Gene-expression controls in tissue, stage and sex specificities are one of the most interesting subjects. Bi-directional promoter is defined as 1) two genes are located bi-directionally 2) both promoters are overlapped each other. Bi-directional promoter is found in lambda phage, bacteria, fruit fly, mouse, and human. None of more than 1,000 candidates in the C. elegans genome was analyzed. We found
spe-17 -like gene (designated spe-TI ) and
sqt-3 (
col-1) -like gene (designated col-TI ) in the cosmid clone C46C8 maps in the center of chromosome V of C. elegans . SPE-TI is a Ser/Thr rich hydrophilic protein and COL-TI is a cuticular collagen similar to SQT-3 at the rate of 84%. Both genes are separated by 1.3 kb and are transcribed bi-directionally. It is of interest to know how and which factors control the gene expression of both. Using the promoter/ lacZ fusion plasmids, we analyzed the promoter activity of each of genes. spe-TI was expressed in about 80 cells of spermatheca after 4th larval stage of hermaphrodite under the control of 0.7kb upstream fragment. col-TI was expressed only in the male tail tip of adult under the control of the 0.6kb upstream fragment. These results suggest that both promoters may overlap to each other and both genes are regulated by sex-specific bi-directional promoter. Transcription factors regulating spe-TI and col-TI were identified with yeast one-hybrid screen system by using the promoter specific regions. After screening approximately 1.4x10 6 yeast transformants for spe-TI and 8.5x10 5 yeast transformants for col-TI , 4 plasmids and 5 plasmids were identified respectively. Surprisingly, all 9 plasmids came from cDNA of
mab-18 isoform II promoting male tail
ray6 formation and are most closely related to the vertebrate Pax-6. In vertebrate, Pax-6 is used for eye crystal formation with HMG class transcription factors but did not use for spermatogenesis nor collagen gene expression. We also found the Pax-6 homeodomain binding sequence on 1.0kb upstream of col-TI and 1.1kb upstream of spe-TI respectively. col-TI::lacZ was not expressed in
mab-18 mutant animals but heat shock promoter derived ectopic MAB-18 cannot activate col-TI::gfp expression ectopically. These results suggest that MAB-18 stimulates col-TI expression directly in tissue or stage-specifically, but another factor(s) are necessary for normal col-TI expression. We believe that bi-directional promoter presented here is a useful example of understanding the sex-specific morphological changes in terminal differentiation.