Roeder made his wish come true. Beginning with his initial discovery, he exposed a string of revelations that illuminated the nature of eukaryotic transcription. This process turned out to differ significantly from bacterial transcription. Instead of finding one enzyme that constructs RNA, as is the case in bacteria, he found three. Yet even these multi-part machines — each more complicated than the single bacterial counterpart — couldn’t identify their appropriate starting points in genes. Each one needed another set of molecular contraptions to make meaningful contact with their respective target DNA sequences. Furthermore, jacking up transcription of particular genes beyond a basal level required yet other proteins. But the demanding transcription apparatus was not finished. The activators refused to talk directly with the other components of the reaction, choosing instead to communicate through an additional intermediary.
The enormous complexity of the eukaryotic transcription machinery furnishes the cell with many tunable steps and components. The ability to tweak the transcription reaction allows the cell to perform feats of sophisticated biology — activating particular constellations of genes at certain times during embryonic development, for example, or in response to chemical cues. Roeder’s work laid the groundwork for these insights and for our current knowledge about how eukaryotic transcription works.
Three’s a charm
By the time Roeder started his graduate work with William Rutter nearly 40 years ago, scientists suspected that bacteria carry a single enzyme, RNA polymerase, that makes protein-encoding RNAs as well as RNAs that perform other functions. Studies of RNA production in isolated nuclei, however, had hinted that the story might be different for eukaryotes.
As a first step toward obtaining protein mixtures that he could test for enzyme activity, Roeder isolated nuclei from sea urchin embryos, but then he faced a challenge. Unlike the bacterial RNA polymerase, which floats free, most of the eukaryotic RNA polymerase binds to DNA and its associated proteins. Stuck to such a large conglomeration, the enzyme (or enzymes) did not dissolve easily. Roeder used some clever biochemical tricks to keep polymerase activity in solution while leaving the DNA-protein amalgam in a clump. He then subjected the dissolved substances to a technique that separates molecules in the mixture from each other. With this approach, he generated many tiny pots of material, each with different collections of proteins from the next.
To test for the presence of RNA polymerase in his samples, he added a portion of each to a test tube containing a DNA template and RNA building blocks — one of which carried a radioactive tag. Any RNA polymerase in the material would presumably sit down on the DNA and link together the building blocks to create a long, radioactive RNA molecule whose sequence reflected that of the DNA. Three separate activities emerged from his samples, and he named them RNA polymerase I, RNA polymerase II, and RNA polymerase III (nicknamed Pol I, Pol II, and Pol III).
Roeder wondered whether these enzymes carried out specialized jobs, and by the time he joined the faculty at Washington University, he was in a position to find out. Other researchers had discovered a mushroom toxin called amanitin, which blocks RNA synthesis when injected into rats. Roeder used this poison to test whether the different polymerase activities perform distinct roles in the cell. In a test tube, tiny amounts of amanitin thwarted Pol II, whereas even large quantities did not perturb Pol I; Pol III’s vulnerability to the toxin lay in the middle. These different sensitivities offered the ability to distinguish the polymerase activities from each other in a setting where all of them were present.
Assuming that the test-tube findings would extend to isolated nuclei, the toxin should profoundly inhibit Pol II activity, partially inhibit Pol III activity, and leave Pol I activity largely unscathed, he reasoned. If the polymerases generated different classes of RNAs, the types of RNA produced would reflect the different amanitin sensitivities. Using this rationale, Roeder showed that Pol I transcribed the bulk of the ribosomal RNAs, Pol II transcribed protein-encoding RNAs, and Pol III transcribed a class of small structural RNAs. With this experiment, he had established that these enzymatic activities produced different classes of RNAs, specified by particular classes of genes.
At the time, however, no one knew whether the three polymerase activities that he had identified represented independent molecular machines; perhaps instead, the cell chemically modified a single core enzyme to render it capable of performing different functions. Roeder next found that each of the three polymerases contained nine or more proteins. Based on their sizes, he concluded that some elements were common to multiple polymerases whereas most were unique to a single enzyme. This work established that a defined selection of proteins composes each of the three polymerases and that they cannot swap identities through small structural alterations. These discoveries added to the growing sense that eukaryotic transcription would prove to be much more complicated than bacterial transcription, where a single enzyme composed of four proteins can manufacture all types of RNA.
Despite Roeder’s triumph in uncovering the general structures and substrates of the polymerases, a problem was gnawing at him. The test-tube reaction that he had been using to measure enzyme activity represented the step at which the polymerases are chugging along, stringing together RNA building blocks. In this system, the DNA template was nicked, which allowed the enzymes to slip onto it; they didn’t have to meet a challenge that existed in the cell — to pick out the gene’s starting point, its so-called promoter. In order to understand how the cell turns on genes at specific times during embryonic development, in certain cell types, and under particular circumstances, Roeder would need to coax his enzymes to faithfully choose where to begin working.
The first breakthrough came in the mid 1970s. Roeder took DNA from immature frog eggs and added purified Pol III. The DNA was special in several ways: It contained multiple copies of the gene for a small RNA called 5S, which was known from the work of Donald Brown to be heavily transcribed in immature frog eggs, and it was not naked. Instead of peeling off its associated proteins, as he had in previous experiments, Roeder dumped it into his reaction mixture along with anything it toted from the frog egg nuclei. This DNA-protein substrate yielded the correct-sized 5S RNA — which began and ended in the right place. Furthermore, the polymerase had homed in on the 5S DNA rather than the other genes that were also present. For the first time, accurate transcription of a eukaryotic gene had occurred in a test tube.
The fact that only protein-covered DNA supported specific transcription indicated that the DNA was carrying along a protein or proteins that help Pol III bind to 5S DNA. Despite the complexity of the polymerase itself, additional elements were necessary for the reaction. Presumably these crucial components were not part of the fundamental DNA-protein coalition called chromatin, which was known to wrap DNA tightly and thought to block access of other DNA-manipulating molecules. Instead, Roeder reasoned, some other ingredients must have clung to the DNA-protein mesh. To identify them, he sought molecular machinery that would allow Pol III to correctly transcribe 5S RNA from pure DNA. In 1979, his lab accomplished this feat. A particular portion of the cell’s contents rendered the purified enzyme able to precisely transcribe 5S RNA. Later that year he similarly reconstituted correct transcription of a protein-encoding gene with Pol II.
The ability to faithfully copy RNA from DNA in a test tube dramatically changed the prospects for studying transcription. Roeder didn’t yet know how many individual modules — or factors — the essential cellular material represented, nor did he know how many proteins composed each factor, but now he had a test system with which to find out. Researchers would be able to use his experimental set-up to tease apart the reaction, identifying its molecular participants and detailing how they influence each other.
A matter of factors
Roeder began exploiting his system. He further divided the required cellular substance, and then added different combinations of his samples back to the polymerases to define the minimal elements required for transcription. In 1980, he found that two factors — which he named general transcription factors — were needed for production of several Pol III-transcribed RNAs and an additional one was needed for production of 5S RNA. He also found evidence for what later turned out to be six general transcription factors that were required for Pol II to accurately transcribe its DNA substrates. The components required for Pol II were distinct from those required by Pol III. He had thus identified polymerase-specific factors, revealing another level of complexity — and potential for control — unique to eukaryotes.
Because 5S RNA production demanded a factor in addition to the general transcription factors, Roeder concluded that something special was necessary to allow Pol III to transcribe the 5S RNA gene. He purified it using his test-tube system. With that accomplishment, Roeder had isolated the first gene-specific transcription factor from a eukaryote, which he called TFIIIA.
The enormous intricacy of eukaryotic RNA transcription was becoming more and more apparent. Not only did eukaryotes employ three rather than one polymerase, each of which transcribed a particular class of genes, but they depended on layered sets of adjunct factors. These factors were required for transcription of all members of a class or for individual genes. General transcription factors triggered a small degree of transcription from pure DNA, and gene-specific activators amplified RNA output.
Roeder subsequently showed that his gene-specific activator — TFIIIA — binds to the 5S RNA gene and recruits one of the general transcription factors. In contrast to the situation in bacteria, where a subunit of the polymerase itself binds to the DNA, the eukaryotic enzyme binds to the gene only after other factors have latched on. Roeder found, for example, that TFIIIA entices a general transcription factor that, in turn, delivers other general transcription factors and the polymerase.
In the years that followed, many laboratories, including Roeder’s, identified gene-specific activators for protein-encoding genes. The mechanism that Roeder worked out for TFIIIA function, in which a gene-specific activator targets general transcription factors rather than the polymerase, has since proved to be the main mechanism for stimulation of protein-encoding genes as well.
Trapping the mediator
Transcription of protein-encoding genes has drawn tremendous interest from the scientific community, because these genes are what render a nerve cell able to fire and make a toe a toe rather than a finger. With his Pol II test-tube reaction, Roeder discovered that a general transcription factor called TFIID touches down on a landing site that marks the start of Pol II genes. Once bound, it recruits the rest of the transcription machinery. Knowing that TFIID lies at the heart of Pol II transcription triggered a race to figure out exactly which of its protein components contacted the DNA.
Philip Sharp and Leonard Guarente along with Pierre Chambon and André Sentenac isolated the DNA-binding element of TFIID from yeast by identifying the protein that would substitute for human TFIID in Roeder’s test-tube system. In 1993 and 1994, structural studies by the groups of Stephen Burley (working with Roeder) and Paul Sigler revealed that this protein pries open the DNA, exposing the normally tightly wound genetic material to the other transcription factors. In 25 years, scientists had proceeded from the most basic questions about eukaryotic transcription—how many polymerases exist—to detail at an atomic level about precisely how the process works.
The field was flourishing, yet major challenges remained. Scientists such as Robert Tjian as well as Roeder were purifying gene-specific activators that were known to ramp up RNA output in the crude test-tube transcription system and in cells. But when they added these regulator molecules to transcription systems that contained extremely pure basic transcription machinery, they lost the ability to boost gene activity above a low level. Cells’ ability to turn on genes under particular circumstances is crucial for their ability to specialize and respond to physiological cues, so scientists invested a great deal of effort into resolving this conundrum. Presumably, they reasoned, their purification schemes were ditching essential components of the reaction, which they would need to retrieve in order to fully define how activation works.
In one of these situations, a protein called thyroid hormone receptor, which activates particular genes in response to thyroid hormone, wouldn’t prod its target genes in a test-tube transcription system composed of purified components. To find the missing piece, Roeder and colleagues tagged thyroid hormone receptor with a molecular handle that allowed them to pull the protein out of a cellular mush — and it brought along a conglomeration of about 25 proteins. This machine proved to be the main constituent of a group of molecular gadgets that Roeder had earlier found to be necessary for the function of other activators. It has since turned out to play an essential role not only in thyroid hormone-activated transcription but in activation of all Pol II-transcribed genes. Through genetic studies and biochemical purification, respectively, Richard Young and Roger Kornberg had previously identified and isolated a similar mélange from yeast, which Kornberg called Mediator; subsequent analysis of Roeder’s thyroid-hormone-receptor-associated protein (TRAP) conglomeration revealed its close relationship to the yeast mediator.
Most eukaryotic genes attract multiple activator proteins — some bound to DNA sites distant from the promoter — and these regulators collaborate to produce more than a sum of their individual stimulatory activities. Today, scientists view the mediator as a multi-part control panel that integrates myriad positive and negative signals. The regulator proteins sit on a piece of DNA and interact with mediator — which is commonly bound to general transcription factors and Pol II — to determine what degree of transcription will occur.
The study of transcription is now booming, owing in large part to Roeder’s initial work. We now know that Pol II plus its general transcription factors are composed of a total of about 44 proteins. The general transcription factors place the polymerase on the promoter and help separate the DNA’s two strands so RNA production can begin. Then they send the polymerase on its way down the gene.
This molecular coalition performs the same enzymatic feat as does the much simpler bacterial polymerase, so one might wonder why evolution has bestowed eukaryotic cells with so many additional proteins. Dozens of studies have revealed the generality of Roeder’s early findings, which in turn address that issue. Unlike the bacterial polymerase, the enzyme can’t initiate contact with DNA on its own; it requires a collection of other proteins. The extra proteins apparently offer the cell additional targets through which it can exert control over the transcription machine. Every step in the assembly process — starting with the binding of TFIID — can be slowed down or sped up, depending on the cell’s needs. Thousands of gene-specific activators and repressors adjust the fervor with which the polymerase and the general transcription machinery clasp DNA.
All of those proteins are doing something else as well. Although the fundamental reaction — stringing together RNA building blocks — is the same in bacteria and eukaryotes, the requirements differ inside the respective cells, in large part because eukaryotic DNA wraps tightly around proteins. The general transcription machinery is sufficient to transcribe naked eukaryotic DNA — but not DNA that is assembled into chromatin. Gene-specific activators not only help the general transcription machinery zero in on genes of interest, but also — with the aid of other molecules — disrobe the heavily protein-clad DNA so that the general machinery can contact the promoter.
Transcription touches all areas of biology, from embryonic development to cell-type specialization to diseases of deviant gene control such as cancer. Scientists are still far from understanding every nuance of the process, but they have come a long way in figuring out how cells read out some genes and not others. Roeder’s work on the fundamental reaction and his development of a system in which to study it has allowed researchers to link together individual facts into the ever-expanding story of eukaryotic transcription.
by Evelyn Strauss
Key publications of Robert Roeder
Roeder, R.G. and Rutter, W.J. (1969). Multiple forms of DNA-dependent RNA polymerase in eukaryotic organisms. Nature. 224,: 234–237.
Parker, C.S. and Roeder, R.G. (1977). Selective and accurate transcription of the Xenopus laevis 5S RNA genes in isolated chromatin by purified RNA polymerase III. Proc. Natl. Acad. Sci. USA. 74, 44–48.
Weil, P.A., Luse, D.S., Segall, J., and Roeder, R.G. (1979). Selective and accurate intiation of transcription at the Ad2 major and late promoter in a soluble system dependent on purified RNA polymerase II and DNA. Cell. 18, 469–484.
Engelke, D.R., Ng, S.-Y., Shastry, B.S., and Roeder, R.G. (1980). Specific interaction of a purified transcription factor with an internal control region of 5S RNA genes. Cell. 19, 717–728.
Matsui, T., Segall, P.A., Weil, D.A., and Roeder, R.G. (1980). Multiple factors required for accurate initiation of transcription by purified RNA polymerase II. J. Biol. Chem. 255, 11992–11996.
Meisterernst, M., Roy, A.L., Lieu, H.M., and Roeder, R.G. (1991). Activation of class II gene transcription by regulatory factors is potentiated by a novel activity. Cell. 66, 981–993.
Luo, Y., Fujii, H., Gerster, T., and Roeder, R.G. (1992). A novel B cell-derived coactivator potentiates the activation of immunoglobulin promoters by octamer-binding transcription factors. Cell. 71, 231–241.
Fondell, J.D., Ge, H., and Roeder, R.G. (1996). Ligand induction of a transcriptionally active thyroid hormone receptor coactivator complex. Proc. Natl. Acad. Sci. USA. 93, 8329–8333.