The Role of c-Jun as a Prospective Modifier In the Regulation of Intestinal Tumorigenesis

Phosphorylation-dependent interaction between c-Jun and TCF4 regulates intestinal tumorigenesis

A review by Anna Wasiuk for Oncogenomics with Charles Brenner PhD

February 20th 2006

* Online subscriptions required to view linked figures

Introduction

The transcription factor AP-1 (activator protein-1) consist of homodimers and heterodimers of the basic region-leucine zipper (bZIP) proteins that belong to the Jun (c-Jun, v-Jun, JunB, JunD), Fos (c-Fos, v-Fos, FosB, Fra1, Fra2) and the related activating transcription factor (ATF2, ATF3/LRF1, B-ATF) subfamilies. The activity of AP-1 is induced by a multitude of stimuli and environmental insults and regulates a considerable number of cellular processes such as cell proliferation, death, survival and differentiation. The appropriate composition of subunits in the AP-1 dimer is determined by the nature of the extracellular stimulus and by the MAPK signaling pathway that is consequently activated (1)(Fig 1). The stimulation of AP-1 function is mediated by the amino-terminal phosphorylation of c-Jun by the c-Jun N-terminal kinases (JNKs) and phosphorylated c-Jun is biologically more active, to some extent, due to the ability to interact with binding partners (2). The paper currently under review investigate the mechanism of c-Jun N-terminal phosphorylation-mediated transcriptional regulation. In doing so Nateri et al. (3) used a yeast genetic screen to identify proteins which interact with c-Jun in a phosphorylation-dependent manner. One of the candidate genes identified by the genetic screens encoded the HMG-box transcription factor TCF4. Upon stimulation, regulation of AP-1 activity occurs by the activation of the appropriate genes as well as by the expression of the c-jun gene itself. The crucial residues involved in the phosphorylation-dependent activation of c-jun are the serine residues at position 63 and 73 and the threonine residues at position 91 and 94 (3).

Figure 1: The Activation of AP-1

TCF4 and WNT Signaling

T-cell factor/lymphoid enhancing factor (TCF/LEF) is a family of transcription factors functioning down-stream of the WNT pathway, a growth stimulating and inter-cell developmental signaling pathway. Unstimulated cell, i.e. without WNT signal, regulate β-catenin levels by the assembly of the so called destruction complex consisting of the adenomatous polyposis coli (APC) tumour suppressor protein, axin and the glycogen synthase kinase (GSK-3). GSK-3 phosphorylates β-catenin and by doing so marks it for ubiquitination and degradation. In the presence of a WNT signal the WNT ligand binds to its receptor Frizzled and a cascade of events, called the canonical WNT signaling, destabilizes the degradation of the complex and allows dephosphorylated β-catenin to build up and translocate to the nucleus. In the nucleus β-catenin functions as a cofactor for the transcription factors of the TCF/LEF family (3). For a more indepth look at WNT signaling and the players involved please visit The Wnt Homepage. WNT signaling has also been shown to activate non-canonical, β-catenin-independent pathways. Some of the identified non-canonical WNT signal transduction pathways include signaling through calcium flux and activation of the JNKs (4).

The Apc^Min mouse model of intestinal cancer

In humans, germline mutations of the APC gene results in familial adenomatous polyposis (FAP) as well as in other tumour syndromes characterized by multiple intestinal adenomas which almost always transform into cancer (3). Additionally, not only the tumors from FAP patients, but also most sporadic colorectal tumors have both APC alleles inactivated. The Min (multiple intestinal neoplasia) mouse model was created by random mutagenesis using ethylnitrosourea (ENU) and is a used extensively as a model for human colon cancer. The mutant carries a nonsense mutation at codon 850 of the Apc gene which leads to a truncated Apc polypeptide. This truncation leads to the development of more than 100 intestinal tumours per animal, mainly located in the upper GI tract, modeling the human colonic adenomas. The transition from benign polyp to maglignant tumour is generally characterized by the loss of the second copy of functionl APC. The Apc^Min model has played an important role in the identification of multiple tumor modifier loci such as Mom-1 and Mom-2. Both genotypic and phenotypic similarities between the Min mouse and FAP patients suggest that the Apc^Min/+ mouse can be used as an critical model for the human disease (5). The Apc^Min mouse and other related APC mouse strains are reviewed extensively by Fodde et al. (5).

Implications

The Apc^Min/+ mouse model has been used, as mentioned, to identify possible modifier genes which can alter tumor progression. One way to indentify such modifier genes is to mate mice carrying an APC mutation with mice carrying a mutation in the gene of interest. Such a cross is a hypothesis driven approach where the gene of interest is thought to be involved in the progression of the tumor. There is no way to know if the gene of interest will act as a modifier, but logical selections can be made based upon temporal and spatial expression patterns of the gene and its involvement in biochemical and physiological pathways.

In the present study by Nateri et al.(3) the authors will utilize the Apc^Min mouse model to investigate defects in c-Jun N-terminal phosphorylation or gut-specific conditional c-jun inactivation in the context of intestinal tumorigenesis. This approach was taken following the observation that c-Jun interacts with TCF4 and from the identification of non-canonical WNT signal transducers, such as the activation of the JNK MAP kinases.

Interaction of phosphorylated c-Jun with TCF4 regulates intestinal cancer development

Abdolrahman S. Nateri, Bradley Spencer-Dene, & Axel Behrens

Nature, Vol 437, 8 September 2005; pp281-285

JNK-dependent interaction between c-Jun and TCF regulates the c-jun promoter

As mentioned above one of the candidate proteins from the yeast genetic screen, which preferentially interacted with N-terminally phosphorylated c-Jun encoded the HMG-box transcription factor TCF4. Additional experiments confirmed that TCF4 interacts with phosphorlyated c-Jun using human embryonic kidney (HEK293T) cells through immunoprecipitation (IP) assays (See Figure 1a,b). In addition, further mapping showed that the required residues for this interaction are the serines at position 63 and 73, but not the threonines at posistion 91 and 93 (See Supplementary Figure 1). TCF mutants lacking the N-terminal β-catenin binding domain preserve the c-jun interaction which suggests that TCF contains two separate bind domains for the interaction with c-jun and β-catenin, further, siRNA-mediated knockdown experiments of TCF4 indicate that TCF4 is able to interact with both c-Jun and β-catenin simultaneously (See Figure 1c,d).

To examine the biological function of the interaction between c-Jun and TCF4 the regulation of the c-jun promoter was studied. The c-jun promoter contains two proximal c-Jun binding sites (jun1 and jun2) for autoregulation of transcription and a TCF consensus binding site located about 3 kb upstream of the jun sites in the human genome. In Figure 2 and Supplementary Figure 2 the authors demonstrate that; 1. JNK inhibition decreases the ability of TCF4 to bind its cognate site in the c-jun promoter, 2. that c-jun requires the presence of its amino-terminal phosphorylation sites for the binding to the TCF and jun upstream promoter regions and 3. that JNK inhibition also reduces the binding of β-catenin to the TCF-site on the c-jun promoter.It was hence concluded that c-Jun and TCF4/β-catenin interact on the c-jun promoter in vivo and in a JNK-dependent manner. As shown through TCF4-mediated c-jun-luciferase activity the TCF site was shown to be absolutely required and mutations in the TCF site reduced but did not completely abolish c-jun-mediated reporter gene activation. Hence the authors conclude that the TCF and the jun sites are both required in cis for the cooperative transcriptional activation by c-Jun and TCF4/β-catenin (See Figure 2 a-f).

Absence of c-Jun N-terminal phosphorylation attenuates intestinal cancer development in Apc^Min mice

In order to determine the physiological relevance of the c-Jun-TCF4 interaction the authors crossed the Apc^Min/- mice to JunAA mice. The JunAA mouse lacks both of the N-terminal JNK phosphorylation serines at position 63 and 73 required for TCF4 binding. Offspring was monitored for survival as well as number of adenomas. Groups of control mice, Apc^Min/+;c-jun^+/+ and Apc^Min/+;JunAA/+ died due to intestinal cancer at the latest after 19 weeks, where the average survival was 15.7 weeks for c-jun^+/+ mice and 16.1 weeks for JunAA/+ mice. The maximal survival time for homozygous Apc^Min/+;JunAA/JunAA mice was exteneded significantly, with an average lifespan of 23.1 weeks (See Figure 3a). Quantitative analysis of the number of adenomas throughout the intestine showed a moderate reduction in numbers in the experimental group as compared to the control groups. The average size of the adenomas, however, was significantly reduced in the experimental mice and the authors therefore conclude that the JunAA mutation greatly reduces the tumour burden of the Apc^Min/+ mice (See Figure 3b-g). Experiments were consequently performed with the intent to determine if the reduced tumour burden is due to an increase in cell death or a decrease in the proliferation of cells as compared to control mice. As decreased proliferation index of the Apc^Min/+;JunAA/JunAA adenoma cells and a comparable amount of cell death in all groups suggests that the main reason for the decrease in tumour size in the homozygous experimental group is reduced proliferation of the adenoma cells. Additionally, parallel with the data from the c-jun promoter analysis (Figure 2), c-jun mRNA and c-Jun protein levels were reduced in the adenoma cells lacking the N-terminal phosphorylation sites. However, the presence of total phosphorylated JNK protein levels were unchanged, as was the expression of c-Jun-independent TCF4 target levels (See Figure 3h-k).

c-Jun is required for intestinal cancer development in Apc^Min/+ mice

To prove a direct relevance of c-jun as a target of the c-Jun-TCF4 interaction during the development of intestinal cancer a floxed allele of c-jun was used. The floxed allele was inactivated specifically in the gut using a villin-cre transgenic mice. This system allowed the authors to eliminate the action of c-Jun in the animals post-development as well as in only the isolated region of interest. The absence of c-jun greatly affected the lifespan of the Apc^Min/+ mice as the Apc^Min/+;c-jun^dG animals showed no clinical signs of cancer development at 9 months of age. At week 18 Apc^Min/+;c-jun^fl/fl control mice were at the end stage of disease with multiple large adenomas, whereas the Apc^Min/+;c-jun^dG animals were characterized only by the presence of cystic structures. Such cysts were not present in WT or c-jun^dG mice, suggesting that they are caused by loss of Apc function, and that β-catenin signaling is unable to trigger tumour development in the absence of c-Jun. This follows the observation that Apc^Min/+;c-jun^fl/fl tumours show large accumulation of β-catenin in the nucleus and the same accumulation can be seen around the cysts in the c-jun^dG animals (See Figure 4a-k).

Summary

Taken together the data presented by Nateri et al. suggest that the molecular mechanism integrating signals for the activation of both the TCF/β-catenin pathway and the JNK pathway may be the phosphorylation-dependent c-Jun-TCF4 interaction. Phospho-c-Jun-TCF4 interaction may stimulate transcription by the recruitment of β-catenin to the proximal jun-sites close to the transcriptional start site (See Figure 4l). The ability of homozygous JunAA mice to develop normally and have a normal lifespan suggests that inhibition of the phospho-c-Jun-TCF4 interaction may be used as a target for therapeutic treatments in the progression or development of human intestinal cancer without substantial side-effects. Furthermore, at this time small molecule and peptide-based JNK inhibitors have been described and, as the authors suggest, it may be interesting to look at what potential beneficial therapeutic effects they might have on intestinal cancer progression (3).

References

1. Shaulian E, Karin M. 2002. AP-1 as a regulator of cell life and death. Nature Cell Bio. 4:E131-E136.
2. Davis RJ. 2000. Signal transduction by the JNK group of of MAP kinases. Cell. 103:239-252.
3. Nateri SA, Spencer-Dene B, Behrens A. 2005. Interaction of phosphorylation c-Jun with TCF4 regulates intestinal cancer development. Nature. 437:281-285.
4. Veeman MT, Axelrod JD, Monn RT. 2003. A second canon. Functions and mechanisms of β-catenin-independent Wnt signaling. Dev. Cell. 5:367-377.
5. Fodde R, Smits R. 2001. Disease model: familial adenomatous polyposis. Trends Mol. Med. 7(8):369-373.