Characterization of repopulating tumor cell subpopulations in recurrent tumors and comparison to primary tumor cells

Introduction - Minimal Residual Disease (MRD) and tumor recurrence

Disseminated disease is the primary cause of mortality from solid tumors (Mehlen P. and Puisieux A, 2006). Curative resection of stage I or II disease is often followed by locoregional or distant recurrences. Cells of the primary tumor tissue enter the circulation and establish distant microscopic metastases (Mocelin S et al, 2006). The occurrence of circulating tumor cells/distant microscopic metastases in bone marrow/minimal residual disease seems to constitute a negative prognostic factor. MRD cells may remain dormant for a very long time or can begin to proliferate after variable times of dormancy in response to signals not completely understood so far. Adjuvant chemotherapy, given in order to eliminate remaining tumor cells, seems to be of limited value in killing dormant tumor cells. Besides resistance of the tumor cells they may be protected furthermore by the interaction with the microenvironment at specific metastatic sites. Therapeutic modalities aiming at the disruption of the interaction of tumor cells with stromal elements or blockade of survival signaling in metastatic cells may be a step necessary to prevent relapse (Wieder R et al. 2005). The methods for the isolation and cultivation of MRD cells/repopulating tumor cells are rapidly developing. For example Lee J et al. 2006 demonstrate marked phenotypic and genotypic differences between primary human tumor-derived stem cells (TSC) and their matched glioma cell lines. Unlike the matched, traditionally grown tumor cell lines, TSCs derived directly from primary glioblastomas harbour extensive similarities to normal neural stem cells and recapitulate the genotype, gene expression patterns, and in vivo biology of human glioblastomas. These findings suggest that TSCs may be a more reliable model than many commonly utilized cancer cell lines for understanding the biology of primary human tumors. Although a lot of experimental work suggests the presence of stem cells in solid tumors, there remain many uncertainties, both theoretical and technical, about the interpretation of these results. The case that a small proportion of cells in solid tumors are specific cancer stem cells and that these cells can be successfully identified and isolated has not yet been proven (Hill RP. 2006). The recurrent, metastatic tumor cells may just recapitalize some of the stem cell properties, which were turned off in the tumor cell population.
In many mammals, including humans, an adult stem cell-like subpopulation termed the "side population" (SP) has been identified (Hirschmann-Jax et al, 2005). SP cells can rapidly efflux lipophilic fluorescent dyes to produce a characteristic profile based on fluorescence-activated flow cytometric analysis. Previous studies have demonstrated SP cells in bone marrow obtained from patients with acute myeloid leukemia, suggesting that these cells might be candidate leukemic stem cells, and recent studies have found a SP of tumor progenitor cells in human solid tumors. These new data indicate that the ability of malignant SP cells to expel anticancer drugs may directly improve their survival and sustain their clonogenicity during exposure to cytostatic drugs, allowing disease recurrence when therapy is withdrawn. Identification of a tumor progenitor population with intrinsic mechanisms for cytostatic drug resistance might also provide clues for improved therapeutic intervention.
A distinct SP was found in neuroblastoma cells from 15 of 23 patients (65%). The SP was capable of sustained expansion ex vivo and showed evidence for asymmetric division, generating both SP and non-SP progeny (Hirschmann-Jax et al, 2004). These cells also expressed high levels of ABCG2 and ABCA3 transporter genes and had a greater capacity to expel cytotoxic drugs, such as mitoxantrone, resulting in better survival. SPs also were detected in breast cancer, lung cancer, and glioblastoma cell lines, suggesting that this phenotype defines a class of cancer stem cells with inherently high resistance to chemotherapeutic agents that should be targeted during the treatment of malignant disease.
Recent work indicates that the growth and behavior of cancers are ultimately determined by a small subpopulation of malignant stem cells and that information about the properties of these cells is urgently needed to enable their targeting for therapeutic elimination (Locke et al, 2005). A key feature of normal stem cells is their asymmetrical division, the mechanism that allows stem cell self-renewal, while producing hierarchies of amplifying and differentiating cells that form the bulk of the tissue. Most cancer deaths result from epithelial malignancies, but the extent to which the hierarchical proliferative stem and amplifying cell patterns of normal epithelia are actually retained in epithelial malignancies has been unclear. Cell lines generated from carcinomas consistently produce in vitro colony patterns unexpectedly similar to those produced by stem and amplifying cells of normal epithelia. From the differing types of colony morphologies formed, it is possible to predict both the growth potential of their constituent cells and their patterns of macromolecular expression. Maintenance of a subpopulation of stem cells during passage of cell lines indicates that the key stem cell property of asymmetrical division persists but is shifted towards enhanced stem cell self-renewal. The presence of malignant epithelial stem cells in vivo has been shown by serial transplantation of primary cancer cells in nude mice and the present observations indicate that stem cell patterns are robust and persist even in cell lines. An understanding of this behavior should facilitate studies directed towards the molecular or pharmacologic manipulation of malignant stem cell survival.
The cooperation between epithelial and mesenchymal cells is essential for embryonic development and probably plays an important role in pathological phenomena such as wound healing and tumor progression (Desmouliere et al., 2004). It is well known that many epithelial tumors are characterized by the local accumulation of connective tissue cells and extracellular material; this phenomenon has been called the stroma reaction. One of the cellular components of the stroma reaction is the myofibroblast, a modulated fibroblast which has acquired the capacity to neoexpress alpha-smooth muscle actin, the actin isoform typical of vascular smooth muscle cells, and to synthesize important amounts of collagen and other extracellular matrix components. Myofibroblasts are capable of remodeling connective tissue but also interact with epithelial cells and other connective tissue cells and may thus control such phenomena as tumor invasion and angiogenesis and therefore may represent a new important target of antitumor therapy.
Carcinoid tumors are slow-growing neuroendocrine neoplasms most commonly associated with the gut and broncho-pulmonary system (Modlin et al., 2004). They often present with pronounced fibrosis in the peritumoral tissues, distant in the heart or lungs, and locally in the peritoneal cavity. Despite medical and therapeutic advances that have alleviated symptoms and prolonged life, a substantial subset of patients develops mesenteric and small bowel carcinoid fibrosis and/or carcinoid heart disease. In the past, individuals with carcinoid disease died of metastasis and uncontrollable symptomatology. The unraveling of the molecular events indicative of fibrosis in these cells and the identification of appropriate therapeutic targets is of considerable patient-care relevance.
The cause of fibrotic diseases, pathologies characterized by excessive production, deposition, and contraction of extracellular matrix, is unknown (Leask and Abraham, 2004). Interactions among the profibrotic proteins transforming growth factor-beta (TGF-beta), connective tissue growth factor (CTGF, CCN2), and ED-A fibronectin (ED-A FN) and the antifibrotic proteins tumor necrosis factor-alpha (TNF-alpha) and gamma-interferon (IFN-gamma) seem to be involved. Malignant tumors induce development of their own stromal tissues during the processes of growth, progression and metastasis (Sugimoto et al., 2005). In an experimental system consisting of a human small-cell lung cancer cell line, WA-ht, and its mouse stromal fibroblast cell line, WA-mFib, both originally derived from a xenograft tumor in a mouse subcutis, co-culture of the cells significantly augmented the plating efficiency of WA-hT cells in vitro, and their co-inoculation in nude mice shortened latency and tumor doubling time. Antisense oligonucleotides to hepatocyte growth factor (HGF) cancelled these augmentation effects with co-culture. The findings highlight the substantial roles of tumor stromal fibroblasts, interacting with soluble growth factors, in promoting the malignant propensity of the tumor (Matsumoto and Nakamura, 2006).
Receptor tyrosine kinase (RTK) targeted agents such as trastuzumab, imatinib, bevacizumab, and gefitinib have illustrated the utility of targeting this protein class for treatment of selected cancers (Christensen et al., 2005). A unique member of the RTK family, c-Met, also represents an intriguing target for cancer therapy that is yet to be explored in a clinical setting. The proto-oncogene, c-Met, encodes the high-affinity receptor for hepatocyte growth factor (HGF) or scatter factor (SF). In support of the clinical data implicating c-Met activation in the pathogenesis of human cancers, introduction of c-Met and HGF (or mutant c-Met) into cells conferred the properties of motility, invasiveness, and tumorgenicity to the transformed cells. Conversely, the inhibition of c-Met with a variety of receptor antagonists inhibited the motility, invasiveness, and tumorgenicity of human tumor cell lines.

Small cell lung cancer (SCLC)

Relapsing tumors are heterogeneous and most likely contain subpopulations of tumor cells, which are involved in reestablishment of cancers and expansion of metastatic tissue. Heterogeneity is expected to be expressed as difference in growth characteristics, cell surface markers, expression of resistance-associated gene products, variable metastatic potential and dye transport/drug resistance. According to the literature subpopulations of cells can be isolated and studied for their repopulating capability and chemosensitivity. The ultimate goal is the detection of chemotherapeutic drugs with highest efficacy against low-proliferating, highly resistant tumor subpopulations, most likely responsible for the recurrences.
Among lung cancers SCLC is characterized by the tendency for early dissemination, high initial response rates to chemotherapy and a high frequency of metastases (Jackman DM and Johnson BE, 2005). Although most patients respond to treatment, more than 95 % of patients eventually die from the cancer. SCLC accounts for approximately 15 % of all new lung cancer cases and the tumors are a consequence of tobacco smoking. An increasing number of lung cancer patients is treated at the oncology unit in Hietzing each year.
Tumor growth in SCLC is effected by autocrine peptide growth factor loops, in many cases involving tyrosine kinase receptor c-Kit (up to 70 % of cell lines and tumors; imatinib shows no clinical effects. In a significant part of the tumors the gene amplification of myc – oncogens are associated with a rapid proliferation (up to 40 % of tumors and cell lines), supplemented by deletion of one of several tumor suppressor genes. The differentiation of SCLC from other lung tumors is accomplished by histology and detection of epithelial/neuroendocrine markers such as chromogranin, synaptophysin and CD56.
Surgical resection in limited disease is rarely possible; most patients have advanced from limited-stage disease to extensive-stage disease at presentation (60-70% extensive disease) with a very low median survival of 7 -12 months (limited-stage median survival 23 month). Systemic treatment with combination chemotherapy (etoposide, cisplatin + chest radiotherapy) results in a complete response in 80% of patients and a median survival > 17 month in limited-stage disease, whereas at extensive-stage the response rate is limited to 20% CR > 7 months. Newer agents targeting c-kit, proteasomes, mTOR, farnesyl transferases had no survival benefit. Receptor tyrosine kinases (RTKs) are important in normal cellular physiology as well in the pathogenesis of a variety of tumors, including lung cancer (Maulik et al. 2003). RTKs are a target for novel therapies currently being investigated. In the clinical situation, EGFR inhibitors and c-Kit inhibitors are already being utilized, and c-Met inhibitors are in development. Even though the RTK inhibitors provide a novel mechanism, it is important to realize that lung cancer etiology is a complex process, and eventually standard chemotherapy may need to be used in conjunction with these novel therapies to make an important difference in response rates. Most patients relapse within one year of starting treatment; second line treatment consists of single-agent topotecan or a combination of cyclophosphamide, doxorubicin, vincristine resulting in a 20 % response rate with median survival of approximately 25 weeks. Iriniotecan and paclitaxel are currently studied.

Carcinoid, colon and breast cancer.

Other tumor entities available at the oncology unit/Hietzing include carcinoids, that are known to share some neuroendocrine markers with SCLC. Colon cancer and breast cancer samples are to be obtained by cooperations within the cluster.
Repopulating subpopulations of tumor cells have been described in colon and breast cancer. A rare population of undifferentiated cells seems to be responsible for tumor formation and maintenance and it has been reported recently that tumorigenic cells in colon cancer are included in the high-density CD133(+) population, which accounts for about 2.5% of the tumour cells (Ricci-Vitiani et al 2006). Subcutaneous injection of colon cancer CD133(+) cells led to induction of the original tumour in immunodeficient mice, whereas CD133(-) cells did not form tumours. CD133(+) colon cancer cells grew exponentially for more than one year in vitro as undifferentiated tumour spheres in serum-free medium, maintaining the ability to induce tumors.
A second group has defined a colon tumor repopulating cell using renal capsule transplantation in immunodeficient NOD/SCID (colon cancer-initiating cell; CC-IC; O’Brien et al, 2006). Purification experiments established that all CC-ICs were CD133(+); the CD133(-) cells that comprised the majority of the tumor were unable to initiate tumour growth. Limiting dilution analysis revealed that there was one CC-IC in 5.7 x 10(4) unfractionated tumour cells, whereas there was one CC-IC in 262 CD133(+) cells, representing >200-fold enrichment. CC-ICs within the CD133(+) population were able to maintain themselves as well as differentiate and re-establish tumour heterogeneity upon serial transplantation.
A tumor subpopulation with repopulating capacity has been defined in breast cancer and that knowledge was applied to the phenotypic analysis of bone marrow micrometastase. The presence of disseminated tumor cells (DTC) in the bone marrow of breast cancer patients is an acknowledged independent prognostic factor. The biological metastatic potential of these cells has not yet been shown. The presence of putative breast cancer stem cells is shown both in primary tumors and distant metastases. These cells with a CD44(+)CD24(-/low) phenotype represent a minor population in primary breast cancer and are associated with self-renewal and tumorigenic potential (Balic et al, 2006).
Therefore within this cell biological part of the cluster project, advantage will be taken of these recently published characterization of the highly malignant repopulating subpopulations in the respective tumors in order to investigate the gene expression, biological properties and drug resistance in detail.


Projected Experiments

Permanent cell lines
A large collection of permanent cell lines obtained from the American Tissue Culture Collection (ATCC) is available, including cell lines, which have been reported to contain repopulating subpopulations (Lee et al. 2005). These cell lines can be used immediately for the in vitro experiments.

Collection of tumor cells
Isolation of tumor cells from pleural effusions, ascitic fluid and biopsies as described in the clinical part of the project. For SCLC paired tumor samples from primary tumor and relapse should be obtained. Informed consent is obtained from all patients as prerequisite. Isolation of tumor cells by tissue digestion (where indicated), centrifugation and cultivation in serum-supplemented or serum-free growth factor-supplemented media (Lee et al., 2006). For repopulating cells exhibiting a CD133+-phenotype magnetic cell separation systems are available commercially. Aliquots of the isolated cell suspensions are stored in liquid nitrogen. Soft agar assays can be used to characterize the capacitiy of the tumor cells to form new tumors. Tumors to be included in the investigation: SCLC, colon cancer and breast cancer. Primary cultures of different tumors have been collected in the past by the participating institutions and can be used immediately for the present project.

Characterization of primary tumor cells
Assessment of repopulating cell markers, such as CD133 for neuroendocrine tumors and colon cancer cells, and CD44/CD24 for breast cancer cells respectively. Gene expression profiling in enriched repopulating tumor cells with comparison to unfractionated whole tumor cell populations. Confirmation of the results using RT-PCR and monoclonal antibodies/functional test as available.
Separation of heterogeneous cell populations from cell lines and tumor specimen:
- Separation of primary tumor cells according to growth morphology. Isolated tumor cells may grow simultaneously as differentiated attached monolayers, as less differentiated floating suspension cultures or as spheroidal colonies. These forms can be separated and tested individually for their drug sensitivity and other properties. (SCLC as exception is growing mainly as cell suspension).
- Separation of the primary tumor cells according to their metastatic potential. Tumor cell suspensions are embedded in ECM (extracellular matrix) drops. Evading cells can be isolated, subcultured and tested further for their phenotype and drug resistance and compared to the bulk population of the tumor cells.
- Separation of primary tumor cells by in vitro drug chemosensitivity. In vitro treatment of isolated tumor cells with new drugs and screening of the phenotype of the surviving resistant populations.
Dye efflux tests (Hoechst 33342 or rhodamine 123) provide informations about heterogeneity and possible drug resistance. Markers of cellular resistance, especially ATP-Binding Cassette transporter isoform G2 (ABCG2) can be further detected by flow cytometry using appropriate monoclonal antibodies. The ABCG2 transporter can be assessed functionally by measurement of topotecan efflux (HPLC test for topotecan available in cooperation).

Isolation of tumor cells with support of activated fibroblasts
Interaction of cancer cells with stromal elements may provide a protected environment. Despite a high shedding of tumor cells from many tumors, the actual formation of metastases is low and seems to be linked to protected tumor cells which are activated and recovered from dormancy. Activated fibroblasts can be generated from fibroblast cell lines pretreated with factors like TGF-beta and isolation of primary tumor cell cultures can be performed in filter insert coculture systems. Activation of the fibroblasts can be tested using the HBA-71 monoclonal antibody directed to the MIC2 antigen, which is expressed on activated tumor fibroblasts (Fellinger et al. 1991).

Gene expression profiling
The different tumor cell subclones may be compared to circulating tumor cells of the same patient in order to investigate which tumor cell subclones in the bulk tumor correspond to the disseminated population of cells and specific subpopulations with unique properties. Profiling is done in cooperation with the LBI for Gynecology and Gynecological Oncology.

In vitro chemosensitivity
The chemosensitivity of the different tumor cell subclones should be tested with a panel of newer experimental therapeutics in order to assess the highest efficient drugs, preventing the generation of metastases from MRD tumor cell clones. The chemotherapeutic drugs to be tested include the classical cytotoxic drugs and the innovative new drugs targeting specific signaling pathways, as summarized in the contribution of the LBI for Applied Cancer Research. In addition inhibitors of specific pathways affecting signal transducers like mTOR, PI3-kinases, c-met and others can be used in vitro to check for the relative contribution of the different growth-promoting mechanisms. Since the classical chemotherapeutic drugs seem not to work for the highly malignant repopulating subpopulations an extensive screen of newer and alternative drugs will be necessary.