Is it possible to design a smart-toxin?

[Hypercube]

I was just watching an episode of Fringe which was about a former Nazi who managed to design a toxin which he was able to custom-tailor to literally only target individuals who had certain genetic traits. For example, he could tailor it to only target people who had brown hair, or brown eyes, brown skin, cleft chins, etc. In fact, the toxin could even be tailored to target only a specific person via their specific DNA, or only people of a certain lineage.

 

So I was wondering; would this be possible to do even in principle? I mean I know that it is possible to target specific genes, but could a toxin be designed that could actually do what it did in Fringe? And I'm just talking in principle, I don't care whether the technology actually exists yet.

[Schrödinger's hat]

I suppose if you were super-good at folding proteins you could create some kind of prion which was based on a protein only a few people had.

Even assuming such a prion/protein exists for a given protein expressed by your target, it would either have to be damaging once it caused the target protein to fold, or the target protein would have to be essential (ie. it might just turn a brown haired person grey). On top of that, finding it would be far harder than curing cancer/aids etc.

 

Other options would be making something that will catalyse/react with the expressed protein, or somehow hijacking the immune system. Not sure how plausible that would be.

[Horza2002]

There are a number of anti-cancer drugs that specifically target certain region of the DNA. For example, Brostalicin specfically targets AT tich sequences to where it binds....so maybe if you could talour a molecule to bind specially to a certain base sequence in the DNA, then you could target specific people. The problem would be that since the DNA code is so long, the chances of there being a sequences similar enough to confuse the toxin in an unrelated characteristic is very high

[Mr Skeptic]

Yes, it is possible. It is easier to target drugs across wide evolutionary gaps, like we do with antibacterials, antifungals, pesticides, herbicides, etc. Our cancer drugs are harder, because we're trying to target our own cells but with some semi-random changes, which is really hard. It is in principle possible to target poisons to specific genes, which is one of the way we target cancer drugs. It is also possible to target drugs to certain DNA. However, I can't think of any way we could do that without having the poison be incredibly fragile and also have to be injected directly into the bloodstream. To target to a specific DNA, you'd need to include the matching DNA (or RNA perhaps), with a mechanism to activate the poison when the match is found. This essentially means you'd have to make the poison out of protein and DNA, and that means that it would rapidly be degraded by just about anything. I suppose that it might be possible to make some other kind of molecule to match specific DNA but I don't know how that could be done even in principle.

 

If you want to know about targeted poisons, check out our latest progress on cancer cells. That's a much more appropriate and profitable use for targeted poisons.

[Dan_Ny]

There are a number of anti-cancer drugs that specifically target certain region of the DNA. For example, Brostalicin specfically targets AT tich sequences to where it binds....so maybe if you could talour a molecule to bind specially to a certain base sequence in the DNA, then you could target specific people. The problem would be that since the DNA code is so long, the chances of there being a sequences similar enough to confuse the toxin in an unrelated characteristic is very high

 

Yeah, another example for a DNA binding (and inhibiting helicase in DNA replication) is the mitomycin family. But this area of DNA bonding is still quite new and a lot of work needs to be done - since the targeting site of these molecules is the minor and major groove of DNA, in principle it should be possible to design a "small molecule" for a specific sequence - but, if that sequence is longer than a few base pairs, the inhibitor of helicase must be damn complex and long - and able to wrap around the DNA a couple of times...

If I had 100 years to synthesize some of these molecules, I would do it, but there must be an easier way

 

It would be cool if we could desing a few enzymes which can be somehow produced on demand and which synthesize such an DNA-intercalating toxin in vivo to adress Mr Skeptic's fragility problem....

 

If you want to know about targeted poisons, check out our latest progress on cancer cells. That's a much more appropriate and profitable use for targeted poisons.

 

Could you post some references? That sounds damn interesting.

Edited March 10, 2011 by Dan_Ny