Hacker News Reader

Comment by Metacelsus | Hacker News Reader

The big problem is that the CRISPR would need to be delivered to 100% of infected cells. And as far as I know (I'm a biologist) nobody has a method that can achieve a delivery rate of 100% in vivo.

View on HN

There are specific latent reservoirs the HIV hides in. It wouldn’t need to get every cell in the body, just have a favorable probability distribution to hit the reservoirs.

If you can get a reasonably high percentage then this might not be a problem. Include something in the crisper treatment that acts as a marker that protects the cells to which the treatment is successfully delivered. Then kill everything without the marker. I don't know what fraction of of your CD4 cells you need to maintain a healthy immune system, but I imagine even a double digit percentage reduction is probably tolerable, especially if the reduction is only temporary.

In-vivo is hard but there are many talented teams working on this. In animals, a company called Verve has discovered how to make in vivo CRISPR work via nanoparticles to the liver which achieves >80% delivery at certain concentrations of nanos. Pretty fascinating

I thought the biggest problem was off-target editing (I'm not a biologist)

That would be for a cure, but not for a recurrent therapy

Do we even need a cure or therapy if we can essentially use this as a vaccine?

Doesn't help people now, but may eliminate it as a threat for future generations.

Well you can’t use it as a vaccine since there would be no hiv to preemptively excise.

What we need is a reliable CRISPR vaccine that confers the CCR5 SNP and essentially give it to everyone.

Yeah, we don’t even need to get every cell at that point, just a sizable number of the CD4 cells that are relatively easily accessible. The HIV would destroy every CD4 that isn’t CCR5 by itself. The antibodies would clean up the rest of the virons. Theoretically that would work as a cure too if you could excise the hiv and insert the CCR5 widely enough.

Could that risk evolutionary pressure on whatever is left behind so that mutations emerge that are more aggressive?

If we assume that the treatment needs to reach the HIV cell to effect it, and that the danger is not 100% of the cells are reached, would it produce evolutionary pressure? The survivors aren't interacting with whatever got its buddies, right?

That itself is an evokutionary pressure. Whatever allowed them to get missed would get amplified by reproduction. Only if it's entirely utterly and only random chance and not something like the method doesn't target certain cells would it not be applying pressure

If you bomb a cow pasture with napalm, you aren't going to be left with napalm resistant cows. Just lucky ones. And luck isn't genetic.

You're viewing the cows in isolation here. Replace the cows with cancer cells with different mutations and it doesn't work anymore.

There's research going with certain types of cancer where the doses of cancer-killing drugs isn't intended to eradicate all cancer cells, but to keep it at a level that's manageable with future treatments.

The theory being that if you go too far, cells that are left with a "lucky" mutation can grow without competition, leaving the patient in an even-less treatable state down the line.

... Cancer cells? This is about HIV. What are you talking about?

you're talking about cows in a HIV discussion, so I could have said "cows and napalm? we're talking about HIV!", but I'm charitably engaging with your analogy to show that evolutionary pressure is an active research area in disease treatment.

But there's a chance that any survivors learn to hide from planes and you've now given an evolutionary pressure to cows to hide when hearing airplanes or helicopters. Along with that a pressure that makes the ones that blend in less likely to be spotted to be bombed. So yes, you're right you probably won't get cows that can survive napalm itself but now you've got cows with camouflage and an instinct to hide.

The selection pressure would have to be some form of resisting the crispr tools from entering the cell or functioning in the cell.

The HIV genome is small and I would be surprised if it were so easy for such a thing to come about.

I'm thinking more of a mutation already present that has less treatable traits but can't spread because neighbouring cells are already infected. But then when the treatment is delivered, it isn't cleared because of the difficulty of the delivery mechanism, but the treatment does clear up the surrounding cells, providing fresh ground for the mutated strain to reinfect.

(This is an active research area I've read about in cancer treatment, where it's been suggested that being too aggressive with clearing certain cancers encourages reservoirs of already mutated aggressive variants to be unleashed after treatment whereas before the avenues they had to spread were already infected with the less aggressive variant).

... and so?

Evolutionary pressure is present in any situation and is unavoidable. It still has an effect even in cases where treatments are effective with a success rate of 100%: the selection of diseases for which there is no cure.

Entertaining such a perspective can lead to an extreme and undesirable conclusion, that the only logical outcome is to do nothing and dismantle the entire modern medical practice.

It's not very constructive.

Other posters have explained why my concerns aren't warranted for this situation, I'm asking the questions to learn a little.

>It's not very constructive. >Entertaining such a perspective can lead to an extreme and undesirable conclusion, that the only logical outcome is to do nothing and dismantle the entire modern medical practice.

Disagree, if you read my parent comment in this thread, I talk about the research going on in cancer to account for evolutionary pressure, and the conclusion they arrived at is nowhere near "do nothing".

It is instead a change in treatment protocols to aim for less aggressive treatment, the clinic in questions motto is for the patients to die from something other than their cancer by having them on a protocol that has them live permanently with their cancer under the thesis that aggressrive attempts at eradication risk a stronger variant resurging.

Obviously it's not applicabale for all types of cancer which is why it's research, and why I'm not telling anyone that evolutionary pressure is a concern here, but am simply asking the question to learn if there are similar concerns between cancer and hiv disease treatments.

This doesn't make sense. HIV virus is generally not "substrate limited"/competing with other virii in its propagation in patients in (temporary) remission -- that's basically all of them that don't have AIDS. HIV retro integrates into host DNA and sporadically resurges on a stochastic basis, the drug cocktails in use inhibits this process.

Generally the viral load of a patient with HIV is miniscule but must always be kept in check.

Note that the point of a crispr treatment is that it would delete dormant HIV sequences that are hiding in the host DNA and not currently producing virii but threatening to pop up at some time in the future.

Obviously this too is subject to evolutionary pressures but almost certainly not the ones you are thinking of.

That's good to know thanks, I'm very much a layperson in my readings and was curious as to how it differed.

Does it need to be 100%? We have dormant viruses that are domant and would only activate when conditions are good, like colds, and we have ways to deal with it. It’s an analogy, for aids we would deal with it with meds and press it back there by turning it into a sort of nuisance like colds. The issue is cost and stigma of passing it. I for one still wouldn’t want aids to be as prevalent as the cold even if it’s super easy to deal with.

Reply on news.ycombinator.com