Skip to the content of the web site.
 brodland collage

Video Force Microscopy

Some years ago we wondered if it would be possible to determine the forces that drive morphogenetic movements from digitized time-lapse images of those processes. Over the course of several years we tried a number of mathematical formulations, but had limited success. Eventually, after nearly giving up several times, we discovered an approach that works. We called this new approach cinemechanometry (CMM) [1] or Video Force Microscopy (VFM) [2]. The basic steps in VFM (Fig. 1) are:

  1. Collect time-lapse images that are of sufficient quality that points on the cells can be tracked over time, ideally using computer algorithms.
  2. Corresponding points in successive images are then digitized. The spatial and temporal densities of these points put a limit on the precision with which the driving forces can be obtained in subsequent steps. For purposes of this explanation, each cell triple junction is assumed to be tracked in each time-lapse image, and the goal is to obtain the equivalent forces that would have to act along each cell edge to produce the observed motions.
  3. The passive mechanical properties of the cytoplasm and the networks and other structures it might contain are assumed to be known and a numerical method, such as finite elements (FE), is used to calculate the forces that would have to be applied to each triple junction (or digitized point) to produce the observed deformations of each cell (or digitized region). The forces needed at each triple junction are obtained by vector addition of the forces needed to deform each of the cells associated with that triple point. One could interpret these forces as those that an external mechanism would have to apply at each triple junction to produce the observed cellular deformations and motions.
  4. In fact, these forces are produced by structures such as microfilaments and microtubules inside the cells, and by adhesions between cells. These structures generate active forces. It can be shown mathematically, that they can be resolved into equivalent edge forces. The fourth step of the method is to use mathematical inverse methods to find the combination of edge forces that will produce the vectorial forces needed at each triple junction, as calculated in Step 3.
Figure 1
Fig. 1 - The basic steps in VFM
 

To assess the accuracy of VFM, we applied it to synthetic data generated using computer simulations of a large number of real and idealized tissue movements. The advantage of using synthetic data is that the true driving forces were known so that the accuracy of the VFM output could be verified [2].

When we then applied VFM to multi-photon cross-sections of Drosophila embryos during the period of time when they form a ventral furrow [1]. VFM was able to determine the equivalent edge forces needed to drive the observed motions, and it was able to do so with sub-minute accuracy.

Movie 1 - A normal embryo. On the left is a multi-photon image of a cross-section of a Drosophila embryo. On the right is the VFM output, which shows the edge forces that must act during the 45sec time interval between each successive pair of images. The result is a detailed map of the forces that drive ventral furrow formation. Forces are shown in colour, with those closest to the red end of the spectrum corresponding to the largest tensile forces. Movies of several mutant strains are also available.

References

1. Cranston, P.G., Veldhuis, J.H., Narasimhan, S. and Brodland, G.W., 2010, "Cinemechanometry (CMM): A Method to Determine the Forces Driving Morphogenetic Movements from Time-lapse Images," Annals of Biomedical Engineering , Vol. 38, pp. 2937-2947. doi: 10.1007/s10439-010-9998-1.

2. Brodland, G.W., Conte, V. Cranston, P.G., Veldhuis, J., Narasimhan, S., Hutson, M.S., Jacinto, A., Ulrich, F., Baum, B., and Miodownik, M., 2010, "Video Force Microscopy Reveals the Mechanics of Ventral Furrow Invagination in Drosophila," Proceedings of the National Academy of Sciences (PNAS), Vol 107, No. 51, pp. 22111-22116. doi: 10.1073/pnas.1006591107 (open access).