Viral locksmith is caught in the act

Interactions between viral and bacterial proteins promise new directions for antibiotics

Rockefeller University scientists, led by Milton H. Werner, Ph.D., were able to visualize the interaction between two proteins, AsiA (red) and sigma70 (blue), which causes the loss of gene expression in the bacterium E. coli. Drugs exploiting AsiA’s ability to inhibit gene activation could hold promise as antibiotics and anti-cancer therapies

How does the molecular machine responsible for activating genes choose which gene to switch on, from among the 30,000 genes contained in each cell of the human body?

In the August 4 issue of the EMBO Journal, researchers at Rockefeller University report that they are beginning to answer that question in bacteria, and the answers are not only surprising, but may also aid in the development of powerful new antibiotics.

If the current research is extended, “we could have a new strategy for developing an antibacterial agent, one that would act to weaken a protein’s structure,” says Milton H. Werner, Ph.D., the Rockefeller scientist leading the study.

“The real value in these findings is that we are learning new ideas about how you would make an antibiotic that would kill off bacteria in the same way that nature’s killers of bacteria already do,” says Werner, associate professor and head of the Laboratory of Molecular Biophysics at Rockefeller. “If we could deliver just those proteins that inhibit bacterial transcription, we would have a truly powerful antibiotic. I am very excited about just that idea.”

The study visualizes an interaction between phage, a virus that infects bacteria, and the proteins that make up the bacterium. The molecular image shows that the structure of the bacterium’s protein is altered as a consequence of the interaction with the phage virus, and the result is both an inhibition of gene activation in bacteria and an increase in the expression of the phage’s genes.

“It is very unexpected that proteins controlling gene expression can do so by significantly altering the organization of an enzyme,” says Werner. “In this case it leads to the loss of all gene expression in E. coli.”

In the bacterium E. coli, proteins called sigma factors monitor the environment of the microbe’s cell and promote the activation, or expression, of certain genes by recruiting a molecular machine called RNA polymerase. Each sigma factor is like a key, and is only able to activate the specific subset of genes that it fits. The sigma factors bring the RNA polymerase to the gene promoters, the segment of DNA that is the start site to transcribe the gene’s instructions. The seven sigma factors present in E. coli can be divided into two families. One of these families, called Sigma70, unlock “housekeeping” genes, which are essential for metabolism and survival of bacteria.

When the phage enters an E. coli cell, it first produces a set of proteins that inhibits all bacterial transcription. With the help of sigma70, the phage hijacks the bacterial RNA polymerase and forces it to only switch on, or transcribe, phage genes.

In a 2001 EMBO Journal paper, Werner and colleagues described the structure of AsiA and suggested that AsiA, expressed as a dimer (two subunits joined together) dissociates to interact with RNA polymerase. The current paper illustrates how AsiA interacts with sigma70 and remodels the domain of sigma70 with which it interacts.

“AsiA influences the polymerase in such a way that it cannot recognize the bacterial promoters anymore,” says Werner, “but instead recognizes a whole new set of phage promoters as a consequence. By understanding the interaction between AsiA and sigma70, which leads to the loss of bacterial gene expression, we can mimic it.”

Antibiotics that interfere with a gene’s transcription of its instructions to a body cell are of great interest because they are less likely to become ineffective. Normally, simple mutations can render a bacterium resistant to an antibiotic. However, transcription depends on the ability of proteins like sigma70 to recognize the promoter DNA, so any mutations in sigma70 must be followed by changes in the DNA. The chances that compatible mutations occur in both the lock and the key to confer antibiotic resistance are very slim, Werner says.

Understanding the function of a protein like AsiA in transcription aids in the development of species-specific antibiotics against diseases such as tuberculosis. In addition, while AsiA is the product of a bacterial pathogen, the same kind of strategy could also be employed by pathogens of animals and humans. Drugs expanding on AsiA’s ability to inhibit transcription could hold promise as anti-cancer therapies as well.

Research by Seth Darst, Ph.D., another Rockefeller University researcher, has examined the interaction between sigma70 and various inhibitory proteins called anti-sigma factors. Anti-sigma factors serve to prevent sigma proteins from interacting with RNA polymerase until the proper time, effectively blocking the sigma key from fitting into its lock.

In this new research paper, Werner and Rockefeller colleagues Lester J. Lambert and Yufeng Wei were able to visualize the interaction between AsiA and sigma70 using a spectroscopic technique called nuclear magnetic resonance (NMR) spectroscopy. They show that in contrast to other anti-sigma factors, when AsiA binds to sigma70, it remodels the key completely, so that now sigma70 only fits phage genes. Sigma70’s new orientation most likely causes subsequent changes in the shape of the RNA polymerase, though that question is still under investigation.

AsiA is now only one of three activators whose interaction with the polymerase has been visualized. Transcription factors are notoriously hard to crystallize because they have so many disordered areas. Werner’s and others’ use of NMR spectroscopy is finally bringing transcription factor structures into light.

“We have almost no pictures of how transcription factors engage polymerase. Lots and lots of people have tried, but they are not amenable to crystallization because they tend to have disordered segments,” explains Werner. “We have an advantage that we can make measurements from proteins that are poorly structured, disordered or unstable, and very often get very useful structural information from. Disorder doesn’t disturb us.”

Co-authors include Virgil Schirf, Ph.D., and Borries Demler, Ph.D., at the University of Texas Health Science Center.

This research was supported by grants from the National Institutes of Health and National Science Foundation. Werner is a W.M. Keck Foundation Distinguished Young Scholar.