Saturday, February 02, 2008

Cool new anti-retroviral

KP1461 is a nucleoside analog which vastly increases the mutation rate of HIV by incorporating into its genome. But really it's the video I dig...

"KP-1212 can replace cytidine when the viral enzyme reverse transcriptase is building a new copy of HIV, and pair normally with guanosine. It does not terminate the DNA chain. But KP-1212 was chemically designed to be a flexible molecule, such that it can also look like thymidine and then pair with adenosine. This introduces an error that then is locked into the viral DNA.

These errors happen at random, anywhere in the virus; and when they do not kill the virus outright, they accumulate over generations in the DNA of the viral population. The result is eventually an error catastrophe that can wipe out the entire quasispecies, at least in laboratory tests. If you then take the drug away, the virus does not come back. And the cells on which the virus grew are still alive -- cured of the infection.

AZT and the other approved nucleoside analogs terminate the growth of the DNA chain, killing the copy of virus being built. But that copy is easily replaced by other copies that do not have an abnormal error accumulation, so the population as a whole is not damaged. In contrast, KP-1212 continues to add new errors to the population, in addition to the errors that are already there due to the very high normal mutation rate of HIV."


Rob said...

That is an awesome idea. I love the approach, although it is doomed to failure.

One percent of mutations are believed to increase viral fitness. So could we create a supervirus? On theoretic grounds it is possible but not likely at all. HIV has evolved to become the supervirus. Mother nature has been testing billions of mutations for decades. For example, compared to Ebola, HIV is successful because it does not kill its host in days. Evolution has selected against the most pathogenic strains that would kill the host too quickly before the virus could be transmitted throughout a human population, which is of course exactly what the virus from its perspective is trying to do. So one could argue that mother nature has in some ways already created this supervirus. The most efficient and effective virus already exists. And 99 of 100 mutations are going to reduce the virus's fitness, not increase it.

So what is being said here is that HIV has evolved to kill slow. So it's already a super virus. Couldn't that also be read that we could make an HIV that kills fast?? Maybe it won't be as fit in a population but within the individual it will be the fittest in the newly mutated batch.
I also like the comment that it's possible but unlikely to create a supervirus. Unlikely becomes likely when you are talking about billions of virus particles.
Am I wrong?

Bayman said...

Rob asks,
Am I wrong?

Yes. The strategy here is to increase the error rate so much that there's no fidelity in the copying process. Without fidelity a viable population cannot be propagated at all. Neither can evolution happen unless mutation is balanced with fidelity. The easiest way to think about this is to imagine the impossibility of life based on a random, rather than template-directed copying of DNA every doesn't work unless enough information is preserved.

I also think this is a great strategy. It'll be interesting to see how well it works.

Anonymous Coward said...

May I point out one of my previous posts about evolution of robustness in viruses...

Anonymous Coward said...

my concern would be toxicity & safety...

Rob said...

I guess I am making two assumptions:
1. The presence of this drug doesn't eliminate ALL fidelity. It's not going to be present at some saturating level. I assume such a level would be toxic to virus and host.
2. I also assume, although I don't know the exact chemistry of this drug, that it is possible for the viral polymerase evolve to use this drug to quickly establish a new mutation rate that benefits an individual virus.
But I guess your right, at some rate of mutation the viral population is going to hit a catastrophe.

Bayman said...

My assumptions were slightly different.

1. To extinguish the viral quasispecies doesn't require elimination of all copying fidelity. You just have to push the error rate past a certain threshold so that virtually all progeny are defective (catastrophe).

2. Viruses have evolved to maximize mutation rate, and therefore replicate at the edge of catastrophe. On the other hand our cells have much lower mutation rates due to proof-reading activities etc. Therefore there is a therapeutic window for mutation-inducing agents - there should be a concentration of drug that pushes viruses over the threshold but that cells are able to tolerate.

If my assumptions are correct, then I think that there is little chance for any evolution of novel variants to happen - superviruses or whatever.

If HIV replication is no more error-prone than that of our cells, then my assumption #2 is wrong and I would have to agree with you.

What do the experiments say?

Kamel said...

In case reading is wondering (as I was) but didn't check out the link, KP-1461 is the prodrug which is intracellularly triphosphorylated to make KP-1212, hence the different nomeclature used in the post.

Rob said...

I'm pretty certain that viruses have not evolved for maximal mutation rate. (your assumption #2)They have evolved a mutation rate that is maximally beneficial, for sure, which for some viruses I'm sure is very low. The post that the AC mentioned kind of addresses that. High fitness peaks favour low mutation rates. A quick survey of a couple of articles suggests that the mutation rate of HIV though is quite high.