We are investigating the function of a homolog of Drosophila mago nashi in C. elegans (
mag-1). The Drosophila mago nashi gene (dm-mago) is required in oogenesis for localization of polar granules and posterior determinants, which in turn affect abdominal segmentation and germ-cell determination (1,2). R. Boswell and co-workers in this department, with whom we are collaborating, initially identified and sequenced a
mag-1 cDNA, predicted it to encode a small protein with 80% identity to the dm-mago product, and showed that, when driven by the dm-mago promoter, the
mag-1 cDNA is capable of rescuing both grandchildless and embryonic lethal mago nashi phenotypes in flies. We have located
mag-1 on a cosmid near
ced-1 ( between
unc-75 and
unc-101 ) on LGI . Northern blot analysis showed that a
mag-1 0.6-kb mRNA is present in all developmental stages and in the hermaphrodite germ line. A
mag-1::lacZ construct with 5 kb of upstream sequence is expressed extensively in embryos and in the dorsal and ventral cord in adults. Antisense RNA experiments suggested that
mag-1 might play role(s) in germ-line development. Injection of
mag-1 antisense RNA into the syncytial gonads of N2 hermaphrodites at L4 stage caused the animals to become sterile, first producing an average of 61 dead embryos each (n=7) and thereafter laying no more eggs. Similar results were obtained following injection of young adult hermaphrodites instead of L4 larvae. After ceasing egg laying, these sterile worms produced only severely defective oocytes, and some of the germ lines also produced excess sperm as shown by DAPI staining, suggesting that
mag-1 might be required in the decision of oogenesis versus spermatogenesis as well as for oogenesis per se. The observation that after injection, some germ-line cells did complete oogenesis and became fertilized may be due to residual
mag-1 activity accumulated before injection. Lethality in these embryos might be caused by either defects of the oocytes or reduction of
mag-1 function in embryogenesis, or a combination of both. In addition, these arrested embryos were not stained with the anti-P-granule MAb K76, suggesting that
mag-1 function may be required in P-granule formation or localization. Surprisingly, dominant negative effects similar to those seen following antisense injection were also observed in transgenic worm lines carrying either 1) a
mag-1 construct containing 0.9 kb of upstream sequence, the entire 0.7-kb coding region and 0.4 kb of 3'-flanking sequence or 2) a
mag-1::lacZ construct with the same 0.9 kb of upstream sequence. One possible explanation is that the 0.9 kb of upstream sequence competes for factors required in normal
mag-1 control. We are currently performing both genetic and Tc1 reverse-genetic screens for mutations in the
mag-1 gene in order to better define its function. (1) Boswell, R., Prout, M. and Steichen, J. (1991) Development 113: 373-384. (2) Newmark, P. and Boswell, R. (1994) Development 120: 1303-1313.