The following text replaces the corresponding text in the Materials and Methods section, which incorrectly described UAS>painlessFL constructs: "The full length painless cDNA from BDGP gold collection (RE03641) has been used as a template. In this cDNA, a TN10 transposon was found to be inserted following nucleotide 1795. To remove the TN10 transposon, we have used a two step PCR strategy. Painless sequences flanking the TN10 sequence were amplified using primers couples (forward/reverse): caaggagacgcagcgcgtcttagccg/ggtaccaacatttgaaattaaatatttactc and gcggccgcggtcgttgtctggatattaa /cggctaagacgcgctgcgtctccttg. Next, the products of these two PCR were mixed to use as a new template and amplified using the following primers couple (forward/reverse): ttaatagcggccgcggtcgttgtctggatattaa/ttaataggtaccaacatttgaaattaaatatttactc. NotI and KpnI restriction sites were respectively included in those forward and reverse primers. The PCR product was verified by sequencing and revealed a single silent mutation at nucleotide 1605 changing C in T. It was subsequently cloned into a KpnI/NotI-cut pUAST-attb transformation vector [46]. This pUAST-attb-pain is available on request."