We use Drosophila oogenesis as a model to explore the developmental regulation of the cell cycle. The long-term goal of the laboratory is to understand how the cell cycle program of the Drosophila ovarian cyst is coordinated with the developmental events of oogenesis. The Drosophila oocyte develops within the context of a 16-cell germline cyst. Individual cells within the cyst are referred to as cystocytes and are connected by actin-rich ring canals. The cyst is produced through a series of four synchronous mitotic divisions during which cytokinesis is incomplete. While all 16 cystocytes enter premeiotic S phase, only a single cell remains in the meiotic cycle and becomes the oocyte. The other 15 cells enter the endocycle and develop as highly polyploid nurse cells. Currently, we are working to understand how cells within ovarian cyst enter and maintain either the meiotic cycle or the endocycle. In addition, we are examining how such cell cycle choice influences the nurse cell/oocyte fate decision.

missing oocyte, a New Gene Required for the Maintenance of the Meiotic Cycle
Iida
To identify the pathways that direct entry into and maintenance of the meiotic cycle in the single pro-oocyte, we screened for mutants in which all 16 cells enter the endocycle and develop as nurse cells. From the screen, we identified a new gene, missing oocyte (mio), that is required for the maintenance of the meiotic cycle. In mio mutants, the oocyte enters the meiotic cycle, forms mature synaptonemal complexes, and progresses to pachytene. However, the meiotic state is not maintained. mio oocytes eventually abandon the meiotic cycle, enter the endocycle, and develop as nurse cells.

We have characterized the mio gene at the molecular level. mio is predicted to encode a protein of 867 amino acids; the protein contains four WD40 repeats and is conserved from yeast to humans. We have determined that the mio protein is present at high levels in the oocyte nucleus soon after it enters prophase of meiosis I. Double labeling with anti-antibodies and an antibody against the synaptonemal complex protein C(3)G indicates that mio specifically localizes to the nucleus of the oocyte soon after the completion of premeiotic S phase. As a result, for the Drosophila oocyte, mio stands apart as one of the earliest nuclear markers that is not a known component of the synaptonemal complex.

Genetic interaction studies indicate that mio functions early in meiosis, before the onset of pachytene. In egalitarian (egl) mutants, all 16-cyst cells enter the meiotic cycle and progress to pachytene as assayed by the presence of mature synaptonemal complexes. However, the meiotic state cannot be maintained, and eventually all the cells exit meiosis and enter the endocycle. Germline cysts from mio, eglnull double mutants have a significantly stronger phenotype. The mio, eglnull double mutants never form mature synaptonemal complexes and do not progress past zygotene. The data indicate that mio influences the oocyte cell cycle in early prophase of meiosis I; however, precisely when mio is required to maintain the meiotic cycle remains to be determined.

In further support of a role for mio in early meiosis, mutations in mei-W68 and mei-P22 suppress the mio 16-nurse–cell phenotype. mei-W68 and mei-P22 are required for the production of double-stranded breaks during meiosis. In addition, mio is partially suppressed by mutations in mei41. mei41 is required for the prophase I meiotic checkpoint arrest. The data suggest that mio may interact with pathways that influence DNA metabolism or the meiotic checkpoint. Further studies of mio may help elucidate the poorly characterized pathways that control meiotic progression and the maintenance of oocyte identity.

Two Novel Functions for the p27KIP1-Like CDK Inhibitor Dacapo in Ovarian Cysts
Hong, Iida, Sugimura
dacapo is a vital gene that encodes a p21CIP/p27KIP1/p57KIP2-like cyclin-dependent kinase inhibitor (cki), which specifically inhibits the activity of CycE/Cdk2 complexes. CycE/Cdk2 activity is required for S phase in Drosophila. Other researchers have been shown that, throughout much of the growth phase of Drosophila oogenesis, the levels of the cki Dacapo oscillate in the 15-polyploid nurse cells but remain persistently high in the single oocyte. We have demonstrated that both modes of Dacapo regulation are functionally important. In the oocyte, the prophase I arrest is lost or not properly established in germline cysts that lack Dacapo. This is the first demonstration of a cip/kip family member functioning in a normal meiotic cycle. In addition, our data indicate that Dacapo is part of the biochemical oscillator that drives the nurse-cell endocycle. Specifically, we find that, in polyploid nurse cells, the oscillations of Dacapo facilitate the relicensing of DNA replication origins during endoreplication by inhibiting CycE/Cdk2 activity at the end of each endocycle S phase. Our data are consistent with recently proposed models that suggest the periodic expression of members of the cip/kip family of Cdk inhibitors’ direct entry into the Gap phase during endoreplicative cycles. We propose that it is through the differential regulation of the cki Dacapo that two dramatically different cell cycles, the meiotic cycle and the endocycle, are independently maintained within the common cytoplasm of the ovarian cyst.

The Drosophila Fusome Contains a Highly Interconnected Endoplasmic Reticulum
Iida; in collaboration with Lippincott-Schwartz
We have entered into collaboration with the laboratory of Jennifer Lippincott-Schwartz to use live imaging of proteins with green fluorescent protein (GFP) to study the structure and function of the Drosophila fusome. The fusome is a membranous organelle comprised of microtubules and an array of membrane-associated cytoskeletal proteins; it grows along the remnants of the mitotic spindle after each cyst division. In the absence of the fusome, which physically connects all the cells in the cyst, mitotic synchrony between cystocytes is lost, the number of cystocyte divisions is reduced, and oocyte differentiation does not occur. However, the molecular mechanisms by which the fusome influences the cystocyte cell cycle and oocyte differentiation are poorly understood.

To understand more fully the origin and function of the membrane component of the fusome, we generated transgenic lines carrying a lysozyme-GFP-KDEL (lys-GFP-KDEL) chimera. In mammalian cells, the chimera is retained in the endoplasmic reticulum (ER) lumen by the COOH-terminal KDEL motif, which, in conjunction with the KDEL receptor, targets soluble proteins to the ER lumen. The targeting system is highly conserved. The Drosophila ortholog of the well-studied ER lumenal protein BiP contains the KDEL targeting sequence. Consistent with the expected location of the ER, our observation of live germaria indicates that the lys-GFP-KDEL protein targets to a series of whip-like tubules in the cytoplasm of germline cells. In addition to localization in cytoplasmic structures, lys-GFP-KDEL targets to thick branching structures that resemble fusomes. We confirmed that the lys-GFP-KDEL chimera targets to fusomal membranes by fixing and staining the lys-GFP-KDEL–expressing ovaries with the anti-fusome marker Hts. The targeting of the lys-GFP-KDEL transgene to the fusome strongly suggests that the fusomal membranes arise from the ER.

Electron micrographs suggest that fusomal membranes are discontinuous fragments. Reports of fragmented ER structures in other organisms are extremely rare, and live cell imaging studies have often revealed these fragmented ER structures to be fixation artifacts. To probe fusome membrane continuity, we performed fluorescence recovery after photo-bleaching (FRAP) of the GFP-chimeras in fusomes. Both visually and quantitatively, the GFP-chimera can be observed rapidly recovering into the photobleached region of the fusome. The fluorescence recovery occurs on both sides of the bleached region, indicating that recovery is not directional. This rapid bidirectional recovery is inconsistent with the fact that the fusome is composed of discrete membranous compart-ments.

Finally, we performed fluorescence loss in photobleaching (FLIP) to assay the continuity of fusome and cytoplasmic ER membranes throughout the cyst. FLIP involves repeatedly photobleaching a small region of interest. If proteins within the photobleached region are completely mobile and the entire structure is continuous with the photobleached region, then the structure’s fluorescence will be depleted over time. When a small region of cytoplasmic ER from a single cystocyte was repeatedly photobleached, the ER in the entire cyst, including the fusome, rapidly lost fluorescence, whereas adjacent cysts were unaffected.

Our work has led to three important conclusions: the fusomal membranes are derived from the ER; the fusome is continuous with the cytoplasmic ER; and all the cells within a single ovarian cyst share a common ER. We predict that the inter-connectivity of the cyst ER, through the fusome, may be important to the synchronization of the mitotic cyst divisions as well as to other signalling events within the cyst.

twin, Putative Ortholog of the Yeast CCR4 Gene
Hong; in collaboration with Lehmann
Mutations in the twin gene cause an array of phenotypes that suggest that the cell cycle program of the ovarian cyst has been altered. First, twin egg chambers frequently contain either too few or two many cells, indicating that the cyst has undergone an inappropriate number of mitotic divisions. Second, approximately five percent of twin 16-cell cysts contain two oocytes. Finally, the nurse cells in twin mutants frequently exhibit abnormal chromatin structure. In collaboration with Ruth Lehmann’s laboratory, we have determined that twin is the putative ortholog of the S. cerivisiae CCR4 gene. In yeast, CCR4 is the catalytic subunit of the major cytoplasmic mRNA deadenylase. Deadenylation influences both mRNA function and stability. In addition, CCR4 has been implicated in trans-criptional initiation and elongation. The catalytic domain, which contains 3'-5'poly(A)RNA and ssDNA exonuclease activity, is highly conserved between CCR4 and twin. We are currently investigating whether the twin phenotype is dominantly modified by mutations in known cell cycle genes and by components of the trans-criptional machinery that genetically interact with CCR4 in yeast.