCs (54). The main features of normal stem cells are the ability to self-renew and to differentiate into many cell types; these features are tightly-regulated by the microenvironment or “niche” (55). A stem cell niche is an interactive structural unit that is organized to facilitate proper cell-fate decisions. Several reports have indicated that microenvironmental niches promote the capacity of CSCs to self-renew, maintain their undifferentiated status and proliferate (56-58). Therefore, modification of the tumor microenvironment, especially the niche, appears to be a potential therapy for preventing the maintenance and growth of CSCs. In the human brain, the sub-ventricular zone (SVZ) and sub-granular layer harbor normal brain stem cells (59, 60), and these regions are believed to contain specific regions of stem cell niches. Evers et al. analyzed 55 patients with high-grade gliomas who were treated with radiotherapy (42). They found that patients whose bilateral SVZ received greater than 43 Gy had a significant improvement in progression-free survival compared to those who received less than 43 Gy (15.0 months vs. 7.2 months, p=0.028). Gupta et al. also indicated that increase of the mean dose of radiation to the ipsilateral SVZ is associated with significantly improved survival in patients with glioblastoma (43). Targeting CSCs along with their niche appears to be a promising approach for future research in combing specific drugs with radiation.

In addition to these above factors, several other factors have been identified as potential targets for CSCs. Piccirillo et al. indicated that bone morphogenetic proteins (BMP), which are soluble factors that normally induce neural precursor cells to differentiate into mature cells, can prompt the differentiation of CD133⁺ brain tumor cells, critically weakening their tumor-forming ability (4). Yawata et al. found that cancer testis antigen (CTA) genes are highly and frequently expressed in CSCs compared with differentiated cells, suggesting that CTA genes may be attractive candidates for targeted vaccine therapy against CSCs in patients with glioma (44). Lomonaco et al. indicated that induction of autophagy contributes to the radioresistance of glioma stem cells and autophagy inhibitors may be employed to increase the sensitivity of CD133⁺ glioma stem cells to radiation (19).

With technological advancements, new innovative therapies have been recently employed. Cui et al. indicated that carbon ion radiotherapy may have an advantage over photon beam therapy due to improved targeting of putative CSCs (49). Kamlah et al. indicated that irradiation with X-rays but not carbon ion irradiation results in a significant increase in blood vessel density. In addition, X-ray irradiation, but not carbon ion irradiation, increases the expression of stem cell factor and subsequently induces phosphorylation of c-KIT in endothelial cells (50). Using internal radiotherapy with copper-64-diacetyl-bis (N4-methythiosemicarbazone) (⁶⁴Cu-ATSM), Yoshii et al. found that ⁶⁴Cu-ATSM administration reduces the tumor volume as well as the percentage of CD133⁺ cells and the metastatic ability of Colo-26 tumor (51). These results indicate that new innovative treatments, such as carbon ion radiotherapy and internal radiotherapy with ⁶⁴Cu-ATSM, appear to be promising treatment modalities targeting CSCs.

Future Perspectives

Several researchers have identified the molecular mechanisms of radioresistance of CSCs. The main mechanisms of radioresistantce of CSCs are reported to be self-renewal capacity, DNA repair capacity and enhanced ROS defenses compared to non-CSCs. These reports have helped us develop a partial understanding of the molecular mechanisms responsible for radioresistance of CSCs. However, the precise mechanisms of radioresistance remain to be elucidated, and other signaling pathways, such as those involving apoptosis and mTOR, may influence the radioresistance of CSCs. The precise mechanisms of radioresistance in CSCs should be further investigated. It seems likely that the field of therapeutic resistance of CSCs will lead to the development of unique targeting agents that may sensitize these cells to radiotherapy for improved cancer care.

Concerning CSC biomarkers, several markers appear to be useful in predicting the prognosis of patients with cancer treated with radiotherapy, and they may reflect the CSC density. In particular, CD133 has emerged as a potential prognostic marker in patients treated with radiotherapy. However, limited information is available regarding CSC biomarkers that will predict clinical outcomes. Therefore, further studies are required to elucidate biomarkers that predict response to radiotherapy at each tumor site. With the establishment of CSC biomarkers, tailor-made therapy, i.e. such modifications of radiotherapy according to the status of the biomarkers, may be possible.

As described earlier, many studies have focused on pathways and conditions associated with CSCs radioresistance. Therefore, the natural tendency is to focus on targeting the same radioresistance mechanisms throughout the course of radiotherapy. Currently, a number of drugs and genetic approaches are being developed that specifically target signaling pathways such as SHH, NOTCH and WNT which are required for stem cell self-renewal and normal cell development (61, 62). Several drugs are being evaluated that are expected to target CSCs via inhibition of stem cell-related signal transduction pathways. Such drugs must be tested pre-clinically in combination with radiotherapy to evaluate their curative potential. In addition, inhibition of DNA repair pathways and abrogation of decreased ROS levels are attractive strategies for targeting CSCs. In the future, radiotherapy combined with drugs that target the pathways such as self-renewal and DNA repair may be a promising strategy for improving clinical outcomes.

As modern radiotherapy techniques allows the delivery of an non-homogenous dose with high precision over the tumor, not only CSCs, but also the niche, are of great interest in the further evolution of radiotherapy. Specific targeting of the niche by combining radiotherapy with drug treatments may be another promising approach to enhancing eradication of CSCs. Current data also suggest that hypoxia may be critical for maintaining the CSC niche (56). CSCs appear to be enriched by hypoxic conditions, which stabilize hypoxia-inducible factor (HIF)-1 in these cells (22). HIF-1 increases the production of vascular endothelial growth factor (VEGF), and has been suggested as a factor that regulates a variety of tumor radioresistance mechanisms. Therefore, HIF-1-mediated radioresistance in tumors may be intimately related to the often hypoxic CSCs, and targeting the CSCs and/or their vasculature niche may have the effect of radiosensitizing CSCs. Hence, combining radiotherapy with antiangiogenic therapies has promise in possibly mediating targeted anti-CSC effects. Recently, several strategies have been initiated to overcome these radioresistance mechanisms, including hyperbaric oxygenation, hypoxic radiosensitizers and anti-angiogenic agents (63-65). Established approaches of targeting hypoxia and the niche can help improve the efficacy of radiation.

Recently, several reports have indicated the possibility of in vivo imaging in tracking and targeting CSCs (48, 66). Tsurumi et al. indicated that non-invasive antibody-based in vivo imaging of tumor-associated CD133 is feasible and that CD133 antibody–based tumor targeting is efficient (66). Vlashi et al. indicated that constitutively reduced 26S proteasome activities, is a general feature of CSCs in glioma and breast cancer cells, and these reduced activities can easily be identified by in vivo imaging (48). In the future, more sophisticated monitoring of CSCs and optimal therapeutic targeting of CSCs should be explored.

Chemotherapy has frequently been used to enhance the effect of radiotherapy, and more recently, molecular-targeted agents, such as erlotinib, have been added to these approaches to improve outcomes (67). Combinations of radiotherapy and CSC targeting drugs in, addition to chemotherapy and/or molecular-targeted agents may improve clinical outcomes, and pre-clinical and clinical studies regarding these combined approaches are warranted. Moreover, with technological advancements, new innovative therapies, such as carbon ion radiotherapy and ⁶⁴Cu-ATSM, have been recently employed (49, 51). Regarding conventional X-ray irradiation, increases in the number of CSCs after sub-lethal doses of radiation (accelerated repopulation) are potentially clinically important, and prevention of this process may lead to improved clinical outcomes (10, 31). These new innovative treatments may prevent accelerated repopulation by targeting CSCs, and appear to be promising treatment modalities that target CSCs.

However, because CSCs possess many of the features of normal stem cells, it will be important to determine if such targeting strategies may be effective in tackling CSCs without unduly harming normal stem cells (70). Therefore, development of specific anti-CSC therapies should target molecules and pathways that are not crucial for normal stem-cell maintenance. Further elucidation of pathways that regulate CSCs will provide insight into the development of individualized therapies and novel therapeutic targets.