Recently, Mark Goebl (University of Washington, Seattle, WA) has pointed out to us that the amino acid sequence of Scl1 shares 28% identity, across the entire coding region, with the 35-kDa subunit of the Drosophila proteasome (Haas et al., 1989), part of a multicatalytic nonlysosomal protease complex. Proteasome proteins are highly conserved in eukaryotes ranging from yeast to man (Tanaka et al., 1988). Thus, it is possible that Scl1 is an essential part of this complex in yeast. In Drosophila, these proteins are developmentally regulated (Haas et al., 1989). We note also that further inspection of the scl1 open reading frame suggests two possible sites of initiation of translation. The first ATG is flanked by nucleotides (nt) (a T at position -3 and an A at +4) that are not preferred for efficient initiation (Cigan and Donahue, 1989). In contrast, a second ATG, located 87 nt downstream, is flanked by favored nt, a G at -3 and an A at +4. The second ATG is located downstream from the secretory signal peptide sequence, suggesting that the Scl1 protein translated from the second ATG would remain cytoplasmic. Recent observations have raised the possibility of the existence of two scl1 gene transcripts, whose translation start sites could correspond to the two ATG mentioned above. © 1990.