42nd Annual Killian Award Lecture—Stephen Lippard

STEVEN HALL: Good afternoon, and welcome to the 42nd Annual James R. Killian Jr. Faculty Achievement Award Lecture. I'm Steven Hall, chair of the MIT faculty. And before we begin, I have two requests. The first is to make sure that your cellphone and any electronic devices that you have are turned off. And the second is, it is crowded and there may be some late arrivals. So if you're at the end and there's seats in the interior of the row, if you could please move in, that'd be greatly appreciated.

The James R. Killian Jr. Faculty Achievement Award was established in the spring of 1971 as a permanent tribute to Dr. James R. Killian, president of MIT from 1948 to 1959, and chairman of the corporation from 1959 to 1971. The purpose of the Killian award is to recognize extraordinary professional achievement by MIT faculty members, and to communicate their accomplishments to members of the Institute community.

The title of Killian Lecturer is the highest honor that MIT faculty may bestow on a colleague. It is our privilege today to be addressed by this year's award winner, Stephen Lippard, the Arthur Amos Noyes Professor of Chemistry. As stated in the citation from the Killian award committee, Professor Lippard is widely acknowledged to be one of the founders of the field of bioinorganic chemistry, a field concerned with the role of inorganic molecules, especially metal ions in their complexes and critical processes in biological systems.

His pioneering contributions include understanding the mechanism of action of the clinically important anti-cancer drug cisplatin in the rational design of new variants to combat the problems of cytotoxicity and resistance. Among many achievements, Professor Lippard's groups discovered the first metal complexes that interpolate into DNA.

Professor Lippard completed his Ph.D. in chemistry at MIT in 1965. And shortly thereafter, joined the faculty of Columbia University in 1966. He's been a member of the MIT faculty since 1983, during which time he has trained many young scientists. He's a teacher of freshman chemistry, and has mentored more than 100 Ph.D. students and an even greater number of post-docs.

He served for 10 years as head of the Department of Chemistry. Professor Lippard has published more than 800 peer reviewed papers, recorded nearly 30 patents, and co-authored the definitive textbook on bioinorganic chemistry. Professor Lippard has received numerous awards for his achievements, far more than I can list in this short introduction. I'll mention just a few.

He's been elected to the National Academy of Sciences, the National Institute of Medicine, and the American Academy of Arts and Sciences. He was recently awarded the 2014 Priestley Medal of the American Chemical Society, which is that society's highest award. and in 2004, he received the National Medal of Science, the nation's highest honor for scientists and engineers. So it's with great pleasure that we recognize Stephen Lippard today with the James R Killian Jr. Faculty Achievement Award.

[APPLAUSE]

We have a certificate of citation for you. And let me read the citation. This certificate is in recognition of groundbreaking work at the frontiers of inorganic chemistry, leading to improvements in human health and conquering of disease. And for having taught, mentored, and trained a generation of scientists at MIT. So congratulations.

STEPHEN J. LIPPARD: Thank you.

[APPLAUSE]

Thank you very much. It's a great honor for me to be here. It's actually very humbling to be elected by one's peers to this honor. It's an unbelievable list of MIT faculty, past and present, and men and women. And it's just great. Some of them are in the audience. And so it's a special pleasure to be here.

I was a student, a graduate student, when Dr. Killian was chair of the corporation. And I did know him. I guess I've spent more time in Killian Hall listening to music than I actually spent with him when I was a student. But it's just a wonderful honor and I'm greatly appreciative. And I thank all of you for coming out to listen. So I also want to thank my sons, Josh and Alex, who've come up to be here.

So we work at the interface of inorganic chemistry and biology. So metals and biology, the two major areas here. One is the understanding of existing metals which function in much of biology. And the other is the introduction of metals as probes and also as pharmaceuticals.

And it's in the last area that I'm going to be discussing exclusively today, talking about understanding and improving platinum anti-cancer drugs. And this work has been supported for a very long period of time by the National Cancer Institute, to whom I'm very grateful. And also in part by the Koch Institute, which we have been an associate member of for some time.

So there are basically two major themes I'm going to talk about today. One is understanding, the extent to which we understand something about platinum compounds. And I'm going to talk about the discovery that platinum compounds can cure cancer. Not something that's particularly an obvious thing.

DNA is the cellular target. And the structure of the adducts that platinum makes on DNA. As well as the functions that are affected. And from that understanding, the ultimate goal was to try to see if we could make improved versions of the compound. And I will talk about dual threat, so-called dual threat constructs to target cancer cells and to try to get around some of the problems that one sees in conventional chemotherapy.

Talk a bit about the strategy for delivery of these compounds in vivo. And then talk about a new class of so-called monofunctional functional compounds, which we think have particularly important potential for drug development. And I'll say a bit more about how we intend to go about that in the near future.

So the discovery that platinum compounds can cure cancer came about when a biophysicist by the name of Barnett Rosenberg, working at Michigan State University, carried out an experiment in which he put E. Coli cells growing in culture in electric fields. And the reason he did that was because the lines of force of electric fields reminded him of a high school biology textbook picture of cells dividing.

So we have physics and biology. I've always said that's kind of a dangerous combination. But understanding requires chemistry, so I have to give a little call out to my colleagues from the department. So here's a picture of Barney. And actually, what happened was, here is electron micrograph of the E. Coli cells growing. And here's what happens when you turn on the electric field.

And it was extremely puzzling. And so after a couple of years, they finally discovered that it had nothing to do with the electric field, this filamentous growth. But a small amount of platinum that was used to supply the electric field leached off the electrode and reacted with ammonium chloride in the buffer, forming this very simple 11 atom compound. The cis-diamminedichloroplatinum(II), later called cisplatin, has no carbon atoms in it, no honorary carbon atoms. And it turns out, that compound is a cure for certain forms of cancer.

And this is the experiment that Rosenberg did. He tried to get the National Cancer Institute and others to carry out trials, human clinical trials, but was very unsuccessful with that until he produced data of this kind. He took a mouse, implanted a sarcoma, and a tumor grew. After three weeks, the mouse died without any treatment. But a single injection of the cis-isomer of diamminedichloroplatinum produced significant necrotic tissue. After about two weeks, the tumor became smaller. It was groomed off and the mouse ultimately lived a normal lifespan.

At that point, the National Cancer Institute got very interested. And it supported a wide range of clinical trials across the country, where they worked on safety, efficacy, and looked at particular types of tumors that were responsive. And of course, cisplatin itself was a known compound. It had been around for quite some time. It was discovered by Pyrone in the mid-19th century. This 1965 discovery of cell division that led to the antitumor activity in 1969 was followed up by FDA approval only about 10 years later.

Now, it is to Rosenberg's credit that he made the connection between filamentous growth and antitumor activity. He knew that agents that caused DNA damage did bring about filamentous growth. That is growth of the cells without cell division. And that was the reason that he took it into the animal trials.

And so now, by today, cisplatin and two derivatives, one called carboplatin, where the chlorides are replaced by the cyclobutanedicarboxylate. And this compound, known as oxaliplatin, where the amines are replaced by the RR isomer-- for the organic chemists, 1,2-diaminocyclohexane oxalate, which is used for colorectal cancer.

These three FDA approved compounds are very common in clinical use. They've been in use for a very long time. And almost all of them are off of patents, so a lot of generics used. And today in 2014, actually, more than half of all cancer patients receiving chemotherapy are now treated with platinum compound, which is an extraordinary thing for an inorganic chemist working at the interface of biology, medicine, and chemistry.

So the question is how do they work? That's a very simple question, but it has a complicated answer. And I would say, like any mechanism in science or in chemistry, you never really prove a mechanism. You just come up with an idea that tells you how it may be happening and wait for somebody to say, oh no, you're wrong. There's another, better explanation. Fortunately that hasn't happened, but if it does, that would also be good fortune as well.

So the focus has been on DNA as the cellular target. And I want to talk a little bit about that and our work on the structure of platinum DNA adducts. So the human cell, from a chemical point of view, is incredibly complicated. Not only because of its different constituents, but because it's constantly changing. And so certain antibiotics, like vancomycin, for example, Merck scientists tell me that this so-called drug of last resort for bacterial, for infection, was actually-- the bacteria became resistant by inventing five totally new biochemical pathways to synthesize themselves around that drug.

And so when we try to understand and begin to look at how any outside chemical can work in this environment, it's a daunting task I found out. I didn't know that at the beginning. We had done a lot of work in the interaction of metals with nucleic acids, because we had spent some years-- and this is when I started at Columbia-- trying to sequence DNA in an electron microscope by attaching it to base-specific sites. Which we succeeded in doing by putting sulphurs at base-specific sites, quantitatively in long stretches of DNA. 1,000 to 10,000 in length.

But we couldn't actually read those beads on the string in the electron microscope because a little too much energy caused blurring of the images, although we certainly tried hard. But with that background in DNA, I thought this would be an interesting problem to study. And I want to start by giving you a little bit of an overview of what happens, and then I'll focus specifically on the DNA.

So there are really, overall, four stages, broad stages, in the cisplatin mechanism of action. One is passage through the bloodstream. It's given as an intravenous injection. And it has to encounter the tumor cell microenvironment and then get into the cancer cell, which it does by passive transport, and possibly by piggybacking on the copper transport. There's reasons for that which I won't have time to talk about today.

But the key thing is that in blood, where the chloride ion concentration's about 0.1 molar, the chlorides are kept on to some extent. Whereas in the cytoplasm, the chloride ion concentration is much smaller. There's an activation and they get lost in a stepwise fashion to produce cationic platinum compounds, which go to DNA, and bind to the strongest electron donating ligands in the DNA, which are the N7 positions of the guanine and adenine basis, forming cross-links because there are two places that it can bind. And then there's a cellular response, which I'll talk about later. And that's incredibly complicated. Thousands of papers a year talk about the cellular response to platinum compounds.

Now, why do we think DNA is the target? Well, for one thing, if you take cells of diminished ability to repair DNA, they're hypersensitive to platinum compounds, particularly cisplatin. And if you want to talk about going all the way to patients, patients who have received platinum infusions as chemotherapy have significantly higher levels of platinum DNA cross-links on the so-called buffy coat, or DNA-containing cells of blood as determined by antibodies that are specific for those particular adducts.

And patients that do not have high levels of platinum DNA adducts do not respond very well to platinum chemotherapy. So all the way from the cell on into the patient being treated with platinum, there's very strong evidence, I think, that DNA is the right target.

Now, we and others looked at some detail at how the adducts on DNA were made, and in particular, the relative amount. So about 90% of all adducts of cisplatin are intrastrand GG cross-links. So if we talk about DNA as having a Watson strand, or a Crick strand, the adduct is making a cross-link on the Watson strand. And the 1,3 cross-link with an intervening nucleotide-- so two guanines with an intervening thymine, maybe about 5% or so, and a smaller number of interstrand cross-links where one links between. So for cisplatin, for example, these are some numbers. And for the carboplatin, they're slightly different.

So the first thing we wanted to do as chemists was to learn something about the structure of these adducts. I mean, if we were the size of the atom, this would all be a visual art. But a lot of chemistry is directed at learning the three dimensional architecture of the micro world and smaller. And so the first thing that we were successful in doing-- this was the work of the graduate student Suzanne Sherman in 1985-- was to synthesize and crystallize deoxy pGpG, a dinucleotide, where platinum binds the N7's, and the two amines are here.

And you can see there's a significant bend at the position of the two guanines, which in a normal double-stranded DNA would be stacked on each other and in a series of parallel planes. And it took us about a decade, because of the difficulty of synthesizing and crystallizing, to get the structure of cisplatin on double-stranded DNA. That was done by Tricia Takahara. And there you see the platinum in the major groove of the DNA where the N7's of the guanine are. Again, making this cross-link.

The picture is very similar to that of the simple dinucleotide. But we can see that it causes a substantial bend in the DNA duplex. And it opens up-- and that's in the major groove of the DNA-- opens up the minor groove. And it opens up a hydrophobic surface to which proteins like to bind.

And even later, we were able to get the structure of the diaminocyclohexane platinum on double-stranded DNA. The diamine platinum is the adduct that's made by cisplatin and carboplatin. With this one, the diaminocyclohexane, we now have the adduct that's made by oxaliplatin. So we now know the structures of the three FDA approved compounds on DNA.

And it was very similar to what we saw for the cisplatin adduct. The one difference being that we could tell why the RR isomer was important in the activity, because there's a specific stabilizing hydrogen bond made with this amino group, with the oxygen atom of the guanine on the 3 prime side of the oligonucleotides on the Watson strand. And in the other isomer, if you have an S-configuration there, you can't make that hydrogen bond.

Now, we come to a much more difficult question, although those were difficult enough at the time. And that is, what are the cellular functions that are affected, and how do those cellular functions lead to the cancer activity, and in particular to the discrimination between normal cells and cancer cells? And a very important paper, at least in my thinking, was one published by Alan Eastman and his colleagues in 1988. It had a couple of papers, actually. Demonstrating that cisplatin arrests the cell cycle at a particular point. That is, just before it undergoes cell division or mitosis.

And so what do he proposed was that it was the inhibition of the transcription of the DNA, that is the conversion of DNA to RNA, which then ultimately makes proteins, that blocks the expression of the enzymes and proteins that are necessary to produce the mitotic spindle. And also, by the way, explains the Rosenberg result of filamentous growth. Although we have to remember that Rosenberg was working on bacteria and cisplatin killing bacteria is not a way to solve cancer. But cisplatin works on humans.

And by the way, I didn't mention-- I meant to mention this earlier. Just to give you an idea of how good it actually is. For testicular cancer, which used to be the leading cause of death of males aged 20 to 40, and would be today, it is essentially a cure. And the poster child is Lance Armstrong. He's not quite as popular a poster child as he once was.

But he had stage four metastatic testicular cancer. It was in his brain. And he had a complete cure, complete remission. Hasn't come back as far as we know. And many, many, many cases like that. In fact, the statistics say-- cancer statistics say there's about 9,000 new cases of testicular cancer each year. And a very small number of deaths-- well under 100.

So what is transcription and how do we think about platinum blocking transcription? And can we actually get evidence for it? After all, the Eastman paper didn't produce direct evidence. It was a hypothesis. And Ryan Todd, a former graduate student, and I described this in a journal known as "Metallomics" recently. And I want to focus specifically on a road block to RNA polymerase.

So here's an example of an experiment carried out by Karen Sandman in the lab, which demonstrates very graphically how transcription is affected. So here we have a construct in which we have two fluorescent dyes-- a coumarin, which fluoresces in the blue, and a fluorescein, which fluoresces in the green. And it's linked by a beta lactam, like the one in penicillin.

And this molecule is protected. And when you put it into a cell, cytoplasmic esterases in the cell will cleave off the protecting groups. And this molecule will fluoresce. So if you shine 409 nanometer light on the molecule, the coumarin will absorb. But then via a process known as FRET, or Fluorescence Resonance Energy Transfer, the photon will be transferred from the coumarin over to the fluorescein, which emits at its own characteristic wavelength of 520 nanometers. And the cells will look green.

Now, if you put into those cells an enzyme known as beta lactamase, that enzyme will cleave the linkage between the coumarin and the fluorescein, so that when you shine 400 nanometer light on the construct, in the cells they'll look blue. But if you have an agent such as platinum, which can block transcription, that is the conversion of the beta lactamase, DNA into the RNA and then ultimately into the protein, it will not be able to cleave because it won't be expressed. And the cells will stay green. And you have enzymatic amplification, which allows one to detect even very low level of gene expression.

And here's the experiment. These are the control cells not treated with platinum. They're blue. Here's cells which are treated with 40 micromolar cisplatin, which are green. And you can clearly see transcription is being affected. In fact, we used this screen one summer to look at the series of potential new platinum compounds that we had made by robotic synthesis, along the lines of what is known to be effective. Namely compounds that had cisamines and good leaving groups.

And I had an undergraduate, a UROP student, who made with that robot about 4,000 compounds over the course of the summer. And my post-doc tested them somewhat manually. And there were four that turned out to be hits. And all four of them had been tested by various drug companies as potential anti-tumor platinum compounds.

Here's another way you can look using an RNA polymerase, an enzyme isolated from T7. And so in this experiment, what we've done is we have taken the RNA polymerase and we have put it onto a magnetic bead. And we have a Watson strand, a Crick strand, and then we have a small piece of RNA as sort of a starting piece. This is one way in which you can carry out transcription.

And then we anneal that to an artificially prepared oligonucleotide-- double-stranded oligonucleotide, containing at a site-specific platinum GG or other cross-link. And so we bring these two together and we ligate them together, paste them together. And now we carry out what's called a transcription walking experiment, where we add nucleoside triphosphates.

If we have a radio label at the starting RNA strand, we can watch the RNA strand be extended, which is transcription. If the platinum blocks the transcription and we run the RNA molecule out on an electrophoresis gel, at the end of the day, we'll see 62 nucleotide piece. So relatively fast moving piece in the gel. Relative to if we had no platinum there, where we're able to bypass the adduct and go all the way to the end, it would have 133 nucleotides.

And here the data. I won't show a lot of data, but this is an example. It's very clean. You can see increasing amounts of the intrastrand cisplatin GG cross-link, or the 1,3 cross-link, or the diaminocyclohexane GG, or 1,3 GTG intrastrand cross-link leads essentially to the complete block of transcription in this particular assay.

OK that's T7 polymerized. What about in humans? What about in mammals? So if we can take mammalian cells, in the mammalian cells, we can put in an expression vector that is a piece of double-stranded DNA which we put into the cell by a process known as transfection, it will go into the nucleus of the cell. And some fair amount of it will actually be transcribed into messenger RNA. And then ultimately translated into a gene that we put into this vector, which is known as Gaussia luciferase, which has the very interesting and nice property that the cells will extrude that.

And if you let that be extruded in the presence of an appropriate substrate, it will turn on a very bright yellow fluorescence. And so you can very nicely see whether or not you're transcribing that particular gene by simply looking with a luminometer at fluorescence that comes out of the cell.

Now, if we repeat that experiment, but we take a platinum compound and we add increasing amounts of the platinum compound to the probe prior to putting it into the cell, what we can do is we can look at cells which are either able to repair or not repair the adduct. Now, I brought a new verb into the discussion, which is repair. And it's a very important verb.

So far I've been talking about transcribing as my verb, but running against that is the fact that cells are able to repair damage on their DNA. And in this particular experiment, we're looking at cells that are either able to repair XPF corrected or not able to repair. And then we actually looked to see whether it mattered whether they were synchronized or not. In other words, were they all in phase in the cell cycle or not? Turns out it doesn't matter, so I won't make any more out of it.

Transfect the probe and collect the cell media at various times. Remember, it's the media around the cells that is going to fluoresce when we put the appropriate substrate on. And assay for this GLuc activity using our luminometer as an auto injector. When I bought that, my post-doc was incredibly happy. It saved him many, many days of work to do the experiments.

And we find the following types of experiments. So if we look at a repair deficient set of cells, XPF-- I'll get to what XP means in a few moments-- and we take, let's say, a certain amount of platinum per plasmid molecule-- let's say roughly 10 platforms per plasmid-- we see that with increasing time-- and so each one of these little diamonds has a different color-- going from an eight hour treatment all the way up to 48 hour treatment, we see that the transcription levels rise as we increase in time.

So here the transcription is knocked out very much after only an eight hour period. But after 48 hours, there's a lot more transcription. And recovery with time indicates that there's repair. Certainly within the first 24 hours. 48, it gets a little more complicated because the cells have divided.

Now on the other hand, if we look at a series of cells which have been corrected, and they're able to carry out nucleotide excision repair, we find that there's a much greater amount of repair. And the recovery with time indicates significant repair compared to limited repair. And that gave us a clue.

We know that cancer cells are inherently deficient in repair. And so already then, is if there is a distinction in the amount of repair, that is a good reason for thinking why, with transcription inhibition versus deficiency in repair, the compound might be active as an anti-tumor agent.

Now, these are experiments where we had the pGLuc templates globally platinated with cisplatin. It just went everywhere. So we didn't always make intrastrand GG cross-links. We made the minor adducts. We made so-called monofunctional adducts. And the transcription levels being higher in the corrected cells indicate that those cells more readily bypass during transcription.

So we repeated it using a site-specifically platinated DNA. Now, site-specific modification of DNA is a really high call. And I have to thank my colleague, John Essigman, who I met shortly after I came to MIT from Columbia. He had done a lot of really beautiful work on site-specific modification of DNA. And we learned a lot from him, and we published a lot of papers together and had a great time looking at specific adducts on DNA. And the intrastrand cross-links are certainly ones that we wanted to look at.

So here is the pGLuc vector. For those who are biologists, we have a promoter, we have the reporter gene. We put the platinum in between the two. Site-specifically platinated. Removed the origin of replication so we didn't replicate out the platinum. And in particular, with a little bit of experimentation, we discovered the best place to put it to get good signals was between the transcription start site and this sequence that directs the translation of the messenger RNA.

So the inventory of site-specifically platinated probes included the intrastrand GG, the diaminocyclohexane 1, the 1,3, and we've also made interstrand-- much more difficult task, but we have it and published on it. And so we could look at how the different adducts were repaired.

And in this particular set of experiments, we use a different set of cells. We use colon carcinoma cancer cells, P53 wild type or mutant. These were mismatched, repaired, deficient. These were mismatched, repaired, proficient. This is the unplatinated plasmid. And these are the platinated plasmids.

So you see very strong inhibition of repair over a period of time. Sorry, inhibition of transcription over a period of time. And very little increase with repair. Mismatched repair does not remove the major cross-links of cisplatin on DNA, which is why those experiments look different from the ones using the XP cells.

Now, this method is applicable to study many cellular contexts and many different types of compounds other than platinum. And we envision a potential application in personalized medicine. Possibly, if at some point down the line, when we're able to get tumor samples from patients in a short period of time-- this is a very fast experiment-- one might be able to take a look at different types of platinum compounds, maybe in conjunction with other things that we put on the platinum, which I'm coming to shortly, to see whether certain people might be able to respond better than others to certain types of platinum-based treatments.

Now, I've mentioned repair. It's time that I tell you a little bit about nucleotide excision repair. So this work was work that we carried out in collaboration with Aziz Sancar at the University of North Carolina, who is a really wonderful colleague and somebody who taught us a great deal about looking at NER. And so all of our genomes are damaged every day.

It's estimated that there's at least a million mutations that are made every day, or damage events that could lead to mutations. And we have repair mechanisms that can take it out. Whether it's walking out in the sun. And we have some today in Cambridge, finally. Or drinking milk and getting a little potassium 40. There's a lot of mechanisms that we have that allow us to repair the damage on the DNA.

And the one that takes out the intrastrand cross-link of cisplatin, essentially exclusively, is nucleotide excision repair. And the way that works is the following. If we have damage on the Watson strand, there's a series of enzymes and proteins which recognize it, and then cut out a piece, which I show down here with the damage on it. It's about 30 nucleotides in length.

It cuts first on the 5 prime end, and then subsequently on the 3 prime end. The cut on the 5 prime end is the first irreversible event in the process. And then you have a gapped DNA duplex, where if you've taken it out of the Watson strand, then the Crick strand has the correct sequence. You can then synthesize against that correct sequence with nucleoside triphosphates and make the repair patch. And the repair takes place.

So the experiment is set up in the following way. We take cell extracts that are repair proficient and repair deficient, and we put a platinated DNA duplex into the cell extract that we synthesize ahead of time. It's about 153 nucleotides long. And if the damage is placed on the Watson strand, when we put this piece in-- let's say it might be 50 nucleotides or so-- we will be sure to have a radio label placed as a phosphate group here.

And so, for example, here is a strand of 156 nucleotides that were put together with this particular sequence. And the platinum piece that was put in was TCTAGG with a 1,2 intrastrand cross-link CCTT. And then in a separate DNA-- this is now separate strand-- we put in TCTGTG so we could look at the 1,3 cross-link.

You make the double-stranded DNA by annealing the complimentary unplatinated Crick strand to it, put it in the repair extract, and look to see whether you get a 30 nucleotide piece-- this is a control lane-- 30 nucleotide piece in your gel. Now, because it's a cell extract, it has nucleases in there other than the ones that take out the repair damage. And so your 30 nucleotide piece is going to be cut to 29, 28, 27, and so forth. So you've got a little ladder of bands.

Now, if we look at HeLa cells, which are very famous cells that are used in cell biology to do experiments of this kind, which are cervical cancer cells-- the whole story related to Henrietta Lacks-- we see that it's proficient in repair and takes out the damage for both the GG, and even better, the GTG. So we see this is a better repair substrate than the major adduct, which is already interesting.

But you see there are these lanes mark F and G. Now F and G stands for the complementation groups F and G. So these cells are derived from patients who have a terrible disease known as xeroderma pigmentosum. It's about 1 in 100,000 individuals. And they lack one or more, usually one, genetically inherited loss of a particular one of those enzymes that's involved in the repair complex.

But when you put an XPF patient's cells together with an XPG patient's cells, they complement each other. And so as a consequence, you get very good repair signals when F and G are put together, but not when either is there alone. And so we can look at the kinetics. That is the rate at which the excision occurs as a function of time. It's shown here. And the GTG is better than the AG or GG instrastrand cross-link.

Now, the question is, how does all of this translate into the killing of cancer cells? And the answer is a battle that goes on in the cancer cell, namely that of transcription inhibition versus the inhibition of nucleotide excision repair. And I have a little cartoon that will perhaps bring this home to you.

Imagine that you have a piece of DNA that's got cisplatin damage. RNA polymerase II, which is the major transcribing molecule in cells, comes along, hit's the damaged site. The stalled RNA polymerase triggers multiple cellular processes, calling out for repair or programmed cell death. So it's going to go down one of those two pathways.

And our lab has studied this. We did some work with Phil Sharp in the past, leading to some insights into this process. And failure to recognize and repair the damage in answer to this 911 distress call is what you want in the cancer cell. Because what will happen otherwise is that the stalled RNA polymerase II will be phosphorylated at the 3 prime end, and then ubiquitylated with the proteins known as ubiquitin, taken to the proteasome, which is basically the garbage can in the cell for taking proteins apart.

The cellular repair machinery will be recruited, particularly nucleotide excision repair. And recognition and repair of the damage in answer to that distress is what you want in healthy cells. And we believe that the anti-cancer activity of the platinum drugs exploits that property.

Now, one of the very interesting questions which I had originally planned to talk about today, but I just don't have time to get it all in-- and we haven't finished the story-- is why it is that testicular cancer cells are so affected? Why is there essentially a cure of testicular cancer and not of so many other cancers?

And I don't know that this is the answer, but we're tracking down a rather interesting lead right now related to a particular protein that is found in all cells. It's in an amazing protein, known as high mobility group protein, that likes to bind to the cisplatin intrastrand cross-link. And when it does that, it blocks nucleotide excision repair. And so it can sensitize the cells to the platinum, because it removes the ability of the cell to repair.

We tried to prove that over the years. And we had a data which did prove it, and we had a data which disproved it. And we published both sets of experiments. And we just didn't-- it was just very frustrating to try to get any degree of consistency.

But recently, a form of the high mobility group protein appeared in testes that was different from that in all other cells. What I think, perhaps, the variable that we were not controlling in all of those earlier experiments was the intracellular redox potential of the cell, which is pegged at a certain value depending upon, if you're a chemist or a biochemist, the glutathione concentration in the cell.

And if the high mobility group protein has a pair of cysteines, and if they make a disulfide bond, they aggregate the interaction of that high mobility group protein with the platinum. Knock it off by about a factor of 1,000 by flipping a phenylalanine side chain that's inserted into the hydrophobic notch opposite the platinum. I have to be careful here, because I might be adding back in the time that I said I was cutting out.

Anyway, the important thing is that in the testes cell, one of those cysteines is a tyrosine. So it can never make the disulfide link. And now we're working very hard to be able to test that. And the big problem that we have is there just aren't enough-- fortunately, it's a good thing, I think-- testicular cancer patients in the world to give us enough cells to be able to test this. But we're meeting soon with some colleagues at Mass General. And we hope to be able to come see whether that hypothesis is the correct one or not.

Now, in the remaining time, I want to talk about attempts to try to improve things. Now, why do we want to improve things? Well, I'm sure you're aware of the fact that chemotherapy has its toxic side effects. And in the case of platinum compounds, there's either acquired or inherent resistance. And in some cases, the tumors, despite every attempt of the physicians, are just simply too aggressive and patient dies of the cancer.

Now, that has led us and others throughout the world to look at so-called non-classical platinum compounds. And I'm going to talk about platinum IV compounds, which are so-called dual threat compounds, and so-called monofunctional platinum compounds. Now, the dual threat strategy is the following. That is, we want to take a platinum compound like cisplatin and add two additional ligands to it by oxidizing it from platinum II the platinum IV.

If we could get that into the cancer cell, it would release the extra ligand, producing cisplatin. And the extra ligand, which if it had its own anti-cancer activity, could serve as a dual threat to try to knock out the cancer cell before it could become resistant or otherwise escape the chemical assault. The other possibility is to package these in nanoparticles that controlled the release of the platinum that platinates the nuclear DNA and operates by the mechanism I described, and/or produces another unit, x in this case, or l in this case, conveys an alternative biological activity.

And our approach was to build a platinum IV compound, which you can do quite easily chemically by oxidizing to the dihydroxide and then attaching a group R. Platinum IV compounds are more inert than platinum II, which allows good synthetic manipulations. And in one case, actually led to a compound that could be administered even orally. Didn't pass the FDA, but it came very close.

And different pharmacological and pharmacokinetic properties-- the spectrum of activity, the R group provides a dual threat. Reduced side effects, the R group could target a cancer cell versus a normal cell. Drug resistance, which can alter the mechanism of uptake, which cells sometimes block the uptake into the cell, which I haven't talked about. Or the compounds could reduce rapidly to potent platinum II compounds.

Now, the first dual threat compound that I thought about came as a result of reading Nature magazine on an airplane flying back to Boston from a trip that I was taking. And I was reading about a molecule known as dichloroacetate, or DCA, the potassium and sodium salts of which had been used to treat lactic acidosis. But cancer patients found that as an experimental cancer drug, it could shrink tumors in animals.

They wanted to do clinical trials in humans, but there was no incentive to do that. There was of a lot of discussion about this in the end of 2000, first decade. Because it had been around for years, its structure couldn't be patented. The drug companies weren't interested. And a lot of the work was done in Canada. And this molecule is still undergoing clinical trials. It's now in phase three in Canada. I looked recently, there's nothing new.

So the plane hit the ground in Logan. I was on my cell phone. And I called my post-doc, Shanta Dahr. And I said, Shanta, it should be very easy for us to attach the dichloroacetate to a platinum IV derivative. See if you can't make this compound.

And literally, before I got to the lab the next day, she had made it. That's how easy it is to make that compound. And we called it mitaplatin. I wanted to call it mitoplatin for mitochondria, but that had been used by I won't say who. Just a very odd thing. So I made up it mitaplatin.

And then one of post-docs says, I really should call it MIT-aplatin, just in case it--

[LAUGHTER].

And I'm willing to fly the flag. I don't have a brass rat, but I'm willing to fly the flag. All right. So mitaplatin, you see, would go into a cell. It would be reduced to platinum II, producing dichloroacetate. Now, what does dichloroacetate do? It takes advantage of something known as the Warburg effect. That is it's known for some time that cancer cells, when they first get going, have somewhat oxygen deprived environments.

And so their mitochondria don't work by the usual oxidative phosphorylation affects converting glucose into ATP. And so they go over and they use another mechanism, similar to what yeast do to anaerobically use-- to convert the glucose. And it turns out that dichloroacetate inhibits a key enzyme in that process known as pyruvate dehydrogenase kinase.

And so when that happens, the potential across the mitochondria, the cell potential, substantially decreases, leading to the release of cytochrome c, which is involved in apoptosis, or the programmed cell death pathway. As well as something known as AIF, the Apoptosis Inducing Factor, which goes over to the nucleus and helps the platinum DNA that's stalled here produce the killing effects themselves.

And so we decided to look at the effects of mitoplatin on cancer cell mitochondria, and we discovered it indeed changed the membrane potential. It released the AIF in cytochrome c. And most importantly, the platinum compound got in there, made cisplatin, or an analog, got onto the DNA, and formed the intrastrand GG cross-link as confirmed by using a monoclonal antibody known as R-C18, which I'll show a little bit later as well. Which is very specific for the intrastrand cross-link made by cisplatin.

So this is kind of a cartoon summary of what it does. And then if we look at the effect against cells, and NTERA-2 testicular cells, HeLa bone, platinum sensitive or non-sensitive cells. We found it was a very potent platinum IV compound.

This is really a proof of principal type idea. We don't think that mitaplatin's actually going to find its way into patients, or necessarily should. But it shows the dual threat, the way in which you can attack two different types of processes in the cell. Which I think would make it much more difficult for the cell to become resistant. And hopefully, lead to clinical success.

Now, the other way, of course, is to deliver platinum drugs in vivo by strategies that involve nanocarriers. And we've been working on this for some time. Many of my colleagues at MIT, including, I'm sure, some in the audience working on this. The idea being, using nanotechnology, to take platinum compounds, or other drugs for that matter, selectively into cancer cells. And perhaps protect them from reactions in the bloodstream that might inactivate them.

One of the ones we looked at very early were single wall carbon nanotubes, which we did in collaboration with Hongjie Dai's group at Stanford University. I call them long boats because he sent us these long tubes that reminded me of the Viking long boats. And I said, Hongjie, how many oars do you have on your boat?

And he said, well, we have about 80. And we had just, at that time, quantitated the amount of platinum that we could put on those oars. And it turned out to be 78. And I said, OK. That stuff is good. I can work with this material. I was very happy about that.

So here's an example of where we actually, using platinum atomic absorption spectroscopy, proved that we have 80. And we put folic acid on one of the two axial positions of platinum. So we oxidized the platinum for dihydroxide. We put a succinic Make acid on both ends, so we make the disuccinate. And one of them we turn into a carbamate with linker chains, ethylene glycol linker chains, which gives water solubility. And out to the end, we put folic acid.

And folic acid is known to target folate receptors on cancer cells. And so when we treat those cells, they will bind to the folate receptor, be internalized, and maybe direct the platinum compound, which would be released inside the cell as cisplatin specifically to that type of cell. This is the immunofluorescence study that indicates, again, using this monoclonal antibody R-C18, that in that experiment, we make the 1,2 intrastrand cross-link.

And then if we look at the cytotoxicity against the different sets of cells, folate cell receptor positive versus-- that's these types of cells-- for the negative one, the testicular cell, we find that the platinum single wall nanotube construct was more than eight times effective in the folate receptor positive cells than in the negative cells. And from the clinical point of view, if you could direct your compound against cancer cells versus normal cells, that would be a very significant achievement.

Now, then at that point, I went over and talked to Bob Langor. Who wouldn't, if you're interested in doing nanoparticles? And I said, Bob, you know, this is what we've got. I would like to take more advantage of nanoparticles. And I'm particularly enamored by your wonderful work using the PLGA-PEG, the Poly Lactic-co-Gycolic Acid.

He said, you know Steve, I've got this wonderful former post-doc named Omid Farokhzad who works with me. He's over at Harvard Medical School at the Brigham. He might be in the audience somewhere. And together with Omid and Bob, our labs have made a number of constructs where using nanoprecipitation, we've put in what would ordinarily be very hydrophilic platinum drugs with hydrophobic side chains-- this again, the work of Shanta Dhar-- into these particles.

And in fact, the three of us have co-founded a company known as Blend Therapeutics, which I'll talk about in a little while, which takes advantage of some of this nanoparticle formulation. But in addition, is very interested in these monofunctional compounds that I'll get to the few moments.

So in this case, what we did was we wanted to put something on the outside of the cell that would direct-- the outside of the polymer, rather, that were direct it specifically towards cancer cells. And so we put a so-called aptamer that was a PSMA aptamer that went after the prostate-specific membrane antigen. Which is overexpressed on the surface of most cancer cells.