Identification of a Cullin Homology Region in a Subunit of the Anaphase-Promoting Complex

See allHide authors and affiliations

Science  20 Feb 1998:
Vol. 279, Issue 5354, pp. 1219-1222
DOI: 10.1126/science.279.5354.1219


The anaphase-promoting complex is composed of eight protein subunits, including BimE (APC1), CDC27 (APC3), CDC16 (APC6), and CDC23 (APC8). The remaining four human APC subunits, APC2, APC4, APC5, and APC7, as well as human CDC23, were cloned. APC7 contains multiple copies of the tetratrico peptide repeat, similar to CDC16, CDC23, and CDC27. Whereas APC4 and APC5 share no similarity to proteins of known function, APC2 contains a region that is similar to a sequence in cullins, a family of proteins implicated in the ubiquitination of G1 phase cyclins and cyclin-dependent kinase inhibitors. The APC2 gene is essential in Saccharomyces cerevisiae, and apc2 mutants arrest at metaphase and are defective in the degradation of Pds1p. APC2 and cullins may be distantly related members of a ubiquitin ligase family that targets cell cycle regulators for degradation.

Two distinct ubiquitin-mediated proteolytic pathways regulate the G1 to S phase and metaphase to anaphase transitions during the cell division cycle (1). In late G1, a CDC34-dependent ubiquitination pathway degrades Sic1p, an inhibitor of G1 cyclin-dependent kinases in budding yeast, enabling the onset of S phase (2); a similar pathway exists in metazoans (3). A distinct ubiquitination machinery governs exit from mitosis through the degradation of mitotic cyclins and other regulatory proteins, such as Pds1p and Ase1p in budding yeast and Cut2 in fission yeast (4-8). In this system, a large protein complex, termed the anaphase-promoting complex (APC) or the cyclosome, functions as a protein ubiquitin ligase (9-12). In the presence of ubiquitin-activating (E1) and ubiquitin-conjugating (E2) enzymes such as UBCx (also called E2-C in clam and UBC-H10 in human) or UBC4, the APC catalyzes the formation of cyclin-ubiquitin conjugates (9,13-15), which are subsequently degraded by the 26S proteasome (16). The APC is the cell cycle– regulated component of the mitotic cyclin degradation system (9, 17, 18). The APC is also thought to participate in substrate recognition, which depends on a sequence element found in all APC substrates, called the destruction box (4, 13, 19). Biochemical studies have shown that the vertebrate APC contains eight subunits, named APC1 to APC8. These subunits include BimE (APC1), CDC27 (APC3), CDC16 (APC6), and CDC23 (APC8) (9, 11, 17).

We report the cDNA cloning of the remaining four human APC subunits on the basis of peptide sequence information obtained from immunopurified Xenopus APC (17, 20). We first identified human expressed sequence tags (ESTs) that encode amino acid sequences more than 80% identical to the XenopusAPC peptides (20). On the basis of the sequences of the ESTs, we cloned the full-length human cDNAs for APC2, APC4, APC5, and APC7 (21) (Fig. 1A). The human homolog of Saccharomyces cerevisiae CDC23 (APC8) was also isolated (21). A CDC23 peptide was isolated from the APC7 band, indicating that APC7 was CDC23 (17). Subsequent sequencing of more peptides from APC7 and APC8 bands and the analysis of the APC7 and APC8 cDNA clones by in vitro translation and immunoblotting showed that APC8 is the CDC23 homolog, whereas APC7 is a new APC subunit. We suspect that the CDC23 peptide was isolated from the APC7 band because of insufficient separation of these subunits by SDS–polyacrylamide gel electrophoresis (SDS-PAGE).

Figure 1

Sequence composition of the APC subunits. (A) Amino acid sequences of human APC2, APC4, APC5, APC7, and CDC23 (APC8). The peptide sequences derived fromXenopus APC subunits are aligned below the human sequences. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. (B) Alignment of the CH domains from APC2 and human (Hs), C. elegans (Ce), and S. cerevisiae (Sc) cullins. The alignment was made with the Clustal method with the program MEGALIGN (DNASTAR, Madison, WI). Residues identical to the consensus are boxed in black. (C) Phylogenetic tree of CH domains.

Database searches revealed that APC4 and APC5 do not share sequence similarity with proteins of known function. APC4 does not appear to have a homolog in S. cerevisiae. However, the NH2-terminal 450 residues of APC4 share limited similarity with the Schizosaccharomyces pombe open reading frame (ORF) Z97209, which in turn is distantly related to the S. cerevisiae ORF YDR118w. The YDR118w protein is a subunit of the yeast APC (APC4) (22). APC5 is similar to a putativeS. cerevisiae protein, ORF YOR249c. Deletion of this ORF from the S. cerevisiae genome indicates that YOR249c is essential for viability, and the null mutants exhibit a terminal G2-M arrest phenotype as would be expected for APC genes (23). The YOR249c protein is found in the purified S. cerevisiae APC (22). Like CDC16, CDC23, and CDC27, APC7 is a tetratrico peptide repeat–containing protein that is most similar to Cdc27p in the yeast genome.

Database searching with the APC2 sequence revealed that it is similar to a recently identified family of proteins, called cullins (24). The similarity between APC2 and cullins is restricted to a 200–amino acid region, which we refer to as the cullin homology (CH) region (Fig. 1, B and C). A cullin protein, Cdc53p, is part of a ubiquitin ligase complex that targets phosphorylated Sic1p and G1 cyclins for degradation in budding yeast (25). Mutations in the Caenorhabditis elegans cul-1 gene cause hyperplasia of all tissues, which would be consistent with a defect in G1 cyclin degradation (24). Several human cullins have also been identified in the EST database (24). Therefore, cullins represent a conserved family of proteins that may be part of the ubiquitin ligases for the degradation of Sic1p, G1 cyclins, and other regulatory proteins (24, 25). Two tandem C. elegans ORFs K06H7.6 and K06H7.5 on cosmid K06H7 are similar to the NH2- and COOH-terminal regions of human APC2, respectively, and probably represent a single gene. Furthermore, we identified a hypothetical yeast protein, ORF YLR127c, which is 18% identical (34% similar) to human APC2 (P = 1.5 × 10 20). To determine whether YLR127c encodes an APC subunit in budding yeast, we cloned the gene encoding YLR127c and inserted a triple hemagglutinin (HA) epitope tag at the NH2-terminus (26). The epitope-tagged protein efficiently rescued an apc2 deletion mutation (apc2::HIS3). HA-tagged yeast Apc2p coimmunoprecipitated with three subunits of the budding yeast APC: Cdc27p, Cdc16p, and Cdc23p (Fig. 2). Apc2p did not coimmunoprecipitate with Cdc34p or Cdc53p, however, suggesting that Apc2p is not a shared component between the APC and the Cdc4p-Cdc34p-Cdc53p complex. Cdc53p is much more closely related to ORFs YGR003w and YJL047c than to Apc2p. A database search with the sequence of Cdc53p identified YGR003w (P = 2 × 10 23) and YJL047c (P = 4 × 10 8), but the same search could not find APC2p.

Figure 2

Coimmunoprecipitation of Cdc16p, Cdc27p, and Cdc23p, but not Cdc34p or Cdc53p, with Apc2-3XHA from budding yeast extracts. Extracts were prepared from logarithmic phase cells of the yeast strain YAP89 expressing Apc2p-3XHA from its own promoter (lanes 4 to 6) or from a congenic strain expressing untagged Apc2p, YAP87 (lanes 1 to 3), and immunoprecipitated with antibody to HA sepharose beads. Crude cell extract (lanes 1 and 4), supernatants (lanes 2 and 5), and the immunoprecipitates (lanes 3 and 6) were separated on a 10% SDS-PAGE gel, transferred to nitrocellulose, and probed with the indicated antibodies.

We tested whether APC2 is an essential gene by replacing the complete coding region of YLR127c with theHIS3 gene (27). Tetrad analysis revealed thatAPC2 is essential. Dead spores arrested as large or multiply budded cells after one to three cell divisions. We then constructed temperature-sensitive alleles of APC2 by mutagenesis in the polymerase chain reaction (PCR) and integrated them into the genome at the LEU2 locus (28). Three hours after a shift to the nonpermissive temperature (37°C), two temperature-sensitive alleles, apc2-1 and apc2-4, caused a substantial increase in cells with a 2n DNA content as compared with wild-type cells (Fig. 3, A and B). Both alleles caused cell cycle arrest primarily as large budded cells with the nucleus at or near the neck, a phenotype characteristic of other known APC mutants. Tubulin staining revealed that the majority of cells arrested with short to midlength spindles, indicating a G2-M arrest (Fig. 3C). The shift to the nonpermissive temperature was lethal; viability dropped to below 7% after 24 hours at 37°C (23). Furthermore, a known APC substrate, Pds1p, was stabilized at 37°C in apc2-1 and apc2-6mutants (Fig. 3D), consistent with results reported forcdc23-1 and cdc16-123 mutants (10). The stabilization of Pds1p, along with the coimmunoprecipitation data, confirms that Apc2p (YLR127c) is a subunit of the yeast APC.

Figure 3

Phenotype of APC2 mutants. (A) Fluorescence-activated cell sorting (FACS) profile ofapc2 alleles. Cells growing in early logarithmic (log) phase at 25°C were shifted to the nonpermissive temperature (37°C). At each time point, about 106 cells were processed for FACS analysis. (B) DNA and spindle morphology of APC2 mutants. Cells were fixed and stained for DNA and tubulin. Fixed cells were examined and counted for DNA morphology (n = 200 per time point). (C) Representative G2-M cells from (B). (D) Defective Pds1p degradation of apc2 alleles at the nonpermissive temperature. Yeast strains YAP100 (APC2), YAP103 (apc2-1), and YAP104 (apc2-6) were transformed with pOCF30 (GAL-PDS1/URA3) and grown to mid logarithmic phase at 25°C in selective media with 2% raffinose. The cells were then arrested with α factor for 2.5 hours and treated with galactose (2%) for 45 min. Galactose induction was repressed by the addition of glucose (2%), and cells were then shifted to 37°C. Equal volumes of cell extract from each time point were run on a 10% SDS-PAGE gel, transferred, and probed with C210 antibody to Pds1 (29).

The identification of the stoichiometric vertebrate APC subunits has been completed. As revealed by our biochemical and genetic studies, the composition of the APC is highly conserved in organisms from yeast to humans. Furthermore, APC2 contains a region that shares sequence similarity with cullins. The CH region may perform similar biochemical functions in both systems, such as binding ubiquitin or the ubiquitin-conjugating enzymes.

  • * Present address: Research Institute for Molecular Pathology, Dr. Bohr-Gasse #7, A-1030 Vienna, Austria.

  • To whom correspondence should be addressed. E-mail: marc{at}


View Abstract

Navigate This Article