The direction in which a cell divides is determined by the orientation of its mitotic spindle at metaphase. Spindle orientation is therefore important for a wide range of developmental processes, ranging from germline stem cell division to epithelial tissue homeostasis and regeneration. In multiple cell types in multiple animals, spindle orientation is controlled by a conserved biological machine that mediates a pulling force on astral microtubules. Restricting the localization of this machine to only specific regions of the cortex can thus determine how the mitotic spindle is oriented. As we review here, recent findings based on studies in tunicate, worm, fly and vertebrate cells have revealed that the mechanisms for mediating this restriction are surprisingly diverse.

We investigated the cell behaviors that drive morphogenesis of the Drosophila follicular epithelium during expansion and elongation of early‐stage egg chambers. We found that cell division is not required for elongation of the early follicular epithelium, but drives the tissue toward optimal geometric packing. We examined the orientation of cell divisions with respect to the planar tissue axis and found a bias toward the primary direction of tissue expansion. However, interphase cell shapes demonstrate the opposite bias. Hertwig's rule, which holds that cell elongation determines division orientation, is therefore broken in this tissue. This observation cannot be explained by the anisotropic activity of the conserved Pins/Mud spindle‐orienting machinery, which controls division orientation in the apical–basal axis and planar division orientation in other epithelial tissues. Rather, cortical tension at the apical surface translates into planar division orientation in a manner dependent on Canoe/Afadin, which links actomyosin to adherens junctions. These findings demonstrate that division orientation in different axes—apical–basal and planar—is controlled by distinct, independent mechanisms in a proliferating epithelium.

In animal cells, mitotic spindles are oriented by the dynein/dynactin motor complex, which exerts a pulling force on astral microtubules. Dynein/dynactin localization depends on Mud/NUMA, which is typically recruited to the cortex by Pins/LGN. In Drosophila neuroblasts, the Inscuteable/Baz/Par-6/aPKC complex recruits Pins apically to induce vertical spindle orientation, whereas in epithelial cells Dlg recruits Pins laterally to orient the spindle horizontally. Here we investigate division orientation in the Drosophila imaginal wing disc epithelium. Live imaging reveals that spindle angles vary widely during prometaphase and metaphase, and therefore do not reliably predict division orientation. This finding prompted us to re-examine mutants that have been reported to disrupt division orientation in this tissue. Loss of Mud misorients divisions, but Inscuteable expression and aPKC, dlg and pins mutants have no effect. Furthermore, Mud localizes to the apical-lateral cortex of the wing epithelium independently of both Pins and cell cycle stage. Thus, Pins is not required in the wing disc because there are parallel mechanisms for Mud localization and hence spindle orientation, making it a more robust system than in other epithelia.

The function of an epithelial tissue is intertwined with its architecture. Epithelial tissues are often described as pseudo-two-dimensional, but this view may be partly attributed to experimental bias: many model epithelia, including cultured cell lines, are easiest to image from the “top-down.” We measured the three-dimensional architecture of epithelial cells in culture and found that it varies dramatically across cultured regions, presenting a challenge for reproducibility and cross-study comparisons. We therefore developed a novel tool (Automated Layer Analysis, “ALAn”) to characterize architecture in an unbiased manner. Using ALAn, we find that cultured epithelial cells can organize into four distinct architectures and that architecture correlates with cell density. Cells exhibit distinct biological properties in each architecture. Organization in the apical-basal axis is determined early in monolayer development by substrate availability, while disorganization in the apical-basal axis arises from an inability to form substrate connections. Our work highlights the need to carefully control for 3D architecture when using cell culture as a model system for epithelial cell biology and introduces a novel tool, built on a set of rules that can be widely applied to epithelial cell culture. [Media: see text] [Media: see text] [Media: see text] [Media: see text] [Media: see text] [Media: see text]

Organ surfaces are lined by epithelial monolayers - sheets of cells that are one-cell thick. This architecture underlies tissue function, and its loss is associated with disease, including cancer. Studies of in-plane epithelial cell behaviors show that a developing epithelium behaves as a fluid in respect to the tissue plane, and can therefore readily adapt to varying mechanical influences during morphogenesis. We asked the question of how monolayer architecture is achieved, and whether it demonstrates the same fluid behavior. To address this problem, we cultured MDCK (Madin-Darby Canine Kidney) cell layers at different densities and timepoints and analyzed their architectures using a novel tool, Automated Layer Analysis (ALAn), which we introduce here. Our experimental and theoretical results lead us to propose that epithelial monolayer architecture is governed by a balance of counteracting forces due to cell-cell and cell-substrate adhesion, and that this balance is influenced by cell density. MDCK cells do not undergo obvious rearrangement along the apical-basal axis; instead, cells that do not contact the substrate aggregate on top of the monolayer. Our findings therefore imply that monolayered architecture is under more rigid control than planar tissue shape in epithelia.