The direction of tracheal tube extension is defined by migration of interconnected precursors toward an FGF signal. But tube morphology is also affected by Dpp and Wnt signaling. The Swiss team wanted to know what cellular processes were being controlled by Dpp and Wnt.
They made movies of living fly embryos with labeled adherens junctions (AJs). The label showed cells that were initially paired, side-by-side, in wider tubes. But gradually individual cells reached around at one end until they made contact with themselves, and then started replacing their intercellular AJs with autocellular AJs. This squeezing process continued until the tube was longer but only one cell circumference around.
Expression of the Wnt target Spalt, a transcription factor, was necessary and sufficient to prevent this intercalation process in tubes that remained larger. Dpp's normal function, by contrast, is to repress Spalt expression and thus allow intercalation.
Cells formed autocellular AJs as long as they had a Spalt-free neighbor. So perhaps it is contraction by the neighbor rather than outgrowth by the autocellular AJ-forming cell that is the driving force. Affolter plans to label old and new AJs to help him determine which cell is controlling the process and whether cell sliding, rolling, or detachment are involved in the reassortment of junctional contacts.