David H. Berger, MD: The Plasmin/Plasminogen System in Colon Tumorigenesis

What I'm going to do is try to tell a little story about metastasis and specifically proteases in the development of colon tumorigenesis.

[ Slide 01 ] The Metastatic Process

As we know, the main problem with cancer is that the primary tumor does not necessarily go on to cause mortality.

However, it's the development of invasive tumor cells which gain access to the blood system and cause metastasis in distant organs that really is responsible for the death of most patients with malignancy.

[ Slide 02 ] Plasmin/Plasminogen System

My lab has been focused on one protease system, the plasmin-plasminogen system, which is very important in the process of cellular invasion, cellular adhesion and angiogenesis. There are several components to the plasmin-plasminogen system and they consist of the urokinase plasminogen activator, which is secreted initially as a pro-uPA or an inactive molecule, which then binds to its receptor, the urokinase plasminogen activator receptor, which is a self-service receptor where it is activated. As with all protease systems, there is a natural inhibitor and actually there are two, and they are known as plasminogen activator inhibitors. And the most important one, really, in epithelial malignancies is plasminogen activator inhibitor type 1, or PAI-1. And what uPA does once it's activated is that it takes plasminogen, which is a very abundant protein, converts it into the active protein plasmin which then can enzymatically degrade collagen, which is the main component of the basement membrane. It also can go on to activate other proteases, importantly the metalloproteases and also plasmin can be degraded down to the active molecule angiostatin, which if you follow the angiogenesis literature is a very important molecule in inhibiting distant metastasis and may be a therapeutic target. So this is a very important system in the development of metastasis by several processes influencing adhesion, invasion and angiogenesis.

[ Slide 03 ] Survival by uPA Status

Well, what is the role of uPA and its receptor in colon cancer? And first of all, studies have shown that the uPA is a prognostic indicator in colon cancer. If you look at all patients with colon cancer and divide them up into patients who have low uPA levels in their tumor versus high uPA levels in their tumor, there's a clear decrease in survival as the amount of uPA increases.

[ Slide 04 ] Survival by uPAR Status

The same thing can be said about the urokinase plasminogen activator receptor. This slide, taken from the JNCI this year, if you look at this slide, the straight line is all Dukes B patients. However if you then divide the Dukes B patients into low uPAR expressers vs. high uPAR expressers, you can see there's an obvious decrease in survival as the expression of uPAR goes up.

[ Slide 05 ] Adenoma-Carcinoma Sequence

Well, the question comes down is when and where is uPA and uPAR expressed in terms of colon tumorigenesis? And as Dan Beauchamp went through nicely, the development from normal colonic epithelium to adenoma and carcinoma, where does uPA and uPAR get up regulated and really what is regulating these proteins?

[ Slide 06 ] beta-Catenin Regulates uPAR - 1

There's an interesting article that appeared in PNAS this past year that showed beta-catenin, which is a protein that is regulated by APC, which is a very early change in colon tumorigenesis, controls uPAR expression. Here you can see directly, this is uPAR mRNA, and as beta-catenin levels go up, uPAR expression, uPAR mRNA goes up from the normal to the tumor to metastasis. And as well if you look at the protein level, there is increased level of uPAR with progression of colon tumorigenesis.

[ Slide 07 ] beta-Catenin Regulates uPAR - 2

And to show that beta-catenin is actually responsible for this increase in uPAR, a cell line without any beta-catenin was transfected with a stable uPAR expression vector and as you can see with beta-catenin expression, uPAR is up regulated. So uPAR is probably expressed very early in colon cancer and it's directly related to loss of APC function.

[ Slide 08 ] uPA Levels During Colon Tumorigenesis

The same thing with uPA. When does uPA begin to be expressed? And as you can see, in normal colonic tissues there is very low levels of uPA protein, however, once you begin to see adenomas or transformation there is an increase in uPA and once you get to cancer there is a further up regulation. But between normal and adenoma that transformation steps. Very early on, uPA is up regulated.

[ Slide 09 ] Methods

Now the question comes down to how is uPA up regulated? Dan Beauchamp went through very well the role of TGF-beta in this process. And we decided because of TGF beta and uPA have very similar expression patterns in colon cancer and the fact that we know that TGF-beta is also associated with poor prognosis in colon cancer we went on to try to determine if TGF-beta was one of the more important mechanisms of regulating uPA expression. What we did is we used the RIE-1 cells and the RIE-Tr cells which are a transformed Rat Intestinal Epithelial cell line that was transformed in Dr. Beauchamp's lab. How they developed these cells was by continuous exposure of the RIE-1 cells to TGF-beta and what they did was they showed that these cells then became tumorigenic with a 95 percent decrease in the TGF-beta Type 2 receptor.

[ Slide 10 ] RIE-TR Cells are Transformed

This is just proof that these cells are transformed. As you can see the already transformed cells proliferate at a much greater rate than the transformed [probably meant parental] cells and in culture the parental cells form monolayers while the transformed cells are all heaped up and if you look at soft agar assays these form colonies in soft agar and also tumors in nude mice. So these are truly transformed epithelial cells.

[ Slide 11 ] TGF-beta1 and COX-2 in RIE Cells

Again if you look at the expression of some of the important proteins that Dr. Beauchamp mentioned, in the transformed cells, there is increase in COX-2 and there is decrease in protein expression of the TGF-beta Type 2 receptor.

[ Slide 12 ] TGF-beta1 - Epithelial Cell Adhesion

As I mentioned, some of the important processes in intestinal metastasis are epithelial cell adhesion and as well, epithelial cell invasion. This is an adhesion assay and what you can see here is that the RIE-1 cells, the parental cells with treatment of TGF-beta, there is no change in adhesion to collagen Type 1, collagen type IV, fibronectin or laminin, which are important basement membrane proteins. However in the transformed cell lines, the RIE-Tr cells in the green and the yellow bar with treatment with TGF-beta, there is a significant down regulation of adhesion to collagen Type 1, type IV, fibronectin and laminin. If the cells are less adhesive, they're more likely to invade through the basement membrane.

[ Slide 13 ] TGF-beta1 - Epithelial Cell Invasion

If you look at invasion, and the effect of TGF-beta on rat intestinal epithelial invasion, as you can see, in the orange bar, the parental cell line, there is no effect with TGF- beta treatment, however, in the transformed cell line, with increasing concentrations of TGF-beta, there is increasing tumor cell invasion.

[ Slide 14 ] TGF-beta1 - Epithelial Cell uPA Expression

We then went on to try to determine what was responsible for this effect on TGF-beta mediated tumor cell invasion and what we first did was looked at the plasmin-plasminogen system and showed that in the RIE-Tr cells, the transformed cells, there was in increase in plasmin activity vs. the parental cell lines in response to TGF-beta. And we looked at the various components of the plasmin-plasminogen system. What we see here is that the transformed cells with TGF-beta there is a marked up regulation in uPA which is not seen in the RIE-1 cells either at baseline or with the addition of TGF-beta.

[ Slide 15 ] TGF-beta1 Effect on uPA - Time Dependant

And this effect of TGF-beta on uPA was shown to be both time-dependent with expression beginning at six hours and also dose-dependent. And these doses correlated to maximal invasion seen in the invasion assays.

[ Slide 16 ] TGF-beta1 - uPAR and PAI-1 Expression

We then looked at uPAR and PAI-1 expression and interestingly there was no PAI-1 expressed in either the transformed or the normal cell lines with or without TGF-beta. There was uPAR expressed, which suggests that uPAR is an even earlier event, I mean that up regulation of uPAR is even an earlier event in the cells than is up regulation of uPA.

[ Slide 17 ] Antibodies - Cell Adhesion

We then, in order to make sure that the TGF-beta mediated cell adhesion and invasion was actually caused by the up regulation of uPA by TGF-beta, we used amino neutralization assays, where we used anti-uPA antibodies and as you can see this is the adhesion of the cells at baseline with TGF-beta and then with increasing doses of the anti-uPA antibodies we restored the adhesion back to baseline, which suggested that the effect of TGF-beta on adhesion was mediated by uPA.

[ Slide 18 ] Antibodies - Cell Invasion

Same thing with invasion. If you look at baseline, the effect of TGF-beta on the transformed cells in terms of invasion, marked up regulation in invasion and this was inhibited back to baseline with anti-uPA antibodies.

[ Slide 19 ] uPA Silencer has SMAD4 Binding Domains

Well the question is, really, how is TGF-beta regulating the expression of uPA? And we have some data, which I because of time will not show, that suggests that the regulation of the uPA gene by TGF-beta is at the transcriptional level. We also have some preliminary data that suggests that SMAD4 may be important and Dr. Ko will later on talk more about SMADs and their role in TGF-beta signaling.

We have identified this region of the uPA regulatory region that is known to be a silencer. And in normal cells, as you see, uPA is usually down regulated by TGF-beta however, in transformed cells it's up regulated, so we believe that normal TGF-beta by binding to this region in the uPA gene known as the silencer, possibly through SMAD4 inhibits uPA. However with transformation and loss of normal TGF-beta signaling, the silencer region is released. These are actually two known SMAD4 binding regions. This is a known TGF-beta response element within the uPA silencer. So this is the region we're beginning to look at in terms of regulation of uPA by TGF-beta.

[ Slide 20 ] Plasmin/Plasminogen System in Colon Tumorigenesis

Well this is a little cartoon of what we think the role of the plasmin-plasminogen system in colon tumorigenesis. As I mentioned, APC normally binds to beta-catenin, which causes degradation of beta-catenin. When you lose APC, which occurs very early in colon tumorigenesis, beta-catenin can then activate the transcription factor Tcf4 which activates uPAR. On the other side, with TGF-beta, TGF-beta normally binds to the Type 2 receptor, phosphorylates the type 1 receptor and then through SMAD2, SMAD3, and SMAD4 there is negative control of transcription of uPA. However, when you lose TGF-beta, normal TGF-beta signaling with transformation, there is loss of this negative influence of SMAD4 and other known TGF-beta signaling pathways such as Ras and NFKappa B lead to transcription of uPA, release of the silencer by SMAD4. You then get uPA secretion, which combines to the urokinase plasminogen activator receptor and lead to tumor invasion and metastasis.

[ Slide 21 ] Acknowledgements

I don't have a slide thanking/acknowledging my collaborators, but I would first of all like to thank Dr. Beauchamp and his lab as well as Dr. Chris O. Mahoney who is a resident in my lab who did a lot of this work. Thank you.