Nucleotide sequence analysis of the similar to 35-kb segment containing interferon-gamma-inducible mouse proteasome activator genes

The proteasome activator PA28 is an interferon-gamma -inducible complex mad e up of two related subunits, named PA28 alpha and PA28 beta, with approxim ately 50% amino acid sequence identity. Accumulated evidence indicates that binding of this complex to the 20S proteasome enhances the generation of c lass I-binding peptides. Previously, we showed that the genes coding for PA 28 alpha and PA28 beta, designated Psmc1 and Psmc2, respectively, are locat ed similar to6 kb apart with their 3' ends pointing toward each other on mo use Chromosome 14. In the present study, we sequenced the regions adjacent to Psmc1 and Psmc2. In a contiguous stretch of similar to 35 kb, we identif ied six genes arranged in the following order: Cg10671-like (a gene similar to Drosophila CG10671)-Psmc1-Cgi112 (a ubiquitously expressed gene with no known function)-Psmc2-Flj10111 (a gene coding for a protein with two RING finger domains)-Isgf3g (an interferon-gamma -inducible gene coding for an i nterferon-dependent, positive-acting transcription factor 3 gamma). Interes tingly, the 3' untranslated region of Psmc1 overlaps with that of Cgi112 by 7 bp. Database analysis indicates that the corresponding human genes also overlap by up to 7 bp in their 3' untranslated regions. The 5' end of the m ouse, but not the human, gene coding for PA28 beta undergoes alternative sp licing that is predicted to alter the N-terminal amino acid sequence. Compa rison of the mouse sequence with a human draft sequence deposited in the NC BI database revealed that the overall organization of the region coding for the interferon-gamma -inducible proteasome activator is conserved between human and mouse.