![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
stormsewerI don't often read about scientific advances that give me a sense of vertigo, but it did happen recently. And no, actually, it's not the idea that we may soon be editing the genomes of human embryos, which has gotten people quite riled up [1]. It is work out of a lab working towards the same or similar goals, though: the Church lab at Harvard [2].
The article is entitled "RNA-guided gene drives can efficiently bias inheritance in wild yeast" [3]. I'm fond of saying that humans will soon be taking control of their own evolution, but the work here was a punch in the face for me regarding with how very possible that is going to be in the near future.
Background
Okay, so what did they actually show? First, two pieces of background. (To skip the background, go here; to skip the technical details entirely, go here.)
(1) CRISPR-Cas. This is what they used to do what they did. You may have heard about it; it's a new genome-editing system that's making a lot of waves. Now, technically there's nothing the CRISPR-Cas system does that we couldn't already do, but it is so much easier than previous methods, and it seems to work in some form or another in essentially every organism that's been tried. That's a big deal.
At the basic level, all CRISPR-Cas does is cut DNA wherever you tell it to. The fun part comes in how cells respond to that. Bacteria usually die [4], which has led to new developments in anti-bacterials by delivering CRISPR-Cas on bacteria-specific viruses [5]. However, in eukaryotes (which in simple terms includes all complex organisms from yeast on up), the cells will look around for intact DNA similar in sequence to the DNA on either side of the cut and stick that in instead. And if the new DNA is a little different, or has completely new stuff snuck in between the parts that actually match, the cell doesn't necessarily care. So by cutting the genome and strategically providing the right genetic band-aid, you can change the genome. Or you can choose not to provide any band-aid, which can leave the targeted gene permanently broken.
(2) Yeast. This is the organism they did it in, but a little background in yeast genetics is necessary to understand their results. We as humans are diploid, meaning that each of our cells contains two slightly different copies of our genome. Haploid organisms, like bacteria, only contain one copy of their genomes. Yeast, however, can switch back and forth. Two haploid yeast cells can fuse to form a diploid cell (this is called "mating"), and a diploid can then divide into two diploid cells or four haploid cells.
What they actually did
As a test case, DiCarlo et al. used a gene (ADE2) that turns the yeast cells red when it is absent and yellow when it is present. So normally, if you take a haploid red cell and mate it with a haploid yellow cell, you'll get a diploid yellow cell, because one copy of the ADE2 cell is present (in other words, it's a dominant gene). If that diploid yellow cell then splits into four haploid cells, you expect half will get the gene and be yellow, and the other half won't and will be red.
What they showed is that if you give a red (ADE2-lacking) haploid cell a CRISPR-Cas construct that cuts the functional version of the ADE2 gene and mate it with another haploid cell harboring a functional copy, all the diploids resulting from the mating will be red, and all the haploid progeny will also be red. This is because the functional gene gets cut and then repaired with a non-functional template, leading to red cells. They showed that they can mate their red cell with all kinds of naturally occurring yeast, and 99% percent of the offspring will end up with the red cell's version of that gene.
What it means
This is a gene drive; a gene that removes other versions of itself from the genomes it inhabits. We now have the ability to make them easily, targeting any gene we want, in essentially any organism we want. CRISPR-Cas9 has already been done in monkey embryos with efficiencies on the order of 50%, apparently without any off-target mutagenesis [6], and you can bet they're working hard on making it even better. My mind threatens to blue screen thinking of the possibilities.
For instance, if you engineer an organism with this construct inside it and release it into the wild, all of its offspring will have the versions of its genes that you specify. Now, such gene drives have already been implemented, with the most notable example being attempts to get mosquitoes engineered to be malaria resistant to take over the mosquito population [7]. But even that makes me wary, and this new method will be so much easier to pull off than the old.
Now, my old criticisms of apocalyptic bioengineering scenarios still stand. If whatever change you're making doesn't actually benefit the survival and reproductive prospects of the organism it's being made in, it won't last. And it is so much easier to break a biological system than improve it. But this still presents us with the possibility of experimenting with the evolution and ecology of the biosphere on an unprecedented scale. And I can easily envision it being done by people who aren't interested in caution or playing by the rules. Today, maybe we make mosquitoes resistant to malaria, but what will we get it in our heads to try after that? The history of human ecological interventions isn't extremely impressive, at least far as the well-being of any species but us is concerned (and the jury's still out even on that).
Gene drives and engineering human embryos
This is also a big step forward in making Gattaca reality, but let me take a moment to explain why the ability to alter human embryos on its own doesn't freak me out nearly as much as gene drives do. Talking about making targeted modifications to the human germ line (in other words, the cells that will be passed on to your children, as opposed to the vast majority of your cells that won't) can be labeled eugenics. This in turn calls to mind early 20th century scenes of forced sterilizations and ethnic cleansing. I consider eugenics to be essentially the institution of a forced (or least coerced) breeding program on the human population, and I think at this point it's self-evident that that is a terrible idea.
Modifying human embryos, then, doesn't qualify as eugenics to me if the parents aren't coerced into doing it, which I think is the most likely outcome. Individual couples would likely be making the decision of whether and how they want their embryos modified. Obviously there are good arguments about why we should be very cautious about this, but if you could ensure that your children didn't have the genes for Alzheimer's or breast cancer or cystic fibrosis, wouldn't you? My expectation is that, current qualms notwithstanding, once one country somewhere manages to implement it successfully, everyone else will need to get on board as well if they don't want to get left in the evolutionary dust. As mentioned at the beginning, it is looking very much like more than one group is working toward doing exactly this [1] right now. I can imagine it not getting a legal green light in our lifetimes, but as long as the technology exists it seems inevitable to me that humans will eventually take on a more direct role in deciding how they evolve as a species. I'm inclined to think that's a good thing [8].
Gene drives, however, change the coercion equation significantly. Even at the molecular level, a gene drive is removing certain alleles from the gene pool, not just tossing some new ones in and letting the chromosomes fall where they may. Putting them into germ-line cells would allow you to control which combinations of the parents' genes actually grow into viable embryos. Not only that, but it would control the combinations in all future generations arising from those embryos, as long as the gene drive remains functional. That does start sounding much more like old-school eugenics.
It doesn't stop there. I can imagine using viruses to deliver gene drives to many members of the population at once, and it would likely not be hard to target them to, say, everyone with blue eyes, or dark skin, or a Y chromosome, or any other inherited characteristic. (And I bet you could engineer versions of CRISPR-Cas that differentiate epigenetic (non-inherited) DNA states, as well.) And the next time something like the Nazi regime comes to power, what's to stop them from having a go? I'm not really sure, but it's something we should be thinking about. This kind of thing has been a staple of science fiction for a while, but for the first time I can envision it actually happening.
Cat, say goodbye to the bag.
PS
Footnotes
[1] The references, in case those links go bad:
Regalado, A. Engineering the Perfect Baby. MIT Technology Review (5 March 2015).
Lanphier E et al. Don't edit the human germ line. Nature 519: 410–411 (26 March 2015). doi:10.1038/519410a
[2] Where I very nearly ended up for my postdoc, but let's not get into that.
[3] DiCarlo JE et al. RNA-guided gene drives can efficiently bias inheritance in wild yeast. bioRxiv: 013896 (16 January 2015). doi: 10.1101/013896
[4] Bikard D et al. CRISPR Interference Can Prevent Natural Transformation and Virulence Acquisition during In Vivo Bacterial Infection. Cell Host & Microbe 12:177-186 (16 August 2012). doi:10.1016/j.chom.2012.06.003
[5] Citorik RJ et al. Sequence-specific antimicrobials using efficiently delivered RNA-guided nucleases. Nat Biotechnol 32:1141-1145 (21 September 2014). doi:10.1038/nbt.3011
Bikard et al. Exploiting CRISPR-Cas nucleases to produce sequence-specific antimicrobials. Nat Biotechnol 32:1146-1150 (5 October 2014). doi:10.1038/nbt.3043
[6] Niu et al. Generation of Gene-Modified Cynomolgus Monkey via Cas9/RNA-Mediated Gene Targeting in One-Cell Embryos. Cell 156:836-843 (13 February 2014). doi:10.1016/j.cell.2014.01.027
[7] Windbichler et al. A synthetic homing endonuclease-based gene drive system in the human malaria mosquito. Nature 473:212-215 (12 May 2011). doi:10.1038/nature09937
[8] Though I can of course imagine it creating nightmarish class divides. As William Gibson said, "The future is already here; it's just not very evenly distributed."
The article is entitled "RNA-guided gene drives can efficiently bias inheritance in wild yeast" [3]. I'm fond of saying that humans will soon be taking control of their own evolution, but the work here was a punch in the face for me regarding with how very possible that is going to be in the near future.
Background
Okay, so what did they actually show? First, two pieces of background. (To skip the background, go here; to skip the technical details entirely, go here.)
(1) CRISPR-Cas. This is what they used to do what they did. You may have heard about it; it's a new genome-editing system that's making a lot of waves. Now, technically there's nothing the CRISPR-Cas system does that we couldn't already do, but it is so much easier than previous methods, and it seems to work in some form or another in essentially every organism that's been tried. That's a big deal.
At the basic level, all CRISPR-Cas does is cut DNA wherever you tell it to. The fun part comes in how cells respond to that. Bacteria usually die [4], which has led to new developments in anti-bacterials by delivering CRISPR-Cas on bacteria-specific viruses [5]. However, in eukaryotes (which in simple terms includes all complex organisms from yeast on up), the cells will look around for intact DNA similar in sequence to the DNA on either side of the cut and stick that in instead. And if the new DNA is a little different, or has completely new stuff snuck in between the parts that actually match, the cell doesn't necessarily care. So by cutting the genome and strategically providing the right genetic band-aid, you can change the genome. Or you can choose not to provide any band-aid, which can leave the targeted gene permanently broken.
(2) Yeast. This is the organism they did it in, but a little background in yeast genetics is necessary to understand their results. We as humans are diploid, meaning that each of our cells contains two slightly different copies of our genome. Haploid organisms, like bacteria, only contain one copy of their genomes. Yeast, however, can switch back and forth. Two haploid yeast cells can fuse to form a diploid cell (this is called "mating"), and a diploid can then divide into two diploid cells or four haploid cells.
What they actually did
As a test case, DiCarlo et al. used a gene (ADE2) that turns the yeast cells red when it is absent and yellow when it is present. So normally, if you take a haploid red cell and mate it with a haploid yellow cell, you'll get a diploid yellow cell, because one copy of the ADE2 cell is present (in other words, it's a dominant gene). If that diploid yellow cell then splits into four haploid cells, you expect half will get the gene and be yellow, and the other half won't and will be red.
What they showed is that if you give a red (ADE2-lacking) haploid cell a CRISPR-Cas construct that cuts the functional version of the ADE2 gene and mate it with another haploid cell harboring a functional copy, all the diploids resulting from the mating will be red, and all the haploid progeny will also be red. This is because the functional gene gets cut and then repaired with a non-functional template, leading to red cells. They showed that they can mate their red cell with all kinds of naturally occurring yeast, and 99% percent of the offspring will end up with the red cell's version of that gene.
What it means
This is a gene drive; a gene that removes other versions of itself from the genomes it inhabits. We now have the ability to make them easily, targeting any gene we want, in essentially any organism we want. CRISPR-Cas9 has already been done in monkey embryos with efficiencies on the order of 50%, apparently without any off-target mutagenesis [6], and you can bet they're working hard on making it even better. My mind threatens to blue screen thinking of the possibilities.
For instance, if you engineer an organism with this construct inside it and release it into the wild, all of its offspring will have the versions of its genes that you specify. Now, such gene drives have already been implemented, with the most notable example being attempts to get mosquitoes engineered to be malaria resistant to take over the mosquito population [7]. But even that makes me wary, and this new method will be so much easier to pull off than the old.
Now, my old criticisms of apocalyptic bioengineering scenarios still stand. If whatever change you're making doesn't actually benefit the survival and reproductive prospects of the organism it's being made in, it won't last. And it is so much easier to break a biological system than improve it. But this still presents us with the possibility of experimenting with the evolution and ecology of the biosphere on an unprecedented scale. And I can easily envision it being done by people who aren't interested in caution or playing by the rules. Today, maybe we make mosquitoes resistant to malaria, but what will we get it in our heads to try after that? The history of human ecological interventions isn't extremely impressive, at least far as the well-being of any species but us is concerned (and the jury's still out even on that).
Gene drives and engineering human embryos
This is also a big step forward in making Gattaca reality, but let me take a moment to explain why the ability to alter human embryos on its own doesn't freak me out nearly as much as gene drives do. Talking about making targeted modifications to the human germ line (in other words, the cells that will be passed on to your children, as opposed to the vast majority of your cells that won't) can be labeled eugenics. This in turn calls to mind early 20th century scenes of forced sterilizations and ethnic cleansing. I consider eugenics to be essentially the institution of a forced (or least coerced) breeding program on the human population, and I think at this point it's self-evident that that is a terrible idea.
Modifying human embryos, then, doesn't qualify as eugenics to me if the parents aren't coerced into doing it, which I think is the most likely outcome. Individual couples would likely be making the decision of whether and how they want their embryos modified. Obviously there are good arguments about why we should be very cautious about this, but if you could ensure that your children didn't have the genes for Alzheimer's or breast cancer or cystic fibrosis, wouldn't you? My expectation is that, current qualms notwithstanding, once one country somewhere manages to implement it successfully, everyone else will need to get on board as well if they don't want to get left in the evolutionary dust. As mentioned at the beginning, it is looking very much like more than one group is working toward doing exactly this [1] right now. I can imagine it not getting a legal green light in our lifetimes, but as long as the technology exists it seems inevitable to me that humans will eventually take on a more direct role in deciding how they evolve as a species. I'm inclined to think that's a good thing [8].
Gene drives, however, change the coercion equation significantly. Even at the molecular level, a gene drive is removing certain alleles from the gene pool, not just tossing some new ones in and letting the chromosomes fall where they may. Putting them into germ-line cells would allow you to control which combinations of the parents' genes actually grow into viable embryos. Not only that, but it would control the combinations in all future generations arising from those embryos, as long as the gene drive remains functional. That does start sounding much more like old-school eugenics.
It doesn't stop there. I can imagine using viruses to deliver gene drives to many members of the population at once, and it would likely not be hard to target them to, say, everyone with blue eyes, or dark skin, or a Y chromosome, or any other inherited characteristic. (And I bet you could engineer versions of CRISPR-Cas that differentiate epigenetic (non-inherited) DNA states, as well.) And the next time something like the Nazi regime comes to power, what's to stop them from having a go? I'm not really sure, but it's something we should be thinking about. This kind of thing has been a staple of science fiction for a while, but for the first time I can envision it actually happening.
Cat, say goodbye to the bag.
PS
Footnotes
[1] The references, in case those links go bad:
Regalado, A. Engineering the Perfect Baby. MIT Technology Review (5 March 2015).
Lanphier E et al. Don't edit the human germ line. Nature 519: 410–411 (26 March 2015). doi:10.1038/519410a
[2] Where I very nearly ended up for my postdoc, but let's not get into that.
[3] DiCarlo JE et al. RNA-guided gene drives can efficiently bias inheritance in wild yeast. bioRxiv: 013896 (16 January 2015). doi: 10.1101/013896
[4] Bikard D et al. CRISPR Interference Can Prevent Natural Transformation and Virulence Acquisition during In Vivo Bacterial Infection. Cell Host & Microbe 12:177-186 (16 August 2012). doi:10.1016/j.chom.2012.06.003
[5] Citorik RJ et al. Sequence-specific antimicrobials using efficiently delivered RNA-guided nucleases. Nat Biotechnol 32:1141-1145 (21 September 2014). doi:10.1038/nbt.3011
Bikard et al. Exploiting CRISPR-Cas nucleases to produce sequence-specific antimicrobials. Nat Biotechnol 32:1146-1150 (5 October 2014). doi:10.1038/nbt.3043
[6] Niu et al. Generation of Gene-Modified Cynomolgus Monkey via Cas9/RNA-Mediated Gene Targeting in One-Cell Embryos. Cell 156:836-843 (13 February 2014). doi:10.1016/j.cell.2014.01.027
[7] Windbichler et al. A synthetic homing endonuclease-based gene drive system in the human malaria mosquito. Nature 473:212-215 (12 May 2011). doi:10.1038/nature09937
[8] Though I can of course imagine it creating nightmarish class divides. As William Gibson said, "The future is already here; it's just not very evenly distributed."
