To confirm the effect of the tissue boundary on the direction of EB3 movement, we carefully observed the EB3 movement in chordamesoderm cells in different positions relative to the notochord, when intercalation was almost complete. We specifically focused on cells that were in contact with the notochord-somite boundary and those that were in the middle of the notochord. Confirming our initial observation, we found that EB3 had a strong tendency to move toward the notochordsomite boundary in cells that were in contact with the boundary. For example, in Fig. 2, in the cell whose right side touches the boundary, most of the EB3 comets moved toward the boundary. The farther away from the boundary the cells were, the less obvious was the directional bias of the EB3 movement. In cells not touching the boundary, i.e., located in the center of the notochord, EB3 moved more randomly and did not show biased movement, even at the same stage. These results show that EB3 moved in an essentially monopolar manner toward the notochord GSK1120212boundary in cells near the boundary. The tracking analysis was performed for three different regions of the cell, and showed that most of the EB3 puncta originating from and passing through the center region, also moved toward the boundary side. Microtubule growth is thus strongly biased toward the notochordsomite boundary in notochord cells near the boundary, revealing one of the mechanisms for attraction. We next examined the localization of Adenomatous poliposis coli, which is known in mammalian cells to anchor polarized MTs to the cell cortex and stabilize MTs. In the presumptive notochord cells in Keller explants, APC was sharply restricted to cell tips facing the boundary. This localization was mostly observed in the cells near the boundary, and cells located at the center region of the notochord lacking the contact with the boundary, did not show polarized APC localization, which is consistent with the polarity of EB3. Furthermore, APC was either randomly localized or undetectable in normal animal cap cells. These results suggest that mesoderm differentiation is a prerequisite for the monopolarized APC localization as well as the biased MTICG-001 company elongation and that the notochord boundary has critical information for the polarization. We next examined the presumptive notochord cells in Keller explants during the early phase of CE, by observing both the cell morphology and EB3 movements over time, to determine when the polarization is initially set up. Around the onset of morphological polarization, we observed the EB3 movement by real-time confocal microscopy. We found that EB3 often showed uni-directional movement even before cell elongation.