New era of fast genetic engineering | Hacker News Reader

New era of fast genetic engineering

newscientist.com

87 pointsprateekj18 comments

>This is not a problem that can be solved by computers. Ultimately, there is only one way to be sure what a particular bit of DNA does – you have to alter it in real, living cells to see what happens.

Ah, the old "comment out this line" style of understanding code.

The debugger hasn't been born yet.

For those who are in this field, I have a question, and will be grateful for an answer. How far-reaching is this technique can be for Cancer treatments? I understand Cancer is not a simple disease that can be delineated.

As a researcher in the field, it's a big question. There are lots of answers. Most of the time, the answer will be, 'maybe, and maybe more if we can get it to the right place at the right time - but it's often hard to know where/when that is'.

More specifically, and more concretely, there are certain kinds of cancers (and diseases/disorders) that this will be very effective against. And some advanced iteration of this technology will be curative - entirely. This is of an entirely different kind of medicine than bathing trillions of cells in a potent chemical in the hopes that one of those molecules might slow another molecule - somewhere. This kind of genetic engineering would enable a biological feature to be entirely restored, changed, or removed.

Those cancers this will be effective against are those where a particular, knowable, and targetable gene has a known and particular mutation. And this is actually a large class of cancers. Cancers often initiate as mutations that destroy the elaborate series of 'checkpoints' your body has instituted to prevent runaway growth. If one of those proteins that guards against runaway growth gets damaged, then there are far more ways to get cancer. If you can repair that checkpoint (via (genetic) therapies), you can halt, slow, or even destroy what caused that (particular) cancer. The analogy of the cells in your body as a society is not a bad one. And cancer, as a 'criminal' is not particularly dangerous to the whole. It only becomes dangerous when the police, the courts, and the military are also crippled. And we have learned a lot about how those regulatory objects function in the past two decades. However being able to restore them or affect them in a living being has been orders of magnitude more difficult than simply watching to see what they do. It would be fair to say we know understand most of the key elements, and many of the details of their roles of what's happening in the society of your body. However making direct changes to that society in a living being is still tricky, and our tools blunt. What you see here is the sharpening of these new tools - tools that offer the opportunity to restore (or improve) those regulatory objects to a broken system.

To say it another way - there is no way I could imagine a cancer to be 'cured' by small molecules (alone). Small molecules work great against 'the different' - but by definition, cancer is not different from its host. I think this is why the drugs of the 20th century were so ineffective against cancer while simultaneously so effective against parasites, bacteria and even viruses for the most part. The only way I could (from first principles) expect to prevent or alter a process like cancer is using tools like the ones talked about here.

Great question. Excellent answer.

The analogies you used made this all much easy to understand.

Thanks.

How can knocking out a certain gene (for all cells in your body) kill a cancer cell? For instance, for many cancer cells with a defect in 'P53' (a gene which can induce cell death), knocking out this defective gene would not help - would it?

Or would you need a two step process, like a molecule which connects to the marker and then use a second step to knock-out a gene essential for cell survival? Or can this technique be used to repair a certain gene?

The most interesting part of the above tool is not its ability to knock out a specific gene, but its ability to knock out a specific gene. The key word is 'specificity'. Knocking out is just the easiest proof of principle thing you can do once you find a sequence. Finding it with specificity in a live organism is the magnificent part.

There is an entire field dedicated to editing genetic code - and it's getting really good. The issue is targeting it to the right spot in a live organism (and conversely, not targeting your editing machinery to the wrong spot). If you can get to it, there's lots you can do besides knocking out a gene. If we call the p53 gene the judge or head executioner in our scheme above, it's finding him that's the problem. If you can pinpoint the corrupt judge, you can easily (relative to finding him) install a replacement. What you don't want to do is accidentally replace your police force or your teachers with judges/executioners (even if they're good ones) - and that ability to target is the greatest issue with engineering the genetics of live cells currently.

(As a side note, this is EXACTLY what the HPV virus does - it runs around assassinating that judge with relentless precision - and thus is a prime cause for allowing cancer to run rampant. The vaccine against HPV prevents these assassinations.)

Further, and only slightly more in the future, once you have access to the genetics, you can start to run logic programs inside the cells. You can start to release the (genetically encoded) tools IFF the cell is $cancerous, where $cancerous is defined as a further set of genetic conditions within the cell. In this way you could surveill the entire society while only targeting the bad guys.

Right! I wasn't aware that the editing process is quite easy once you can target a specific gene. Awesome! :)

Specifically, you can target just the defective version of the gene. In theory, your good copies could be ignored, only removing the bad genes from the cancer tissue.

However, I don't think that this will ever be applied on a whole-body scale because the off-target effects aren't well characterized at this point. That is, of course, unless you have a bad cancer and have no other options. At that point, having off-target effects is the least of your concerns.

The problem with gene editing in this context is that the changes are permanent. So if you miss and edit a good cell, that cell is now damaged, or may only have 1 good copy of a gene, and thus be more susceptible in the future. In some cases, shRNA to silence a gene may prove more effective because it is temporary.

Right now these are only used in the lab for research.

So how hard would it be to fix hereditary illnesses due to known malformed genes?

Initially, these new genome editing techniques are going to be (and already are becoming) extremely useful to more efficiently set up genome-wide screens and do basic science to discover new targets and understand the biology of cancer.

Applying genome editing directly as a therapy is a more speculative proposition for the time being.

Scientist may never discover a single "cure". Cancer cells tend to exhibit a greater variety of traits than the healthy cells in your body. In other words, cancer cells represent a more diverse population of cells in your body than your healthy cells, which should be near identical copies of each other (assuming that the healthy cells are all part of the same tissue of course). That is one reason why cancer is so hard to effectively treat. How do you find a drug that kills all the cancer cells in your body given that there is so much variety among the cancer cells? Even if you discover a drug that kills most of your cancer cells, and you use the drug, you will then be left with cancer cells that are immune to your drug.

As a futurist, I love reading quotes like this:

"What used to take two years or more can now be done in six weeks, says Zhang. "That's a big difference." For those who have spent years trying to make just one or two specific changes to plants or animals, this is revolutionary.

What's a "futurist"?

Someone who loves reading quotes like that.

The parent is obviously a "social scientist whose specialty is to attempt to systematically explore predictions and possibilities about the future and how they can emerge from the present, whether that of human society in particular or of life on earth in general." Obviously. http://en.wikipedia.org/wiki/Futurist

I find it amusing that the underlying technology that made this study possible was patented (end of page 1), but that a royalty free alternative solution was found. In science people are often quick to patent each new breakthrough, which sometimes just prevents other scientist from building on that initial breakthrough. I think it is an important lesson to realize that not every new discovery is worth patenting. Perhaps only when a technology is truly ready to be brought to market in the foreseeable future should a patent be considered. Otherwise, the patent could become a distraction to the importance of the breakthrough itself.

All the technologies described in this article have many competing patent claims by academics at Harvard, MIT, Berkeley, and a few others. Obviously it's hard to tell right now which patents will go through, but I'm pretty sure they won't be royalty free. The main difference is that the procedure is now so easy that most academics will just infringe.

Reply on news.ycombinator.com