in Lethal Metastatic Prostate Cancer

Vasan Yegnasubramanian, MD, PhD; The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University

Our major goal in our research is to try and understand how genetic and epigenetic processes

can cooperate in making a cancer cell be very aggressive or even get the cancer cell started.

The DNA in a cell is what tells a cell what to do and each cell in our body has the same

genetic code but they actually behave very differently. And so there's actually a level

of coding beyond the genetic sequence which is called epigenetics--beyond genetics -- that

actually helps those cells figure out what to do and when in order to behave the way

they should so that a brain cell doesn't behave like a liver cell, so on and so forth.

In cancer cells we know very well that they can change their genetic sequence so that

they become different and actually can escape the rules that normal cells live by. They

also change their epigenetic code, so they change the interpretation of that genetic

sequence so that they can actually do many new functions that they weren't doing previously.

One of the main questions in epigenetics is that the process is very plastic so a given

cell can actually change the way it behaves by changing its own epigenetic behavior. And

this can happen in even normal cells. And so a lot of people question whether these

epigenetic processes can really be selected for in a cancer cell and really stay stable

enough to drive the cancer cell to become initiated and then progress to very aggressive

forms. And that remains sort of an open question. And this research really tried to address

that in a very serious way.

What we found that was quite surprisingly actually is that the epigenetic changes were

almost as closely maintained as the genetic changes. And this is actually something that

I think will challenge a lot of notions that are out there that--that only genetic changes

are very stable and that epigenetic changes are too dynamic to really be selected for

in this way. This work really shows that the epigenetic changes can really also contribute

in the same way that genetic changes can. There's an implication for that. Mutations

that change a sequence in the DNA, we're actually finding there may not be as many

recurrent changes. Any given change is not so frequent in all different prostate cancers.

But interestingly, some of these epigenetic changes like DNA methylation, they can occur

in the vast majority of prostate cancers, the same exact change present over and over

again in many different people with prostate cancer. That's the making of an ideal biomarker.

Right; so you can actually use the same exact test on many different people and tell whether

they have cancer or not and in this case we're also finding markers that look like they might

be better for identifying those with aggressive cancer or not, in a way that you don't have

to have an individual biomarker for each person.

if you think about the genetic sequence in the cells, it's actually a vast volume.

So if you think about it as a book of letters, A, G, C, and T, it's six billion letters

long in a single cell. And that cell has to figure out which parts of those letters it

needs to use at a given time, and so one way it does that is it actually puts sticky notes

and little marks and bookmarks and--and little tabs on those pages so that it actually knows

to do something at a given time and--and sort of indexes it in that way. And those little

tabs and sticky notes we call epigenetic processes. And one such process is called DNA methylation;

it's sort of a sticky note which I put on as these little lollipops in this figure that

basically can mark these little fragments of DNA and we know that cancer cells can move

these sticky notes in ways that they shouldn't and put them in new places and erase them

from old places. And our goal is to try and study both the genetic sequence, but also

where these sticky notes were placed, these DNA methylation sticky notes were placed on

these DNA fragments.

And we had a technology that would help us do that and that technology is based on a

protein that actually exists in our cells called methyl-binding domain protein. And

this protein is actually very good at finding those sticky notes and binding to them. And

so we can use that protein to capture all the places in the genome where these DNAs

were marked by these DNA methylation marks and in a similar process just take all the

DNA, and process it, and analyze it by some of the latest innovations in micro-array technology

that allows us to interrogate the entire genome all at once.

Any time you look at the entire genome at once it's a real challenge because we have

six billion base pairs; it's this enormous space of information. And to really coalesce

it in a way that we can actually understand and sort of think about it is difficult. But

we finally think we have a nice way of really looking at all of the data in a way that can

be informative to us. And we call these the DNA methylation cityscapes or epigenetic cityscapes

of lethal prostate cancer. Each of these little blocks represent where a single chromosome

was sort of arranged into a neighborhood from the genome. So this red block here represents

a neighborhood of chromosome one; this one represents a neighborhood of chromosome two,

three, four, five, so on and so forth.

And overlaid on top of that is our actual cityscape where the data was plotted and what

we see is basically that some regions in the genome are really like skyscrapers, if it's

really tall on this cityscape is that it's very frequently changed in prostate cancer

compared to normal tissues.

Not only do we see the height of the buildings, we also see the color of these little buildings.

And we see that there's lots of buildings that are red and // those are clearly important.

They're things that occur very frequently in people and they're maintained in all

of the metastases from those people. And so there's genes like GSTP1 and APC that are

actually well known to be changed in prostate cancer that we see but we also see dozens

of other skyscrapers that are red that are giving us new biomarkers and new avenues to

actually pursue.

But what's interesting to me actually is that typically we would only use frequency

meaning how many people have a change as a measure of how important something is. But

that's not always the best measure because in some cases you might have something that's

not very frequent but for that one person it was very important. And we have examples

of that as well, like for instance, this MLH1 gene and the ESR1 gene, the arrows are pointing

to little huts, and those little huts are actually red and what that means is that very

few people had that change but when they did have that change every metastases from that

person had that change. That suggested that those changes really needed to be made upfront

in order for that to become a metastasis in that person. The fact that these changes were

maintained in every metastases suggested that it was really a driver of establishing metastatic

disease, and so we actually find numerous red huts across the cityscape as well.

Things that we may not be very interested in following up on any further are huts that

are white where there's really no maintenance across the individuals and across the metastases

within an individual but there's some small frequency or even lightly colored skyscrapers

like this one for instance where it really doesn't seem to be maintained across all

the mets from a given person.

And so we're actually following up on these types of findings further in another set of

human clinical specimens. This one is actually going to be a much larger set; it includes

550 cases and controls of people that had very highly aggressive cancer versus not as

aggressive cancer in different ways, so either by pathological definitions or by functional

definitions like whether they recurred later on after therapy or not and--and we're wondering

whether the changes that we're finding in the lethal metastatic cancers from these individuals

are actually going to help us identify biomarkers and that's ongoing work that--that we'll

be hopefully finishing up in the next few months.