Joe Derisi and virus detection

Joe Derisi is an associate professor at UC San Francisco and a Macarthur grant recipient. He’s interested in viruses – specifically the detection of viruses. He begins by showing us a bacteriophage, a virus that attacks bacteria. There are 10 million of these in the average drop of sea water, and there’s an amazing genetic diversity of them.

There’s lots of diseases caused by viruses. If you go to a doctor with a respiratory infection, 60% of the time you’ll go undiagnosed. But patients want a cure, and usually you’ll get prescribed antibiotics. They’re totally ineffective, and they lead to antibiotic resistant bacteria, which is a bad, bad thing.

What would be great is to have a chip that could detect what virus you were infected with. This is what Derisi works on – he’s finding RNA fingerprints for different viruses and building DNA chips that detect them. DNA chips are pretty easy – they’re glass plates with small pits filled with DNA, which match to different pieces of RNA in a virus. Different areas will “light up” when it detects a particular virus. The result is a “bar code”, a distinct signature of a specific virus.

(Joe has put a guide to building your own DNA chip fab in your garage – he says it costs less than building your own Toyota Camry. The guide is step by step and well illustrated – someone let Make magazine know…)

To test this, Joe shows us an experiment where he gets a subject infected with rhinovirus (shooting virus up the patients nose!) The chip detects nothing on the first day, clearly detects rhinovirus on the second day. Trying the same chip with patients off the street, it’s also capable of accurately diagnosing virus-based diseases.

Joe has now build a chip that has every known virus discovered on it. A chip this big requires some informatics attached to it – it needs an expert system to help you offer a diagnosis. He gives us a “House”-type story – a young woman who shows up at a hospital with a severe fever and breathing problems. They treat her with different antibiotics, but eight days into the hospital, she can’t breathe without mechanical assistance and is cleary dying. Using a little fluid from the lungs, they detect Parainfluenza 4, a very little-known disease that no one would ever detect. (No word on whether the patient survived, or whether House and Wilson are secretly having an affair…)

Showing us some new research, Joe lets us know about a possible link between a new retrovirus – similar to a mouse virus – that appears in roughly 40% of prostate cancer cases. It’s unclear whether this link is causal – it could simply be that these patients are very vulnerable to viruses – but it’s a door that allows investigation that would otherwise not be able to take place… and this detection couldn’t have happened without Joe’s gene chip strategy.