1The budding of tubular organs from flat epithelial sheets is a vital morphogenetic 2 process. Cell behaviours that drive such processes are only starting to be unraveled.
3Using live imaging and novel morphometric methods we show that in addition to 4 apical constriction, radially oriented directional intercalation of placodal cells plays a 5 major contribution to the early stages of invagination of the salivary gland tube in the 6 Drosophila embryo. Extending analyses in 3D, we find that near the pit of 7 invagination, isotropic apical constriction leads to strong cell wedging, and further 8 from the pit cells interleave circumferentially, suggesting apically driven behaviours.
9Supporting this, junctional myosin is enriched in, and neighbour exchanges biased 10 towards the circumferential orientation. In a mutant failing pit specification, neither 11 are biased due to an inactive pit. Thus, tube budding depends on a radially polarised 12 pattern of apical myosin leading to radially oriented 3D cell behaviours, with a close 13 mechanical interplay between invagination and intercalation. 14 15 6 apical-basal depths in the cells, and uncover cell behaviours in a 3D context: near 1 the pit of invagination, where medial myosin II is strong (Booth et al., 2014), cells are 2 isotropically constricting apically leading to cell wedging, and with distance from the 3 pit cells progressively tilt towards the pit. Cells also interleave apically in a 4 circumferential direction, i.e. they contact different neighbours along their length, a 5 process that can be compared to a T1 transition in depth. This strongly suggests 6 apically driven behaviours, and we show that across the placode junctional myosin II 7 is enriched in circumferential junctions leading to polarised initiation of cell 8 intercalation. This is followed by polarised resolution of exchanges towards the pit, 9 thereby contributing to tissue invagination. forkhead mutants, that fail to form an 10 invagination, only show unproductive intercalations that fail to resolve directionally, 11 likely due to the lack of an active pit. Thus, tube budding depends on a radial pattern 12 of 3D cell behaviours, that are themselves patterned by the radially polarised activity 13 of apical myosin II pools. The continued initiation of cell intercalation but lack of 14 polarised resolution in the fkh mutant, where the invagination is lost, suggest that a 15 tissue-intrinsic mechanical interplay also contributes to successful tube budding. 16 17