Self-Generated Chemoattractant Gradients: Attractant Depletion Extends the Range and Robustness of Chemotaxis

Abstract

Chemotaxis is fundamentally important, but the sources of gradients in vivo are rarely well understood. Here, we analyse self-generated chemotaxis, in which cells respond to gradients they have made themselves by breaking down globally available attractants, using both computational simulations and experiments. We show that chemoattractant degradation creates steep local gradients. This leads to surprising results, in particular the existence of a leading population of cells that moves highly directionally, while cells behind this group are undirected. This leading cell population is denser than those following, especially at high attractant concentrations. The local gradient moves with the leading cells as they interact with their surroundings, giving directed movement that is unusually robust and can operate over long distances. Even when gradients are applied from external sources, attractant breakdown greatly changes cells' responses and increases robustness. We also consider alternative mechanisms for directional decision-making and show that they do not predict the features of population migration we observe experimentally. Our findings provide useful diagnostics to allow identification of self-generated gradients and suggest that self-generated chemotaxis is unexpectedly universal in biology and medicine.

Inference of the drivers of collective movement in two cell types:Dictyosteliumand melanoma

“…As the melanocytes that mutate into melanoma cells develop from neural crest cells [39], it is likely that melanoma cells will also exhibit CIL, which may be a contributing factor to the selection of the overcrowding model for the melanoma dataset. Previous results simulated from an individual-based cell movement model suggested that CIL may also play a role in Dictyostelium movements in the system investigated here [7]. Our inability to detect this effect in Dictyostelium through a preference for the overcrowding model over the receptor saturation model may be a result of the loss of information incurred in moving from an individual-based modelling approach, where the movement path of each cell is known, to the population-based approach used in our study, where individual movement paths are not analysed.…”

“…For Dictyostelium , the form of the predicted folate distribution gives a relatively close visual match to that measured experimentally by Tweedy et al . [7], using the same assay but with a higher initial folate concentration.…”

“…Our study systems here are relatively simple in comparison with the levels of complexity often observed in vivo , where multiple cell types may be interacting within a considerably more complex environment. However, they are nonetheless examples of real cell movement behaviour, one of which is of great medical relevance, in which we have been able to detect the presence of self-generated chemotactic gradients; a movement driver recently found to be important in many systems, including in vivo [3–7]. This success is an encouraging first step, indicating that model inference has the potential to support targeted experimental work in increasing our understanding of collective cell movement in a range of systems.…”

“…Data on the collective movement of Dictyostelium cells during an under-agarose assay [19] were collected by Tweedy et al [7]. The agarose under which the cells moved contained folate, a chemoattractant that the cells can deplete from their environment, at an initially homogeneous concentration of 10 µM.…”

“…The agarose under which the cells moved contained folate, a chemoattractant that the cells can deplete from their environment, at an initially homogeneous concentration of 10 µM. Under these conditions, Dictyostelium cells create a gradient in folate through depletion, and then collectively move up this gradient [7]. …”