Isolation of human cyclin K by a functional selection strategy with yeast cells. (A) Selection scheme for isolation of CPR genes. Galactose induces high levels of the Gα subunit Ste4, which binds to and inhibits the Gα subunit Gpa1, thus freeing Gα for signaling and activation of cell cycle arrest and transcription. To eliminate selection of cDNAs that interfere with the upstream signal transduction cascade, CPR gene-harboring yeast cells are required to grow on galactose while simultaneously maintaining an active FUS1-HIS3 gene, thus ensuring the activity of the transcriptional branch. The level of activity of the FUS1 promoter is quantified with a FUS1-lacZ reporter. (B) Yeast strain SY2227 (15) expressing CPR4 (human cyclin K) is resistant to Ste4 overproduction by GAL-STE4 without interference of FUS-HIS3 expression. SY2227 strains harboring either pCPR4 or pUN80 (a URA3 control plasmid lacking CPR4/cyclin K) were streaked onto SC plates lacking uracil with 2% glucose as a carbon source (SC-ura) or SC plates lacking histidine and uracil (SC-ura-his) and containing 3 mM 3-aminotriazol (3AT) and 2% galactose. The 3-aminotriazole is used to select for expression of the FUS1-HIS3 fusion. (C) CPR4/cyclin K rescues the lethality of a yeast strain with cln1, cln2, and cln3 deletions. The yeast strain Y145 has a deletion of cln1, cln2, and cln3 and is kept alive by an integrated GAL-CLN3 gene. On galactose medium, CLN3 is transcribed and the cells live. On glucose, the GAL promoter is shut off and the cells fail to grow. Y145 expressing CPR4/cyclin K is able to grow on glucose, while the same strain harboring a control plasmid, pUN80, fails to grow on glucose.

The cyclin K gene encodes a new class of metazoan cyclin that is related to cyclins C and H. (A) cDNA and protein sequence of the CPR4 gene, which encodes a new human cyclin, designated cyclin K. (B) An evolutionary tree of the cyclin C, H, and K subfamilies from humans and yeast. Hs, human; Dm, D. melanogaster; Sc, S. cerevisiae; Sp, S. pombe. Protein sequence similarities between all cyclin types were compared by using the program Pileup from the GCG (Genetics Computer Group, Madison, Wis.) sequence analysis package. The vertical lines indicate divergence; the horizontal lines indicate the relative time between each occurrence. (C) Homology alignment with human cyclins C and H and yeast cyclins Pch1 and Srb11. Black boxes indicate identities, and gray boxes indicate similarities. K, human cyclin K; P, S. pombe Pch1; H, human cyclin H; C, human cyclin C; S, S. cerevisiae Srb11. (D) Cyclin box alignment of human, mouse, and Drosophila cyclin K. The GenBank accession numbers for the cyclin K genes are as follows: human, AF060515; mouse, AF060517; and Drosophila, AF060516. (E) Genomic structure of human cyclin K. Human P1 clones containing cyclin K were subcloned into pBluescript KS II+ and sequenced. (F) The human cyclin K gene is located on chromosome 14q32. In situ hybridization was performed on mitotic chromosomes by using fluorescently labeled cyclin K P1 clones. Arrows indicate the localization of cyclin K on chromosome 14 at 14q32.

Expression pattern of cyclin K mRNA. (A) Northern blots containing poly(A)⁺ RNA (2 μg per lane) from the indicated tissues were probed with full-length human cyclin K cDNA as described in Materials and Methods. PBLs, peripheral blood lymphocytes. Exposure times were 1 h. (B) Sagittal section from a 15.5-day-postcoitus C57 Black embryo that was subjected to in situ hybridization with the mouse cyclin K ³⁵S-labeled riboprobe. (C) Transverse section through an adult mouse testis. (D) Transverse section through an adult mouse ovary.

Human cyclin K is associated with a potent CTD kinase activity. (A) Western blot analysis of human cell lysates probed with antibodies (α) to cyclin K. Sixty micrograms of protein from each of the following whole-cell extracts was probed with affinity-purified α-K^pep antibodies: U-2 OS, osteogenic sarcoma; WI-38, human diploid fibroblast; HepG2, hepatocellular carcinoma; and RPE (gift of A. Davis). (B) Immunoprecipitation of endogenous cyclin K protein from the [³⁵S]methionine-labeled RPE cell line. Two microliters (2 μg) of affinity-purified α-K^pep antibodies or 5 μl of NRS was used to immunoprecipitate protein from lysates (500 μg). One microgram of cyclin K peptide (80-fold molar excess) was used to inhibit 2 μg of α-K^pep antibody. Asterisks correspond to bands associated with the cyclin K immunoprecipitation. (C) GST-CTD in vitro kinase activity associated with cyclin K. RPE whole-cell extracts (100 μg) were immunoprecipitated with either affinity-purified α-K^pep (lanes 1 to 6) or preimmune (P.I.) sera (lanes 7 and 8) that had been preincubated with (lanes 5 and 6) or without (lanes 1, 2, 3, 4, 7, and 8) peptide competitor, and the resultant complexes were used for in vitro kinase assays with (lanes 1, 2, 3, 5, and 7) or without (lanes 4, 6, and 8) the addition of a bacterially produced GST-CTD fusion protein. (D) Association of cyclin K with RNAP II in RPE. Immunoprecipitations were performed with 100 μg of RPE cell extract by using 20 μl each of protein A-Sepharose beads, protein A-Sepharose beads coupled with preimmune (P.I.) sera, and protein A-Sepharose beads covalently cross-linked with crude α-K^pep antibodies, with or without preincubation with cyclin K peptide competitor. Western blots were then performed with anti-RNAP II (ARNA3) antibodies (upper panel) and α-K^FL antibodies (lower panel) as probes.

Cyclin K is a CAK in vitro. (A) Cyclin K is associated with a CAK activity towards cyclin A-Cdk2. Immunoprecipitations were performed from RPE and WI-38 cell lysates by using 2 μg each of NRS, affinity-purified α-K^pep antibodies, affinity-purified α-K^pep antibodies preincubated with peptide competitor, and anti-Cdk7 (Santa Cruz Biotechnology, Inc.) antibodies as described in Materials and Methods. IPs were incubated with 0.1 μg of bacterially expressed and purified cyclin A, HA-Cdk2, and ATP. After activation, the complexes were assayed for H1 kinase activity with [γ-³²P]ATP as previously described (9), and reaction products were electrophoresed by SDS-PAGE (10% polyacrylamide) and visualized by autoradiography. (B) The cyclin K-associated CAK activity requires T161 of Cdk2. CAK assays were performed as described for panel A with IPs from RPE cell extracts, except that instead of HA-Cdk2, the T161A mutant of HA-Cdk2 was used.