Today’s ration of scientific confusion comes courtesy of Wired, in an article that talks about using a modified form of TMV (tobacco mosaic virus) for delivering silencing RNAs. A group at Maryland has used the virus to deliver various siRNAs to cell lines in vitro, which is an interesting idea. But then it gets the Wired treatment:

The short, double-stranded RNA molecules known as siRNA can program cells to destroy disease-causing proteins. Their molecules turn on a cell's own built-in disease-fighting mechanisms. They can be programmed for a wide range of ailments -- from cancers to viruses -- and because they use the cell's own defense mechanisms, they produce minimal side effects.

In addition to treating cancers and genetic disorders, siRNA could prove useful against a variety of rare diseases that have, and always will be, overlooked by big pharmaceutical companies -- the long tail of disease.

People suffering from similar, exotic maladies could band together and recruit a small team of scientists, as if they were the Seven Samurai, to champion their cause and quickly design a cure.

Let’s unravel some of that yarn. What siRNA does, actually, is cause proteins not to be produced, rather than “program cells” to destroy them. The effect lasts for as long as the siRNA is present, so I wouldn’t use the analogy to programming. And it’s true that siRNAs can “turn on a cell’s own built-in disease-fighting mechanisms”, but that’s mostly considered an undesired off-target side effect, which people are still trying to get a handle on. You don’t want to set off immune responses to your RNA therapies, believe me.

And in the next sentence, we get to hear more about programming. But what’s glossed over is that we don’t know how to “program” siRNAs for a wide range of ailments yet, because in most cases we don’t know what causes a wide range of ailments to start with. If you don’t know what protein you want to knock down, you’re not going to get very far with siRNA. And what about the diseases that aren’t caused by single proteins (which is most of them?) Putting cancer in a list like that is a sure sign that the author is either exaggerating or doesn’t understand what’s going on, because cancer is not a disease. It’s several thousand diseases, each of which may need to be addressed differently if we’re going to use the word “cure”.

The next paragraph works in the “long tail” concept, another hook for the intended audience. But look, for example, at something like Gaucher’s disease, which you’d think was pretty far down that tail. Genzyme is doing tremendous business there, because they actually have something – basically the only thing - that helps. For many of these obscure conditions, it’s not so much that we in the drug industry don’t do anything, it’s that we don’t know what to do. And if we’re going to work on something that we’re not sure we can treat or not, which is the usual situation, we’d rather take our chances on something more potentially lucrative.

And that last line, with the Kurosawa reference, is just great. Programming, long-tail, classic foreign movies – this piece must have gone through the editorial process at Wired in about ten minutes. I’ll bet my readers in the drug industry are wondering how they can get together in small teams, whip out their samurai swords, and quickly design cures – admit, you are, aren’t you? Well, the next paragraph of the piece quotes Stephen Hyde of Oxford:

” “The speed with which you develop siRNA drugs is truly amazing,” said Stephen Hyde. “In the past, a traditional small molecule drug might take several years of intensive research effort by a large team of scientists to develop. Today, with siRNA technology, it is possible for a single researcher to develop a drug candidate in a few weeks.”

It’s hard to know which end of that statement to untangle first. If you know exactly which protein you want to target for a disease, then yes, you can then know what sort of siRNA sequence you want to try to knock it down. But is that a drug, as the first line suggests? Nowhere near. Sad to say, you still have those years and years of clinical testing for safety and efficacy to go through.

Now, where Prof. Hyde’s statement makes some sense is in the preclinical world. It does take longer for a team of chemists and biologists to come up with a small-molecule drug candidate, and that’s where the promises of siRNA (and antisense DNA) come in. If you’re targeting the expression of a particular protein (a big if, as I’ve said), then you immediately have a relatively short list of sequences to try, as opposed to the wide-open world of small molecule screening. Chemistry really is only one way to get to a drug candidate, and just because it’s been the way for most drugs until now doesn’t mean it always will be.

But it’s not going to go away, either. Small molecules can do things that changes in protein expression can’t – we can make agonists and antagonists of receptors, for one thing, and we can make inhibitors with varying selectivities across related targets. And there will always be diseases – the majority of diseases – where several things will have to be affected at the same time for any kind of cure to be realized. We’re going to need all the modes of attack we can get.

The rest of the Wired article, to its credit, does mention the single biggest problem with siRNAs: their delivery in vivo. And if you get down to the last few sentences, you can find out that the TMV delivery system has not yet been shown to work in a living animal, could cause immune responses even when it does, and has (as yet) no way to target its delivery to a specific cell population. It is, in other words, an ingenious idea – one of many – that has a long way to go before it sees a sick patient. And we have a long way to go before we have seven-scientist samurai teams cranking out cures in a few weeks. Perhaps we’ll live long enough to see it.