Fig. 5. Olfactory precursors in amos and ato mutants. These discs should be compared with the corresponding wild-type discs in Figs 3 and 4. (A,D) The early precursors are specifically lost in ato mutants. The remaining olfactory precursors correspond to the third wave (B,C) and align very closely with the amos expression domains (E,F). In the second segment, the chordotonal precursors are also missing and only a few bristle precursors remain (*, A-C). (G-L) The late precursors are specifically lost in amos mutants. (G,J) Early precursor pattern resembles wild type, with mutant Amos¹ protein detectable between the rows of precursors (brackets). Caret in G indicates chordotonal precursors. (H,K) At 8 hours APF, the pattern remains unchanged as the third wave SOPs are not formed (c.f. Fig. 4C). These early precursors mostly express Ato, although a number of non-Ato expressing SOPs appear between the early rows, which could correspond to the bristle SOPs (arrows in K). (I,L) The early pattern is still apparent at 16 hours APF as it has not been obliterated by the third wave of SOPs (the early SOPs have now been replaced by PSCs, some of which are ringed). (M,N) Cut expression appears prematurely activated in amos mutants. (M) Wild type at 8 hours APF. No Cut expression is detectable in the third segment SOPs; however, Cut stains very strongly in the surrounding tissue (caret). (N) amos¹ mutant at 8 hours APF. Some Cut labelling (arrows) appears in SOPs derived from the Amos-expressing domains (in this case expressing non-functional Amos¹ protein). These cells seem to correspond to the ato-independent cells in K.