The destruction box of human Geminin is critical for proliferation and tumor growth in human colon cancer cells

Article metrics


A domain-specific disruption was performed on the destruction box sequence of endogenous Geminin gene, an inhibitor of the DNA replication initiation complex, in a human cancer cell line HCT116 resulting in the formation of a protein that was stable in the G1 phase of the cell cycle. Although the total amount of Geminin in asynchronous cultures was not elevated, the G1-specific stabilization of Geminin, diminished chromatin loading of minichromosome maintenance complex, inhibited DNA replication, and resulted in the accumulation of cells in G1. The mutated Geminin suppressed in vivo tumorigenicity and in vitro cell growth. Cells carrying this mutation failed to support the replication of a plasmid bearing the oriP replicator of Epstein–Barr virus. The DNA damage checkpoint pathway was activated in the mutated cells with increased levels of p53 protein and its target, the p21 protein. All these deficits were rescued by overexpression of Cdt1, a replication initiator protein that binds to Geminin. Therefore, alteration of the cell cycle-dependent regulation of endogenous Geminin in human cells without increasing total protein level inhibits DNA replication and suppresses tumor growth.


In late mitosis and early G1 phase of the cell cycle, replication origins are ‘licensed’ for replication by the loading of a complex of six minichromosome maintenance (MCM) proteins, MCM2–7 (Blow and Laskey, 1988; Diffley, 2001; Lei and Tye, 2001; Blow and Hodgson, 2002). The origin recognition complex (ORC) first binds to each replication origin, and then recruits two other proteins, Cdc6 and Cdt1 during G1 (Bell and Dutta, 2002). Subsequently, MCM complex is bound at replication origins to license the origins for replication fork initiation at the onset of S phase (Diffley et al., 1994; Labib et al., 2001). Two regulatory proteins, cyclin-dependent kinase (cdk) and Geminin play a critical role to prevent re-licensing by inhibiting loading of new MCM complex onto the origins during S, G2, and early M phase (Diffley, 2001; Lei and Tye, 2001; Blow and Hodgson, 2002).

Geminin, an inhibitor of DNA replication, was originally identified as a substrate for the anaphase-promoting complex (APC) that degraded in mitotic Xenopus egg extracts. Geminin contains a destruction box sequence near its amino terminus that is necessary for this degradation during the cell cycle (McGarry and Kirschner, 1998). Geminin is degraded as the cells exit from mitosis in synchronized Hela cells, then Geminin is absent during the G1 phase and accumulates during the S, G2, and M phases (McGarry and Kirschner, 1998). Cdt1, an essential replication factor, must be incorporated into prereplication complex to recruit MCM complex (Maiorano et al., 2000; Nishitani et al., 2000). Geminin inhibits DNA replication by preventing incorporation of the MCM complex into the prereplication complex (Wohlschlegel et al., 2000; Tada et al., 2001).

The biological function of Geminin has been characterized in higher organisms by suppressing gene transcription, depleting endogenous protein, or overexpressing exogenous protein. In Xenopus embryos, Geminin induces uncommitted embryonic cells to differentiate as neurons (Kroll et al., 1998). Geminin mRNA elimination by using antisense techniques arrested the embryos in the G2 phase immediately after the midblastula transition indicating that Geminin has an essential function. Loss of this function prevents entry of cells into mitosis by activation of a mechanism that depends on Chk1, the effector kinase of the DNA damage checkpoint pathway (McGarry, 2002). The Drosophila homolog of Geminin was shown to have properties similar to Xenopus Geminin. Overexpression of Drosophila Geminin in embryos inhibits DNA replication, induces ectopic neural differentiation, and undergoes apoptosis (Quinn et al., 2001). The silencing of Geminin expression in Drosophila Schneider D2 cells indicates that Geminin is required for the suppression of over-replication and for genome stability in Drosophila cells (Mihaylov et al., 2002).

Overexpressing a nondegradable form of Geminin in which the destruction box was mutated in human cancer cells U2OS activated an intra-S checkpoint, arrested cell proliferation, and triggered apoptosis (Shreeram et al., 2002; Wohlschlegel et al., 2002). These results suggest that overexpression of a stabilized form of Geminin inhibits cancer cell proliferation. It is, however, unclear whether overexpression of the mutant protein produces nonspecific effects.

In this paper, we show that cells with a targeted disruption of the destruction box of human Geminin survive with only a slight prolongation of the G1 phase of the cell cycle. They do not permit EBNA1-dependent replication of episomes from oriP, but this replication is restored upon stable overexpression of Cdt1 in these cells. In vivo tumorigenicity as well as in vitro cell proliferation in the mutated cells was suppressed. The diminished level of DNA replication in mutated cells is accompanied by the activation of a DNA damage checkpoint pathway. The increased level of p21 protein, presumably resulting from the stabilization of p53, is rescued by overexpression of Cdt1. Taken together, these data suggest that even at physiological levels of Geminin, the mutation of the destruction box is sufficient to inhibit DNA replication and tumor growth without inducing apoptosis. Disrupting the normal G1 destruction of endogenous Geminin also appears to activate the DNA damage checkpoint pathway in human cells.


Mutation of Geminin

The biological function of Geminin in mammalian cells has been assessed only through overexpression experiments, so that it has not been possible to eliminate pleiotropic effects due to the overexpressed exogenous protein. To circumvent this problem, a homologous recombination was used to replace the second exon of Geminin (encoding the initiator ATG) in HCT116 colon carcinoma cells with a neomycin phosphotransferase (Neo) gene (Figure 1a). PCR screening of genomic DNA with the indicated primers was performed. For the left homologous arm, G17 and Cre1 primer set, which both are included in the targeting vector, produces a 1.0 kbp product both in Geminin heterozygous (+/−) and neo-resistant wild-type clones in which the targeting vector was nonhomologously integrated in the chromosome (see Figure 1b, right lane of 5′PCR panel). G18 and Cre1 or G1 and Cre1 primer sets produce 1.1 or 1.68 kbp products from the − allele, respectively. The 1.68 kbp product is cut by HindIII into fragments of 865, 468, and 350 bp and EcoRV into fragments of 1420 and 263 bp (Figure 1b, 5′PCR). For the right homologous region, Cre2 and G7 primer set, which both are included in the targeting vector, produces an about 4.2 kbp product both in Geminin heterozygous (+/−) and Neo-resistant wild-type clones in which the targeting vector was randomly integrated in the chromosome (see Figure 1b, Neo-R lane of 3′PCR panel), whereas Cre2 and G23 primer set produces a 4556 bp product from the − allele that is cut by EcoRI into fragments of 3098 and 1458 bp (Figure 1b, 3′PCR). Consequently, PCR screening with the G1 and Cre1 plus Cre2 and G23 primer sets identified 10% of G418-resistant colonies as Geminin +/− clones. The integration of a single copy of the targeting vector into Geminin locus was confirmed by Southern blot after digestion with EcoRI. A single 6.0 kbp band was identified by a probe designed in Neo gene (Figure 1c).

Figure 1

Targeted disruption of the human Geminin gene. (a) Wild-type Geminin locus with the targeting vector. Solid boxes represent the exons. Primers and enzyme sites used for PCR screening are indicated. Recombination between the targeting vector and endogenous locus deletes 51 bp of exon 2 containing the ATG (gray box). (b) Genomic PCR of the cell clones to identify products of the homologous recombination. Genomic DNA prepared from parental +/+ clone and from +/− clone was subjected to PCR with primer pairs shown in (a). Gel electrophoresis and ethidium bromide stain to visualize the products of the PCR reaction and the fragments produced by HindIII or EcoRV digestion (left panel for 5′PCR) and EcoRI digestion (right panel for 3′PCR) of the PCR products. (c) Southern blot of EcoRI-digested genomic DNA from targeted and wild-type HCT116 cells hybridized with the Neo probe. The 4.8 kbp EcoRI fragment recognized by the Neo probe indicates the single integration event in the mutated (–) allele of +/− clone. (d) Cre-mediated recombination of the +/− cells obtained in (b) produces +/LoxP cells where the excision of the Neo cassette leaves a LoxP site in exon 2 (LoxP). The second round of targeting can either create a +/− or a −/LoxP cell. Otherwise, targeting vector can randomly integrate into irrelevant genomic locus (+/LoxP random). PCR with indicated primers produce the diagnostic PCR products shown to the right of each allele. The gray box in exon 2 indicates the region to be deleted. (e) Genomic PCR of the cell clones to identify products of the Cre-mediated recombination (left) and the second round of homologous recombination (right). Genomic DNA prepared from indicated clones was subjected to PCR with primer pairs shown in (d). Gel electrophoresis and ethidium bromide stain to visualize the products of the PCR reaction. Neomycin sensitive (Neo-S) and resistance (Neo-R) were noted at the bottom of the figures

Cre-mediated recombination of the LoxP sites, flanking the Neo gene excised the Neo cassette, made the cells susceptible to G418, and left a LoxP site in place of the latter half of exon 2 in the allele called LoxP. PCR screening of genomic DNA with the indicated primers (G25 and G24) produces a 327 bp product from the + allele (Figure 1d and e). Cre-mediated recombination was distinguished by PCR screening because it produces a 231 bp product from the LoxP allele instead of a 1.4 kbp product from the − allele (Figure 1e, left). The second allele of Geminin was targeted with the same vector followed by G418 selection. Retargeting of the LoxP marked allele recreates the − allele in cells (Figure 1d). Targeting of the remaining wild-type Geminin allele, on the other hand, creates the desired cell line (−/LoxP) (Figure 1d). PCR screening produces a 231 bp product from the LoxP allele and a 1.4 kbp product from the – allele (Figure 1e, right). PCR analysis identified about 7% of the G418-resistant clones as –/LoxP mutants at the Geminin locus.

LoxP-Geminin allele produces a low level of mRNA but a physiological level of N terminally truncated stable Geminin protein

In a conventional gene deletion experiment, transcription and polyadenylation of the drug resistance cassettes inserted in the two alleles of the target gene prevents residual expression of the latter. In the strategy employed here, however, only a LoxP site was left in the latter half of exon 2 in the LoxP allele. Although the initiator ATG was removed, low levels of a variant mRNA produced by read-through transcription or alternative splicing could express an N terminally deleted protein (Dhar et al., 2001). The RT–PCR analysis revealed that the expression of Geminin mRNA in the mutant cells (−/LoxP) significantly decreased when compared to that of wild-type (+/+) cells (Figure 2a). Owing to the absence of a splice donor site in the LoxP site, a cryptic splice site in intron 2 could have been used to produce an alternatively spliced Geminin mRNA in the mutant cells (−/LoxP) that is less stable than wild-type Geminin mRNA.

Figure 2

Characterization of Geminin in the −/LoxP cells. (a) RT–PCR analysis of total RNA prepared from wild-type (+/+) and −/LoxP cells. Geminin mRNA expression level was detected using primers (arrows) indicated in the exon (boxes). Human β-actin primer set was used as a control. (b) Western blot analysis of total cell lysates using anti-Geminin antibody. Total cell lysates of +/+ or −/LoxP cells loaded per lane prepared from asynchronous (AS) or G1–S phase blocked cell culture. Schematic representation of wild-type and truncated protein is given on the right side. The conserved destruction box (D box) is indicated as gray color. The Orc2 protein was immunoblotted as a control. (c) The truncated Geminin is stable during mitosis. Total cell lysates of +/+ or −/LoxP cells prepared from nocodazole (Noc) block or 3 h release from block (M–G1 phase) cell culture were loaded per lane. The cyclin B protein was immunoblotted as a control. (d) Detection of DNA replication licensing proteins in the chromatin fraction of +/+ and −/LoxP cells. The extracted protein from the chromatin fraction immunoblotted with antibodies to MCM3, MCM6, MCM7, and ORC2 proteins. Histograms of a representative blot are calculated (Scion Image Beta 4.02) and shown below. Left (black) and right bar (gray) for each blot represents +/+ and −/LoxP, respectively

Although an anti-Geminin antibody did not recognize the full-length 33 kDa Geminin protein in the −/LoxP cells, a mutant Geminin of about 28 kDa (GemininDEL) was detected at almost equal protein level as wild-type cells when asynchronously cultured cell lysates were analysed (Figure 2b). As the 28 kDa protein, GemininDEL, was seen only in cells with the LoxP allele, the polypeptide was most likely the product of the LoxP-marked Geminin locus in which the internal methionine-28 encoded by exon 3 is used for translation initiation. The next internal methionine is positioned at 89 and would produce a much shorter protein. The truncated N terminal 27-amino-acid of Geminin includes most of the destruction box RRTLKMIQP (23–31 amino-acid residues).

Geminin is ubiquitinated by APC, which specifically targets B-type cyclins and other proteins containing destruction box motif(s) for degradation by ubiquitin-mediated proteolysis (Zachariae and Nasmyth, 1999). Since APC becomes active during mitosis and remains active throughout the G1 phase of the cell cycle, we examined whether GemininDEL was degraded in mitosis and G1. The wild-type cells blocked at the G1–S phase of cell cycle by mimosine has less Geminin protein when compared with the GemininDEL protein prepared from −/LoxP cells (Figure 2b). The Orc2 protein level was not changed between +/+ and −/LoxP cells when the G1–S blocked cell lysates were used for immunoblotting (Figure 2b). To further define the stability of GemininDEL protein during cell cycle, a culture of HCT116 cells was synchronized with nocodazole at the M phase, and the cells were collected at 3 h after release from the block. Note that the protein level of wild-type Geminin and GemininDEL at prometaphase is unchanged in nocodazole-treated cells (Figure 2c). At 3 h after release from nocodazole block, the cells progressed through mitosis and started entering G1 as assessed by cyclin B degradation (Figure 2c, bottom). In the same cells, the GemininDEL was much more stable compared to the wild-type Geminin consistent with the destruction box being critical for degradation of Geminin (Figure 2c).

Targeting of destruction box of Geminin inhibits chromatin loading of MCM complex and prolongs G1 phase of the cell cycle

Geminin was shown to inhibit Cdt1, which plays a role in loading the MCM complex onto chromatin (Maiorano et al., 2000; Wohlschlegel et al., 2000; Tada et al., 2001). To investigate the effect of targeting the destruction box of Geminin on the chromatin loading of DNA licensing factors, immunoblotting of proteins from the chromatin fraction of cells was performed. The amount of chromatin-associated MCM3, MCM6, and MCM7 proteins was decreased to less than 50% in the −/LoxP cells, whereas chromatin-associated Orc2 protein level was not changed (Figure 2d).

The proliferation rate of the Geminin −/LoxP cells was decreased by 30% (Figure 3a). DNA replication activity based on BrdU incorporation of −/LoxP cells was approximately 50% that of wild-type cells and +/− cells (Figure 3b). FACS used to measure the DNA content of asynchronously growing cells indicated almost no change in the G1 phase population in the wild-type and −/LoxP cells (data not shown). Contrary to the apoptosis observed upon overexpression of stabilized Geminin (Shreeram et al., 2002), there was no sub-G1 population in the −/LoxP cells. More direct assays for apoptosis-like terminal deoxynucleotidyltransferase-mediated dUTP-end labeling (TUNEL) assay also did not show an increase in apoptosis (data not shown). The percentage of cells remaining in the G1 phase following release for 24 h from mimosine induced G1–S block increased in −/LoxP cells by about 15% compared to wild-type and +/− cells, suggesting a slight delay in entry into the S phase (Figure 3c). The percentage of cells that remained in the G1 phase of cell cycle did not vary by more than 5% between the repeated experiments.

Figure 3

Geminin −/LoxP cells are compromised in cell proliferation. (a) Cell proliferation of the indicated cell lines. The data are shown as mean and standard deviation of four measurements. (b) BrdU incorporation of +/+, +/−, and −/LoxP cells. Comparison of BrdU incorporation of asynchronous cells growing 2 h in culture, with the incorporation in +/+ cells held at 100%. The data are shown as mean and standard deviation of four measurements. (c) FACS analysis for DNA content of +/+, +/−, and −/LoxP cells. Cells synchronized at the G1–S boundary by mimosine (top) or 24 h following release from the mimosine block (bottom) were subjected to FACS analysis. G1 population accumulated in −/LoxP cells are marked

The destruction box of Geminin is required for efficient tumorigenicity

The importance of APC-mediated destruction of cell cycle regulators has not been tested in animals. We, therefore, examined whether the absence of the destruction box of endogenous Geminin had any effect on tumorigenesis in vivo. Wild-type and −/LoxP HCT116 cells were injected subcutaneously and grown as xenografts in nude mice for up to 4 weeks. There was a dramatic reduction in tumor establishment and growth of the −/LoxP cells compared with wild-type HCT116 cells (Figure 4a and b). Histological analysis of the tumors has revealed the presence of two distinct zones in both tumors; namely, central necrotic zone (central pinkish area) that was surrounded by the circular marginal zone (purple) composed of active tumor cells (Figure 4 a–c and e, f). High-power view analysis has revealed increased size of nuclei in −/LoxP mutated cells, hence increased nuclear–cytoplasmic ratio (Figure 4c, c vs d), and prominent mitotic figures are seen in wild-type but not in mutated cells (arrows). To test whether the suppression of tumor growth in mutated cells was a result of increased apoptosis, we performed TUNEL staining on tumor sections. Numerous TUNEL-positive cells were equally observed in the central zone of both the tumors (Figure 4c, g and h). We could not, however, observe any significant difference in the TUNEL positivity in cells located in the marginal zone of tumor, suggesting that increased apoptosis is not responsible for the tumor growth suppression in the mutated cells. Immunohistochemical staining for Ki-67, which is preferentially expressed during the late G1-, S-, M-, and G2 phase of the cell cycle, showed significantly reduced numbers of proliferating cells in −/LoxP tumors as compared to the wild-type tumor (Figure 4c, i vs j). The tumors from −/LoxP cells showed a more than twofold decrease in Ki-67-labeling index (24.8%±8.9 for −/LoxP vs 56.0%±9.3 for wild type, P<0.002), suggesting that the targeted disruption of destruction box of Geminin leads to a slowing of tumor cell proliferation. We also noticed that Ki-67-labeled cells had tendency to locate predominantly on the outer side of the marginal zone and did not exist in the central zone of the tumors. This implicates that local environmental factor such as oxygen supply played a role in regulating cell growth/death. Thus, massive cell death observed in the central zone could be presumably caused by a consequence of intratumoral hypoxia. Taken together, these results indicate that the major cause of tumor mass reduction upon stabilization of endogenous Geminin is not due to increased apoptosis but conceivably due to a defect in DNA replication activity.

Figure 4

The destruction box of Geminin is required for efficient tumorigenicity. HCT116 wild-type (+/+) and −/LoxP cells were injected as xenografts in nude mice and killed after 25 days. (a) Clinical appearance of tumor formation after subcutaneous injection of cells. Representative mice (tumors; arrowheads) at day 20 are shown. (b) Growth curves of wild-type and −/LoxP xenografts. Each data point for wild-type and −/LoxP cells represent the average tumor size of four mice. Error bars represent the standard deviation (n=4 mice per group). Tumors that did not grow were assigned a volume of zero. (c) Histological examination of tumors. H&E staining of transverse sections of the tumors of wild-type (+/+) (a, c, e) and Geminin destruction box-deficient (–/LoxP) (b, d, f) mutated HCT116 cells are shown at a magnification of × 4 (a, b), × 600 (c, d), and × 200 (e, f), respectively. In the low power view, two distinct areas, namely, central zone (cz; central pinkish area) and circular marginal zone (mz, purple) are evident in both the tumors (a, b). In the high power view, mitotic figures (arrows) are prominent in wild-type cells (c vs d). e, g, i and f, h, j, each corresponds to the same view filed at the same magnification on serially sectioned specimen and stained with TUNEL (g, h), and Ki-67 (i, j), respectively (outer and inner border of the marginal zone are indicated by solid and dashed lines, respectively). TUNEL analysis has revealed no significant differences between wild-type and mutated cells (g vs h: note that very few TUNEL +ve cells (purple) in the marginal zone and numerous apoptotic cells are present in the central zone). Ki-67-labeled cells (brown) tend to locate at the outer side of the marginal zone and do not exist in the central zone in both the tumors (i, j). Increased numbers of Ki-67-labeled cells are seen in wild-type cells (i) as compared to mutated cells (j)

Activation of checkpoint and cell cycle proteins

Recent work has shown that overexpression of a nondegradable form of Geminin resulted in the activation of an intra-S checkpoint (Shreeram et al., 2002). HCT116 cells are known to maintain intact p53 and retinoblastoma (Rb) protein. To determine whether endogenous nondegradable form of Geminin is sufficient to activate a DNA damage checkpoint signal in HCT116 cells, cell lysates were immunoblotted for detecting proteins involved in DNA damage checkpoint and cell cycle pathways. The −/LoxP cells contained high levels of p53 and a target of p53, the Cdk inhibitor p21Waf1/Cip1 (Figure 5). In support of inhibition of Cdk2, phosphorylation of serine 795 of Rb was low and cyclin A level slightly decreased in −/LoxP cells (Figure 5). The level of p53 phosphorylation at serine 6, 9, 15, 37, 46, and 392 and the level of p16INK4a remained at background levels or were unchanged relative to wild-type cells (data not shown). In contrast, phosphorylation of p53 on serine 20 was modestly elevated in −/LoxP cells, indicating stabilization of the protein possibly due to phosphorylation by Chk1 or Chk2 protein kinases (Chehab et al., 2000; Shieh et al., 2000; Abraham, 2001). HU treatment (1.5 mM) of the same cells showed that Chk1-mediated phosphorylation of serine 20 of p53 was not hyperactive under all circumstances in the −/LoxP cells (Figure 5). The Chk1 and Chk2 protein levels were similar between wild-type and −/LoxP cells (Figure 5). The status of Chk1 phosphorylation at serine 345 and that of Chk2 at threonine 68 could not be detected (data not shown). No difference was observed in p34Cdc2 and Akt/PKB protein levels between wild-type and −/LoxP cells (Figure 5 and data not shown).

Figure 5

Biochemical characterization of checkpoint and cell cycle proteins in the −/LoxP cells. Western blot analysis of total cell lysates using anti-Chk1, anti-Chk2, anti-p53, anti-phospho p53 (Ser 20), anti-p21, anti-phospho Rb (Ser 795), anti-Cyclin A, and anti-cdc2/p34 antibodies. Total cell lysates of wild-type (+/+) or −/LoxP cells were loaded per lane. For p53 (Ser 20), hydroxyurea (HU)-treated cell lysates were also used

Stabilized Geminin impairs replication from the oriP of EBV

We have previously reported that replication of the Epstein–Barr virus (EBV) oriP episome in HCT116 cells was inhibited by transient overexpression of wild-type Geminin and mutant Geminin lacking a destruction box (Dhar et al., 2001). Therefore, it was of interest to investigate whether the slight stabilization of endogenous Geminin is sufficient to interrupt oriP-based replication. A plasmid carrying oriP, EBNA-1, and a hygromycin resistance gene (p220) was transfected into wild-type and −/LoxP cells and the transfected cells were selected in hygromycin for up to 12 days (Figure 6b). Replication of p220 allowed robust hygromycin-resistant colonies to emerge in the wild-type cells, but not in the −/LoxP cells, indicating that the stabilization of endogenous Geminin is sufficient to inhibit replication of oriP of EBV. p396, a similar plasmid to p220 except for a deletion in the EBNA-1 binding sequence, did not support any colony formation both in the wild-type and −/LoxP cells, indicating that colony formation required EBNA-1-dependent replication of the plasmid (Figure 6c). pBabe-puro, a plasmid that does not have oriP or EBNA-1 and is integrated in the host cell chromosome, was used as a control and produced almost the same size and number of puromycin-resistant colonies in wild-type and −/LoxP cells (Figure 6a). pSG5, a mammalian expression vector without any drug-resistant gene, produced no colonies under hygromycin selection (Figure 6d). Therefore, wild-type and −/LoxP cells are not inherently different in their transfection efficiency or in their expression of drug resistance markers. The observation that the size and number of the puromycin-resistant colonies are similar in wild-type and −/LoxP cells also indicates that the inability of the plasmid p220 to replicate cannot be due to the difference in proliferation rates between the two cells. Collectively, these results imply that the cell cycle-regulated degradation of Geminin is important to permit for EBV oriP-dependent replication of plasmids in human cells.

Figure 6

Geminin −/LoxP cells are defective in DNA replication from oriP of EBV. Establishment of drug-resistant colonies of +/+ and −/LoxP cells after transfection of pBabe-puro (a), a plasmid containing a puromycin resistance gene as a control, plasmid p220 (b), carrying EBV oriP, EBNA-1, and the hygromycin-resistance marker, p396 (c), similar to p220 but lacks EBNA-1 binding site, or pSG5 (d), an expression plasmid containing no drug resistance gene as a control. Following selection with puromycin (a) or hygromycin (b–d) up to 2 weeks, colonies were visualized by staining with crystal violet

Rescue of −/LoxP cells phenotype by overexpressing Cdt1 protein

As −/LoxP cells could have a deficiency in Cdt1 function due to the stabilized GemininDEL, we investigated whether overexpression of Cdt1 can rescue the phenotypes seen in −/LoxP cells. Cdt1 expressing vector was stably transfected into −/LoxP cells. The expression level of Cdt1 mRNA was semiquantitatively detected by RT–PCR. As shown in Figure 7a, transfection of exogenous Cdt1, but not vector alone, resulted in an increase in Cdt1 mRNA level. Importantly, this resulted in attenuation of the p21 protein level (Figure 7b). Overexpression of large-T antigen of simian virus 40 (SV40-T), but not expression vector alone, also decreased the p21 level in −/LoxP cells (Figure 7b) because SV40-T associates with and inactivates the cellular p53 protein (DeCaprio et al., 1988). To evaluate the effect of p21 degradation on the cell phenotype, we examined BrdU incorporation and cell growth. Overexpression of Cdt1, but not SV40-T, caused a partial recovery of DNA replication activity in −/LoxP cells (Figure 7c). Overexpression of Cdt1 and SV40-T partly rescued the cell proliferation defect of −/LoxP cells after 3 days culture (Figure 7d). The G1 delay was also partially rescued by Cdt1 overexpression but not by SV40-T overexpression (data not shown).

Figure 7

Overexpression of Cdt1 rescued the phenotype of −/LoxP cells. (a) RT–PCR analysis of Cdt1 mRNA level in −/LoxP cells stably transfected with pcDNA3 or pSPORT-Cdt1. Human Cdt1 primers were set in the 3′ untranslated region. Human β-actin primer set was used as a control. (b) Immunoblotting of p21 in the lysates prepared from −/LoxP cells stably transfected with pSPORT-Cdt1, pcDNA3-SV40-T, or pcDNA3 alone. (c) BrdU incorporation of the cells described in (b). Asterisk represents statistical significance (P<0.01) in a Student's t-test in comparison between pcDNA3 and Cdt1. Mean and standard deviation of four measurements are shown. (d) In vitro cell proliferation assay. Cell numbers are counted after 3 days of culture. Asterisk represents statistical significance (P<0.05). Mean and standard deviation of four measurements are shown. (e) Partial rescue of oriP-based plasmid replication in −/LoxP cells by stably expressing Cdt1. Establishment of hygromycin-resistant colonies of −/LoxP cells after transfection of plasmid p220. (a) pcDNA3, (b) pSPORT-Cdt1, (c) pcDNA3-SV40-T, (d) pLXSN, (e) pLXSN-HPV E6, and (f) pLXSN-HPV E7 were used. Histograms of a representative colony blot is calculated (Scion Image Beta 4.02) and shown in the right

If the defect in oriP-based plasmid replication in −/LoxP cells is due to the stabilization of Geminin, overexpression of Cdt1 protein should rescue the replication suppression. Cdt1 or viral oncogenes were stably introduced into the −/LoxP cells by cotransfection with pBabe-puro and selection for stable clones under puromycin. p220 oriP-based plasmid was then stably transfected into the puromycin-resistant cells and selected with hygromycin. The increase in hygromycin-resistant colonies in −/LoxP cells stably expressing Cdt1 indicates that the expression of Cdt1 in −/LoxP cells restored replication from oriP of EBV (Figure 7e and b). The SV40-T and HPV E6, which are capable of binding and inactivating the cellular p53 protein (Werness et al., 1990), had little effect on the colony formation in −/LoxP cells, suggesting that the activated p53 pathway is not responsible for inhibiting the replication and maintenance of p220 in −/LoxP cells. In addition, the E7 protein of HPV, which physically interacts with and inactivates the Rb protein and then abrogates Rb-mediated growth suppression (Dyson et al., 1989), had almost no effect on p220 replication (Figure 7e).


Overexpression experiments potentially have pleiotropic effects partly due to the high expression level far from the physiological range that could lead to overestimation of gene function. Here, we employ a more subtle approach: the targeted deletion of the destruction box of human Geminin gene in a cancer cell line by homologous recombination. The genetically altered cell line did not produce a grossly increased amount of Geminin protein compared to the parental HCT116 cells. This enables us to examine the importance of the destruction box of Geminin in human cells at near physiological levels of protein expression. In agreement with what has been reported previously, the results clearly indicate that destruction of Geminin is important for moving cells from G1 into S. A previous study has shown that overexpression of a nondegradable mutant of Geminin elicited apoptosis in U2OS cells (Shreeram et al., 2002). Since we did not observe any apoptosis in the −/LoxP cells, we conclude that an endogenous nondegradable mutant of Geminin is sufficient to inhibit Cdt1 function, but not sufficient for the induction of apoptosis in the HCT116 colon cancer cell lines. The stabilization of endogenous Geminin in G1 leads to a reduction in the quantity of chromatin-bound MCMs, but not of ORC2, and a decrease in the DNA replication activity. It is noteworthy that the inhibition of DNA replication was reversed by overexpression of Cdt1 but not with SV40 large T, suggesting that the inhibition was mediated by a reduction in origin licensing. Targeted deletion of the destruction box of Geminin in human cells affected the degradation of Geminin in mitosis and G1. This is the first evidence that the importance of a destruction box has been tested in human cells without overexpressing the mutated protein.

It is interesting that the slight stabilization of endogenous Geminin was sufficient to activate a DNA damage checkpoint pathway. The specific induction of p21 was evidenced by the fact that overexpression of Cdt1 reverses the p21 protein expression. We initially hypothesized that the increase in cell cycle inhibitors was responsible for the phenotypes observed in Geminin-mutated cells. To test this hypothesis, we made stable cell lines in which viral oncogenes SV40-T, HPV E6, or E7 are stably expressed in Geminin-mutated cells to inactivate the p53 and Rb pathways. The inhibition of DNA replication, prolongation of G1 phase, and support of oriP-based replication were not rescued by the simple inactivation of p53, Rb, or p21. In contrast, overexpression of Cdt1 rescued these phenotypes, suggesting that the alteration of the Geminin–Cdt1 balance in favor of Geminin could be implicated in the phenotypes seen in the Geminin-mutated cells.

Although the functions of destruction boxes have been tested in cells in culture, we were curious whether the phenotypes observed due to the stabilization of Geminin will persist in animals. In vivo tumorigenicity assay provides a means of assessing the functional importance of the destruction box of Geminin in animals. Our results clearly showed that the tumor volume and the numbers of Ki-67-positive cells decreased in Geminin-mutated cells, suggesting that the destruction box of Geminin is required for normal cell proliferation even in the context of tumors in animals.

It remains to be determined how Geminin regulates the levels of p53 and p21. It was previously shown that p53 is phosphorylated at serine 15 after adenovirus-mediated overexpression of nondegradable Geminin (Shreeram et al., 2002). We now show that stabilization of Geminin in G1 without gross overexpression also stabilizes p53. Whenever mammalian cells are prevented from completing synthesis from early replicons, ATM (ataxia telangiectasia mutated) independent, checkpoint signal stabilizes components of existing replicons and prevents initiation of replication from late-firing origins (Dimitrova and Gilbert, 2000). Chk1 responds to stalled replication forks as a necessary component for an intra-S phase checkpoint (Feijoo et al., 2001). Thus, one possibility is that DNA damage from incomplete replication initiation is responsible for activating p53 in the −/loxP cells. A slight increase in serine 20 phosphorylation of p53 is seen, but in the absence of any evidence of Chk1 or Chk2 protein activation, we cannot yet conclude that DNA damage is responsible for the activation of the DNA damage checkpoint pathways and constitutive stabilization of p53.

Finally, we provide additional evidence that cellular Cdt1 is necessary for replication from oriP of EBV using these genetically engineered cells. Geminin interacts tightly with Cdt1, a replication initiation factor necessary for MCM loading (Wohlschlegel et al., 2000; Tada et al., 2001). Inhibition of DNA replication by Geminin in cell-free DNA replication extracts could be reversed by the addition of excess Cdt1 (Wohlschlegel et al., 2000). We previously found that ORC2 is associated with EBNA1 and the extreme sensitivity of the oriP-dependent plasmids replication in the ORC2 hypomorphic cells (Dhar et al., 2001). The association of MCM2 and ORC2 with the replicator of oriP of EBV has also been reported (Chaudhuri et al., 2001). It is conceivable that Cdt1 forms a complex with the replicator of oriP via the interactions with other proteins. Such a possibility is consistent with the fact that Cdt1 associates not only with Geminin but also with ORC2 and MCM6 (Yanagi et al., 2002). Consistent with this evidence, our results clearly indicate a defect in oriP-dependent replication upon the stabilization of Geminin in the −/LoxP cells. In addition, the defect was selectively rescued by overexpression of Cdt1.

The possibility that Geminin could involve in tumorigenicity was tested in nude mice where Geminin mutant cells were injected as xenografts. Geminin mutant cells led to a marked reduction of tumor formation. Given the in vitro and in vivo growth suppression of the Geminin-mutated cells, it is likely that the stabilizing Geminin, even at physiological protein levels during the G1 phase of the cell cycle, might efficiently regulate both the aberrant cell proliferation and latent infection of EBV. It is interesting to note that in a study with human lymphomas, however, increased levels of Geminin are correlated with increased tumor cell proliferation (Wohlschlegel et al., 2002). The difference between the natural tumors and the xenografts reported here suggests that in the natural tumors, the increased Geminin was either countered by increased Cdt1, or the Geminin was seen only in the S and G2 phases of the cell cycle when it has little effect on cell proliferation.

Materials and methods

Tissue culture and transfections

The human colorectal cancer cell line HCT116 (American Type Culture Collection) and its derivatives were grown in 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in McCoy's 5A modified media (Invitrogen) and maintained at 37°C in 5% CO2. Cells were transfected with Lipofectamine Plus (Invitrogen), following the manufacturer's protocol and selected by using hygromycin or geneticin (Invitrogen) at concentrations of 0.1 mg/ml and 0.5 mg/ml, respectively.

Geminin targeting construct and screening for recombinants

A promoterless strategy was adapted for targeting the Geminin gene (Waldman et al., 1995; Dhar et al., 2001). A PAC clone (clone ID: RP3-369A17; GenBank accession number AL133264) containing human Geminin gene was purchased (Invitrogen) and used as the source for homologous arms. The two PCR-amplified fragments, one 1.0 kbp and the second 4.2 kbp, were used to construct the 5′ and 3′ arms of the targeting vector, respectively. The 1.0 kbp subclone contained the region immediately 5′ of the initiation codon located in exon 2 of the Geminin coding region. The 4.2 kbp subclone contained a region beginning 131 bp distal to the initiation codon. Two fragments were assembled in pKO plasmid (Stratagene) surrounding a promoterless geneticin (G418)-resistant gene containing simian virus 40 polyadenylation signals (Neo cassette) (Figure 1a). LoxP sites surrounding the Neo cassette were incorporated into the vector. All sequence was confirmed by sequencing. Transfection was performed with NotI linearized targeting vector.

For first allele targeting, G418-resistant clones were screened by PCR (Figure 1a). Genomic DNA was prepared from G418-resistant clones and homologous recombinants identified by PCR were further confirmed by Southern blot using Neo probe. A clone carrying a homologous recombinant and no additional random integrants was then infected with recombinant adenovirus expressing Cre recombinase (purchased from Gene Transfer Vector Core, University of Iowa, IA, USA) to yield G418-sensitive clones. One of these heterozygous clones without Neo (+/LoxP) was then used for second round homologous recombination using the same targeting vector and several clones with both alleles of the Geminin disrupted (–/LoxP) were obtained and used for further experiments. Details of the constructs and PCR conditions, primer sequences are available upon request.


We isolated total RNA using TRIZOL reagent (Invitrogen) according to the manufacture's instructions. The reverse transcription (RT) step was done according to the manufacturer's directions (Invitrogen). Briefly, 1 μg of extracted RNA and oligo(dT) primer were diluted in 13 μl of RNase-free water, heated to 65°C for 5 min, and then chilled on ice. For first-strand cDNA synthesis, heat-denatured RNA solution, along with deoxynucleoside triphosphates, 10 mm dithiothreitol (DTT), and SuperScript II were added to make up 20 μl of reaction mixture, followed by incubation at 42°C for 50 min, and then heated to 70°C for 15 min and chilled on ice.

In order to amplify the Geminin and Cdt1 cDNA, 513 and 240 bp, respectively, the following primers were designed; Geminin: sense, 5′-IndexTermCATCTACAACTTCCAGCCCTGGG-GTTA-3′; antisense, 5′-IndexTermGAACTAGTGGAGGTAAACTTC- GGCAGT-3′, Cdt1: sense, 5′-IndexTermTGGGCACCTGTCGTCCCAGCTACTAGGGAG-3′; antisense, 5′-IndexTermTTCAAAGCTGG-CTGGCTCTGGCCCTGTCAT-3′. PCR was performed as follows: 2 min at 94°C; followed by 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s; followed by an 5 min extension at 72°C. As a control, β-actin primer set was used (Invitrogen).

Western blot analysis and chromatin fraction preparation

Cells were harvested and lysed in RIPA lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM PMSF, 1 mM Na3VO4, 1 mM NaF, and 1 μg/ml each of aprotinin, pepstatin, and leupeptin) for 20 min on ice. The cell lysates were centrifuged and protein concentration was determined by the Bio-Rad protein assay kit (Bio-Rad laboratories). Before being subjected to SDS–PAGE, the reaction was stopped by adding Laemmli sample buffer containing 100 mM DTT. Equal amounts of cellular protein (30–100 μg) were electrophoresed in NuPAGE 4–12% Bis-Tris gel with MES running buffer (Invitrogen) and transferred to a Hybond-PVDF membrane (Amersham). The membrane was first blocked in TBS containing 0.1% Tween 20 and 5% nonfat dried milk and then incubated with the following primary antibodies: rabbit anti-human Geminin (sc-13015); goat anti-human Orc2 (sc-13238); goat anti-human MCM3 (sc-9850); goat anti-human MCM6 (sc-9843); mouse anti-human MCM7 (sc-9966); rabbit anti-human Chk1 (sc-7898); rabbit anti-human Chk2 (sc-9064); rabbit anti-human phospho Chk2-Thr 68 (sc-16297); rabbit anti-human p53 (sc-6243); rabbit anti-human phospho p53-Ser 15 (sc-11764); rabbit anti-human phospho p53-Ser 20 (sc-18079); rabbit anti-human p21Waf1/Cip1 (sc-397); mouse anti-human Rb (sc-102); rabbit anti-human Rb (sc-50); rabbit anti-human phospho Rb-Ser 795 (sc-7986); mouse anti-human p34Cdc2 (sc-054); rabbit anti-human cyclin B (sc-055); rabbit anti-human cyclin A (sc-751); and goat anti-human Akt1/PKB (sc-1618) from Santa Cruz Biotechnology, Cell cycle/Checkpoint sampler kit (including rabbit anti-human phospho Chk1-Ser 345, rabbit anti-human phospho Rb-Ser 807/811) and Phospho-p53 antibody sampler kit (including rabbit anti-human phospho p53-Ser 6, 9, 15, 20, 37, 46, and 392) obtained from Cell Signaling Technology: and then with peroxidase-linked anti-mouse immunoglobulin, anti-rabbit immunoglobulin, or anti-goat immunoglobulin antibody (Amersham). Enhanced chemiluminescence reagents were used to detect the signals according to the manufacturer's instruction (ECL Plus, Amersham).

For preparation of chromatin fraction, cells were treated prior to lysis as described (Todorov et al., 1995; Shreeram et al., 2002). Briefly, cells were washed with PBS and with cytoskeleton (CSK) buffer (10 mM PIPES, pH 7.0, 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2). Then the cells were extracted twice with CSK buffer containing 0.5% Triton X-100, 0.5 mM PMSF, 1 μg/ml of each of leupeptin, pepstatin, and aprotinin for 5 min at 20°C. The supernatant corresponding to the nonchromatin-bound fraction was separated by centrifugation at 1500 r.p.m. for 5 min. The remaining pellet containing the chromatin-bound fraction was washed with the same buffer and lysed in RIPA lysis buffer as above. For quantifying immunoblotting, Scion Image 4.02 software ( was used.

Cell proliferation and cell cycle analysis

Cells were counted after trypan blue treatment. Cells were synchronized to G1–S by 0.4 mM mimosine (Sigma) for 24 h (Gilbert et al., 1995). To prepare the M–G1 phase cell lysates, the cells blocked for 24 h by nocodazole (200 ng/ml) were harvested after the 3 h of release. DNA replication activity was measured by bromodeoxyuridine (BrdU) incorporation into newly synthesized DNA using Cell Proliferation ELISA (Roche). BrdU labeling was conducted by adding BrdU to the tissue culture medium before harvest for a period of 2 h. Standard methods were used for flow-cytometry analysis. Briefly, cells were harvested and washed once with PBS and fixed in ice-cold 70% ethanol with vigorous mixing. Cells were pelleted, washed once with PBS, and resuspended in 1 ml of PBS at 5 × 105 cells/ml containing 25 μg/ml propidium iodide and 0.1 mg/ml RNase A for 30 min in the dark at 37°C. The cells were analysed with a FACScan flow cytometer (Becton Dickinson) with Cell Quest software. The data from 30 000 cells were collected for analysis. Student's t-test was used for determining statistical significance throughout the study.

Assay for rescue and viral plasmid replication and maintenance

Mammalian expression vectors encoding full-length Cdt1 (Wohlschlegel et al., 2000) and SV40 large T antigen (SV40-T) under CMV promoter and the equivalent mock plasmid, pcDNA3 (Invitrogen) were used. Retroviral expression plasmid pLXSN encoding human papillomavirus type16 E6 or E7 (Halbert et al., 1992) was kindly provided by Dr Galloway (Fred Hutchinson Cancer Research Center, Seattle, USA). These vectors were cotransfected into −/LoxP cells with pBabe-puro (10 : 1) and stable clones were isolated after 2–3 weeks of puromycin (1 μg/ml; Invitrogen) selection.

For long-term assay of episomal plasmid replication, cells (2.5 × 105 per 35-mm dish) were transfected with a plasmid (1 μg). Cells were washed extensively 1 day after transfection and 10% of the cells were returned to culture. At 48 h after transfection, hygromycin (100 μg/ml) was added and cells were selected for up to 2 weeks. Colonies were stained with crystal violet. The minimal replicator of EBV oriP plasmid, p396 is identical to p220 except that it has deletion in the EBNA1 DNA binding domain (Yates et al., 2000), was kindly provided by Dr Yates (Roswell Park Cancer Institute, Buffalo, USA).

In vivo assays

Cells were harvested before inoculation and resuspended in serum-free medium at a concentration of 5 × 107 cells per milliliter. Cells (5 × 106 cells in 0.1 ml) were then inoculated subcutaneously (s.c.) at the proximal dorsal midline into 4-week-old female athymic BALB/c-nu/nu mice (Japan Clea), where inoculations were made in the right flanks. Tumor sizes in two dimensions were measured weekly, and volumes were calculated with the formula (L × W2) × 0.5, where L is the length and W the width. Mice were housed in barrier environments, with food and water provided ad libitum. Animal care and use were in accordance with Institutional and the University of Tokyo guidelines.

Histological analysis

Formalin-fixed, paraffin-embedded sections were used for hematoxylin–eosin (H&E) staining and for immunohisto-chemical analysis. To assess proliferative activity of tumor cells, immunohistochemical staining for Ki-67, a proliferation-associated antigen, was performed by standard peroxidase method. In brief, deparaffinized sections were pretreated by autoclave heating (120°C for 5 min) followed by anti-Ki-67 mAb (clone MIB-1, DAKO) treatment. Antibody-binding sites were detected using peroxidase-conjugated polyvalent detection system (ENVISION kit, DAKO). The chromogen used was 3,3-diaminobenzidine tetrahydrochloride (DAB, brown). For detection of apoptosis, TUNEL method (Gavrieli et al., 1992) was applied by using commercially available kit (ApopTag In situ apoptosis detection kit, Intergen). TUNEL was performed according to the manufacturer's suggested method with slight modifications. In brief, deparaffinized sections were treated with proteinase K (20 μg/ml for 15 min at 37°C), and incubated in TUNEL mixture containing FITC-labeled dUTP and recombinant terminal deoxynucleotidyltransferase (TdT). Detection of incorporated FITC-labeled dUTP was done using peroxidase-conjugated rabbit anti-FITC F(ab′) antibody (DAKO). The chromogen used was VIP (Vector VIP substrate, Vector Labs, purple).



anaphase-promoting complex






Epstein–Barr virus


cyclin-dependent kinases


minichromosome maintenance proteins


origin recognition complex


  1. Abraham RT . (2001). Genes Dev., 15, 2177–2196.

  2. Bell SP and Dutta A . (2002). Annu. Rev. Biochem., 71, 333–374.

  3. Blow JJ and Hodgson B . (2002). Trends Cell Biol., 12, 72–78.

  4. Blow JJ and Laskey RA . (1988). Nature, 332, 546–548.

  5. Chaudhuri B, Xu H, Todorov I, Dutta A and Yates JL . (2001). Proc. Natl. Acad. Sci. USA, 98, 10085–10089.

  6. Chehab NH, Malikzay A, Appel M and Halazonetis TD . (2000). Genes Dev., 14, 278–288.

  7. DeCaprio JA, Ludlow JW, Figge J, Shew J-Y, Huang C-M, Lee W-H, Marsilio E, Paucha E and Livingston DM . (1988). Cell, 54, 275–282.

  8. Dhar SK, Yoshida K, Machida Y, Khaira P, Chaudhuri B, Wohlschlegel JA, Leffak M, Yates J and Dutta A . (2001). Cell, 106, 287–296.

  9. Diffley JF . (2001). Curr. Biol., 11, R367–R370.

  10. Diffley JF, Cocker JH, Dowell SJ and Rowley A . (1994). Cell, 78, 303–316.

  11. Dimitrova DS and Gilbert DM . (2000). Nat. Cell Biol., 2, 686–694.

  12. Dyson NP, Howley M, Munger K and Harlow E . (1989). Science, 243, 934–937.

  13. Feijoo C, Hall-Jackson C, Wu R, Jenkins D, Leitch J, Gilbert DM and Smythe C . (2001). J. Cell Biol., 154, 913–923.

  14. Gavrieli Y, Sherman Y and Ben-Sasson SA . (1992). J. Cell Biol., 119, 493–501.

  15. Gilbert DM, Neilson A, Miyazawa H, DePamphilis ML and Burhans WC . (1995). J. Biol. Chem., 270, 9597–9606.

  16. Halbert CL, Demers GW and Galloway DA . (1992). J. Virol., 66, 2125–2134.

  17. Kroll KL, Salic AN, Evans LM and Kirschner MW . (1998). Development, 125, 3247–3258.

  18. Labib K, Kearsey SE and Diffley JF . (2001). Mol. Biol. Cell, 12, 3658–3667.

  19. Lei M and Tye BK . (2001). J. Cell Sci., 114, 1447–1454.

  20. Maiorano D, Moreau J and Mechali M . (2000). Nature, 404, 622–625.

  21. McGarry TJ . (2002). Mol. Biol. Cell, 13, 3662–3671.

  22. McGarry TJ and Kirschner MW . (1998). Cell, 93, 1043–1053.

  23. Mihaylov IS, Kondo T, Jones L, Ryzhikov S, Tanaka J, Zheng J, Higa LA, Minamino N, Cooley L and Zhang H . (2002). Mol. Cell. Biol., 22, 1868–1880.

  24. Nishitani H, Lygerou Z, Nishimoto T and Nurse P . (2000). Nature, 404, 625–628.

  25. Quinn LM, Herr A, McGarry TJ and Richardson H . (2001). Genes Dev., 15, 2741–2754.

  26. Shieh SY, Ahn J, Tamai K, Taya Y and Prives C . (2000). Genes Dev., 14, 289–300.

  27. Shreeram S, Sparks A, Lane DP and Blow JJ . (2002). Oncogene, 21, 6624–6632.

  28. Tada S, Li A, Maiorano D, Mechali M and Blow JJ . (2001). Nat. Cell Biol., 3, 107–113.

  29. Todorov IT, Attaran A and Kearsey SE . (1995). J Cell Biol., 129, 1433–1445.

  30. Waldman T, Kinzler KW and Vogelstein B . (1995). Cancer Res., 55, 5187–5190.

  31. Werness BA, Levine AJ and Howley PM . (1990). Science, 248, 76–79.

  32. Wohlschlegel JA, Dwyer BT, Dhar SK, Cvetic C, Walter JC and Dutta A . (2000). Science, 290, 2309–2312.

  33. Wohlschlegel JA, Kutok JL, Weng AP and Dutta A . (2002). Am. J. Pathol., 161, 267–273.

  34. Yanagi K, Mizuno T, You Z and Hanaoka F . (2002). J. Biol. Chem., 277, 40871–40880.

  35. Yates JL, Camiolo SM and Bashaw JM . (2000). J. Virol., 74, 4512–4522.

  36. Zachariae W and Nasmyth K . (1999). Genes Dev., 13, 2039–2058.

Download references


We are grateful to Yoko Kanamori for technical assistance, Haruo Onoda for histochemical staining and Shigeo Mori for help with histology. We thank the members of our laboratory for their helpful discussions during the course of this work. This study was supported in part by a grant-in-aid from the Ministry of Education, Science, Sports and Culture, Japan.

Author information

Correspondence to Kenichi Yoshida.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Yoshida, K., Oyaizu, N., Dutta, A. et al. The destruction box of human Geminin is critical for proliferation and tumor growth in human colon cancer cells. Oncogene 23, 58–70 (2004) doi:10.1038/sj.onc.1206987

Download citation


  • Geminin
  • destruction box
  • DNA replication initiation
  • tumor growth

Further reading