June 12, 2006
When it comes to gene transcription, random pauses aren't quite so random, study finds
By Mark Shwartz
Of the thousands of proteins produced in our cells, few are as important as the enzyme RNA polymerase (RNAP), which has the unique ability to faithfully copy genetic information from DNA. In fact, all organismsfrom bacteria to peopledepend on RNAP to initiate the complex process of protein synthesis. Despite its crucial role in cell biology, fundamental questions remain about how the RNAP enzyme actually works.
Now scientists from Stanford University and the University of Wisconsin-Madison have solved part of puzzle. Writing in the June 16 edition of the journal Cell, the research team found that a molecule of RNAP makes frequent pauses at specific sites along the DNA double helix. This finding comes on the heels of the team's 2003 discovery that RNAP enzymes routinely make thousands of brief stops ("ubiquitous pauses") when carrying out the vital task of transcribing genetic information from DNA to RNAa process called transcription.
"Transcription of genes is terribly important," said study co-author Steven M. Block, professor of biological sciences and of applied physics at Stanford. "It's what determines the difference between the cells in your brain or your heart or your liver. All of your cells have exactly the same DNA, but what makes them different is that they transcribe different genes that code for different proteins."
From DNA to RNA to protein
Protein synthesis is strikingly similar in all organisms. It starts with DNAthe famous ladder-shaped double helix, whose rungs (or "bases") consist of four chemical units known by the abbreviations A, T, G and C.
A typical DNA molecule contains thousands of genes that encode thousands of proteins, which are essential for life. Each gene consists of a set of DNA bases arranged in a unique sequence that carries explicit instructions for building a specific protein. But one misplaced letter in that sequencea T substituted for a C, for examplecould produce a damaged protein that causes a serious disease or birth defect.
Transcription, the first step in protein synthesis, begins when an RNAP enzyme unzips a small section of the DNA double helix where a gene is located. The enzyme then builds a new complementary strand of RNA by chemically copying ("transcribing") the gene one base at a time. RNAP will continue moving along the DNA strand until the entire gene sequence is transcribed onto the encoded RNA, which then serves as a template for building the actual protein.
To observe RNAP in action, Block and his colleagues use a custom-built "optical trap" housed in his Stanford lab. This sensitive instrument allows researchers to observe transcription in real time by trapping individual molecules of DNA and RNAP in minute beams of infrared light.
"Our measurements are accurate to one-tenth of a nanometerthe width of a single hydrogen atom," Block explained. "When you study an RNAP enzyme at that scale, you discover that it moves along the DNA for a while, and then for no apparent reason it appears to stop. Some pauses we've already figured out. The really long ones, which happen every 1,000 bases or so and last up to 30 minutes, often occur when the enzyme makes a mistake. Then, it's got to back up and correct the mistake. But for every one of those, there are roughly 10 ubiquitous pauses that only last about 1 second and occur every 100 bases or soand their role is really something of a mystery."
To find the answer, Kristina M. Herbert, a graduate student in Block's lab and lead author of the Cell study, created experimental DNA templates using a special 240-base pair sequence that triggers one of two types of long pauses in RNAPa "backtracking pause" associated with gene regulation in which the enzyme reverses direction briefly; or a "hairpin pause, " named for tiny hairpin-shaped structures that sometimes form when an RNA strand binds to itself.
"Kristina made these totally cool DNA templates that have the same 240-base pair sequence repeated over and over again eight times in a row," Block said. When molecules of RNAP were attached to the templates, they behaved as predicted, pausing briefly at all of the backtrack and hairpin pause sites, but not actually backtracking or forming hairpins.
"That's great," Block said. "It's telling us that the enzyme is doing just what it should. After all, it's seen the same sequence eight times in a row, so it had better do the same thing eight times in a row. It also paused at several other sites as well, which is interesting. Sometimes it paused longer, sometimes shorter, but the average was remarkably the sameabout a second or so. We also discovered that it just didn't stop at any old sequence but at very specific places where there's a signal in the DNA that basically says, 'Pause here.'"
That signal, he added, occurred for specific sequences in the DNA. "We found that there is always a G near a specific pause position, and always a T or a C at another nearby position," he said. "So the pause seems to be sequence dependent. It's not always the same duration every time, but it's more likely to pause at one of these sites than at any other sites in between, so it's not just some random phenomenon that happens every once in a while. If I'm running down the road and I trip, that would be a random phenomenon. But if I run down the road and every time I trip there's a pothole, then that's not random."
Some researchers have argued that all pauses might be associated with either hairpin formation or backtracking, but the Cell study contradicts that assumption. "Most ubiquitous pauses have nothing whatsoever to do with backtracking or hairpins," Block said. "We think ubiquitous pauses are the most common and probably most important kind of pause, and the models that some biochemists have been using are just wrong."
What causes pauses?
The study also addresses a long-standing question about enzyme memory: If an enzyme pauses at one DNA site, will it alter its behavior when it encounters the same sequence again? "It turns out that RNAP enzymes may have individual personalities but no memories," Herbert explained. "That is, they exhibit a distinct behaviorthey tend to pause more or less, but they don't seem to remember having paused."
Why does RNAP make these fitful stops and starts? "No one really knows what all these ubiquitous pauses are doing nor what really causes pauses. They may be there to act as some kind of a governor to control the speed of transcription," Block said. "The control of gene transcription is one of the most fundamental ways to regulate gene expression in general, and one way to control transcription is to control pausing. Cancer is an example of gene control gone amok, so understanding the regulation of genes is critical to understanding cancer. Our results provide fresh insights into the control mechanisms that cells have for regulating the genes they express."
Other Stanford co-authors of the study are postdoctoral fellow Arthur La Porta, former undergraduate Becky J. Wong and former postdoctoral fellow Keir C. Neuman. Rachel A. Mooney and Robert Landick of the University of Wisconsin-Madison also are co-authors. Financial support was provided by the National Institutes of Health, the Stanford Undergraduate Research Grant Program and the Howard Hughes Medical Institute.
"Sequence-Resolved Detection of Pausing by Single RNA Polymerase Molecules," by K. M. Herbert et al., will be published in the June 16 edition of Cell. Images are available at http://newsphotos.stanford.edu/ (slug: "rna").