Gene Editing Tool Hailed as a Breakthrough, and It Really Is One | Hacker News Reader

Gene Editing Tool Hailed as a Breakthrough, and It Really Is One

npr.org

84 pointsevo_945 comments

Something to consider: it is now technically possible and within the grasp of hundreds of clinics worldwide to run gene therapy. The only reason they aren't yet doing this is that it takes some increment of time to get things lined up; funding, customers, propagation of knowledge, etc. But all the pieces are in place.

Myostatin or follistatin gene therapy is about as proven as these things get. It exists in heavily muscled natural mutants, including humans (look up brute whippets). Antibody myostatin blockade has just completed phase 2 clinical trials, adding muscle mass to 75 year olds. Chinese labs are turning out myostatin-free dogs. The BioViva CEO organized her own follistatin gene therapy earlier this year in an overseas lab. Five years from now it won't take starting a company level of networking to fly somewhere and do this.

But myostatin for muscle mass is just the best risk/benefit/knowledge picture out of hundreds of alterations that could in theory now be run up in humans, many of which are high risk in the sense that they've only been tried a few times in mice and not followed for a long time. For example adding additional lysosomal receptors to mice restores liver function in old mice to be the same as in young mice. Increased GDF-11 sends stem cells back to work and improves health in old mice.

And so on and so forth. The eager early adopter with a hundred thousand dollars to spend ten years from now could spend the time between now and then mining the research literature and organizing collaborators, making connections with labs overseas and setting up for multiple gene therapies down the line, many of which are associated with interventions shown to delay or turn back specific measures of aging in mice.

This isn't SENS, it isn't repair of underlying damage, but it is a step ahead of what is currently possible.

The greatest barrier here is the FDA. They don't let you perform medical interventions (even breakthroughs like this) without going through a rigorous process. Companies like Editas ($43M Series A) are going through those motions right now.

The inevitable consequence will be either the abolition of the FDA; or a chasm between advanced medical therapy readily available in some places outside of the US and sclerotic, bureaucratic "healthcare" in the US.

Long-term, where I live will depend on the above outcome.

CRISPR is far from being the first technology to be held up in the United States while being available elsewhere. There is very little chance the FDA will materially budge on their standards, and even less chance the FDA will be abolished. Most people have a high degree of trust in the FDA (they have a 65% favorability rating according to Pew).

Note that the FDA does allow some compassionate use of experimental medications and technologies in extreme cases. I'm not familiar with Editas's regulatory strategy but it wouldn't be crazy for them to pursue that avenue for, say, a child cancer patient as an early step. But even they will be cautious, as any misstep (i.e., harm to a patient) will have major negative consequences for the technology as a whole.

I'm not denying running these experiments is becoming much more feasible for small institutions but you have a pretty naive understanding on how to turn these kinds of experiments into successful therapies.

There is often a huge disconnect between what we can reliably and reproducibly show in a lab, and what can be turned into therapy. And that's not just as simple as mining the literature and running more trials.

One of my points is that adoption is going to precede the current consensus definition of "successful therapy" by a decade, in much the same way as happened for stem cell therapies following the turn of the century.

Although not usually mentioned, it is important to note that CRISPR is a natural process that bacteria use as part of their immune system. For what is worth, the mechanistic process and a specific functional result - defend against viruses - have stood the proof of time.

"bacteria use Cas enzymes to grab fragments of viral DNA. They then insert the virus fragments into their own CRISPR sequences. Later, when another virus comes along, the bacteria can use the CRISPR sequence as a cheat sheet to recognize the invader." [1]

[1] https://www.quantamagazine.org/20150206-crispr-dna-editor-ba...

It's funny that this should appear on HN ... Sunday I just completed a bit of software that looks for candidate CRISPR sites in a particular mouse genome for my daughter. In our case, the real trick is to find repeats in the section of the chromosome she wants to manipulate that don't also exist in the rest of the chromosome - or any of the other chromosomes.

Like any "big data" project, the trick is representing your data in a way that makes searching it trivial as well as parallelizable.

That sounds like a very interesting and I'm assuming, a sad story. Do you mind if I ask if you are sharing it publicly?

It took me a few rereads of my original post to realize what might be sad about it ... then I got it.

It's actually a happy story (for me). My daughter is a PhD candidate at the Johns Hopkins school of medicine and is working towards becoming a researcher (molecular genetics). I'm not sure I should comment on which lab she's working in or what they're trying to accomplish.

It might not be fun for the mouse but I'm thankful for projects like this because we get to work on something together. It's almost like going back to when she was a girl doing a science project in elementary school.

Radiolab had a great episode about CRISPR earlier this year: http://www.radiolab.org/story/antibodies-part-1-crispr/

A quote that stuck with me from a molecular biology class taken a year ago, "It's not who wins a Nobel Prize off of CRISPR work, it's how many there will end up being".

From what I understand (from a PhD in biotech), the price for utilizing CRISPR/Cas9 is extraordinary. So much so that only multi-million dollar labs can afford to use it in any meaningful quantity. The price "per reaction" is in the thousands.

This quote, from a Nature news article, suggests otherwise:

> Biologists have long been able to edit genomes with molecular tools. About ten years ago, they became excited by enzymes called zinc finger nucleases that promised to do this accurately and efficiently. But zinc fingers, which cost US$5,000 or more to order, were not widely adopted because they are difficult to engineer and expensive, says James Haber, a molecular biologist at Brandeis University in Waltham, Massachusetts. CRISPR works differently: it relies on an enzyme called Cas9 that uses a guide RNA molecule to home in on its target DNA, then edits the DNA to disrupt genes or insert desired sequences. Researchers often need to order only the RNA fragment; the other components can be bought off the shelf. Total cost: as little as $30. “That effectively democratized the technology so that everyone is using it,” says Haber. “It’s a huge revolution.” [1]

[1] http://www.nature.com/news/crispr-the-disruptor-1.17673

More or less... but it might be a country thing.

I am from Brazil, and my sister, and a lab partner of hers are the first persons in Brazil to develop and use a local version of CRISPR, the bugdet of their lab is not that big, and finding reagents was a bit difficult, but not too much, but I think maybe her lab didn't have to pay all the patent licenses, I am not sure.

My sister is currently at MIT learning direct from the source (it was invented at MIT) how to do "real" CRISPR (in the sense that the Brazillian version was made by her and another lab mate reading papers and replicating them, it worked, but in a inconsistant manner, sometimes it worked perfectly, sometimes it failed for no obvious reason, the Brazillian government then decided to pay to send my sister to MIT so she can learn how to do it properly and then come back and teach the others here).

It's worth drawing the distinction between academic and commercial users, as academics don't have to pay the licensing fees. You can get the plasmid for $65 [1], which is basically all you need (other than some cheap reagents and some standard lab equipment).

[1] https://www.addgene.org/42230/

Any insight into the source of those costs?

licensing the tech, plus no competition = charge what the market will bear

http://labiotech.eu/crispr-patent-war-end-discovery-new-edit...

So patents. That was going to be my guess.

So in 20 years it'll be pretty cheap. Meanwhile, China and other parts of the world that happily ignore patents will use the technology without restriction.

You're not even kidding, they've already been experimenting with it on human embryos and creating modified dog breeds. Have a look:

http://www.nature.com/news/chinese-scientists-genetically-mo...

http://www.telegraph.co.uk/technology/news/11943103/Chinese-...

There's a common sci-fi theme about China being way ahead in the future. I've never heard a plausible explanation in book but ignoring patents is a really interesting angle.

That's how it always works. Early American industrialists weren't exactly diligent at paying royalties to British and continental European patent holders... and now it's China's turn to prosper by doing the same thing to us. More power to them.

Patents are a racket that governments run against their own people, exclusively for the benefit of a very few inventors and a whole lot of lawyers.

"However, like every biotech revolution, there then follows a patent war."

But hey, continuing to do the same thing and expecting different results sounds like the righteous path.

I've often wondered how CRISPR is actually useful, because at its core it's just a nuclease. It allows to you precisely cut DNA where you want to cut it. But for that to be useful you still have to put the DNA back together afterwards. I found this article useful for understanding the different ways that happens in practice:

https://www.addgene.org/CRISPR/guide/

Turns out you get some small mutations in the most common mechanism (non-homologous end joining or NHEJ). This could be a problem for use in therapies, but definitely not a huge issue for most research-based applications where you can control which mutants you allow to proceed to the next experiment.

This is one reason why Cpf1 could be interesting, since it leaves sticky ends when it cuts: http://www.cell.com/cell/abstract/S0092-8674%2815%2901200-3

I am excited and worried about what CRISPR will bring about.

The application of CRISPR/Cas to gene editing, while a breakthrough, is more about technology than science. It just makes things easier- similar to the way compilers make it easier to program computers than assembly language.

I also find it interesting that we consider this tool a breakthrough when nearly all our tools for gene editing are just copped off bacteria that have been using them for billions of years.

Yeah! Like, antibiotics, man. It was basically mold that invented that! Totally not a medical breakthrough at all... no idea why so many people were impressed by it.

Fusion was invented by the sun (at least in this solar system -- there are theories that it was independently invented elsewhere too)!

Pfft, the sun just cribbed that idea from the big bang. Typical, unoriginal sun...

The difference is that antibiotics weren't a tool to make more discoveries. They were an immediately applicable medical breakthrough (which also happens to be a useful tool, for example as a genetic marker). CRISPR/Cas9 is just a tool, not an immediately applicable therapy.

In the debate over the continued utility of human labor in the face of automation, I suspect gene editing (and/or embryo selection) is the most ignored countervailing factor.

So all these truckdrivers that are being replaced with self driving rigs are going to get into biotech now?

Nope. Self driving cars are gonna be a huge problem. I don't think many people realize that nearly 9% of the US Employed workforce has no skill other then driving a vehicle.

We'll still need people to evaluate/compare/vote whether any automated process (driving, pricing, placing, whatever) is good or not. This is where the human cannot be replaced.

Possibly big portion of the employable human population would be simply evaluators, high-level instructors, QA, real-life testing conductors, etc. Silly, but no machines can replace that yet.

What exactly is to stop them from learning a new trade?

Nothing except time and money, but that's a lot of people. Even occurring over a decade or more it's going to cause the wages of a lot of those trades to plummet.

The specific instance here was drivers, who in this hypothetical, will be outcompeted by self driving cars, and and naturally find themselves with more time on their hands whether they want it or not.

As to money, there are a number of trades you can learn without spending money. And there are also jobs where you get on-the-job training.

I'm not saying it won't be a problem for these people, but eventually, most if not all will adjust to the new situation and find themselves in a new job.

That's less of a problem than what happens to the spaces those people move into. Fewer jobs for more people will be the mid-range social problem for automation.

In the U.S.? - It was predicted that as jobs became automated, the population would "share the wealth" and we'd become a leisure society. What's really happened is that the economic gains have gone to a relatively small ultra-wealthy class. If things had worked as predicted, we'd have had to reduce our work week already due to a lack of work.

> we'd have had to reduce our work week already due to a lack of work.

This makes no sense.

It assumes some subset of required work. There is no required work. The work that is done is determined by demand for that work and the demand is determined by availability of resources (time, money, etc.).

If we have more resources, such as time, we won't just sit on our asses doing nothing. That makes no sense. We'll develop new desires, which will lead to new types of work. There will always be work and there will always be enough work to go around. It may be very different from the type of work we do now, but it will still be work and it will still be around.

I'll agree that the original premise assumes that all jobs would be magically automated at the same rate.

what kind of paid work? http://strikemag.org/bullshit-jobs/

> Ever had the feeling that your job might be made up? That the world would keep on turning if you weren’t doing that thing you do 9-5?

Who said that every job needs to be super-duper-important and that you need to change the world with every breath you take? That's some new-age bullshit right there.

If you're getting paid, it means someone is deriving value from your work. There are no bullshit jobs.

Eventually, you're going to run out of trades to learn, and/or jobs in those trades.

Reply on news.ycombinator.com