Efraim Racker

Transcript

 

[ Crowd Noises ]

>> Rick Cerione

Good evening.

I would like to welcome all of you to this year's Efraim Racker Lectures in Biology and Medicine.

This lecture series began in 1992 with the intention of bringing both to the scientific community of Cornell but also to the broader public examples of some of the major developments in biology and medicine by those individuals who have been responsible for some rather remarkable accomplishments, and in that vein, we've been very fortunate over the years right from the very beginning to have had some truly outstanding lecturers, including the first lecturer Jim Watson and then others including Sidney Brenner, David Baltimore, Howard Varmus, Robert Weinberg, Robert Lefkowitz, and tonight's lecturer, Lewis Cantley.

What we've tried to do with this series is to identify individuals whose research interests really epitomize the breadth of Ef's work through his career and his devotion both to basic science and to a molecular understanding of disease.

Ef came to Ithaca in 1966 as the Albert Einstein Professor of Biochemistry after a rather extraordinary career, where he'd spent time at NYU in the Public Health Institute of New York and would continue on an amazing career while he was at Cornell.

For those who knew Ef, you know that his research style was one that drew from really an extraordinary creativity and imagination every bit as much as drawing from technology developments or literature details.

His science was really like his art and his painting, which was of course another one of his great loves, and I know he'd be particularly pleased with this year's selection as a lecturer for a number of reasons.

One, Ef knew Lew Cantley when he was a graduate student here in chemistry at Cornell with Gordon Thomas, and because Lew's discovery of the PI 3-kinase, and really what that has meant to growth factor signaling in cancer, and Lew's more recent interest in the roles of metabolism in cancer and disease perfectly parallel Ef's life work in terms of trying to really understand how a Barrett metabolism can give rise to disease states and especially how altered glycolysis, the well-known Warburg Effect, plays a role in cancer progression.

On top of this, I think Ef would be particularly pleased because Lew has, in effect, really now come to his senses and after a very distinguished career at Harvard, has come back home, so to speak.

He is now going to be a Director of Cancer Institute at Weill Cornell, so he has come back to Cornell.

Now before David Shalloway formally introduces Lew, I have one other addition to read, and that is next year, 2013, is going to be the Centennial year of Ef's birth, and so to commemorate that, we are all very pleased to announce that in conjunction with Ef's daughter, Dr. Ann Costello, who's here with her husband John and her daughter, they're going to be developing a new website that will make past Racker lectures available online and will allow you to view the 78 art albums that were created by Ef from 1936 right up to immediately before his death in 1991.

Also photographs and stories of life on the Cornell campus during Ef's time in Cornell will also be included, and we are really reaching out to those who might have copies of Racker paintings or stories or remembrances that they might want to share, and for those who might, and would be willing to contribute, you could contact any of the members of this organizing committee whose emails are on the pamphlets, I think, or on the posters that are outside the lecture hall.

So with that, now David Shalloway will formally introduced this year's Racker lecturer.

>> David Shalloway

Up to a few weeks ago, Lew Cantley was the William Bosworth Castle Chair in Medicine, and Professor of Systems Biology at Harvard, and the Director of the Harvard Associated Beth Israel Deaconess Medical Center as Chief of the Division of Signal Transduction.

No more. As Rick said, we're delighted and honored that two weeks ago, he moved to Cornell and became the Director of the Weill Center -- Weill Cornell Cancer Center in New York City.

He is one of the world's most preeminent scientists in both basic and clinical research and, as you heard, a Cornell University alumnus.

So tonight's Racker lecturer is just the beginning of a closer scientific relationship.

Dr. Cantley graduated summa cum laude in 1971 with a Bachelor of Science degree in chemistry from West Virginia Wesleyan College, and obtained a PhD in biophysical chemistry from Cornell in 1975.

He conducted postdoctoral research at Harvard from 1975 to '78 when he was appointed Assistant Professor of Biochemistry and Molecular Biology there.

He then went to Tufts in 1985, but returned to Harvard Medical School as a Professor of Cell Biology in 1992.

He has made some very important advances in cancer research stemming from his discovery of the signaling pathway involving Phosphoinositide 3-kinase, PI 3-kinase, in the mid-80s, which explains a lot about the signal transduction of the growth of the cell and has major implications for cancer.

His pioneering research discovered that human cancer is frequently a PI 3-kinase mutation and for the past three decades, he has worked to identify new treatments for cancer that result from defects in the pathway.

This discovery has led to promising avenues in the development of personalized cancer therapies and has resulted in novel treatments for cancer, diabetes, and autoimmune diseases.

He is quite interested, as you'll hear, in translational research and his research is funded in part by the charity Stand Up To Cancer, and has a strong clinical as well as basic focus.

Dr. Cantley is a member of the American Academy of Arts and Sciences, the National Academy of Arts and Sciences, and has received many awards that you can read about in your program.

What isn't mentioned there is that he was recently featured, along with three others, in the documentary "Inspiring West Virginians," aired on West Virginia Public radio couple of weeks ago.

A tribute from his home state.

We are delighted to have him with us tonight to speak on the future of cancer prevention and treatment, moving from discovery to implementation.

[ Applause ]

>> Lewis Cantley

Thank you.

Thank you, David.

Let me just turn this on so.

So it's a pleasure to be back here.

Actually I typically make it back to Cornell every two or three years, so it's not like it's been such a long time, but it certainly is -- always brings back great memories of the times that I was here.

And I want to start out, my lecture really has three parts to it.

First I want to start with a few little vignettes of my personal interactions with Ef Racker when I was a graduate student here.

I was incredibly -- in many ways he was a mentor to me.

I worked for Gordon Hammes in the Chemistry Department, but my research was really an extension and inspired by work that Ef Racker initiated, and I talked to him virtually every week the time that I was here, and we became very close friends and even long after I left, we continued to correspond very frequently.

This is a picture of Ef, and my guess is this was about almost the time that I arrived here that this picture was taken, either late 60s or early 70s.

That's probably in his office because I remember that I would go over to Wing Hall, that was my lab, Gordon Hammes's lab was in Baker, but I was working on the F1-ATPase, the enzyme synthesized ATP from mitochondria, and to purify that enzyme, I had to start out with about at least one B part full of, one Cal full of B part mitochondria to isolate the protein in order to do the experiments, and Ef was kind enough to allow me to use the mitochondria the data isolated in his laboratory to start my preparation.

So I would go over to Wing Hall with my ice bucket and spin down the mitochondria and sonicate them and start the first steps of the purification there, but while I was waiting for all that to happen, almost invariably, Ef, who was always in the laboratory, would bump into me, and as a young graduate student, he would pull me into his office and start talking to me about how is ATP really synthesized by this enzyme.

And it was very controversial at that time, 1971, as to how you could actually make ATP in mitochondria.

In fact, it was not just controversial; it was really contentious, because various people had very strong ideas about how it worked.

But Ef had a very open mind about this, and he was considering all possible ways by which this could happen.

The Chemiosmotic idea was becoming popular, but not yet generally accepted, and certainly not proven, and so we talked a lot about that, and I was incredibly impressed, as a first year graduate student, he would spend an hour in his laboratory, even though I didn't work for him, discussing these ideas with me while I was waiting for the centrifuge to finish spinning down my enzyme preparation.

So that -- those were very impressionable times in those early days, and I should say that in about the fourth year of, I guess all three and half years into my degree, Peter Mitchell, who ultimately received the Nobel Prize for discovering the Chemiosmotic idea of how proton gradient could be used to synthesize ATP, came to visit Cornell.

He had at that time not received the Nobel Prize.

In fact the definitive experiment to prove his idea had yet to be done, and he visited -- Ef Racker brought him over to visit Gordon Hammes and myself, and I spent about an hour and a half presenting my thesis work to the two of them, so you can imagine, here's Peter Mitchell, here's Efraim Racker, Gordon Hammes, and then me presenting slides of my nucleotide-binding to the F1-ATPase, and every single slide I would show, Peter Mitchell would say, that's consistent with the Chemiosmotic Hypothesis, and Ef Racker would say, but what about this possibility?

And actually I don't think I got seven words in in the hour and a half, the two of them argued over what my data meant.

But that was the level of excitement there was and the level of contention and the level of important style about that work.

So in the end of course, Ef Racker was the one who actually did the definitive experiment to prove that the Chemiosmotic theory was correct and once that was generally accepted among the community, Peter Mitchell got the Nobel Prize.

I think today, without any doubt, Ef Racker would have gotten the prize for what he contributed to that work.

After all, he purified all the components and proved that that's how it worked.

The idea that it was a proton gradient really came from Peter Mitchell.

So that was my exciting years as a graduate student.

I'd I like to think that I brought something to the theory, but in fact, if you go back and read my PhD work, it was actually pretty boring.

In fact, Rick Cerione keeps explaining to me how he went back and looked through my notebooks and decided it was totally worthless and dumped it all in the wastebasket, as he followed up all my research.

>> That's a personal story.

>> Oh, I'm not supposed to tell you that.

So, but the other thing that impressed me about Ef Racker is that he was -- he was always -- he was an incredibly creative guy, but he was always coming up with very interesting quotes, and these were meant to sometimes be humorous, but there's always enough truth in them that -- that they really stuck with you, and so, in 1963, when the controversy about how you made ATP and mitochondria was at its peak, the -- at one of the meetings on this, Ef Racker is famous for having said that, quoting Albert Einstein, that "Nature may be difficult, but she is never malicious." Ef Racker's response was, well, obviously Einstein never worked on oxidative phosphorylation.

And then in a follow-up on that is that anyone who is not thoroughly confused by oxidative phosphorylation just does not understand the situation, so that was the level of complexity of this, and I should say that as we try to teach this to biochemistry students, graduate students, undergraduate students, and medical students as I sometimes do now, I think they still find that it's incredibly complicated and confusing.

Now I'll come back to these kinds of quotes again in a minute.

Another quote that really stuck in my mind, and this was a quote -- I think I first heard from him when I came back to give a lecture at Cornell after I had become an assistant professor at Harvard, and I remember joking with Ef before I gave my talk.

I said well, the last time I talked here was you and Peter Mitchell were in the audience and I actually couldn't get to say a single word during my talk, and Ef was very famous for interrupting the speaker periodically.

What did you mean by this?

Or, what about this?

And so I gave my talk and he didn't say a single thing through the entire talk, and at the end of the talk I was offended that he hadn't interrupted me.

I felt like maybe I hadn't said anything interesting at all to him, and I said, well, Ef, don't you have a question?

He said, well you told me not to ask anything.

But he did, of course; at that point he then started chiming in.

But one of the points that he was making, I can't even remember what my talk was about, but I must have mentioned something about organelles and how they were specialized for different functions, and he made the point that to solve the complexity of generating life, nature has taken the course recommended by Philip the Second.

Some people claim that it was Julius Caesar who said this: divide and conquer, and this is really what -- what helps a cell accomplishes this complex feat of dividing tasks, of lipid synthesis, protein synthesis, nucleic acid synthesis, in to different compartments in order to solve the complexity of using sometimes the same components for different purposes.

And this turned out to be really important.

To me it was an insightful thought.

I always wondered, what the hell are all those organelles out there anyway?

I should say I never had a biology course.

I was trained entirely as a chemist and so to me to look at a cell with all the stuff in it was totally confusing.

I wondered why isn't it just a bag of enzymes, and so Ef Racker's comment about this got me thinking about oh, okay, that makes sense.

So I didn't like to paraphrase that some of Racker's quotes with regard to cancer, and Ef Racker was very interested in cancer and was very much persuaded by Otto Warburg, whom I will bring up in more detail in a moment, and his thoughts about cancer.

But to paraphrase Ef Racker in the context of cancer, again if you consider Albert Einstein's comments about nature being difficult but not malicious, my response to that would be obviously, Einstein did not work with cancer.

Cancer is a malicious disease without any doubt.

And actually I'd also say that, again to paraphrase another one of Racker's comments, that anyone who thinks we will cure all cancers, in other words, anyone who thinks that cancer is not complicated and we'll cure it all in the next 10 years as we cure -- we continue -- I think every 10 years, somebody states we're going to cure cancer in the next 10 years.

Anyone who thinks that doesn't really understand the disease.

It's an incredibly complicated disease.

And then finally I would say that to have an impact on cancer, we must take a lesson from nature and divide and conquer, so this idea of dividing and conquer really is the theme of what I'll be talking about tonight with regard to cancer.

We have to divide and conquer cancer to conquer cancer.

So I'm going to give a very brief history, a modern history of cancer, and I'm going to go into the detailed history of how cancer was initially found in the ancient times, but if you are interested in a very detailed, very -- going all the way back to ancient times about how cancer has been observed and treated over the years and the modern -- the modern breakthroughs in it, there's a really very nice book by Sid Mukherjee, who actually worked with one of my trainees when he was a fellow at -- in the Harvard system, and it's called The Emperor of All Maladies and it won a Pulitzer Prize.

It's extremely well written and a very accurate history of cancer, for those of you who want -- who, after hearing me talk [inaudible] tonight want to read it in more detail.

But I want to start just 90 years ago really, or 100 years ago, to one of the first breakthroughs with our understanding cancer at a more molecular level, and that came from work of a fellow named Peyton Rous, and what's interesting about Peyton Rous to me, one of the things that's interesting, is that the temporary office I currently have at Weill Cornell looks straight on to the building where he worked and made the discovery there that I'm about to tell you about.

So he was at the Rockefeller and he made the observation that chickens come down with cancers and that there was something infectious about the cancer that they got and that you could extract particles from that cancer and give them to another chicken and that chicken would come down with cancer, too.

And he ultimately purified that component that was transmitting the cancer from one chicken to the next, and concluded there was a virus.

And so this had a major impact in -- in -- in the early part of the 20th century and our thoughts about cancer so Peyton Rous was saying it's, you know, cancer is a transmissible disease.

I should say ultimately, Peyton Rous got the Nobel prize for this discovery, but not until 1966, and as I tell a little bit more history you'll see why it took so much later before his -- the -- he was acknowledged for having impacted the field.

In fact, the reason that his discovery did not have a bigger impact, or that the impact was relatively short-lived, is that as people looked at human cancers and looked for transmissible agents, they found that it was extremely rare that you could find -- it wasn't impossible we know, there are viruses like papilloma virus, even HIV can predispose you to cancer, so there are vir -- there are cancers that clearly are affected by viruses, leukemias, but they are extremely rare, so the general concept that all cancers are caused by viruses is clearly not true, and since cancer was really thought of as a single disease, if you found a single exception to its cause, you would conclude that therefore that is an epiphenomenom, is not related to cancer, it's not a causative agent.

So his ideas didn't really have much traction in the middle part of the 20th century.

Instead, another scientist raised, really rose to prominence in the cancer field.

This was a fellow named Otto Warburg, and Warburg was a classic German biochemist and in the day when the German biochemist really ruled the world, in 1924 he was looking into the biochemistry metabolism in cancers and he took out the tumor, tumors from various animals and compared the metabolism of the tumor to the metabolism of the normal tissue in which the tumor arose, and he was particularly interested in glucose metabolism, so what happens to sugar when it goes in to the tumor versus when it goes in to the normal tissue?

And he found that most of the sugar that went into the normal tissue was metabolized in an oxidative way.

In other words, oxygen was being consumed, he took the glucose away, oxygen consumption went down, you added glucose, oxygen consumption went up, and the glucose got converted to carbon dioxide.

That's the way most normal tissues did.

But when he did this with a tumor, he found that most of the glucose, first of all, the tumor was eating glucose about 20 to 50 fold faster than the normal tissue, and secondly, it wasn't burning it.

It wasn't converting it to carbon dioxide the way the normal tissue was.

It was what he called fermenting it, much like what yeasts do to convert sugar into alcohol.

The tumor was converting the sugar into lactate, and that wasn't consuming any oxygen.

This was a process which ultimately became called glycolysis that did not involve burning, but rather converting the chemical energy of glucose into the chemical energy of ATP.

And so what we later learned about this process is that it's an incredibly inefficient way to make ATP, and it became a rather mystery of why the cancer should have such a very inefficient way to make ATP.

But this observation held up very strongly throughout the 20th century.

Almost every cancer that one looked at had this phenotype, this so-called, it became known as the Warburg Effect, of eating glucose at a very high rate, but not burning it, but rather converting to ATP and spitting out lactate.

Now ultimately, Warburg got the Nobel Prize as well, in fact he got it before Peyton Rous did, not for this discovery but for other work that he did in metabolism.

But he was convinced throughout his career that the secret to understanding cancer was to understand this shift from oxidative phosphorylation to glycolysis.

Now the reason I tell you this story, it's something to tell, is because Ef Racker was really the person who figured out oxidative phosphorylation, purified the whole component, and he, throughout his career, as he was trying to understand how ATP was made by this process, was continually intrigued by why cancers don't do this efficient mechanism for ATP production, but rather do glycolysis.

And so many of my discussions with Ef Racker as a graduate student really centered around this question, why is this happening, and as many of you know, he thought about this throughout his career.

So then let's speed forward to more modern times, and I call them modern times, 1971 is really modern times because that's when I started as a graduate student, and so some of you may think this is actually ancient history, 1971, but to me it sounds like yesterday.

So in 1971, Richard Nixon declared war on cancer, and I -- and I think, at some point in that announcement, he expected -- stated that we would cure cancer in 10 years.

Almost everyone who's ever had a prominent position in cancer at the NCI has at some point said we will cure cancer in 10 years.

Obviously it didn't happen, and I'm not going to say it tonight either.

But the war on cancer was declared and at the time the war on cancer was declared, Peyton Rous would come back into favor again.

He'd just gotten the Nobel Prize because in the '50s, and later in the '60s, more and more viruses were being found in mice and chickens that caused cancer, like the Rous sarcoma virus, and -- and so at the time that Nixon declared war on cancer, there was a much greater enthusiasm for embracing these viruses.

DNA had been discovered.

RNA had been discovered, and the viruses clearly were primarily DNA and RNA, so we felt, well we can figure out what DNA and RNA does, so let's focus on how DNA and RNA imported into cells cause cancer.

And so that's -- a lot of the early funding NCI had went into this area, and it was a useful time to do it and it made a major breakthrough in that in 1976, Mike Bishop and Harold Varmus, at the UCSF at that time, discovered, working with Peyton Rous's sarcoma virus, that the mechanism, like which that virus was causing tumors in chickens, was that it had picked up an endogenous chicken gene, so the virus infected the chicken, reverse transcribed the messenger RNA from one of the genes that the chicken normally makes called, which ultimately was called the Sarc gene, but in the process of reverse transcribing it, it mutated it, and that mutation made it much more active than the wild type gene, and that activity ultimately explained why the virus caused the cancer.

David Shalloway has spent much of his career trying to understand how that -- the normal Sarc gene really does cause cancer and how it's regulated.

So that discovery was enormous because it told us that you could mutate a normal human gene, or at least a normal chicken gene, we assume something similar could happen in humans, and turn it into what they then called an Onco gene, a cancer gene.

Now -- yet there was still some reluctance to -- to believe that this was generally going to be relative to cancer because, again, most cancers were not caused by viruses.

But in 1982, Bob Weinberg, who was a former Racker lecturer here, so those of you who've been to Tinney's every year have already heard this story, made the discovery that you can get the same kind of oncogenic mutations by sporadic mutational events that can be caused by mutagens, reactive oxidant species, radiation UV light, can mutate genes, and you end up -- you can end up with exactly the same mutation sporadically that the virus induces when it infects, and that was first observed with regard to a gene called Ras that caused tumors in mice.

And so at that time, by the mid-1980s with Weinberg's discovery and Bishop and Varmus, it was generally accepted that cancer really was a mutagenic disease, and the question then was, well, what do these oncogenes do?

And -- and so at -- by this time, I was actually even an associate professor by that time, and was very excited by convert -- moving my lab and to trying to understand what these oncogenic do.

Now I'll rush to the future and tell you that now that we can sequence genes at a very high rate, it costs about $10,000 now to sequence an entire genome of a tumor, and there are thousands of tumors that are being sequenced, and within a year it will only cost about $1000 to do it, and so 10 years from now, everyone will wonder, well why didn't you sequence, you know, you went to the doctor yesterday, why didn't you sequence your genome?

It only cost 50 bucks.

So this is going to be a very routine process.

But what we've already learned from sequencing a few thousand tumors is that there are only about, the good news is, that there are only about 100 or so, maybe 150 at most, genes whose mutations are deriving cancers.

That's good, that's good news because we have 20,000, you know, more than 20, 22,000 genes in our body, and we only have to worry about 150 to 200, and in fact, most of the cancers that we look at are really confined to mutations in only about a few dozen genes.

But let's say roughly 100 genes we have to worry about, on the average, as being responsible for the majority of cancers that we run into.

So that's the good news.

Now the other good news is that mutating a single gene is, at least in humans, is almost certainly not enough to cause a cancer.

You have to have several mutational events, or we now know some types epigenetic events, before a cancer can emerge, and so that's good news as well because that says that these are relatively infrequent events and to have three of them happen in the same cell is going to be rare, and that's why we could live to be 60, 70 years old before we are likely to get cancers.

So that's the good news.

The bad news is the same math kind of works against us, because if you really do have to have at least three genes, three to five genes, and if you're -- and any of these roughly 100 genes can mutate to form cancer, can occur almost in any random order and still result in a cancer, and as we're sequencing more and more, we see that things tend to be fairly random in the set of genes that get mutated.

Then if you do the math on this, then the number of different types of cancers that are likely to exist are roughly 100 times 99 times 98, so 1-2-3 in any order, and that comes out to a million.

So, you know, a rough estimate is that there are a million different cancers.

Now most clinicians say there are about 20 different cancers.

They are calling cancers based on the tissue from which they emerged.

Actually, if you look at the more rare ones, there may be as many as 50, but not a million, and so this is actually rather scary.

It says that no two people who have breast cancer are probably going to have exactly the same disease, and -- so we have -- if we really start going to the molecular level to define our cancers, we're going to find that they're really breaking down into many, many, many subsets.

So that could be a problem.

So the next issue that came up, and this was of course very central back in the 70s and 80s as these oncogenes were being discovered, is what do these cancer genes do?

A lot of them turn out to be kinases, some of them turn out to be transcription factors, they were in signaling networks, so-called signaling networks, nobody knows what that means.

That means that they're talking to each other in some complex wiring diagram.

But the real question is what do they do to cause the cancer?

And as only in the last 10 years or so have we really begun to figure out what the downstream events are that are critical in these oncogenes control that convert a normal cell into a cancer cell, and for the most part, most of these events turn out to be changing the metabolism in the cell.

So what we've discovered is after nine years of research, is we've come full circle, and we've finally come back to Otto Warburg again, that what all these oncogenes do is shift metabolism to a different state.

And so it seems a little frustrating that we've made it back to 1924 after all this work, but we have learned a lot along the pathway because we now understand a lot about the wiring mechanism by which these oncogenic mutations switch metabolism so that cells that are normally designed to be static can only grow if there's an injury, will now grow continuously as though they had an injury continuously, so these oncogenic mutations tend to shift the metabolism into doing growth patterns using glucose and amino acids to make DNA, RNA, proteins and lipids in order for a cell to grow and divide, rather than just making ATP in order to keep the cell alive.

So that's the major thing these oncogenes are doing.

So the question then is if there's a million different ways, mutations that can potentially do this, how do we figure out how to conduct therapies?

How can we reduce this to a simple enough problem that we can attack it?

So for those of you who are not scientists in the audience, this slide is totally meaningless.

It looks like a wiring diagram, something you would see in the manual to your refrigerator that explains how the fan and the compressor all work and where the resistors and capacitors are.

And in some ways, this wiring diagram kind of works like the wiring diagram for your refrigerator.

But in the end, at the very bottom you'll see protein synthesis, glycolysis, and in fact metabolism in general being regulated by this network.

This is the network that my lab has worked on for more than 20 years, PI 3-kinase, which David Shalloway introduced in the introduction, is an enzyme that we discovered back in its called PI3K here that we identify back in the late 1980s.

It's central in this pathway, and it's regulated by a host of different receptors, molecules in the surface of the cell that respond to growth factors, things even like insulin or insulin like growth factor, factors that come in to tell a cell to repair itself if there's an injury in the tissue, and so it is one of the components of the signaling network.

We discovered it because it was associated with a number of viral oncogenes.

In fact it itself turns out to be a virally encoded oncogene.

Ras, the gene that Bob Weinberg discovered, is mutated in human cancers but is also picked up by viruses, is also its component, in fact, Ras and PI3K directly touch each other and are involved in -- in regulation in that way.

You notice that Ras is much smaller than PI3K, in this pathway, just to tell you the relative importance of those two genes.

So you may recognize some other genes, too, RAF, and some of you may have heard it, B-raf, the gene that's most frequently mutated in melanoma, which there's a lot of excitement about drugs that target this, so what we and others have worked on in many years is trying to figure out how all of these various oncogenes, including the -- everything in red here I should say is an oncogene, that -- that's -- many of which were picked up in visually in viruses that are sporadically mutated or amplified in human disease, and how they all interact with each other to communicate signals that ultimately regulate metabolism, including this increasing glycolysis that Otto Warburg described in 1924.

So we're finally getting to understand why this is happening with this disease.

I should say everything in blue is a tumor suppressor gene, and that means that if you lose these genes, you get cancer.

If you gain function in these genes, you get cancer, so it's a very complex array but we're beginning to make sense of how it all fits together.

Now that tells us, once we understand the wiring diagram, we don't necessarily have to have a different drug for every single mutation, because we can focus on key nodes that all come together and maybe we only need a few dozen therapies.

We don't need a million therapies, if we can understand the wiring diagram.

So that's the hope and that's -- that's how we hope, that's how we think that we will ultimately make progress in cancer is by understanding the wiring diagram that approach and attacking it logically.

The one thing I should caution about these wiring diagrams is that you notice a lot of redundancies.

This pathway circles back to here, this pathway circles back to here, and so you can imagine that you shut off this branch, you may not block the cancer because the other branch can accomplish the same feat, so these redundancies make single agent drugs not necessarily effective, and so we have to understand the drug -- the wiring diagram to figure out how to use drug combinations to treat cancers.

An illustration of the Warburg Effect is actually shown here in a mouse in which we introduce the PI3K gene; we introduced actually a mutant form of the PI3K gene, using the mutation that is very frequently found in human breast cancer and to some extent, also in lung cancer.

And so putting the human mutated gene into the mouse lung alveolar tissue and turning it on in a drug dependent manner, we found that after about three months of this gene being on, the mouse developed the cancer.

Now this cancer, this gene alone is not enough to cause the cancer, but during that three month period of time, additional events happened, other mutations that we don't yet understand, occurred that allowed the cancer to occur in a single cell, even though every single cell in the lung had this gene in it, only one other ultimately became a cancer because it acquired the additional mutations necessary to transform it.

But, what we learned was that this cancer was still addicted to that particular oncogene because we could turn the gene back off and the cancer would go away.

Or we could add a drug that inhibited the activity of PI 3K, and the cancer would go away.

But the first thing we noticed was that the cancer, first of all, takes up glucose at an incredibly high rate.

This is what Warburg observed in his Ras back 80 years ago, 85 years ago, and we see the same thing when we put a PI 3K into a tumor.

It drives glucose uptake, and we can see this when using a radioactive form of glucose, fluorodeoxyglucose, so if you -- may know relatives or friends who've actually had this radioactive molecule injected into their bloodstream in order to visualize their cancer.

Doctors still today use the Warburg Effect as a way to figure out where your cancer is in your body because it's taking up glucose so much more than the surrounding normal tissue.

You turn off this enzyme and within 48 hours the glucose uptake goes away, so this cancer is eating glucose at a very high rate because it has a mutated PI 3K.

You turn PI 3K off, the glucose uptake goes away, and by three days the tumor, four days, the tumor is massively reduced and by three weeks, the tumor has gone away.

So in this simple model, we can actually cure this mouse by reversing the functions of the PI 3K.

Can we accomplish this in humans?

The first evidence that we can do it really came from a drug called Herceptin, which some of you may have heard of.

Very frequently used in breast cancer.

It's used for a very specific subset of breast cancer called HER2 Positive breast cancer, and it was -- went into clinical trials from Genentech in the 1990s, because of the observation that HER2, which is a tyrosine kinase, one of these receptor types at the top, it's right here, so it's an activator of PI 3-kinase, and you can get a sense, if it's on the soft surface an antibody will attack it, and so Genentech developed an antibody that attacked this tyrosine kinase, they put it in the clinical trials and it got approved, relatively minor affect, like three month life extension in end stage disease, but that got it approved.

Once it went into adjuvant therapy and following it up in the next 10 years, we've noticed about a 70-80% reduction in recurrence of this disease if you take this monoclonal antibody after surgery, and so this is not that effective in curing disease once it's in metastatic state, but preventing the metastasis from occurring by treatment after surgery has been remarkable with this drug.

It's saved many, many lives.

We could say these have been cures; they've been cures by preventing relapse, not at the end stage disease level.

But that's exciting.

A second very strong excitement that came slightly after that was a drug called Gleevec, which is another tyrosine kinase inhibitor, and this attacked a drug -- a target called ABL, actually BCR ABL, that was molecularly defined as a -- as a -- as a translocation back in the 1960s, so-called the Philadelphia chromosome translocation, which observed in a microscope that ultimately later was found to be an activation of an oncogene called ABL.

This drug attacks the ABL enzyme and it went into phase 1 clinical trials back in -- in the late 1990s, and what was found in the phase 1 trial is that every single patient who took the drug, their disease disappeared.

Virtually every one.

So typically to get a drug approved, even in the case of Herceptin, you're talking about a 5000, 10,000 patient, 5, 6, 7 year trial to try to see a three-month life extension.

This drug, by the time 50 people had taken it, everybody responded.

It was unethical not to give everybody the drug because it was proved in a remarkably short period of time and I know people who have been on this drug for 10, more than 10 years, that still have had -- not had a recurrence.

It's not yet a cure though.

You take the drug away, and the disease comes back.

And some people become resistant to it.

But these two examples inspired pharmaceutical companies to go down this route scientifically approaching cancers based on what's going on in the mutations, because of these two early successes.

We now have a number of targeted therapies, including a therapy that attacks the EGF receptor, another one of these tyrosine kinase on the surface of cells very much like HER2 and this is a really a good lesson for us to learn with regard to clinical trials.

So this drug, much like Herceptin attacking HER2, this drug was attacking [inaudible] molecules attacking EGF receptor.

It went into the first drug was a drug called [inaudible], went into a large phase three clinical trial and this is roughly what the trial looked like.

I made up this data I should say; it's roughly what the trial looked like, and the red line is the drug, the patients who are treated with the drug, and the blue line is a placebo.

Now the FDA decides whether to approve the drug.

As to what the difference is for the 50% survival rate is how many years that is, or how many months that is, so in the case of Herceptin, that was about three months and that was enough to get it approved.

This was about two months, not enough to get it approved.

But what the doctors recognized was that there were these subgroup of patients, about 5-7% of the patients on the trial, who had miraculous responses.

They were end stage disease, they were on oxygen, they couldn't even walk, they took the drug and a week later they're walking without their oxygen and three months later they're out jogging.

That had never been seen in lung cancer before.

But they couldn't understand why is it only 5%, and why is it on average not much going on at all?

It turned out that these 5% or so all turned out to have mutations in the gene.

They were giving the drug to everyone, whether or not they had a mutation, because they didn't even know the mutations existed in this gene, and clearly the majority of people did not respond at all, but a subset did, so that still didn't get the drug approved.

They had to go back and do a second trial, and the second trial looked more like this.

They now selected patients that had the mutation, and now you see a year or two differentials between the placebo and the drug treatment.

Now ultimately this drug was approved.

So this tells you the difference between doing blind therapies the way we've been doing them for the last 30 years, which is large placebo control in unselected patients, 10,000, five or six years, and then you unblind it at the end and you discover maybe a three-month life extension, maybe nothing, and actually selecting patients that you know are going to respond, and everybody responds.

After 50 patients you have the drug approved.

Obviously you want to do the latter, not the former, but surprisingly, it's still the majority of clinical trials are done today like this.

Not like that.

That's what we have to change.

And that's what I'm excited about, and that's the reason I took the job at Weill Cornell because I think we can put into place the infrastructure there to ensure that all the trials we'll be doing at least at Weill Cornell fall into this latter category.

How much time do I have left?

>> Ten minutes.

>> Ten minutes.

Okay, so I'm going to very briefly tell you about -- David mentioned in my introduction that I received this $15 million grant from Stand Up To Cancer.

This is a group of Hollywood producers and ad agency executives who wanted to implement an organization analogous sort of to the AIDS activists where they would accelerate drug combinations and treatments for cancer to get things into the clinic much faster and get them approved much faster, and so I put together a team and we applied for this money and we got $15 million.

Now the team that we put together, $15 million sounds like a lot of money, but actually we have 65 people on our team and seven institutions.

By the time you divide that up, it's not a lot of money.

But per person, it's clearly not a lot of money, but because it held us together as a team, and this team includes surgeons, oncologists, molecular pathologists who know how to look for these mutations, obviously we need the surgeons and the oncologists to enroll the patients, to help us design the trial, we clinical trial experts, we have people who are experts in the pathway that I told you about, and people who are experts in, excuse me, and people who are experts in mouse models for these cancers, and so we all got together and worked as a team to figure out which drugs are most likely to work, particularly how to find drug combinations we combine with PI 3-kinase inhibitors that could be useful in treating cancers.

And the reason we focused on PI 3-kinase is not just because that's discovered in my laboratory but rather because we were discovering that in women's cancers by far, PI 3-kinase is mutated much more frequently than any other gene.

Far more frequently than HER2, for example, and so how might we figure out how to develop drugs that we get to start with?

We also have patient advocates, in fact one of our patient advocates are also experts in clinical trial design and nurses who've worked with -- in cancer clinical trials.

So this was the pathway we really wanted to target.

We wanted to figure out how to target multiple nodes in the pathway to have a better effect than a single agent drug would likely have because these single agent drugs, even EGF receptor inhibitor, and the and the ABL inhibitor and the BRAF inhibitors which were recently approved, these don't cure people.

They extend life, they definitely extend quality-of-life, but ultimately the tumor comes back, so how do we figure out how to prevent the tumors from coming back?

One of the things we did upfront was just assess what is the frequency of mutations in those nodes in that network that I showed you in women's cancers?

So we looked at all the tumors that we could get a sequence from, from all the seven institutions, and added up all of our numbers and asked what's the frequency of the PI3K mutations?

You can see in bold, you don't need to see it back there but everything in bold is at least 20% of cancers have that particular gene mutated.

PI 3-kinase is clearly the champion here.

A third of the -- of the -- of the ER positive breast cancers have this, 25% of HER2 Positive breast cancers have PIK3CA mutations.

In others, you can see very frequently other events in this network.

Now pharmaceutical companies have already begun to develop inhibitors for every node in this pathway, everything that they can figure out a way to drug, they were putting drugs in the clinical.

There are 20 PI 3-kinase inhibitors currently in clinical trials.

There are also ATB inhibitors.

You look at almost every node in this pathway and I left off all the drugs that hit this pathway, the BRAF inhibitors, the neck inhibitors, etc. And of course they're all the tyrosine kinase inhibitors which have already been approved, so we have really a plethora of drugs to choose from, but if we randomly chose every possible combination, we wouldn't have enough patients on earth to do all the trials you would need to do to get these drugs approved, so there has to be a scientific logic to decide which is the better accommodation to use.

How do we arrive at the logic?

That was the task that I assigned our team to figure out.

What are the best drug combinations to use based on the mutational events that are going on in a particular patient.

Now we have some early data on single agent trials with P13K inhibitors.

This is a drug that has a lot of promise still in early, it's about to go into phase 2 trials.

A drug from Novartis.

This is what's called a waterfall plot.

Each one of these bars represents a patient that, some of these are breast, some are colorectal, head and neck cancers or other cancers.

And everything below the line here means that the tumor is shrinking while on the drug and everything above the line means that the tumor is growing while the patient's on the drug.

You can see that half the patients that went on this drug had some responses, half continued to progress, and there were some we would call a progression free disease.

Clinically, to call it a true response, it has to be below the dotted line here, more than a 30% reduction in tumor size.

There were a few of those occurring.

The question was why, what's different about this group of patients from this group?

Why did they respond, and why did they not?

We had to figure that out so that we could select for our trial the patients like this one, then it would be approved very rapidly, but if we ended up diluting with patients like this, first of all we would do the patient's no good.

You don't want a patient on a trial if you know they're not going to respond, so we have to get these patients off of this trial and on to some trial that is going to work for them.

So that was our task.

And I won't go into great detail about how we approached this, but basically it required a lot of work to figure out what you can actually do in a clinical setting, getting so called cleared proved biomarker analyses, we were short half the time, they gave us only three years to cure cancer, I told them we needed 10, they gave us three, actually they just gave us the fourth year, we're just going into our fourth year.

We haven't yet cured cancer, but we definitely think we've made progress in figuring out what trials to do, and what combinations to use, and one of the techniques that has really been extremely powerful is to engineer mice -- mouse models to have exactly the same mutations that we see in our human patients.

So we have mice that have PIK3CA mutations in the breast that also have HER2 mutations in the breast, or HER2 implications in the breast, and the various other events that you saw on that graph.

We just re-created mice that had those same events, and then we asked, how did those various mice respond to single agent drugs or combinations of drugs that hit nodes in that pathway, and once we found combinations that worked in that in vivo setting, we would then set up a trial and continue with the mouse work in parallel with the human disease so that our mice would really be slightly ahead of our human clinical trials and tell us what we expect to see next in our patients and how we can anticipate and prevent the resistance from occurring.

And that's what, I just want to end with one very last comment, because I had in my title how might we also prevent cancer?

Everything I've talked about today is once you have a cancer, how do you treat it, and how do you prevent it from coming back as a metastasis after surgery, but in fact our greatest chance of curing cancer or preventing, is to prevent it.

If you can prevent the cancer from ever appearing, obviously you can save a huge amount of money and a huge amount of agony and obviously prevent deaths.

The danger is that right now, and I should say that we've had, we've had progress in reducing cancer rates in some areas because of mainly by reducing smoking.

You know, 30 years ago if we're sitting in this room, half the people would be smoking during my lecture, there be a huge cloud in here, you'd barely be able to see the slides.

I don't see anybody smoking in here right now.

It's amazing.

The smokers are all outside I hope.

But, so that's how to make a big impact in lung cancer, and other cancers as well, but the scary thing is, that as we are reducing smoking, we are also increasing rates of obesity.

There's dramatic increases in rates of obesity particularly in the red states.

I don't know why the red states, but that seems to be a perfect correlation with obesity.

And that's leading to diabetes and surprisingly, because no one was really anticipating this, a dramatic increase in cancer.

There's a strong correlation.

So why is that true?

It turns out, we think, this again is a hypothesis, and I'll just end with this idea, that the reason comes back to PI 3-kinase again, so we discovered PI 3-kinase because of this implication in cancer, but as we began to study, we realized that PI 3-kinase is actually what mediates everything insulin does.

And as a consequence, even if you have too much PI 3-kinase activity, you can get in epithelial cells, you can get cancers, that's what I was just telling you about.

However, if you have too low of PI 3-kinase activity, particularly liver, muscle, and fat tissue, you get diabetes.

You first get insulin resistance, and ultimately you get diabetes, so it's a failure of insulin to be able to activate PI 3-kinase that causes diabetes.

And this is caused by lack of exercise, by overeating, and there's also a genetic predisposition to it.

So two of the major diseases we worry about right now are both involve PI 3-kinase, in the one case too much of it, and the other case too little of it, and the surprising thing is the two are linked, and we think the link is that as you go from obesity to insulin resistance, and I don't have time to get into tell as to why overeating can cause insulin resistance, but we know that it does.

At that stage, insulin levels become very high in your serum, and insulin is the best possible way to activate PI 3-kinase.

And what we're discovering is that a lot of the cancers, particularly these cancers in breast, endometrial, a lot of the cancers we're studying in our dream team have insulin receptors on the surface of the cancer, so that period of time of insulin resistance where your body is making 100 times more insulin than it should be making, it's driving the growth of those cancer cells.

It's the best way to make them grow.

Now by the time you get type II diabetes, the ability to make insulin begins to drop because the aisle itself begins to poop out.

They can't quite keep up with the demand.

But the dangerous time is this period of insulin resistance.

So we think that to get to the model that this insulin resistance raising serum insulin levels is driving these micro tumors into growing at a very high rate, and this is what's predisposing obesity to cancer, and the question is, how can we prevent this from happening?

Obviously, the best way is to get people to eat less and to exercise so that their insulin levels will come down.

Tell them if they don't eat rapid release carbohydrates, the best way to raise your insulin level is to drink Coca-Cola that has sugar in it.

Your insulin level is going to go hundredfold in the next hour.

If you have cancer there that has insulin receptors on it, I would be pretty worried about that.

So lowering the intake of rapid release carbohydrate, keeping insulin levels from fluctuating so high, we think is going to be really important in preventing these types of cancers.

Now I'll just finish by saying that once you have insulin resistance, actually insulin resistance typically is a sign of disease, it's often not treated at all.

You don't even know you have it.

So type II diabetes the doctor tends to do something about it, and the doctor has several choices.

In some cases they can inject you with insulin, you know, if you can't make enough insulin to keep your glucose level down then you're going to have to be injected with it, but you are going to be it injected with very high levels of insulin if your insulin if you're insulin resistance in order to bring the tumor back, I mean to bring the glucose levels back down, so as a consequence, if it's an injection that probably raises insulin as it controls glucose.

The endocrinologist isn't worried about that.

I'm worried about it and I think a lot of oncologists are beginning to worry about it.

Endocrinologists typically don't think about this.

Their job is to get the glucose down.

Now there are other ways to control this.

There are agents that stimulate insulin released from the aisle itself.

They also raise insulin levels they break glucose down, so to the endocrinologist, they've accomplished their feat.

There are other approaches, like Metformin, where you can bring insulin down by suppressing gluconeogenesis.

In that case you're bringing both insulin and glucose down.

Sounds like a better way to do it, and there are also comparing [inaudible] factors that can perhaps accomplish this same event.

So even to the -- although the endocrinologist might consider all of these equally valuable therapies, in fact we would argue that this is the preferable approach to take over this one, if you have the option and can make it happen.

Obviously the best approach is to do exercise and reduce carbohydrate intake, but if you have to be medicated, let's think about what you take.

Now that raises a question, that is whether Metformin, the drug that suppresses insulin levels and glucose, might be one of our best therapies, and it turns out that a number of retrospective studies done in independent countries, independent investigators, all arrived at the conclusion that the subset of type II diabetics who take Metformin have a 25 to 30% reduction in cancer deaths compared to those who take other therapies.

And so this actually raises the question whether Metformin is actually the best cancer drug currently on the market.

It's probably saved more lives than any other cancer drug.

Now the question is, is this really working by the mechanism that we're proposing, or are there other alternative explanations for this?

I don't have time to get into them, but I think it's, as a consequence of these observations, there are more than 100 prospective trials now going on in which Metformin is being given to patients who are not diabetic, in order to see whether it reduces their cancer rates.

So with that, I'll finish and be glad to take questions if we have any time left.

Thank you very much.

[ Applause ]