Implication of Cancer Stem Cells in Cancer Drug Development and Drug Delivery

Abstract

Conventional therapy against cancer by chemotherapy and radiation therapy has proven successful at decreasing bulk tumor size when cancer is diagnosed early enough. These therapeutic approaches, however, often result in recurrence years later. A number of studies have begun to identify a subpopulation of tumor-initiating cells that may be critical to early tumorigenesis and subsequent recurrence. These cells have been termed cancer stem cells (CSCs) and have been demonstrated to have properties of self-renewal, differentiation into other tumor cell types, and enhanced drug resistance. Given the growing body of evidence that CSCs play a key role in tumor biology in many cancer types, it is important to take into account the properties of these cells when developing novel cancer drugs and designing enhanced methods of drug delivery. This review covers some of the mechanisms by which CSCs escape conventional therapy as well as the potential approaches to targeting CSCs that may be used during cancer drug development.

Introduction

Cancer, the abnormal growth of tumor cells that can invade other tissues and disrupt normal biological processes, is thought to arise from a series of genetic alterations in normal cells that allows these cells to ignore internal and external signals that inhibit their proliferation. Approaches to the treatment of cancer have traditionally sought to enhance overall killing of tumor cells and impair the growth of the general population of tumor cells. This approach has resulted in prolonged survival rates of cancer patients in many cancer types. Intrinsic and acquired drug resistance, however, continues to contribute to tumor recurrence and often death in cancer patients. As such, new approaches to cancer drug development are needed to improve the way in which cancer patients are treated.

Determining the best approaches to developing novel therapeutics for specific cancers requires an understanding of the tumor-initiating properties of tumor cells as well as the mechanisms by which tumor cells escape current methods of treatment. Many cancers consist of a fairly heterogeneous population in which different tumor cells contain different genetic and epigenetic alterations. ¹ This heterogeneity lends itself to the proposal of two major models of cancer initiation and progression. The stochastic model suggests that although the tumor cells may appear phenotypically heterogeneous, they share key biological properties that allow a number of different cells within a tumor to have equivalent tumor-initiating capacity. In this model, the microenvironment, immune response, and internal gene regulation signals would determine which tumor cell initiates tumor growth. The hierarchical model suggests that different populations of tumor cells within a tumor population have varying tumor-initiating capabilities, with some cells being much more capable of initiating tumor growth than others. These tumor cells with enhanced tumor-initiating capability have been identified as cancer stem cells (CSCs). In addition to having the ability to initiate and drive tumor population expansion, these CSCs are identified as being capable of self-renewal and differentiation into other cell types within the tumor.^2,3 Recent work in hematological cancers and solid tumors suggests that both models may fit depending on the cancer type and the initiating oncogenic events. For example, work done in melanoma models would suggest that even phenotypically different tumor cells in melanoma maintain equivalent tumor-initiating capabilities, whereas extensive reports in the identification and characterization of CSCs in colon, breast, and hepatic cancers suggest that these solid tumors often follow the hierarchical model.^4–8

Conventional cancer therapy often results in enhanced cell death of the overall tumor population and a reduction in the bulk tumor. For cancers that follow the hierarchical model, identifying and understanding the CSC population of the tumor is important as complete eradication of cancer requires the use of therapies that can effectively target and kill this tumor-initiating population. As such, new approaches to cancer therapy will be needed to better treat cancers in which CSCs are the key tumor-initiating cell population. One approach is to specifically target biological processes that are unique and required for the maintenance of CSC viability. This approach would result in a loss of CSCs, and because of the limited proliferative potential of the rest of the tumor cells, the bulk tumor would eventually shrink. In cancers in which differentiation of tumor cells is unidirectional from CSC to the rest of the cancer cell population, this may be a viable method of treatment. Recent work in breast cancer, however, suggests that differentiation between non-CSCs and CSCs may be bidirectional in some cancers and would thus require combinatorial treatment of both CSCs as well as the bulk tumor population. ⁹ In this review, we discuss some of the mechanisms by which CSCs escape conventional therapy, as well as discuss promising CSC therapeutic targets and approaches to targeted CSC drug delivery that may be useful in drug development and delivery against cancers.

Mechanisms of Drug Resistance in CSCs

A key property of CSCs that has been identified in a number of studies is enhanced chemoresistance by CSC populations compared with non-CSC populations in tumors. This is particularly intriguing as this provides one potential explanation for the recurrence of cancer following a decrease in bulk tumor or remission. A major mechanism of resistance that has been identified in a number of tumor models is the increased expression of ABC transporter proteins, including P-glycoprotein (MDR1) and ABCG2, in CSCs compared with other populations in the tumor.^10,11 These proteins normally function to protect cells from toxins and xenobiotics.^12–14 In cancer cells, however, these proteins can confer increased resistance to chemotherapeutics. This mechanism of chemoresistance can also be used to enrich for CSC cells based on the differential efflux of Hoechst 33342 dye by CSCs compared with non-CSCs. ¹⁵ Hoechst 33342 dye-effluxing populations, termed the side population (SP), were initially identified as a method for enriching for hematopoietic stem cells. ¹⁶ First used to isolate CSCs in the C6 glioma cancer cell line, SP analysis has enriched for tumor-initiating CSCs in a number of solid tumors, including colon, hepatic, ovarian, and breast.^10,17–20 Work in solid tumor models such as hepatic tumors also suggests that this mechanism of chemoresistance is not uniformly shared among all cancer cell lines and tumor populations of similar tissue origin.^7,8 Work in our laboratory suggests that the initial driving oncogene may play a significant role in determining the expression of ABC transporter proteins and subsequent SP phenotype in CSCs. ¹¹ Another internal phenotypic method for enriching and isolating CSCs is through aldehyde dehydrogenase (ALDH) activity. Interestingly, studies with colon cancer cells suggested that the canonical CD44+ colon CSC population exhibited enhanced ALDH activity and that ALDH1 contributed to resistance to cyclophosphamide.^6,21

In addition to the expression of multidrug resistance proteins and enzymes, activation of specific signaling pathways can result in the expression of antiapoptotic genes, such as Bcl-2, that also contribute to chemoresistance. ²² Ma et al. ²³ demonstrated that CD133-positive Huh7 hepatic cancer cells represented a chemoresistant CSC population. Inhibition of AKT1 resulted in the downregulation of the antiapoptotic gene Bcl-2 and decreased chemoresistance by these CSCs. Bcl-2 and Bcl-xL have been found to be highly expressed in CSCs through other factors, including Aurora A, STAT3, and IL-4.^23–26 In addition, the DNA damage-checkpoint activation protein CHK1 has been found to be more strongly activated in CD133-positive colon CSCs compared with their CD133-negative counterparts following treatment with DNA interstrand-crosslinking chemotherapy agents such as cisplatin. ²⁷ This represents yet another mechanism by which certain CSCs could escape conventional chemotherapy treatment. Understanding these mechanisms of chemoresistance by CSCs can provide insight into improved drug development and therapeutic approaches to treating cancers that follow a hierarchical model.

Potential Approaches to Treating CSCs

Increased understanding of the biology of CSCs has led to a number of promising approaches to treating cancer while considering the importance of eradicating CSCs. One approach that has shown promise is differentiation therapy. Differentiation of CSCs into non-CSCs that have limited tumor-initiating and chemoresistant ability may be effective when combined with traditional therapy against the bulk tumor. An effective example of this approach is the treatment of chronic myeloid leukemia (CML) by interferon (IFN)–α. The most effective treatment of CML currently is with the tyrosine kinase inhibitor, imatinib. Imatinib can significantly reduce the presence of the disease but fails to fully eradicate CML. A number of studies have demonstrated that treatment of CML with IFN has a much slower response but is significantly more effective at killing the CML progenitor cells. ²⁸ Furthermore, predifferentiation of CML progenitors with IFN combined with the treatment of CML with tyrosine kinase inhibitors such as imatinib has been demonstrated in multiple studies to be a more effective method of treating CML than single-agent therapies alone.^29,30 Other studies with CSCs have provided further evidence that differentiation therapy is a viable approach to treating certain cancer. Piccirillo et al. ³¹ demonstrated that treatment of glioma tumor cells with bone morphogenetic proteins (BMPs) was effective in differentiating glioma CSCs into nonmalignant cells and impairing tumor growth in vitro and in vivo.

In addition to promoting differentiation of CSCs into nonmalignant cancer cells through the exploitation of their response to external maturation cues, another promising method of treating CSCs is through the development of drugs that target and block CSC-specific signaling pathways. One such signaling pathway is the Sonic hedgehog (Shh) signaling pathway. The Shh signaling pathway has been demonstrated to play a key role in early embryonic development as well as in the maintenance of CSCs in a number of different cancer models, including CML, medulloblastoma, and lung, breast, pancreatic, and liver cancer.^32–37 GDC-0449, a small molecule targeting this pathway, was recently developed by Genentech and was first demonstrated in human trials to be an effective antitumor therapeutic against medulloblastoma. ³⁸ Additional studies have confirmed these results as well as demonstrated encouraging results of this molecule in basal cell carcinoma. ³⁹ These studies provide evidence that developing drugs that target signaling pathways that are key to CSC self-renewal and maintenance may serve as a reasonable approach to improving cancer therapeutics. Additional signaling pathways that have also been identified in various CSC studies include Wnt/β-catenin, Notch, STAT3, PI3K/AKT, Bmi-1, and c-Met.^{10,23,40–43} Further development of therapeutics that specifically target these pathways may also prove beneficial in the treatment of cancers in which CSCs require these signaling pathways.

Enhanced Drug Delivery against CSCs

In addition to new approaches to drug development against CSCs, enhanced drug delivery may enhance the efficacy of existing drugs against CSCs as well as provide methods for combinatorial treatment to single cells. As discussed, chemoresistance by increased expression of multidrug resistance proteins is a common mechanism by which CSCs escape conventional chemotherapy treatment. Because CSCs of varying origin also differ in the multidrug resistance proteins expressed, a nonspecific, nontoxic method for overcoming their activity may be an effective method for enhancing the effect of conventional chemotherapy against CSCs. Nanodiamonds (NDs) represent one such potential drug delivery platform against CSCs. At 2 to 5 nm in size, NDs are nontoxic nanoparticles that can be conjugated with a number of therapeutic and targeting agents, including small molecules, proteins, and genetic material.^44–47 We had previously demonstrated that delivery of doxorubicin (Dox) by NDs enhances Dox efficacy in Dox-resistant tumor models. ⁴⁴ In fact, in MDR1-overexpressing cell lines, Dox-conjugated NDs (NDX) were effective at reducing efflux of Dox by MDR1 and enhancing retention of Dox in the cells. Similar work has also been performed with increased efficacy over conventional therapeutics with other carbon nanoparticles, including single-walled carbon nanotubes (SWNTs). ⁴⁸ Because of the versatile capabilities of NDs and other carbon nanoparticles, further refinement of NDs as a drug delivery platform against CSCs is possible with the introduction of targeting and combinatorial drug delivery options.

The most well-developed nanoparticle drug delivery platform against cancer and cancer stem cells is liposomal delivery. Liu et al ⁴⁹ demonstrated that combinatorial delivery of liposomal encapsulated therapeutics can enhance the efficacy of these drugs against putative CSCs in established human breast cancer models. Liposomes have also demonstrated enhanced drug delivery through the addition of targeting components.^50–53 One target of particular interest is CD44, which has been shown to be highly expressed by CSCs from the colon and breast.^5,54 Eliaz et al. ⁵⁵ demonstrated that the addition of the CD44 ligand, hyaluronic acid (HA), to the surface of liposomes can enhance delivery and efficacy of Dox to CD44-overexpressing tumor cells. Other additional nanoparticle drug delivery methods that have shown promise for targeted drug delivery against cancer also include nanoparticle-aptamer bioconjugates in which RNA aptamers target the extracellular domain of the prostate-specific antigen (PSMA), an antigen that is commonly expressed in prostate cancer cells.^56,57 Similar applications with nanoparticle-aptamer conjugates can be applied to targeting CSCs based on unique receptors that are highly expressed in CSCs and currently used to isolate them. In addition, peptides can also be attached to nanoparticle drug carriers to deliver drugs to specific cancer cells. ⁵⁸

Drug resistance contributes to treatment failure in 90% of metastatic cancers. ⁵⁹ As such, understanding how CSCs contribute to this drug resistance as well as how CSCs initiate and maintain long-term tumor growth will provide insight into novel methods for targeting and treating cancers. Although there have been promising advances in this area, including the development of GDC-0049, more work remains. Of particular interest will be the use of nanotechnology-based drug delivery platforms as a method for targeting and treating cancer. A number of platforms have demonstrated promise in model systems and initial clinical trials. Increased use of nanotechnology-based drug delivery platforms, paired with drug development that takes the role of CSCs in cancer progression and drug resistance into consideration, will likely result in improved therapeutic options for cancer patients.