Tumours are made up of a heterogeneous population of cells which are distinct in terms of their differentiation competencies, proliferative capabilities as well as functional properties. [A] The mechanisms responsible for such heterogeneity are the subject of research, and two models have been put forth in order to explain the phenomenon – Cancer stem cells (CSCs) and clonal evolution. [A]

Cancer stem cells are a subset of the total population of cells in a tumour that have the ability to undergo self-renewal, as well as to differentiate into the different types of cells that comprise the tumour. [A] These CSCs are said to be responsible for tumorigenesis as well as for driving tumour growth. [U]

Evidence supporting the existence of cancer stem cells

Differences in clonogenicity among cancer cells were first documented in cases of leukaemia and multiple myeloma. It was found that 0.01 – 1% of the cells were capable of extensive proliferation, and able to establish colonies when grown in vitro (Park, C. H., Bergsagel, D. E. & McCulloch, E. A. Mouse myeloma tumor stem cells: a primary cell culture assay. J. Natl Cancer Inst. 46, 411–422 (1971). Two possible explanations existed for this – either leukemic cells had a low overall capacity for proliferation, or only a definite subset of these cells were clonogenic.

In 1994, John Dick and his group of researchers carried out a landmark study where CSCs were isolated from a mouse model that had been transplanted with human AML cells. [Z15] This was the first conclusive evidence for the existence of a subset of the leukaemia cells that were highly clonogenic, in comparison to the remaining cancer cells.

It was later observed that a similar condition exists in the case of solid tumours, where only a small subset of the total cell population is tumorigenic. [G]

Origins of cancer stem cells

Several contradictory theories exist regarding the cellular origins of cancer stem cells. Some state that these cells are derived from normal stem cells that have acquired oncogenic mutations [G], others refute this with the claim that cancer stem cells can arise from a committed progenitor cell that has acquired the properties of a stem-cell during its cancerous transformation [Z1], while yet others suggest that these cells could arise as a result of a fusion event between a stem cell and a tumour cell. [N]

The idea that cancers could arise from normal stem cells is highly plausible because not only do they continuously undergo divisions, but they are also long lived, allowing them to accumulate multiple mutations, as is required for a cancerous transformation. [B]

Apart from the accumulation of mutations, the interaction of a cell with its local microenvironment also influences the tumorigenic process. Mouse leukaemia models have been able to provide evidence that given suitable niche conditions, a progenitor cell is capable of de-differentiating to form a CSC. [V] However, since most progenitor and mature cells have a relatively short life-span, it seems unlikely that will be able to acquire the oncogenic mutations required to render them tumorigenic. [I]

Despite these explanations, the exact origin of most tumours and cancer stem cells remains unknown, and can only be speculated based on experimental findings. [A] Additionally, irrespective of the origin, the identification and isolation of CSCs in a tumour indicates that there exists a functional hierarchy exists within the tumour tissue. [L]

Properties of cancer stem cells

These cells can undergo symmetric as well as asymmetric divisions, which results in the expansion of the cancer stem cell population itself, as well as an increase in the number of differentiated cells that constitute the bulk of the tumour. [Z1]

THE CANCER STEM CELL MODEL

As previously mentioned, two models have been put forth to explain the heterogeneity of a tumour cell population. The first model is the CSC model, also known as the hierarchical model, which states that within a tumour, there exist different classes of cells and that the CSCs represent a biologically distinct subpopulation of cells that are capable of propagating the tumour. [C] It suggests that the characteristics of the cells within the tumour are intrinsically determined and therefore only certain cells possess the ability to undergo extensive proliferation to initiate tumour formation, these cells are called the CSCs; while the remaining cells are incapable of tumorigenisis.

According to the CSC model, although most cancers arise from a group of cells that are genetically monoclonal in nature, the high level of tumour heterogeneity is a result of the interaction between cells that are in different states of differentiation after have initiated from a common precursor. [Z12]

Evidence supporting this hypothesis emerges from the observation that though tumours may initially respond well to chemotherapy, there is often a case of relapse; which could occur due to the CSCs that persist post-treatment and are then able to re-initiate tumour formation. [Z13]

There are, however, limitations to the CSC model; the first being that all studies that support it have only addressed the potential of the cells to proliferate and give rise to tumours, but not the actual fate. [D] Since the conditions applied to test the tumorigenic potential of these cells may vary considerably from the conditions experienced by the cells in vivo, we do not know which of these cells actually contribute to the establishment and growth of the primary tumour. It is also noteworthy that it has been found that if the population of cancer cells acquires an immense number of mutations and aberrations, then almost all of them begin to show stem-like properties. [C] In such a case, the CSC model becomes irrelevant.

On the other hand, the stochastic model states that cells in a tumour are biologically equivalent, and that each cell has the ability to act as a CSC, given the right circumstances. A combination of intrinsic and extrinsic influences is said to determine the proliferative capacity and the ultimate fate of a cell. [C] Behaviour of a cell is therefore not pre-determined by intrinsic characters alone and tumour initiating cells cannot be enriched.

It is however likely for both these paradigms to be observed in vivo, in different cancers. Some cancers may follow the CSC model, while others may not. Based on transplantation studies in mice, it has been found that only in a fraction of cases, does AML follow the CSC model while in others there is no evidence for the existence of a highly tumorigenic sub-population of cells that continue to display CSC activity upon serial propagation. [F] Therefore, although CSCs may be responsible for driving the growth of a majority of tumours and cancers, there are studies which indicate that certain malignancies may be sustained primarily by the bulk of the tumour cells. [F]

IDENTIFICATION AND ISOLATION OF CANCER STEM CELLS

Cancer stem cell assays

Purification and enrichment techniques

CSC markers

CSCs in various cancers

HETEROGENEITY IN CANCER STEM CELL POPULATIONS

Based on the study of the CSC model, the question arises as to whether similar hierarchical subpopulations of tumorigenic and non-tumorigenic are observed in patients suffering from the same type of cancer; and whether these tumorigenic CSCs can be isolated based on conserved cell- surface markers. However it has been found that there are often phenotypic differences in CSCs even within the same cancer sub-type. [E]

For example, though it has been found that the CD44+/CD24- population of breast cancer cells are generally tumorigenic, this is not universally the case and in certain cases, it has been found that cells of diverse phenotypes are able to act as CSCs. (Al-Hajj, M., Wicha, M.S., Benito-Hernandez, A., Morrison, S.J., and Clarke, M.F. (2003). Proc. Natl. Acad. Sci. USA 100, 3983–3988.) Similarly, in gliomas CD133 expression is not always associated with CSCs, and in certain instances CD133- cells have also been found to be tumorigenic. (Beier, D. et al. CD133+ and CD133– glioblastoma derived cancer stem cells show differential growth characteristics and molecular profiles. Cancer Res. 67, 4010–4015 (2007).

Apart from the phenotype, the frequency of CSCs in a solid tumour or population of cancer cells is also variable. In melanomas, CSCs constitute anything between 1.6 to 20% of the total cells, while in the case of colorectal carcinomas, they represent between 1.8 to 24.5% of the cells. [A] Additionally, in general, the percentage of CSCs in solid tumours has been found to be significantly higher than the percentage of leukemic stem cells. [Z15]

This heterogeneity has implications on the prognosis of the disease as well as the outcomes of various therapeutic interventions. It is envisioned that with the identification of more refined markers and improved methods for determination of CSC frequency, we may eventually be able to correlate the percentage of CSCs with the tumour grade and the outcome. [A]