SPEAKER: In the last several years, there have been major breakthroughs in our understanding of cancer, and yet I find that most laypeople, and also even most scientists, haven't really understood the significance of the ramifications of this new understanding. And so this lecture today is devoted towards the quantal theory of immunity, which is extremely important for the understanding of what the breakthroughs have actually been.
So to understand leukemogenesis-- the genesis of leukemia-- as the prototypic cancer in this instance, really, it boils down to two molecular events that you need to know about. The first is what are the molecular events inside the cell that control the normal cell cycle progression or cellular proliferation, because it's very tightly regulated. The second is we need to know what the molecular events are that when there is a translocation within the chromosomes that lead to a specific kind of a chromosomal abnormality that was labeled the Philadelphia chromosome. If we know about the molecular aspects of these two events, then you'll understand cancer.
The story starts more than 50 years ago in 1960, when Peter Nowell, who was a young professor at the University of Pennsylvania, discovered that phytohemagglutinin could actually cause lymphocytes, which are the primary cell that make up our lymph nodes-- would cause the lymphocytes to become big, large cells that he called blastoid cells-- lymphocyte transform blast cells and undergo mitosis. Up until that time, in most textbooks it was thought that lymphocytes could not undergo self-renewal and were end-stage, terminally differentiated cells.
The second thing that Peter Nowell reported in 1960 in the Journal of the National Cancer Institute was his discovery of a specific chromosomal abnormality in cells from patients with Chronic Myelocytic Leukemia-- CML, which he named the Philadelphia chromosome, because that's whence the city of where the University of Pennsylvania was. And it wasn't until 1973, when Janet Rowley discovered through better techniques to look at the chromosomes that it wasn't just a disappearance of the long arm of chromosome 22, which is what Peter Nowell thought it was, but rather, there was a reciprocal translocation from this long arm of chromosome 22 to the long arm of chromosome 9. And that the reciprocal translocation of genetic material from here to here and from here to here, it turns out is very important in the pathogenesis of leukemia. Then in 1977, Clara Bloomfield, another investigator, found that acute lymphoblastic leukemia-- and not all patients, but a significant fraction also had this Philadelphia chromosome.
Now, this is what chronic myelocytic leukemia looks like to the physician, to the hematologist. If you do a bone marrow biopsy and look in the microscope at a low power, what you find is that there is a hypercellularity, or what the hematologists called a packed marrow of predominantly immature myeloid cells, but also mature myeloid cells, polymorphonuclear leukocytes, or granulocytes. And that became pathognomonic for the identification of a patient with chronic myelocytic leukemia.
Now in 1960, there were basically two theories as to what might be wrong with the cells that made them cancerous. One was-- because the cells looked immature under the microscope-- a lot of their cells did, the thought was that there was some sort of a maturation arrest. Now, the alternative theory was that there was something that had gone wrong in terms of the normal growth control.
At the time around 1960, there had been the discovery of tumor viruses in animals that ultimately were found to be ribonucleic acid tumor viruses. First in 1908, with Ellerman and Bang, in Germany found a cell-free filtrate that would cause leukemia in chickens that could passage from one chicken to another. And then Peyton Rous here at The Rockefeller in New York identified what became known as Rous sarcoma virus, which would cause solid tumors in chicken. And then over the course of the '50s and the '60s, there were just multiple reports of different kinds of leukemia and sarcomas that were discovered to be due to viruses-- up until even 1972, the Gibbon ape leukemia virus. So in primates, as well as cats, as well as mice, there were multiple viruses discovered that would cause leukemic transformation.
And at the time, Howard Temin, working at the University of Wisconsin in Madison, felt that if RNA viruses could cause a transformed cell where you could also transfer the transformed cell, there must be some way that the RNA copied itself into DNA. And so he came up with what he called the protovirus hypothesis in 1964. And on the basis of this hypothesis-- this was really against the central dogma, as coined by Sir Francis Crick of the Watson and Crick fame of the discovery of the structure of DNA.
They felt that the information flow could only be from DNA to RNA to protein, and Temin felt that it must go the other way. And if it did, there was a particular enzyme that would be an RNA-dependent DNA polymerase, which got shortened to reverse transcriptase. Temin had the first evidence that there was such an enzyme, and David Baltimore almost simultaneously found evidence, as well, and there were two back to back papers published in 1970. And this was really an earthshaking event in biology and molecular biology.
Then just a year later, Hanafusa here at The Rockefeller across the street found-- by using techniques of hybridizing DNA from normal chicken cells and viral RNA-- that there were some sequences that had to be similar. And that led then Temin to alter his provirus hypothesis to the protovirus hypothesis, because he hypothesized that what was happening is that if RNA from viruses could copy itself into DNA and insert itself into the cellular genome, that it might then pick up parts of genes from the normal cell and incorporate them into the virus, and that could very well cause cancer. And so that was the protovirus hypothesis.
Now, also at the same time or simultaneously in the '60s and '70s, there were people that were trying to understand normal cell growth control, and they focused on what became known as 3T3 cells, which were reported by George Todaro and Howard Green in a series of papers in the early '60s. And these were mouse embryo-- embryo fibroblasts that if they were kept at a very sparse culture-- so that if they were passaged 3,000 cells-- 3T-- every three days, they never would have contact with one another, and they would stop growing. If they did have contact with one another, they would inhibit their growth, which was called contact inhibition, which came to be a characteristic of normal cell growth, and which was lost by transformed cells.
And they knew from the basis of their experiments that there in serum were growth stimuli-- the possibility was that they were growth factors in the serum, but they never really looked for them because it was so complicated, that were causing the cells to grow. And now we know, many years later, that there's no such thing as contact inhibition. It wasn't the fact that the cells were contacting one another that stopped them from growing. Rather than when they got to a pretty high density, normal cells would deplete the medium of their growth factors, and then they would stop growing. And because transformed, malignant cells did not depend on the growth factors-- they were growth factor independent-- they just kept on growing.
In 1974, Arthur Pardee did a series of experiments that were seminal in this field. And he coined the term the restriction point, or thee R point, which was a point in the late G1 phase of the cell cycle when you could take the serum away, and the cells would still undergo the transition into the DNA replication phase or the S-phase-- the DNA synthesis phase. Then in 1977, Jack Pledger and Chuck Stiles separated the serum dependent phase into two different phases in G1, one that they called competence, which was the G0 to G1 transition. And they found and showed that there was a platelet extract from blood platelets that would cause this conversion to occur, and then the cells would not undergo progression through G1 into the S-phase until they got something else from platelet-poor plasma, and that was the G1 S-phase transition. So we had competence and progression for the first time.
Then simultaneously in the hematology fields, people were looking at the growth of lymphocytes after Peter Nowell had shown that they could undergo blastogenesis and proliferation. And there were two papers that came out back to back in Nature in 1965, both from the same place at McGill University in Montreal-- one from the hematology division and another one from the experimental surgery department, both showing the same thing. And that was there was some sort of factor that would stimulate DNA synthesis of leukocytes-- white blood cells-- that they could find in the medium, and they called this lymphocyte-stimulating factor. They called it blastogenic factor in 1965, and that was probably-- that was the first cytokine or lymphokine discovered and reported upon.
Having some problems here. Also about the same time in 1965 and '66, Leo Sachs in Israel and Don Metcalf in Israel were looking at the growth of hematopoietic precursor cells in soft agar. So that they would let them grow for a week or so, or even 10 days to two weeks, and they would see macroscopic colonies. So they talked about colony-stimulating activity, which they got from a lung or macrophage condition medium.
Then in 1974, Cline and Golde found that if they used PHA-lymphocyte conditioned medium from growing lymphocyte, which they called LyCM, they could also get hematopoietic growth factors out of. And then in the mid '60s, there were several groups that reported that you could get long-term lymphocyte growth in suspension cultures, as long as you continued to add allogeneic lymphocytes in a mixed lymphocyte culture. And up until this time, people had only been able to grow lymphocytes for a few days.
Then in 1976, Doris Morgan, working in Bob Gallo's laboratory at the NCI, was looking for a leukemia cell growth factor, and because of these earlier reports of hematopoietic growth factors, she took PHA lymphocyte condition medium and she found that she could get normal T cells to grow long-term up to 13 weeks, as long as she kept them in this condition medium. However, because she had used leukemic bone marrow for the source of the cells, there was a possibility-- it was probable, actually, that the cells that were growing were really immature T cells and not the mature T cells. And most of the immunologic community really ignored this paper.
At the time, I was trying to generate cytotoxic T cells that would lyse specific leukemia-specific cells. And so once I found out about Morgan's report, I thought well, perhaps we could use this condition medium to not only just grow T cells, but to grow antigen-specific T cells. And the very first experiments worked, and this is a picture of what the cells looked like. They were large, blastoid cells filled with lots of granules that turned out ultimately to be cytotoxic granules. And we reported this in Nature in July 14tH-- Bastille Day in 1977.
So at the time-- I should go back one slide. This was also against the dogma, because the dogma was that the only thing that would make lymphocytes, particularly T cells, proliferate was antigen itself. And so that became the question. The question was-- was antigen the thing that was causing these cells to proliferate long-term, or was it was it T cell growth factor, which we called our factor in 1978.
And I felt that in order to be able to answer this question, we really needed to do reductionist science, and we needed to clone the T cells-- make monoclonal T cells so that we didn't have to worry about cell-cell interaction between different cell types. We needed to generate monoclonal antibodies to the T cell growth factor, which allowed us to purify to homageneity and then rename it to interleukin 2, as I'll get into. And then we needed large quantities of pure molecules that we could use in our experiments, and we needed to find out how the growth factor caused the cells to grow, and that led us to the discovery of the first cytokine receptors-- T cell growth factor receptors.
Now, this is a slide from our paper in 1979 where we first showed monoclonal cytotoxic T cell lines, and these are cytotoxicity assays, where we have the percentage of cytotoxicity on the y-axis, and then increasing ratios of killer cells to target cells. And at the time, we generated our killer ourselves against allogeneic tumors, or histo-incompatible tumors, and they had allogeneic specificity in terms of their killing, but they would also kill syngeneic tumors. And so we felt that if we could clone these cells and devise the asexual progeny of single cells that we would get cells that would only recognize allo antigens and other clones that would only recognize tumor-specific antigens, and that was exactly the case.
And that took the ambiguity and the heterogeneity out of the cell population. And then we used these cells to generate a bioassay for the growth factor. And with that bioassay, we were able to screen for monoclonal antibodies reactive to the growth factor, and those experiments all worked wonderfully, as well. And this is just some data from our very first paper, where we had purified interleukin 2 to homogeneity using the monoclonal antibodies, and then we analyzed it by HPLC and found a single peak denoting that the fact is that it was a single protein. And we also did SDS-PAGE gels and found the molecular size of about 15 and 1/2 thousand daltons, and it allowed us then to do the N-terminal sequence, proving that we had homogeneous molecules uncontaminated by other molecules.
Now, that was very important because the next thing we wanted to do was to radio labeled the growth factor, which we did with s-35 methionine by adding the radio labeled isotopic labeled methionine into the culture medium. It was incorporated into newly synthesized protein, and then we used our monoclonal antibodies to purify the protein. And here you see a single radio labeled protein on an SDS-PAGE gel. And when we use this to do binding assays-- the radio receptor assays, we found the characteristic data showing that there was saturation of a single species of receptor, which then led to a straight line Scatchard plot denoting the fact that we had a very high affinity and saturable receptor that was stereo-specific only for interleukin 2 or T cell growth factor.
Now, with a radio labeled growth factor assay in hand, then the next thing we could do-- we collaborated with Tom Waldmann's group, who thought they had an antibody reactive to the growth factor, and we did a series of experiments that proved that they actually did. That antibody was called anti-TAC, and this is the first flow cytometry profile that we got up at Dartmouth on using anti-TAC against cells that we thought had IL2 receptors, and these experiments were done by Doreen Cantrell, who was a post-doctoral fellow at the time. And you can see the typical lognormal distribution of receptors in a receptor-positive population. There are some cells with a high fluorescence intensity and a very high density of receptors. And then as you go down to here, there's lower and lower, on the order of two to three orders of magnitude.
Now at the time, we knew that when we did a dose response of the concentration curve of the ligand and then assayed for the cells for the growth characteristics of the cells, that we would see that there were some cells that would seem to be proliferating at very low concentrations of the growth factor, but then as we increased the concentration, we got more and more and more, until we reached maximum. Now at a time, what we didn't know was that if you looked at the half max 50% point right here-- would that mean that all of the cells were incorporating thymidine, which was our assay over here, but only 50% as well at this concentration of growth factor, or whether 50% of the cells had left the G1 phase of the cell cycle and had gone into the DNA synthetic phase.
Now, with a flow cytometer, we could do single cell analysis and quantitative staining with propidium iodide, and here you see a single peak of DNA at this concentration of growth factor, whereas you see two peaks of DNA-- this is the diploid and this is the tetraploid DNA peak here, where we've duplicated the DNA. And so that meant that as we were going up the concentration curve, that at a single cell level, cells were individually making a decision to start to synthesize DNA. But when we were here at very high density of receptors on cells, that we felt that those were probably the cells that were leaving G1 and incorporating DNA at very low concentrations of the growth factor.
And as we then go down here into a lower receptor density, we need greater and greater and greater concentrations of the ligand to fire the response. Now, in order to examine that, we used the fluorescence-activated cell sorter, and we separated-- this is another linear plot of the IL2 receptor antibody, with some of the cells off scale here.
And what we did was we separated the cells into a low density or a high density of receptors. We synchronized the cells by taking them out of the growth factor overnight and then added the growth factor the next day. And then we did rapid pulses of thymidine over the next day or so. And we found that the population of cells that had the lowest density of receptors were the ones that took the longest to undergo the G1 s-phase transition and start to synthesize DNA, whereas the cells that had high densities of receptors did this reaction much more rapidly, and the whole population was about halfway in between.
The other thing that we found was is that it would take-- the timing aspect of this led us to think that perhaps there was some message that was building up in these cells that was causing them to ultimately undergo the proliferative response. And one of the first things we did was to collaborate with Ellis Reinherz and his group at the Dana-Farber, because he had used our technique of cloning T cells to derive the first human T cell antigen-specific clones, which he then raised monoclonal antibodies to that reacted to the T cell antigen receptor. And so he had different clones and different antibodies that we could use to find out what was really the stimulus causing the proliferative response.
And so we could use macrophages with the specific antigen on the cell surface, or he could use his monoclonal antibodies reactive to specific clones, activated the cells, and what they did then was they changed and they started to release interleukin 2 in these dots here. And also at the same time, they started to express on the cell surface interleukin 2 receptors. And it was the interleukin 2 receptor interaction that caused the cells to replicate their DNA and undergo cytokinesis into two daughter cells.
So this looked just like the Pledger and Stiles competence and progression that they'd found in fibroblasts. A few years later with Julie Stern as a post-doctoral fellow, we did a series of experiments that really showed us that this was the case. And what she did was she took lymphocytes and grew them up into larger quantities, and then took them out of the growth factor so that they would synchronize into small cells and then restimulated their T cell receptor here at minus 12 hours. And then at zero hour, she added back interleukin 2 and then looked at the cells under the microscope at 5 hours, 10 hours, and 20 hours.
And you can see that at the beginning, they're all small lymphocyte. And even at 5 hours after interleukin 2, they were still small lymphocytes. But by 10 hours, we saw the big lymphoblasts that Peter Nowell had seen back in 1960. And by 20 hours, now we start to see mytosis.
Now, if we label the cells with pulses of thymidine or pulses of uridine and follow them over time, if we did not add the interleukin 2 at time zero, the cells did not incorporate any of the nucleic acid precursors. However, if we did in the solid symbols here, the uridine incorporation into newly synthesized RNA preceded the change and the incorporation of thymidine, so that the uridine is showing us the same thing that the microscope did, which was the cell gets big and blastoid before it starts to replicate its DNA.
And this is just a pictorial representation of that. When we start here on the top with the T cell receptor activating small lymphocytes with antigen, what happens is the cells then start to produce interleukin 2 in the green squares and also express interleukin 2 receptors. And that interaction then causes the cell to undergo blastogenesis and DNA replication. So now we have two chromosomes in this cell, pictorially. And then the cell undergoes cytokinesis into two daughter cells. And in the cell cycle terms, we're starting with cells that are in G0 and that enter G1 because of the T cell receptor stimulation and then progress through the cell cycle to the s-phase because of the cytokine interaction, and here is the restriction point in late G1.
Now, the same thing happens in the other hematopoietic cells, and here I've depicted myeloid precursor cells with a hematopoietic stem cell here, which is the counterpart to the resting lymphocyte before the T cell receptor activates it, and we don't know what the receptor is for hematopoietic stem cells yet. But what we do know is that once the cell makes this transition from G0 into early G1, it is now susceptible to different growth factors, and here I've placed C-CFS-R-- Granulocyte Colony-Stimulating factor. They express the receptors, and then the cytokine receptor interaction causes the blastogenesis, the DNA replication, and the cytokinesis, so it's a very similar process.
And all of those data over the course of more than 20 years led me to theorize a new way that the immune system works, which I called the quantal theory of immunity. And this theory simply states that the cells have the ability to count the number of receptor hits that they get, and it accumulates until it gets to a critical number of receptor hits. And then leads to a quantal, which means in cell cycle terminology all or none-- the cell either does it or does not have a cellular response.
And depending on which receptor you're talking about-- if you're talking about the T cell receptor, which causes competence, you get an all or none expression of the IL2 genes and the IL2 receptor genes as a consequence of triggering the T cell receptor. When you trigger the IL2 receptor, you get progression through G1 into S beyond the R point, and then you get DNA replication. And this quantal theory has been published in several papers and finally, a book last year published by World Scientific Publishing from Singapore.
Now at the time that I was thinking about these theories, I was reading a new biography about Albert Einstein, and I came across this statement of his. He says, "a theory is more impressive the greater the simplicity of its premises, the more different things it relates, and the more expanded its area of applicability." And so Einstein has really set the bar pretty high in terms of claiming that you have a very good theory.
And I felt at the time that-- I wondered whether or not cancer-- what we were talking about when we look at cancer-- whether it's really the loss of this quantal cytokine receptor growth control, and the cell undergoes a transition or a transformation whereby it doesn't need the cytokine or the cytokine receptor interaction anymore. Well, in order to delve into that question, there had to be a series of very important discoveries that were made over the course of 20 years, 25 years. The first one was about the theory about cellular proto-oncogene that Temin had hypothesized back in 1971.
Peter Nowell in a review in Science in 1976, and 16 years after he had made his groundbreaking discoveries of lymphocyte blastic transformation and the Philadelphia chromosome, he hypothesized that what was going on, using CML and the Philadelphia chromosome as his model, that there were mutations that were occurring-- spontaneous somatic mutations that would lead to escape-- somehow allow the cell to escape from normal growth control, that would give it a Darwinian selective growth advantage. And then over time, additional mutations would be selected for that would accelerate the whole process.
And, of course, nowadays we know that this is exactly what the case was, but back in 1976, I think most people ignored Peter Nowell's hypothesis, because right at the same time in 1976, Mike Bishop and Herold Varmus found the first evidence for cellular proto-oncogene that Temin had hypothesized five years earlier. And what they did was-- this is before DNA cloning and sequencing. The only way where you could detect homology between genes was to do kinetic hybridisation, and so they took the viral DNA from Roul's sarcoma virus and chopped it up into smaller pieces, and then looked in normal genomic DNA of normal chickens to see whether or not they could show any hybridization. And they did, and they had a paper out in Nature, and that was the first demonstration of, really, a cellular proto-oncogene that was incorporated into a retrovirus-- an RNA tumor virus. And that gene then became src-- what we call now src these days.
Now just a few years later-- so this set everybody off looking for more oncogenes in retroviruses. They had all these retroviruses. And for the next decade, people were taking apart the viruses looking for more oncogenes.
However, in 1979, Bob Weinberg and Jeff Cooper in 1980, did some very seminal experiments where they didn't have to use any viruses. They would take tumor cells from human tumors that had no evidence of any viruses whatsoever. They would grind it up and give the RNA to 3T3 cells from the experiments of Greene and so forth. And they found that there was something in normal DNA-- our cellular DNA from tumor cells that would transform these cells.
And then in 1978, Erikson and 1982 Owen Witte, found that the src gene first was a protein tyrosine kinase. That was a very important discovery. And then Owen Witte found that the Abelson gene from the Abelson leukemia virus was also a protein tyrosine kinase.
Then in the early '80s, Grosveld working at the NCI-- now that you could clone DNA and sequence DNA, he sequenced through the Philadelphia chromosome, and he found that it was encoding for the c-abl. And the proto-oncogene for c-abl mapped to chromosome 9, the long arm. And On chromosome 22, where it seemed to be having this reciprocal translocation, there was a breakpoint cluster region, so that became known as BCR-ABL.
And he found that the BCR-ABL region encoded a larger product of 210,000 daltons that coded for a constitutively active protein tyrosine kinase, whereas the normal c-abl gene only codes for a 145 kDa protein, and it's inactive unless activated in a normal process. This was extremely important because of the fact that this was the first animal virus oncogene that had been found, and it would have been found to be involved in a genetically altered human cancer, connecting the two fields for the very first time.
And it wasn't too much longer before people took the Abelson virus and added that to IL 2-- interleukin 2-dependent antigen-specific long-term T cell lines, and they were able to get transformation of these cell lines so that they were no longer dependent on interleukin 2. Here's the normal cells that with an increase in concentration of interleukin 2 and then thymidine incorporation, and you can see that as you add more IL2, you get more DNA replication. And if you had transformed the cells first with Abelson virus, they were growth factor independent. They didn't really need the growth factor anymore.
Exactly the same thing was found with interleukin 3-dependent cell lines, as well as GM-CSF-dependent cell lines. And these cytokine-dependent cell lines then became the workhorse-- essentially, the new 3T3 cells-- looking for things that would transform these cells so that they were growth factor independent.
Now, we still didn't understand how it all worked, however. And the next thing that happened was the discovery of new proteins called cyclins and also other data coming from the cytokine field. In the early to mid '80s, Tim Hunt, John Ruderman, and David Beach found proteins that were cyclically expressed during a synchronized growth of first, sea urchin eggs and then clam eggs that had been fertilized, and then yeast. And that set everybody in turn to try to find the same kinds of cyclic proteins in mammalian cells.
And in 1991, Sherr and Ruderman and Peters and Meeker all found evidence for mammalian G1 cyclins, which were called the D cyclins. And there are three different varieties-- 1, 2, and 3. And then in 1993, Julia Turner, who was a graduate student with Tim Hunt at Cambridge, collaborated with-- Doreen Cantrell had gone back to England, and Doreen taught Julia Turner how to culture the lymphocytes. And she was able to show very, very nicely in a seminal paper that IL2 induces the expression of cyclin D2 and D3 messenger RNA.
And so this is a northern blot from this paper that was published in International Immunology. And what she's done is synchronized the cell as Doreen taught her, and then added back to the interleukin 2 in different hours afterwards, would harvest aliquots, and then do northern blots for cyclin D2, D3, cyclin A, and this is a loading control for MHC. And you can see that as the cells progressed through the cell cycle that there is a sequential expression of first cyclin D2 and then D3, and then during the s-phase, cyclin A.
So it was clear that there that cyclin Ds looked to be the intracellular sensor of what was going on at the cell surface, but exactly how the message was being transferred was really unknown. And another series of proteins and genes needed to be discovered, and they were the JAKs and the STATs. And in the late '80s and early '90s, Wilks in Australia Krolewski here in New York at Columbia discovered four new enzymes that they call the Janus kinases 1, 2-- actually, Wilks discovered 1 and 2, and Krolewski another one that he called tick to and then Tyk-2. And then the next year, John O'Shea, and Jim Ihle found JAK3 and showed that JAK3 was involved in interleukin 2 signaling. And so that there were four gene products that contained two kinase domains, and that's oriented oppositely, and that's why they named them Janus kinases, for the January or the Janus Roman god.
While this was going on, Jim Darnell at the Rockefeller was looking about how interferon activates new gene expression and identified four new proteins that he labeled STAT 1 through 4. And STAT stands for Signal Transducer and Activator of Transcription. And then Groner in Germany found STAT5 induced by the hormone prolactin. And then McKnight, in 1994, found STAT6 that was found to be induced by interleukin 4. And then Pelligrini and Stark and Kerr and Darnell, in a series of very elegant experiments, really showed the JAKs activated the STATs.
That all happened very fast in the first half of the '90s. And one of the experiments that was done early on was done by Carol Beadling, who was a graduate student from my laboratory who went to work with Doreen Cantrell in England. And they collaborated with Jim Ihle, because he had antibodies to the JAKs and the STATs, as well as Ian Kerr at the ICRF in London.
And they wanted to know whether or not the T cell receptor would activate the JAKs and the STATs, or whether only interferons or cytokines would do that. And they found, as in the title of this paper, that the JAKs and the STATs were only activated by cytokines, IL 2, and interferon in this particular paper, but not by the T cell antigen receptor in human T cells. And this is just a Western blot showing JAK1 activation by IL 2 and interferon, but not the T cell receptor. And then these are electrophoretic mobility shift assays for the response elements for the STATs that happened after IL 2 stimulation over time, but not after the T cell receptor stimulation.
So the story was almost complete at this stage, but we still didn't know the exact series of events. What was carrying the signal through the cell? And then what happened were the discovery of Protein Tyrosine kinase inhibitors.
The first was reported in 1986 by a Japanese group, who found a natural product that would inhibit in micromolar concentrations the phosphorylation of the EGF receptor in vitro. And then Levitzki from Israel, in 1988, found additional EGF receptor inhibitors. Again, at the micromolar concentration range they would block the enzyme in vitro, but he also showed that they would block the proliferation of the cells. And then in 1993 in Science, he showed that he found BCR-ABL inhibitors in micromolar concentrations.
Now, a micromolar is a pretty low concentration-- 10 to the minus 6 molar, but it's high enough so that you can get a lot of nonspecific effects. And to really get a drug, you need nanomolar-- you need a drug in a nanomolar range, and that was accomplished by Brian Druker, who reported in 1996. Ciba-Geigy had had a drug discovery program, and it found compounds, but they really put them on the shelf, and they didn't know what to do with them.
Brian, who was trained at the Dana-Farber in kinases, he knew what to do with them. And he showed that these compounds, which are now called G0ec, would block the tyrosine kinase activity at a 25 nanomolar concentration and block the proliferation of the cells at a 250 nanomolar concentration. And this is just a picture of the molecule that Brian published in 1996, and then two different concentrations blocking the viable cell proliferation of a BCR-ABL transformed cell line.
So it was clear that the protein tyrosine kinase inhibitors were blocking the proliferative response of the leukemia cells, and it was shown that they would block the phosphorylation of many, many proteins. And the question is, was it just one protein that was being blocked that was important, or many proteins? And the answer was really found in the cytokines.
There was a classic paper now that was published in 1999 by Richard Morriggl and Jim Ihle's lab, where they knocked out the STAT5 genes. There's two different genes that are 95% homologous-- STAT5A and STAT5B And when they did this, they got mice, and they got cells that they could get from these mice, but when they activated them in vitro with anti-CD3 and added interleukin 2, they didn't proliferate. And that was very, very important.
Then in 2001, Brad Nelson showed that STAT5 bound to response elements upstream of the cyclin D2 gene, and so that connected everything from the receptor at the cell service, through the JAKs, into the STATs, and into cyclin D2. Then in the very next year, Carol Beadling, back in my lab, showed that cyclin D2 is an immediate early IL 2-induced gene, and that really wrapped up the story at that point. This is the 1999 paper from Morriggl and Jim Ihle's lab, where they put cells from these STAT5 knockout animals in vitro and activated them with the anti-CD3, the signaling component of the T cell receptor.
And you can see that the wild type or heterozygote cells proliferate quite nicely. This is thymidine incorporation, whereas the knockout cells do not. If you add IL2, it really doesn't cause any proliferation. And then if you looked at the expression of the genes important for cell cycle progression in the knockout-- cyclin D2, D3, cyclin E, cyclin A-- the knockout basically-- no matter what kind of a stimulus you gave them, you did not see any gene expression, whereas you do, particularly when you have anti-CD3 and IL 2 in terms of the cyclins.
So that was consistent with the fact that this whole process was very important for not only normal T cell proliferation, but also for leukemogenesis and the BCR-ABL effect. And there was a series of papers that were published between 1995 up to 2002, where they all suggested the same thing, and that was you could get malignancy if you had something wrong with a cell such that you got persistent STAT activation. The first experiments were done by Richard Jove at the Rockefeller, and then Jackie Bromberg and Jim Darnell's lab at the Rockefeller, and they found that the src kinase phosphorylates and persistently activates STAT3 in transformed cells.
Then Paul Rothman at Columbia found the same thing when he transformed cells with v-abl. It persistently activated STAT5. And then Bernard and Marynan showed another translocation TEL with JAK2 persistently activated STAT5. And then Skorski found that the BCR-ABL did not do this directly, but rather, it activated a src family kinase, which then persistently phosphorylates and activates STAT5.
And so this seemed to be the answer to cancer. However, Veronika Sexl, working with Jim Ihle and her husband Richard Morriggl, wanted to find out whether v-abl would transform lymphocytes from these mice that had STAT5 knocked out. And basically, she found that the answer to the question was yes. Grossly, as well as microscopically, v-abl could transform these cells.
And that really threw a monkey wrench into the whole process, and people started to wonder what was really going on. And there were papers that were published that they hypothesized that the activation of STAT5 was the result of malignant transformation, not necessarily the cause of the malignant transformation. Well, it turns out that the Stat5 knockout was not a complete knockout. It was only 99% complete. And it was found in the intervening years between 2000 and 2006 that there's still some residual STAT5 protein.
And so Veronika Sexl went back to Vienna and started her own laboratory and knocked out the entire STAT5 genes and then repeated the experiments. And these are in vitro transformation assays. This is a wild type, so that you can see small colonies here, and here's a close up of one of the colonies of transformed cells. This is the heterozygote, and here we have the complete knockout, and there were no colonies. And she also did exactly the same experiment in-vivo. Vivo.
So here we are, up to the present time in 2011-- 50 years later. And now we know that what cancer is is the mutation-driven loss of normal cytokine receptor quantal cell self-growth control. And if you're interested, I published that in 2008 in the Journal of Clinical Investigation, along with Jim Griffin from the Dana-Farber at Harvard.
If you look back, then, at what the critical aspects of this whole process is, it's clear the competence and progression are two different molecular events in cells that cause two different changes in the cell. As I've said, the cellular competence of the change from G0 into G1 is triggered by the T cell receptor, but the proliferative response-- the replication of the blastogenesis and the replication of the DNA and the cytokinesis is all caused by the growth factor, so that the important cell growth aspects of normal cell growth control is being controlled by the cytokine factor family. And in fact, if you look at the structure of the T cell receptor and the interleukin 2 receptor, you find that they're quite different. The proteins that form the antigen receptor are different. They activate different enzyme, different kinases on the inside of the cell, and they cause different sets of transcription factors to migrate into the nucleus, and they cause different genes to be activated. The most important in T cells is the interleukin 2 gene and the interleukin two receptor gene.
By comparison, if you look at the IL 2 receptor, it has three chains-- the alpha, beta, and gamma-- that complex with two different JAK species-- JAK1 and JAK3. And that when you add the ligand, it brings all three of these proteins together, and that brings the JAKs together, which start to transphosphorylate one another. And once they do that, they phosphorylate certain residues on the beta chain of the IL 2 receptor, which then serves as a dockings side for phospho-STAT5. Once STAT5 becomes phosphorylated, it dimerizes, translocates to the nucleus, and activates the expression of cyclin D as a cell cycle progression gene and Bcl-X, for example, as a cell survival gene.
And exactly the same thing happens with the hematopoietic factors-- the G-CSF or granulocytes, the epo receptor for red blood cells, and the thrombopoietin receptor for platelets. The difference is that this is only one chain, and it dimerizes-- it makes a homodimer with a ligand. And it only complexes with one JAK-- happens to be JAK2, but then it phosphorolates the receptor chain, which serves as a docking site for phospho-STAT5 step five again, which then translocates to the nucleus and does exactly the same thing that IL 2 does.
And so what's going on in BCR-ABL is that BCR-ABL is activating, not through JAK2, but through HCK, which then causes the phosphorylation of STAT5, and it circumvents the whole cytokine receptor process. Now, just to bring it back to the lymphocytes again-- just within the past few years, here in this paper that was published in 2009 by a group from Belgium, acute lymphoblastic leukemia-- they found JAK1 mutants that are activating mutants. They activate the JAK-STAT pathway, and they use the interleukin 9 receptor homodimers to dock on and activate STAT5.
Then another paper now, again, from Veronika Sexl's group, together with her husband Richard Morriggl in Vienna, that basically proved that STAT5 activation is indispensable for the maintenance of BCR-ABL positive leukemia. And then just recently this year and just last month in the Journal of Experimental Medicine, gain-of-function mutations in interleukin 7 receptor alpha in childhood acute lymphoblastic leukemia. And then I think this is a very, very important paper from David Frank's lab just recently in March in Blood, where he's found a STAT5 inhibitor-- a small inhibitor called pimozide, which decreases the survival of chronic CML cells that are resistant to kinase inhibitors. So this is just the beginning of a whole new series of compounds that are going to be targeted onto the STAT molecules, and in particular STAT5.
So after 50 years, we can look back over time and say, OK, so what were the critical contributions in this whole process? Well, it's clear that Peter Nowell's discovery of the Philadelphia chromosome and lymphocyte blastic transformation were absolutely seminal. The oncogenic retroviruses-- the RNA tumor viruses and Temin's provirus hypothesis was important, not because RNA tumor viruses are important for human cancer-- they're not, but because it showed us what some of the genes were that were involved in the oncogenic process.
And the mutation of cellular proto-oncogenes-- Weinberg and Cooper, and then Temin's protovirus hypothesis were very, very important. And then Pardee's restriction point, the competence and progression coming from the 3T3 cell scientists was very important, because it pointed out the differences in terms of G0 to G1 transition versus G1 to s-phase transition. And for sure, the cytokines-- if we had not discovered the cytokines, their receptors, the signaling pathways, that JAKs and the STATs and the genetic programs that the cytokines activate, we really would have many pieces missing in the puzzle.
The idea that there's quantal control-- all or none control over the decision of the cell either to proliferate or not to proliferate is very, very important. And it's my feeling that it's not just T cells and CML cells that behave in this fashion, rather all cells are operating under the same quantal growth control of individual cytokines and their receptors. And what happens is that mutations occur that give the cell a cytokine independence to the proliferative pathway. The protein tyrosine kinases and their inhibitors, Brian Druker in particular, was just absolutely very, very important for our current understanding of the whole process.
So what about the future? The future is bright for the first time in 50 years-- first time in 100 years, really. It's clear that we need more new drugs, more new Gleevecs to target the final cell cycle pathways. We need drugs that are anti-JAKs, anti-STATs, and anti-cyclins or cyclin-dependent kinases, and that's going to be the combination chemotherapy of the future, with targeted drugs that target each of these molecules that's in the growth proliferative pathway.
We need to define our diseases, not by microscopy, which we've been doing for 350 years. We need to define it by the molecules, by the intracellular signaling pathway abnormalities. And for example, breast cancer that is due to problems in the prolactin pathway-- persistently activated STAT5 is very important. Same thing with lung cancer and colon cancer, because the EGF receptor hooks up with STAT5. And so STAT5 and also STAT3 are going to be, really, the nodes in this whole pathway.
We also have to begin to start talking about diseases as to their drug sensitivity in the same way we talk about bacterial infections-- as mold methicillin-resistant staph aureus, for example. We already have Gleevec-resistant CML. We need to be able to talk about them in that fashion. We need to be able to diagnose that.
And the other thing is is we need to be able to diagnose early, after the first mutation. And Peter Nowell told us this in 1976, when we first see the Philadelphia chromosome. But we need new biomarkers to be able to identify. We were lucky with the Philadelphia chromosome, because you could see it under the microscope.
But we need to make the transition from microscopy at the micromolar level down to the molecular level at the nanometer level-- 1,000 fold smaller, and we need molecular markers of these different mutations. And once we have them, we need to treat it very early after the first mutation, before the cell has accumulated that second and third and fourth mutation and now is a very virulent malignancy that's unresponsive to all therapies. And I think that what the importance is for the story that I've told you this afternoon is that this is a paradigm for all cancers.
This is not an isolated case of just CML. This is going to be the prototype for all cancers, and now that we know this, we can proceed with-- we can take the blinders off, and we can see exactly what's going on. And so I'll stop there, and thank you for your attention.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
Dr. Kendall Smith, the Rochelle Belfer Professor of Immunology and Medicine at Weill Cornell Medical College, reviews the progress that has been made in the past 50 years in our understanding of the molecular control and regulation of normal cell proliferation, and how this understanding has led for the first time to a new insight into the molecular pathogenesis of abnormalities of cellular growth regulation, a.k.a. cancer.