Evolution of cancer stem cells: A problem with the cancer stem cell hypothesis

A99

The cancer stem cell hypothesis asserts that there is a subset of neoplastic cells, the so-called cancer stem cells, which have acquired both the hallmarks of cancer and the ability to self-renew. These cells divide both symmetrically, increasing the number of stem cells, and asymmetrically, to generate non-stem cells (cells that cannot replenish the neoplasm).
 >Using agent-based computational modeling, we simulated the evolution of the stem cell self-renewal rate (the proportion of stem cell divisions that were symmetric vs. asymmetric). In this model, we defined two classes of cells: stem cells and transient amplifying cells. The stem cells had slower growth and apoptosis rates than the transient amplifying cells. Furthermore, stem cells symmetrically divided and increased the stem cell population at an initial rate of 0.0018 per time step; otherwise stem cells divided asymmetrically. The self-renewal rate of new stem cells was allowed to mutate randomly.
 >Transient amplifying cells could divide 10 times before undergoing apoptosis or senescence. To keep the model tractable, there were only enough nutrients and space for the tumor to reach a carrying capacity of 100,000 cells. When the tumor size exceeded the carrying capacity, cells underwent apoptosis at random until sufficient nutrients existed to maintain the tumor.
 >We found that the mean frequency of the stem cells in the population reached 74.6% (standard deviation 0.4%, N=50 replicates); the control population in which the self-renewal rate was not allowed to mutate had a much lower mean frequency of stem cells (0.9%, standard deviation 0.3%, N= 50). The self-renewal rate of the stem cell population evolved to a mean value of 0.9359 (standard deviation 0.0007, N=50). Thus, there is natural selection for an increase in the self-renewal rate leading to a high frequency of stem cells in a neoplasm.
 >These results lead to an intriguing observation. Under conditions in which there is no form of extrinsic inhibition on the rate of symmetric stem cell division, stem cells should take over the neoplasm and reach a very high frequency (at a mutation-selection equilibrium). However, experimental evidence suggests that only on the order of 1 in a thousand cells in a neoplasm are stem cells. This presents a paradox to the cancer stem cell hypothesis that warrants further exploration.