This analysis showed that, indeed, BCs follow the long axis rule in Y-27632 but not H2O-injected embryos (Figures S2ICS2K). we have discovered a population of cells in the embryonic epidermis whose mitoses CHUK do not follow the long axis rule. These cells are located at the parasegmental boundaries (PSBs) and divide perpendicular to a contractile actomyosin cable that forms at the boundary cell-cell interfaces (Monier et?al., 2010). We provide evidence that the orientation of the division plane of the boundary cells is governed directly by local tension anisotropy rather than by cell geometry or genetic cues. Results Cells Dividing at Parasegment Boundaries Do Not Follow the Long Axis Rule During embryogenesis, the epidermis undergoes waves of cell divisions at extended germband stages 9 to 11 (Foe, 1989, Martinez-Arias, 1993). PSBs form through patterning mechanisms and prevent cells or their descendants from changing compartments (Monier et?al., 2010, Vincent and O’Farrell, 1992) (Figure?1A). Here, we find that at these stages, boundary cells (BCs; cells with an edge contributing to a boundary) bias their orientation of division differently from non-boundary cells (NBCs) (Figures 1AC1C). Note that all angles are given relative to the antero-posterior (AP) axis throughout the manuscript (angle measurements are described in Figures S1A and S1B and STAR Methods). In fixed embryos, NBCs divide predominantly perpendicular to the AP axis of the embryo (Figures 1B and 1D). In contrast, BCs predominantly orient their divisions parallel to the AP axis of the embryo, perpendicular to the PSBs (Figures 1C and 1E). Moreover, this bias is AG-1478 (Tyrphostin AG-1478) the same on either side of the boundary (either or embryo when the germband (blue) is extended (stages 9 to 11). Cell divisions occur throughout the extended germband epidermis. The metameric subdivisions are the parasegments, separated by parasegment boundaries (PSBs, pink). BC, boundary cells; NBC, non-boundary cells. Examples of the planar cell division biases found in non-boundary (B) and boundary cells (C). VM, ventral midline. Scale bar, 10?m. (D) Quantification AG-1478 (Tyrphostin AG-1478) of the angle of cell division in fixed embryos relative to the antero-posterior (AP) axis in NBC (n?= 391 cell divisions) and BC (E) (n?= 289 cell divisions; Mann-Whitney test, embryo. was used to identify PSBs (not shown) and (green) to label the mitotic spindle. The orientation of cell division (pink vector) versus the orientation of interphase cell shape (white vector) is shown. Scale bar, 5?m. (G) In NBC, there is a correlation between these two angles, suggesting that these cells follow the long axis rule (n?= 77; Spearmans rho test, (Fink et?al., 2011) as well as in tissues (Campinho et?al., 2013, Mao et?al., 2013, Wyatt et?al., 2015), we hypothesized that the actomyosin cable at PSBs might act as a source of anisotropic tension during mitosis. As previously reported (Monier et?al., 2010), live imaging using GFP-tagged Myosin II Regulatory Light Chain (MRLC-GFP) and quantification of fluorescence intensity at boundary versus non-boundary interfaces of dividing cells showed that the actomyosin cable-like enrichment persists at the cortex of boundary cells during division (Figures 2C, 2D, and S2A). We asked whether the actomyosin cable is required for the division orientation bias we observed in these cells. We examined null mutant embryos, where actomyosin fails to accumulate at PSBs (Monier et?al., 2010, Tetley et?al., 2016, Urbano et?al., 2018) (Figure?2E). Strikingly, the majority of BCs now divide perpendicular to AP like NBCs (Figures 2E, S2B, and S2C). To test if this difference was caused by the loss of actomyosin enrichment in mutants, we inhibited Myosin II activity in two different ways. First, we injected wild-type embryos AG-1478 (Tyrphostin AG-1478) with a concentration of the Rok inhibitor Y-27632 that does not affect cell division but does disrupt boundary function (Monier et?al., 2010, Urbano et?al., 2018). Second, we overexpressed a dominant-negative form of the Myosin II Heavy Chain in the epidermis (Franke et?al.,.