A Screen for Modifiers of Cyclin E Function in Drosophila melanogaster Identifies Cdk2 Mutations, Revealing the Insignificance of Putative Phosphorylation Sites in Cdk2

Corresponding author: Christian F. Lehner, Department of Genetics, University of Bayreuth, 95440 Bayreuth, Germany. E-mail:chle@uni-bayreuth.de

Communicating editor: T. SCHÜPBACH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In higher eukaryotes, cyclin E is thought to control the progression from G1 into S phase of the cell cycle by associating as a regulatory subunit with cdk2. To identify genes interacting with cyclin E, we have screened in Drosophila melanogaster for mutations that act as dominant modifiers of an eye phenotype caused by a Sevenless-CycE transgene that directs ectopic Cyclin E expression in postmitotic cells of eye imaginal disc and causes a rough eye phenotype in adult flies. The majority of the EMS-induced mutations that we have identified fall into four complementation groups corresponding to the genes split ends, dacapo, dE2F1, and Cdk2(Cdc2c). The Cdk2 mutations in combination with mutant Cdk2 transgenes have allowed us to address the regulatory significance of potential phosphorylation sites in Cdk2 (Thr 18 and Tyr 19). The corresponding sites in the closely related Cdk1 (Thr 14 and Tyr 15) are of crucial importance for regulation of the G2/M transition by myt1 and wee1 kinases and cdc25 phosphatases. In contrast, our results demonstrate that the equivalent sites in Cdk2 play no essential role.

ENTRY into or exit from the cell division cycle occurs in general during the G1 phase in eukaryotes. The regulation of progression through the G1 phase, therefore, is crucial for the control of cell proliferation. Progression through the G1 phase is governed by G1 cyclins in association with cyclin-dependent protein kinases (cdks). The understanding of the precise roles of these G1 cyclin/cdk complexes is most advanced in yeast where efficient genetic methodology is available. However, metazoans have a distinct set of G1 cyclin/cdk complexes. The genetic characterization of these higher eukaryote regulators is feasible in Drosophila melanogaster. Moreover, genetic screens for mutations modifying the effects of another mutation have recently been used with great success in this species for the dissection of various signal transduction pathways. Here, we describe a genetic screen designed to identify components involved in the function of cyclin E, one of the higher eukaryote G1 cyclins.

Mammalian cyclin E was originally identified because of its ability to complement the loss of G1 cyclin function in yeast cells (LEW et al. 1991 ). Cyclin E was found to bind to cdk2, which was originally designated as Cdc2c in Drosophila (LEHNER and O'FARRELL 1990 ). Protein kinase activity of cyclin E/cdk2 complexes is maximal in mammalian cells during late G1 and early S phase (DULIC et al. 1992 ; KOFF et al. 1992 ), and the phenotypic characterization of Cyclin E (CycE) mutations in Drosophila has demonstrated that Cyclin E is required for progression into S phase (KNOBLICH et al. 1994 ).

The physiological substrates that need to be phosphorylated by cyclin E/cdk2 for progression into S phase are not yet known very well. The protein product of the retinoblastoma tumor suppressor gene (pRB) appears to be an important physiological target of cyclin E/cdk2 (for a review see MITTNACHT 1998 ). A family of pRB-related genes (pRB, p107, and p130) is present in mammals (WEINBERG 1995 ) and a pRB-related gene (RBF) has been described in Drosophila as well (DU et al. 1996 ; DU and DYSON 1999 ). All RB family proteins bind to E2F/DP heterodimers (for review see DYSON 1998 ; HELIN 1998 ). In mammals, the complexity of E2F and DP genes is considerable (DYSON 1998 ; HELIN 1998 ). In Drosphila, two E2F genes and one DP gene have been identified (DYNLACHT et al. 1994 ; OHTANI and NEVINS 1994 ; HAO et al. 1995 ; SAWADO et al. 1998 ). E2F/DP heterodimers bind to specific DNA sequences in promoters of target genes encoding proteins required for DNA synthesis and cell cycle progression like cyclin E, cyclin A, cdk1, ribonucleotide reductase, DNA polimerase , and proliferating cell nuclear antigen. In many promoters, E2F-binding sites confer repression during G0 or G1. In these promoters, E2F/DP presumably recruits pRB protein, which is responsible for inhibition of target gene transcription. pRB appears to repress transcription by altering chromatin structure as it can bind to histone deacetylases and E2F/DP simultaneously (BREHM et al. 1998 ; LUO et al. 1998 ; MAGNAGHI-JAULIN et al. 1998 ; ZHANG et al. 1999 ). In some target genes, however, E2F/DP-mediated transcriptional activation has been observed, and pRB can interfere with activation at these promoters independent of histone deacetylase recruitment (HARBOUR et al. 1999 ).

As a result of sequential pRB phosphorylation in G1, hyperphosphorylated pRB dissociates from histone deacetylase and E2F/DP and inhibition of target gene expression is relieved. While complexes of D-type cyclins with cdks 4 or 6 are thought to provide the major pRB kinase activity in mid G1, cyclin E/cdk2 contributes to pRB phosphorylation in late G1 as well (HINDS et al. 1992 ; KITAGAWA et al. 1996 ; ZARKOWSKA and MITTNACHT 1997 ; KELLY et al. 1998 ; LUNDBERG and WEINBERG 1998 ; MITTNACHT 1998 ; BROWN et al. 1999 ; HARBOUR et al. 1999 ). In Drosophila, RBF is phosphorylated by Cyclin E/Cdk2 in vitro and genetic interactions have been demonstrated in vivo (DU et al. 1996 ; DU and DYSON 1999 ). Moreover, ectopic Cyclin E expression can generate E2F activity and induce expression of E2F/DP target genes in a number of cell types (DURONIO and O'FARRELL 1995 ; SAUER et al. 1995 ; WEINBERG 1995 ).

Evidence from a number of organisms has indicated that pRB is not the only physiological target of cyclin E/cdk2 activity (RESNITZKY and REED 1995 ; DURONIO et al. 1996 ; LENG et al. 1997 ; LUKAS et al. 1997 ; GENG et al. 1999 ). Moreover, experiments with Xenopus egg extracts have demonstrated that cyclin E/cdk2 controls entry into S phase at a post-transcriptional level (JACKSON et al. 1995 ; STRAUSFELD et al. 1996 ; HUA et al. 1997 ; HUA and NEWPORT 1998 ). Recently, centrosome duplication was found to be dependent on cyclin E/cdk2 in mammalian cells and Xenopus egg extracts (HINCHCLIFFE et al. 1999 ; LACEY et al. 1999 ; MATSUMOTO et al. 1999 ). In these extracts, activation of cyclin B/cdk1 complexes is also dependent on cyclin E/cdk2 (GUADAGNO and NEWPORT 1996 ). The relevant substrates regulated by cyclin E/cdk2 in these cases are completely unknown.

Interestingly, cyclin E and p27kip1, a cdk inhibitor (cdi) that inhibits the activity of cdk2 complexes, are phosphorylated by cyclin E/cdk2 (CLURMAN et al. 1996 ; WON and REED 1996 ; XU and BURKE 1996 ; SHEAFF et al. 1997 ; VLACH et al. 1997 ; MONTAGNOLI et al. 1999 ). In both proteins, this phosphorylation triggers their subsequent proteolytic destruction in cultured mammalian cells. It is not known whether this control of cyclin E and cdi degradation also occurs in Drosophila, where dacapo (dap) has been shown to encode a p27kip1-related cdi specific for Cyclin E/cdk2 complexes (DENOOIJ et al. 1996 ; LANE et al. 1996 ).

Additional regulation of cyclin E/cdk2 activity involves phosphorylation and dephosphorylation of the cdk2 subunit. Apart from the activating phosphorylation of a threonine residue in the activation loop by cdk activating kinase activity, inhibitory phosphorylation of vertebrate cdk2 on a tyrosine residue and on a neighboring threonine residue has also been observed (GU et al. 1992 ; HOFFMANN et al. 1994 ; JINNO et al. 1994 ; BLOMBERG and HOFFMANN 1999 ; KIM et al. 1999 ). Phosphorylation at the corresponding site in cdk1 is brought about by myt1 and wee1 kinases and is reversed by the activity of cdc25 phosphatases. While this inhibitory phosphorylation is known to be of crucial importance for the regulation of cdk1 in vivo, the physiological significance in the case of cdk2 is not understood. Vertebrate cdc25A phosphatase can dephosphorylate and activate cyclin E/cdk2 complexes in vitro and is expressed late in G1 and required for progression into S phase (HOFFMANN et al. 1994 ; BLOMBERG and HOFFMANN 1999 ; KIM et al. 1999 ). Experiments involving premature cdc25A expression during G1 to test the idea that cdk2 is a relevant physiological substrate of cdc25A appear to have given ambivalent results (BLOMBERG and HOFFMANN 1999 ; SEXL et al. 1999 ).

In summary, the functional characterization of cyclin E has clearly established its central role in cell cycle regulation. Moreover, multiple positive and negative autoregulatory loops involving RB family proteins, E2F/DP heterodimers, cdc25A phosphatase, and cdk inhibitors have been implicated in the control of cyclin E/cdk2 activity. However, the significance of the different regulatory loops in vivo is poorly understood. Finally, except for pRB, p27kip1, and cyclin E, little is known about physiological substrates of cyclin E/cdk2 activity.

The design of our genetic approach in Drosophila was based on the finding that ectopic Cyclin E expression in embryos prevents the arrest of cell cycle progression at the appropriate developmental stage (KNOBLICH et al. 1994 ). Therefore, we expected that ectopic Cyclin E expression in postmitotic cells of developing ommatidia would also result in extra cell cycle progression and thereby cause defects in the extremely regular eye pattern that is observed in wild-type flies. Ectopic Cyclin E expression in postmitotic cells during eye development was achieved with a transgene (Sev-CycE) regulated by an enhancer from the sevenless gene. By screening for mutations that dominantly modified the resulting rough eye phenotype, we identified genes encoding proteins known to interact with Cyclin E, like dap, dE2F1, and Cdk2, as well as additional genes like split ends (spen). Moreover, the Cdk2 mutations allowed a demonstration that Cdk2 phosphorylation on threonine 18 and tyrosine 19 (corresponding to the phosphorylation sites controlling Cdk1 activity) does not play an essential role during Drosophila development.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila stocks:
Hs-Cdk1(Cdc2) (STERN et al. 1993 ), Df(3R)H81 (STERN et al. 1993 ), dap⁴ (LANE et al. 1996 ), dE2F1⁹¹ and dE2F1⁷¹⁷² (DURONIO et al. 1995 ), and dDP^a2 and dDP^a4 (ROYZMAN et al. 1997 ) have been described previously as well as the spen alleles E(Raf)2A^16H1 and E(Raf)2A^16T1 (DICKSON et al. 1996 ), poc³⁶¹ and poc²³¹ (GELLON et al. 1997 ), and l(2)03350 and l(2)k13624 (SPRADLING et al. 1995 ).

Fly stocks carrying Sev-CycE transgenes were obtained after P-element-mediated germ-line transformation with pKB267 constructs. pKB267 (kindly provided by Konrad Basler, University of Zurich) contains between a 5' and a 3' P-element end two copies of the sevenless enhancer (BASLER et al. 1991 ), the hsp 70 promoter/leader, a tubulin trailer, and a mini-white⁺ gene. Cyclin E type I or type II cDNA fragments (RICHARDSON et al. 1993 ) comprising the entire coding regions were inserted into the unique KpnI and EcoRI sites between leader and trailer. Several independent lines carrying either the Sev-CycE I or the Sev-CycE II transgene were established. Both transgenes were found to cause a rough eye phenotype, although its expressivity varied with different transgene insertions. The Sev-CycE I insertion III.1 on the third chromosome, which results in a strong phenotype, was used in the interaction screen and in the experiments described here.

Fly stocks carrying a Cdk2⁺ transgene were obtained by P-element-mediated germ-line transformation with a pCaSpeR 4 (PIRROTTA 1988 ) construct containing a 10-kb genomic SalI fragment inserted into the XhoI site of the vector.

Fly stocks with transgenes allowing the expression of wild-type Cdk2 under the control of the heat-shock promoter (Hs-Cdk2(Cdc2c)) have been described previously (STERN et al. 1993 ). Site-directed mutagenesis following the method of CHEN and PRZYBYLA 1994 was used to mutate the codons for potential phosphorylation sites in the parental pCaSpeR-hs construct, which contained an EcoRI-SnaBI Cdk2 cDNA fragment inserted into the EcoRI and StuI sites of the vector. In a first mutant transgene (Hs-Cdk2T18A), the ACC codon for threonine 18 was changed into GCC encoding alanine. In a second mutant transgene (Hs-Cdk2Y19F), the TTC codon for tyrosine 19 was changed into TAC encoding phenylalanine. In the third mutant transgene (Hs-Cdk2AF), we introduced both these codon changes. In the first step of the mutagenesis, we used the primer P1 (5'-CAAA GAATTC GTT TAT TTT GCC AAC ATC-3') in combination with either the primer PmT18A (5'-ACC GTA GGC GCC CTC GCC AAT TCT-3'), PmY19F (5'-TAT ACC GAA GGT GCC CTC GCC AAT-3'), or PmAF (5'-TAT ACC GAA GGC GCC CTC GCC AAT TCT-3') for enzymatic amplification of short fragments from the Cdk2 cDNA. These fragments were extended during a second polymerase chain reaction (PCR) using a third primer P2 (5'-GCC CAA CAG AAT CTC TGG AGC-3'). The resulting fragments were digested with EcoRI and KpnI and used to replace the corresponding fragment in the Hs-Cdk2 construct. The replaced regions were sequenced to verify their correctness. Several independent transgene insertions were established and none of the mutant transgenes were found to induce lethality after expression during embryogenesis. In subsequent experiments, we only characterized Hs-Cdk2AF transgene insertions. Fly stocks (Hs-Cdk1(Cdc2)AF(III.72) and (III.74)) with an analogous transgene allowing the expression of mutant Cdk1 with alanine and phenylalanine instead of threonine and tyrosine at position 14 and 15, respectively, were kindly provided by Patrick O'Farrell, University of California, San Francisco (SPRENGER et al. 1997 ).

CyO and TM3 balancer chromosomes with Act-GFP and Hs-hid transgenes were obtained from the Bloomington Stock Center (Indiana University).

Screen for dominant Sev-CycE modifier mutations:
For EMS mutagenesis, w; iso2; iso3 males were starved for 12 hr and then fed a 1% sucrose solution containing 25 mM EMS for 20 hr. Mutagenized males were crossed to Sev-CycE I (III.1) females. The eye phenotype of 38,000 F₁ progeny was scored under a dissection microscope. Flies with either a stronger or weaker phenotype than normally observed with Sev-CycE I (III.1) were backcrossed individually to w; +/CyO; Sev-CycE I (III.1)/TM2 flies. In cases where eye phenotypes of progeny fell into two distinct classes, linkage tests were performed and balanced lines were established. Further characterization was restricted to chromosomes associated with recessive lethality. Therefore, three chromosomes resulting in enhancement were eliminated from subsequent analyses. By inter se crosses, the recessive lethal mutations were assigned to different complementation groups. The enhancer mutations E(Sev-CycE) A14, E(Sev-CycE)E93, E(Sev-CycE)G36, and E(Sev-CycE)G66 on the second chromosome were found to behave as single hits. The enhancer mutations E(Sev-CycE)D45, E(Sev-CycE)D50, E(Sev-CycE)E19, and E(Sev-CycE)E44 as well as the suppressor mutations S(Sev-CycE)D2, S(Sev-CycE)D16, S(Sev-CycE)D17, S(Sev-CycE)D28, and S(Sev-CycE)D30 on the third chromosome were also found to behave as single hits. Representative alleles of complementation groups with multiple alleles were meiotically mapped by analyzing linkage between recessive visible markers and the dominant interaction with Sev-CycE. An al b c sp chromosome and a ru h th st cu sr e ca chromosome were used to map mutations on the second and third chromosome, respectively. Deficiency mapping was used to verify meiotic map positions.

BrdU pulse labeling, scanning electron microscopy, and immunolabeling:
BrdU pulse labeling of embryos and eye imaginal discs followed by immunolabeling with anti-BrdU antibodies (Becton-Dickinson, San Jose, CA) was done as described (LEHNER et al. 1991 ; STAEHLING-HAMPTON et al. 1999 ). The monoclonal antibody AXD5 against a sperm tail antigen (KARR 1991 ) was used for immunolabeling and Hoechst 33258 for DNA staining of early embryos (LEHNER and O'FARRELL 1989 ). Scanning electron micrographs and plastic sections of adult eyes were prepared as described (BASLER et al. 1991 ).

Induction of heat-inducible transgenes:
To test whether periodic expression of the transgenes Hs-Cdk2 and Hs-Cdk2AF suppresses the lethality resulting from the lack of endogenous Cdk2 function, we recombined transgene insertions on the third chromosome with Df(3R)H81. Flies with the recombinant chromosomes balanced over TM2 were crossed to either Cdk2²/TM3, Sb or Cdk2³/TM3, Sb. Progeny were exposed to periodic heat shocks as described previously (STERN et al. 1993 ). In initial experiments, heat shocks were applied at 3-hr intervals, and the fraction of progeny flies without balancer chromosomes was determined. Under these conditions, rescue of Cdk2/Df(3R)H81 mutants was found to be inefficient, largely because the Df(3R)H81 chromosome dominantly reduced survival. Therefore, we crossed Cdk2²/TM3, Sb flies carrying either Hs-Cdk2 (III.70.1) or Hs-Cdk2AF (II.2) to Cdk2¹/TM3, Sb or Cdk2³/TM3, Sb for the experiments described in detail (see Table 3 and Fig 4).

View larger version (125K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Sev-CycE induces ectopic cell cycle progression. Eye-antennal imaginal discs from wild-type (A) and Sev-CycE (B) larvae at the wandering stage were labeled with BrdU. In Sev-CycE discs, an additional band of BrdU incorporation is observed in the posterior region and is indicated by the solid arrow. The anterior of the discs is oriented to the left, the posterior to the right.

View larger version (184K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Sev-CycE induces a rough eye phenotype sensitive to transgene dose and genetic interactions. Scanning electron micrographs of eyes from adult flies with the genotypes Sev-CycE I (III.1)/+ (A), Sev-CycE I (III.1)/Sev-CycE I (III.1) (B), Sev-CycE I (III.1), Sev-CycE I (III.2)/Sev-CycE I (III.1), Sev-CycE I (III.2) (C), wild type (D), Sev-CycE I (III.1)/Cdk22 (E), and spenE(Sev-CycE)E9/+; Sev-CycE I (III.1)/+ (F).

View larger version (169K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Sev-CycE induces ommatidial irregularities including alterations in photoreceptor cell numbers. Tangential sections of either wild-type (A) or Sev-CycE I (III.1)/Sev-CycE I (III.1) (B) flies are shown. Some of the ommatidial clusters with supernumerary photoreceptor cells are indicated by solid arrows and some with decreased photoreceptor cell numbers are indicated by solid arrowheads.

View larger version (15K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Zygotic Cdk2 function is required for larval growth and long-term survival. (A) Cdk22, Hs-Cdk2/TM3, Sb flies were crossed with Cdk23/TM3, Sb and the progeny collected during a 4-hr egg collection was exposed to periodic heat shocks at 6-hr intervals starting at variable times of development at 25°. The onset of the periodic heat shocks in hours after the end of the egg collection is indicated below the bars. The percentage of Cdk22, Hs-Cdk2/Cdk23 among the total eclosing flies was determined. While early onset of periodic heat pulses resulted in a full rescue of Cdk22, Hs-Cdk2/Cdk23 vitality (33% of Sb+ flies theoretically), no rescue was observed after late onset of periodic heat pulses. At least 200 flies were scored for each time point. (B) The growth of Cdk22/Cdk23 (triangles), Cdk21/Cdk23 (squares), and heterozygous sibling larvae marked by an Act-GFP balancer chromosome (circles) was analyzed as described in MATERIAL AND METHODS. At least 18 larvae were analyzed for each genotype and time point.

To address the role of Cdk2 during oogenesis and early embryogenesis, we crossed Cdk2², Hs-Cdk2/TM3, Hs-hid with either Cdk2¹/TM3, Hs-hid or Cdk2³/TM3, Hs-hid and exposed the progeny to 30-min heat shocks at 12-hr intervals. The resulting Cdk2², Hs-Cdk2/Cdk2¹ or Cdk2², Hs-Cdk2/Cdk2³ females were crossed to male flies that had been raised at 25° to ensure maximal male fertility. For control experiments, such males were also crossed to w females that had been exposed to 30-min heat shocks at 12-hr intervals during development. One-half of the crosses were kept at 25°, while the other half of the crosses was continuously exposed to periodic heat shocks at 12-hr intervals. After 4 days, the flies exposed to periodic heat shocks were distributed into two fresh bottles. One of these bottles was subsequently kept at 25°, while the other was further subjected to periodic heat shocks. During each day of the experiment, egg collections were performed for 5 hr on apple agar plates for counts of laid eggs and for the subsequent 7 hr for fixation and analysis by double labeling with anti-sperm tail antibodies and Hoechst 33258.

To determine the lethality of Hs-transgene expression during embryogenesis, we crossed males homozygous for one of the transgenes (see Table 3) to w virgins. From these crosses, eggs were collected for 2 hr at 25°. After aging for 3 hr, a heat shock was applied by floating the collection plates on a 37° water bath for 30 min. The collection plates were returned to 25° and after a 3-hr incubation, one-half of the collection was exposed to a second heat shock. After an additional incubation for 24 hr at 25°, the fraction of unhatched eggs was determined.

Larval growth measurement:
Flies carrying different Cdk2 alleles balanced with TM3, Act-GFP were crossed and eggs were collected during 3 hr at 25°. Larval progeny was isolated at different time points after egg deposition as described (BRITTON and EDGAR 1998 ). After collection of larvae in 2 M sucrose and freezing in 86% glycerol, green fluorescent protein-positive Cdk2⁺ and GFP-negative Cdk2 mutant larvae were sorted using a Leica MZFLIII microscope equipped with a video camera. Video frames were captured using IPLab Spectrum software and the pixel area covered by individual larvae was determined.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A screen for modifiers of the rough eye phenotype resulting from ectopic expression of Cyclin E under the control of the sevenless enhancer:
During wild-type development, Cyclin E expression is no longer observed in eye imaginal disc cells after their recruitment into differentiating ommatidia (RICHARDSON et al. 1995 ). To achieve ectopic Cyclin E expression in these postmitotic cells of differentiating ommatidia, we used an enhancer element from the sevenless (sev) gene that directs expression in a subset of photoreceptor cells and in cone cells (DICKSON and HAFEN 1993 ). Several lines with a Sev-CycE transgene were established and pulse labeling with BrdU was used to determine whether the ectopic Cyclin E expression from this transgene resulted in extra proliferation in eye imaginal discs. In wild-type eye discs from third instar larvae, we observed the expected, characteristic pattern of BrdU incorporation [ Fig 1A; for a description of cell proliferation during Drosophila eye development see WOLFF and READY 1993 ]. In eye discs from transgenic lines, we obtained an additional posterior band of BrdU incorporation (Fig 1B, arrow). Double labeling of Sev-CycE eye imaginal discs with a DNA stain and antibodies against Elav, a marker for neuronal differentiation that is only expressed in postmitotic photoreceptor cells during wild-type development, revealed mitotic figures in some of the elav-expressing cells (data not shown). These observations indicate that Sev-CycE transgene expression interferes with the arrest of cell cycle progression at the correct developmental stage. The ectopic Cyclin E expression forces many cells into an extra S phase and at least some also through an extra M phase.

In addition, the Sev-CycE transgene was found to cause pattern defects in the eyes of adult flies (Fig 2, A–C). Ommatidia containing more than the normal complement of photoreceptor cells were frequently observed in sections through these eyes (Fig 3B). However, ommatidia with fewer photoreceptor cells were also observed, as well as other pattern irregularities. Loss of cells from ommatidia might result as a secondary consequence of mitotic divisions of differentiating cells, perhaps because extra cells cause difficulties for the establishment of the correct cell-cell contacts required for cell survival. Alternative explanations are not excluded and neomorphic activity of ectopic Cyclin E that is unrelated to its physiological role in proliferating cells might contribute to the rough eye phenotype. Pattern defects in the adult eye have also been observed after ectopic Cyclin E expression from a heat-inducible transgene (RICHARDSON et al. 1995 ).

While the expressivity of the rough eye phenotype was very similar in flies carrying a particular Sev-CycE insertion, eye appearance clearly varied with different insertions, suggesting that the eye phenotype was sensitive to differences in expression levels resulting from transgene position effects. The enhancement of eye roughening resulting from increases in transgene copy number confirmed this notion (Fig 2, A–C). Conversely, heterozygosity for the deficiency Df(3R)H81, which deletes the Cdk2 gene, suppressed the eye roughening in Sev-CycE flies partially. Since a Cdk2⁺ transgene reversed this partial suppression by Df(3R)H81, we conclude that the Sev-CycE phenotype is also sensitive to the Cdk2⁺ dose (data not shown, but see below). The sensitivity of the Sev-CycE phenotype, therefore, appeared suitable for the identification of mutations interacting with Cyclin E/Cdk2 function.

By screening 38,000 flies carrying Sev-CycE and a set of EMS-mutagenized chromosomes, we recovered 32 chromosomes that dominantly modified the rough eye phenotype and were also associated with recessive lethality. A total of 24 mutations acted as enhancers [E(Sev-CycE); for an example see Fig 2F]. A total of 8 mutations behaved as dominant suppressors [S(Sev-CycE); for an example see Fig 2E]. Based on linkage and complementation tests, 13 mutations qualified as single hits, while the rest could be assigned to one of three different complementation groups with multiple alleles (Table 1). One of the single hits and the three complementation groups could be assigned to defined genes as described in the following.

split ends (spen):
The largest complementation group E(Sev-CycE)2A comprised 11 mutations. Meiotic recombination mapping placed this complementation group close to the left end of chromosome 2. Subsequent complementation tests with a series of deficiencies refined the map location to 21B. Additional complementation tests revealed that several alleles of this complementation group failed to complement two P-element insertions, l(2)03350 and l(2)k13624, which had been mapped to this region. These two P elements were found to be inserted 1.5 kb apart from each other and 5' of a large open reading frame. Complementation tests with additional mutations that had been previously assigned to this region revealed that E(Sev-CycE)2A is allelic to E(Raf)2A (DICKSON et al. 1996 ), E(E2F)2A (STAEHLING-HAMPTON et al. 1999 ), and polycephalon (GELLON et al. 1997 ), which has recently been characterized molecularly and found to correspond to spen (WIELLETTE et al. 1999 ; see DISCUSSION).

dE2F1:
The complementation group E(Sev-CycE)3A with five alleles was mapped meiotically to the region that contains the Drosophila dE2F1 gene. Complementation tests indicated that the two analyzed E(Sev-CycE)3A mutations failed to complement dE2F1⁹¹ and dE2F1⁷¹⁷².

dap:
One E(Sev-CycE) mutation (G36) failed to complement the P-element insertion B13-2nd-10 (BIER et al. 1989 ). We have already described a detailed analysis of the corresponding gene, which we have named dacapo (dap) since epidermal cells in dap mutant embryos resume progression through an additional division cycle instead of becoming postmitotic (DENOOIJ et al. 1996 ; LANE et al. 1996 ). dap encodes a cdk inhibitor related to vertebrate CIP/KIP-type inhibitors and the DAP protein has been shown to bind and inhibit specifically Cyclin E/Cdk2 complexes (DENOOIJ et al. 1996 ; LANE et al. 1996 ).

Cdk2:
The complementation group S(Sev-CycE)3A with three alleles was mapped meiotically to the right arm of chromosome III in the genomic region containing the Cdk2 gene. Therefore, we tested whether the lethality that was caused by hemizygosity of these alleles over Df(3R)H81 could be prevented by a Cdk2⁺ transgene. The Cdk2⁺ transgene was found to confer full viability to the three different hemizygotes. Moreover, periodic expression of a heat-inducible transgene (Hs-Cdk2) containing a Cdk2 cDNA under control of a heat-shock promoter was found to prevent the lethality associated with transheterozygous combinations of the three S(Sev-CycE)3A alleles (Table 3). Finally, sequence analysis of the Cdk2 sequence isolated from one of the S(Sev-CycE)3A chromosomes revealed the presence of a premature stop codon (TAG) instead of a glutamine codon (CAG) at position 134 of the predicted amino acid sequence. Based on these results, we conclude that S(Sev-CycE)3A corresponds to the Cdk2 gene. In the following, the alleles are designated as Cdk2¹, Cdk2², and the sequenced allele Cdk2³. The Cdk2³ mutation is likely to eliminate Cdk2 function completely, since translational termination at the premature stop codon at position 134 results in deletion of the C-terminal lobe of the protein kinase domain.

Cdk2 is required zygotically for larval growth and maternally for early embryogenesis:
To characterize the effects of Cdk2 mutations on development, we first analyzed embryogenesis. However, we were unable to identify a requirement for zygotic Cdk2 expression during this developmental phase. Embryos hemizygous or transheterozygous for Cdk2 allele combinations appeared to have wild-type morphology. They incorporated BrdU with normal efficiency and in normal patterns throughout embryogenesis. Hatching of larvae was observed to occur at the same rate as in control embryos.

For the characterization of larval development, we analyzed progeny from parents with Cdk2 alleles over a balancer chromosome carrying Tb. Scoring for the Tb phenotype, which can readily be distinguished from wild type after development beyond the second larval instar, suggested that Cdk2 mutants do not reach this stage. However, experiments involving the expression of Hs-Cdk2 in Cdk2 mutant larvae indicated that these mutants survive for longer time periods than what is normally required to reach second instar (Fig 4A). As indicated above, periodic Hs-Cdk2 expression by heat shocks allowed Cdk2 mutants to develop into morphologically normal and fertile adults. Some Cdk2 mutants were still observed to develop into adults even if the onset of the periodic Hs-Cdk2 expression was delayed until 116 hr after egg deposition (Fig 4A). Cdk2 mutant larvae, therefore, fail to grow but some survive for several days and can be rescued by providing Cdk2 again. The dependence of larval growth on zygotic Cdk2 expression was confirmed by comparing the size of Cdk2⁺ and Cdk2 mutant larvae derived from parents with Cdk2 alleles over a balancer chromosome marked by an Act-GFP transgene. GFP-negative Cdk2 mutant larvae were observed in decreasing numbers for at least 4 days after egg deposition, but their size failed to increase significantly after 1.5 days (Fig 4B).

The normal initial development that was observed in the absence of zygotic Cdk2 function could be explained by maternally derived Cdk2. The presence of a maternal Cdk2 contribution in the Drosophila egg has been demonstrated (LEHNER and O'FARRELL 1990 ). To demonstrate the functional role of this maternal contribution, we analyzed the development of eggs derived from Cdk2 mutant females that had been rescued by periodic Hs-Cdk2 expression. These mutant females (Cdk2², Hs-Cdk2/Cdk2³ or Cdk2², Hs-Cdk2/Cdk2¹) readily laid eggs as long as they were subjected to periodic heat shocks (Table 2). However, after termination of periodic heat shocks, egg deposition decreased rapidly and stopped completely within 2–3 days (Table 2). This arrest of egg deposition was readily reversed within 7 days after resumption of periodic heat shocks.

The eggs from mutant females collected 1 day after the termination of periodic Hs-Cdk2 expression were fixed and stained for DNA. For comparison, we also analyzed the eggs from mutant females that had been maintained with periodic Hs-Cdk2 expression. In addition, we analyzed eggs from w control females exposed to periodic heat shocks or 1 day after termination of these heat shocks. The great majority of the eggs from these w control females revealed normal DNA staining patterns (Fig 5A). Conversely, the majority of the eggs collected from mutant females (Cdk2², Hs-Cdk2/Cdk2³ or Cdk2², Hs-Cdk2/Cdk2¹) displayed abnormal DNA staining patterns (Fig 5, B–D, data not shown). The spatial distribution of nuclei and the appearance of chromatin was often aberrant, indicating that progression through the syncytial division cycles was severely perturbed in these embryos. This finding suggests that the maternal contribution is required during the syncytial division cycles. In addition, a significant fraction of embryos contained very few nuclei (Fig 5E and Fig F), suggesting that they had failed to commence progression through the syncytial divisions (although it is not excluded that a minor fraction of these eggs were fixed while progressing normally through the first three cycles). Double labeling with an antibody recognizing a sperm tail epitope (Fig 5G and Fig H) indicated that about two-thirds of these eggs with less than five nuclei were not fertilized. Compared to continuously heat-pulsed Cdk2 mutant females, those withdrawn from heat-shock treatment generated a higher fraction of eggs containing less than five nuclei at the expense of the eggs with normal appearance. Termination of periodic Hs-Cdk2 expression in Cdk2 mutant females, therefore, is accompanied by a transient production of eggs that cannot be fertilized followed by a rapid arrest of egg laying. A significant fraction (60%) of abnormal eggs were produced even when periodic Hs-Cdk2 expression was maintained. The production of abnormal eggs despite periodic Hs-Cdk2 expression is likely to reflect the fact that the germ line is refractory to induction of heat-shock genes during stages 10–12 of oogenesis (WANG and LINDQUIST 1998 ). Our observations demonstrate that Cdk2 expression is crucial for oogenesis and early embryogenesis.

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Maternal Cdk2 function is required for early embryonic cell cycle progression. Eggs from wild-type females (A) and Cdk22, Hs-Cdk2/Cdk23 females collected 1 day after the last of periodic heat shocks stained for DNA (A–F) and with anti-sperm tail antibodies (G and H) are shown to illustrate the various phenotypic classes observed. E, G and F, H, respectively, show the same egg. For further details see text.

Cdk2 lacking the conserved putative phosphorylation sites Thr18 and Tyr19 can provide all essential functions in vivo:
Inhibition of Cyclin E/Cdk2 appears to be required for the arrest of cell proliferation at the appropriate developmental stage. In the embryonic epidermis, upregulation of dap expression is known to occur immediately before the epidermal cells exit from the mitotic cell cycle. In dap mutants an additional division cycle is observed instead of a proliferation arrest (DENOOIJ et al. 1996 ; LANE et al. 1996 ). In parallel to dap upregulation, downregulation of Cyclin E expression is observed in the embryonic epidermis (KNOBLICH et al. 1994 ). It is not known what triggers this simultaneous up- and downregulation and, given the complexity of the regulatory loops controlling progression into the cell cycle, it was not excluded that inhibition of Cyclin E/Cdk2 by inhibitory phosphorylation of the Cdk2 subunit, as known to occur in cultured mammalian cells, might contribute to Cyclin E/Cdk2 inactivation at the stage when the embryonic epidermal cells exit from the cell cycle. Moreover, upregulation of dap expression does not always precede cell proliferation arrest during Drosophila development. A cell proliferation arrest independent of dap upregulation, for instance, is observed in the region of the wing margin in third instar imaginal discs and in the region in front of the advancing morphogenetic furrow in the eye imaginal disc (DENOOIJ et al. 1996 ; LANE et al. 1996 ; JOHNSTON and EDGAR 1998 ). Inhibitory Cdk2 phosphorylation, therefore, might potentially replace inhibition by DAP.

To address the role of inhibitory Cdk2 phosphorylation, we mutated the conserved amino acid residues at the positions corresponding to the sites of inhibitory phosphorylation in Cdk1 and mammalian cdk2. By mutating the threonine and tyrosine codons at positions 18 and 19 into alanine and phenylalanine codons, respectively, the potential phosphate acceptor sites were eliminated. We established transgenes, allowing conditional expression of the mutant kinase under control of the heat-shock promoter.

In a first experiment, we analyzed whether expression of mutant Cdk2 at the stage of embryogenesis when epidermal cells exit from the cell cycle prevents a timely arrest of cell proliferation. BrdU pulse labeling failed to reveal extra proliferation (data not shown), suggesting that the cell cycle arrest in the embryonic epidermis is not dependent on inhibitory Cdk2 phosphorylation. In further experiments, we addressed whether expression of the mutant kinase at other developmental stages caused lethality or morphological abnormalities. However, we failed to observe effects of Hs-Cdk2AF (Table 3). These negative results raised the question whether the mutant kinase was expressed and active in vivo. Therefore, we crossed the Hs-Cdk2AF transgene into a Cdk2 mutant background and tested whether periodic expression could rescue the lethality associated with the Cdk2 mutant background. As shown in Table 3, Hs-Cdk2AF rescued the lethality with the same efficiency as Hs-Cdk2. We conclude therefore that Cdk2, which cannot be phosphorylated on the sites corresponding to those that are of crucial importance in the case of Cdk1 regulation, can provide all the essential functions during Drosophila development.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Although the crucial role of cyclin E/cdk2 complexes in cell cycle control has been clearly demonstrated, their precise functions have yet to be defined. Thus we have designed a genetic approach in D. melanogaster for the identification of genes involved in Cyclin E/cdk2 function. By screening for mutations that act as dominant modifiers of the rough eye phenotype caused by ectopic Cyclin E expression under the control of the sevenless enhancer, we have identified the genes dE2F1, dap, and Cdk2(Cdc2c), which are known to encode proteins that are intimately involved in Cyclin E function. The future analysis of the additional uncharacterized modifier mutations that we have identified, therefore, can be expected to define novel components involved in Cyclin E/Cdk2 function.

The largest complementation group identified in our screen, E(Sev-CycE)2A, has not previously been implicated in Cyclin E function. Our genetic and molecular analysis of E(Sev-CycE)2A suggested that it corresponds to the spen gene. Independent work has proven this suggestion (WIELLETTE et al. 1999 ). spen encodes a 600-kD ubiquituously expressed nuclear protein containing three RNP-type RNA binding domains and a novel characteristic C-terminal domain defining a family of homologous metazoan genes. Mutations in spen result in peripheral nervous system defects (KOLODZIEJ et al. 1995 ; KUANG et al. 1999 ) and interact with raf kinase signaling (DICKSON et al. 1996 ) and the function of HOX (GELLON et al. 1997 ; WIELLETTE et al. 1999 ) and E2F/DP transcription factors (STAEHLING-HAMPTON et al. 1999 ) and, as shown here, with Cyclin E. It is attractive to speculate that spen is particularly important for the transition from cell proliferation to terminal differentiation. spen mutant ommatidia in the eye (DICKSON et al. 1996 ) display the same defects as those resulting from the Sev-CycE transgene. The affected ommatidia are of variable composition, often lacking either R7 or one or more other photoreceptors, but also occasionally containing extra photoreceptors. Sev-CycE transgene expression forces differentiating cells through an extra cell cycle, presumably explaining the presence of extra cells. In addition, extra divisions of differentiating cells are likely to disturb the regular arrangement of the ommatidial cluster and consequently might cause apoptosis, potentially explaining the observed loss of cells as well. Similarly, coexpression of GMR-E2F1 and GMR-DP transgenes in all eye imaginal disc cells posterior to the morphogenetic furrow has been shown to result in ectopic BrdU incorporation and apoptosis. spen mutations dominantly enhance both the Sev-CycE and GMR-E2F1/DP rough eye phenotype. Conversely, spen mutations suppress the eye phenotypes resulting from GMR-dap expression in a CycE heterozygous background (STAEHLING-HAMPTON et al. 1999 ). While spen function opposes the mitogenic activity of CycE and dE2F1, it remains to be analyzed whether the phenotypic interactions observed between spen and Hox and Raf involve deregulated cell proliferation as well.

The identification of mutations in Drosophila dE2F1 in our screen was expected on the basis of the large body of evidence demonstrating the tight functional relationship between Cyclin E and E2F/DP transcription factors. However, the fact that dE2F1 mutations resulted in enhancement rather than suppression of the Sev-CycE phenotype would not necessarily have been predicted since the results of genetic analysis in Drosophila so far have suggested that E2F/DP activity has a positive role in stimulating the transcription of S phase genes (Cyclin E, RNR2, DNA pol , PCNA, and Orc1) and cell proliferation (DURONIO et al. 1995 , DURONIO et al. 1998 ; ROYZMAN et al. 1997 ; ASANO and WHARTON 1999 ; SECOMBE et al. 1999 ). In contrast, the enhancement of the Sev-CycE phenotype observed with dE2F1 alleles points to a growth-suppressive role of dE2F1. Similarly, the E2F1 knock-out phenotype observed in mice has clearly demonstrated a tumor-suppressing function (FIELD et al. 1996 ; YAMASAKI et al. 1996 ). Moreover, while vertebrate E2F/DP functions as a transcriptional activator in some promoters, it acts as a corepressor in conjunction with pRB in many other promoters (for a review see DYSON 1998 ). A decrease in E2F/DP levels, therefore, might also result in derepression of unknown proliferation-stimulating genes and synergy with ectopic Cyclin E expression.

An understanding of the observed enhancement will require extensive and careful analysis given the complexity of the interactions between Cyclin E, E2F/DP, and RBF. We point out, however, that enhancement of the Sev-CycE phenotype was also observed with the putative null allele dE2F1⁹¹, suggesting that we have not selected for dominant alleles in our screen (C. LEHNER, data not shown). dDP alleles were not recovered in our screen, and a modification of the Sev-CycE phenotype by putative dDP null alleles could not be observed (C. LEHNER, data not shown). With the exception of spen alleles, we also did not recover mutations in other genes identified in the screen for E2/DP interactors (STAEHLING-HAMPTON et al. 1999 ; C. LEHNER, data not shown).

Cdk2 alleles were isolated as suppressor mutations, as expected. Our initial characterization of the Cdk2 mutant phenotypes clearly demonstrates that Cdk2 is required in females for oogenesis and for provision of a maternal contribution to the egg. This maternal contribution allows development in the absence of zygotic function until the first larval instar. After that stage, zygotic Cdk2 expression is clearly required for larval growth and survival.

The identification of mutations in Cdk2 has allowed us to evaluate the physiological significance of potential phosphorylation sites (Thr 18 and Tyr 19) that have been implicated in negative regulation of Cdk2 activity. The corresponding sites in Cdk1 (Thr 14 and Tyr 15) are known to be of paramount importance for physiological regulation. Expression of Cdk1 mutant protein (Cdk1AF) that can no longer be phosphorylated results in premature mitosis and inability to delay entry into mitosis in the presence of unreplicated or damaged DNA. Cdk1AF expression from a heat-inducible transgene in Drosophila embryogenesis results in lethality. Conversely, expression of an analogous Cdk2 mutant protein (Cdk2AF) has apparently no effect during Drosophila development. Moreover, Hs-Cdk2AF expression can restore normal development in mutants lacking endogenous Cdk2 completely, demonstrating that Cdk2AF is functional. Our results therefore clearly demonstrate that phosphorylation on Thr 18 and Tyr 19 does not represent an essential level of Cdk2 regulation. The dap-dependent cell cycle arrest in G1 after the terminal mitosis 16 occurs at the correct developmental stage in the embryonic epidermis in the presence of Cdk2AF. In addition, the dap-independent G1 arrest in front of the approaching morphogenetic furrow in eye imaginal discs does not appear to be abolished by Hs-Cdk2AF expression. Failure to arrest in G1 ahead of the advancing morphogenetic furrow is known to result in a rough eye phenotype (THOMAS et al. 1994 ), which is not observed after Cdk2AF expression.

Our demonstration that phosphorylation of Cdk2 on Thr 18 and Tyr 19 has no essential role during normal development does not exclude its involvement in subtle or stress regulation. Moreover, we also point out that vertebrate cells, in which Cdk2 phosphorylation on Thr 18 and Tyr 19 has been demonstrated to occur, express A-type cdc25 phosphatases that have been implicated in Cdk2 dephosphorylation and that do not appear to exist in Drosophila.

	FOOTNOTES

¹ Present address: Department of Pharmacology, University of Massachusetts Medical School, 55 Lake Ave. North, Worcester, MA 01655.
² Present address: Institut für Molekularbiologie und Tumorforschung, Emil-Mannkopff-Str. 2, 35037 Marburg, Germany.

	ACKNOWLEDGMENTS

We thank Nick Dyson, Elizabeth Wiellette, and William McGinnis for sharing unpublished results; Pat O'Farrell, Bob Duronio, Terry Orr-Weaver, Ernst Hafen, and Konrad Basler for providing fly stocks and plasmids; and Eva Illgen for technical help. Ernst Hafen also generously assisted in the characterization of eye phenotypes. Moreover, we acknowledge the help of Petra Roeglin during mapping experiments and of Hubert Jaeger and the great value of the Berkeley Drosophila Genome Project in the molecular characterization of E(Sev-CycE)2A. This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG Le/1-1 and DFG Le/2-1).

Manuscript received November 23, 1999; Accepted for publication January 18, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ASANO, M. and R. P. WHARTON, 1999  E2F mediates developmental and cell cycle regulation of ORC1 in Drosophila. EMBO J. 18:2435-2448[Medline].

BASLER, K., B. CHRISTEN, and E. HAFEN, 1991  Ligand-independent activation of the sevenless receptor tyrosine kinase changes the fate of cells in the developing Drosophila eye. Cell 64:1069-1081[Medline].

BIER, E., H. VAESSIN, S. SHEPHERD, K. LEE, and K. MCCALL et al., 1989  Searching for pattern and mutation in the Drosophila genome with a P-lacZ vector. Genes Dev. 3:1273-1287[Abstract/Free Full Text].

BLOMBERG, I. and I. HOFFMANN, 1999  Ectopic expression of Cdc25A accelerates the G1/S transition and leads to premature activation of cyclin E- and cyclin A-dependent kinases. Mol. Cell. Biol. 19:6183-6194[Abstract/Free Full Text].

BREHM, A., E. A. MISKA, D. J. MCCANCE, J. L. REID, and A. J. BANNISTER et al., 1998  Retinoblastoma protein recruits histone deacetylase to repress transcription. Nature 391:597-601[Medline].

BRITTON, J. S. and B. A. EDGAR, 1998  Environmental control of the cell cycle in Drosophila: nutrition activates mitotic and endoreplicative cells by distinct mechanisms. Development 125:2149-2158[Abstract].

BROWN, V. D., R. A. PHILLIPS, and B. L. GALLIE, 1999  Cumulative effect of phosphorylation of pRB on regulation of E2F activity. Mol. Cell. Biol. 19:3246-3256[Abstract/Free Full Text].

CHEN, B. and A. E. PRZYBYLA, 1994  An efficient site-directed mutagenesis method based on PCR. Biotechniques 17:657-659[Medline].

CLURMAN, B. E., R. J. SHEAFF, K. THRESS, M. GROUDINE, and J. M. ROBERTS, 1996  Turnover of cyclin E by the ubiquitin-proteasome pathway is regulated by cdk2 binding and cyclin phosphorylation. Genes Dev. 10:1979-1990[Abstract/Free Full Text].

DENOOIJ, 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].

DICKSON, B., and E. HAFEN, 1993 Genetic dissection of eye development in Drosophila, pp. 1327–1362 in The Development of Drosophila, edited by M. BATE and A. MARTINEZ ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

DICKSON, B. J., A. VANDERSTRATEN, M. DOMINGUEZ, and E. HAFEN, 1996  Mutations modulating raf signaling in Drosophila eye development. Genetics 142:163-171[Abstract].

DU, W. and N. DYSON, 1999  The role of RBF in the introduction of G1 regulation during Drosophila embryogenesis. EMBO J. 18:916-925[Medline].

DU, W., M. VIDAL, J. E. XIE, and N. DYSON, 1996  RBF, a novel RB-related gene that regulates E2F activity and interacts with cyclin E in Drosophila. Genes Dev. 10:1206-1218[Abstract/Free Full Text].

DULIC, V., E. LEES, and S. I. REED, 1992  Association of human cyclin E with a periodic G(1)-S phase protein-kinase. Science 257:1958-1961[Abstract/Free Full Text].

DURONIO, R. J. and P. H. O'FARRELL, 1995  Developmental control of the G1 to S transition in Drosophila—cyclin E is a limiting downstream target of E2F. Genes Dev. 9:1456-1468[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].

DURONIO, R. J., A. BROOK, N. DYSON, and P. H. O'FARRELL, 1996  E2F-induced S phase requires cyclin E. Genes Dev. 10:2505-2513[Abstract/Free Full Text].

DURONIO, R. J., P. C. BONNETTE, and P. H. O'FARRELL, 1998  Mutations of the Drosophila dDP, dE2F, and cyclin E genes reveal distinct roles for the E2F-DP transcription factor and cyclin E during the G1-S transition. Mol. Cell. Biol. 18:141-151[Abstract/Free Full Text].

DYNLACHT, B. D., M. S. BROOK, M. S. DEMBSKI, L. YENUSH, and N. DYSON, 1994  DNA-binding and trans-activation properties by Drosophila E2F and DP proteins. Proc. Natl. Acad. Sci. USA 91:6359-6363[Abstract/Free Full Text].

DYSON, N., 1998  The regulation of E2F by pRB-family proteins. Genes Dev. 12:2245-2262[Free Full Text].

FIELD, S. J., F. Y. TSAI, F. KUO, A. M. ZUBIAGA, and W. G. KAELIN, JR. et al., 1996  E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell 85:549-561[Medline].

GELLON, G., K. W. HARDING, N. MCGINNIS, M. M. MARTIN, and W. MCGINNIS, 1997  A genetic screen for modifiers of Deformed homeotic function identifies novel genes required for head development. Development 124:3321-3331[Abstract].

GENG, Y., W. WHORISKEY, M. Y. PARK, R. T. BRONSON, and R. H. MEDEMA et al., 1999  Rescue of cyclin D1 deficiency by knockin cyclin E. Cell 97:767-777[Medline].

GU, Y., J. ROSENBLATT, and D. O. MORGAN, 1992  Cell cycle regulation of cdk2 activity by phosphorylation of Thr160 and Tyr15. EMBO J. 11:3995-4005[Medline].

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].

HAO, X. F., L. ALPHEY, L. R. BANDARA, E. W. F. LAM, and D. GLOVER et al., 1995  Functional conservation of the cell cycle-regulating transcription factor DRTF1/E2F and its pathway of control in Drosophila melanogaster.. J. Cell Sci. 108:2945-2954[Abstract].

HARBOUR, J. W., R. X. LUO, A. DEI SANTI, A. A. POSTIGO, and D. C. DEAN, 1999  Cdk phosphorylation triggers sequential intramolecular interactions that progressively block Rb functions as cells move through G1. Cell 98:859-869[Medline].

HELIN, K., 1998  Regulation of cell proliferation by the E2F transcription factors. Curr. Opin. Genet. Dev. 8:28-35[Medline].

HINCHCLIFFE, E. H., C. LI, E. A. THOMPSON, J. L. MALLER, and G. SLUDER, 1999  Requirement of Cdk2-cyclin E activity for repeated centrosome reproduction in Xenopus egg extracts. Science 283:851-854[Abstract/Free Full Text].

HINDS, P. W., S. MITTNACHT, V. DULIC, A. ARNOLD, and S. I. REED et al., 1992  Regulation of retinoblastoma protein functions by ectopic expression of human cyclins. Cell 70:993-1006[Medline].

HOFFMANN, I., G. DRAETTA, and E. KARSENTI, 1994  Activation of the phosphatase activity of human cdc25A by a cdk2/cyclin E-dependent phosphorylation at the G1/S transition. EMBO J. 13:4302-4310[Medline].

HUA, X. H. and J. NEWPORT, 1998  Identification of a preinitiation step in DNA replication that is independent of origin recognition complex and cdc6, but dependent on cdk2. J. Cell. Biol. 140:271-281[Abstract/Free Full Text].

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].

JACKSON, P. K., S. CHEVALIER, M. PHILIPPE, and M. W. KIRSCHNER, 1995  Early events in DNA replication require cyclin E and are blocked by p21^Cip1. J. Cell Biol. 130:755-769[Abstract/Free Full Text].

JINNO, S., K. SUTO, A. NAGATA, M. IGARASHI, and Y. KANAOKA et al., 1994  Cdc25A is a novel phosphatase functioning early in the cell cycle. EMBO J. 13:1549-1556[Medline].

JOHNSTON, L. A. and B. A. EDGAR, 1998  Wingless and Notch regulate cell-cycle arrest in the developing Drosophila wing. Nature 394:82-84[Medline].

KARR, T. L., 1991  Intracellular sperm egg interactions in Drosophila: a 3-dimensional structural analysis of a paternal product in the developing egg. Mech. Dev. 34:101-111[Medline].

KELLY, B. L., K. G. WOLFE, and J. M. ROBERTS, 1998  Identification of a substrate-targeting domain in cyclin E necessary for phosphorylation of the retinoblastoma protein. Proc. Natl. Acad. Sci. USA 95:2535-2540[Abstract/Free Full Text].

KIM, S. H., C. LI, and J. L. MALLER, 1999  A maternal form of the phosphatase Cdc25A regulates early embryonic cell cycles in Xenopus laevis. Dev. Biol. 212:381-391[Medline].

KITAGAWA, M., H. HIGASHI, H. K. JUNG, I. SUZUKITAKAHASHI, and M. IKEDA et al., 1996  The consensus motif for phosphorylation by cyclin D1-Cdk4 is different from that for phosphorylation by cyclin A/E- Cdk2. EMBO J. 15:7060-7069[Medline].

KNOBLICH, J. A., K. SAUER, L. JONES, H. RICHARDSON, and R. SAINT et al., 1994  Cyclin E controls S phase progression and its downregulation during Drosophila embryogenesis is required for the arrest of cell proliferation. Cell 77:107-120[Medline].

KOFF, A., A. GIORDANO, D. DESAI, K. YAMASHITA, and J. W. HARPER et al., 1992  Formation and activation of a cyclin E-cdk2 complex during the G1-phase of the human cell cycle. Science 257:1689-1694[Abstract/Free Full Text].

KOLODZIEJ, P. A., L. Y. JAN, and Y. N. JAN, 1995  Mutations that affect the length, fasciculation, or ventral orientation of specific sensory axons in the Drosophila embryo. Neuron 15:273-286[Medline].

KUANG, B., S. WU, Y. SHIN, K.A. MACE, L. LUO, and P. KOLODZIEJ, 1999  split ends is required for axon growth and guidance in Drosophila embryos and encodes a member of a new family of proteins containing RNP motifs. Development 127:1517-1529[Abstract].

LACEY, K. R., P. K. JACKSON, and T. STEARNS, 1999  Cyclin-dependent kinase control of centrosome duplication. Proc. Natl. Acad. Sci. USA 96:2817-2822