Cancer-associated fibroblasts (CAFs) and the associated extracellular matrix
(ECM) constitute a significant part of the tumour microenvironment (TME), playing
an important role in the invasive potential of the tumour. The alignment of CAFs
and the corresponding ECM which they produce and organise is linked with
increased cancer invasion. Additionally, massive variation in the physical
architecture of the ECM is observed in both normal and pathological tissues for
example swirling, diffuse or porous patterns. How these mesoscale patterns arise
remains largely unexplored.
An agent-based flocking model was developed to investigate CAF properties and
their involvement in emergent alignment. The model established that aligning cells
had a requirement of highly persistent migration coupled with an active cell-cell
collision guidance mechanism. The model predicted that alignment was a fragile
state which could be easily destroyed in a heterogeneous population. These
findings were confirmed experimentally.
The model was then extended to include a second underlying layer of ECM fibres
that the CAFs could produce, degrade and rearrange but were also instructed to
follow, constituting a CAF-ECM feedback loop. This mechanism was capable of
generating diverse matrix patterns, reminiscent of those seen in vivo. The model
was challenged to unpick the process of interconversion between matrix patterns
as seen in cancer, wound healing and ageing, which it elucidated with considerable
success.
Finally, clinical samples of ECM were quantified to establish if certain metrics of
ECM architecture could be useful clinical prognostic factors. Early results suggest
this to be true. Matrix patterns were quantified by a carefully constructed software
pipeline suitable for use by other researchers on versatile data samples.