Fig. 2. hlh-14 gene structure, predicted protein sequence and comparison with other A S family members. (A) Schematic representation of hlh-14 mRNA. By 5' and 3' RACE, we detected three different hlh-14 cDNAs, predicting three different forms of HLH-14 protein (Materials and methods). Analyses of these cDNA species indicate that hlh-14 mRNA is not trans-spliced. The three forms of mature hlh-14 mRNA have unique start codons (AUG) but encode the same bHLH domain and the same novel C terminus. The shortest form encodes a 148 amino acid protein. The 5' end of the corresponding short cDNA is the only 5' end we could amplify out of a RACE cDNA library without performing a second, nested PCR reamplification (Materials and methods). Therefore, it is likely that this form is the most abundant form of hlh-14 mRNA in C. elegans. In the corresponding protein, there are only eight amino acids N-terminal of the bHLH domain (MAKKNQVA). (B) Primary amino acid sequences of HLH-14. The three different start methionines are enlarged. For the two longer forms of HLH-14, alternative beginning peptides are boxed. The leucine and glutamine residues changed, respectively, by the hlh-14(gm34) missense mutation and the hlh-14(ju243) nonsense mutation are enlarged and underlined. Both lesions are in the conserved bHLH domain. (C) Alignment of the bHLH domains of HLH-14, HLH-3 (C. elegans), Achaete (Drosophila), Scute (Drosophila) and Mash1 (Ascl1; mouse). Residues shaded in black are identical. Lightly shaded residues are similar to corresponding black residues. Consensus residues are present in at least three of the five proteins aligned. The residues affected by the hlh-14(gm34) and hlh-14(ju243) mutations are detailed.