Cliff Mead: This year's legacy award winner is Dr. Roger Kornberg, a Stanford University biochemist. Roger Kornberg was awarded in 2006 the Nobel Prize in Chemistry for his fundamental studies of the molecular basis of eukaryotic transcription, the process by which DNA is copied. Kornberg's 1974 discovery of the nucleosome the basic protein complex packaging of chromosomal DNA in the nucleus of eukaryotic cells marked the beginning of his work on DNA. Coupled with his most recent discovery of the mediator protein complex, Kornberg's impressive research has added sequentially to the understanding of the mechanisms of eukaryotic transcription. Dr. Kornberg received his B.A. in Chemistry from Harvard University and his Ph. D. in Chemical Physics from Stanford University. He completed a post-doctoral fellowship at the laboratory of molecular biology in Cambridge of England before joining the Stanford Faculty. He has since cofounded Stanford's department of structural biology, the first of its kind in the United States. In 1993 he was elected to membership of the National Academy of Sciences. I'm here to tell you that Roger Kornberg is not the only one in his family involved in the sciences. His brother Thomas is a biochemist and his brother Kenneth is an architect, who designs biomedical buildings. Their mother Sylvy Ruth was a chemist and his father Arthur Kornberg won the Nobel Prize in Medicine in 1959 with Spaniard Severo Ochoa for their work on how genetic information was transferred from one DNA molecule to another. This was a family, who believed in science. Arthur Kornberg would take his children to the lab on weekends, they would talk about science at dinner, and they would take part in scientific activities during leisure time. When Roger was eight or nine years old his parents asked him what he wanted for Christmas and he said a week in the lab. Please help me welcome Dr. Roger Kornberg the man who has spent his life opening Christmas presents. [3:28]

Roger Kornberg: Thank you for such a warm and generous introduction and I thank all of those responsible for the honor of the occasion and the privilege of being here to speak with you. What I will try and do tonight is tell you the scientific story behind what has been so well described by Clifford Mead and his introduction. I mentioned to some of those earlier on this evening the several ways in which I am a scientific descendent of Linus Pauling and I might just repeat one of those examples I was in many respects a student of Linus Pauling while in chemical physics at Stanford, only then as you've heard I moved to join the competition, which was the group at the laboratory of molecular biology at Cambridge England. But I did so because it was the best place to continue in the Pauling tradition. At that time X-ray crystallography was most intensively practiced there and thus the reason for my move. I learned about it in very many aspects and it was the application of what I learned that leads to what I describe this evening. In the spirit of acknowledgment that we heard earlier on I'm going to begin recognizing the contributions of more than 100 remarkable students and fellows who've worked with me over the years and who are really responsible for the findings that I will describe to you in the next several minutes. So the work of some thirty years by some hundred people compressed into this short space of time. So if you could give me the first slide - so this is actually only a partial list of all of those who contributed, but in bold face, the individuals responsible for some of the most recent work that I'll tell you about tonight, if time permits. It includes my wife Yahli Lorch, my longest and closest collaborator, and also my son, who contributed to some of the most recent findings, but especially my longest student and fellow and then subsequent colleague in the laboratory, David Bushnell, and other very important contributors.

Now as you've already heard, when I moved to Cambridge England, I became interested in the structure of chromosomal materials. Indeed, it was the discovery of the nucleosome, shown on the next slide, that was my introduction to the subject of this evening's lecture. From a combination of biochemical and X-ray diffraction studies I was led to propose the wrapping of DNA around a set of eight protein molecules in the nucleosome, this fundamental particle of the eukaryote chromosome, the basic repeat unit of the structure. It was apparent from the outset that the wrapping of a DNA molecule around a set of proteins would interfere with many DNA transactions, and in particular, transcription, the process by which the information from DNA is read out for directing the form and function of every living organism.

Indeed, in the mid 1980's, some 10 years after the discovery of the nucleosome, Yahli Lorch and I showed that packaging a gene in this way would prevent its expression in the test tube. Not long after, Michael Grunstein and his colleagues at UCLA showed a similar inhibitory effect of a nucleosome particle upon gene expression in a model organism, the baker's yeast, en vivo. From that point on we all appreciated, the nucleosome is a general gene repressor. It prevents the activity, the expression, of all the many thousands of genes in an organism, such as a human, except those whose expression is brought about by specific positive regulatory mechanisms. All the work since that time on these lines was directed towards the discovery of those positive mechanisms by which inhibition of transcription or of expression by the nucleosome is somehow relieved. This has been the work of very many people over a long period of time. I'll summarize only very briefly the twists and turns of the particular path that has been followed.

So at the beginning, very early on, it was thought that the solution to the problem was the removal of the so-called histone proteins from the DNA, and exposure in a naked form. The reason for that, it was discovered very early on that the DNA of genes engaged in transcription is particularly or especially exposed to attack by enzymes that cleave the DNA nucleuses by chemical agents and what have you. Then, some years later, that view was overturned by the discovery that the proteins, as I say, called histones, are still associated with genes that are active, that are undergoing transcription, except they are in an extensively modified form. In the next slide - So they're modified by methylation, by acetylation, and in other ways, especially on the tails of the proteins that protrude from the central region of this particle. On this basis, it came to be believed, and the view was widely held throughout the 1990's and well into the current decade, on the next slide, that histones are not dislodged from the DNA, for the purpose of turning genes on, their activation. But rather the nucleosomal particle is altered in some matter that makes the DNA available for attack, by enzymes as I mentioned, and makes the DNA available for expression, or transcription. And it was with this idea in mind that two recent colleagues, whose names are on the previous slide, Boeger and Griesenbeck, undertook to isolate the altered form of the nucleosome in order to understand the mechanism of gene activation. [10:53]

Now, I'll only summarize that extensive work with two slides, one which illustrates the approach they used and the other which states the conclusion from the work. But it's again, it's quite straightforward. So the approach they used, which is illustrated in the next slide, please Jill, was to isolate a gene of interest, so a gene that would serve as an example for the purposes of illustration. The gene name is PHO5, but that's not important, it's a gene of the baker's years, that model organism to which I referred before. The gene contains, consists of a region called the promoter, where transcription begins, and then a region that actually codes for the protein product of the gene. The method was to place DNA elements on either side of the gene that would be recognized by an enzyme called the recombinase that could join them together and excise this region, as a circular entity free from the rest of the chromosomal material, permitting its isolation and analysis. Now in this particular case, the region in front of the protein coding part, is packaged in three nucleosomes, illustrated by these blue ovals. It is this region that was known from previous studies using for example nucleasis to attack the DNA, this region that was known to undergo the alteration of structure that we wished to understand. Now the conclusion from analyzing the isolated material, what emerged from our studies, both of the previous ideas were partly right, and both, in some way incorrect. It turned out that nucleosomes are removed, histone proteins are dislodged from the DNA, in the course of gene activation, but they are also rapidly reassembled. The difference between an inactive gene, one that is silent, and one that undergoes transcription, is not the presence or absence of the proteins that package the DNA, but rather a conversion from a static state in which nothing happens to a dynamic one, in which the packaging proteins are continuously removed and reassembled. As such the DNA is then transiently available for a brief interval of time in the naked for that can be accessed, that is able to interact with the transcription machinery. May I have the next slide.

Now in our four studies the transcription machinery was that for the central enzyme called RNA polymerase II abbreviated pol II for the reason that is shown in the next slide. RNA polymerase II is the enzyme responsible for all transcription of DNA called messenger RNA, which codes for proteins. RNA polymerase II catalyzes the first step in the pathway of express of genetic information as such the action of this enzyme is a focal point of cellular regulation it is for example it is the intricate regulation of RNA polymerase II transcription that underlies cell differentiation and development of a complex multicellular organism. The effort to define the components and understand the mechanism of this machinery began long ago. It was first undertaken successfully by Robert Roeder and his colleagues at Washington University in St. Louis in the 1970s. They prepared an extract from human cells they were so called hiela cells that could be grown in a laboratory in flasks in a culture, from which I should say an extract of the cells that was capable of accurate gene transcription. So they derived the first cell free system for analyzing this process. Due to the limitation of the amount of the material, human cells and culture are very scare, it was difficult for them to go on dissect this machinery, purify components from the extract, and actually identify the molecules that were responsible for transcription. That challenge was most successfully taken up by Ronald and Joan Conaway at the Oklahoma Medical Research Foundation, who began not with human cells, but with rat liver, an abundant source of material for this purpose. [15:59]

In parallel with that work my colleagues at Stanford and I undertook the same type of study starting from baker's yeast, the model organism to which I referred, that others have exploited beginning very long ago. Now there was a particular challenge in the case of studies with baker's yeast for this purpose. We and others in fact every laboratory around the world interested in this problem it had been attempted unsuccessfully to make an extract like that originally derived by the Roeder group at Washington University from human cells. It was only in the late 1980s when a graduate student in our group, actually an M.D. Ph.D. student at Stanford discovered the solution of that problem, could make an active extract, take advantage of its many virtues for the purpose of study and that led rapidly then to the dissection of the machinery, the identification of all of the molecules involved.

So in around 1991, 1992, if I may have the next slide, we, the Conaways at the Oklahoma Medical Research Foundation and others arrived at the conclusion that is shown here, that we could establish, could identify six molecules that could be combined to bring about accurate gene transcription in the test tube. These six molecules include the central enzyme that is responsible for the generation of the RNA, the RNA polymerase itself, and then five additional molecules that go by the letter names B, D, E, F, and H. Now I will show you shortly in atomic detail the RNA polymerase, pol II, is itself alone capable of unwinding the DNA double helix and then of the assembly of an RNA transcript, the messenger RNA, which is built along one strand of the DNA. The RNA polymerase is by itself even capable of correcting errors that are made in this process to assure the faithful readout of the genetic information. But the RNA polymerase enzyme by itself is incapable of locating a gene to begin with. So incapable of finding what is called the promoter, the start site. And also incapable of initiating the process. And it is for these crucial functions that the general transcription factors by these letter names are required.

Now in the beginning we thought that this set of six molecules are not only necessary for the process of transcription, but sufficient in all aspects. We imagined, others believed, it was widely thought that this constellation of protein molecules was capable of not only transcription, but a response to regulatory influences. The kind of regulation that is the key to understanding the processes to which I alluded RNA polymerase is so essential, cell differentiation, cell development, response to environmental influences, and what have you.

We believed that this system was complete and in that sense capable of responding to regulatory influences because, if I may have the next slide, of many experimental findings to indicate gene activator proteins, the molecules that turn on expression of genes infringe direction upon this machinery to bring about the readout of the genetic information. Now in spite of those experimental proofs of such direct interaction, a student in our laboratory Ray Kelleher nevertheless exceeded in showing, around 1990, that there was an additional component required, one whose name has already been mentioned. [20:38]

So if I may have the next slide. What Kelleher showed is in the context of the proteins derived from the baker's yeast another crude fraction derived also from the yeast was also necessary for turning on gene transcription. The activity in that crude fraction was given the name mediator because it evidently serves as a go between, the trigger protein, the activator, and the transcription machinery. The pursuit of this mysterious component is a long story, another one that I won't have time to tell, but again I'll just come to the conclusion.

In 1994 two post doctoral colleagues Stefan Bjorklund, a postdoc coming from Sweden who has since returned, and Young-Joon Kim, a Korean post doctoral fellow who now leads a leading laboratory a very fine lab in Korea, succeeded in the isolation of mediator in a homogeneous form. They discovered it was an assemblage, a vast assemblage, made up of 21 individual proteins with a molecular weight greater than one million dalton units. It was for almost a decade viewed as an anomaly a kind of idiosyncrasy of yeast cells because it continued to be believed for reasons that I won't have time to explain that the process of turning genes on is direct in human cells and it results from the trigger protein, the activator, making as I say direct contact with the transcription machinery. It was only around the year 2000 that we and several other groups succeed in human, other mammalian, and now we know also plant and additional higher cells the presence of an exact counterpart of the mediator originally identified in yeast.

The purpose of the next slide is just to emphasize both the complexity and evolutionary conservation. There is for every one of the 21 proteins in yeast a demonstrable counterpart in higher cells such as the fruit fly drosophila or humans cells or as I've told you in plant cells. Now we've only just begun to fathom the complexity of this vast mediator assemblage. A few of the main points to have emerged from the recent years are listed in the next slide.

The first is that mediator is not only required for control of transcription, for regulation of the process, for turning genes on and off, it is required for all the technical transcription of all genes that are transcribed, that are read by this RNA polymerase enzyme in all eukaryotic cells from yeast to humans. In human cells every activator isolated proves to be in the form of a tight complex of this mediator assemblage. In this way we now understand mediator communicates information from DNA sequences in the genome called enhancers to the genes that are undergoing transcription, whose activity needs to be controlled either turned on or off. A mediator is commonly called a co-activator because it collaborates with the trigger protein that turns on genes. There won't be time to tell you in detail, but mediator is equally responsible for turning off transcription. So it's an activator, collaborator, co-repressor, and also a general factor for transcription. Mediator then is a kind of molecular computer. It receives multiple inputs of regulatory information from within the cell, from within the organism, and the environment and in some way processes that information and then delivers a signal to the enzyme associated with the gene, whether to read the information or not or to transcribe or not and to what extent to do so. [25:07]

Now it's a dictum attributed to Francis Crick, but could as well or at least is truly in the spirit of Linus Pauling, “if you wish to understand function, study structure.” I think it would be fair to say the tradition which Crick embodied with that statement certainly originated with Linus Pauling. And if I may have the next slide. The challenge in this case is the size and the complexity of the structure. Some 60 individual protein molecules assemble in a giant complex with a mass larger than a ribosome, larger than three million dalton units. On every gene prior to every round of readout of transcription of the genetic information. Now at the time when we began to pursue structure for the purpose as I say of understanding function an object of this size lay hopelessly beyond the reach of the methods available. We started our work with the polymerase component, itself an assembly of 12 individual proteins of mass half a million daltons some ten times larger than any protein previously solved by structural studies. It was in retrospect a fortunate choice had we begun with one of the smaller simpler general factors, to which I have alluded, we would have been unsuccessful. We now know they adopt defined structures that one could hope to determine only in the complexes they form when they interact with the RNA polymerase. This molecule is the platform which all of the other components assemble and as I will try and explain and I think you will understand, knowledge of the structure of this enzyme is really the key to appreciating to fully grasping and understanding the mechanism and the regulation of the gene readout, the transcription process.

Now as I say when we began the methods at the time were really inadequate to the task. Neither X-ray crystallography, the main means by which protein structural information was derived, nor another method some of you will know called magnetic resonance was really appropriate for something of such size. So we took another approach, which employs not X-rays, but rather electrons. The essential instrument, the electron microscope, is employed for crystal structure determination only in this case the crystals must be very thin preferably one layer of molecules thick or two dimensional to allow transmission of the electrons from the source to the photographic film or other recording device used to measure the outcome of the experiment. Now the challenge was how to make single layer thick, two dimensional, crystalline arrays and we devised an approach that is shown in the next slide.

So the idea is illustrated here. It was to adsorb a protein of interest to a monolayer of so called lipid molecules, fat molecules, shown here spread at an air-water interphase binding, adsorbing, the protein to a monolayer would constrain the molecules in two dimensions. Then the hope was that the rapid movement of the molecules laterally within such a layer, which I had discovered as a graduate student in the studies I did, which I briefly alluded to before at Stanford, would enable crystallization. Well in 1979-1980 two of my colleagues, Seth Garth and Allen Edwards succeeded in making single layer thick crystals of RNA polymerase II in this way. It emerged essential to remove by genetic means very small subunits from this enzyme to for very large well ordered two dimensional crystals. Then something quite unexpected happened, if you could give me the next slide, that even at the very very low concentration of protein that was used in such an experiment and under conditions which are close to natural, the physiological, that would not be expected to induce crystallization. Nevertheless 2-D crystals adsorb additional layers of protein in register with the first. The 2-D crystals promote a process, which is called epitaxial growth. The 2-D crystals could be used to see the formation of large single crystals, next slide, that would never otherwise have been obtained, but which lend themselves to structure determination by the time honored method, to which I alluded, X-ray diffraction. So in this way, we, by a circuitous route, found our way through to finally the application of the time honored way of deriving atomic structural information. [31:04]

Now another period, a long period, of trial and error ensued and the next slide simply summarizes some of the difficulties that were unique to this system in consequence of which, if I may have the next slide, ten years from my original graduate work on lateral diffusion to the formation of single layer thick crystals, another ten years to the discovery of how to form large single crystals, macroscopic crystals, suitable for X-ray analysis, and then another ten years, if I may have the next slide, 10,000 liters of yeast and one graduate student went into the solution of this very challenging problem. Actually you see here the most recent student Ken Westover being lowered by my longtime colleague David Bushnell, who I've mentioned, to repair a valve at the bottom of this fermenter, in which all of the yeast, 10,000 liters were grown over this period of time to do the work. Ken Westover was an M.D. Ph.D. student and has since gone on to the Brigham, the women's, I can't recall which of the major hospitals, the Harvard hospitals in Boston where he is today a medical resident. I don't know if we'll ever find a student again of his caliber.

So we finally arrived in the year 2000 at the structure of this remarkable, essential enzyme molecule shown in the next slide. This is as I say a constellation of 10 proteins made up of 3,500 building blocks or amino acids, 28,000 carbon, hydrogen, excuse me, carbon, nitrogen, and oxygen atoms, 28,000 non-hydrogen atoms solved to initially a resolution of 2.8 angstroms. This is a view of the entire structure with each of the individual protein components in a different color. The one in red for example in the lower left is the fifth largest protein and it is a 25,000 molecular weight protein, which is atypical of most individual and commonly studied protein molecules. You will also see here a pink sphere which symbolizes a magnesium atom that marks the active center of the enzyme where the actual process of gene transcription occurs. Now what I'm going to do is just rotate this around to give you a sense of the size and complexity. This is a view, which we say is from the top. First we'll rotate so you can see from the side. Then it'll turn around and show you from the other side and then finally come back to the top.

So midway through that brief animation you will have seen a view actually right down the central channel and I'll come back to that in a moment. Indeed the first thing you wonder when you see this picture is, where does the gene enter? And where does the messenger RNA that is read out from the gene exit? And you might imagine already from what I've told you well perhaps it will enter through this capacious central channel leading to what I told you is the active center where the RNA molecule is made. Well proof of that point came from determination by a particularly long suffering of a post doctoral fellow, who actually succeeded in finally solving this structure, not of the enzyme alone, but in the form of an actively transcribing complex with the gene DNA present and the RNA being made. Indeed he made crystals of that, in which the enzyme is still active and if the building blocks are added it will make more RNA. [35:45]

So this was the work of Avi Gantt and I'll preface it with a slide for the purpose of just introducing the color code, if you could give me the next slide. So I just need to explain, that at the active center of a polymerase molecule engaged in this process of transcription, the DNA strands were known for many years to be separated in the form of what is called a transcription bubble for obvious reasons. It was also known that the RNA product is associated with one of the DNA strands so to speak hybridized in the center of that level with the growing end of the RNA adjacent to that magnesium atom that I mentioned just before the junction between the bubble and the double helical DNA. Now in this picture and all the ones to follow, the strand of the DNA that directs the synthesis of the RNA will be blue and the opposite strand of the DNA of the double helix green, and the RNA is always red. And now with the next slide.

First of all you have here a view down that central channel, one of the ones you saw when I rotated the molecule to begin with, but now the only difference is that it's all one color except for one portion, instead of different colors for each protein. Now what I'm going to show you here is a movie and it's made of only two frames and each frame is an actual crystal structure, it's the actual arrangement of atoms and molecules determined by the method of crystallography. So in this second frame of the movie you see the structure determined by Gantt of the RNA polymerase engaged in the process of transcription with the DNA and RNA both present and indeed the DNA blue and green does occupy the central channel and the associated RNA as well. And you may further more have noticed this colored protein element, which is itself massive and has components of four different individual proteins of the entire complex, swings a large distance over the DNA and RNA in the course of forming the active complex and for obvious reasons we call this a clamp that hold the DNA in place so that you can read from the beginning to the end of the gene. So you can see I have rotated around 90 degrees and removed some of the protein in the front that was white just because it obscures the DNA and RNA and gives a better view of this enzyme in the act of reading genetic information, truly at atomic resolution. You don't see here all of the individual atoms, that would be too complicated, but what you see is the path of the individual protein chains and of the individual DNA and RNA chains traced through the locations of the atoms determined by crystallography. And what will be apparent to you is that the DNA double helix enters from the right and then unwinds in the manner that was suspected from biochemical studies. Three building blocks of the DNA, three nucleotides before the active center, after which there is sharp bend in the coding strand of the DNA in consequence of which the next DNA unit or base is flipped 90 degrees points down towards the floor of the active center for readout in the transcription process. The component of the RNA, the ribonucleotide as it is call, just added is still paired to the corresponding DNA and there are eight more DNA RNA base pairs of the hybrid region resolved in this structure. This really is the essential outcome of the work after all that time. And then the rest of what followed during the next ten years was really to try and learn the lessons that it taught us about transcription. [40:14]

One of the principle points to emerge from this analysis was understanding the basis of the fidelity of transcription, like the biological specificity that occupied so much of Linus' interest in his career. The question was how does the RNA polymerase select the correct building block of RNA to add in response to the sequence of the DNA? What is the basis of the accurate readout of the genetic code, which is after all the essence of transcription. Well the answer to that question has emerged from this work and I can explain in a bit more detail with the use of a couple of the diagrams if someone would like later on. Then after that we've also learned from structures such as the one shown here how after a nucleotide the, after adding a component to the RNA, the machinery advances to the next position. Finally we've learned how after the RNA molecule is made it is separated from the template DNA that directs it synthesis. In a separate line of work we have investigated the interaction of this core of the transcription machinery with the many additional components, to which I alluded, the general factors, whose letter names I mentioned, B, D, E, F, and H and the massive mediator that controls or that computes the control of the process. I thought I would just conclude by just giving you a sense of what has emerged from that work still in progress.

This is sort of as I say leaping to the conclusion. So here we have structures of all the components of the machinery, to which I alluded at the beginning, the polymerase, and then the other molecules that go by the letter names B, this is the key component of D, E, F, and H. Those missing small subunits of the polymerase as well. And of course the DNA that contains the beginning of the gene to start transcription. Now in this series of slides you'll see how it all comes together. So the next slide shows where the two small subunits that we left out at the beginning actually add on to the RNA polymerase. I might just mention for clarity, this is the same view that I showed you originally, except this is in a surface representation, previously what you saw was a trace of the polypeptide or the protein chains all in the different color, but this is the same view from the top as before. And then we'll just advance the slides. Next you'll see where factor F goes and the one half of factor B goes right in here and I could explain why it performs a remarkable and key role in the beginning of the process. Amongst other things its other half shown here brings the component of D and the gene DNA itself into the picture. E interacts with H, next slide, which brings it to the surface of the complex. This is what it all looks like when transcription is about to begin. The key next step is H contains an enzyme that separates the DNA strands to create the transcription bubble. And once that has happened transcription can finally begin.

This is a picture, still very fuzzy, of just the beginnings of an outline of the massive mediator. Which is shown here in blue that partly surrounds the back side of the RNA polymerase enzyme. If you could go a slide forward now, we've crystallized one third of this massive mediator and its structure will soon be derived and now you can just in advance and that will be the end. And we anticipate, I should say we hope, we look forward when the structure of mediator is finally known to not only understanding the mechanism of transcription, how it begins, but what decides when it happens, what controls the beginning, and thus the long sought purpose of all the exercise, the control of eukaryotic transcription. And with that I thank you for your attention.

[Audience Question] [45:10]

One of the central questions about the mechanism, what propels the enzyme forward? I mean there are many aspects to your question, but to focus on one central one, how does it move in the right direction, what makes it move forward? And the answer is interesting. It was directly demonstrated by a colleague in the biology department at Stanford named Steven Blocking. It was long known that the polymerase enzyme after adding a nucleotide, a building block to the RNA, is capable of then either moving forward or back. It diffuses freely, it oscillates between both directions. What is responsible for propelling it in the forward direction is simply entry of the next building block, not its addition, but its entry and occupation of the active center of the enzyme. As long as there's an abundance of building blocks then the enzyme will repeatedly adsorb that component and then add it to the growing chain. It can as well, as I have said, diffuse backwards, but that's unproductive and when it happens to move forward it's captured in that state, adds a nucleotide, and then moves ahead. So it's a, exactly, it's what's called the brownian ratchet.

[Audience Question] [46:45]

So the question is, about the third of this giant mediator that we've made some progress towards analyzing in structural terms and the young man who asked the question is aware that it may play a very important role in that key initial step in transcription, where the DNA strands are melted apart, a step that really sets the stage for the process. And the unfortunate answer is no we don't yet know from the structure so far as it has been derived how that works, but there's still hope because the structure isn't really done. So we'll see, but I can't answer your question. I'm sorry.

[Audience Question] [47:33]

So the young man who did that work, Neal Lu, came to me one day, I should say that we had discussed this of course and he had all sorts of ideas about how to make it work and there were many straight forward things and of course still it didn't work, and then one day he came along and said what did I think if he were to change the salt solution, something I believed was perfectly irrelevant, from sodium chloride to sodium acetate. I assured him it would make no difference, and that was the trick that made it work. So it's very fortunate that my students don't listen to me.

Audience Member: The human extracts early on seemed to work without the mediator, even though later on you discovered the mediator was important. [48:30]

Ah, okay that's a very sharp question. The perception that there's an internal contradiction in what I've said. Because on the one hand I commented that the mediator is essential for all transcription, but then at the same time I had previously told you that transcription was studied in its absence. So the answer, we think, is that the transcription that was studied before that was the basis for all of the work to identify these components was a product of, so to speak, leaky regulation. It wasn't actually supposed to happen. In fact it's a terribly inefficient reaction. At best one percent of all the DNA one introduces into such a biochemical reaction actually engages in transcription. And it is my belief that that one percent is a mistake. That it was supposed to be zero and that in nature, in our cells, in vivo it is truly zero and you require mediator and that is important since you should require a stimulus for that to happen. So this point has not be proven, but the basis for the second statement, that all transcription requires mediator, comes from yeast genetics, where it has been shown that indeed in cells in vivo if one knocks out mediator, one observes no measurable transcription from any yeast gene.

[Audience Question] [50:24]

Yes please. So the question was is there a single type of mediator or are there multiple forms for different genes or what have you. The answer interesting, on the one hand there a unique homogeneous mediator and the best evidence for that we've succeeded actually in growing single crystals of the entire yeast mediator. So it must be very uniform. And that is all of the mediator that one isolates from a yeast cell. So on the one hand the answer to your question is a general factor and it is the same molecule for every gene. However, at the same time it's been shown in human cells that mediator from different tissues may include some additional subunits, up to as many as 14 additional subunits. The total may be as large as 35 proteins. So the mediator in its action in a particular cell type may be general, but there can be variation from one type of cell to another, important for biological specificity.

[Audience Question] [51:46]

So this is a question that pertains to the interface between genetic complexity and the generality of the underlying mechanism. At some point there must be some divergence between the general components common to all genes and specific components that regulate the activity of individual genes. And the answer is quite clearly, what I have said, that the mediator is general, but then the trigger proteins, the activators, and also the converse the negative regulatory proteins the repressors. Those are gene specific. Now then the mediator has to be the interface from the general apparatus to this bewildering variety of individual gene regulatory molecules. Well we don't know much about that except to say that first of all we believe that must at least partly explain the size and complexity of mediator. Bear in mind that this interface must be capable of a simultaneous response to many different inputs any one gene is regulated in regard to its location in the body, the particular time, all the different environmental conditions that impinge upon its activity. And all of that information has to get funneled through this mediator to the underlying transcription machinery. I think the final answer to your question, which would probably make sense in terms of what you've studied in biochemistry, there may be thousands of different regulatory proteins associated with different genes. They then have features, so called activation domains, for the activator proteins which fall in broad categories. And then for each of those categories there's an appropriate receptor or surface of the mediator. So the complexity funnels down from the broadest at the level of the individual genes finally to the level of the mediator.

[Audience Question] [54:49]

So this question alludes to something which I kind of mentioned, but also I glossed over. There are, as you many realize as you reflect on what I've said, two important levels of the controls of gene activity. One, as I've mentioned in the beginning, unraveling the chromosomal material, exposing the DNA, and then a second, once it is available, to interacting with the transcription machinery, then control through mediator to respond to all of the many fine and the nuances of transcription control via the environmental influences and so forth. Now what this question alludes to a component of that first step, what if a part of that machinery for unraveling chromosomal material were making the DNA available for the second stage, which is transcription. We don't yet know whether that very first stage is also in some way controlled and modulated by mediator, is the answer to your question. We suspect it is, indeed we have evidence for interaction of mediator directly with the chromosomal material, we and others I should say. But that hasn't been demonstrated in the conclusive fashion that you might like.

Cliff Mead: Perhaps one or two more questions then we can end the session.

[Audience Question]

Roger Kornberg: There are both immediate opportunities for application of this knowledge and ones that lie further in the future. The immediate opportunities include for example the direct inhibition of RNA polymerase molecules in unwanted organisms. So on the basis of the structure that I showed you, one can readily design small molecules that could be used as drugs that inhibit fungal growth or that would inhibit bacterial growth to make better antibiotics. And this work is being done, it's being done on a commercial basis now taking advantage of this information. At the next level one could anticipate something finer by way of controls. So not such a blunt instrument as simply blocking the process but actually regulating, controlling, the process. That will require more knowledge of the mediator that I mentioned and of these molecules with which it interacts to influence the transcription process. So there's an immediate benefit from the standpoint of human health, but there will be much greater ones in the future. And if you like the enthusiasm for stem cells depends upon our ability to direct their fate. And that in turn depends very much on greater knowledge and thus better control over the transcription of the genes that determine those fates.

[Audience Question] [58:58]

The question is about the other major forms of RNA polymerase. There are three in eukaryotic cells. The reason that polymerase II goes by that name is because there is also and one and it turns out there is a three. One and three are responsible for the synthesis of the structural RNA molecules involved in the expression of genetic information. One, RNA polymerase I, makes the RNA component of ribosomes that in turn make proteins that read the messenger RNA translated sequence that of protein molecules. Polymerase III does something closely related, it makes all of the individual small so called transport molecules, that escort the building blocks of the protein to the ribosome. The point here is that one and three recognize a very limited number of promoters. There's a mark contrast between those enzymes and polymerase II. One transcribes only two promoters, those are the large and small RNA component of the ribosome. Three only a few dozen promoters at most. Where as two, the enzyme I've been speaking of, transcribes literally thousands of promoters. So one and three don't require the fine regulation, but two on the other hand requires enormous capacity for discrimination as a consequence one and three have no mediator and are not subject to control in the same way. Two, polymerase II, the enzyme spoken of is unique in nature, neither does the bacterial RNA polymerase, similar or related in structure, which does transcribe thousands of genes, even have a mediator. It is only the eukaryotic RNA polymerase II, which must be capable of this orders of magnitude greater diversity of response. Only for the RNA polymerase II is this additional assemblage derived and acquired.

Cliff Mead: This concludes the 2010 Pauling Legacy Award Lecture.