CRISPR update

I believe that all we have to do to make ourselves younger is to turn on the genes that were expressed when we were young, and turn off the genes that are expressed when we are old. This will require both knowledge and technique; (1) knowing which genes these are, and (2) having a targeted mechanism for turning specific genes on and off in vivo. At this time, our technique is advancing nicely, outpacing the knowledge, thanks to CRISPR / Cas9.

I wrote nearly two years ago that CRISPR was a third-generation technology for editing the genome, which could also be adapted to “edit the epigenome” by turning genes off.  Turning genes on was, and still is more difficult.  The work-around is to add extra copies of the gene, which can be a higher-risk operation, because the body has no evolved mechanisms for deciding when to turn the extra copy on and off.

The older technology of AAV (Adeno-Associated Virus) can be deployed within the living body, transfecting large numbers of cells and inserting a payload gene (of limited size); but there is no control over where in the genome the gene is inserted, and so there is no assurance that it is turned on or off at appropriate times and places.  The newer technology of CRISPR is precisely targeted, can remove or insert a gene, can turn a gene off (but not on).  But so far it is only possible in cell cultures in the lab, and not for large numbers of cells within a living organism.

There are thousands of clinical trials worldwide for AAV gene therapies, and last week the first clinical trial was announced for CRISPR as a cancer therapy.  The protocol is to extract the patient’s own T cells from a blood sample, then modify the cells in lab culture using CRISPR.

The researchers will remove T cells from 18 patients with several types of cancers and perform three CRISPR edits on them. One edit will insert a gene for a protein engineered to detect cancer cells and instruct the T cells to target them, and a second edit removes a natural T-cell protein that could interfere with this process. The third is defensive: it will remove the gene for a protein that identifies the T cells as immune cells and prevent the cancer cells from disabling them. The researchers will then infuse the edited cells back into the patient.

This is a modest first effort in many ways–not just that it is limited to 18 patients and nominally seeks only safety data.  There is great potential for sensitizing the T cells to the patient’s particular cancer.  It would also be logical to combine CRISPR with stem cell therapy.  A patient’s bone marrow stem cells could be harvested, modified with CRISPR, and re-injected, whereupon they would create an ongoing supply of sensitized T cells.  Neither of these ideas will be attempted in this first trial.

The Nature article goes on to recall the tragedy of the first gene therapy trial to kill an 18-year-old patient in 1999, and how gene therapy research lost a decade dealing with safety issues after that.

Hydrodynamic Gene Therapy

This is a kind of brute force method for delivering a genetic payload.  A large volume of dissolved DNA is injected directly into a vein, rapidly enough to raise blood pressure system-wide for a few seconds.  The pressure pushes some of the payload through capillary walls.  This system has been widely adapted for rodent experiments, with a tail vein used as the delivery point.  It has even been tried in humans.  But it is crude and untargeted.  Penetration rates remain low, and collateral damage is unavoidable.


Incorporating CRISPR into Gene Therapy

Of course, what we would really like is the specificity of CRISPR combined with the wide in vitro delivery provided by AAV gene therapy.  The complete machinery for Cas9 to break the DNA strand in a chosen location is too large a payload to fit within the AAV virus.  So marrying CRISPR to AAV has been the subject of some ingenious research just in the last two years.  The first successful experiment was announced this past winter in Nature Biotech.  An MIT-based research team reports that in a single treatment, they are able to make targeted modifications to 6% of white blood cells in a lab mouse.  I’m out of my depth reading about their technique, but from what I understand, there are separate delivery systems for the gene (via AAV virus) and for the targeting (via nano-particles of Cas9 enzyme dissolved in organic fats).  The former makes its way efficiently to the cell nucleus, because that it is what the virus was evolved to do.  The latter must be relied upon to diffuse into the nucleus at random, and the microencapsulation facilitates its transit.

Once inside the nucleus, the Cas9 breaks the chromosome in just the right place, and the virus seizes the opportunity to insert its payload conveniently.  The authors emphasize the importance of eliminating the Cas9 promptly.  If it were part incorporated in the virus, there would be a danger of ongoing, long-term DNA breaks; but the nano-particles containing Cas9 are short-lived.

Here is a recent review of progress in combining CRISPR with viral vectors for gene therapy, written at a technical level.  The Concluding Remarks section speculates on the possibility of combining three separate viruses for delivery of the CRISPR template, the Cas9 enzyme, and the genetic payload.  The obvious problem with such a system is going to be that each of these three viruses has a limited penetration, and it will only be in cells that all three viruses have transfected that the right thing happens (double-stranded break in just the right place, followed by insertion of the payload gene).  In the much larger number of cells that receive one or two of the viruses, there is the probability of damaging side-effects.


A Tangle of Signals

In the fable of the Sorcerer’s Apprentice (and a hundred myths from ancient Africa, Europe and the Orient), the protagonist is attracted to the quick acquisition of power, less interested in the slow acquisition of wisdom.  Exercise of power without wisdom is the classical gateway to tragedy.

And so we see our labs acquiring control over the genome and over gene expresssion proceeding apace, while understanding of the tangle of signaling pathways lags behind.  Many of us are now convinced that aging is controlled by epigenetic signals.  We are beginning to map difference in gene experesssion that occur with age.  But which genes are upstream and which are downstream?  Which are cause and which are effects?  Which are tissue-specific, and which are systemic signal molecules?

It now appears that the technology to modify gene expression will be ours before we know how to use it.

27 thoughts on “CRISPR update

  1. It is CRISPR/Cas9 that will enable us to understand the function of genes and gene expression through the activation and deactivation of code segments. CRISPR/Cas9 will first act as the tool to gain the genetic understanding that will later make it the tool for the cure.

  2. Compile the code. Set a hook at the end of the AAV payload length limit to be picked up by the next one.
    Safety measures will indeed be needed to avoid mentioned side-effects, probably involving offsetting the promoter code by a non-multiple of 3 number (base pairs units). Effectively making everything look as “junk DNA”, which should sound familiar, although, of course, the question then becomes how the AAV will be able to find the previously prepared hook efficiently. Bioinformatics will be needed to choose a sequence that won’t have a sub-string that will look like valid code, before the whole thing is finally assembled.

    • Yes, as you suspect, this article is out of date.

      There are two ways I know of reading the epigenome. One is gene expression profiling. Some of the proteins that are in circulation at any given time can be captured with specific antibodies. The end result of epigenetics is read as the set of proteins that are produced. The second is methylation maps. First, the full genome is sequenced. Then a chemical is added that changes all the methylated C’s to T’s, but leaves the unmethylated C’s alone. Then the full genome is sequenced a second time. By “subtracting” the two sequences, a computer program can determine which C’s were methylated.

      Methylation is one epigenetic marker on the DNA. There are others, and there are yet more that work on the histone spool that the DNA is wound around, not on the DNA directly.

  3. Josh, do you have any idea if the epigenetic clock is proven for short lived animals, such as C. Elegans for example?

    • In a month, I’m going to China to work in a worm lab where my host is asking exactly this question. I might have an answer for you when I return in November.

      In the meantime, I do know three things: (1) C elegans doesn’t have stem cells. The worm has the same 1,000 cells through its adult life. So cell aging is likely to be less a factor. (2) C elegans doesn’t use DNA methylation as an epigenetic marker, and (3) Gene expression changes substantially through the worm’s life history.

      • Thanks a lot. #2 is very interesting as DNA methylation is really widespread even plants use it.
        However if true that would underpin my hypothesis that DNA methylation changes effect the plasticity of stem cells thus effecting organismal aging.
        All research into the aging differentiated cells may thus be an uphill battle compared to studies on stem cells(in case of higher animals).

    • We’ve been burned with gene therapy that kills. AAV has been proved ultra-safe over many years. Lentiviral vectors will have to go through a long, slow process of testing and acceptance.

  4. Thanks, Josh!

    This landmark paper might be of interest:

    The transcriptional landscape of age in human peripheral blood

    It’s just all about 1500 changes in gene expression! A puzzle with 1500 pieces should be solvable, right?

    OK, it looks complicated in the first place, but in fact there are only 4 upregulated gene network clusters and 3 downregulated clusters we should take care off.

    • Very interesting paper thank you very much for sharing. I know I should have a more than cursory look at it. However my impression is that this strategy is destined to fail. I think organismal aging is the result of stem cell aging. But this study is investigating differentiated cells and not stem cells. The difference between stem and differentiated cells is mainly gene expression, whereas the methylome controls the tissue type and the age. The methylation signature of aging can be found in both rapidly dividing and non dividing cells (neurons) but gene expression can be very different. So the transcriptome is more noisy than the methylome with regards to aging.
      I have collected some papers and thoughts here on my blog

    • Thank you! My information was out of date. Apparently there is a new success in turning genes on as well as off. Turning genes on is more difficult because when the DNA is tightly spooled (turned off, heterochromatin) it is less accessible to CRISPR.

  5. How complicated and expensive is (and will be) using CRISPR, is this something we can expect people to tinker with in their garages? (Not saying they should or shouldn’t.)

Leave a Reply

Your email address will not be published. Required fields are marked *