A Genetic Screen for Suppressors and Enhancers of the Drosophila PAN GU Cell Cycle Kinase Identifies Cyclin B as a Target

Corresponding author: Terry L. Orr-Weaver, Whitehead Institute, 9 Cambridge Ctr., Cambridge, MA 02142., weaver{at}wi.mit.edu (E-mail)

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The early cell cycles of Drosophila embryogenesis involve rapid oscillations between S phase and mitosis. These unique S-M cycles are driven by maternal stockpiles of components necessary for DNA replication and mitosis. Three genes, pan gu (png), plutonium (plu), and giant nuclei (gnu) are required to control the cell cycle specifically at the onset of Drosophila development by inhibiting DNA replication and promoting mitosis. PNG is a protein kinase that is in a complex with PLU. We employed a sensitized png mutant phenotype to screen for genes that when reduced in dosage would dominantly suppress or enhance png. We screened deficiencies covering over 50% of the autosomes and identified both enhancers and suppressors. Mutations in eIF-5A and PP1 87B dominantly suppress png. Cyclin B was shown to be a key PNG target. Mutations in cyclin B dominantly enhance png, whereas png is suppressed by cyclin B overexpression. Suppression occurs via restoration of Cyclin B protein levels that are decreased in png mutants. The plu and gnu phenotypes are also suppressed by cyclin B overexpression. These studies demonstrate that a crucial function of PNG in controlling the cell cycle is to permit the accumulation of adequate levels of Cyclin B protein.

DROSOPHILA, like many organisms that require rapid embryogenesis in order for their young to reach a motile stage, utilizes a streamlined cell cycle for its early embryonic divisions. The first 13 division cycles occur without cell growth; they are controlled by maternally deposited components and thus are independent of transcription (FOE et al. 1993 ). In these cell cycles, S phase, during which DNA replication occurs, oscillates directly with mitosis without intervening gap phases. In addition, the nuclei divide in a shared cytoplasm with nearly synchronous cycles during these divisions.

Although many cell cycle proteins used during the archetypal G₁-S-G₂-M cycle also function during the S-M cycles, their regulation must be modified, because they are not regulated transcriptionally during the S-M cycles. The entry into S phase and M phase is controlled by cyclin-dependent kinases (CDKs), a complex of a CDK subunit and a cyclin (for review see MURRAY and HUNT 1993 ). In metazoans undergoing the archetypal cell cycle, Cyclin E complexed with CDK2 is necessary for the onset of S phase, and sufficient levels of Cyclin E must be transcribed for the G₁-S transition to occur (KNOBLICH et al. 1994 ). Adequate amounts of the Cyclin B transcript must be accumulated in G₂ for the Cdc2/Cyclin B complex to drive the onset of mitosis. In Drosophila, the mitotic Cyclins A, B, and B3 can each form complexes with Cdc2, and they appear to act synergistically to promote progression through mitosis (LEHNER and O'FARRELL 1989 , LEHNER and O'FARRELL 1990 ; KNOBLICH and LEHNER 1993 ; JACOBS et al. 1998 ).

During the embryonic S-M cycles, CDK2/Cyclin E, Cdc2/Cyclin B, Cdc2/Cyclin B3, and Cdc2/Cyclin A complexes are present and have kinase activity, but the high levels of maternal transcripts do not appear to fluctuate (WHITFIELD et al. 1990 ; RICHARDSON et al. 1993 ; EDGAR et al. 1994 ; JACOBS et al. 1998 ). Thus S-M cycling does not involve changes in cyclin transcript levels. A requirement for CDK2/Cyclin E is inferred from the presence of active kinase (RICHARDSON et al. 1993 ). The mitotic Cyclins A and B have been shown to be required for normal S-M cycling in the early divisions by analysis of mutants (JACOBS et al. 1998 ; STIFFLER et al. 1999 ).

Another critical cell cycle decision point is exit from mitosis, which is initiated at the metaphase/anaphase transition and requires the degradation of mitotic cyclins (for review see ZACHARIAE and NASMYTH 1999 ). Cyclin degradation is signaled by ubiquitination of the proteins by the anaphase-promoting complex/cyclosome (APC/C), and APC/C is activated by FIZZY (FZY) or FIZZY-RELATED (FZR) proteins. In single embryos in interphase or mitosis of the early cleavage cycles, changes in the levels of Cyclin A or B were not detected by immunoblotting, and Cdc2 H1 kinase activity was continuously high (EDGAR et al. 1994 ). Despite the overall constancy of cyclin protein levels and Cdc2 activity within each syncytial embryo, localized degradation of Cyclin B occurs in the vicinity of the mitotic spindle at the end of mitosis. This was shown by the ability of injected nondegradable Cyclin B to cause a mitotic arrest and by immunostaining of endogenous Cyclin B in syncytial embryos (SU et al. 1998 ; HUANG and RAFF 1999 ). Although Cyclin B protein levels remain high in the cytoplasm of early embryos, localized degradation in the vicinity of the spindle causes a decrease in Cdc2/Cyclin B activity at the end of mitosis. Similar experiments have not been done to examine Cyclin A and B3 degradation during the early cycles, but Cyclin A and B3 have been shown to be degraded under the control of FZY later in embryogenesis (SIGRIST et al. 1995 ). Thus, it is likely that the same mechanism of localized degradation occurs for these other mitotic cyclins.

These unique aspects of cell cycle regulation predict that specialized cell cycle regulators might be employed specifically for the S-M cycles. Indeed, the pan gu (png), plutonium (plu), and giant nuclei (gnu) mutations affect the cell cycle solely at the onset of development (FREEMAN et al. 1986 ; FREEMAN and GLOVER 1987 ; SHAMANSKI and ORR-WEAVER 1991 ). Eggs from females homozygous for any of these three mutations are unable to exit the interphase that normally transiently follows meiosis. In these mutants, the meiotic products remain in interphase in unfertilized eggs and undergo inappropriate DNA replication. Similarly, in fertilized embryos from these mutant mothers, mitosis occurs very infrequently, and the nuclei undergo extensive DNA replication to become highly polyploid. In contrast to the near absence of nuclear division in strong png alleles, mutations predicted to retain some PNG activity permit a transient linkage of the S-M cycles. In these weak png mutant embryos, up to six mitotic divisions can occur before the nuclei become polyploid. We have found that PNG is a serine/threonine protein kinase that physically associates with PLU, a novel 19-kD ankyrin repeat-containing protein (AXTON et al. 1994 ; FENGER et al. 2000 ). Additionally, we found that the levels of Cyclin A and B proteins are reduced in png, plu, and gnu mutants (FENGER et al. 2000 ). The degree of reduction is proportional to the severity of the png phenotype.

Genes that regulate the same biological process can be identified via genetic enhancement and suppression screens. Although screening for recessive mutations that genetically interact is labor intensive, it is possible to obtain dominant enhancement or suppression of a mutant phenotype by reducing the level of another protein that participates in the same process. This can be achieved by decreasing the dosage of a gene via a deletion or a null mutation or by reducing levels of gene transcript (for example, by a P-element insertion into regulatory regions of the gene). Recovery of dominant suppressors or enhancers is facilitated by screening with a sensitized mutant phenotype near a threshold such that changes in severity can be detected (SIMON et al. 1991 ). In Drosophila, there is a collection of deficiencies available that permit systematic screening for genomic intervals that when deleted show dominant suppression or enhancement of mutant phenotypes. This permits large numbers of genes to be screened rapidly. Once an interacting region is identified, the individual locus responsible can be identified by testing smaller deficiencies and known complementation groups in the region.

We undertook a dominant enhancer/suppressor screen in order to identify genes that interact with png to control the S-M cycles. Weak png mutations permit several transient S-M cycles to occur, producing embryos with a characteristic phenotype of multiple polyploid nuclei (SHAMANSKI and ORR-WEAVER 1991 ; FENGER et al. 2000 ). Mutations in plu and gnu were shown previously to dominantly enhance this phenotype such that the embryos had one or a few polyploid nuclei (SHAMANSKI and ORR-WEAVER 1991 ). We expected to identify suppressors by their ability to increase the number of embryos with multiple nuclei; similarly, enhancers would decrease the number of embryos with multiple nuclei. By screening for interacting genes, we would be able to recover either downstream targets (possibly substrates) of PNG or upstream regulators. If a substrate were activated by PNG, then reducing its dosage might enhance the png phenotype, whereas if the substrate were inactivated by PNG, it would suppress. Reducing the dosage of an upstream activator of png would enhance the phenotype, but an upstream repressor would appear as a suppressor. It is also possible to recover genes as suppressors or enhancers that act in parallel to png, rather than in the same pathway.

We screened a collection of deficiencies and identified suppressors and enhancers of png. In addition, we used this approach to test known cell cycle regulators for interactions with png. These experiments demonstrated that Cyclin B is a key target of PNG.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila stocks:
Crosses were carried out at 25° using standard techniques unless otherwise noted (GREENSPAN 1997 ). The original png alleles (png^13-1058 and png^12-3318) are described in SHAMANSKI and ORR-WEAVER 1991 ; for simplicity, only the last four digits of the allele designations are used in this article. y w recombinants of the original png alleles were generated as previously described (FENGER et al. 2000 ). plu⁶ (SHAMANSKI and ORR-WEAVER 1991 ; AXTON et al. 1994 ) and gnu³⁰⁵ (FREEMAN et al. 1986 ; FREEMAN and GLOVER 1987 ) were crossed into w^- backgrounds for experiments described in this article. The deficiency kit and mutant strains were obtained from the Bloomington Stock Center. Strains for overexpression of cyclins by heat shock (hs-cycA and hs-cycB), strains carrying eight copies of cyclin B or cyclin B3 transgenes, and cyclin B and cyclin B3 mutant strains were provided by Christian Lehner (LEHNER and O'FARRELL 1990 ; KNOBLICH and LEHNER 1993 ; JACOBS et al. 1998 ). A nanos-Gal4: VP16 transgenic line for germline expression was provided by Pernille Rorth (RORTH 1998 ). The eIF-5a allele l(2)10530 was provided by Amy Beaton [Berkeley Drosophila Genome Project (BDGP)].

Deficiency screen:
The strategy used for the screen to identify deficiencies that dominantly modify png is outlined in Fig 1. The P[w⁺] transgenes used in this scheme to identify the autosomes carrying deficiencies originated from a cosmid walk performed in this laboratory during the cloning of the ord gene (BICKEL et al. 1996 ). The transgene T-021 is a 3.6-kb ord^- w⁺ cosmid on the second chromosome, and the transgene T-014 is a 16-kb ord^- w⁺ cosmid on the third chromosome. These transgenes have been shown to have no effect on the png mutant phenotype. The following 101 deficiency lines were tested for their ability to dominantly modify the png phenotype.

View larger version (19K): In this window In a new window Download PPT slide   Figure 1. Screen for deficiencies that dominantly modify png. Males carrying a balanced deficiency (Df) on either the second or third chromosome were mated to females heterozygous for y png1058 w and homozygous for a P-element insertion marked by the white+ gene (P[w+]) on either the second or third chromosome, respectively. Male progeny resulting from this cross carrying y png1058 w on the X and a deficiency on one of the autosomes in trans to P[w+] were mated to females heterozygous for y png3318 w. Female y png3318 w/y png1058 w progeny carrying an autosomal deficiency were identified by the absence of the w+ marker. These females were mated to wild-type males, and embryos were collected for analysis as described above and in MATERIALS AND METHODS.

Chromosomal arm 2L: Df(2L)net-PMF; Df(2L)al; Df(2L)ast2; Df(2L)dp-79b; Df(2L)C144; Df(2L)JS32; Df(2L)cl-h3; Df(2L)E110; Df(2L)spd; Df(2L)30A-C; Df(2L)Prl; Df(2L)esc10; Df(2L)b87e25; Df(2L)osp29; Df(2L)r10; Df(2L)H20; Df(2L)TW137; Df(2L)TW50; Df(2L)pr76; Df(2L)E55; Df(2L)TW84.

Chromosomal arm 2R: Df(2R)M41A4; In(2R)bw^VDe2L; Df(2R)nap1; Df(2R)nap9; Df(2R)cn88b; Df(2R)pk78s; Df(2R)cn9; Df(2R)H3C1; Df(2R)44CE; Df(2R)H3E1; Df(2R)B5; Df(2R)X1; Df(2R)GJ68-36; Df(2R)en-B; Df(2R)en30; Df(2R)vg135; Df(2R)vg-C; Df(2R)CX1; Df(2R)vg-B; Df(2R)vg56; Df(2R)trix; Df(2R)Jp1; Df(2R)Jp8; Df(2R)l13; Df(2R)Pcl7B; Df(2R)PC4; Df(2R)Pcl11B; Df(2R)P34; Df(2R)173; Df(2R)AA21, In(2R)AA21; Df(2R)Pu-D17; Df(2R)X58-7; Df(2R)X58-12; Df(2R)59AB; Df(2R)59AD; In(2LR)Px4; Df(2R)Px2; Df(2R)M60E.

Chromosomal arm 3L: Df(3L)emc-E12; Df(3L)emc5; Df(3L)R-G5; Df(3L)R-G7; Df(3L)HR119; Df(3L)GN50; Df(3L)GN24; Df(3L)66C-G28; Df(3L)h-i22; Df(3L)29A6; Df(3L)vin7; Df(3L)Ly; Df(3L)fz-GF3b; Df(3L)fz-M21; Df(3L)BK10; Df(3L)brm11; Df(3L)st-f13; Df(3L)81K19; Df(3L)W10; Df(3L)Cat; Df(3L)W4; Df(3L)VW3; Df(3L)ri-79C.

Chromosomal arm 3R: Dp(3;1)2-2, Df(3R)2-2; Df(3R)Tp110; Df(3R)Scr; Df(3R)p712; Df(3R)p-XT103; Df(3R)ry615; Df(3R)C4; Df(3R)P14; Df(3R)Cha7; Df(3R)Dl-BX12; Df(3R)e-N19; Df(3R)e-R1; Df(3R)e-BS2; Df(3R)crb87-4; Df(3R)crb87-5; Df(3R)XS, Dp(3;3)XS; Df(3R)3450; Df(3R)L127; Df(3R)awd-KRB.

Embryo fixation, staining, and microscopy:
Embryos were collected, dechorionated, fixed, 4',6-diamidino-2-phenylindole (DAPI)-stained, and prepared for microscopy as described by SHAMANSKI and ORR-WEAVER 1991 . In some experiments, embryos were stained with propidium iodide instead of DAPI to visualize DNA (FENGER et al. 2000 ). A Zeiss Axiophot microscope with a Plan-neofluar x10 objective was used to examine fluorescence staining. Confocal images were taken with a Zeiss LSM510 confocal laser system mounted on a Zeiss Axiovert 100M microscope with a x40/1.2 W Korr C-APOCHROMAT water objective. Optical sections were taken and projected into a single plane.

Construction and germline expression of UASp-D46cycB:
The plasmid p-sE-hsp-D46cycB (SPRENGER et al. 1997 ) contains a cyclin B cDNA insert that encodes a stable form of Cyclin B lacking the N-terminal 46 amino acids. (The cyclin destruction box lies within this deleted region.) A XhoI/NotI fragment containing the D46cycB insert was subcloned into XhoI/NotI-digested pBlueScript SK+ in order to add a KpnI site to the 5' end. A KpnI/NotI fragment containing the D46cycB insert was subsequently subcloned into KpnI/NotI-digested UASp (RORTH 1998 ). The resulting UASp-D46cycB construct was transformed into y w flies using standard techniques (SPRADLING 1986 ). Strains carrying the UASp-D46cycB transgene on both the second and third chromosomes were obtained. Transgenes containing UASp-D46cycB and nanos-Gal4: VP16 (RORTH 1998 ) were crossed into the png mutant background (both y png³³¹⁸ w and y png¹⁰⁵⁸ w) for experiments summarized in Table 2.

Heat-shock protocol:
For Cyclin A and Cyclin B overexpression experiments summarized in Table 2, hs-cycA and hs-cycB transgenes (KNOBLICH and LEHNER 1993 ) were crossed into the png mutant background. Homozygous mutant png females (y png³³¹⁸ w/y png¹⁰⁵⁸ w as well as y png¹⁰⁵⁸ w/ y png¹⁰⁵⁸ w) carrying one copy of hs-cycA or hs-cycB were used in these experiments; sibling homozygous mutant png females lacking hs-cycA or hs-cycB transgenes served as controls in these experiments. Newly eclosed females of the desired genotypes were collected over 5 days and then mated with wild-type males over 2 days at 25°. Flies were subjected to a heat-shock pulse (by incubating vials in a 37° H₂O bath) for 1 hr per day for 6 days and otherwise maintained at 25°. Immediately following the final heat-shock pulse, 0- to 4-hr embryos were collected for DNA staining.

Embryo extracts and immunoblots:
Preparation of embryonic protein extracts and immunoblotting was performed as described in TANG et al. 1998 except that embryos were homogenized at 3:1 USB/embryo (v:v). Mouse monoclonal anti-Cyclin B antibody F24F (LEHNER and O'FARRELL 1990 ) was used at a dilution of 1:200. Equal amounts of protein were loaded in each lane as confirmed by reprobing the Cyclin B immunoblot with antibodies against -tubulin using rat anti--tubulin (YL1/2, Harlan Sera-lab, 1:200).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of deficiencies that suppress or enhance png mutations:
We initiated a screen to identify mutants that enhance or suppress weak png mutations by using the deficiency collection available from the Bloomington Stock Center to survey about 50% of the genome using a minimum number of stocks and crosses. We scored for deficiencies that when heterozygous would dominantly suppress or enhance weak png mutations by examining the nuclear phenotype of embryos. Females transheterozygous for the weak png³³¹⁸ allele and the strong png¹⁰⁵⁸ allele produce embryos in which some mitotic divisions occur before the nuclei ultimately become polyploid. These embryos contain up to 16 giant, polyploid nuclei (Fig 2A and Fig 3A). In the same collections, there are embryos that have a single, multilobed nucleus that results from the female meiotic products and male pronucleus undergoing DNA replication but no nuclear division. As these nuclei become polyploid they can fuse together. Thus, in the absence of mitosis and nuclear division, between one and five giant polyploid nuclei are produced. We scored these two classes of embryos as those with five or fewer nuclei vs. those with greater than five (multinucleated). The relative percentages of these two classes produced by png³³¹⁸/png¹⁰⁵⁸ females is affected by genetic background and can vary between 30 and 55%. Thus we defined suppression as genotypes producing >60% multinucleated embryos and enhancement as <30%.

View larger version (42K): In this window In a new window Download PPT slide   Figure 2. Identification of deficiencies that dominantly enhance or suppress png. Embryos from png mutant females mated to wild-type males were fixed, stained with DAPI to visualize the DNA, and viewed as whole mounts. (A) Embryos from y png3318 w/y png1058 w females contain giant, polyploid nuclei. Typically, 30–55% of these embryos contain greater than five nuclei (multinucleated phenotype), which indicates that some mitotic divisions occurred. (B) Embryos from y png3318 w/y png1058 w; Df(2R)59AB/+ females have a dominantly enhanced png phenotype, as evidenced by the decreased percentage (9.7%) of multinucleated embryos. (C) Embryos from y png3318 w/y png1058 w; Df(3R)ry615/+ females have a dominantly suppressed png phenotype, as evidenced by the increased percentage (74.2%) of multinucleated embryos.

View larger version (25K): In this window In a new window Download PPT slide   Figure 3. Dominant modification of png by cyclin B and eIF-5A. Embryos from png mutant females mated to wild-type males were fixed, stained with propidium iodide to visualize the DNA, and viewed as whole mounts. (A) Trans-heterozygous y png3318 w/ y png1058 w females produce an average of 30–55% multinucleated embryos. (B) Homozygous y png1058 w/ y png1058 w females produce an average of 5–9% multinucleated embryos. (C) Embryos from trans-heterozygous y png3318 w/ y png1058 w; cycB2/+ females have a dominantly enhanced png phenotype, as evidenced by the decreased percentage (13.4%) of multinucleated embryos (compare to A). (D) Embryos from homozygous y png1058 w/ y png1058 w; l(2)01296/+ females have a dominantly suppressed png phenotype, as shown by the increased percentage (50.4%) of multinucleated embryos (compare to B). l(2)01296 is an allele of eIF-5A.

Embryos were collected for 0–4 hr from y png³³¹⁸ w/y png¹⁰⁵⁸ w females carrying a deficiency on the second or third chromosome. The embryos were fixed and stained with DAPI. The females with the deficiency were distinguished from their sibling controls by the absence of a dominant P[w⁺] marker carried on the autosome (Fig 1). Enhancement of the png mutant phenotype was evident by a reduction in the number of embryos with greater than five nuclei (Fig 2B). Suppression was identified by the appearance of an increased number of embryos with multiple nuclei; strong suppressors also produced some embryos that had a greater number of nuclei than the maximum of 16 typically seen in embryos from png³³¹⁸/png¹⁰⁵⁸ females (Fig 2C). Some of the deficiencies that were identified as suppressors were subsequently tested for their ability to suppress females homozygous for the strong allele, png¹⁰⁵⁸/png¹⁰⁵⁸, to produce embryos with multiple nuclei (Table 1).

We screened 59 deficiency intervals on the second chromosome and 42 intervals on the third chromosome. These deficiencies spanned 62% of the second and 47% of the third chromosome. As a control that the screen was effective we recovered a deficiency for plu (Df(2R)173) as an enhancer of the png phenotype. We identified 8 deficiencies that strongly enhance the png phenotype, 12 that weakly enhance, and 8 that suppress png mutations (Table 1).

To identify the locus within each deficiency responsible for the genetic interaction with png, we tested lethal P-element insertions within or adjacent to each interval for interaction with png and, when possible, examined deficiencies that dissect the interval. This screen led to the identification of five single complementation groups that interact genetically with png. The strongest enhancer we identified was cyclin B, discussed in detail below. The strongest suppressors were eIF-5A and PP1 87B, described below. Two weaker enhancers were recovered: l(2)s4989 is a P-element insertion that disrupts a novel gene and l(2)10481 is a P-element insertion 5' to the Tkr gene (Tyrosine kinase-related).

Cyclin B is a target of png:
Two deficiencies that remove the 59AB region on chromosome 2R strongly enhanced the png mutant phenotype (Table 1; compare Fig 2B and A). This region contains the cyclin B gene (KNOBLICH and LEHNER 1993 ), a strong candidate for the interacting locus given the observation that Cyclin B protein levels are decreased in png mutants in proportion to the strength of the allele (FENGER et al. 2000 ). Further reduction of Cyclin B protein by mutation of one copy of the gene would be predicted to enhance the png phenotype. To test whether the enhancement was due to a decreased gene dosage of Cyclin B, we obtained an allele of cyclin B that is a small deletion generated by imprecise excision of a P element (JACOBS et al. 1998 ). This mutation strongly enhanced the png phenotype (Table 2; compare Fig 3C and A).

Conversely, we wanted to test whether we could suppress the phenotype by increasing the level of Cyclin B protein in the png mutant embryos, so we crossed multiple copies (six to eight) of the wild-type cyclin B gene into the png mutant background. We showed by immunoblotting that strains with extra copies of the cyclin B gene in the png mutant background produced increased amounts of Cyclin B protein (Fig 4). Overexpression of cyclin B strongly suppressed the png mutants. In the weak png³³¹⁸ mutants with extra copies of the cyclin B gene, there was a striking increase in the number of nuclei in the embryos as well as in the number of embryos with multiple nuclei (Table 2; compare Fig 5C and A). In addition, the polar bodies contained condensed chromosomes with a normal rosette arrangement (compare Fig 5H and Fig G). This extent of suppression was not observed with any other deficiency or mutant tested. Moreover, the chromosomes of the zygotic nuclei were condensed, and mitotic figures were visible (Fig 5E and Fig F). The suppression of the zygotic nuclei was not complete, however, because all of the embryos contained polyploid nuclei (Fig 5C), and none survived past embryogenesis. Thus, the linkage between S and M phases was ultimately broken, and nuclear division ceased. Overexpression of cyclin B dramatically suppressed the strong png¹⁰⁵⁸ mutants as well, resulting in multiple nuclei in these embryos (Table 2; compare Fig 5D and Fig B). However, unlike the weak png³³¹⁸ mutants with extra cyclin B, normal polar bodies were not observed.

View larger version (19K): In this window In a new window Download PPT slide   Figure 4. Restoration of Cyclin B protein levels in png mutants by increased cyclin B gene copies. An immunoblot of embryonic extracts probed with anti-Cyclin B and anti--tubulin is shown. Extracts were prepared from 0- to 2.5-hr embryos laid by wild-type (OR, Oregon-R) females and homozygous png females with or without extra gene copies (six to eight) of cycB. The increased cycB copy number restores the low levels of Cyclin B protein in png mutants (for both the weak png3318 and strong png1058 alleles) to nearly normal levels.

View larger version (19K): In this window In a new window Download PPT slide   Figure 5. Suppression of png by increased cyclin B gene copies. Embryos from png mutant females mated to wild-type males were fixed, stained with propidium iodide to visualize the DNA, and viewed as whole mounts. (A) Embryos from homozygous y png3318 w/ y png3318 w females. (B) Embryos from homozygous y png1058 w/ y png1058 w females. (C) Embryos from homozygous y png3318 w/y png3318 w females carrying six to eight extra copies of cycB have a strongly suppressed png phenotype, as evidenced by the increased percentage of multinucleated embryos (from 30–55 to 93.9%) and an increased maximum number of nuclei per embryo (compare to A). (D) Embryos from homozygous y png1058 w/ y png1058 w females carrying extra copies of cycB also have a strongly suppressed png phenotype, as shown by the increased percentage of multinucleated embryos (from 5–9 to 94.4%; compare to B). Confocal imaging of an embryo from homozygous y png3318 w/y png3318 w females carrying extra copies of cycB reveals zygotic nuclei with condensed chromosomes in metaphase (E), anaphase figures (F), and a rosette structure of the polar bodies (H). For comparison, a polar body rosette in an embryo from a wild-type female is shown (G).

Cyclin B levels were also increased in png mutants by heat-shock induction and by induction in the germline using GAL4 under the control of the nanos promoter. The cyclin B gene in the latter construct lacks the N-terminal destruction box and thus would promote elevated levels of Cyclin B both from transcriptional induction as well as lack of degradation by the ubiquitin-APC/C pathway (GLOTZER et al. 1991 ). Both the weak and strong png phenotypes were suppressed by overexpression of cyclin B by either of these two methods (Table 2). However, the degree of suppression of the zygotic nuclear phenotype of png was less than that for lines carrying extra copies of the wild-type cyclin B gene, and restoration of normal polar body morphology was not observed.

Increased cyclin B suppresses the plutonium and gnu phenotypes:
Given the observation that png, plu, and gnu appear to function in a common pathway, the genetic interactions between png and cyclin B, and the decreased Cyclin B protein levels in plu and gnu mutants, we wanted to test whether or not the plu and gnu phenotypes could be modified by altering the cyclin B gene dosage. Unlike png, all of the existing mutations in plu and the single gnu mutation appear to be null alleles (FREEMAN et al. 1986 ; FREEMAN and GLOVER 1987 ; SHAMANSKI and ORR-WEAVER 1991 ; AXTON et al. 1994 ). Because mitosis occurs only very rarely in embryos from plu and gnu mutant females, it is not possible to identify enhancers of this strong phenotype. Deficiencies identified in this screen as suppressors of the png phenotype (Table 1) did not suppress the plu phenotype (data not shown), perhaps due to a lack of residual PLU function. Surprisingly, increasing the level of Cyclin B protein in plu⁶ and gnu³⁰⁵ mutants by increasing the cyclin B gene dosage (four extra copies) resulted in suppression of the mutant phenotypes. For plu females with extra cyclin B, 39% of their embryos were multinucleated compared to 1.5% in sibling controls; for gnu females with extra cyclin B, 65% of their embryos were multinucleated compared to 0.8% in sibling controls. Similar results were obtained by inducing nondegradable Cyclin B in the germline of plu mutant females (data not shown). Unlike the weak png mutation, overexpression of cyclin B did not correct the polar body defects associated with plu and gnu mutations.

Effects of mutations in other cyclin genes on the png phenotype:
In addition to Cyclin B, the levels of Cyclin A protein are decreased in proportion to the strength of the png allele. Surprisingly, neither a deficiency that uncovers cyclin A nor mutations in the gene enhanced the png mutant phenotype (Table 2 and data not shown). A negative result in this test does not exclude the possibility of interaction because a twofold reduction in gene dosage may not decrease the level of gene product below a threshold required to detect an effect. However, loss of one copy of the cyclin A gene in the mother does affect cell cycle parameters and nuclear division in the embryo (STIFFLER et al. 1999 ), so one copy of the gene does not produce sufficient levels of the protein for normal cell division. We tested whether overexpression of cyclin A could suppress the png mutant phenotype. This experiment produced an unexpected result. When cyclin A was induced by heat shock, the png phenotype was enhanced (Table 2). This may be due to the ability of Cyclin A when overexpressed to drive cells inappropriately into S phase (LEHNER et al. 1991 ; SPRENGER et al. 1997 ). Thus, this result may be a consequence of artifactually high levels of Cyclin A and may not accurately reflect the function of Cyclin A in the early embryonic divisions.

Mutation of the cyclin B3 gene (JACOBS et al. 1998 ) had no effect on the png phenotype, but crossing eight extra copies of the wild-type cyclin B3 gene into the png mutant background resulted in suppression (Table 2). A deficiency that uncovers cyclin J did not modify the png mutant phenotype (Table 2). The cyclin E gene is required for S phase in Drosophila (KNOBLICH et al. 1994 ). We tested whether a deficiency that uncovers cyclin E or a mutation in the cyclin E gene would dominantly suppress png. Unexpectedly, we found that a decreased gene dosage of cyclin E resulted in strong enhancement of the png phenotype (Table 2 and data not shown). As an independent test for potential interactions between cyclin E and png, we tested a mutation in dacapo, an inhibitor of the CDK2/Cyclin E kinase (DE NOOIJ et al. 1996 ; LANE et al. 1996 ). Both the dacapo mutation and a deficiency that uncovers dacapo (Df(2R)B5) slightly suppressed the png mutant phenotype (Table 1 and Table 2).

The eIF-5A and PP1 87B genes suppress png mutations:
The deficiency Df(2R)Px2 strongly suppresses the png mutant phenotype. To identify the gene responsible for this interaction, we tested available lethal P element insertions both within the genomic interval of this deficiency and in immediately adjacent-lettered divisions. We found that a P-element inserted within the first intron of the eIF-5A gene suppresses png (Table 3; compare Fig 3D and Fig B). The eIF-5A mutation lies in the region of Df(2R)Px2, but it complements the deficiency and thus lies outside of it. Consequently, both eIF-5A and another as yet unidentified locus inside the deficiency are suppressors of png. We confirmed that the eIF-5A mutation is a suppressor by testing an independent P-element insertion in the eIF-5A gene and observing that it also caused suppression (Table 3). The lethal phenotype of these mutations has not been analyzed, and the gene is defined on the basis of its homology to eukaryotic orthologs of the eIF-5A protein. Although the eIF-5A protein was identified by its ability to promote peptide bond formation in vitro, its role in translation and gene expression is not well understood (KANG and HERSHEY 1994 ). It may contribute to RNA stability rather than to translation initiation. We could not detect an increase in Cyclin B protein levels in embryos from png¹⁰⁵⁸/png¹⁰⁵⁸; eIF-5A/+ females by immunoblotting (data not shown).

To test whether the effect of eIF-5A on png could be due to perturbations in translation initiation, we crossed mutations in the translation initiation factors eIF-4a, eIF-4E, and eIF-3p40 into png mutants, but did not observe enhancement or suppression of the phenotype (Table 3). We also examined mutations in two other genes known to be involved in translation, poney and plume (GALLONI and EDGAR 1999 ), and found they did not genetically interact with png. Although these results do not eliminate the possibility of a link between translation levels and the png phenotype, they suggest that the effect of eIF-5A is not due to a global decrease in translation.

Reduction of the gene dosage of the type 1 serine/threonine protein phosphatase PP1 87B appears to be responsible for suppression of the png phenotype by Df(3R)ry615 (compare Fig 2C and A). We tested the j6E7 mutation in the PP1 87B gene and found dominant suppression of the png³³¹⁸/png¹⁰⁵⁸ phenotype with 78% multinucleated embryos compared to 41% in sibling controls. PP1 87B also suppressed females homozygous for the strong allele, png¹⁰⁵⁸, to produce multinucleated embryos (data not shown). This suppression could be direct in that PP1 could be responsible for dephosphorylating the substrates of the PNG kinase. It is possible, however, given the pleiotrophic nature of PP1 87B phenotypes that it is suppressing indirectly (AXTON et al. 1990 ; BAKSA et al. 1993 ). A lethal P-element insertion (l(2)k09822) in another serine/threonine protein phosphatase gene, PP2A 28D, the protein phosphatase 2A catalytic subunit gene at 28D (SNAITH et al. 1996 ), also dominantly suppressed the png³³¹⁸/png¹⁰⁵⁸ phenotype with 89% multinucleated embryos compared to 31% in sibling controls. A similar degree of suppression was observed for the l(2)s5286 allele of PP2A 28D, and both alleles suppressed the strong png¹⁰⁵⁸ phenotype as well (data not shown). Reducing the activity of a phosphatase might generally increase mitotic activities in the mutant embryos, thereby suppressing the png phenotype.

Analysis of candidate cell cycle regulators:
In addition to the deficiency screen, we tested for interactions between png and known cell cycle regulators. The kinase activity of Cdc2 is decreased in png mutant embryos, presumably due to the reduction in Cyclin A and Cyclin B protein levels (FENGER et al. 2000 ). We tested three alleles of cdc2, including the putative protein null cdc2^B47 (STERN et al. 1993 ), but none of these dominantly enhanced the png phenotype (data not shown). This is not surprising, given that Cdc2 activity is controlled by phosphorylation and association with cyclin subunits rather than by alteration of Cdc2 protein levels. Other cell cycle genes that did not show an interaction with png (data not shown) were the Cyclin A regulator rca1 (DONG et al. 1997 ), the S phase inhibitor escargot (HAYASHI et al. 1993 ), the grapes checkpoint gene (FOGARTY et al. 1997 ; SIBON et al. 1997 ), dE2F1 (DURONIO et al. 1995 ), dDP (ROYZMAN et al. 1997 ), three rows (D'ANDREA et al. 1993 ; PHILP et al. 1993 ), pimples (LEISMANN et al. 2000 ), the cullin gft (ASHBURNER et al. 1999 ), morula (REED and ORR-WEAVER 1997 ), and double parked (WHITTAKER et al. 2000 ).

To test whether png might affect mitotic cyclin levels by controlling cyclin protein degradation, we examined mutations in genes encoding the APC/C regulators fizzy (fzy) and fzy-related (fzr) as well as mutations in an APC/C subunit and many components of the ubiquitination pathway. We found that none of five mutations in fzy tested affected the png phenotype (data not shown; DAWSON et al. 1993 , DAWSON et al. 1995 ; SIGRIST et al. 1995 ). We made several recombinant chromosomes containing both the png¹⁰⁵⁸ allele and the Df(1)bi-D3 deficiency that removes fzr (SIGRIST and LEHNER 1997 ), but these did not have an altered phenotype in trans to png³³¹⁸ or png¹⁰⁵⁸ (data not shown). Mutations in APC5 (SPRADLING et al. 1999 ), effete (UbcD1; CASTRILLON et al. 1993 ; CENCI et al. 1997 ), and lesswright (Ubc9; SPRADLING et al. 1999 ) did not affect png (data not shown).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The use of a weak allele of png in trans to a strong allele provided a sensitized system that permitted us to screen for genes whose reduced dosage would suppress or enhance the png mutant phenotype. Suppression or enhancement resulting from reduced gene dosage requires less specificity than does an interaction with a missense mutation. In addition, genetic interactions arising from changes in gene dosage are more likely to be dominant. These two features made it feasible to screen for suppressors and enhancers of a phenotype that required collection and staining of embryos for microscopic examination.

We found that a mutation in cyclin B dominantly enhanced png. The addition of extra copies of the wild-type cyclin B gene increased the level of Cyclin B protein in the embryos to near wild-type levels and dramatically suppressed png. Increased Cyclin B suppressed both of the major defects seen in png mutants: inappropriate replication in the polar bodies and the near absence of mitosis in the zygotic nuclei. It is striking that increased Cyclin B also suppressed the plu and gnu mutants such that zygotic nuclei underwent some mitotic divisions. We had never previously observed suppression of the plu phenotype (e.g., by deficiencies or mutations that dominantly suppress png). Although the function of GNU is not known, nor is it known where it acts in the PNG pathway, the suppression by increased copies of cyclin B suggests that it also critically controls Cyclin B levels. Thus, Cyclin B is a key target of both the PNG/PLU complex and GNU.

Suppression of the weak png mutation by increased Cyclin B restored the normal rosette structure and ploidy to the polar bodies. In wild-type embryos, the unused female meiotic products exit the postmeiotic interphase, their chromosomes condense, and these nuclei remain arrested in a metaphase-like state despite sharing the same cytoplasm with dividing nuclei. These observations lead to two conclusions about the regulation of the cell cycle in the polar bodies. First, condensation of the chromosomes is likely to prevent DNA replication and, second, Cyclin B is the sole PNG target necessary for condensation and inhibition of DNA replication in the polar bodies.

The suppression of png, plu, and gnu by overexpressing cyclin B is not complete because ultimately nuclear divisions fail, and the nuclei continue to replicate and become polyploid. It is possible that Cyclin B is the sole target of the PNG/PLU Complex and GNU and that the levels of increased Cyclin B protein in the png, plu, and gnu mutants (via increased copies of the cyclin B gene) are not adequate for completion of all the S-M cycles. However, it seems more likely that, although Cyclin B is a key target, other targets of the PNG/PLU complex and GNU are also important. Cyclin A is a particularly good candidate for two reasons. Cyclin A protein levels are decreased in png, plu, and gnu mutants; for png, the decrease is in proportion to the strength of the allele (FENGER et al. 2000 ). Decreasing the dosage of maternal cyclin A to one copy caused an increase in cycle time during the early embryonic divisions (STIFFLER et al. 1999 ). In contrast, decreasing the dosage of cyclin B did not affect the timing of nuclear cycles, whereas it did affect microtubule dynamics (STIFFLER et al. 1999 ). These observations led Stiffler et al. to conclude that Cyclin B controls cytoskeletal events during the S-M cycles, but Cyclin A controls the nuclear cycles. If this model is correct, Cyclin A may be a critical target for the influence of PNG, PLU, and GNU on the nuclear cycles.

We did not observe an enhancement of the png phenotype by mutations in cyclin A, and overexpression of cyclin A unexpectedly enhanced the phenotype. This latter result likely reflects the ability of excess Cyclin A to promote DNA replication (LEHNER et al. 1991 ; SPRENGER et al. 1997 ), but it is not clear why a reduction in Cyclin A did not affect the png phenotype. Further delineation of the role of Cyclin A levels will likely emerge from identification of PNG kinase substrates and elucidation of the mechanism by which PNG influences Cyclin A and B protein levels.

There are several mechanisms by which PNG, PLU, and GNU could affect Cyclin A and B protein levels including maternal transcription, mRNA stability or processing, translation, and cyclin protein stability. We examined mutations in the pathway that targets mitotic cyclin proteins for destruction. We did not observe suppression of the png mutant phenotype by reducing the dosage of the two known activators of cyclin destruction, fzy or fzr. Similarly, mutation of an APC/C subunit or several genes affecting the ubiquitin pathway did not alter the png mutant phenotype. These negative results do not exclude a role for PNG in controlling APC/C-mediated protein degradation, as the dosage reductions may not have reduced protein activity below a crucial threshold. Additional experiments will be required to evaluate how PNG affects Cyclin A and B protein levels.

In addition to cyclin A and B, we examined the effects of alterations in the gene dosage of other cyclins on the png phenotype. Embryos from mutant png females appear to have normal Cyclin B3 protein levels by immunoblotting (our unpublished observations), and decreasing the cyclin B3 gene dosage had no effect on the png phenotype. However, addition of eight extra copies of the wild-type cyclin B3 gene resulted in suppression of the weak png phenotype. We considered the possibility that overexpression of cyclin B3 might suppress png by causing an increase in Cyclin B (e.g., via competition for APC/C-mediated protein degradation), but immunoblotting revealed no change in Cyclin B protein levels. The suppression of png by overexpression of cyclin B3 could be a consequence of overlapping functions of Cyclins A, B, and B3 during mitosis (JACOBS et al. 1998 ); high levels of Cyclin B3 might compensate for reductions in Cyclins A and B. The png phenotype was not altered by a deficiency that uncovers cyclin J. The cell cycle function of Cyclin J has not been defined, and mutations are not available (FINLEY et al. 1996 ).

Given the well-established role for Cyclin E in progression through S phase, we were surprised to find that decreasing the cyclin E gene dosage enhanced the png phenotype. Consistent with this observation, mutation of dacapo, a CDK2/Cyclin E kinase inhibitor, slightly suppressed png. In addition to promotion of S phase, other cell cycle roles have been ascribed to Cyclin E that might account for the enhancement of png by decreased cyclin E. For example, in Xenopus egg extracts, Cdk2/Cyclin E has been implicated in activation of Cdc2/Cyclin B for progression into mitosis (GUADAGNO and NEWPORT 1996 ). In Drosophila, ectopic expression of cyclin E induces post-transcriptional accumulation of mitotic cyclins (KNOBLICH et al. 1994 ). We tested whether decreasing the cyclin E gene dosage enhanced png by causing a further decrease in Cyclin B, but we detected no additional decrease in Cyclin B protein levels in embryos from png³³¹⁸/png¹⁰⁵⁸; l(2)05206/+ females by immunoblotting (our unpublished observations).

There are two precedents for how Cyclin E could affect DNA replication in a manner that would enhance png. A paradoxical role for Cdk2/Cyclin E as a negative regulator of DNA replication in Xenopus extracts has been described (HUA et al. 1997 ). In Drosophila, Cyclin E also may negatively regulate DNA replication, because hypomorphic cyclin E mutations cause endocycling nurse cells to undergo late DNA replication with accumulation of increased amounts of heterochromatic DNA (LILLY and SPRADLING 1996 ). Thus decreased cyclin E could permit more unregulated DNA replication in png mutant embryos, possibly further impeding entry into mitosis and resulting in enhancement of the png phenotype.

Mutation of eIF-5A strongly suppressed png mutants. We cannot at present distinguish whether this is so because eIF-5A directly influences the same process as png (i.e., Cyclin B protein levels) or whether the suppression results from compensatory changes in the cell cycle. Although we did not detect an increase in Cyclin B protein in png¹⁰⁵⁸/png¹⁰⁵⁸; eIF-5A/+ strains by immunoblotting, it is possible that Cyclin B was increased sufficiently to account for the suppression seen or that its local concentration in the vicinity of the zygotic nuclei was increased.

The difficulty in defining the relationship between eIF-5A and png is in large part due to the fact that the biological function of eIF-5A is not understood, and it is not clear that its primary function is in translation initiation. Consistent with this, we did not observe a genetic interaction between genes that encode known translation initiation proteins or other proteins in the translational machinery. Mutation of eIF-5A in yeast decreases general protein translation by 30% (KANG and HERSHEY 1994 ). It is possible that a general reduction of translation permits more efficient translation of specific messages that could suppress png. However, eIF-5A has been implicated in several other processes. Mutation of eIF-5A stabilizes an mRNA with a nonsense mutation, suggesting eIF-5A may be involved in transcript decay (ZUK and JACOBSON 1998 ). Thus the suppression of png by mutation of eIF-5A could be due to stabilization of cyclin B mRNA in the early Drosophila embryo. eIF-5A also has been implicated in nuclear export of RNA (ROSORIUS et al. 1999 ; LIPOWSKY et al. 2000 ), but it is less clear how such a function could account for png suppression.

An alternative explanation is that mutation of eIF-5A suppresses png mutants as an indirect consequence of eIF-5A's role in promoting the onset of S phase. The activities of eIF-5A in translation initiation in vitro and RNA transport require a unique post-translational modification of the eIF-5A protein, hypusination (SCHMIER et al. 1991 ). Hypusination can be inhibited by the drug mimosine, which is thought to affect eIF-5A specifically (HANAUSKE-ABEL et al. 1994 , HANAUSKE-ABEL et al. 1995 ). Treatment of cultured mammalian cells with mimosine blocks the onset of S phase and affects the distribution of mRNAs on polysomes, raising the possibility that eIF-5A promotes S phase by directing the translation of factors required for DNA replication. Consistent with this, overexpression of mammalian eIF-5A was shown to drive the G₁-S transition in yeast cells by causing the accumulation of the G₁ cyclin CLN3, although the mechanism by which CLN3 levels were increased was not determined (EDWARDS et al. 1997 ). If eIF-5A promotes DNA replication in Drosophila, then reducing its activity could suppress png indirectly. It is interesting that the deoxyhypusine synthase gene required for hypusination of eIF-5A maps within the deficiency Df(3L)66C-G28 that suppresses png. This gene may be responsible for the suppression observed by this deficiency, but a test of this hypothesis awaits the identification of mutations in the deoxyhypusine synthase gene.

The third interacting gene we recovered from this screen encodes the serine/threonine protein phosphatase PP1 87B. This type 1 phosphatase might dephosphorylate PNG substrates, such that reducing the activity of the phosphatase suppresses png mutations. The suppression could be indirect, however. Mutations in PP1 87B cause a mitotic arrest with predicted high levels of mitotic cyclin proteins, which could alleviate the png phenotype (AXTON et al. 1990 ). Another serine/threonine protein phosphatase gene, PP2A 28D, was also identified as a suppressor of png. Mutations in PP2A 28D cause a mitotic arrest with overcondensed chromatin (SNAITH et al. 1996 ). In Xenopus egg extracts, protein phosphatase 2A antagonizes Cdc2/Cyclin B activity (LEE 1995 ) and is required for the initiation of DNA replication (LIN et al. 1998 ). A decrease in either of these PP2A functions could account for the suppression of png by mutations in PP2A 28D. Once PNG substrates are identified, it will be of interest to determine whether they are dephosphorylated by PP1 87B or PP2A 28D.

This genetic screen led to the important identification of Cyclin B as a PNG target. In addition, an unanticipated interaction between png and eIF-5A was uncovered. Identification of the genes responsible for the suppression or enhancement observed with the other deficiencies will provide key insights into the mechanism by which the PNG/PLU protein kinase complex controls the S-M cell cycles.

	FOOTNOTES

¹ These authors contributed equally to this work.
² Present address: Department of Biochemistry, 1041 E. Lowell St., University of Arizona, Tucson, AZ 85721.

	ACKNOWLEDGMENTS

We acknowledge technical assistance on the screen provided by Doug Van Hoewyk and Colleen Raymond. We are grateful to the Bloomington Stock Center for providing the deficiency kit and mutant strains. We thank Christian Lehner for the cyclin A, B, and B3 overexpression strains as well as cyclin B and cyclin B3 mutant strains; Pat O'Farrell for Cyclin B antibodies and the D46 cyclin B cDNA; Pernille Rorth for the UASp vector and nanos-Gal4: VP16 transgenic line; and Amy Beaton (BDGP) for the eIF-5A allele l(2)10530. We benefited from helpful discussions with Doug Fenger, Tom Dever, and Steve Elledge. Confocal microscopy was done in the Keck Imaging Center at the Whitehead Institute. L.A.L. was supported by a postdoctoral fellowship from the National Institutes of Health (NIH), L.K.E. was supported by a postdoctoral fellowship from the American Cancer Society, and G.B. is a postdoctoral fellow of the Damon Runyon-Walter Winchell Cancer Research Fund (DRG 1537) and a Margaret and Herman Sokol Fellow of the Whitehead Institute. This work was supported by NIH grant GM39341 to T.O.-W.

Manuscript received November 21, 2000; Accepted for publication May 7, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ASHBURNER, M., S. MISRA, J. ROOTE, S. E. LEWIS, and R. BLAZEJ et al., 1999  An exploration of the sequence of a 2.9-Mb region of the genome of Drosophila melanogaster. The Adh region. Genetics 153:179-219[Abstract/Free Full Text].

AXTON, J. M., V. DOMBRADI, P. T. W. COHEN, and D. M. GLOVER, 1990  One of the protein phosphatase 1 isoenzymes in Drosophila is essential for mitosis. Cell 63:33-46[Medline].

AXTON, J. M., F. L. SHAMANSKI, L. M. YOUNG, D. S. HENDERSON, and J. B. BOYD et al., 1994  The inhibitor of DNA replication encoded by the Drosophila gene plutonium is a small, ankyrin repeat protein. EMBO J. 13:462-470[Medline].

BAKSA, K., H. MORAWIETZ, V. DOMBRADI, M. AXTON, and H. TAUBERT et al., 1993  Mutations in the protein phosphatase 1 gene at 87B can differentially affect suppression of position-effect variegation and mitosis in Drosophila melanogaster.. Genetics 135:117-125[Abstract].

BICKEL, S. E., D. W. WYMAN, W. Y. MIYAZAKI, D. P. MOORE, and T. L. ORR-WEAVER, 1996  Identification of ORD, a Drosophila protein essential for sister-chromatid cohesion. EMBO J. 15:1451-1459[Medline].

CASTRILLON, D. H., P. GONCZY, S. ALEXANDER, R. RAWSON, and C. G. EBERHART et al., 1993  Toward a molecular genetic analysis of spermatogenesis in Drosophila melanogaster: characterization of male-sterile mutants generated by single P element mutagenesis. Genetics 135:489-505[Abstract].

CENCI, G., R. B. RAWSON, G. BELLONI, D. H. CASTRILLON, and M. TUDOR et al., 1997  UbcD1, a Drosophila ubiquitin-conjugating enzyme required for proper telomere behavior. Genes Dev. 11:863-875[Abstract/Free Full Text].

D'ANDREA, R. J., R. STRATMANN, C. F. LEHNER, U. P. JOHN, and R. SAINT, 1993  The three rows gene of Drosophila melanogaster encodes a novel protein that is required for chromosome disjunction during mitosis. Mol. Biol. Cell 4:1161-1174[Abstract].

DAWSON, I. A., S. ROTH, M. AKAM, and S. ARTAVANIS-TSAKONAS, 1993  Mutations at the fizzy locus cause metaphase arrest in Drosophila melanogaster embryos. Development 117:359-376[Abstract].

DAWSON, I. A., S. ROTH, and S. ARTAVANIS-TSAKONAS, 1995  The Drosophila cell cycle gene fizzy is required for normal degradation of cyclins A and B during mitosis and has homology to the CDC20 gene of Saccharomyces cerevisiae.. J. Cell Biol. 129:725-737[Abstract/Free Full Text].

DE NOOIJ, J. C., M. A. LETENDRE, and I. K. HARIHARAN, 1996  A cyclin-dependent kinase inhibitor, Dacapo, is necessary for timely exit from the cell cycle during Drosophila embryogenesis. Cell 87:1237-1247[Medline].

DONG, X., K. H. ZAVITZ, B. J. THOMAS, M. LIN, and S. CAMPBELL et al., 1997  Control of G1 in the developing Drosophila eye: rca1 regulates Cyclin A. Genes Dev. 11:94-105[Abstract/Free Full Text].

DURONIO, R. J., P. H. O'FARRELL, J.-E. XIE, A. BROOK, and N. DYSON, 1995  The transcription factor E2F is required for S phase during Drosophila embryogenesis. Genes Dev. 9:1445-1455[Abstract/Free Full Text].

EDGAR, B. A., F. SPRENGER, R. J. DURONIO, P. LEOPOLD, and P. H. O'FARRELL, 1994  Distinct molecular mechanisms regulate cell cycle timing at successive stages of Drosophila embryogenesis. Genes Dev. 8:440-452[Abstract/Free Full Text].

EDWARDS, M. C., N. LIEGEOIS, J. HORECKA, R. A. DE PINHO, and G. F. SPRAGUE et al., 1997  Human CPR (cell cycle progression restoration) genes impart a Far-phenotype on yeast cells. Genetics 147:1063-1076[Abstract].

FENGER, D. D., J. L. CARMINATI, D. L. BURNEY-SIGMAN, H. KASHEVSKY, and J. L. DINES et al., 2000  PAN GU: a protein kinase that inhibits S phase and promotes mitosis in early Drosophila development. Development 127:4763-4774[Abstract].

FINLEY, R. L., B. J. THOMAS, S. L. ZIPURSKY, and R. BRENT, 1996  Isolation of Drosophila cyclin D, a protein expressed in the morphogenetic furrow before entry into S phase. Proc. Natl. Acad. Sci. USA 93:3011-3015[Abstract/Free Full Text].

FOE, V. E., G. M. ODELL and B. A. EDGAR, 1993 Mitosis and morphogenesis in the Drosophila embryo: point and counterpoint, pp. 149–300 in The Development of Drosophila melanogaster, edited by M. BATE and A. MARTINEZ ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

FOGARTY, P., S. CAMPBELL, R. ABU-SHUMAYS, B. PHALLE, and K. YU et al., 1997  The Drosophila grapes gene is related to checkpoint gene chk1/rad27 and is required for late syncytial division fidelity. Curr. Biol. 7:418-426[Medline].

FREEMAN, M. and D. GLOVER, 1987  The gnu mutation of Drosophila causes inappropriate DNA synthesis in unfertilized and fertilized eggs. Genes Dev. 1:924-930.

FREEMAN, M., C. NUSSLEIN-VOLHARD, and D. GLOVER, 1986  The dissociation of nuclear and centrosomal division in gnu, a mutation causing giant nuclei in Drosophila. Cell 46:457-468[Medline].

GALLONI, M. and B. A. EDGAR, 1999  Cell-autonomous and non-autonomous growth-defective mutants of Drosophila melanogaster. Development 126:2365-2375[Abstract].

GLOTZER, M., A. W. MURRAY, and M. W. KIRSCHNER, 1991  Cyclin is degraded by the ubiquitin pathway. Nature 349:132-138[Medline].

GREENSPAN, R. J., 1997 Fly Pushing: The Theory and Practice of Drosophila Genetics. Cold Spring Harbor Laboratory Press, Plainview, NY.

GUADAGNO, T. M. and J. W. NEWPORT, 1996  Cdk2 kinase is required for entry into mitosis as a positive regulator of Cdc2-cyclin B kinase activity. Cell 84:73-82[Medline].

HANAUSKE-ABEL, H. M., M. H. PARK, A. R. HANAUSKE, A. M. POPOWICZ, and M. LALANDE et al., 1994  Inhibition of the G1-S transition of the cell cycle by inhibitors of deoxyhypusine hydroxylation. Biochim. Biophys. Acta 1221:115-124[Medline].

HANAUSKE-ABEL, H. M., B. SLOWINSKA, S. ZAGULSKA, R. C. WILSON, and L. STAIANO-COICO et al., 1995  Detection of a sub-set of polysomal mRNAs associated with modulation of hypusine formation at the G1-S boundary. Proposal of a role for eIF-5A in onset of DNA replication. FEBS Lett. 366:92-98[Medline].

HAYASHI, S., S. HIROSE, T. METCALFE, and A. D. SHIRRAS, 1993  Control of imaginal cell development by the escargot gene of Drosophila.. Development 118:105-115[Abstract].

HUA, X. H., H. YAN, and J. NEWPORT, 1997  A role for Cdk2 kinase in negatively regulating DNA replication during S phase of the cell cycle. J. Cell Biol. 137:183-192[Abstract/Free Full Text].

HUANG, J. and J. W. RAFF, 1999  The disappearance of cyclin B at the end of mitosis is regulated spatially in Drosophila cells. EMBO J. 18:2184-2195[Medline].

JACOBS, H., J. KNOBLICH, and C. LEHNER, 1998  Drosophila Cyclin B3 is required for female fertility and is dispensable for mitosis like Cyclin B. Genes Dev. 12:3741-3751[Abstract/Free Full Text].

KANG, A. H. and J. W. B. HERSHEY, 1994  Effect of initiation factor eIF-5A depletion on protein synthesis and proliferation of Saccharomyces cerevisiae.. J. Biol. Chem. 269:3934-3940[Abstract/Free Full Text].

K