Super Awesome HIV antibodies: Its the same old story, same old song and dance, my friend

Antibody Discovered for 91% of HIV Strains


For all intents and purposes, Ive talked about this before.


Your adaptive immune system operates by, what I would call, picoevolution. It is chance + random mutations + natural selection.

Lets say both you and I get infected with the exact same variant of swine flu. We both fight it off successfully, and generate a potent neutralizing antibody response to swine flu. My neutralizing antibodies are nothing like your neutralizing antibodies. Chance, randomization, and selection led to different 'answers' to swine flu in both of us... and everyone else infected with swine flu.

Lets say you and I both got the swine flu vaccine last year, thus we never got sick. The antibodies my body generated to the vaccine are nothing like the antibodies your body generated to the vaccine-- even if we are both equally protected. Chance, randomization, and selection led to different reactions to the swine flu vaccine in both of us... and everyone else that got the vaccine.

But lets say, by chance, my body generated an antibody that neutralized 100% of all influenzas. I would never get the flu ever ever ever again. There is no way to force anyone elses body to create that antibody. The patients who they isolate these super awesome antibodies from? They have had HIV-1 for a long time. Which means there have been many many many rounds of chance, randomization, selection, chance, randomization, selection, chance, randomization, selection. You cant get these kinds of antibodies from a vaccine, which are normally one hit (proteins), one round (pseudoviruses, dummy viruses). Live attenuated viruses are never going to happen with HIV-1. There is just no way to make people make this antibody, no matter how awesome it is... short of gene therapy:

Finally, there are experimental methods that employ tactics such as gene therapy. Nobel laureate David Baltimore, with funding from the Bill and Melinda Gates Foundation, is working on one such approach.

His team at the California Institute of Technology in Pasadena, Calif., has stitched genes that code for antibodies into a harmless virus, which they then inject into mice. The virus infects mouse cells, turning them into factories that produce the antibodies.

That isnt exactly a viable option in most cases, or with most people at risk for HIV-1 infection.

The best use of this antibody I see is one they already thought of-- Put the antibody in spermacides/microbicides.

The antibody could also be tested in a "microbicide," a gel that women and receptive partners in gay male pairings could apply before sex to prevent infection.

A big problem with HIV-1 vaccines is, sure you might make great anti-HIV antibodies or anti-HIV cytotoxic T-cells-- but they have to be in the right place at the right time. If the anti-HIV CTLs arent hanging out at the site of infection, and HIV gets a foothold, youre still screwed. But putting this antibody AT the site of infection would be the way to go.

But still, no, super awesome antibodies arent a cure for HIV/AIDS.

More like this

There you have to go and get all "sciency" and let your "facts"* get in the way of euphoria.


*As Colbert said, it's well known that facts have a liberal bias.

I see another problem with these antibodies, aside from the ones you point out.

91% effectiveness, in regards to retroviral therapies, is not much better than 0% effectiveness. The reason we use HAART therapy is because most of the anti-HIV drugs are "only" effective against 99.9% of HIV, allowing for the evolution of resistant strains.

Use of this antibody as a therapeutic, or preventative, will have the same eventual outcome - temporary effectiveness, followed by the proliferation of resistant strains.


Optimus-- Very good, young padawan (mixing movie references, *wince*)

Alex-- They dont call me Debbie Downer for nuthin!

Bryan-- I would agree, BUT...

Using it in microbicides is a good idea. Youre taking a very rare event, and making it more-rarerer. Yes, an antibody resistant virus would make it through, but if you werent using that antibody as an antiviral therapy, the quasispecies wont keep that particular bit of variation (especially if escape comes at a fitness cost, which they havent studied). Thus you could still try to keep THAT person from passing it on to someone else!

I do still think this has potential.

I think the idea is sound for vaccination, you don't need everyone to make THAT antibody, just an antibody that targets the same region. Unfortunately, it looks like this one has a binding sight that's over a fairly broad area of the protein structure, but that doesn't mean that this approach can't yield a binding site that could be replicated with a small peptide.

As for getting enough somatic hypermutation to get it going, that's what multiple doses are for.

The reason that antibodies are made out of proteins is because that is all that cells know how to make. Given that factories can make stuff out of other things too, like metal, plastic and ceramics, can't antibody-like bits be made that will bind to this HIV "magic spot"?

Presumably they know the shape of this antibody, and so they know the shape of the "magic spot" on HIV.

@ daedalus2u - The problem is, no drug or synthetic molecule is going to hang out in the body for very long. If you get a vaccine working, memory B-cells will hang out for ages pumping out antibodies, and if they actually see a virus, they'll pump out more for good measure. Plus, antibodies have special properties that allow them to hang out in the blood-stream for a long time.

Recombinant (grown in a dish) antibodies and other proteins ARE used to treat certain diseases (any medication that ends in "-mab" is a Monoclonal AntiBody), but this would require transfusions every couple weeks/months, and isn't realistic for a prophylactic.

My impression from people I know working in this area is that the "magic spot" on HIV is actually pretty floppy and not well-defined.

While I don't know much about how the immune system works, an artificial antibody would also need interact with cell receptors which could be tricky as well.

But I think the main problem would be the kind of the target somebody would be going for. It may be my background hanging around bioengineers, but it seems if somebody is looking to design something that will have some kind specific molecular interaction - the material of choice would probably be biological.

Kevin, I agree with you in the body, but Abbie was talking about using this in microbicides, which are only in the body for a short time.

But transmission of HIV as an STD might not be via free virus, but as infected cells (which a microbicide targeting the magic spot of HIV probably wouldn't work on).

daedalus - I see what you mean. But antibodies are pretty cheap/easy to make once you know what sequence you want, so churning out tons of antibodies for use in a topical microbicide is probably easier than trying to devise something synthetic.

Abbie - did you see this recent Nature Reviews?
I didn't read it super closely, but it seems like they're arguing that trying to get Ab - based vaccines against conserved motifs is the best way to go. The paper you talk about here seems to suggest that if we got that antibody response, it could be effective. I'm pretty ignorant about HIV though, do you have reason to think this won't work?

They're not that cheap/easy to make... 100 mg of antibody is almost always $300. Rituximab, another clinical mAb, is $568 for 100 mg. Granted, microbicidal antibodies will not need to be as free of contaminants as antibodies for IV treatment. But still, as a potential prevention of infection, it's a nonstarter I think. The mean IC50 (half-effective concentration) for all the tested viral isolates was 0.33 ug/mL. Meaning that a tube of semi-effective microbicide would be about $1/mL (a tube about the size of a travel sized toothpaste tube would be $30). Especially in Africa - circumcision would be a far more cost effective measure even if these antibodies were literally 100% effective.

@ Sylphs - Fair enough, I admit I'm thinking cheap/easy in a laboratory setting. When we have hybridomas we can churn out liters of antibody in relatively short order. Give that to a pharmaceutical company and add the sort of purity/quality control you'd need for therapeutic use, I can see how it might get expensive.

The Nature news story has some interesting background on the extent of affinity maturation that led to the generation of VRC01:

I'd been trying to find out what the normal average number of mutations is in antibody-producing genes during affinity maturation and had found a reference to nine, in the article Peter Kwong says it's 10-15. There were 66 in the B cell that made VRC01, and according to the paper they reverted them all in stepwise fashion and saw the neutralization potency erode to zero as they did so.

By Richard Jefferys (not verified) on 09 Jul 2010 #permalink

Oh, for Pete's sake, move to Boston already, so I can hang out with you and absorb your smarts via osmosis. As always, wonderful post.

Abbie, quick question from an interested layperson: given that HIV has such a fondness for CCR5, has anyone tried making an externally-applied gel that contained fake CCR5 receptors that could bind to the virus before it has time to infect? Or would that just not be effective since it only takes one unbound virion to infect the first T cell?

@Chronos - Lots of "fake" receptors for HIV have been made: alone, in synthetic structures, fused to the Fc portion of antibodies etc. A small molecule drug inhibiting the interaction between HIV and CCR5 has even been approved (Maraviroc), but the whole topical application/gel business is a bit of a distant dream.

How do you keep a drug stable in the conditions of the vagina/anus, whilst also ensuring there's no immune activation or damage to the musoca? Trials so far haven't had any success with any microbicides.

In terms of "fake" protein receptors or antibodies as microbicies, this problem of stability is like 1000x more difficult and any resulting gel or insertable ring 1000x more expensive.

Given that passive immunization of broadly neutralizing antibodies to patients hasn't even been particularly successful, i kind of feel like the whole HIV antibody community is grappling for slightly stupid excuses to justify our work, but of all of the stupid suggestions, I do actually like what Peter Kwong is saying about these highly matured nAbs - that we need to use repeated vaccination to "guide" antibody maturation.

@daedalus2u , Kevin , Sylphs, and whoever else was talking about synthetic antibodies -

Antigen binding drugs don't "have" to be antibodies, but antibodies are so crazy-good at binding antigens on account of their flexible structure, and crazy-good at eliminating pathogens because they don't just bind stuff, but recruit other proteins and cells to get rid of the shit they've bound.

In the case of HIV it's beginning to look like the ability to recruit certain cells to the virus, or virally infected cells is pretty key for induction of sterilizing immunity.

Designing stuff from scratch to do all of this is not easy, but there was recently a paper in JACS outlining the use of "Molecularly Imprinted Polymer Nanoparticles" aka "plastic antibodies", so who knows...

I recently read a very interesting article about an African American whose antibody neutralizes 91% of the virus. I began to feel elated, until I read ur post, Thanks for the dose of reality. Makes me wonder will a cure ever be found?

The reason that antibodies are made out of proteins is because that is all that cells know how to make. Given that factories can make stuff out of other things too, like metal, plastic and ceramics, can't antibody-like bits be made that will bind to this HIV "magic spot"?

Presumably they know the shape of this antibody, and so they know the shape of the "magic spot" on HIV.

i now very happily retract my suggestion that microbicides always damage the host more than the virus :)))

This post was inspired by a long weekend I spent browing your blog! So thanks for what you do, and thanks for your comments here.

By Activating tra… (not verified) on 31 Aug 2010 #permalink