I was just reading through this really encouraging sum-up of progress on preventing and treating HIV:
On the whole, it's great news. I mean, as much as I dislike making circumcision mainstream, if it saves lives in a continent where AIDS is at full epidemic strength, it's a no-brainer. I just hope we can move past it with more advanced methods at some point in the future.
However, the bit that caught my attention was the ongoing (though low-level) debate as to whether a practical cure for HIV will ever be possible. That's because HIV, like many common viruses, integrates itself into the genome of the host cell when it infects, and often goes dormant for weeks, months or years in an individual cell. This means that, functionally speaking, there is no viral activity in the cell anymore. Only a string of DNA instructions that could, at some point, emerge as a virus. There's no currently available method of treating a string of unwanted DNA, and some people will argue that there never will be.
Not so! There is quite a great deal you can do about DNA, and hopefully we'll see progress on it soon. It falls under the heading of DNA logic; methods of reading and responding to DNA in an intelligent manner. This sort of technology isn't as well-developed as you might expect from the movies, but it's entering a phase of rapid advancement at the moment with falling prices in prototyping costs for DNA. When you can test your idea for only 200 euro, whereas before it would cost either a year of work or 2000 euro, you can make a lot more progress in less time.
What's DNA Logic?
DNA logic can either work by protein-mediated reading of DNA, or by RNA-mediated reading of the RNA copies of DNA that are used to make new viruses. The second one could be used to make treatments that prevent viruses from emerging, silencing the nascent virus before it can awaken. The first one, though, is more interesting; when a protein “reads” target DNA, it can be designed to do anything from activating it to cutting it out entirely. This is where the cures will be found.
Here's how it would work: A researcher uses web-tools such as those provided by the Zinc Finger Consortium (Open source biotech baby!) to design a pair of small proteins that each bind at either end of a crucial part of the HIV genome that doesn't change much. When used together, these proteins will naturally meet one another and interact only (statistically speaking) when they have bound to the DNA of a secret virus genome embedded within the genome of an infected cell. Great!
Payloads for DNA-targeting Proteins
Now, to do something with them. Traditionally, the researcher might add bits of other natural proteins called “nuclease domains” to the little DNA-binding proteins, which only work when they are brought close together by successful DNA binding. These domains cut DNA, creating a lethal break that will probably be repaired to the severe detriment of the virus. Creative selection of nuclease domains might even cut the virus out entirely, rather than just poking holes in its genome.
However, you can be more creative if the nucleases don't work well: you can activate the virus. Sounds bad, right? Not exactly. The problem with HIV is precisely that it doesn't activate! When you treat with antiretrovirals, you're preventing the virus from successfully producing baby viruses, while the immune system tracks down cells that are trying to produce new viruses and kills them. However, the immune system fails to find cells that carry the virus but aren't actively producing it. If you combine anti-retroviral drugs with a treatment that activates virus genomes fully, you can slash the number of cells carrying a dormant genome, and approach a cure.
Delivering the Treatment
How to get the proteins into the cells you need? That's actually more challenging than designing and testing the proteins, and poses more of an ethical problem. You might try to use a virus to carry them into the cells, but you'll have to fight the immune system to cure it. You could try using synthetic carriers that bring the DNA coding for the proteins, or the proteins themselves, into the cell..but sometimes they're toxic, and the efficiency is pretty low.
Probably the best method is to coat the proteins in something that'll hide them from the immune system but which will shed inside the cell, and to include what's called a “Nuclear Localisation Sequence” in the proteins so that the cell will carry them into the Genome immediately on entry. Then you blanket-bomb the body with them, particularly the lymph system, and cross your fingers.
Get Back To Work Cathal
For all my talk, you'd think I'd be doing this.. The take home message is that HIV will see a cure, and it'll be from Open-Source Biotech if we're lucky. My job at the moment is making OSB a reality, so perhaps stuff like the above can happen quickly rather than relying on companies that don't want to cure a multi-billion-dollar-treatment pandemic. Hold that thought, I'm going back to work.