REGULATION OF CYCLIN E BY UBIQUITIN-MEDIATED PROTEOLYSIS: IMPLICATIONS IN CARCINOGENESIS

Charles H. Spruck¹, Adrian P. L. Smith¹, Susanna Ekholm Reed^1,4, Olle Sangfelt¹,  Jaimie Keck¹, Heimo Strohmaier¹, Juan Mendez², Martin Widschwendter³, Bruce Stillman², Anders Zetterberg⁴ and Steven I. Reed^1*

¹ The Scripps Research Institute, La Jolla, CA, USA
² Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, USA
³ Innsbruck University Hospital, Innsbruck, Austria
⁴ The Karolinska Institute, Stockholm, Sweden
^*  sreed@scripps.edu

INTRODUCTION. Cyclin E is a positive regulator of Cdk2 in metazoan species.  In somatic mammalian cells, cyclin E is expressed periodically during an interval extending from the latter part of G1 until mid-S phase.  Consistent with these kinetics, cyclin E associated Cdk2 kinase activity is thought to mediate S phase events.  On the other hand, deregulation of cyclin E has been correlated with a broad spectrum of human malignancies   

RESULTS AND DISCUSSION. We sought to determine if deregulation of cyclin E could be mediated by defects in the cyclin E degradation pathway.   SCF protein-ubiquitin ligases are characterized by specificity factors, known as F-box proteins that target specific subgroups of proteins.  We therefore analyzed the F-box protein that targets cyclin E, hCdc4, for mutations in tumor-derived cell lines and in primary tumors(1).  One out of 8 breast cancer derived cell lines tested was found to have mutated hCDC4 and loss of heterozygosity (LOH) correlating with stable cyclin E(1).  Next, endometrial tumors were analyzed by SSCP hCDC4 mutations were found in 16% (8/51) and LOH could be confirmed in most cases(2). 

Immunofluorescence analysis of cyclin E in the breast tumor-derived cell line found to be mutated for hCDC4 (SUM149PT) indicated that all cells in an asynchronous population were cyclin E-positive.  This is in contrast to other tumor-derived cell lines and non-transformed cells where only 30-60% of cells stain positively for cyclin E.  The normal cell cycle pattern of cyclin E expression was restored when an hCDC4 cDNA was introduced by retroviral transduction.  Therefore, loss of Cdc4 function, presumably leading to loss of SCF^Cdc4 protein-ubiquitin ligase activity, results in deregulation of cyclin E relative to cell cycle progression.  This finding was confirmed in endometrial tumors.

In a tissue culture model, deregulation of cyclin E has been shown to promote genomic instability, including increased levels of chromosome loss and polyploidy(3).  In order to test whether cyclin E-mediated chromosome loss could indeed synergize with tumor suppressor mutations by accelerating LOH, transgenic mice were constructed that were heterozygously mutated at the p53 locus and that carried a hormone-responsive degradation-resistant cyclin E transgene expressed during pregnancy and lactation in the mammary epithelium.  While the cyclin E transgene and the heterozygous p53 mutation alone each conferred mammary tumorigenesis at a rate of 10% or less, approximately 60% of mice carrying both the p53 mutation and the cyclin E transgene developed mammary tumors.  This strong significant genetic interaction is consistent with the hypothesis that cyclin E-mediated chromosome instability causes increased levels of LOH at tumor suppressor loci. 

Cell lines deregulated for cyclin E expression exhibit several cell cycle perturbations that might account for the observed genomic instability.  Flow cytometric analysis reveals a shortened G1 phase consistent with cyclin E driving cells into S phase prematurely, and lengthened S and G2/M phases, respectively.     A hypothesis that could potentially link cyclin E deregulation to impairment of DNA replication is that inappropriate cyclin E/Cdk2 activity at the M/G1 boundary could interfere with pre-replication complex assembly.   In telophase cells analyzed for pre-replication complex status, all proteins were loaded onto chromatin equivalently in cyclin E-deregulated and control cells except for Mcm4, which was underrepresented on the chromatin of cyclin E-deregulated cells.  Therefore, it is likely that the inefficient S phase observed in response to cyclin E deregulation is a consequence of a reduced number of competent pre-replication complexes resulting from an inhibition of Mcm4 loading.   We speculate that occasional failure of checkpoints that monitor the completion of DNA replication will allow cells with incompletely replicated genomes to enter mitosis, leading to chromatid non-disjunction events and chromosome losses.  

Cyclin E deregulation also leads to an accumulation of G2 and/or M phase cells.   Comparing progression through mitotic phases in cyclin E-deregulated and control cells, cells take longer to traverse prometaphase and metaphase when cyclin E is deregulated.    At high levels of cyclin E expression, some cells appear to be blocked at prometaphase and/or metaphase.    We speculate that the increase in polyploidy observed when cyclin E is deregulated is a consequence of cells failing at the metaphase-anaphase transition and eventually exiting mitosis without division. 

REFERENCES

1. Strohmaier, H., Spruck, C. H., Kaiser, P., Won, K. A., Sangfelt, O. & Reed, S. I. (2001) Nature 413, 316-22.
2. Spruck, C. H., Strohmaier, H., Sangfelt, O., Muller, H. M., Hubalek, M., Muller-Holzner, E., Marth, C., Widschwendter, M. & Reed, S. I. (2002) Cancer Res 62, 4535-9.
3. Spruck, C. H., Won, K. A. & Reed, S. I. (1999) Nature 401, 297-300.