Most cell types in multicellular eukaryotes exit from the mitotic cell cycle before terminal differentiation. We show that the dacapo gene is required to arrest the epidermal cell proliferation at the correct developmental stage during Drosophila embryogenesis. dacapo encodes an inhibitor of cyclin E/cdk2 complexes with similarity to the vertebrate Cip/Kip inhibitors. dacapo is transiently expressed beginning late in the G2 phase preceding the terminal division (mitosis 16). Mutants unable to express the inhibitor fail to arrest cell proliferation after mitosis 16 and progress through an extra division cycle. Conversely, premature dacapo expression in transgenic embryos results in a precocious G1 arrest.

Reduced protein kinase A (PKA) activity in anterior imaginal disc cells leads to cell-autonomous induction of decapentaplegic (dpp), wingless (wg), and patched (ptc) transcription that is independent of hedgehog (hh) gene activity. The resulting nonautonomous adult wing and leg pattern duplications are largely due to induced dpp and wg expression and resemble phenotypes elicited by ectopic hh expression. Inhibition of PKA in anterior cells close to the posterior compartment can substitute for hh activity to promote growth of imaginal discs, whereas overexpression of PKA can counteract transcriptional induction of ptc by hh in these cells. PKA therefore appears to be an integral component of the mechanism by which hh regulates the expression of key patterning molecules in imaginal discs.

The cAMP-dependent protein kinase (PKA) has been shown to mediate the vast majority of cellular responses to the intracellular second messenger, cAMP, in eukaryotes. To study the role of cAMP signal transduction in Drosophila development, we have isolated and molecularly characterized mutations of varying severity in the Drosophila PKA gene, DC0. Biochemical measurements indicate that DC0 is either the sole or the major PKA catalytic subunit gene in Drosophila. Adult females heterozygous for a strong and a weak DC0 allele fail to lay eggs and show a striking and novel defect in oogenesis that includes the formation of egg chambers containing multinucleate nurse cells. Females heterozygous for two weak DC0 alleles are fertile but produce offspring showing a variety of defects in embryogenesis, including preblastoderm arrest and alterations in cuticular patterning. Animals zygotically null for DC0 die as morphologically normal first-instar larvae, implying that maternally encoded protein, which perdures for at least 12 hr, suffices for embryogenesis. Animals hemizygous for weak DC0 alleles survive for several days as larvae but grow slowly. Mitotic recombination experiments in the adult eye indicate that the DC0 gene is not required autonomously either for cell viability or normal growth rates. These results argue that cAMP-mediated signal transduction is essential at a variety of stages during the development of a metazoan.

Mixed lineage kinases (MLKs) are MAPKKK members that activate JNK and reportedly lead to cell death. However, the agonist(s) that regulate MLK activity remain unknown. Here, we demonstrate ceramide as the activator of Drosophila MLK (dMLK) and identify ceramide and TNF-alpha as agonists of mammalian MLK3. dMLK and MLK3 are activated by a ceramide analog and bacterial sphingomyelinase in vivo, whereas a low nanomolar concentration of natural ceramide activates them in vitro. Specific inhibition of dMLK and MLK3 significantly attenuates activation of JNK by ceramide in vivo without affecting ceramide-induced p38 or ERK activation. In addition, TNF-alpha also activates MLK3 and evidently leads to JNK activation in vivo. Thus, the ceramide serves as a common agonist of dMLK and MLK3, and MLK3 contributes to JNK activation induced by TNF-alpha.

Cell cycle arrest in G1 at the onset of patterning in the Drosophila eye is mediated by roughex. In roughex mutants, cells accumulate Cyclin A protein in early GI and progress into S phase precociously. When Roughex is overexpressed in S/G2 cells, Cyclin A is mislocalized to the nucleus and degraded, preventing mitosis. Whereas Roughex inhibits Cyclin A accumulation, Cyclin E down-regulates Roughex protein in vivo. Roughex binds to Cyclin E and is a substrate for a Cyclin E-Cdk complex in vitro. These data argue that Roughex inhibits Cyclin A accumulation in early G 1 by targeting Cyclin A for destruction. In late G1, Roughex is destabilized in a Cyclin E-dependent process, releasing Cyclin A for its role in S/G2.