Targeting breast cancer stem cells: fishing season open!

Studies describing the tumor as a hierarchically organized cell population have changed the classical oncogenesis view and propose new therapeutic strategies. Cancer stem cells (CSCs) are thought to sustain tumor initiation/maintenance, therapy resistance, and systemic metastases. Targeting this tumor cell population is crucial to achieve a true cancer cure. A large research effort is now aiming to develop drugs targeting CSCs, based either on a priori understanding of key pathways regulating CSC biology or on high-throughput screening to identify novel targets and compounds.


Background
In recent years, tumor-initiating cells -so-called cancer stem cells (CSCs) -have been characterized in multiple cancers, including breast cancer [1]. Th is component of cancer cells retains key stem cell properties, including selfrenewal (which initiates and drives tumorigenesis) and diff erentiation, albeit aberrant (which contributes to cellular heterogeneity). Moreover, CSCs are thought to be the seed for the distant metastasis responsible for poor clinical outcome [2,3]. Th e discovery of CSCs provides an explanation for why cancer may be so diffi cult to cure, and suggests new therapeutic strategies. Several studies demonstrate that breast CSCs are resistant to conventional therapeutic strategies such as radiotherapy or chemotherapy [2]. Neoadjuvant chemotherapy thus leads to an increase in breast CD24 -/CD44 + or ALDH high CSCs and tumorsphere-initiating cells [4,5]. If these cells are the tumor root, then they are the cells to be killed.
Two approaches have been developed to design the best therapeutic strategies targeting CSCs. Th e fi rst approach is based on targeting key pathways regulating CSC survival, diff erentiation, and self-renewal. Several master pathways (Hedgehog, NOTCH, and AKT/WNT/ β-catenin signaling) commonly involved in self-renewal of embryonic and adult stem cells are known to be deregulated in CSCs and to induce an expansion of this population [6]. A number of agents targeting these pathways are currently being tested preclinically, and some have entered clinical trials. Meanwhile, studies of CSCenriched populations using omics technologies are rapidly defi ning additional regulatory pathways and networks regulating CSC biology. We recently established a gene expression signature that allowed the identifi cation of CXCR1/IL-8 signaling as a key regulator pathway of breast CSC biology [7]. Utilizing a small molecule inhibitor of CXCR1, repertaxin, we were able to specifi cally target the CSC population in human breast cancer xenografts, retarding tumor growth and reducing metastasis [8].

The article
To identify novel drugs that target specifi cally CSCs, researchers from Ciliberto's group have privileged the second approach based on unbiased high-throughput screening (HTS) of small-molecule libraries on CSCenriched populations [9]. Because a tumor cell population could contain very few CSCs, HTS needs to be redesigned to specifi cally measure gene inhibition or drug eff ects on the CSC population.
In the MCF7 breast cancer cell line, the authors described a cell population staining pale toluidine blue (light cells) enriched in CSCs. Light cells presented an increase in tumorsphere-forming effi ciency and were enriched for ALDH bright cells, described to exclusively contain the CSC population [7]. When transplanted in immunodefi cient mice, light cells were highly tumorigenic compared with bulk MCF7 cells. Utilizing this experimental system, Cioce and colleagues performed a drug-screen assay. A total of 26 compounds were screened for their ability to kill specifi cally the light cells at a greater rate than the bulk MCF7 cells. Th e screening assay identifi ed four such compounds, which all interfered with NF-κB signaling [9].

Abstract
Studies describing the tumor as a hierarchically organized cell population have changed the classical oncogenesis view and propose new therapeutic strategies. Cancer stem cells (CSCs) are thought to sustain tumor initiation/maintenance, therapy resistance, and systemic metastases. Targeting this tumor cell population is crucial to achieve a true cancer cure. A large research eff ort is now aiming to develop drugs targeting CSCs, based either on a priori understanding of key pathways regulating CSC biology or on high-throughput screening to identify novel targets and compounds.

The viewpoint
Th is unbiased drug-screen strategy on a CSC-enriched population was initially developed in the laboratories of Weinberg and Lander. Th e model involved experimentally transformed HMLER breast cancer cells modifi ed by shRNA-mediated inhibition of the human E-cadherin gene. Inhibition of E-cadherin expression induced an epithelial-mesenchymal transition, resulting in an increase in CD44 high /CD24 low cancer cells. A total of 32 compounds in a library of 16,000 chemicals had selective toxicity for these artifi cially enriched breast CSCs. Among these compounds, salinomycin was the most potent. Th e use of this potassium ionophore inhibitor as a potential cancer drug is novel and was validated in vivo using breast cancer cell line xenografts, with a decrease in tumor growth and metastasis formation [10].
A similar approach has been developed for human brain tumors with the establishment of several glioma neural stem cell lines stably enriched in CSCs. Utilizing a cell imaging-based chemical screen (comprising 450 US Food and Drug Administration-approved drugs), Dirks' group identifi ed both diff erential sensitivities of CSCs and a common susceptibility to perturbation of serotonin signaling [11]. Th ese observation suggests that CSCs might be highly susceptible to metabolic changes and may open new therapeutic possibilities.
Other than testing selective drug toxicity on an enriched-CSC population compared with bulk cancer cells, HTS can be designed to directly measure the drug eff ect on CSC function. Exploiting the relationship between neural stem cell self-renewal and neurosphere proliferation, a screen of more than 1,200 compounds identifi ed several neuromodulators as key regulators of stem cell biology [12]. A similar approach may be envisaged with tumorspheres from breast cancer cell lines.
Instead of using chemicals, RNA interference libraries can be screened to identify factors that control CSC tumorigenicity and stimulate the development of novel anti-CSC therapies. A recent kinome-wide RNA interference screen identifi ed factors that control the balance between maintenance and diff erentiation of glioblastoma CSCs. For example, silencing of TRRAP was described to increase diff erentiation of glioblastoma CSCs in vitro and also suppressed tumor formation in vivo [13].
In conclusion, HTS assays of CSCs provide opportunities to identify multiple compounds that could represent new revolutionary therapies. Because these novel compounds will be selected in vitro, it is crucial to extensively validate in vivo the selective toxicity of these drugs toward CSCs, utilizing primary tumor xenografts as a preclinical step. Moreover, serial transplantation of the residual cells isolated from treated tumors will be needed to prove the complete eradication of the tumorinitiating cell population. If the fi shing season is offi cially open, the question remains how to choose the best bait. Developing therapeutic strategies to target CSCs will hence need a thorough and rigorous eff ort as many challenges remain to be overcome, such as the evaluation of drug effi ciency in cancer patients [14].