Modeling How Heterogeneity in Cell Cycle Length Affects Cancer Cell Growth Dynamics in Response to Treatment

Abstract

Tumors are complex, dynamic, and adaptive biological systems characterized by high heterogeneity at genetic, epigenetic, phenotypic, as well as tissue microenvironmental level. In this work, utilizing cellular automata methods, we focus on intrinsic heterogeneity with respect to cell cycle duration and explore whether and to what extent this heterogeneity affects cancer cell growth dynamics when cytotoxic treatment is applied. We assume that treatment acts on cancer cells specifically during mitosis and compare it with a (cell cycle-non-specific) cytotoxic treatment that acts randomly regardless of the cell cycle phase. We simulate the spatiotemporal evolution of tumor cells with different initial spatial configurations and different cell length probability distributions. We observed that in heterogeneous populations, strong selection forces act on cancer cells favoring the faster cells, when the death rates are lower than the proliferation rates. However, at higher mitotic death rates, selection of the slower proliferative cells is favored, leading to slower post-treatment regrowth rates, as compared to untreated growth. Of note, random cell death progressively eliminates the slower proliferative cells, consistently, favoring highly proliferative phenotypes. Interestingly, compared to the monoclonal populations that exhibit complete response at high random death rates, emergent resistance arises naturally in heterogeneous populations during treatment. As divergent selection forces may act on a heterogeneous cancer cell population, we argue that treatment plan selection can considerably alter the post-treatment tumor dynamics, cell survival, and emergence of resistance, proving its significant biological and therapeutic impact.

PD-L1 Expression Fluctuates Concurrently with Cyclin D in Glioblastoma Cells

“…Lack of normal growth control is not only operative in early tumorigenesis, but also during invasion and metastasis. The natural variability in the proliferative capacity of the cell population plays a relevant role in tumor evolution [ 16 ].…”

Stochastic fluctuations drive non-genetic evolution of proliferation in clonal cancer cell populations

“…One of the novel aspects of this work was the spontaneous increase in average growth rate with time. A reduction in cell cycle duration over time has previously been proposed theoretically under therapy-induced cell death [36]. In that work it was assumed that a tumour population initially heterogeneous consisted of cells with different, albeit intrinsic and fixed, proliferation rates.…”

“…This wellknown fact is typically attributed to uncontrollable changes in experimental conditions. Some authors have previously accounted for these baseline variations in cell cycle duration among heterogeneous cancer cell populations [26,27] that could be indeed tightly linked to tumour response to therapy [28]. By nature, phenotypic plasticity could affect any cell trait [29], but here we will focus on proliferation as a key phenotypic characteristic in cancers.…”

“…Genetic changes seem to be 54 necessary [26] but would not suffice to generate such accelerated tumour growth since they would 55 require either an extremely high mutation rate, which would be restricted by cell viability, or a outcome that cannot be attributted only to differences in the number of cells seeded initially. Some authors have previously accounted for these baseline variations in cell cycle duration among heterogeneous cancer cell populations [34,35] that could be indeed tightly linked to tumour response to therapy [36]. Phenotypic plasticity could affect any cell trait [37], but here we will focus on proliferation as a key phenotypic characteristic in cancers.…”

Modeling Progression of Single Cell Populations Through the Cell Cycle as a Sequence of Switches

“…Some mathematical models of the cell cycle try to tackle these questions. For example, agent-based or cellular automaton cell cycle models focus on the optimization of cancer drug delivery (Altinok et al, 2007), competition of fast and slow cell cycles within a tumor under treatment (Tzamali et al, 2020), or cell confluence and elongation of the G1 phase (Bernard et al, 2019). However, most of the existing models remain limited to describe the behavior of cell cycle during tumorigenesis at full complexity because of the existing discrepancy between the nature of the available molecular data and the level of the details of these models.…”