10 Cancer stem cells

10.1 Discovery of tumour initiating cells in solid tumours

When scientists isolated cancer cells from solid tumours and analysed them individually using flow cytometry and fluorescent markers it was observed that these cells were not homogenous and depending on the nature of the markers used so-called side populations could sometimes be observed, e.g., cells which would maintain a lower intracellular concentration of a fluorescent dye than the majority of cells. This specific observation was linked to the comparatively higher activity of membrane efflux pumps (ABC, pgp) in those cells which continuously pumped out substrate dye. Similar side populations can be observed when probing the cancer cell population of a tumour with antibodies for certain cellular surface markers.

To find out whether such sub-populations differ in human tumours the cells in the breast cancer lumps of human patients were examined after surgery (Figure 10.1). A combination of a tissue-specific and lineage-negative (LIN−) markers allows the isolation of tumour cells which can then be separated in different subpopulations on the basis of cell surface marker expression or enzymatic activity. To evaluate whether the presence or absence of certain surface markers predicts tumour growth a certain number of cells (limiting dilutions) from these populations are transplanted under the skin of immunodeficient mice to see whether a tumour develops (Al-Hajj et al. (2003)).

The isolation of tumour initiating cancer cells from human breast cancers reveals heterogenous cell populations some of which are able to give rise to a new tumour while the majority will not. When the process is subsequently applied to the mouse tumour the same populations can again be identified, etc.

Figure 10.1: The isolation of tumour initiating cancer cells from human breast cancers reveals heterogenous cell populations some of which are able to give rise to a new tumour while the majority will not. When the process is subsequently applied to the mouse tumour the same populations can again be identified, etc.

Subsets of tumour cells giving rise to secondary tumours more frequently than unsorted tumour cells because thet are enriched in cells that have the ability to initiate or propagate these tumours [tumour propagating (TPCs) or tumour initiating cells (TICs)]. When the enriched PTC/TIC populations are injected in mice just a few of these cells are requried to give rise to a tumour whereas those cells lacking the correct surface marker combination may not give rise to a tumour or onlya after injcetion of many more cells.

When the tumour that has develope in the mouse is then disaggregated and sorted in the same way as the orignial patient tumor the same sub-populations can be identified; when the enriched population is again isolated and transplanted into antoher mouse the populations show the same ability to propagate or initiate tumour growth as in the previous round. in fact, with the TICs/TPCs this process can be repeated again and again, while cells that do not have those surface markers tumours fail to develop.

Interestingly, the ability to give rise to new tumour is dependent on specific lineage markers, i.e. surface markers that characterise a cell type and a certain stage of differentiation with in the differentation sequence typical for the cell type. Only some types of cancer cells are able to induce tumours repeatedly - these cells carry surface markers similar to adult stem cells.

10.2 Adult stem cells

Adult stem cells are tissue specific and long lived cells whose role is to replenish the tissue specific differentiated cells that are lost; each adult stem cell has the ability to recapitulate its complete tissue specific cell lineage. Within the tissue these special cells are maintained in a specific niche; they are normally dormant and only divide to produce progenitor cells when those are needed. They have the ability to divide indefinitely – but will only divide occasionally and ‘on demand.’

As with other cells their cell division produces two daughter cells; however, stem cells actually undergo asymmetric cell division because one of the daughter cells remains behind in the niche and retains all the parental traits while the second daughter cell moves away from the niche and becomes a progenitor cell which continues rapidly and repeatedly to replenish the tissue. Only the cell that has remained in the niche will be able to undergo further asymmetric cell divisions (Figure 10.2)..

Caption required

Figure 10.2: Caption required

While progenitor cells have the ability to divide rapidly they can only do so for a limited number of times; ultimately they produce terminally differentiated cells which have lost their proliferation capabilities. In contrast, the ability of the adult stem cells that remain in the niche to indefinite continue asymmetric division requires self-renewal, i.e., the ability to undergo an indefinite number of cell division while remaining in a fully undifferentiated state.

While most other cells in tissues are fully differentiated and are replaced regularly with new cells adult stem cells persist throughout our whole life. Because of their long life-span they are at a higher risk of being damaged by exposure to toxic compounds. This is probably the reason why they are typically also better protected e.g. by having higher levels of efflux pumps.

10.2.1 Adult stem cells and cancer stem cells share properties

As discussed abobe, tumour cell populations can be selected on based on surface (lineage) markers. When the surfaces markers that differentiate between the adult stem cell and the progenitor cells for a tissue are applied to the cancer cells ‘progenitor-like’ and ‘stem-cell-like’ sub-populations can be identified (Figure 10.1). * ‘Progenitor’-like cancer cells can proliferate for a limited time but can not recapitulate the complete tumour.

  • ‘Stem cell’-like cancer cells selected from a tumour can give rise to a new tumour – the process can be repeated multiple times. ‘Stem cell’-like cancer cells also have the ability to self-renew.

Those cancer cells which have the stem cell like properties and the ability to propagate or initiate a secondary tumour (TPCs or TICs) are therefore also known as cancer stem cells (Batlle and Clevers (2017)).

10.3 Cancer stem cells

The sub-populations of cancer cells that has stem cell like properties and which are able to initiate or propagate tumours (known as CSCs, TICs, or TPCs) have been identified in a wide range of human tumours already; these include breast, colon, pancreatic, brain cancers and an ever expanding list of other cancer types. Their relative proportion with in the overall cancer cell population differs in different tumours and ranges from in some cases 25% (e.g. some malignant melanomas), to 1 per 1,000 to 1 per 100,000. Their presence has not been confirmed for every type of cancer and it is not clear whether every cancer is sustained by CSCs. However, as the CSC subpopulation appears to sustain cancer development in those cancers where they exist this has important implications for their role in cancer progression and therapy.

If CSCs are essential to propagate or initiate tumours this suggests that the original transformed cells from which a tumour arises are CSCs. This suggests that CSC’s

  • have a stem cell origin and mutations/de-differntiations occur that transform adult stem cells to become cancer cells, or
  • could arise from more differentiated / progenitor type cells that have been transformed acquire mutations that give them the stem cell like phenotype.
  • One of the reasons in favour of the stem cell origin of cancers is the fact that the long life-span of adult stem cells would mean such cells could be exposed to multiple mutations over a longer periods of time, consistent with the long term multi-stage nature of progression typically observed in cancer.

Regardless, once CSCs have the ability for indefinite proliferation this allows accumulation of mutations and supports clonal evolution.

10.3.1 CSCs contribute to tumour heterogeneity

Genetic heterogeneity and somatic evolution are key influences that shape the heterogeneity of cancer cell populations (Figure 10.3(Fulawka, Donizy, and Halon (2014)): Repeated mutations (and (genomic instability)[#genomic instability]) lead to the development of genetically and linked to this phenotypically heterogeneous cell populations which due to selection undergo clonal evolution with expansion of the fittest clones with in a cancer cell populations. A second source of heterogeneity is linked to the fact that CSCs repopulate a lineage of cells at different stages of maturation, including progenitor and mature / differentiated cells. In the tumour both sources of heterogeneity are relevant and occurring in parallel; this could mean that clonal evolution of CSCs is the source of heterogeneity relevant for the generation of metastases or the repopulation of tumours after therapy while at the same time each clone contributes to heterogeneity by producing a range of cells of different maturation stages.

Tumours have evolving heterogeneous cell populations (right). This is due to a mix of two process: the on-going process of somatic clonal evoluation (left) and the proliferative maturation from cancer stem cells to more mature cells.

Figure 10.3: Tumours have evolving heterogeneous cell populations (right). This is due to a mix of two process: the on-going process of somatic clonal evoluation (left) and the proliferative maturation from cancer stem cells to more mature cells.

10.4 Clinical consequences of CSCs

Some of the properties of cancer stem cells are potentially important for cancer development. Their longevity (if derived from adult stem cells) facilitates the accumulation of mutations and thus allows for a prolonged clonal evolution with consequential acquisition of various cancer hallmarks.

The ability of CSCs to initiate tumour growth and their ability to recapitulate the different cell maturity stages (stem, progenitor, differentiated cell) would suggest that the cells are most likely to give rise to metastasis if carried away from the primary tumour.

The physiologically high levels of efflux pump activity found in adult stem cells would if replicated CSCs protect these against drugs that are substrates of these pumps. Furthermore, the fact that such cells typically undergo asymmetric division would allow them to divide less frequently and thus not be rapidly proliferating. This, again, would reduce the sensitivity to many chemotherapeutic agents.

Anti-cancer therapies may not kill all tumour cells equally, e.g. cancer stem cells (CSCs) that sustain tumour growth or another population of more slowly cycling tumour cells may be responsible for tumour resistance to therapies and tumour relapse.

The CSC model suggests that inhibiting CSC renewal or promoting their differentiation should induce tumour regression. Drugs could impair CSC self-renewal, induce their specific cell death, induce their differentiation or target their niche. All of these strategies would lead to the depletion of the pool of CSCs and subsequent tumour regression. However, if the CSC potential is reversible, or if newly acquired mutations confer resistance to therapy, then tumour regression would only be transient, leading to cancer relapse.

Caption required

Figure 10.4: Caption required



–Drug sensitivity