Due to the existence of GH146-negative adPNs (Figure S1A), we repeated the above mosaic analysis with Acj6-Gal4, which permits the labeling of all adPNs ( Lai et al., 2008). We again observed loss of the same seven glomeruli accompanied by stronger labeling of the DM3 and D glomeruli (data not shown). However, we were not sure whether the polyglomerular adPNs remained Docetaxel clinical trial intact in these large NB clones, given that their diffuse dendritic elaboration is overshadowed by the much more dense uniglomerular projections of most adPNs ( Yu et al., 2010). To detect the polyglomerular PNs derived near the end of the

lineage, we went on generating NB clones at midlarval stage. We could reproducibly observe the diffuse dendritic processes characteristic of the polyglomerular PNs in mutant clones made following the birth of last GH146-positive VA1lm-targeting adPN ( Figure 1D). INCB024360 solubility dmso This concludes no involvement of Chinmo in the specification of any GH146-negative temporal cell fates, including polyglomerular PN fates, justifying the use of GAL4-GH146 in further phenotypic analysis of chinmo mutant adPN clones. Taken together, loss of Chinmo selectively eliminates eight temporal cell fates in the serial production of 40 adPN types. Intriguingly, Chinmo is required in two adjacent windows of adPN lineage

development that are interrupted by a single DM3-targeting adPN. The embryonic-derived DM4, DL5,

and VM3(a) fates reside in the first Chinmo-required window. By contrast, the subsequent window covers the VM3(b) and DL4 fates before the NB becomes quiescent at the end of embryogenesis and also includes the DL1, DA3, and DC2 fates after the quiescent NB resumes proliferation in late first-instar larvae. In addition, the loss of seven glomerular targets (eight temporal fates) was consistently accompanied by enlargement of the DM3 and D glomeruli, whose corresponding adPN fates follow the Chinmo-required windows. This implies that the missing adPN types have probably been transformed into their next Chinmo-independent neuron types in mutant clones, arguing for Chinmo as a temporal fate regulator. Knocking down Chinmo from GMCs made around the Chinmo-required windows should provide clues about the natures of suspected temporal Resminostat fate transformations. By using twin-spot MARCM, each GFP-labeled mutant GMC clone, containing only one neuron in the adPN hemilineage, is paired with an RFP-marked multicellular wild-type NB clone. This allows us to deduce the prospective cell fates for mutant GMC clones based on the subsequently derived neuron types present in their accompanying wild-type NB clones (Figure 1A, right). One can thus determine whether the GMC progeny was born with incorrect temporal identity as reflected by the actual neurite trajectories.