(722f) Ranking the Relative Dominance of Migration Cues When Guiding Individual Fibroblasts Faced with a Directional Decision in Simple Microfluidic Bifurcations

Introduction: Directed cell migration in
complex micro-environments, such as in vivo pores, is important for predicting
locations of artificial tissue growth and optimizing scaffold architectures. However,
forecasting cell migration through microscopic pores is challenging because
their directional decisions are affected by a combination of different factors:
such as, physical confinement, varying levels of biochemical agonist gradients,
mitosis, contact guidance, etc. Therefore, the interplay between multiple
factors is crucial to unraveling the underlying biology mechanisms behind the
migration directionality and consequently, to achieving control over cell
behavior. To the best of our knowledge, previous studies tend to quantify the
directional decisions (i.e., a choice of whether to go left or right) of the
migrating cells as a function of isolated factors (e.g., channel length, width,
concentration gradient, contact guidance, etc.).  At most just two of the migration cues are
explored at a time, while the directional decisions of cells facing multiple (i.e. > 2) simultaneous physiochemical
cues have not been characterized.  In order
to address that niche, we rank the relative importance of four different chemical
and physical cues that influence fibroblast directional making decision
processes.  The cell choice was made due
to our interest in tissue generation and repair. To the best of our knowledge,
this is the first such study.

Materials and Methods: Polydimethylsiloxane (PDMS) devices with bifurcated branches of different widths were
created using a replica molding technique (see Fig. 1a).  Normal human
dermal fibroblast cells (NHDF) and mouse embryonic (NIH/3T3) fibroblasts were then induced to travel across these geometries by establishing a
gradient of chemoattractant “PDGF-BB” between the entrance and the exit of the
device. Time-lapse phase-contrast imaging of the fibroblast migration
was performed using a fully automated Olympus IX83 microscope. The migrating cells and the directional
decisions made by them were tracked using the Manual Tracking plug-in for
ImageJ software. Quantitative statistics of the cell decisions were then
calculated using an in-house Matlab® 2016b code. Significance level was then determined
by using a non-parametric test for a binomial distribution and the
nonparametric Mann-Whitney U test. Furthermore,
we integrated a novel image-based modeling approach, in order to assess the
localized chemoattractant consumption by the individual cell.  Eventually, a combination of statistical
analysis and diffusion modeling was used to report how the presence of multiple
complex migration cues, including cell-cell influences, affect the fibroblast
decision-making.

Results and
Discussion:  Overall, we found that the cells alternate between the bifurcated branches
when their width is equal, but prefer the wider channels over the narrower ones
when choosing between asymmetric branches
(Fig. 1a). This is evidenced by the trend in the bias confidence
level, which for NHDF cells is p = 1.6´10^-7 for 1.5X branch widths ratio (BWR) and
5.5´10^-17 for 3X
BWR.  The only exception to this occurred when the cells mitosed at the
first bifurcation.  In that case the
daughter cells went in the opposite directions, with one of them going into the
narrow channel (Fig. 1b). In fact, the
probability for one daughter cell to enter the narrow channel and the other
daughter to go into the wider one is about 50:50, which indicates that the cells
are unaffected by the presence of the width asymmetry when making their
directional decisions.  Therefore, the mitosis
effects supersede those of the BWR, and by association those of the weaker cues
as well.

Only when the branches were symmetric
in width (see Fig. 1c), did the chemoattractant gradient become relevant
in directing the cells towards the path that they take.  In this case, the fibroblasts alternate between
the left and the right paths that they encounter at the bifurcation.  In our previous study, we showed that this
effect occurs due to localized gradient differences created by the leading
cells consuming the chemoattractant. 
This causes their followers to take the alternative path, which is
consequently richer in the PDGF-BB.  In
this study, we confirmed that effect holds for progressively wider symmetric branches, in order to rule out
the possibility that it is the width alone (and not the width asymmetry) that guides the fibroblasts’
choices.  Finally, when both the gradient
and the channels were symmetric, contact guidance became important for guiding
the cells in making directional choices (Fig. 1d). Among ~90% of the NHDF
cells display wall contact, ~75% commit to the same branch as the wall that they
were touching, while only ~25% move to the opposite branch.

Conclusions: Based on these results, we have
ranked the relative importance of the different directional cues affecting the
migration of fibroblasts from most influential to the least as follows: mitosis
> channel width asymmetry > chemoattractant gradient difference > and
contact-guidance.  Furthermore, the
uncovered patterns hold for fibroblasts originating from two different species,
which demonstrates the robustness of our results. Consequently, this study has
potential widespread implications to the field of cell migration, as it
provides an insight into the understanding of cellular behavior in the presence
of complex combinations of different directional cues, which is the crucial activity
in wound healing, developmental biology, and regenerative medicine. It also
opens a potential avenue to those who aim to control the cellular behavior
(e.g., minimizing the product variability for the sake of successful
biomanufacturing of artificial tissues).