Cyclin D1 - Nature

Oncogene (2006) 25, 1620–1628

& 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc

REVIEW

Cyclin D1: polymorphism, aberrant splicing and cancer risk KE Knudsen1,2,3, J Alan Diehl4, CA Haiman5 and ES Knudsen1,2,3 1

Department of Cell Biology, University of Cincinnati, Cincinnati, OH, USA; 2University of Cincinnati Cancer Center, University of Cincinnati, Cincinnati, OH, USA; 3Center for Environmental Genetics, University of Cincinnati, Cincinnati, OH, USA; 4Abramson Family Cancer Research Institute, University of Pennsylvania Cancer Center, University of Pennsylvania, Philadelphia, PA, USA and 5 Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA

The cyclin D1 proto-oncogene exercises powerful control over the mechanisms that regulate the mitotic cell cycle, and excessive cyclin D1 expression and/or activity is common in human cancers. Although somatic mutations of the cyclin D1 locus are rarely observed, mounting evidence demonstrates that a specific polymorphism of cyclin D1 (G/A870) and a protein product of a potentially related alternate splicing event (cyclin D1b) may influence cancer risk and outcome. Herein, we review the epidemiological and functional literatures that link these alterations of cyclin D1 to human tumor development and progression. Oncogene (2006) 25, 1620–1628. doi:10.1038/sj.onc.1209371 Keywords: cell cycle; cyclin-dependent kinase; alternative splicing; nuclear localization; retinoblastoma tumor suppressor

Introduction The cancer relevance of the cyclin D1 gene was apparent upon its identification in 1991; the cyclin D1 locus (also called PRAD1 or CCND1) was initially identified based on its involvement in a chromosomal rearrangement of benign parathyroid tumors (Motokura et al., 1991). Cyclin D1 was simultaneously identified by two groups; the first identified cyclin D1 in a subtractive screen for growth factor inducible genes (Matsushime et al., 1991) and the second group identified cyclin D1 through its ability to complement G1 cyclin loss in S. cerevisae (Xiong et al., 1991). These initial findings suggested that cyclin D1 harbors the capacity to modulate cell cycle progression. From these beginnings, a wealth of studies systematically defined the manner in which cyclin D1 functions to regulate cell cycle and oncogenic transformation in mammalian cells (Sherr, 1995; Diehl, 2002; Gladden and Diehl, 2005). The cyclin D1 locus is known to be amplified in specific tumor types (Jiang et al., 1992; Buckley et al., 1993; Zhang et al., 1993; Gillett et al., Correspondence: Dr E Knudsen, Department of Cell Biology, University of Cincinnati Cancer Center, Center for Environmental Genetics, University of Cincinnati, 3125 Eden Avenue, Cincinnati, OH 45267, USA. E-mail: [email protected]

1994; Sherr, 1995; Palmero and Peters, 1996), and it is thought that this event renders a net increase in the proto-oncogenic functions of the cyclin D1 protein. Recent evidence indicates that the action of cyclin D1 in cancer may also be influenced by a common polymorphism that is believed to alter splicing of the cyclin D1 transcript and influence cancer risk and/or outcome. The present review is focused on the role of this polymorphism in cancer risk, the alternative splicing of the cyclin D1 RNA, and role of the resultant cyclin D1 variant protein in human cancer.

Cyclin D1 and cell cycle control Cyclin D1 expression is highly regulated, as would be expected based on its powerful role in proliferative control. Cyclin D1 expression is induced as a delayed early response to many mitogenic signals (Sherr et al., 1992; Sherr, 1995), and is universally associated with the transition from quiescence into the proliferative cycle (Won et al., 1992; Winston and Pledger, 1993; Lukas et al., 1994b; Ladha et al., 1998; Diehl, 2002). Among the cyclins that regulate G1 progression, it is hypothesized that stimulation of cyclin D1 expression represents the point at which mitogenic signal transduction cascades are integrated to mediate engagement of the cell cycle machinery (Figure 1). After induction, both cyclin D1 mRNA and protein levels are under stringent regulation. First, cyclin D1 mRNA levels are dramatically increased following mitogenic stimulation. This occurs in part through direct induction of the cyclin D1 promoter (Herber et al., 1994; Muller et al., 1994), such as after estrogen stimulation in breast cancer cells (Sabbah et al., 1999). Additionally, many oncogenes (e.g. Ras and b-catenin) harbor the capacity to induce cyclin D1 promoter activity (Albanese et al., 1995; Tetsu and McCormick, 1999). However, the cyclin D1 promoter is very complex and in general is only weakly induced in in vitro reporter assays after mitogen stimulation. This result suggests that much of the regulation of cyclin D1 action could occur either through intrinsic complexities associated with the transcription initiation/elongation, such as chromatin structure or long-range enhancer elements or via coordinate effects on mRNA stability. In fact, it has been shown that cyclin D1

Cyclin D1 variants and cancer KE Knudsen et al

1621 Anti-proliferative Signals

Mitogenic Signals

Cdk4/6 Cyc D1

p16ink4a

RB E2F

P RB

P E2F

Cell Cycle

Figure 1 Schematic role for cyclin D1 in the cell cycle. Mitogenic and antiproliferative signals are integrated to dictate the levels of cyclin D1 in the cell. Cyclin D1 can associate with the CDK4 or CDK6 proteins to activate catalytic activity. This action of cyclin D1 is antagonized by p16ink4a, which binds to the CDK-moiety and disrupts the association with cyclin D1. Active CDK/cyclin D1 complexes initiate phosphorylation of the retinoblastoma tumor suppressor, which disrupts ability RB-mediated transcriptional repression of E2F and facilitates cell cycle progression.

mRNA regulation occurs through mitogen-induced stabilization of the mRNA transcript (Dufourny et al., 2000; Lin et al., 2000). Once produced, the cyclin D1 protein is intrinsically unstable, thus providing an additional level of regulation. Instability of the protein has been attributed to a C-terminal PEST sequence (Dragnev et al., 2001) and the presence of a specific GSK-3B phosphorylation site (threonine 286) (Diehl and Sherr, 1997; Diehl et al., 1997). Although highly invoked in the literature, there has not been a detailed analysis of the role for the PEST sequence in regulating cyclin D1 protein stability. In contrast, phosphorylation of the T286 residue has been shown to target the cyclin D1 protein for nuclear export and proteasomal degradation (Alt et al., 2000). Indeed, phosphorylationdependent destruction of cyclin D1 provides a critical barometer to prevent overexpression of cyclin D1 resulting from continuous mitogen-dependent cyclin D1 transcription. As such, mitogen-dependent induction of cyclin D1 and the resultant cell cycle progression reflects the induction of a threshold of cyclin D1 protein, which is determined by the rate of expression versus destruction. Lastly, cyclin D1 is marked for degradation after genomic insult (Agami and Bernards, 2000), thus preventing cell cycle progression in the presence of DNA damage. Combined, these mechanisms contribute to the accumulation of cyclin D1 protein and are believed to be critical for ensuring the progression into the cell cycle is restricted to the appropriate mitogenic context. The predominant action of cyclin D1 in cell cycle control is believed to be manifest through interactions with CDK4 and/or CDK6 (Matsushime et al., 1992, 1994; Meyerson and Harlow, 1994; Sherr, 1995). CDK4 and CDK6 belong to a distinct subarm of the family of cyclin-dependent kinases, which specifically bind D-type cyclins. Association with cyclin D1 activates the CDK moiety and promotes phosphorylation of specific substrates (Figure 1). CDK/cyclin D1 complexes act on a limited repertoire of substrates (Matsushime et al., 1992, 1994), relative to other promiscuous CDK/cyclin com-

plexes. The first substrate defined for CDK4/cyclin D1 complexes was the retinoblastoma tumor suppressor protein, RB (Ewen et al., 1993; Kato et al., 1993; Matsushime et al., 1992, 1994). In its active state, RB inhibits cell cycle progression through its ability to repress transactivation of genes required for DNA replication and G2/M progression (reviewed in Sherr and McCormick, 2002; Cam and Dynlacht, 2003; Cobrinik, 2005). Activated CDK4/cyclin D1 complexes initiate phosphorylation of RB in response to mitogens (Mittnacht, 1998; Harbour et al., 1999). This phosphorylation event disrupts the ability of RB to mediate transcriptional repression and therefore nullifies its antiproliferative function (Hinds et al., 1992; Dowdy et al., 1993; Ewen et al., 1993; Harbour et al., 1999). Early evidence suggested that the ability of cyclin D1 to promote RB phosphorylation is crucial for its cell cycle function, as cancer cells that are defective in RB are refractory to cyclin D1 action (Lukas et al., 1994a, 1995a). This hypothesis is further supported by the finding that cyclin D1 expression is often attenuated in RB-deficient tumors (Bartkova et al., 1994; Muller et al., 1994), suggesting that the two events confer redundant growth advantages. The concept that RB is a critical downstream target of cyclin D1 was reinforced by study of the p16ink4a tumor suppressor. p16ink4a binds directly to CDK4 and CDK6 and displaces cyclin D1 from the complex (Serrano et al., 1993; Koh et al., 1995; Russo et al., 1998; Jeffrey et al., 2000). Through this ability to antagonize cyclin D1 function, p16ink4a blocks cell cycle progression, and the antiproliferative action of p16ink4a is dependent both on the presence of CDK4/6 activity and RB (Koh et al., 1995; Lukas et al., 1995b). Clinical analyses support a model wherein p16ink4a, cyclin D1 and RB fall in a common pathway (Figure 1), as loss of p16ink4a, loss of RB, or excessive activation of CDK4/cyclin D1 are mutually exclusive events in human cancers (Otterson et al., 1994; Kelley et al., 1995; Palmero and Peters, 1996). Disruption of this ‘p16-cyclin D1-RB’ axis is known to occur in the majority of human tumors (Palmero and Peters, 1996; Sherr and McCormick, 2002), thus underscoring the importance of this pathway in growth regulation. Several additional substrates of CDK4/cyclin D1 have been identified, including Smad3 (Liu and Matsuura, 2005), CDT1 (Liu et al., 2004), and the RB-related proteins p107 (Matsushime et al., 1992; Leng et al., 2002) and p130 (Canhoto et al., 2000). Each putative substrate harbors an established role in cell cycle control and therefore may influence tumorigenesis. Identification of the comprehensive cohort of CDK/cyclin D1 substrates and assessing each specific contribution to cyclin D1 action will reveal the biological consequence of CDK-mediated cyclin D1 function. Transcriptional roles for Cyclin D1 Although cyclin D1 acts through CDK4 to regulate cell cycle progression, a growing body of evidence demonstrates that cyclin D1 harbors CDK-independent Oncogene


1622

activities (reviewed in Coqueret, 2002; Ewen and Lamb, 2004). Evidence for CDK-independent functions was first observed in clinical samples, wherein breast cancers overexpressing cyclin D1 failed to show enhanced proliferation rates, and control of the RB pathway was retained (Oyama et al., 1998; Lamb et al., 2003). In addition, B-cell lymphomas overexpressing cyclin D1 retained RB in its activated state (Zukerberg et al., 1996). Thus, it was suspected that cyclin D1 likely harbors activities outside the cell cycle. Subsequent studies strikingly demonstrated that cyclin D1 is a modifier of gene transcription, and that this function of cyclin D1 has important implications in tumor biology. Specifically, cyclin D1 has been shown to regulate a number of sequence-specific transcription factors, including C/EBPb (Lamb et al., 2003), STAT3 (Bienvenu et al., 2001), DMP1 (Inoue and Sherr, 1998), and BETA2/NeuroD (Ratineau et al., 2002). The largest class of transcription factors regulated by cyclin D1 belong to the nuclear receptor superfamily, and include the estrogen receptor (Zwijsen et al., 1998; Lamb et al., 2000), androgen receptor (Knudsen et al., 1999; Reutens et al., 2001; Petre et al., 2002; Burd et al., 2005), thyroid hormone receptor (Pibiri et al., 2001), and peroxisome proliferator activated receptor-g (Qin et al., 2003). In many cases cyclin D1 was shown to directly associate with the transcription factors, independent of CDK4 association, and modify transcription factor action through cell-cycle independent mechanisms. In this capacity, cyclin D1 can enhance transcriptional activation (e.g. through recruitment of coactivators such as P/CAF), or induce transcriptional repression (e.g. through association with histone deacetylases such as HDAC3). Each of these activities have been mapped to specific regions of the cyclin D1 protein. Specifically, transcriptional activation is associated with an LxxLL motif that is important for coactivator recruitment (Zwijsen et al., 1998; McMahon et al., 1999), whereas transcriptional repression is associated with the presence of a repressor domain for selected transcription factors (Petre-Draviam et al., 2005). The biological significance of the cyclin D1 transcriptional regulatory function was solidified in a large scale, comprehensive analysis of genes deregulated in breast cancer specimens (Lamb et al., 2003). This study unexpectedly revealed that cyclin D1 overexpression does not disrupt the RB function in breast cancer, and the cellular changes associated with cyclin D1 overexpression were not attributed to alterations in cell cycle. Rather, the principal action of cyclin D1 in this tissue type was shown to occur via direct modulation of C/EBPb action. Thus, emerging evidence demonstrates that the ability of cyclin D1 to regulate transcription may strongly influence its function in cancer. Cyclin D1 alterations in cancer Given its role in promoting cellular proliferation and modulating transcription, it is not surprising that cyclin D1 is deregulated in cancer. Interestingly, cyclin D1 Oncogene

alterations are tumor-type specific, and occur through chromosomal translocation, gene amplification, or excessive accumulation of the protein. Chromosomal translocation of cyclin D1 is relatively rare in most cancers, but occurs in greater than 90% of Mantle Cell lymphoma, a mature B-cell malignancy (Bosch et al., 1994; De Wolf-Peeters and Pittaluga, 1994; Bigoni et al., 1996). This same translocation (11;14)(q13;q32) is also observed at a lower frequency mutliple myeloma and specific leukemias (Bosch et al., 1994; Bergsagel and Kuehl, 2005). By contrast, cyclin D1 gene amplification or overexpression occurs with significant frequency in other tumor types, such as breast cancer and esophageal carcinoma, wherein 40–50% of primary carcinomas overexpress cyclin D1 (Jiang et al., 1992, 1993; Buckley et al., 1993; Gillett et al., 1994). In addition to these mechanisms, it is tempting to speculate that activating mutations of cyclin D1 may also be selected for in human cancers, which may enhance association with CDK4 or alter the ability of cyclin D1 to modulate gene transcription. This hypothesis is not supported by the current literature, wherein surprisingly few tumorderived mutations of cyclin D1 have been reported (Hosokawa et al., 1995). However, recent evidence has indicated that a polymorphism of cyclin D1 (G/A 870) may influence cancer risk and outcome. Cyclin D1 polymorphisms and cancer risk Over 100 single nucleotide polymorphisms have been identified spanning the cyclin D1 locus and catalogued in public single nucleotide polymorphism databases (dbSNP: www.ncbi.nlm.nih.gov/SNP/; HapMap: www. hapmap.org; or GeneSNPs: www.genome.utah.edu/ genesnps/). No common missense polymorphisms have been identified through these discovery efforts. However, whether this gene contains rare missense polymorphisms (p1% frequency) will require deeper resequencing in large numbers of subjects with and without cancer phenotypes. Of the polymorphisms identified, the cyclin D1 G/A870 polymorphism has received the most investigation. The polymorphism frequency in the Caucasion population is approximately 44% A and 56% G, depending on the study (Simpson et al., 2001; Sanyal et al., 2004), but large variances between racial and ethnic groups have been reported. Owing to the significance of cyclin D1 in human cancer, a large number of epidemiological studies have challenged the influence of this particular polymorphism in cancer susceptibility and disease outcome. These studies generally compare the allelic frequency of G/A870 in disease affected or unaffected individuals, and assess correlations with clinical parameters (e.g. stage at diagnosis or overall survival). As summarized in Table 1, studies have shown specific genotypes harbor increased cancer risk or poor outcome for a large number of tumor types. The majority of studies link the A-allele to increased cancer risk and poor disease outcome, with the largest associations observed with the A/A genotype. In these

Table 1 Published studies investigating G/A870 polymorphism in cancer Effect polymorphism

Reference

Basal cell carcinoma

Increased risk with A/A

(Zhang et al., 2004)

Bladder Bladder Bladder Bladder

Increased risk for disease progression with A/A No effect of the polymorphism No effect of the polymorphism Increased risk with A/A

(Ito et al., 2004) (Sanyal et al., 2004) (Cortessis et al., 2003) (Wang et al., 2002)

Breast cancer Breast cancer Breast cancer

Increased risk with A/A (specific conditions) No effect of the polymorphism No effect of the polymorphism

(Shu et al., 2005) (Krippl et al., 2003) (Grieu et al., 2003)

Cardiac cancer


(Wang et al., 2003b)

Cervical cancer

Increased risk with G/G

(Catarino et al., 2005)

Colorectal Colorectal Colorectal Colorectal Colorectal Colorectal Colorectal

A/A associated with early disease onset No influence of the polymorphism Increased risk with G/G Increased risk with A/A or A/G Increased risk with A/A and A/G A/A or A/G associated with early onset Increased risk with A/A or A/G

(Kong et al., 2001) (McKay et al., 2000) (Hong et al., 2005) (Lewis et al., 2003) (Le Marchand et al., 2003) (Kong et al., 2000) (Porter et al., 2002)

Esophagealadenocarcinoma, gastroesophageal reflux disease, Barrett esophagus, Esophageal cancer Esophageal and gastric cardiac carcinoma

Increased risk with A/A Increased risk with A/A Increased with risk A/G and A/A

(Casson et al., 2005) (Wang et al., 2003b) (Zhang et al., 2003)

Gastro-intestinal Gastro-intestinal Gastro-intestinal Gastro-intestinal

Increased risk with A/A No effect of the polymorphism Increased risk with A/A and A/G Increased risk with A/A and A/G

(Buch et al., 2005) (Geddert et al., 2005) (Nishimoto et al., 2004) (Nishimoto et al., 2004)

Increased risk with A/A Increased risk with A/A Increased risk with G/G Increased risk with G/G or G/A No effect of the polymorphism G/G associated with reduced disease free survival

(Zheng et al., 2001) (Rydzanicz et al., 2005) (Holley et al., 2005b) (Deng et al., 2002) (Wong et al., 2003) (Matthias et al., 1998)

Hemangioblastoma Hemangioblastoma (VHL-associated)

No effect of the polymorphism Increased risk with G/A or G/G

(Gijtenbeek et al., 2005) (Zatyka et al., 2002)

Hepatoblastoma Hepatocellular cancer

Early onset with A/A No effect of the polymorphism

(Pakakasama et al., 2004) (Zhang et al., 2002)

Leukemia (acute lymphoblastic) Leukemia (acute lymphoblastic)

Increased risk with A/A A/A associated with reduced disease free survival

(Hou et al., 2005) (Costea et al., 2003)

Lung cancer Lung cancer (non-small cell)

Increased risk with A/A A/A associated with reduced disease free survival

(Qiuling et al., 2003) (Betticher et al., 1995)

Ovarian cancer

A/A associated with early disease progression

(Dhar et al., 1999)

Prostatic adenocarcinoma Prostatic adenocarcinoma (familial)

Increased risk with A/A Increased risk with A/A (specific conditions)

(Wang et al., 2003a) (Koike et al., 2003)

Renal cell carcinoma (sporadic)


(Yu et al., 2004)

Squamous cell carcinoma of the uterine cervix

No effect of the polymorphism

(Jeon et al., 2005)

Head Head Head Head Head Head

carcinoma (in situ) cancer cancer cancer

and and and and and and

cancer cancer cancer cancer cancer cancer cancer

neck neck neck neck neck neck

(advanced) (sporadic) (advanced) (heriditary nonpolyposis) (familial and sporadic)

(upper aero-digestive tract cancer) (adenocarcinoma of the upper tract) (squamous cells carcinoma upper aero-digestive tract) (squamous cells carcinoma upper aero-digestive tract)

(squamous cell carcinoma) (squamous cell carcinoma of the larynx) (oral squamous cell carcinoma (nasopharyngeal carcinoma) (oral squamous cell carcinoma) (squamous cell carcinoma)


Tumor Type

1623

Oncogene


1624

studies, relative risks were significant but typically modest, with many studies reporting less than a twofold effect. Such a result is in part to be expected for a common allele (i.e. A-870) that contributes to a complex phenotype. However, results have been inconsistent and some studies have implicated the G-allele with increased cancer risk, and others have ascribed no significant value to any allele of the G/A870 polymorphism. Combined, these results indicate that individual alleles may harbor differential effects in distinct tumor types. However, even within a specific tumor type (e.g. colorectal cancer) there have been disparate conclusions regarding the significance of the polymorphism. In part these disparities could be reflective of the patient population under study and the possible involvement of external factors (e.g. smoking or obesity) that have been suggested to cooperate with the polymorphism in specific studies (Buch et al., 2005; Shu et al., 2005). Thus, although the majority of observations are indicative of a role for the polymorphism in cancer risk, analysis of larger patient cohorts such as the NCI’s Breast and Prostate Cancer Cohort Consortium will be required to resolve inconsistencies in the published literature. Important issues that remain to be resolved are whether the G/A870 polymorphism is the specific causal variant and whether there are other polymorphisms at this locus that are biologically relevant. It is formally possible that G/A870 is a proxy that is in linkage disequilibrium with the actual functional variant that is modulating cancer risk. Such a possibility may in part explain some of the discrepancies associated with the role of G/A870 in cancer in different study populations and emphasizes the importance of conducting linkage disequilibrium or haplotype-based investigations of common genetic variation across the entire locus. One group has coordinately analysed both the G/A870 and the G/C1722 polymorphism of cyclin D1 (Holley et al., 2001). This study indicated that the two polymorphisms are in linkage disequilibrium, such that individuals harboring the A870-allele are most likely also carrying the C1722-allele. In spite of this observation, each polymorphism had a distinct influence on disease (Holley et al., 2001), suggesting that other variants in addition to the G/A870 variant in cyclin D1 may be important. Therefore, although more exhaustive studies will be required to conclusively determine the involvement of the G/A870 polymorphism in cancer, there is significant evidence that it alters risk, thus necessitating study of the functional significance of the polymorphism.

have detected a cyclin D1 message that derives from alternative splicing (Betticher et al., 1995; Hosokawa et al., 1997, 1999; Bala and Peltomaki, 2001). This isoform was initially defined following the identification of the G/A870 polymorphism, that occurs at the intron 4/exon 5 boundary. As shown in Figure 2, the G870 allele creates an optimal splice donor site, resulting in the well-described transcript for cyclin D1 (‘transcript a’). By contrast, the A870 allele is predicted to hinder the splicing event, thus allowing for read-through into intron 4 and production of a variant splice product of cyclin D1, termed ‘transcript b’. Consistent with this hypothesis, Betticher et al. (1995), Holley et al. (2001), and Howe and Lynas (2001) demonstrated that the A-allele of cyclin D1 was preferentially associated with transcript b production. However, individuals with the A/A genotype can still produce transcript a (Bala and Peltomaki, 2001; Holley et al., 2001), thus indicating that the 870-A allele is not completely penetrant for transcript b production, and individuals with the G/G genotype can produce transcript b. It should be noted that the presence of the A-nucleotide at the splice donor site does represent a change from the consensus (Figure 2), wherein the G-Allele represents 91% consensus using a weighted splice site identification matrix and the A-Allele is 85% (Shapiro and Senapathy, 1987). Importantly, both sites would be predicted to be functional and mutations in splice donor sequence analgous to the A-allele are appropriately spliced (Wieringa et al., 1983). Thus, modifiers in addition to the polymorphic 870-allele must influence the splicing event. In principle these could be additional polymorphisms within the cyclin D1 gene or trans-acting factors, which have yet to be defined. Interpretation of epidemiological studies comparing the G/A870 polymorphism to cancer risk may be confounded by the dysjunction between the genotype and transcript b production. To date, no study has directly examined the relationship of the G/A870 polymorphism to transcript b production using quantitative/mechanistic methodologies. However, individual assessment of the prognostic values of the polymorphism and transcript b production in colon cancers revealed that while the polymorphism held no predictive

Exon

Consensus: AG GTRAGT

CG GTAAGT

Intron 4

not defined

91%

G/A870 polymorphism and splicing A870:

Oncogene

Exon

(Y)nCAG GT

Cyclin D1 Exon 4

G870:

The cyclin D1 gene is comprised of five distinct exons and until recently, splicing of cyclin D1 was not considered a significant regulatory factor. Two transcripts of cyclin D1 can be detected in normal cells (Howe and Lynas, 2001), although regulation of splicing appears to be most relevant in cancer. Numerous groups

Intron

YNYURAC

CA GTAAGT 85%

Exon 5

(Y)nCAG GA 93%

not defined

(Y)nCAG GA 93%

Figure 2 Cyclin D1 polymorphism and splicing of exon 4 and exon 5. Depicted is a consensus splicing site and the sequences surrounding the exon/intron junctions of cyclin D1. The program splice site was utilized to determine the score for the splice donor and acceptor sites relative to the consensus sequence.


1625

value, alterations in transcript b production were of significance (Bala and Peltomaki, 2001). Therefore, possible sources of the heterogeneous influence of the cyclin D1 polymorphism in cancer are factors which influence transcript ‘a’ versus ‘b’ splicing independently of the polymorphic nucleotide. Thus, there is an urgent need to solidify the relationship of the polymorphism to transcript b production, and to correspondingly assess the impact of transcript b in human disease. The cyclin D1b variant protein As depicted in Figure 3, transcript b gives rise to a cyclin D1 protein product with a completely divergent C-terminal domain, termed ‘cyclin D1b’. The divergent C-terminal domain is defective in several motifs that typically regulate cyclin D1 turnover, including both the PEST motif and the residue that controls nuclear export and protein stability (threonine 286). Thus, it would be predicted that cyclin D1b is a more stable, constitutively nuclear protein, harboring increased capability to regulate CDK activity and cell cycle control. Several groups have recently assessed the action of cyclin D1b on these parameters. The expectation that cyclin D1b would exhibit increased protein stability (as compared to cyclin D1a) was not significantly observed (Lu et al., 2003; Solomon et al., 2003). Rather, the half-lives and relative accumulation of cyclin D1 and cyclin D1b were indistinguishable in most model systems (Lu et al., 2003; Solomon et al., 2003). By contrast, these same studies validated the prediction that cyclin D1b is constitutively nuclear in localization (Lu et al., 2003; Solomon et al., 2003). This was a critical observation, as constitutive nuclear localization of cyclin D1 (as achieved through mutation of T286) is known to promote oncogenic transformation (Alt et al., 2000). Strikingly, it was demonstrated that cyclin D1b harbors increased transforming capability (as compared to cyclin D1) (Lu et al., 2003; Solomon et al., 2003), thus indicating that cyclin D1b may serve as a more potent oncogene in human cancers. In light of this enhanced oncogenic action, substantial effort has been directed at delineating the mechanisms

underlying cyclin D1b function. Cyclin D1b maintains its capacity to associate with CDK4 (Lu et al., 2003; Solomon et al., 2003), yet may have an altered capacity to drive CDK4 activity. In one set of reported experiments, cyclin D1b/CDK4 complexes immunopurified from cells phosphorylated recombinant retinoblastoma substrate as efficiently as cyclin D1/CDK4 complexes (Lu et al., 2003). Conversely, in a different approach, cyclin D1b was relatively inefficient in its capacity to drive RB inactivation in cells as determined by assessing accumulation of phospho-RB in cells (Solomon et al., 2003). As there is no detailed structural data on the CDK4-cyclin D1 kinase, it is hard to predict why alterations in the extreme C-terminus would effect kinase activity. Based on the crystal structure of other cyclin-CDK complexes, the Cterminus of the cyclin is not predicted to affect interactions with the CDK subunit (Russo et al., 1998; Jeffrey et al., 2000). Thus, one intriguing possibility is that the C-terminal alterations influence substrate recognition and thus substrate specificity. If so, one might anticipate disparate biological properties of the two molecules. Indeed, it was also observed that cyclin D1 but not cyclin D1b can efficiently advance cell cycle progression under conditions of RB-mediated cell cycle arrest or low serum (Solomon et al., 2003; Holley et al., 2005a). Thus, while cyclin D1b has enhanced transforming properties, this function of cyclin D1b may be attributed to novel oncogenic functions. In addition, it will be important to determine if the cyclin oncogenic function of cyclin D1b is limited (or related) to tissue types wherein the A870-allele increases cancer risk. In addition to functions on classical cell cycle control, it would be predicted that the loss of the LxxLL-motif and changes in nuclear localization may alter the transcriptional activity of cyclin D1b. Consistent with this idea, recent studies have demonstrated that in specific tumor types, the ability of cyclin D1b to regulate transcription factor action may be significantly altered and contribute to proliferative control (Burd CJ and Knudsen KE, unpublished data). Future studies to delineate the mechanisms of cyclin D1b function and the contribution of the novel C-terminus to cyclin D1b

G

Transcript a E1

E2

E3

E4

E5

Cyclin D1

RB Binding Motif: 5-9 Cyclin Box: 56-145 T286-nuclear export/destabilization

A

Transcript b E1

E2

E3

E4

I4

PEST Motif: 241-290 LxxLL Motif: 251-255 Repressor Domain: 142-253 Intron 4 encoded: 241-275

Cyclin D1b Figure 3 Domain structure of cyclin D1 and D1b. The predominant domains and their localization within the primary sequence of cyclin D1 and D1b are provided. Oncogene


1626

action will be critical to elucidate its contribution to cancer development and progression. To date, the number of human cancers that may be influenced by cyclin D1b activity have been identified largely through association with the G/A870 polymorphism. Given the unreliable relationship between the genotype and cyclin D1b expression, there is a current emphasis on developing tools to screen for cyclin D1b expression. Although the transcript is relatively easy to detect by PCR based methods (Bala and Peltomaki, 2001; Howe and Lynas, 2001), detection of cyclin D1b protein is subject to the limitation that most available cyclin D1 antibodies were either raised against or preferentially recognize the C-terminus of cyclin D1a and therefore do not detect cyclin D1b. However, Hosokawa et al. (1999) detected cyclin D1b protein in B-lymphoid cell lines employing the DCS-6 monoclonal antibody, which is directed against a domain conserved within both cyclin D1 and D1b. Antibodies have also been developed against the unique sequences of cyclin D1b, so as to distinctly assess its expression levels in human specimens (Lu et al., 2003; Solomon et al., 2003). This approach proved successful in at least one study, which specifically detected the cyclin D1b protein in esophageal tumors (Lu et al., 2003). Analysis of cyclin D1b in alternate tumor types is essential to ascertain its putative impact in tumorigenesis. The prediction that alternative splicing might contribute to cancer-specific alterations in cyclin expression is not without precedent. Alternative splicing of cyclin E has been documented in both cancer-derived cell lines and primary cancer tissue (Mumberg et al., 1997). In addition to regulated splicing, novel cyclin E isoforms that appear to be generated via proteolytic processing in cancer tissue have also been documented (Porter and Keyomarsi, 2000). These novel cyclin E isoforms contribute to a functionally active CDK2 kinase with altered substrate specificity (Porter and Keyomarsi, 2000; Porter et al., 2001) and expression of these isoforms correlates with poor prognosis in breast cancer (Keyomarsi et al., 2002). What remains unique with regard to cyclin D1 is the association between alternative splicing, the polymorphic 870 site and the unique expression of the variant D1b isoform in cancer. Still, the question as to whether alternative cyclin D1 splicing is a cause or an effect of cancer initiation and/or progression is of critical importance.

human disease. First, large multiethnic studies are needed to assist in delineating the precise impact of the polymorphic alleles on tissue-specific cancers. Given the disparity in alleleic distribution across populations and the possibility that the G/A870 polymorphism may be a marker of another variant in the cyclin D1 gene, it may be necessary to analyse each group separately to designate prognostic value to the polymorphism. Second, the relationship between the polymorphism and cyclin D1b production should be clarified. Although it is apparent that specific polymorphic alleles may facilitate generation of transcript b, it is clear that additional factors influence the splicing event. When possible, concurrent analysis of genotype and cyclin D1b expression would clarify the relationship between the two parameters and assign their relative prognostic values in individual tumor types. Third, it will be of significant benefit to identify the mechanisms underlying the enhanced oncogenic action of cyclin D1b. The present data demonstrate that cyclin D1b incorporates novel mechanisms to regulate cellular transformation and transcriptional control, with the contribution of these functions to cancer development and progression as the focus of future studies. Finally, it will be essential to determine whether expression of the alternatively spliced cyclin D1b influences therapeutic efficacy. Although numerous small molecule inhibitors have been developed with a more pointed interest in targeting the CDK2 kinase, recent data suggest that the cyclin D1/ CDK4 kinase may represent a good therapeutic target. CDK4/CDK6-specific small molecule inhibitors have been developed and one such molecule, PD 0332991, exhibits an IC50 in the nanomolar range, with excellent target specificity (Fry et al., 2004; Toogood et al., 2005). In addition, the cytostatic property of PD 0332991 is restricted to tumor-derived cell lines, which retain wildtype Rb and depend upon the cyclin D1/CDK kinase for proliferation (Fry et al., 2004). However, if the cyclin D1b/CDK complexes exhibit alternative substrate specificities that reflect altered structural properties, expression of cyclin D1b will have a significant impact on therapeutic outcome. The current data support a model wherein cyclin D1b displays potentially novel mechanisms to regulate cellular proliferation, transcription factor action, and cellular transformation. Assessing the contribution of each function to cancer development and therapeutic intervention will contribute to our ability to develop cancer therapies and utilize these therapies in an appropriately targeted manner.

Conclusions and lingering questions Cyclin D1 is a critical regulator of the cell cycle and transcriptional processes which is overexpressed at high frequency in human cancer. A large number of studies have implicated the cyclin D1 G/A870 polymorphism as a modulator of cancer risk and/or poor prognosis in human disease, and it has been suggested that specific polymorphic alleles may facilitate production of the cyclin D1b oncogene. However, significant questions remain regarding the influence of this polymorphism in Oncogene

Acknowledgements We appreciate our colleagues in the field, and regret any omissions which may have occurred in the preparation of this review. The laboratories of Karen Knudsen and Erik Knudsen provided assistance for editorial review of the manuscript. Studies on the cyclin D1 polymorphism are supported by grants KEK: RO1-CA099996, JAD: 1R01CA111360, ESK: Comparative Mouse Genomics Centers Consortium U01ES011038.


1627 References Agami R, Bernards R. (2000). Cell 102: 55–66. Albanese C, Johnson J, Watanabe G, Eklund N, Vu D, Arnold A et al. (1995). J Biol Chem 270: 23589–23597. Alt JR, Cleveland JL, Hannink M, Diehl JA. (2000). Genes Dev 14: 3102–3114. Bala S, Peltomaki P. (2001). Cancer Res 61: 6042–6045. Bartkova J, Lukas J, Muller H, Lutzhoft D, Strauss M, Bartek J. (1994). Int J Cancer 57: 353–361. Bergsagel PL, Kuehl WM. (2005). J Clin Oncol 23: 6333–6338. Betticher DC, Thatcher N, Altermatt HJ, Hoban P, Ryder WD, Heighway J. (1995). Oncogene 11: 1005–1011. Bienvenu F, Gascan H, Coqueret O. (2001). J Biol Chem 276: 16840–16847. Bigoni R, Negrini M, Veronese ML, Cuneo A, Castoldi GL, Croce CM. (1996). Oncogene 13: 797–802. Bosch F, Jares P, Campo E, Lopez-Guillermo A, Piris MA, Villamor N et al. (1994). Blood 84: 2726–2732. Buch S, Zhu B, Davis AG, Odom D, Siegfried JM, Grandis JR et al. (2005). Mol Carcinog 42: 222–228. Buckley MF, Sweeney KJ, Hamilton JA, Sini RL, Manning DL, Nicholson RI et al. (1993). Oncogene 8: 2127–2133. Burd CJ, Petre CE, Moghadam H, Wilson EM, Knudsen KE. (2005). Mol Endocrinol 19: 607–620. Cam H, Dynlacht BD. (2003). Cancer Cell 3: 311–316. Canhoto AJ, Chestukhin A, Litovchick L, DeCaprio JA. (2000). Oncogene 19: 5116–5122. Casson AG, Zheng Z, Evans SC, Geldenhuys L, van Zanten SV, Veugelers PJ et al. (2005). Cancer 104: 730–739. Catarino R, Matos A, Pinto D, Pereira D, Craveiro R, Vasconcelos A et al. (2005). Cancer Genet Cytogenet 160: 49–54. Cobrinik D. (2005). Oncogene 24: 2796–2809. Coqueret O. (2002). Gene 299: 35–55. Cortessis VK, Siegmund K, Xue S, Ross RK, Yu MC. (2003). Carcinogenesis 24: 1645–1650. Costea I, Moghrabi A, Krajinovic M. (2003). Pharmacogenetics 13: 577–580. De Wolf-Peeters C, Pittaluga S. (1994). Ann Oncol 5(Suppl 1): 35–37. Deng L, Zhao XR, Pan KF, Wang Y, Deng XY, Lu YY et al. (2002). Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai) 34: 16–20. Dhar KK, Branigan K, Howells RE, Musgrove C, Jones PW, Strange RC et al. (1999). Int J Gynecol Cancer 9: 342–347. Diehl JA. (2002). Cancer Biol Ther 1: 226–231. Diehl JA, Sherr CJ. (1997). Mol Cell Biol 17: 7362–7374. Diehl JA, Zindy F, Sherr CJ. (1997). Genes Dev 11: 957–972. Dowdy SF, Hinds PW, Louie K, Reed SI, Arnold A, Weinberg RA. (1993). Cell 73: 499–511. Dragnev KH, Freemantle SJ, Spinella MJ, Dmitrovsky E. (2001). Ann N Y Acad Sci 952: 13–22. Dufourny B, van Teeffelen HA, Hamelers IH, Sussenbach JS, Steenbergh PH. (2000). J Endocrinol 166: 329–338. Ewen ME, Lamb J. (2004). Trends Mol Med 10: 158–162. Ewen ME, Sluss HK, Sherr CJ, Matsushime H, Kato J, Livingston DM. (1993). Cell 73: 487–497. Fry DW, Harvey PJ, Keller PR, Elliott WL, Meade M, Trachet E et al. (2004). Mol Cancer Ther 3: 1427–1438. Geddert H, Kiel S, Zotz RB, Zhang J, Willers R, Gabbert HE et al. (2005). J Cancer Res Clin Oncol 131: 1–6. Gijtenbeek JM, Boots-Sprenger SH, Franke B, Wesseling P, Jeuken JW. (2005). J Neurooncol 74: 261–266. Gillett C, Fantl V, Smith R, Fisher C, Bartek J, Dickson C et al. (1994). Cancer Res 54: 1812–1817. Gladden AB, Diehl JA. (2005). J Cell Biochem 96: 906–913.

Grieu F, Malaney S, Ward R, Joseph D, Iacopetta B. (2003). Anticancer Res 23: 4257–4259. Harbour JW, Luo RX, Dei Santi A, Postigo AA, Dean DC. (1999). Cell 98: 859–869. Herber B, Truss M, Beato M, Muller R. (1994). Oncogene 9: 1295–1304. Hinds PW, Mittnacht S, Dulic V, Arnold A, Reed SI, Weinberg RA. (1992). Cell 70: 993–1006. Holley SL, Heighway J, Hoban PR. (2005a). Int J Cancer 114: 364–370. Holley SL, Matthias C, Jahnke V, Fryer AA, Strange RC, Hoban PR. (2005b). Oral Oncol 41: 156–160. Holley SL, Parkes G, Matthias C, Bockmuhl U, Jahnke V, Leder K et al. (2001). Am J Pathol 159: 1917–1924. Hong Y, Eu KW, Seow-Choen F, Fook-Chong S, Cheah PY. (2005). Eur J Cancer 41: 1037–1044. Hosokawa Y, Gadd M, Smith AP, Koerner FC, Schmidt EV, Arnold A. (1997). Cancer Lett 113: 123–130. Hosokawa Y, Joh T, Maeda Y, Arnold A, Seto M. (1999). Int J Cancer 81: 616–619. Hosokawa Y, Tu T, Tahara H, Smith AP, Arnold A. (1995). Cancer Lett 93: 165–170. Hou X, Wang S, Zhou Y, Xu Z, Zou Y, Zhu X et al. (2005). Int J Hematol 82: 206–209. Howe D, Lynas C. (2001). Haematologica 86: 563–569. Inoue K, Sherr CJ. (1998). Mol Cell Biol 18: 1590–1600. Ito M, Habuchi T, Watanabe J, Higashi S, Nishiyama H, Wang L et al. (2004). Urology 64: 74–78. Jeffrey PD, Tong L, Pavletich NP. (2000). Genes Dev 14: 3115–3125. Jeon YT, Kim JW, Song JH, Park NH, Song YS, Kang SB et al. (2005). Cancer Lett 223: 259–263. Jiang W, Kahn SM, Tomita N, Zhang YJ, Lu SH, Weinstein IB. (1992). Cancer Res 52: 2980–2983. Jiang W, Zhang YJ, Kahn SM, Hollstein MC, Santella RM, Lu SH et al. (1993). Proc Natl Acad Sci USA 90: 9026–9030. Kato J, Matsushime H, Hiebert SW, Ewen ME, Sherr CJ. (1993). Genes Dev 7: 331–342. Kelley MJ, Nakagawa K, Steinberg SM, Mulshine JL, Kamb A, Johnson BE. (1995). J Natl Cancer Inst 87: 756–761. Keyomarsi K, Tucker SL, Buchholz TA, Callister M, Ding Y, Hortobagyi GN et al. (2002). N Engl J Med 347: 1566–1575. Knudsen KE, Cavenee WK, Arden KC. (1999). Cancer Res 59: 2297–2301. Koh J, Enders GH, Dynlacht BD, Harlow E. (1995). Nature 375: 506–510. Koike H, Suzuki K, Satoh T, Ohtake N, Takei T, Nakata S et al. (2003). Anticancer Res 23: 4947–4951. Kong S, Amos CI, Luthra R, Lynch PM, Levin B, Frazier ML. (2000). Cancer Res 60: 249–252. Kong S, Wei Q, Amos CI, Lynch PM, Levin B, Zong J et al. (2001). J Natl Cancer Inst 93: 1106–1108. Krippl P, Langsenlehner U, Renner W, Yazdani-Biuki B, Wolf G, Wascher TC et al. (2003). Breast Cancer Res Treat 82: 165–168. Ladha MH, Lee KY, Upton TM, Reed MF, Ewen ME. (1998). Mol Cell Biol 18: 6605–6615. Lamb J, Ladha MH, McMahon C, Sutherland RL, Ewen ME. (2000). Mol Cell Biol 20: 8667–8675. Lamb J, Ramaswamy S, Ford HL, Contreras B, Martinez RV, Kittrell FS et al. (2003). Cell 114: 323–334. Le Marchand L, Seifried A, Lum-Jones A, Donlon T, Wilkens LR. (2003). Jama 290: 2843–2848. Leng X, Noble M, Adams PD, Qin J, Harper JW. (2002). Mol Cell Biol 22: 2242–2254. Oncogene


1628 Lewis RC, Bostick RM, Xie D, Deng Z, Wargovich MJ, Fina MF et al. (2003). Cancer Res 63: 8549–8553. Lin S, Wang W, Wilson GM, Yang X, Brewer G, Holbrook NJ et al. (2000). Mol Cell Biol 20: 7903–7913. Liu E, Li X, Yan F, Zhao Q, Wu X. (2004). J Biol Chem 279: 17283–17288. Liu F, Matsuura I. (2005). Cell Cycle 4: 63–66. Lu F, Gladden AB, Diehl JA. (2003). Cancer Res 63: 7056–7061. Lukas J, Bartkova J, Rohde M, Strauss M, Bartek J. (1995a). Mol Cell Biol 15: 2600–2611. Lukas J, Muller H, Bartkova J, Spitkovsky D, Kjerulff AA, Jansen-Durr P et al. (1994a). J Cell Biol 125: 625–638. Lukas J, Pagano M, Staskova Z, Draetta G, Bartek J. (1994b). Oncogene 9: 707–718. Lukas J, Parry D, Aagaard L, Mann DJ, Bartkova J, Strauss M et al. (1995b). Nature 375: 503–506. Matsushime H, Ewen ME, Strom DK, Kato JY, Hanks SK, Roussel MF et al. (1992). Cell 71: 323–334. Matsushime H, Quelle DE, Shurtleff SA, Shibuya M, Sherr CJ, Kato JY. (1994). Mol Cell Biol 14: 2066–2076. Matsushime H, Roussel MF, Ashmun RA, Sherr CJ. (1991). Cell 65: 701–713. Matthias C, Branigan K, Jahnke V, Leder K, Haas J, Heighway J et al. (1998). Clin Cancer Res 4: 2411–2418. McKay JA, Douglas JJ, Ross VG, Curran S, Murray GI, Cassidy J et al. (2000). Int J Cancer 88: 77–81. McMahon C, Suthiphongchai T, DiRenzo J, Ewen ME. (1999). Proc Natl Acad Sci USA 96: 5382–5387. Meyerson M, Harlow E. (1994). Mol Cell Biol 14: 2077–2086. Mittnacht S. (1998). Curr Opin Genet Dev 8: 21–27. Motokura T, Bloom T, Kim HG, Juppner H, Ruderman JV, Kronenberg HM et al. (1991). Nature 350: 512–515. Muller H, Lukas J, Schneider A, Warthoe P, Bartek J, Eilers M et al. (1994). Proc Natl Acad Sci U S A 91: 2945–2949. Mumberg D, Wick M, Burger C, Haas K, Funk M, Muller R. (1997). Nucleic Acids Res 25: 2098–2105. Nishimoto IN, Pinheiro NA, Rogatto SR, Carvalho AL, Simpson AJ, Caballero OL et al. (2004). Oral Oncol 40: 604–610. Otterson GA, Kratzke RA, Coxon A, Kim YW, Kaye FJ. (1994). Oncogene 9: 3375–3378. Oyama T, Kashiwabara K, Yoshimoto K, Arnold A, Koerner F. (1998). Cancer Res 58: 2876–2880. Pakakasama S, Chen TT, Frawley W, Muller CY, Douglass EC, Lee R et al. (2004). Oncogene 23: 4789–4792. Palmero I, Peters G. (1996). Cancer Surv 27: 351–367. Petre CE, Wetherill YB, Danielsen M, Knudsen KE. (2002). J Biol Chem 277: 2207–2215. Petre-Draviam CE, Williams EB, Burd CJ, Gladden A, Moghadam H, Meller J et al. (2005). Oncogene 24: 431–444. Pibiri M, Ledda-Columbano GM, Cossu C, Simbula G, Menegazzi M, Shinozuka H et al. (2001). Faseb J 15: 1006–1013. Porter DC, Keyomarsi K. (2000). Nucleic Acids Res 28: E101. Porter DC, Zhang N, Danes C, McGahren MJ, Harwell RM, Faruki S et al. (2001). Mol Cell Biol 21: 6254–6269. Porter TR, Richards FM, Houlston RS, Evans DG, Jankowski JA, Macdonald F et al. (2002). Oncogene 21: 1928–1933. Qin C, Burghardt R, Smith R, Wormke M, Stewart J, Safe S. (2003). Cancer Res 63: 958–964. Qiuling S, Yuxin Z, Suhua Z, Cheng X, Shuguang L, Fengsheng H. (2003). Carcinogenesis 24: 1499–1503.

Oncogene

Ratineau C, Petry MW, Mutoh H, Leiter AB. (2002). J Biol Chem 277: 8847–8853. Reutens AT, Fu M, Wang C, Albanese C, McPhaul MJ, Sun Z et al. (2001). Mol Endocrinol 15: 797–811. Russo AA, Tong L, Lee JO, Jeffrey PD, Pavletich NP. (1998). Nature 395: 237–243. Rydzanicz M, Golusinski P, Mielcarek-Kuchta D, Golusinski W, Szyfter K. (2005). Eur Arch Otorhinolaryngol. Sabbah M, Courilleau D, Mester J, Redeuilh G. (1999). Proc Natl Acad Sci USA 96: 11217–11222. Sanyal S, Festa F, Sakano S, Zhang Z, Steineck G, Norming U et al. (2004). Carcinogenesis 25: 729–734. Serrano M, Hannon GJ, Beach D. (1993). Nature 366: 704–707. Shapiro MB, Senapathy P. (1987). Nucleic Acids Res 15: 7155–7174. Sherr CJ. (1995). Trends Biochem Sci 20: 187–190. Sherr CJ, Matsushime H, Roussel MF. (1992). Ciba Found Symp 170: 209–219; discussion 219–226. Sherr CJ, McCormick F. (2002). Cancer Cell 2: 103–112. Shu XO, Moore DB, Cai Q, Cheng J, Wen W, Pierce L et al. (2005). Cancer Epidemiol Biomarkers Prev 14: 91–97. Simpson DJ, Fryer AA, Grossman AB, Wass JA, Pfeifer M, Kros JM et al. (2001). Carcinogenesis 22: 1801–1807. Solomon DA, Wang Y, Fox SR, Lambeck TC, Giesting S, Lan Z et al. (2003). J Biol Chem 278: 30339–30347. Tetsu O, McCormick F. (1999). Nature 398: 422–426. Toogood PL, Harvey PJ, Repine JT, Sheehan DJ, VanderWel SN, Zhou H et al. (2005). J Med Chem 48: 2388–2406. Wang L, Habuchi T, Mitsumori K, Li Z, Kamoto T, Kinoshita H et al. (2003a). Int J Cancer 103: 116–120. Wang L, Habuchi T, Takahashi T, Mitsumori K, Kamoto T, Kakehi Y et al. (2002). Carcinogenesis 23: 257–264. Wang R, Zhang JH, Li Y, Wen DG, He M, Wei LZ. (2003b). Zhonghua Yi Xue Za Zhi 83: 1089–1092. Wieringa B, Meyer F, Reiser J, Weissmann C. (1983). Nature 301: 38–43. Winston JT, Pledger WJ. (1993). Mol Biol Cell 4: 1133–1144. Won KA, Xiong Y, Beach D, Gilman MZ. (1992). Proc Natl Acad Sci USA 89: 9910–9914. Wong YK, Lin SC, Chang CS, Tseng YH, Liu CJ, Lin HC et al. (2003). J Oral Pathol Med 32: 265–270. Xiong Y, Connolly T, Futcher B, Beach D. (1991). Cell 65: 691–699. Yu J, Habuchi T, Tsuchiya N, Nakamura E, Kakinuma H, Horikawa Y et al. (2004). J Urol 172: 2410–2413. Zatyka M, da Silva NF, Clifford SC, Morris MR, Wiesener MS, Eckardt KU et al. (2002). Cancer Res 62: 3803–3811. Zhang J, Li Y, Wang R, Wen D, Sarbia M, Kuang G et al. (2003). Int J Cancer 105: 281–284. Zhang YJ, Chen SY, Chen CJ, Santella RM. (2002). Mol Carcinog 33: 125–129. Zhang YJ, Jiang W, Chen CJ, Lee CS, Kahn SM, Santella RM et al. (1993). Biochem Biophys Res Commun 196: 1010–1016. Zhang Z, Liu W, Jia X, Gao Y, Hemminki K, Lindholm B. (2004). Clin Chim Acta 342: 137–143. Zheng Y, Shen H, Sturgis EM, Wang LE, Eicher SA, Strom SS et al. (2001). Carcinogenesis 22: 1195–1199. Zukerberg LR, Benedict WF, Arnold A, Dyson N, Harlow E, Harris NL. (1996). Blood 88: 268–276. Zwijsen RM, Buckle RS, Hijmans EM, Loomans CJ, Bernards R. (1998). Genes Dev 12: 3488–3498.