The toxin, called cytolethal distending toxin, or CDT, is used by bacteria that cause a range of diseases, from typhoid fever to diarrhea. And unlike any other bacterial toxin discovered to date, CDT attacks the DNA in human cells, creating lesions and breaks that cause cells to stop dividing and eventually die.
Principal investigator C. Erec Stebbins, Ph.D., who conducted the research with Rockefeller colleagues Dragana Nesic, Ph.D., and Yun Hsu, says that having a chemical model provides a visual blueprint for understanding how the toxin damages DNA.
"CDT may be a carcinogen because it damages DNA," says Stebbins, assistant professor and head of the Laboratory of Structural Microbiology at Rockefeller. "That makes CDT interesting in terms of both infectious disease and cancer."
While normal cells regularly replicate, or make exact copies of themselves, cells exposed to CDT go into "cell cycle arrest." While some CDT-invaded cells continue to grow in size without dividing, others commit suicide through a process known as apoptosis.
These cellular reactions to CDT invasion are not surprising because cells have "checkpoint" mechanisms to insure that DNA sequences are copied correctly during the process of replication, says Stebbins. When a "mistake" or a break in DNA is detected, cells either repair the DNA, or, if it is irreparable, they stop replicating and often commit suicide. This process is a protection mechanism to stop mistakes from being propagated.
Stebbins says that CDT-containing bacteria might induce cell cycle arrest because it stops cells from being sloughed off in areas such as the intestines where bacteria seek to make a home.
"CDT may also act as an immunosuppressant," says Stebbins, "since the immune system requires cell division to respond to microbial infection."
There are nearly 10 different species of disease-causing bacteria that use CDT, including Salmonella typhi, a bacteria that causes typhoid fever; Haemophilus ducreyi, a bacteria that causes genital ulcers; Campylobacter jejuni, a common cause of food poisoning; certain strains of Escherichia coli that cause diarrhea, and a host of other pathogenic bacteria.
"More CDT-containing bacteria are discovered each year," says Stebbins, "Many of these bacteria cause very different kinds of diseases and colonize different tissues. But they all have CDT. To me, that argues that it's playing an important role."
Stebbins' structure of CDT visually confirms that this genotoxin is made up of three subunits, including one called CdtB that cleaves, or cuts, DNA.
According to Stebbins' model, the three-unit toxin contains a long, deep groove, a cluster of ring-shaped molecules, called the "aromatic patch," and a dangling protein tail that can block a key portion of the CdtB subunit that is necessary for DNA cleavage.
"We're not sure what the role of the cleavage-blocking protein tail is, but the structure helps us to understand how to interact with the active site of CdtB to impair its activity, which could give us some ideas for achieving the same thing with a drug molecule," said Stebbins.
Armed with an atomic model of a protein, scientists can program computers to screen virtual representations of millions of compounds to see if they have a good chance of interacting with the target. Compounds are narrowed down to a selected few, which are then tested to determine how well they interact with the drug target. By screening compounds by computer before physically testing them, drug makers save time and money.
Stebbins' laboratory includes a drug-design team, which is currently designing molecules to thwart bioterrorism agents such as anthrax and plague. In the future, the lab might design a drug against the CDT toxin, Stebbins says.
"We can screen three million molecules against a particular target in two weeks by computer," says Stebbins. "That's good for finding about 500 molecules that are potential drugs that can be examined for activity against a protein. But after that come other hurdles, such as whether molecules can enter and work within cells and organisms."
The hardest part of solving CDT's structure was isolating hundreds of milligrams of the toxin in a pure form, says first author Nesic, who went through more than 24 gallons of buffer solution per week in order to purify the toxin into a form that could be used for crystallization. It took Nesic nearly two years to obtain the crystals she needed for X-ray crystallography - a technique that records the pattern produced when X-rays bounce off a target protein crystal.
"You have to find the right conditions and the right concentrations for crystallization," said Nesic. "You start with liters, and you end up with a single drop."
Stebbins has solved the structures of over ten other proteins, including the cancer-related VHL tumor-suppressor and several other bacterial toxins, before solving the structure of CDT.