Figure 12. Model of membrane-dependent and membrane-independent Hh protein signaling from posterior producing cells to distant receiving cells and their potential congruency.

(a) Schematic of a third instar wing disc with the Hh-producing posterior compartment labeled in green and the Ptc-receptor-producing anterior compartment labeled in orange. The a/p compartment border crossed by spreading Hh is shown as a dashed line. (b) Line drawing of a vertical section of the wing disc at a site marked with a red line in (a). Two pools of Hh are secreted from posterior columnar cells of the wing imaginal disc by different mechanisms (dashed circles). One pool is released from the apical side of the polar epithelium using lipophorins as hydrophilic carriers (b′) (Panáková et al., 2005) or via proteolytic processing of lipidated Hh membrane anchors, as suggested in this work (b′′). Subsequently, unprocessed lipidated Hh or processed ectodomains diffuse through the fluid-filled peripodial space (labeled blue) to bind to Ptc receptors expressed in the anterior compartment. Another pool of apical cell-surface Hh is internalized and re-secreted apically (D'Angelo et al., 2015) or basolaterally (Callejo et al., 2011) using exosomes (b′′′) (Matusek et al., 2014) or long cellular protrusions, known as cytonemes, as carriers. Cytonemes can extend from posterior Hh-producing cells to deliver Hh to cell surface receptors on receiving anterior cells in their close vicinity, or can meet ‘receiving’ cytonemes extending from more distant anterior cells at defined contact sites (b′′′′) (González-Méndez et al., 2017). Receiving anterior cytonemes that take up Hh from basal subcellular sites of expressing posterior cells for subsequent intracellular transport to the apical pole of anterior epithelial cells have also been described (b′′′′′) (Chen et al., 2017). However, an explanation is needed about how lipidated Hh can ‘switch’ between sending and receiving cytonemes, or relay from vesicular or cytoneme membranes to their receptors. This problem may be solved by proteolytic Hh processing, resulting in its relay between cytonemes or from producing cell membranes to Ptc receptors on receiving anterior cells. We note that the findings presented in this work do not support the alternative possibility, that is, that different transport modes work in parallel, because HA insertion into the N-terminal Hh processing site abolished (most) Hh-dependent patterning of the L3-L4 intervein region in a dominant-negative manner.