illustration by Nick Tassone

When Worlds Collide

The right time and place led to a new RNAi-like pathway in bacteria.

Though their labs at University of California, Berkeley, are only a few buildings apart, biochemist Jennifer Doudna and geobiologist Jill Banfield were strangers. But when Banfield needed expertise outside her scope, their worlds converged.

Banfield studies microbes that grow in extreme environments. She needed a partner who knew about the gene-silencing process known as RNA interference (RNAi).

Doudna, an HHMI investigator who studies RNAi in eukaryotic cells, had solved the crystal structure of Dicer, a key enzyme in that process.

Realizing her luck at having a resource so close by, Banfield called Doudna to discuss what appeared to be a microbial gene-silencing strategy analogous to RNAi that was accomplished by highly repetitive DNA sequence elements known as CRISPRs—an acronym for clustered regularly interspaced short palindromic repeats.

Banfield and others suspected that microbes were using CRISPRs to fend off invaders. But nobody knew how CRISPRs could function as an immune system, and scientists were intrigued by what else the tiny bits of RNA derived from CRISPRs might be doing.

The two researchers met for coffee a few times in 2006 and explored the overlap between their worlds. Those conversations and the well-timed arrival of two new lab members led to Doudna's team capturing the first snapshot of a CRISPR enzyme in action. They reported in September 2010 that the CRISPR-associated enzyme Csy4 has the unusual property of recognizing, binding to, and cleaving RNA in a sequence- and structure-specific way.

"As often happens in science, there is a lot of serendipity involved," Doudna acknowledges.

The Timing was Right

Just as Doudna was becoming keen on the idea of CRISPRs, Blake Wiedenheft appeared. While earning his Ph.D. in Montana, he had become intrigued by the odd repetitive snippets of DNA he found in the microbes of Yellowstone National Park's boiling acid pools. Wiedenheft was convinced that there might be a system in play in these organisms similar to RNAi in eukaryotes. He wanted to pursue the question in Doudna's lab.

Soon after Wiedenheft joined the Doudna lab, graduate student Rachel Haurwitz arrived and became fascinated with Wiedenheft's project. "CRISPR in general is a pretty new field," says Haurwitz. "When I started my project, there had been a lot of computational groundwork done and it was really on the biologists to test the hypotheses."

In 2007, the first report surfaced on CRISPR-mediated acquired immunity. The Danish food production company Danisco needed a way to protect its yogurt-producing Streptococcus thermophilus cultures from bacteriophages, viruses that attack bacteria. Its scientists established the relationship between CRISPR and phage resistance. The immunity depended on the bacteria acquiring new DNA in the CRISPR region of their genome upon phage infection. The DNA sequences were identical to sequences in the infecting phages, and subsequently the bacteria were immune to infection with the same phages.

Armed with that new information, Haurwitz began her thesis project, asking: What enzyme in the CRISPR system is responsible for making the small bits of RNA that are later used to target a phage, and how does it function?

The team worked with Pseudomonas aeruginosa, a human bacterial pathogen with a CRISPR that had not been implicated directly in defending against phage infection. As such, it offered an opportunity to study its enzymes and mechanisms and then compare the findings with what was already known from Escherichia coli and Streptococcus. Hints of CRISPR involvement in biofilm formation suggested that it could be a potential target for controlling the persistent Pseudomonas infections in the lungs of cystic fibrosis patients.

A Discriminating Enzyme

Haurwitz identified Csy4 as the protein responsible for cutting up the RNA transcribed from a CRISPR sequence into a segment specific to Pseudomonas. Then, Haurwitz and Martin Jinek, a structural biologist and postdoc in Doudna's lab, set out to learn how Csy4 works.

Haurwitz designed a mutated version of Csy4's RNA target that allowed the enzyme to bind but prevented cleavage of the RNA, thus immobilizing the enzyme in action. Haurwitz and Jinek devised a method to create individual crystals of the RNA/enzyme complex, so that its three-dimensional structure could be determined by x-ray crystallography.

The team discovered that Csy4 recognized both the stem-loop, or hairpin, structure of the RNA molecule and the sequence of the RNA at the base of the stem (the double-stranded portion of the hairpin). "This exquisite combination of sequence- and shape-specific recognition allows the Csy4 protein to discriminate its target RNA from all the other cellular RNAs," says Jinek. The group published its results in the September 10, 2010, issue of Science.

"It has been quite a surprise to find that bacteria might be using a pathway similar to RNA interference but with a distinctly different set of enzymes," says Doudna. The rewards of venturing into this parallel universe, she says, are considerable. She can envision controlling microbial gene expression, gene silencing in eukaryotic cells without interfering with the cells' RNAi machinery, and developing diagnostic tools and therapeutics that rely on Csy4's ability to specifically target and destroy RNA.