Fig. 1. rdx gene structure. (A) The rdx genomic locus spans 64 kb in cytological band 88A4. Exons are represented by blue boxes and numbered underneath; lines indicate splices; light blue indicates start codon; red indicates stop codon; arrows indicate direction of transcription (towards the centromere). Four alternative transcriptional start sites generate five transcripts (forms A-E) that share exons 7-13. Exons 1 and 2 are unique to the B form. Exons 3 and 4 are unique to the C/D forms. C differs from D by lacking seven nucleotides in the 5'UTR at the end of exon 3. Exon 5 is unique to the E form (represented by our cDNA 5-2; predicted to be the nested gene CG12537). Exon 6 is unique to the A form. (B) Mutants are mapped onto the gene structure. Exons are indicated by boxes, where grey indicates UTRs, red indicates coding sequence unique to each form, green indicates coding sequence for the conserved MATH domain and magenta indicates coding sequence for the conserved BTB domain. The asterisk identifies alternative termination for 3' UTRs. rdx¹ is a LacW insertion 33 nt upstream of the start of exon 5. rdx⁴ is a complex rearrangement of the region upstream of exon 5, generated by imprecise excision of rdx¹. Black arrowheads indicate the missense mutations in rdx⁵ (G to T in exon 5 encoding Q193H and A to T in exon 10 encoding D653V), and the splice junction mutation in rdx⁶ (an A to G change in the 23rd nt of the intron between exons 12 and 13). (C) Block diagrams indicating the four proteins encoded by the rdx locus. The numbers above each indicates the amino acid coordinates. Unique sequences are in red; the MATH domain is in green; the BTB domain is in magenta. The arrow indicates the convergence of amino acid sequences.