BIOLOGICAL OVERVIEW

Orderly progression through a typical somatic cell cycle requires the cyclic activity of several distinct cyclin-dependent kinase (Cdk)/cyclin complexes. Entry into the DNA synthesis (S) phase requires the activity of Cdk2, which forms complexes with cyclin E in late G1 phase (Knoblich, 1994). A number of other Drosophila Cdks and cyclins have been identified but their roles in regulation of the cell cycle have not been determined. Some of these uncharacterized Cdk/cyclin complexes may contribute to cell cycle regulation in specific developmental contexts. Cyclin J has an RNA expression pattern that suggests a possible role in early embryogenesis. Cyclin J protein is present in early embryos where it forms active kinase complexes with cyclin-dependent kinase 2 (Cdc2c, also known as Cdk2). To determine whether Cyclin J plays a role in controlling the early nuclear cycles peptide, aptamers (compounds that bind specific proteins) were isolated that specifically bind to Cyclin J and inhibit its ability to activate Cdks. The inhibitory aptamers were injected into syncytial Drosophila embryos and they cause defects in chromosome segregation and progression through mitosis. These results suggest that a Cyclin J-associated kinase activity is required for the early embryonic division cycles, regulating the rapid S phases in the early embryo. The Cyclin J aptamers inhibit Cyclin J kinase with efficiencies comparable to that of a natural Cdk inhibitor. It is suggested that Cyclin J/Cdk2 dimers fulfil the same function in regulating S phase in Drosophila as do CyclinA/Cdk2 dimers in mammals (Kolonin, 2000).

Cdk1 and Cdk2 are both abundant in the early embryo, as are cyclins A, B, B3, and E. Maternal Cdk1 and cyclin B have been shown to be required for the early embryonic cycles, as has Cdk2 (Lane, 2000), whereas a requirement for the other Cdks and cyclins has not yet been demonstrated. At least some of the unique features of the nuclear division cycles may be mediated by novel regulation of these cell cycle proteins. For example, in contrast to later cycles, the overall levels of cyclin A and cyclin B do not oscillate during the early nuclear division cycles. Instead, exit from mitosis coincides with localized degradation of mitotic cyclins and localized Cdk1 inactivation along the chromosomes (Kolonin, 2000 and references therein). In addition to this embryo-specific regulation of known Cdks and cyclins, the early embryo contains Cyclin J, which may contribute to some of the unique features of the nuclear division cycles. In contrast with other cyclins, Cyclin J is not expressed later in embryogenesis, but instead is supplied maternally. Cyclin J has been shown to interact with Cdk2 (Finley, 1996).

A useful approach to studying the early embryonic division cycles has been to inject embryos with antibodies or small molecules that specifically inhibit maternally deposited proteins. To investigate the function of Cyclin J in early development, attempts were made to isolate peptides that could be injected into living embryos to interfere with Cyclin J function. First, peptide aptamers were identified that bind specifically to Cyclin J. A yeast two-hybrid was screened with a library that expresses random 20 amino acid peptides on the surface of a platform molecule, E. coli thioredoxin (trxA) (Colas, 1996), and a number of trxA-fused peptides that interact with Cyclin J were isolated. Several of the cyclin J-interacting peptides (JA1, JA2, and JA3) interact specifically with Cyclin J but not with Drosophila cyclins A, D, or C or cyclins E or H. A control trxA-fused peptide, PepC2 (Colas, 1996), or a peptide aptamer specific for cyclin D, DA2, did not interact with Cyclin J. The sequences of the Cyclin J aptamers share no strong similarity with each other or with previously identified proteins (Kolonin, 2000).

To determine whether the Cyclin J peptide aptamers interfere with Cyclin J function, their ability to inhibit Cyclin J-associated kinase activity was tested. Cyclin complexes were immunoprecipitated from embryo extracts with affinity-purified Cyclin J antibodies and the histone H1 kinase activity of the complexes was assayed in the presence or absence of purified trxA-fused peptides. In these experiments the peptides were incubated with the complexes after immunoprecipitation. The Cyclin J-associated kinase activity in the precipitates is inhibited by aptamer JA1 and JA2 at peptide concentrations of 1 to 10 mM, but not by the control peptide, pepC2. JA1 and JA2 inhibit the Cyclin J kinase immunoprecipitated from syncytial embryos and from later embryos that ectopically express Cyclin J; the later extracts may contain Cyclin J in complexes with Cdk1 in addition to Cdk2. To compare the effects of JA1 and JA2 with those of known Cdk inhibitors, Cyclin J- and E-associated kinases were tested for inhibition by a Cdk2 inhibitor Dacapo. Dap is known to efficiently inhibit cyclin E-associated kinase activity. Dap also inhibits the Cyclin J-associated kinase with half-maximal inhibition at 1-3 mM (Kolonin, 2000).

A functional assay was used in yeast to test whether the Cyclin J aptamers could inhibit Cyclin J in living cells. Like many other cyclins, Drosophila Cyclin J is able to complement cln^- mutant yeast which lack the three yeast G1 cyclins, CLN1, CLN2, and CLN3. TrxA-fused JA1 or JA2 were expressed in strains of cln^- yeast that depended on a Drosophila cyclin for growth. The Cyclin J peptide aptamers dramatically inhibit the growth of Cyclin J-dependent yeast. The observed functional inhibition of Cyclin J by the aptamers may result from the aptamer blocking the Cyclin J-Cdk interaction, as suggested by two-hybrid experiments for JA1. Alternatively, an aptamer may inhibit kinase activity by another mechanism, for example, by blocking access to the substrate. Combined, these results indicate that JA1 and JA2 are efficient and specific inhibitors of Cyclin J function. Both aptamers bind to Cyclin J and not to other cyclins; both specifically inhibit Cyclin J-associated kinase activity from embryo extracts, and both are able to specifically inhibit Cyclin J function in a yeast assay (Kolonin, 2000).

To determine the effects of inhibiting Cyclin J function in vivo, Cyclin J aptamers or affinity-purified Cyclin J antibodies were injected into syncytial wild-type Drosophila embryos. Purified peptides fused to the trxA platform were injected because previous results had suggested that trxA peptides have higher affinity for their targets than the peptides alone. After injection with the aptamers or antibodies, embryos were incubated for variable times before fixation to allow nuclei to proceed through additional cycles. Localization of the injected peptides or antibodies was determined by immunofluorescence. For each injected agent, approximately 200 embryos were stained with a DNA-staining reagent, Hoechst 33258. This allowed investigators to look for cell cycle defects that colocalize with the injection site (Kolonin, 2000).

Injection of the Cyclin J aptamers or antibodies results in profound effects on the syncytial cycles. Telophase chromatin bridges near the injection site are induced at high frequency by both the aptamers and the Cyclin J antibodies. In addition to the chromosome segregation defects, the Cyclin J aptamers and antibodies cause an apparent delay in progression through mitosis. In many of the embryos that display localized staining for the aptamer or antibody, nuclei proximal to the injection site are at an early mitotic stage, whereas nuclei distal to the injection site have proceeded to progressively later stages. The gradient of mitotic progression and the higher frequency of chromatin bridges near the injection site suggest that the effect of the Cyclin J inhibitors is dosage-dependent. Injected trxA-fused peptides rapidly diffuse throughout the embryo and are no longer localized to the injection site within ~ 20 min after injection. Consistent with this, local cell cycle defects are observed only when embryos are fixed within 10 min of peptide injection. To quantify the effects on cell cycle progression in embryos fixed later than 10 min after injection, the cell cycle stages and the frequencies of chromatin bridges among all injected embryos were scored. Cyclin J aptamers and antibodies each cause chromatin bridges even in the embryos that do not display localized defects. The Cyclin J aptamers and antibodies also significantly increase the overall frequency of embryos with nuclei in mitosis. These results suggest that a cyclin J-associated kinase activity is required for normal nuclear divisions during early embryonic development (Kolonin, 2000).

It is concluded that Cyclin J is a candidate early embryo-specific cell cycle regulator. Consistent with this possibility, injection of early embryos with specific Cyclin J inhibitors disrupts progression through the rapid division cycles. Also, whereas other cyclins are able to induce progression through the cell cycle when overexpressed in late embryogenesis, no abnormal cell divisions is detected when Cyclin J is ectopically expressed in the embryo after the endogenous protein is degraded. Combined, these results suggest that Cyclin J is required for early embryogenesis but not for cell divisions after cellularization (Kolonin, 2000).

The primary Cdk partner of Cyclin J in the early embryo is Cdk2. A view supported by studies in a number of organisms is that the functional specificity of a Cdk/cyclin complex resides in the cyclin subunit. However, the Cdk subunit can also determine the biological function of the complex. In vertebrates, for example, Cdk1/cyclin A complexes control mitotic events, whereas Cdk2/cyclin A complexes are required for S phase. Moreover, although Cdk1 and Cdk2 each complex with more than one cyclin, all known Cdk1 complexes function in promoting M phase while all known Cdk2 complexes function in promoting S phase. Thus, the association of Cyclin J with Cdk2 in the Drosophila embryo raises the possibility that Cdk2/Cyclin J complexes may function during S phase, similar to vertebrate Cdk2/cyclin A complexes. In this regard it may be significant that the closest homolog of Cyclin J is cyclin A (Kolonin, 2000).

The phenotypic consequences of inhibiting Cyclin J are consistent with a possible role for Cyclin J in regulating the rapid S phases in the early embryo. Inhibition of cyclin J-associated kinase activity results in a delay in progression through mitosis and abnormal chromosome segregation. Nuclei entered mitosis with chromosomes remaining unresolved, leading to telophase daughter nuclei connected by chromatin bridges. A similar phenotype has been observed upon injection of agents that inhibit proteins required for DNA synthesis such as DNA polymerase alpha and MCM proteins, but not by inhibitors of transcription, protein synthesis, or microtubule dynamics. Telophase chromatin bridges arise presumably because Drosophila embryos lack the checkpoint that in later cycles prevents entry into mitosis in the presence of unreplicated or damaged DNA. In addition to inducing chromatin bridges, DNA replication inhibitors, such as aphidicolin, can slow progression through mitosis in the early embryo. This can have the net effect of increasing the number of nuclei in various stages of mitosis, which is observed upon injection of Cyclin J inhibitors. In contrast, inhibition of proteins required during mitosis causes an arrest at a specific mitotic stage, without chromatin bridges. The possibility of Cyclin J being an early embryo-specific S phase regulator seems particularly intriguing given the fact that genome duplication is proceeding at a uniquely rapid pace in the syncytium. The phenotypes resulting from Cyclin J loss-of-function mutations may provide additional information on the precise cell cycle function of this unique cyclin (Kolonin, 2000).

GENE STRUCTURE

There is an overlapping repeat (gaaatccaaatagcttagaagtgaaa) in the 3' UTR of the mRNA (Finley, 1996)

Exons - 4

Bases in 3' UTR - 230

PROTEIN STRUCTURE

Amino Acids -

Structural Domains391

The amino acid sequence of Cyclin J is most similar to that of the A-type cyclins, but lacks the destruction box and other consensus motifs found in all mitotic cyclins. The Cdk interactor (Cdi) Cdi5 (the term applied to Cyclin J in this study) also has strong similarity to cyclins but defines a protein of a new sequence type. From residue 74 to 223, Cdi5 contains 16 matches to the 20 residues conserved in most cyclins. While Cdi5 is most similar to certain A cyclins and B cyclins, it does not fit the consensus for the A or B class; Cdi5 lacks 50 of the 86 residues common to known A cyclins, and 16 of the 35 residues common to known B cyclins. These facts suggests that Cdi5 defines a new cyclin class, Cyclin J. Cyclin D (Cdi3) and Cyclin J (Cdi5) are functional; both complement a yeast strain that lacks Cln1-3 G1 cyclin activity required for growth. The fact that Cyclin D and Cyclin J can function as cyclins in yeast is consistent with the fact that they can interact with the S. cerevisiae Cdk, Cdc28, in two-hybrid assays (Findley, 1996),

date revised: 28 January 2001

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.