Precision Gene Editing

mpthompson writes "NewScientist.com is reporting that scientists at Sangamo Biosciences have developed a method of editing DNA mutations with unprecedented precision without weaving in potentially harmful foreign genetic material. Different combinations of amino acids are designed to latch on and cut the DNA at exactly the place where the mutated gene lies. This triggers the body's natural repair process which corrects the gene where the DNA was cut. The technique will be used to target diseases caused by single-gene mutations such as combined immune deficiency (X-SCID) - or bubble boy disease - and sickle cell anaemia."

Re:I don't care what they say..

[Prophetic_Truth]

I would say that's a bit paraniod, and possibly based on some educational time spent watching the sci-fi channel.

PFFFT!!! GIVE ME A BREAK! You want the truth!?

I predict the following Prophecy:

Years from now we will have enhanced ourselves to the point where our skulls grow large and our eyes turn black. We no longer need to speak with our tounges, rather our minds. We no longer use sex as a means to reproduce. We have geneticly engineered ourselves to do things even I, Prophetic Truth, can not invision. BUT THERE'S A FLAW! A horrible flaw which can not be fixed by our future selves. A flaw of such consiquence that it will wipe out our species.

The solution?

Time travel, anal probes, and sperm collection.

The Prophecy has already been fulfilled

In the case of specific genetic diseases

[MichaelPenne]

like the 'bubble boy' defect mentioned in the article, we often know the specific bit of code that causes the problem.

"IL-7 signalling pathway

Most cases of SCID are derived from mutations in the c chain in the receptors for interleukins IL-2, IL-4, IL-7, IL-9 and IL-15. These interleukins and their receptors form part of the IL-7 signalling pathway.

The IL-2 receptor (IL-2R) gene is located on the X chromosome and mutation of this gene causes X-linked SCID.

Janus kinase-3 (JAK3) is an enzyme that mediates transduction of the c signal. Mutation of its gene also causes SCID."

http://en.wikipedia.org/wiki/Severe_combined_immun odeficiency [wikipedia.org]

Not specific enough for safety (yet)

[G4from128k]

TFA noted that the zinc fingers cue in on two sets of 6 base pairs to find the site that needs correction. Assuming randomness in the base-pair sequences, this 12 base-pair key will bind with approximately 1 out of every 16.8 million (actually 1 out of every 8.4 million due to complementarity of the base pairs). Given that the human genome has about 3.2 billion base pairs, this means that the modifier will match in 381 positions more or less.

Thus, this method will fix the error in one place and introduce an error in 380 other locations. The key needs more than 16 base pairs to be statistically assured of homing in on a unique mutation (depending on the statistics of DNA, it may need more or less).

with a PhD in Genetic engineering

[cinnamon colbert]

I have not read the article, but repair processes can be "error prone". That is, the mechanisms cells use to repair DNA often involve high error rates.

The human genome is 3e9 BP long (roughly..not counting indels, the unsequenced centromeres, etc etc)

So the chemical process of identifying the one single mutated basepair has to have a chemical specificity of >>1e9, because there are >>1e6 cells that are exsposed. That is, lets say you feed the reagent to a person. Millions of cells, each with 1e9 bp, are expsosed. Say the process has an error rate of 1e10 - many, many cells will have incorrect repairs done
This is just like error rates in, say, reading data from a harddrive: the larger the file, the lower the error rte has to be

What /.ers may not appreciate is that typically, it is VERY, repeat VERY hard to get chemcial reaction specificity of anywhere close to 1e9 for reactions invovling DNA.

I will rtfa,

Re:Clarification

[saytan]

While I haven't read the article, I have heard a presentation on this from one of the researchers involved.

The old technology involves the use of a retrovirus containing the correct copy of the X chromosome gene involved. This copy inserts itself (nearly randomly) into the DNA. The problem with this was that you couldn't control the point of insertion, causing a whole new set of diseases.

The new technology involves repairing the endogenous gene sequence rather than inserting a good copy at another locus. By doing this, you get around the problems caused by random retroviral insertion. The key breakthrough in the new technology was the ability to make proteins that can cleave highly specific sequences. Researchers at Sangamo can custom make a protein to bind at only one place in a genome of 3 billion base pairs.

Both of these techniques work by taking out some stem cells from your body, transforming them, and placing them back in with your normal stem cells. This means that the DNA sequence of your germ cells, the cells that pass down your DNA to your children, is not changed.

Re:Clarification

[Fadeproof69]

In order to answer your question, i'm going to have to give a little background...

contrary to popular belief, 99.99% of the body's cells don't keep dividing. The somatic cells of the body are replenished by stem cells and progenitor cells which act as the main copy from which all the "backup" cells are made. These cells specialize into skin cells, blood cells, and possibly nerve cells. The only way to have a permanent effect with this treatment would be to fix the mutation in the stem cells/progenitor cells, so that future specialized cells will all have the fix incorporated.

To make this change heritable, you need to fix the mutation in the sperm or egg which is eventually used to create an embryo. Otherwise, the mutation will be passed on.

From what the article says, there's only an 18% transformation efficiency so of all of the cells treated (this would never be the entire human body, just the cells collected), only 18% will be fixed.

We are a long way off from doing the 100% effective gene therapy you see on Star Trek.

Re:Homologous Recombination

[Anonymous Coward]

Yes, it does have to do with homologous recombination. Creating a double strand break in the chromosomal DNA induces various DNA repair pathways including homologous recombination. The break can be healed by "copying" information off of the non-broken chromosome as you suggest.

If, however, you introduce a piece of "foriegn DNA" into the system at the same time that you make the chromosomal break, and that foreign DNA has homology to the DNA sequence flanking the chromosomal break, then the forien DNA can by recombined into the chromosome at the break point. Thus, one can insert any gene into a specific place on any chromosome (in theory).

Re:Precise Gene Editing = Hex Editor

[harvardian]

Great, now the gene splicers have the equivalent of a hex editor, but still have no clue what they are editing. It's like hacking binary code out of one program and inserting into another program and somehow getting it to work.

This isn't entirely true. We can figure out where a gene starts in DNA, and we know how to read the DNA into a protein. We know that from the start point, DNA is broken up into 3's such that each set of three DNA bases code for one amino acid. To use the case of sickle cell anemia, the DNA sequence GAG is replaced by GTG. This causes a glutamine amino acid to be incorporated into the Hemoglobin beta chain instead of a valine (this can be predicted since we know the entire triplicate-to-amino acid dictionary). Partly because glutamine is a charged amino acid and valine isn't, this causes Hemoglobin with this mutated beta chain to clump together when deoxygenated -- hence the sickle cell phenotype.

So in this case it isn't true that we're hacking binary code. We're hacking a DNA code that we know enough about to fix simple point mutations like the one found in sickle cell anemia. As for other, more complicated, diseases, we are indeed still poking in the dark. But that doesn't mean progress isn't being made...

Re:Not specific enough for safety (yet)

[Anonymous Coward]

Upon careful reading of the paper, it seems from Fig. 1a, the Introduction, and the Materials and Methods, that two zinc fingers, each recognizing 12 bp are required for editing to work. The boolean sum of the recognition sequences of the two zinc fingers -- 24 bp in total -- is unique in the human genome.

Re:having RTFA,

[lockholm]

Actually, the simple process of removing blood from the body is not mutagenic - for example, think of blood transfusions, where blood is not only removed from the body, but frozen and stored.

Also, the large percentage of blood consisting of the red blood cells and platelets don't actually have any DNA in them to be mutated - these cells don't have nuclei.

Finally, in bone marrow transplants, one method of collecting the marrow cells to transplant is to hook the donor up to a machine through which their blood flows. In the machine, the stem cells (the cells that divide to produce all the elements of blood, including red blood cells and immune cells) are separated out, and these are the cells that are then transferred as the marrow transplant. You can find out more about this process here [health-alliance.com]. The objective with this treatment is to cure the cancer - so if simply removing the cells from the body causes cancer, it would be a very counter-productive treatment.

Precise Gene Editing = Patch Files

[cookie_cutter]

Your right in that this doesn't give us the ability to do really novel gene manipulation.

But it does give us the ability to create the equivalent of patch files for bad/defective genes when a good/functional version of the gene is available.

There are many genetic diseases where the mistake in the DNA is well characterized, and it is very clear exactly what difference between the normal version of the gene and the defective version causes the disease, even if we don't have a full understanding of what the hell gene does; we just know to a high degree of certainty that a particular error causes a particular phenotype.

This new technology, if it lives up to the hype it's given here, could mean we can fix these kinds of diseases.

Re:Not specific enough for safety (yet)

[Anonymous Coward]

The point is that it induces a break which will be repaired by homologous recombination, which is error free. This is in contrast to non homologous end joining which is also a common repair pathway, but introduces errors.

Homologous recombination needs a template which is usually a sister chromosome, however, in this case the template should be engineered with the non-mutated gene. Breaks other places in the genome will not be repaired by the engineered template since it is no homologous, it will be repaired using the sister chromosome as template. Therefor it does not introduce the gene in unexpected places. Easy.

Myotonic Muscular Dystrophy cure

[RhettLivingston]

If they were to concentrate this work on Myotonic Muscular Dystrophy, they could likely achieve a success very quickly. It is caused by an unstable CTG sequence of DNA that expands in length when replicated. The progression of the disease is characterized by the number of expansions. Since it is an unstable sequence and of little use, simply cutting it out of all DNA should "cure" the disease. I put the "cure" in quotes because reversing the damage is likely not possible, but it could at least eliminate it from future generations and stop the progression.

the article

[bikerguy99]

Highly efficient endogenous human gene correction using designed zinc-finger nucleases

FYODOR D. URNOV1, JEFFREY C. MILLER1, YA-LI LEE1, CHRISTIAN M. BEAUSEJOUR1, JEREMY M. ROCK1, SHELDON AUGUSTUS1, ANDREW C. JAMIESON1, MATTHEW H. PORTEUS2, PHILIP D. GREGORY1 & MICHAEL C. HOLMES1

1 Sangamo BioSciences, Inc. Pt. Richmond Tech Center 501, Canal Blvd, Suite A100 Richmond, California 94804, USA
2 Department of Pediatrics and Department of Biochemistry, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, Texas 75390, USA

Correspondence should be addressed to M.C.H. ([e-mail]) or M.H.P. ([e-mail]); requests for materials should be addressed to M.C.H.

Permanent modification of the human genome in vivo is impractical owing to the low frequency of homologous recombination in human cells, a fact that hampers biomedical research and progress towards safe and effective gene therapy. Here we report a general solution using two fundamental biological processes: DNA recognition by C2H2 zinc-finger proteins and homology-directed repair of DNA double-strand breaks. Zinc-finger proteins engineered to recognize a unique chromosomal site can be fused to a nuclease domain, and a double-strand break induced by the resulting zinc-finger nuclease can create specific sequence alterations by stimulating homologous recombination between the chromosome and an extrachromosomal DNA donor. We show that zinc-finger nucleases designed against an X-linked severe combined immune deficiency (SCID) mutation in the IL2Rbold italic gamma gene yielded more than 18% gene-modified human cells without selection. Remarkably, about 7% of the cells acquired the desired genetic modification on both X chromosomes, with cell genotype accurately reflected at the messenger RNA and protein levels. We observe comparably high frequencies in human T cells, raising the possibility of strategies based on zinc-finger nucleases for the treatment of disease.

Most human monogenic disorders remain difficult to treat because therapeutic transgenes do not undergo homologous recombination (HR) into the mutated locus1, 2, and gene addition by virus-driven random integration remains a challenge owing to transgene silencing, improper activity or misintegration3, 4. Furthermore, targeted alteration of DNA sequence in vivo--in principle, a powerful basic research technique for studying genome function--in mammals requires sophisticated targeting vectors and drug-based selection1, 2, which limits the use of this approach5-7.

The C2H2 zinc-finger, originally discovered in Xenopus8, is the most common DNA binding motif in all metazoa9. Each finger recognizes 3-4 base pairs of DNA via a single alpha-helix10, 11, and several fingers can be linked in tandem to recognize a broad spectrum of DNA sequences with high specificity12-14. Engineered zinc-finger protein (ZFP)-based DNA binding domains with novel specificities have been extensively applied in vivo to target various effector domains12, 15. Work from the Chandrasegaran laboratory has shown that a ZFP can be coupled to the nonspecific DNA cleavage domain of the Type IIS restriction enzyme, FokI, to produce a zinc-finger nuclease (ZFN)16, which then cuts the DNA sequence determined by the ZFP16, 17. An important specificity mechanism derives from the requirement that two ZFNs bind the same locus, in a precise orientation and spacing relative to each other, to create a double-strand break (DSB; Fig. 1a)17. One mechanism by which eukaryotic cells heal DSBs is homology-directed repair (Fig. 1b)18-20, which transfers information missing at the break from a homologous DNA molecule (Fig. 1b). Work from the Jasin laboratory21, followed by that of others22, 23, demonstrated that the endonuclease I-SceI can potentiate HR into loci previously engineered to contain its own recognition site, and the Carroll24, 25 and Baltimore26 laboratories have shown that a ZFN-invoked DSB increases the rate of HR in model systems.

Figure