Dr. Frank Slack and Dr. Joanne Weidhaas, Cancer
Genomics
July 18, 2010
Welcome to Yale Cancer Center Answers with Dr. Ed Chu and Dr. Francine Foss, I am Bruce Barber. Dr. Chu is Deputy Director and Chief of Medical Oncology at Yale Cancer Center and Dr. Foss is a Professor of Medical Oncology and Dermatology specializing in the treatment of lymphomas. If you would like to join the conversation, you can contact the doctors directly. The address is canceranswers@yale.edu and the phone number is 1888-234-4YCC. This evening Francine is joined by Dr. Frank Slack and Dr. Joanne Weidhaas for a conversation about cancer genomics. Dr. Slack is a Professor of Molecular, Cellular, and Developmental Biology and Dr.Weidhass is an Assistant Professor of Therapeutic Radiology. Here is Francine Foss.
Foss
Could we start off by asking you both for a basic description of
what the cancer genome is?
Slack
It's a very exciting year as it's actually the 10th anniversary of
the elucidation of a human genome. A human genome is
essentially the complete set of letters in our DNA that codes for
all of the things that make us human. It also includes the
genes and pieces of DNA between the genes. We now have the
ability to actually figure out the sequence of letters in every
single human being and we can find in that sequence certain
anomalous DNA changes, which we thought of as mutations, that have
occurred that are often found in genes that are important for cell
cycle control and cell mobility control, all things that are
important for normal bodies, but when mutated, can lead to
cancer. It's a road map that let's determine whether somebody
has mutations in the DNA that might lead to cancer or may have
caused the cancer.
Foss
So, basically DNA is kind of like the reading code. Joanne,
can you talk a little bit about how we actually get from DNA to the
level of proteins actually in the cell? What happens between the
DNA and the protein that exerts an affect in a cell?
Weidhass
I tell people that every cell has the same DNA, it's kind of like
the text book and if it's a liver cell, it reads a chapter on being
a liver cell. It will take the information from the DNA and
basically build the building blocks of the cell to be the function
of that cell. That's a very basic understanding of how we
have understood DNA, how it turns ultimately into RNA, and then
protein, although there is really a whole new level of
understanding now with the discovery of microRNAs.
Foss
Frank, you mentioned the importance of the Human Genome Project and
studying these genes to understand cancer because of these
mutations. How frequently do these mutations occur for a normal
DNA, and how often do people get cancer if they develop a mutation?
I know that's a bit of a tough question.
Slack
I will answer the first part. We are continually being
bombarded with things that damage our DNA. I mean even if you
lived in a cave, in a hermetically-sealed bubble, you would still
be damaging your DNA in some way or another; there are just natural
occurrences, every time
3:23 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
a cell divides it has to replicate its DNA and there are just
natural errors that occur in that process, but then we actually
enhance that by exposing ourselves to environmental mutagens and
carcinogens. These are compounds like UV light, cigarette
smoke, or automobile exhaust that do damage DNA and when that
occurs it occurs randomly, pretty much, throughout the genome, but
if you get unlucky enough that it hits one of these important
cancer genes that can contribute to a cancer, maybe Joanne can
answer how frequent that is in the population.
Weidhass
That's a great question. I think that everyone's damaged all the
time and what we are starting to realize is that your risk for
developing cancer may have something to do with your ability to
repair that damage well. As an oncologist, I see patient's
and everyone always wants to know, why me, why me? There are some
very healthy people with cancer that have just done everything
perfectly and we are starting to find some information that there
could be things that just tip the scale a little bit where things
might seem to be fine, you are a perfectly healthy individual,
however, you may not do quite as well of a job in repairing your
DNA. Some of the findings that Frank and I have found
together have shed some new light on this. People have been
interested in this for years, but there are really very few
currently recognized syndromes where people get cancer and I would
postulate that this is much more common.
Slack
It's been predicted that about one in two males will get cancer in
the United States, so that gives you a sense of how common it is
right now, and one in three females.
Foss
And that includes all different types of cancers, like somebody's
benign skin cancer that usually don't cause a problem. We
have talked on other shows about various cancer genes and family
syndromes, but Joanne, you bring up an excellent point, which is
that most patient's who develop cancer don't have one of these
defined syndromes and that we're really just scratching the surface
in terms of understanding all the different mechanisms that lead to
cancer.
Weidhass
Exactly, and I think Frank can explain a little bit about how we
have shed some new light on this with his understanding of some new
biology. I will just say briefly that the syndromes we have
known of were built on the paradigm that all DNA turns into protein
and the areas of the genome that have been studied are those
areas. With new understanding we have looked in different
areas of the DNA and we found new markers of cancer risk.
Foss
In particular, you mentioned microRNAs. That seems to be a hot
topic; Frank could you talk a little bit about that?
Slack
MicroRNAs are actually made from our own genes. They are
very small RNAs, so as Joanne was mentioning earlier, most genes
are made into RNA and then RNA gets translated into protein.
But 15 years ago, it was discovered that there are certain kinds of
genes that make an RNA that is not
6:37 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
made into protein, in fact this RNA by itself has a function and
that function is to go and negatively regulate other genes, so
basically squelch the activity of other genes. About ten
years ago, it was discovered that humans have many of these
microRNA genes. We have about a thousand of these microRNA
genes, and then about seven years ago, it was discovered that these
little RNA genes get mis-regulated in cancer, so they get mutated
in cancer, they get deleted in cancer, they get up-regulated in
cancer, meaning you have too much of these products being made, and
we now have the ability to use these little RNAs, these little
microRNAs as very important diagnostics and prognostic tools and
ultimately we hope therapeutic tools in the fight against
cancer.
Foss
Why do you think it took us so long to find these microRNAs?
Slack
Well they have been referred to as the dark matter of the genome
because they are so tiny. People for many years have this
misconception that the genome was made up of about 25 thousand
protein coding genes and everything else was junk. That work
junk is fast finding itself on the junk heap. We now realize
that much of this information between the genes is actually being
used and in some cases being used to make these thousand microRNA
genes. They were just very hard to find because the genes are
so small and we did not really know they were going to exist like
that, but working model organisms really paved the way in this case
and proved that these little RNAs were in fact working and that
they were useful and that they could be damaged. And when
they were discovered in humans, it was not that big a leap to show
the same was true in humans.
Foss
Joanne, can you tell us how these microRNAs have been important in
specific cancers that we are studying now?
Weidhass
Frank was really a pioneer in the microRNA world, one of the first
to show that microRNAs were important. The first cancer where
it was really found was lung cancer and there has been a fair
amount of work on microRNAs in lung cancer. First are the
differences in microRNA levels, predicting response to treatment or
outcome in lung cancer. There has been extensive work now
showing microRNAs are abnormal in every cancer, and then Frank and
I really joined forces and were some of the first to show that
inherited changes in microRNAs and the regions where they bind and
exert their influence can be important markers in developing cancer
risk, really predictive markers, so that's kind of a new area for
microRNAs.
Foss
Are there multiple of these microRNAs, multiple families of
these?
Weidhass
There are probably about a thousand microRNAs known at this point,
many are grouped together, but we think even though they look
similar, they probably have unique functions and we are just
beginning to understand that.
9:49 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
Foss
You mentioned that these were important in a number of different
types of cancer, pretty much all different types of cancer.
Are there specific ones associated with specific tumor types?
Weidhass
Absolutely, it seems that each tissue really has its set of
microRNAs it uses and in a tumor from that tissue it will be that
set of microRNAs that are mis-regulated and important, some seem to
be universal though. For example Let-7 is a very commonly
mis-regulated microRNA in almost all forms of cancer, particularly
solid tumors, but it really creates this similarity across tumor
types. So inherited changes that might lead to one cancer
type, will now tie together cancer types you would not normally
group together, which is really changing how we as oncologists view
cancer. Right now it's separated really by the sections of
the body, for example, the thoracic cancers are in the chest.
There are certain groups based on where things are, but
biologically, they might be more similar, for example, a skin
cancer might be more similar to lung cancer that we didn't realize
and some it's just understanding the underlying biology, which
microRNAs let us do.
Foss
Frank, that sounds very interesting. Can you tell us how you
actually apply that in the lab, in your research? Do you look for
these common microRNAs among these different tumor types or are you
actually looking for unique ones, or both?
Slack
For example, in collaboration with Joanne, we will profile various
different types of cancer. So we will take the tumor tissue
from the wonderful donors that have donated their tumors here at
Yale, and we will ask whether those tumors have a common set of
microRNAs being made that, for example, are different from normal
tissue taken from tissue surrounding that tumor. And we are
trying to identify a signature for each type of tumor, and as
Joanne was saying, we find that in general, every tumor type has
its own signature and our goal is to build these signatures to the
point where a physician one day might be able to read out the
microRNA signature and know exactly what tumor type that patient
has, and potentially what the response to therapy might be for that
patient, and what the potential outcome for that patient would be.
We are building that area. The second area is we identify
these microRNAs that are mis-regulated in these cancer types.
In general, we find two different classes of microRNAs. We
find those that are lost, or down-regulated in the cancer, so
invariably almost all the patients in that sub-class of cancer have
a few of these microRNAs that have been lost. They also have
a few microRNAs that have been up-regulated and our goal then is to
try and replace the microRNAs that have been lost in those patients
and somehow knock out the microRNAs that up-regulated. We
first try and show that indeed in cell lines and tissue culture
work that if you give microRNAs back that have been lost, or if you
knock out microRNAs that are too expressed in those patients that
we see an affect on cells; usually the cells die or the cells stop
growing. Then we try the same experiments in the mouse model where
we put the human cells into a mouse and we can show that those
cells will now stop growing in the presence of the microRNAs, or we
actually build the mouse model that mimics the human cancer and we
show that those microRNAs are therapeutic in that mouse model with
the
13:29 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
ultimate goal of going to clinical trials at some point in the
next few years to actually deliver these agents as therapeutics,
natural therapeutics.
Foss
It sounds like now we are moving towards manipulating the level of
these microRNAs and you think that's actually going to get into
people from these animal models?
Weidhass
We certainly hope so. What's beautiful about microRNAs, because
they are a natural element, they are not toxic, in general. You
deliver them to the body, they're something the body has, but they
are not at regular levels in the tumor so there is a real potential
to use them. And it's not a chemical like much of the other
therapies that we use, and it would be very specific to the cancer
that has lost or has over expression of specific microRNAs that we
are targeting
Foss
That sounds really exciting and I would like to talk a little bit
more about that after the break. We are going to take a short
break for a medical minute now. Please stay tuned to learn
more about cancer genomics with our guests, Drs. Frank Slack and
Joanne Weidhaas.
Foss
Welcome back to Yale Cancer Center Answers. This is Dr.
Francine Foss and I am here today with my guests Dr. Slack and Dr.
Weidhaas to talk about cancer genomics. We talked before the
break about some exciting advances with using these microRNAs to
identify specific tumor types and also possibly to modulate those
to affect cancer. I would like to go back to the point about
using those as a signature for tumors. Frank, you had mentioned
that we can now identify patterns of microRNAs and I am wondering,
are those assays available yet in the clinic and will those be
applicable to fresh tissue from patients? Or will we be able to use
the archived tissue that we have in our tissue banks?
Slack
Those sorts of assays are already for sale and we expect to see an
explosion of those sorts of assays over the next few years.
You can already purchase an assay that will allow a physician
to
16:29 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
understand where a metastatic tumor arose from. In many
cases oncologists have a difficult time identifying exactly where
the primary tumor arose from in a metastatic patient, but because
these metastatic tumors retain that signature of the originating
tissue, it is possible just by profiling their microRNAs that we
can determine what the primary tissue is and then of course the
physician has an easier time getting and trying to identify where
the primary tumor is.
Foss
Joanne, you had mentioned that we are now understanding that some
of these tumors are more related than we think they are based on
the microRNA signature, so that certainly could change the way we
treat some of these tumors.
Weidhaas
Absolutely, and I think another thing that is tying it together
with microRNAs and understanding the inner relationship between
tumors is this new area where we, and others, have been looking at
inherited changes in your germline DNA, not just in the tumor, that
might predispose you to developing certain cancers and how they
might be cancers that we wouldn't necessarily group together, but
there might be risks for both of them.
Foss
This is really an important area that you bring up, actually
looking at the germline mutations of these microRNAs and how that
could potentially be a new marker for cancer risk.
Weidhaas
Absolutely, so that's an area we are very interested in. My
sense has been that after finding some of the big protein coding
sequence inherited mutations, people have somewhat given up on
finding other markers of inherited cancer risk, while I think that
is actually much more common then we realize. It's important
to help predict who is going to develop cancer and understanding
microRNAs and how important they are in cancer and as regulators,
it was very logical to start looking in those regions of your DNA
to see if there are changes you inherited from your parents, from
your family that disrupted microRNAs and how they worked that put
you at risk for cancers.
Foss
Previously on another show we talked about the SNP analysis,
genomic analysis to look at cancer risk. In those kinds of
analyses do they look at these microRNAs, or is this is a
completely different area?
Slack
There are a number of studies, even some done here at Yale Cancer
Center where SNP analysis has been performed in the microRNA genes
themselves, and also in the places where the microRNAs tend to
bind. Some very interesting results have already come from
that. Our work has been slightly different in that we have
looked for new types of genetic changes that aren't in the
traditional SNP databases with the idea that all SNPs have not been
discovered yet and that's one of the wonderful advantages of having
the human genome available to us, is that we can quickly compare
back to the reference human genome any changes that we find and ask
whether
19:31 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
they are truly different from undiagnosed people. And so,
yes, those SNP analysis have been useful but we think that they
have to be limited and in fact there is more out there that needs
to be discovered.
Foss
Can you talk a little bit, both of you, about this whole area of
cancer genomics and how much information is out there among all
these different sites that are doing these studies and how that all
gets integrated to help patients? I think, from a patient's
point of view, it is hard to really figure out how this all comes
together.
Weidhaas
That's a great question and it has been difficult. There have been
approaches with SNP analysis where they do genome-wide SNP
Association Studies and each marker is found to have just a little
bit increased risk, and how would you combine that into meaningful
information for a patient is I think undetermined. As Frank
mentioned, we have taken approaches to look a little more carefully
at certain regions of DNA and in doing so, we have found some
markers that are more predictive of cancer risks, stronger markers,
because they are functional. A lot of SNP work has looked at
markers that are not themselves functional, so they might just be
associated with something else and we think that is perhaps why
they are not quite as powerful. Some of the markers that we
have been studying are themselves the marker that predicts cancer
risk and they are stronger. I think there is a lot of
interest and it is really important how this is communicated to
people and how it is used and plugged into the existing clinical
paradigms. Our goal is to have the medical community weigh in
heavily on this and to do it very responsibly because we think some
of these markers are important and should really be plugged into
the medical paradigms and communicated appropriately to
patients. We are currently doing it from the top down
starting with the physician community and having it move through
that way.
Foss
Certainly, as these microRNAs are used now for diagnosis, that will
touch the patient in a more direct way.
Slack
Yes, there are many areas from just risk assessment, through
diagnosis, through prognosis, through ultimately treatment, where
microRNAs are probably going to play a role.
Foss
Could we switch gears a little bit now and talk about some other
areas of cancer genomics, specifically I wanted to ask you guys if
you could define an oncogene and tell us how an oncogene
contributes to cancer?
Slack
An oncogene is really just a mutated proto-oncogene. I will
explain what a proto-oncogene is. We actually need these
genes in our normal existence. It's when they get mutated in
such a way that they are not functioning correctly and can cause
cancer, then they are called oncogenes. Think about how the
single celled embryo needs to become the trillion cells that we
have in our body,
22:36 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
well when we are sitting in our mother's wombs, we actually use
these proto-oncogenes to drive lots of cell division. We grow
from a single cell to a trillion cells in about three months and we
actually need these proto-oncogenes. But as we get more
developed in the womb and our organs start to form and our limbs
start to form, these proto-oncogenes get switched off slowly and
cell division gets switched off to the point where when we are
adults, we have very little cell division going on in our
body. Only in certain places, like our skin and the lining of
our intestine where we actually damage a lot of cells and we
replace those cells on a frequent basis do we actually have lots of
cell division. So these proto-oncogenes are very important
for our development and for our normal existence. But when
they get damaged, they no longer regulate appropriately and they
are actually driving cell division inappropriately, then they
become oncogenes. Now some of the traditional oncogenes have
been considered to be the protein coding genes. Some famous
ones include genes like MIC and ras but we are discovering that in
fact microRNAs themselves can be oncogenes as well.
Foss
That's really interesting. These oncogenes again are normal
genes that have gone awry essentially?
Weidhaas
Exactly.
Slack
A bit of a misnomer, when the press refers to a person as having a
gene for cancer, in fact everybody has got that gene. It's
just that it's not mutated in everybody; it is mutated in the
cancer patients.
Foss
This mutation occurs at some point as the patient is basically
evolving, not before the patient was born?
Weidhaas
It depends, and that is a big controversy. Certainly there
can be acquired mutations during life that could put you at cancer
risk. We really think that there are inherited subtle changes
in the regulation of some of these oncogenes or proto-oncogenes
that somehow leads to an accumulation of damage or just a loss of
control at some time in life that would put you at increased cancer
risk.
Foss
That's an important point when you think about the issue of
identical twins who basically have the same genetics, at least to
start off with, and then one of them will go on to develop a cancer
where the other one won't.
Weidhaas
Cancer is very multi-factorial. That's the thing about all of these
even inherited risks, nothing is 100% and certainly there are the
effects of living life, even a clean good life, which might
interact
in a certain way with some of what you are born with and lead to
cancer for you and maybe not someone that is identical to you.
25: 20 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
Foss
We have talked a lot about the specific genetic issues in cancer
and I know both of you are working in the lab on a number of these
issues. Could you just tell us briefly what you are working on and
what you think the most exciting areas are in cancer genetics?
Frank, do you want to start?
Slack
I think if I had unlimited money, which is of course something
that all scientists would love to have, I would start very quickly
just getting the whole genome sequence of every single cancer
patient that Yale could get their hands on and I think that is
really the exciting part in genomics right now. It is trying
to determine the entire set of changes within each individual and
then trying to map those changes back to increased risks of various
diseases. At the moment it's pretty expensive to sequence an
entire person's genome, it costs about $20,000, but that cost is
declining and I think one of the really exciting areas in the next
few years is going to be the fact that a lot of human genome
sequencing is going to be done. In fact, there will be a time
when almost everybody will have the genome sequence done just as
part of their medical record and hopefully there will be a time
when we can read out from that sequence of DNA exactly what risks
these people will have coming down the road. I would really
like to be moving in that direction if I could, but at this point,
it is a little bit too expensive. In lieu of that we have
taken the approach of trying to look at more limited parts of the
human genome, the parts that we think contain some of that
regulatory information where the microRNAs bind and in the microRNA
gene. We are actually mapping all of the mutations in a whole set
of different cancer patients for mutations in just those
regions. We also have a few experiments going on in the lab
where we are trying to build mouse models of cancer taking some of
these microRNAs that we think are oncogenes and over expressing
them in mice to try and prove that they can indeed cause cancer by
themselves. Then using those mouse models then as models for
therapy, we can try and treat those mice with therapies, maybe
novel therapies that we might ultimately move into the clinic one
day that could help patients that also have over expression of
these same microRNAs. And then we also have mouse models
where we try to cure the mice using these microRNA genes themselves
both as genes and also just by giving them the RNAs. I am
trying to inject these RNAs into either the tumor site or into the
bloodstream or try and give it to them orally or intranasally and
see if we can cure these particular mice.
Foss
And Joanne?
Weidhaas
There is a lot certainly that overlaps because Frank and I
collaborate closely on many of these projects and have a lot of the
same interests. One thing I would add is that there is
certainly sequencing going on, although the focus has really been
the sequence of tumors instead of normal germline DNA, and I feel
very strongly that there is a wealth of information that is in your
normal DNA and for people that develop cancer, it is searching for
the hidden treasure. There is something there that is going
to be meaningful for them, and for their children, not only in the
development of cancer but ultimately in how that tumor biologically
behaves, their outcome and
28:56 into mp3 file
http://yalecancercenter.org/podcast/july1810-cancer-answers-slack-weidhaas.mp3
what therapy is the very best therapy for them. I do not
think that is going to be in just trying to look at the tumor,
which is so complicated by the time we look at it, there are so
many events that take place as a tumor progresses, but instead, in
what tipped the scales for them in the pathway that their tumor
took to become a tumor, I think that is going to be the Achilles'
heel of those tumors. We do have some data suggesting that
with some of these markers that we found that do predict for
a cancer risk that those same markers tell them how those people
are going to do, what therapies are going to work or not
work for those patients. Ultimately that's everybody's goal,
we can't prevent all cancers, but to get this fundamental
understanding we could certainly do a better job treating. I
mean everyone does their best, and everyone gets standard of care,
certainly, but there are some things that works differently and
better for other people and there is such diversity across
cancer.
Foss
It certainly does sound like there is a lot of exciting new
research going on and that there will be a marrying of these two
different aspects of the susceptibility of the patient as well as
identifying specific mutations in the tumors that ultimately will
hopefully lead to cure for more patients. It has been very
exciting having this discussion with both of you today. Until
next week, this is Dr. Francine Foss from Yale Cancer Center
wishing you a safe and healthy week.
If you have questions or would like to share your comments, visit yalecancecenter.org where you can also subscribe to our podcast and find written transcripts of past programs. I am Bruce Barber and you are listening to the WNPR Health Forum on the Connecticut Public Broadcasting Network.