Gene Drives: mendelian genetics has just been overwritten.

[Streetwind]

Guys, we need to talk about this. Right now. This is one of the most amazing things I've seen in all of 2016 so far - more amazing than rocket landings and all that jazz.

It's also one of the most frightening things I've seen in... well, my whole life, probably. Holy crap.

 

A method has been invented that assures that a custom-written genetic trait is passed on to offspring. And I mean assured. The success chance is 100%. It's a "global search-and-replace tool for genetic information" - and the tool passes itself on along with the custom traits it is designed to pass on. This is called a "gene drive". Everytime an organism so modified reproduces, it force-overwrites the genetic information of all offspring to match the gene drive's design specifications. The other partner's genetic information at this juncture is irrelevant. The new offspring, of course, now carries the gene drive too...

And it's easy to do. It only needs moderately specialized equipment, and a skilled highschooler to operate it. There's no secret, patented formula - the approach uses existing tools that everyone, worldwide, has access to.

Now, there's caveats. It doesn't work for viruses or bacteria, among other things (thank heavens!). And while it's easy to build, the act of first designing the blueprint of the genetic change you want to make is actually a really elaborate and difficult process. And it's potentially possible to build targeted "reverse drives".

But still, the mere implications of this! The example given in the video is that you could eliminate Malaria - yes, eliminate - in the span of a single year. At the same time: just one modified organism accidentally escaping from a research lab could potentially end up overwriting an entire species, if it finds a chance to reproduce before it dies.

The presenter rightfully asks: "Have we become God?"

I don't know. But... Fate help us all.

 

 

Edited June 2, 2016 by Streetwind

[Shpaget]

Sounds like a very bad idea.

[Streetwind]

It does, doesn't it?

But it also sounds like such a good idea. So many viruses that cause hundreds of thousands of deaths each year could be eliminated, just by tweaking their carrier species so that the virus can no longer survive inside of them...

[KerikBalm]

this is over a year old

[Streetwind]

Of course it is. You don't get invited to speak at TED unless you publish something first. That conference is not a medium to pull out surprises, but rather a means to spread awareness of things that are "worth spreading".

Which only underlines the fact that we should talk about it, because everyone should know this exists, and clearly not nearly enough people do. This may be the "atom bomb of biology" - something with equal potential to end, or save, millions of lives just through the mere fact that it exists.

[Shpaget]

Yeah, it would be a great idea if we were capable of foresight. How many times do we need to repeat the history? We've already both unintentionally and intentionally introduced invasive species that it's not even funny. We even have genetically modified species running amok bringing imbalance to the wild.

The only word that comes to my mind to describe someone considering using this technology is arrogance.

[Scotius]

Oh my stars...my imagination just went into full-out overdrive. This IS a biological equivalent of splitting the atom. Double edged sword - a miraculous tool and an extinction-level weapon wrapped in one, terrifying package. I can just hope we actually learned something, when we opened gates of Hell over the Hiroshima and Nagasaki.

We can only hope...

[PB666]

Your talking about CRISPR, its been out there for about 2 years, longer than that.

https://en.wikipedia.org/wiki/CRISPR

Notice you haven't seen me panicking about it. Actually they are using it to edit antibody genes so that they can fight cancer and autoimmune disease, so . . . . .

But even so, this just makes it a bit easier than before, we edit genes all the time, you put His tags or antibody tags, move them from human, to bacteria, to chinese hamsters overy cells, back to human cells again if you like.

The point about CRISPR is that it offers the opportunity to remove some really bad diseases from human society while still at the embryonic stage. For example suppose you knew you carried a really bad gene, but you didn't know until after the child was say in its 20s. Well to bad for him/her. Something like Huntington's disease. Now you could do genetic testing in the womb, but then the parents have to make that dreaded A decision, and be you pro-life or pro-choice, if you are the parent and you have to make that decision, there is nothing pro about it. So now you can test embryos' but aren't you playing god with your genes, cause you are selecting out eggs just because of one defect on one chromosome. So a minimalist approach is just to fix that really, really bad guy and don't fix anything else.

Here's another situation.

You are a 55 year old male who spent most of his time studying science and deferring having children as you pushed forth your great new sciency stuff to change the world. But now you have decided to have children. But hold on a minute, while women are more at risk for chromosomal abnormalities after their 40s, mens proto-sperm cells are cooking with fire since the age of puberty, and by the time you are 55, those sperm are well done. So the child's risk for autism and schizophrenia go way up past the age for 40s for males. One thing we can do now at the 8 cell stage is do genome sequencing of both parents and match runs of identity. Then we can look for defects, most are harmless but some are nothing but bad news, so that these can be corrected at the 8 cell stage and then one cell can be used to develop a whole new being without the one or 2 defects. Win, win.

Here's a third thing that can be done. In your body are plueripotent stem cells, they evolve into your immune system and blood cells. There are occasionally defects in these cells that affect the entire body, even though it only takes a few cells in the body to produce these. What we can do for affected individuals is to take the plueripotent stem cells from bone marrow and and fix the defective gene, often a premature stop codon, after expanding those cells in the lab we can put them back, walla, individual is not normal, but they can survive. Imagine giving the bubble boy (remember the boy trapped in a bubble at TCH in Texas medical center back in the 70s, John Travolta made a movie about him), a chance to leave the bubble.

A forth thing that we can do. Suppose that you have been diagnosed with a lymphoma, and they have to irradiate all your immune cells to kill the cancer and give you a bone marrow graft from an unrelated individual, they have an HLA type that is close but not identical. With CRISPR we can sometimes make the HLA major antigen genes identicle. This means that we can stop the possibility of rejection. We can also remove some of your bone marrow cells, and fix the defect that causes the cancer, and put them back. In addition we can use CRISPR to specifically redesign  a few of your natural killer cells to go after the lymphoma when a particular antibody, that we add, is present. In this way the Killer cells will identify one part of the antibody and the other part of the antibody will recognize the lymphoma and only the lymphoma. Does this sound wild, but such bivalent antibodies already exist and are being used. All we have to do is to alter the CD8+ cells so that they recognize the other part.

Most cancers will be a thing of the past in 10 years. Radiation and these harsh therapies are not going to be the first line therapy. You are already seeing major increases in the life expectancy of melanoma and lung cancer patients.

CRISPR is that which makes a hard job easier, it does not do something we could not do before, So now its possible to treat a patient who has months to live before they die instead of after they die, which is kind of useless. 'Hey sir, I manage to insert the right gene into your defective cells, Sir, Sir? Sir? Oh, no.'

 

 

 

[PB666]

Imagine a very long space mission. We have to carry embryos of course, but the mission is so long that humans would not be able to repeat, and at the site we can only create so many people.

By the time the generational voyage is over, the chromsomes have undergone fixation, you emplant embryos but this is not enough, you have at the end the genetic diversity of about 100 individuals. But the target is around 10,000. So that as your population grows you want to get this information back.

Guess what, you don't need to send people, the genetic information is light speed distance to earth away, just send the information from 10,000 diverse individuals and maybe the 3 or 4th child an individual has add some genetic diversity so over time you can, with out sending a molecule, reconstitute the entire useful amount of human genetic diversity. Let recombination do the rest.

Here's an example, you get to the planet, and you find, unbeknownst to you that a deadly virus is very popular in the environment and humans have no natural defenses to that virus, they cannot spread the disease but neither can they defend themselves (sort of like microbial neurotoxins). A molecular analysis of that virus reveals it has some similarities to a bacterial virus that infects bacteria of a certain species of Marsupials. But the marsupials are uneffected. So you examine their MHC class I antigens and you find that they have one molecule and a TLR receptor that provide the defense. Structural modeling reveals that altering the structure of these homologs in humans of two of these allows the production of an innate immune response. Synthesi of a small amount of viral antigen cause the immune system to produce alot of antibody, providing an immediate humoral defense to the environmental toxin. Problem solved.

Another example, the environment has a high level of a certain heavy metal that causes premature aging and senility. That metal is pervasive as a consequence of a deep space asteroid collision with a moon of the planet that continues drop sufficient amounts to produce a health hazard. Remedial measure are quite expensive. A study of benthic worms on earth finds that certain worms living around volcanic outcrops have developed a resistance to this metal by producing a variant of Cytochrome P450 that specifically binds this metal and causes the metal to be excreted. You add a liver, bile duct targeting signal and it is excreted in the waste. It does not remove the toxin from the body, there is but the rate of turnover is faster and does less damage. The sequence of the P450 of human that can be modified can be sent, and engineered into embryos of individuals who are likely to end up exposed. Eventually the gene perculates through the population.

Remember the saying god does not place dice with the universe. The same can be said about DNA, your chromosome is yours to play with, for good or for bad, but those who chose to play will be at selective advantages in most cases. Its not like inserting crystaline proteins from b. thurinigenis in to maize, this technology only tweeks what you have to make it do something ever so slightly different.

I

 

[tater]

Gene drives seem to be an application of crispr. The idea above is putting crispr inside the cells of the target as I understood the ted talk (beat me up if I'm wrong). Meaning putting the the molecular machinery to cut and paste your desired genes an animal that will target any cells it comes in contact with via sexual reproduction. So that with a small % of these individuals introduced to a population, you can rapidly (for things like insects) change the entire population. Add 1% mosquitos that cannot carry malaria to the general population, and in a year no mosquitos carry malaria any more. That was my take away, anyway, that you could change an entire species with this technique over fairly short time intervals (basically X generations, so it's more effective when the generation time is short).

My neighbor is a molecular biologist, I'll ask him about it next time we're having cocktails, we talked about crispr a while ago, and he was pretty excited about it as a tool. Like I said, I welcome our new masters  

Edited June 2, 2016 by tater

[PB666]

17 minutes ago, tater said:

Gene drives seem to be an application of crispr. The idea above is putting crispr inside the cells of the target as I understood the ted talk (beat me up if I'm wrong). Meaning putting the the molecular machinery to cut and paste your desired genes an animal that will target any cells it comes in contact with via sexual reproduction. So that with a small % of these individuals introduced to a population, you can rapidly (for things like insects) change the entire population. Add 1% mosquitos that cannot carry malaria to the general population, and in a year no mosquitos carry malaria any more. That was my take away, anyway, that you could change an entire species with this technique over fairly short time intervals (basically X generations, so it's more effective when the generation time is short).

My neighbor is a molecular biologist, I'll ask him about it next time we're having cocktails, we talked about crispr a while ago, and he was pretty excited about it as a tool. Like I said, I welcome our new masters  

Thats kind of what KRISPR does, except you just KRISPR the KRISPR genes. It does not go every where it can only move serially through the population. There are many variations. BTW we have been similR things since sthe seventies. One of my FFA projects was parasitized by screw worms, which was supposed irradicated in the US. The next week txA&M ag ext service had air drop boxes of sterile bluebottle flies to irradicate the females. This was 1979. 

There are all kinds of schemes. You could have schemes in which you could create a mosquito born parasite that eats falciparum. 

[Bill Phil]

3 hours ago, Scotius said:

Oh my stars...my imagination just went into full-out overdrive. This IS a biological equivalent of splitting the atom. Double edged sword - a miraculous tool and an extinction-level weapon wrapped in one, terrifying package. I can just hope we actually learned something, when we opened gates of Hell over the Hiroshima and Nagasaki.

We can only hope...

Except the atom bomb could never drive us to extinction. Humans are like roaches. We always survive. 

But this could potentially cause extinction. It's far more powerful than splitting the atom.

[KSK]

Whilst I'm not disagreeing PB666, gene drives are another thing you can do with CRISPR and I can understand Streetwind's alarm about those. From reading up on this a little, it seems that a CRISPR based gene drive would basically be a construct which includes the gene of interest plus the cas9 gene and guide RNA needed to insert it. You insert the entire package into the target genome (again using CRISPR) and end up with what's essentially a self-editing cell. If that cell is a germline cell then that self-editing capability can spread throughout the population. Sexual reproduction (what are the bets that gets edited out by the forum software) is required, hence the drive won't work for bacteria or viruses.

I think.

OK, it's just another gene editing tool and so can be used for whatever purposes we desire, benign or otherwise. But it's also a genetic tool that's deliberately designed to be replication competent, which means, in my opinion, that a long hard look at how we plan to use it is certainly warranted.

[Streetwind]

Indeed @KSK, that's precisely the point. It's not that CRISPR is new; much rather, the fact that it is not new is one of the things that worry me most. Because CRISPR is a tool that is available to anyone, worldwide. And now there is an application of this tool that allows anyone to play god and not only create customized species, but also to replace every non-customized individual worldwide with their custom version.

Knowing human nature, it will be abused. It's not a matter of if, it's a matter of when. And when it happens, I hope there will be safeguards and reverse drives available. The one good takeaway from that video is that at least so far, scientists are now talking about the implications and consequences first instead of going ahead and trying it out in the field and worrying about the cleanup afterwards.

[magnemoe]

6 hours ago, PB666 said:

Thats kind of what KRISPR does, except you just KRISPR the KRISPR genes. It does not go every where it can only move serially through the population. There are many variations. BTW we have been similR things since sthe seventies. One of my FFA projects was parasitized by screw worms, which was supposed irradicated in the US. The next week txA&M ag ext service had air drop boxes of sterile bluebottle flies to irradicate the females. This was 1979. 

There are all kinds of schemes. You could have schemes in which you could create a mosquito born parasite that eats falciparum. 

Yes, this is how I understand it, its replace both dna chains, in practical it works just like an perfectly dominant gene. It will always be carried on and will always be active.
It will only progress trough generations so I don't get the moskito experiment in just two generations, yes as the gene start to become pretty common it would rapidly take over. say 25% has the gene and its random mating, 75% does not have the gene and chance for them mating is 56%, 44% would have the gene, next generation only 0.56^2 or 31% would not be affected, then 10% then 1% 

[RainDreamer]

Well, when we discover a way to sequence genomes, we have found the source codes of life. Now with this, we can make mods! I hope we will get good mods instead of ones that crash the whole thing.

[PB666]

4 hours ago, magnemoe said:

Yes, this is how I understand it, its replace both dna chains, in practical it works just like an perfectly dominant gene. It will always be carried on and will always be active.
It will only progress trough generations so I don't get the moskito experiment in just two generations, yes as the gene start to become pretty common it would rapidly take over. say 25% has the gene and its random mating, 75% does not have the gene and chance for them mating is 56%, 44% would have the gene, next generation only 0.56^2 or 31% would not be affected, then 10% then 1% 

But population replacement follows the 2 n rule, so if the population has a size of billions or trillions and you have 20 generations per year, it still takes a long time. Second over many generations you have three fates

1. The mutation is slightly deleterious and alters tona point in all copies that it can no longer complete its task

2. Its neutral and accumulates mutation and heads toward one a bit slower

3. Its selective and maintains purifying selection moving faster through the population. 

[YNM]

How susceptible is the inserted gene to mutation ? If it's low enough then probably you don't need to worry about anything bad happening in the next few decades, maybe millenias.

Unless it doesn't. Then be aware.

[PB666]

9 minutes ago, YNM said:

How susceptible is the inserted gene to mutation ? If it's low enough then probably you don't need to worry about anything bad happening in the next few decades, maybe millenias.

Unless it doesn't. Then be aware.

Mutation occurs at the rate of gene cell line replication. 

 

[KerikBalm]

Its not the ability to insert a gene, and for a new gene to mutate that is the problem.

The potential problem here is the ability to lower genetic diversity in a population. A point mutation in the target PAM sequence, for instance... or likewise in the DNA complementary to the guide sequence, will yield a gene that this gene drive won't work against.

Thus we could target specific alleles, and only remove the "bad diversity" while leaving the rest intact.... like removing debilitating mutations that cause mitochondrial diseases, and defective p53.

You can also design it so that you can remove the gene drive cassette at will with induction of a promoter by a drug, if you are really worried

[PB666]

http://www.lifescience.net/news/864/new-crispr-system-for-targeting-rna/#.V1Fo8yHEacA.reddit

 

[magnemoe]

1 hour ago, PB666 said:

But population replacement follows the 2 n rule, so if the population has a size of billions or trillions and you have 20 generations per year, it still takes a long time. Second over many generations you have three fates

1. The mutation is slightly deleterious and alters tona point in all copies that it can no longer complete its task

2. Its neutral and accumulates mutation and heads toward one a bit slower

3. Its selective and maintains purifying selection moving faster through the population. 

My assumption (unnafected faction)^2 was with random mating, then it would take 9 generation from 1% to 99% affected. 
In reality it would be far slower as most animals are fairly stationary inside an area so you would get edge effects much like how diseases spread but slower as its limited by mating and offspring. 

Mutation of this will probably be like normal mutations, exception is an mutation of only the gene drive payload would be bad. 
Another issue would be if you had multiple and even overlapping gene drive edits. 


 

[PB666]

1 hour ago, magnemoe said:

My assumption (unnafected faction)^2 was with random mating, then it would take 9 generation from 1% to 99% affected. 
In reality it would be far slower as most animals are fairly stationary inside an area so you would get edge effects much like how diseases spread but slower as its limited by mating and offspring. 

Mutation of this will probably be like normal mutations, exception is an mutation of only the gene drive payload would be bad. 
Another issue would be if you had multiple and even overlapping gene drive edits. 


 

Your estimation is not even close, multiply generations by a billion. 

Edited June 3, 2016 by PB666

[KerikBalm]

the target gene has to be a sequence match withthe guide RNA. If your gene drive guide RNA mutates... it stops driving anything.

If two target the same sequence, the first to insert wins, because it changes the sequence and the 2nd has nothing that it recognizes.

I fail to see how your scenarios are problematic.

[PB666]

Before these responses go on into never-never land.

 

https://en.m.wikipedia.org/wiki/Fixation_(population_genetics)#Probability_of_fixation

 

Edited June 3, 2016 by PB666