Molecular Networks Uncovered in Bacterial Infection, Nerve Cell Communication

RNA Interference Enables Process-wide Studies

Life near the cell surface is anything but superficial. The plundering entry of a bacterium, the release of chemical messengers into a synapse—each requires the coordinated effort of hundreds of proteins. Until recently, scientists had to doggedly piece together a picture of these events one protein at a time, either by genetic means, creating mutants and knockouts, or biochemically, targeting individual proteins with small molecules. In the 1990s, researchers stumbled upon the existence in plant and worm cells of a completely unexpected class of molecules—snippets of double-stranded RNA—that could effectively turn down or off the expression of genes. It soon became apparent that double-stranded RNAs (dsRNAs) could be artificially created for every gene in an organism, raising the possibility that whole genomes might be interrogated gene by gene.

Photo by Graham Ramsay

“RNA interference is exciting because we finally have a tool that allows us to do global analyses,” says Norbert Perrimon (left). “This is the new genetics.” Clockwise from left are collaborators Jennifer Philips, Laura Burrack, Hervé Agaisse, and Darren Higgins.

Labs all over the country raced to develop genomewide dsRNA libraries and assays in worms, flies, mice, and even humans. Many of these efforts have come to fruition. Researchers are using them to identify entire ensembles of genes and proteins. The approach, generally known as RNA interference (RNAi), is bringing the prospect of answering long-held questions about complex processes, such as bacterial infection and synaptic transmission, into the realm of possibility.

“It’s like the Fourth of July, Christmas, and Chinese New Year all wrapped up into one,” said Darren Higgins, HMS assistant professor of microbiology and molecular genetics, about the excitement generated by the technique. Higgins, Hervé Agaisse, and Laura Burrack, along with Jennifer Philips, Eric Rubin, and Norbert Perrimon, recently conducted a genomewide RNAi study to see how two different species of bacteria—Listeria monocytogenes and Mycobacterium fortuitum—enter and get a foothold inside Drosophila cells. Their findings, which include a long-sought receptor gene, appeared in a pair of papers on July 14 in the online edition of Science. Across town at Massachusetts General Hospital, Derek Sieburth, QueeLim Ch’ng, Joshua Kaplan, and colleagues have been feeding dsRNAs to worms in an effort to understand how message-conveying chemicals are released into synapses. Their investigation, reported in the July 28 Nature, is the first large-scale RNAi screen in neurons.

Burden of Riches
While the sheer abundance of data, well over a hundred genes in each study, is allowing researchers to contemplate big-picture questions and potential new therapies, it is also presenting challenges. Identifying genes is an important first step, but a full understanding of biological function requires knowing when and where proteins appear in the cell and how they interact. “Right now we are creating these catalogs of genes,” said Perrimon, Howard Hughes investigator and HMS professor of genetics. “What we would like to understand is networks. Before we would talk about three or four genes in a system. Now we have to understand how hundreds of genes are working together.”

“RNAi is an excellent method for defining the components—this is the list of the proteins we should deal with... The idea is to then build additional correlations by bringing information from protein–protein interactions and gene expression.”

In 2000, after the Drosophila genome was sequenced, Perrimon, whose main research interest is in signaling pathways, knew that he needed to think larger scale. “We realized the approaches we had so far were too limited in scope,” he said. With funding from the National Institutes of Health, he and his colleagues worked for four years to set up the Drosophila RNAi Screening Center, essentially a high-throughput facility that systematically knocks down each of the 21,000 genes in the Drosophila genome (see “Genomics,”Focus, Jan. 23, 2004). Since May 2004, when the center opened, he and his colleagues have used the facility to identify proteins involved in cell motility, mitosis, and numerous signaling pathways (see “Research Briefs,” Focus, April 22, 2005).

Differential Invasion
The foray into bacterial entry and infection grew out of a common research concern shared by Philips and Higgins. Philips, a research fellow in Perrimon’s lab, is interested in Mycobacteria. Higgins has studied Listeria extensively. Years of research had revealed a general sketch of the invasive habits of each, but detailed pictures were lacking. Both microbes latch onto receptors on the host cell and are then engulfed in a membrane-bound vesicle. While Listeria soon escapes into the cytosol, Mycobacteria remain and replicate in the harsh environment of the vacuole. “We knew little bits and pieces, but we really did not know the bigger picture,” said Higgins. “For example, how is the bacterium getting into the cell in the first place? What is the host receptor?”

Image courtesy of Darren Higgins

Knock-down punch. Researchers took 21,000 double-stranded RNAs (dsRNAs), one for each gene in the Drosophila genome and put them into well plates, one dsRNA per well. They added cultured fly cells to each well and waited four days for the dsRNAs to knock down gene expression. Cells were then exposed to bacteria expressing green fluorescent protein. Infected cells became filled with green bacteria (a). Cells in which the dsRNA turned down a critical gene showed little or no green bacteria (b and c).

Rather than focus on the pathogens, the researchers, including Rubin, an associate professor of immunology and infectious diseases at HSPH, looked at the host cells and asked which genes and their proteins are necessary to support infection. Philips, along with Agaisse and Burrack, researchers in Higgins’s lab, carried out parallel RNAi screens on M. fortuitum and L. monocytogenes (see figure). “We said, let’s see what specific host factors are required by each bacterium. But let’s also see what are the general things pathogens need to get into a cell and replicate,” Higgins said. “Those would potentially be targets for the development of new therapies on the far end.” The screens netted a total of 305 genes in L. monocytogenes and 86 in M. fortuitum. These were genes that when knocked down altered the ability of the pathogen to infect. Of these, several were critical to infection by both M. fortuitum and L. monocytogenes. One of these, including the host receptor, Pes, was more critical to M. fortuitum. The researchers then grouped the proteins into known functions. Not surprisingly, many of the host factors, both shared and specific, are thought to be associated with the host cell’s vesicular and cytoskeletal systems.

Figuring out how the host proteins are coopted by the invading microbes—that is, how host and pathogen interact to make infection possible—will take further sleuthing. “RNAi is an excellent method for defining the components—this is the list of the proteins we should deal with,” said Perrimon. “The idea is to then build additional correlations by bringing information from protein–protein interactions and gene expression.” In addition to interacting with bacterial proteins, the host proteins are presumably interacting with one another, either to build or maintain vacuoles and the cytoskeleton or to carry out other cellular functions. “By looking at how bacterial factors are specifically interacting with host proteins to get certain phenomena, you get great insight into what the host protein is doing for the cell,” said Higgins.

Neuron Screen
It would help such efforts to know where in the cell particular proteins are located. Sieburth, Kaplan, and colleagues actually took this step in their large-scale RNAi screen in neurons. It was not easy, considering neurons have been resistant to the RNAi method. Using a mutant strain of nematode worms developed in the lab of HMS professor of genetics Gary Ruvkun and reported in the July 28 Nature, the researchers were able to knock down more than 2,000 genes. Their goal was to identify factors involved in the transmission of the neurotransmitter acetylcholine across the neuromuscular junction, part of a larger project on synaptic transmission. “We’re very interested in how synapses work, how they are built, how presynaptic cells recognize postsynaptic cells, and how connections are built between them,” said Kaplan, HMS professor of genetics.

Photo by Marc Raila; inset courtesy of Joshua Kaplan

“We have missed many of the components that act at the synapse because until now we have not been able to apply these broader approaches of looking at many molecules,” says Joshua Kaplan (inset), who worked with QueeLim Ch’ng (left) and Derek Sieburth.

Like the bacterial researchers, they netted a large number of genes, but when it came time to organize their results, they faced an even greater problem. Of the 185 genes they found, 132 had not been previously associated with any aspect of synaptic transmission. To figure out what these might be doing, Sieburth, HMS research fellow in genetics, and his colleagues took a subset of 60 genes and repeated the RNAi screen under seven different conditions, each of which is known to stimulate acetylcholine secretion. Then they asked, Does knocking down this gene decrease acetylcholine secretion? For each inhibited gene, they got a reading—yes, a reduction results or no, it does not. They clustered the genes based on similarity of profile over the seven conditions into three main functions—synaptic vesicle release, neuropeptide signaling, and responsiveness to phorbol ester, a substance that promotes acetyl-choline secretion.

Using a fluorescence assay developed by Marc Vidal, HMS associate professor of genetics at the Dana–Farber Cancer Institute, they were able to locate the proteins to either the presynaptic axon or postsynaptic dendrite. “So we have 24 new presynaptic markers to look at, and now we can ask how does the abundance of each of these presynaptic proteins change when we manipulate the system either pharmacologically or genetically,” said Kaplan.

“The hope is to take the roughly 200 genes that we found and do a combination of experiments where we can continue to look broad-band at a large set of molecules as an ensemble— how they regulate each other or act together to control synapses—as well as start to do more in-depth studies of individual molecules in a more traditional way, asking how the biochemical mechanism of each might be playing off the others.”

—Misia Landau

top