Ann Hochschild, Keith Joung (center), a resident in clinical
pathology at MGH, and postdoctoral fellow Simon Dove have devised a two-hybrid
screen for bacteria, a powerful tool that could uncover new interactions
between mammalian proteins.Microbiology
To many who are not specialists in the field, gene transcription rings of confusion and complexity. Promoters, repressors, and operators buzz through geneticists' discussions. Enhancers and regulators don't much clarify matters for the uninitiated. Yet gene transcription--the first step in making a protein from a gene--is fundamental to just about anything that happens in cells and, by extension, a tissue or an organism.
Those without a knack for genetics can take heart: Sometimes, gene transcription turns out to be surprisingly straightforward. In the April 10 Nature, Ann Hochschild, HMS associate professor of microbiology and molecular genetics, and her coworkers report that getting a gene transcribed can be simpler than had been suspected. They showed that all it takes to set in motion the transcription of a gene is for a protein--any protein--to touch the RNA polymerase perched at the start of the gene. (The RNA polymerase is the enzyme that transcribes DNA into RNA.)
Ann Hochschild, Keith Joung (center), a resident in clinical
pathology at MGH, and postdoctoral fellow Simon Dove have devised a two-hybrid
screen for bacteria, a powerful tool that could uncover new interactions
between mammalian proteins.
The finding advances researchers' understanding of how cells regulate gene expression, a prerequisite for being able to manipulate this process therapeutically when it goes awry. Moreover, the work has spun off an unexpected technology: It sets the stage for using bacteria as a vessel for identifying new interactions among human proteins. A similar technique currently available in yeast--called the two-hybrid screen--is widely used throughout academia and industry.
Hochschild's team set out to define what it really takes to turn on a gene. Researchers had long known that so-called activator proteins must settle down on the DNA near the gene to be transcribed, and that the activators then somehow touch the RNA polymerase. But they wondered if a simple touch suffices to activate the enzyme or if something more complex is going on. Can any protein interaction do the trick, or are only special areas on the activators and the polymerase "competent" to turn physical contact into transcription?
To address these questions, the researchers removed part of the protein apparatus that normally switches on genes in bacteria and replaced it with better-known parts. In essence, the researchers devised a genetic setup that recreated the scenario of transcriptional activation with stand-in players. In the lead, they cast the protein *cI, which Hochschild had studied extensively before. One end of *cI--its C-terminal domain--tends to associate with other *cI C-terminal domains. Four of them bind one another, like four rolls baked together. Taking advantage of this feature, the researchers attached two such domains to the DNA, close to a test gene. They tethered two more to the polymerase (see diagram).
Touching Off Transcription: In this genetic contraption, the C-terminal domain of the protein *cI (red) initiates gene transcription. How? It appears *cI domains recognize one another, therefore, two domains bound to the proper spot on the DNA (pink) touch two domains tethered to a subunit of the RNA polymerase (light gray). This contrived contact triggers the polymerase to start producing RNA, showing that simple touching between proteins suffices to initiate gene transcription.
Having set the stage for the four *cI domains to touch
one another, the researchers discovered that once the domains did so, the
test gene was actually transcribed. After repeating this experiment with
mutant versions of the *cI domain, they also found that the difference between
a light touch and a strong grip matters to gene activation. The more strongly
the *cI domains touched their counterparts on the RNA polymerase, the more
RNA was produced.
This means that the touch probably helps pull the enzyme down onto the DNA, enabling it to read the gene more efficiently, says Hochschild. But not all natural activators in bacteria work that simply, she adds. Others use more complicated ways to put the polymerase to work.
The practical side of this basic study is already taking shape. By measuring gene transcription as the flag that signals protein binding, Hochschild's genetic contraption makes it possible to use E. coli to identify novel protein-protein interactions.
Such a technique, the two-hybrid screen, already exists for yeast cells and has become a mainstay of biomedical research. For example, it has greatly accelerated the deciphering of signal transduction cascades, one of many fields in which progress hinges on a scientist's ability to identify which protein interacts with which. "Protein-protein interactions are important in DNA replication, transcription, any cellular process you can think about," says Hochschild.
The test has not been available in E. coli until now, yet this lowly creature offers several advantages over yeast, says Hochschild. The great evolutionary distance between bacteria and mammals could make this prokaryote a more neutral environment for studying human proteins. Genetic manipulation of bacteria is easier and faster than that of yeast. Two-hybrid assays would benefit from standard techniques in bacterial genetics, such as using antibiotic-resistance genes to "fish" one's clone of interest out of a sea of genes. "The problem always is: How do you sift through all the garbage and find that rare thing that has the desired properties? Well, here you have the bacterium do the work for you," says Hochschild.
Pharmaceutical and biotechnology companies may be interested in adopting her technique, since it can help identify candidate drugs that disrupt proteins involved in disease. It enables the rapid screening of "libraries" of natural proteins or even artificially created peptides. Harvard's Office of Technology Licensing is working on commercializing the technique for diagnostic and therapeutic applications.
--Gabrielle Strobel
Focus April 18, 1997