I'll be summarizing a 2014 review from the journal Cell. The title of the review is "Development and Applications of CRISPR-Cas9 for Genome Engineering" (unfortunately, the review is behind a paywall).
As you can imagine, gene editing is somewhat of a holy grail. To be able to erase undesired mutations in DNA would be a dream for many clinicians/doctors. But there are many different applications that don't necessarily have to do with erasing what we don't want, rather, we could introduce variations that we want: creating an animal model for a disease, developing crops with desired traits, etc.
The paper starts by providing a brief overview of how genome engineering has been done in the past (focusing primarily on mammalian cells), outlining the pros and cons of each method. The difficulties inherent in current methods has led to the development of programmable gene editing technologies, the most promising of which is the CRISPR-Cas9 system which is the focus of the paper.
Spouse: I need to clarify something before I continue. The system isn't quite as simple as I'll describe below. There are additional DNA/RNA sequences and enzymes that are part of the complex, and I'll be omitting them because I could hear your voice in my head complaining about all the acronyms and jargon. So consider this to be the uber-abridged Cliffs Notes on gene editing.
Cas9 is an endonuclease, meaning that it's an enzyme that can cut both strands of DNA's double helix. Endonucleases can be random, meaning that they can cut anywhere along the length of the DNA. In fact, one of the fears of scientists working with DNA is nuclease contamination, which can render your samples to DNA dust. Other endonucleases, such as restriction enzymes, search for a specific DNA sequence and cut at that site. However, the cut site cannot be specified and can be found frequently in a genome (i.e. restriction enzymes can cut thousands of times).
Unlike restriction enzymes, the bacterial Cas9 cuts at a specific site but the DNA sequence where it cuts can be specified. Cas9 is associated with the CRISPR system, which guides Cas9 to its target using a small piece of RNA. In nature, this small piece of RNA generally encodes for a viral (phage) sequence. Cas9 searches for the viral sequence and then hacks it up, which is why the CRISPR system is part of the bacteria's antiviral defense mechanism. However, the small piece of RNA that guides Cas9 can be replaced with a sequence of the researcher's choice. Part of the system's benefit is the fact that you can provide more than one guide molecule, meaning that you could direct the system to cut more than one place, if desired. The system can be specific in the DNA sequence that it cuts, however, as this paper highlights, off-target edits can occur and are an area of ongoing study/research.
The paper provides a history of the many discoveries surrounding CRISPR-Cas9. There are different mechanisms by which the CRISPR system gets activated, and after many years of research, it was decided that the CRISPR-Cas9 was the most promising in terms of trying to find a programmable system for gene editing. By 2013, researchers successfully engineered the CRISPR system from two types of bacteria, including the one used to make yogurt, to edit genes in mammalian cells.
So far, I've described how to get the system to cut where you want it to cut. But then what? If you think of editing as deleting something that's incorrect and typing in something else, how do you get "what's right" or "what you want" into DNA?
Once the DNA is cut, the cell's natural repair mechanism kicks in and one of two things can happen (see my bee-u-ti-ful graphic below):
- The two loose ends of the DNA get glued back together again. The system is error prone but easy to use, so if your goal is to create a protein that doesn't function or to delete it altogether, this may be the way to go. This process is known as non-homologous end joining.
- The break is detected by enzymes that look around for the proper template to use to fill in the gap. If that template is provided artificially, then it will copy in that sequence. The template that researchers provide can contain the desired sequence, additional sequence, etc. This process is known as homology directed repair.
CRISPR-Cas9 can also be modified so that the "search" function of the system remains intact, but the cutting function is disabled. As such, researchers can create a complex where they guide their enzyme of choice to a specific region. I think that the simplest example that the paper provides is one where Cas9 was fluorescently labelled/tagged, so researchers could visualize the location of the DNA sequence they were studying.
In reviewing this article, the spouse asked if we could write a movie script where the villian sprinkled Cas9 along with the DNA specific to his arch-nemesis' genome into the latter's cereal. Would it be the perfect crime? Would the CRISPR-Cas9 enter the arch-nemesis' body and hack up his DNA? Unfortunately, no. Keep in mind that our DNA is within the nucleus of our cells and isn't very accessible. Additionally, as non-bacterial species, we don't have CRISPR-Cas9 in our cells. Getting the CRISPR-Cas9 into the nucleus isn't all that simple and requires a bit of fancy lab work.
To date, there's no medical therapy on the market developed using gene editing. Likewise, there's no crop on the market that has been engineered using CRISPR-Cas9. However, studies have demonstrated that crops can be modified using the system (this paper provides an example of successful gene editing using CRISPR-Cas9 in rice). Consequently, many wonder whether crops generated through gene editing would be considered GMOs.
As I've previously described, what are currently known as a GMOs or genetically modified organisms are transgenic crops, meaning that a gene from a different species has been added to their genome (NOTE: in this post, I note that anti-GMO activists have a very different definition of a GMO). But in the case of crops modified using CRISPR-Cas9, what's edited was there to begin with. Technically, nothing has been added from a different species. So how will regulatory agencies categorize these crops?
This paper published just last month provides a great summary: it states that the USDA has concluded that if you cannot distinguish an edit from a naturally occurring mutation, then it's not a GMO. Additionally, if a gene is deleted using the cell's own repair mechanism (as is the case with non-homologous end joining), then it isn't a GMO either. Interestingly, the paper states that the USDA has waived regulations on two crops generated using gene editing, because they fell within these categories. The European Union has yet to determine how these crops will be classified, because they consider something to be genetically modified if "it is altered in a way that does not occur naturally by mating and/or natural recombination" (although crops generated through mutagenesis are not regulated in the EU. Please see my previous post on mutagenesis for more information on this technique). There are two additional points that the paper makes that I completely agree with: 1) if the EU's definition of a GMO does not end up aligning with the USDA's, the regulation of these crops for import will be very difficult since there will not be an easy way to detect if the crop is a product of gene editing. 2) If the EU's definition of a GMO does not end up aligning with the USDA's then the cost of getting a crop through regulatory hurdles will limit the development of these plants to large biotech companies, which will stifle innovation; i.e: if you want someone other than Monsanto, Syngenta, et al to make a biotech crop, these crops should not be considered GMOs.
To conclude, here's my first "infographic" on the different methods or ways used to develop new traits in crops and feedback would be appreciated. My perspective on gene editing is the same as it is on transgenesis and mutagenesis: crops should be regulated based on the trait introduced/modified, not on the way that the introduction/modification was generated.