It was Tuesday evening, June 7. A frightening outbreak of food-borne bacteria was killing dozens of people in Germany and sickening hundreds. And the five doctors having dinner at Da Marco Cucina e Vino, a restaurant in Houston, could not stop talking about it.
What would they do if something like that happened in Houston? Suppose a patient came in, dying of a rapidly progressing infection of unknown origin? How could they figure out the cause and prevent an epidemic? They talked for hours, finally agreeing on a strategy.
That night one of the doctors, James M Musser, chairman of pathology and genomic medicine at the Methodist Hospital System, heard from a worried resident. A patient had just died from what looked like inhalation anthrax. What should she do?
The questions were: Was it anthrax? If so, was it a genetically engineered bioterrorism strain, or a strain that normally lives in the soil? How dangerous was it? And the answers, Musser realised, could come very quickly from technology that would allow investigators to determine the entire genome sequence of the suspect micro-organism.
It is the start of a new age in microbiology, Musser and others say. And the sort of molecular epidemiology he and his colleagues wanted to do is only a small part of it. New methods of quickly sequencing entire microbial genomes are revolutionising the field.
The first bacterial genome was sequenced in 1995 – a triumph at the time, requiring 13 months of work. Today researchers can sequence the DNA that constitutes a micro-organism’s genome in a few days or even, with the latest equipment, a day. They can simultaneously get sequences of all the microbes on a tooth or in saliva or in a sample of sewage. And the cost has dropped to about $1,000 per genome, from more than $1 million.
In a recent review, Dr David A Relman, a professor of medicine, microbiology and immunology at Stanford, wrote that researchers had published 1,554 complete bacterial genome sequences and were working on 4,800 more. They have sequences of 2,675 virus species, and within those species they have sequences for tens of thousands of strains – 40,000 strains of flu viruses, more than 300,000 strains of HIV, for example.
With rapid genome sequencing, “we are able to look at the master blueprint of a microbe,” Relman said in a telephone interview. Matthew K Waldor of Harvard Medical School said the new technology “is changing all aspects of microbiology – it’s just transformative.” One group is starting to develop what it calls disease weather maps.
The idea is to get swabs or samples from sewage treatment plants or places like subways or hospitals and quickly sequence the genomes of all the micro-organisms. That will tell them exactly what bacteria and viruses are present and how prevalent they are.
With those tools, investigators can create a kind of weather map of disease patterns. And they can take precautions against ones that are starting to emerge – flu or food-borne diseases or SARS, for example, or antibiotic-resistant strains of bacteria in a hospital.
Others are sequencing bacterial genomes to find where diseases originated. Still others, including Relman, are examining micro-organisms that live peacefully on and in the human body. He finds, for example, that the bacteria in saliva are different from those on teeth and that the bacteria on one tooth are different from those on adjacent teeth. Those mouth bacteria, researchers say, hold clues to tooth decay and gum disease.
Real-world test
For Musser and his colleagues, the real-world test of what they could do came on that June evening. The patient was a 39-year-old man who lived about 75 miles from Houston in a relatively rural area. He had been welding at home when, suddenly, he could not catch his breath. He began coughing up blood and vomiting. He had a headache and pain in his upper abdomen and chest. In the emergency room, his blood pressure was dangerously low and his heart was beating fast. Doctors gave him an IV antibiotic and rushed him to Methodist Hospital in Houston. He arrived on Saturday night, June 4.
Despite heroic efforts, he died two and a half days later.
Now it was Tuesday night. From an autopsy, the cause looked for all the world like anthrax, in the same unusual form – so-called inhalation anthrax – that terrified the nation in 2001. Even before the man died, researchers had been suspicious because washings from his lungs were teeming with the rod-shaped bacteria characteristic of anthrax. Investigators grew the bacteria in the lab, noticing that the colonies looked like piles of ground glass, typical of anthrax but also other bacillus microbes.
“We had to know precisely what we were dealing with. That’s when we put into play a plan to sequence the genome, Musser said.” A few days later they had their answer. The bacteria were not anthrax, but were closely related. They were a different strain of bacillus: cereus rather than anthracis.
The bacteria had many of the same toxin genes as anthrax bacteria but had only one of the four viruses that inhabit anthrax bacteria and contribute to their toxicity. And they lacked a miniature chromosome – a plasmid – found in anthrax bacteria that also carries toxin genes.
The conclusion was that the lethal bacteria were naturally occurring and, though closely related to anthrax, not usually as dangerous. So why did this man get so ill?
He was a welder, Musser noted, and welders are unusually susceptible to lung infections, perhaps because their lungs are chronically irritated by fine metal particles. So his fatal illness was most likely due to a confluence of events: welding, living in a rural area where the bacteria lived in the soil and happening to breathe in this toxin-containing species of bacteria.
Waldor and his colleagues asked a slightly different question when Haiti was swept by cholera after last year’s earthquake. Cholera had not been seen in Haiti for more than a century. Why the sudden epidemic?
The scientists quickly sequenced the genome of the bacteria in Haiti and compared them with known cholera strains from around the world. It turned out that the Haitian strain was different from cholera bacteria in Latin America and Africa, but was identical to those in South Asia. So the researchers concluded that the earthquake was indirectly responsible for the epidemic. Many relief workers who came to Haiti lived in South Asia, where cholera was endemic.