Selfish Genes

 Original Image: Mariuswalter CC BY-SA 4.0

I've just finished reading "A Crack In Creation" by Jennifer Doudna and Samuel Sternberg.  Doudna is the co-creator of CRISPR, the gene editing tool that genuinely is a Cracking Creation.  (I don't think the pun was intended since the word Cracking is not used much outside of and Wallace and Grommit appreciation societies, however, it would have been incredibly apt if it were!)

If you are interested in How-Things-Work then this is a book for you.  I knew next to nothing about cell biology before starting it and now I feel like a pro.  The authors' writing style is incredibly clear, but what really saves you from the sense of drowning in acronym soup are the excellent illustrations throughout the book.  So, if you have heard the terms: DNA, RNA, base pair, ribosome, amino acid, protein, phage, virus, prokaryote, ... and so on, but aren't really sure what these things are or how they relate to each other then this book is much better than an introductory text book.

However, A Crack In Creation is much more than an introduction to cell biology.  It describes in detail the gene editing tool CRISPR that is likely to change the world forever.  The first half of the book is dedicated to explaining how it was discovered (for it was 95% discovered and only 5% engineered - bacteria have used CRISPR to fight off viruses for billions of years), how it works, and how it has been modified for our purposes.  The second half covers the potential applications together with the risks and rewards.

It is clear that Doudna feels a great deal of personal responsibility for CRISPR and therefore a duty to ensure that her creation will only be used for good and not for ill.  The difficulty with this is that everyone has a different idea of where the boundary between these two lies.  So a significant portion of Part II is a call to arms to scientists, politicians, and other thought leaders, to build a worldwide consensus on what applications of the tool should and should not be permitted.  On one level the place we're at is similar to where we were after nuclear fission was discovered.  However, unlike with the A-bomb there are no tell tale signs when someone is not following the rules (like centrifuge plants, imports of large quantities of Uranium, or tests that set off earthquake detectors).  And bio-hacking kits that gives you the power to precisely edit a genome start at under $100.  So, if we failed to contain the nuclear weapon genie, then the chances of doing better with the bio-hacking genie don't look great.  Not without a global surveillance/police state, that is.

There are many worrying potential applications of CRISPR technology, and although the book tries to mention them all it focuses (I think) disproportionally on human germline editing.  Certainly this is something we should not rush into without careful consideration, but in such a tightly regulated field that is not very likely.  The authors point out that germline editing would be consciously directing our future as a species in a way we have never done in the past.  Maybe.  But medicine has been unconsciously shaping the gene pool since its inception.  Every time a human life is saved by modern medicine, and that human goes on to reproduce, medicine has shaped the gene pool.  And if that life was originally at threat because of a heritable problem which the child inherits, then medicine has caused the problem to be more prevalent in the gene pool.  In this way medicine makes itself more indispensable each generation.  So, if germline editing were used to fix inherited diseases, I would argue that it would simply be undoing some of the damage it had previously caused.  The worst outcome would be if gene editing were only allowed on somatic (i.e. non-germline cells).  This would effectively remove all natural selection against harmful conditions without replacing this (natural, but cruel) selection mechanism with another (unnatural but humane) one.

For me, there are two risks associated with applications of CRISPR that are far more worrying than human germline editing.  The first is the risk that changes made for farming have an unexpected greater ecological impact.  The authors give as an example application a modification made in Salmon.  Researchers modified one gene responsible for creating a growth hormone and found that the fish took much less time to grow to its adult size.  If farmed, this gene-edited Salmon would be responsible for 1/25th the level of CO2 emissions.  However, the book does not mention whether or not any research has been done on what effect escapee fish might have on the wider ecosystem.  It may be that figuring this out would actually be much harder than figuring out how to make the original change!

The second risk is the one that really animates me.  In general genes proliferate through the mechanism of natural selection only if they confer some advantage.  However, with CRISPR it is now possible to build what is known as a selfish gene.  The idea is that there is a location within a chromosome which contains

1. The Cas9 gene which builds the "programmable" DNA cutting protein
2. A spacer sequence which builds the "guide RNA" that "programs" Cas9
3. A payload gene - this is the gene we wish to spread through the species despite the fact it confers no advantage
4. Two bookend sequences each of which match one half of the spacer

If the above DNA sequence is present on a chromosome inherited from one parent, it will magically copy itself onto the corresponding chromosome inherited from the other parent.  The idea is that the Cas9 protein and the guide RNA finds the sequence on the peer chromosome matching the spacer DNA and cuts it there.  At this point the cell repairs the cut chromosome using a mechanism called Homology Directed Repair, whereby it finds some nearby DNA that matches the loose ends and uses that to fill in the gap.  The bookend sequences on the original chromosome match the loose ends and so it uses that to repair the peer chromosome.  The result is an identical copy if the DNA described in the list above is copied onto the peer chromosome.  This mechanism is known as Gene Drive and has been proven in the lab.  The picture below illustrates this mechanism, and the picture above illustrates how this causes the payload gene to spread through the entire population.

Picture by Thomas Julou CC BY-SA 4.0

What this means is that anyone with $100 and a good understanding of CRISPR can modify a gene, make it selfish, modify a single individual - be it a fly, a parasite, or whatever - and release it into the wild.  And nothing else is required to ensure that more or less the whole population of that species adopts the gene.  This is enormous power.  While we worry about 3D printing of guns, science has made something more like nuclear weapons just about as freely available.

Gene Drives could wipe out malaria by conferring malaria resistance on all mosquitoes. But they could also modify the Plasmodium parasite that carries malaria to make the it more infectious, or to make it affect only people with certain racial profiles.  Whether or not that particular nefarious use is possible I do not know, but there is no doubt that Gene Drive in particular and CRISPR in general could be weaponized in some way or another.  And there is no doubt that governments and other groups worldwide already know this.

A global surveillance police state is starting to look like a good idea.