Novel Image Analysis for Trajectory of Microtubules Gliding on Kinesins with Tip Detection

Abstract

Control of the gliding directions of kinesin-driven microtubules (MTs) in vitro has good feasibility for the development of nano-scale transport systems. A requirement for the control of transporters in these systems includes detecting the positions of gliding MTs; however, no studies have reported on the monitoring of the positions of gliding MTs. Here, we suggest an algorithm to detect tip coordinates of gliding MTs by binarization, skeletonization, and filtration of fluorescent images of MTs. The algorithm was first applied to artificially drawn segments with given lengths (10-80 pixels), widths (1-10 pixels), and curvature radii (20-120 pixels) to verify the effect of the sizes of MTs on accuracy of tip coordinates extracted by the algorithm, and error was estimated by referring to the true coordinates. The estimated errors were as small as 2 pixels in the width and were not affected by the length and the curvature radius, indicating that our algorithm is useful to extract the tips of MTs. The algorithm was subsequently applied to images of gliding MTs. Since distances from the trajectories of the MTs to the centers of gravity of the MTs (3.7 ± 2.1 pixels) were significantly larger than those to the tips (1.9 ± 0.5 pixels), the use of the tips as representative points of gliding MTs was verified. A detection method using tips of MTs, as suggested in this study, may be a useful technique for monitoring each MT in nanoscale transport systems.

Keywords

• image analysis
• tip detection
• microtubule
• kinesin
• nano-scale transport system