Microtubules, Take the Wheel: A Neuron’s Journey to its Target


Kahmina Ford

Faculty Mentor(s)

Erin Craig (Physics)


Neuroplasticity, the ability for neurons to reorganize and form new connections, has major implications for brain trauma recovery and neurological disorders. During nervous system development, neurons extend long narrow fibers called axons that must grow in the correct direction to form neural connections. The microscopic navigators that lead axon extension are called growth cones. These sensitive motile structures at the tip of axons respond to various chemical cues for navigation. In this study, we develop computer-based models to investigate how growth cones respond to external chemical “traffic signals” and guide axon outgrowth. Our model framework builds on previous studies by assuming that a dense network of dynamic filaments called f-actin operate as the engine of the growth cone vehicle. We use this framework to investigate the less well understood growth cone steering mechanism. We introduce long filaments of dynamic biopolymer proteins termed “exploratory” microtubules (MTs) that extend from the axon into the growth cone, randomly switching between states of growth and shrinkage as they “explore” the growth cone leading-edge. Our leading hypothesis is that microtubules respond to external guidance cues by triggering a biochemical reaction, causing the f-actin growth cone engine to engage. The objective of the current model is to investigate the physical role of the exploratory MTs and the mechanisms that bridge chemotactic cues to growth cone turning. By computationally reconstructing the mechanics of the growth cone, we are addressing an essential step to understanding the underlying causes of neurological diseases.

Keywords: Neuroscience, Cytoskeleton, Morphogenesis


7 thoughts on “Microtubules, Take the Wheel: A Neuron’s Journey to its Target”

  1. Hi Kahmina! Really nice job describing this project. I have two questions:
    1) What do the different colors indicate in the movie of your results near the end of the presentation?
    2) How do you set the chemical gradients in the model? Do the chemicals have some interesting spatial distribution? Is there a good way to use MATLAB to visualize that? (I guess that is more than one question.)

    1. Hello Dr. Snowden,
      Those are some really good questions!
      1) The different coloration in the simulation videos allows me to visualize and collect rates specific to the different quadrants of the growth cone (Q1: 0-45 degrees; Q2: 46-90…etc). I initially modeled the growth cone this way so that I could fit the model to the experimental data obtained in Lee & Suter 2008. The authors of this paper were particularly interested in analyzing microtubules’ role in adhesion-mediated growth cone guidance. When assessing MT translocation and polymerization rates they found that dynamic microtubules explore the peripheral domain 67% more frequently in the side quadrants (Q1 & Q4) than the center quadrants (Q2 & Q3). I thought this was really interesting because they also found that steady-state growth cones had higher densities of dynamic MTs on the sides. One possible explanation is that the increased MT density could increase the sensing potential for cues that are more “off-path” for the growth cone. However, a consequence of setting up my model this way was that the membrane of each quadrant moved independently from the others. As a temporary fix, I took a similar approach to Oelz, et. al. 2018 for modeling the cell membrane boundary. Each time a coordinate of the membrane protrudes or shrinks I have a few lines of code built in that will also protrude or shrink the membrane of the two surrounding coordinates at 0.25 times the rate of the center coordinate. This is a simple way to model membrane elasticity, but we are looking into other methods for modeling membranes for moving cells.
      2) The chemical gradient is modeled as the following term: (1-G*cos(theta)). We can plot this chemical “signal” as y=1-G*cos(theta) from theta at 0 degrees to 180 degrees, where the G parameter influences the total magnitude of the signal. I wish I could attach a picture to explain but the general shape of this gradient is it increases concave up from 0 degrees, has an inflection point at ~90 degrees, and then increases concave down where it eventually levels out at 180 degrees. This would show you a stronger signal on the left side of the growth cone at 180 degrees. If we changed the “-” to positive this would flip the signal, giving a stronger signal on the right side at 0 degrees. However, it would be interesting to try different “distributions” of chemicals, kind of like we did in our previous models for optical signals. I was thinking to help visualize the spatial distribution of chemicals/signal molecules across the growth cone, I could overlay a “heatmap” across the growth cone, where the color intensity/hues could represent the strength of the signal. Hopefully, that answers your questions!

      1. Hi Kahmina. You’ve answered my questions. Of course, now I am curious about other things too. I’d love to read the paper that comes out of this work. Really great job!

Leave a Reply

Your email address will not be published. Required fields are marked *