HOME | RESEARCH | PEOPLE | PUBLICATIONS | CONTACT | BOOKMARKS | FILESERVER

Intrinsically, many cellular processes, such as growth, differentiation, movement, senescence and death, are linked to the cell division cycle. Deregulations in cell cycle control could have detrimental consequences to the cells and lead to various human proliferative diseases, including cancer. The major goal of this laboratory is to combine genetic, cellular, biochemical and proteomic approaches to determine the mechanisms controlling the cell cycle in normal human cells, in both stem and somatic cells, and how this control is altered during tumorigenesis. Three major areas of our current research are described below.

1. Cancer and stem cell cycle control by CDK inhibitors

Eukaryotic cell cycle progression is primarily controlled by a family of protein kinases, known as cyclin-dependent kinases (CDKs), that consist of an activating cyclin subunit and a catalytic subunit CDK. The principle negative regulation of CDKs is provided by two families of CDK inhibitors, the p21 and INK4 families, which link diverse cellular processes to the central G1 cell cycle control. We are taking a genetic approach towards these issues by targeting specific CDK inhibitor genes in mice to determine their in vivo function. Our current research in this area is focused on two closely related questions: (1) Determining the function of INK4 genes in stem and progenitor cell cycle control and tumor suppression, and (2) Determining the transcriptional regulation of INK4 genes by different cell pathways. (Cell 71:505; Nature 366:701; Genes & Dev 8:2939; Genes & Dev 12:2899; Mol. Cell Bio.20:6147; Cancer Res. 67:4732; Cancer Cell 15:289).

(A) Combinatorial interactions of CDKs and cyclins promote cells progressing through different phases of the cell cycle. INK4 and p21 families of CDK inhibitors negatively regulate CDKs. (B) Mice lacking an INK4 gene, p18INK4c, exhibit a series of growth defects including gigantism as shown here resulting in part from increased proliferation of stem and progenitor cells in many tissues and organs. (C) Identification of a putative lung bronchioalveolar stem cell (indicated by a yellow arrow, upper panel), a mammary stem cell (yellow arrow, lower pane) and luminal progenitor cells (green arrows, lower panel) in p18-deficient mice, which develop lung and mammary tumors.

2. Control of p53 ubiquitylation 

p53 mediates multiple checkpoints in response to a range of cell stresses by causing either cell cycle arrest or apoptosis, and is the most frequently mutated gene in human cancer. Elucidating p53-mediated checkpoint pathways—how a specific cell stress signal is detected and transduced to p53—not only helps to understand tumor development, but also to identify potential targets for therapeutic intervention. Our current research in this area is focused on two specific questions. (1) How do oncogenic signals derepress epigenetically silenced p15-ARF-p16 tumor suppressor gene clusters, leading to the activation of both p53 as well as Rb tumor suppression pathways? (2) How is p53 regulated by CUL7 and CUL9 E3 ligases? (Cell 92:725; Science 292:1910; Genes & Dev 21:49).


(A) Human chromosome 9p21 encodes three tumor suppressor genes and is frequently mutated in tumors. INK4A and INK4B inhibit CDK4 and CDK6 to retain the growth suppressive activity of RB and ARF, translated in an alterative reading frame of INK4A, inhibits MDM2 and thereby activates p53. The INK4A-ARF-INK4B gene cluster is epigetically silenced by polycomb and activated by oncogenic insults or during aging. (B) p53 is activated by a variety of cellular stresses such as DNA damage or oncogenic insults and is inhibited during normal cell growth by MDM2-promoted ubiquitylation and cytoplasmic degradation.

3. Cullin-RING family E3 ubiquitin ligases

Most, if not all, cellular processes, including notably cell cycle control and p53 control, are regulated by ubiquitin-mediated modification and degradation. The mechanisms targeting specific proteins for ubiquitylation, in most cases, are poorly understood. We discovered two novel RING finger proteins, ROC1 and ROC2, which constitute active ubiquitin ligases with members of the cullin family. We also discovered that Cullins 3 and 4 could assemble in vivo as many as 200 and 100 distinct E3 ubiquitin ligases, respectively. Our current research in this area is focused on two issues. (1) Developing a strategy to systematically identify the substrates of the cullin-RING E3 ligases, and (2) Elucidating the mechanism of CUL4 E3 ligases in control of gene expression and chromatin structure. (Mol. Cell 3:535; Mol. Cell 10:1511; Nat. Cell Biol. 5:1001, Nat Cell Biol 6:1003; Genes Dev. 20:2949; Genes Dev. 22:866; Science 324:261).


(A, B) Cullins bind to a RING finger protein, ROC1 or ROC2, which brings in and activates E2 ubiquitin conjugating enzymes. CUL3 binds with BTB domain and CUL4 interacts with WD40 repeats via DDB1. Various BTB and WD40 containing proteins recruit different substrates for the ubiquitylation by CUL3-ROC and CUL4-ROC E3 ligases. Mammalian cells express 200 BTB and 300 WD40 proteins. (C) CUL7 and CUL9 both localize in the cytoplasm, and bind with ROC and p53. The functions and mechanisms of both CUL7 and CUL9 in the cell cycle and p53 control are not clear.