In 2012, Jennifer Doudna and Emmanuelle Charpentier made a Nobel prize-winning discovery: crispr cas9. This system allows for specific, targeted editing within a genome, and has been expanded from bacteria into eukaryotes, including – in a shockingly unethical stunt – in humans. Crispr can be used to advance synthetic biology, agricultural research, drug development, and human gene therapy. Clearly, this is a tool that can be used for good or evil. But how does it work?
When the Doudna & Charpentier team published in 2012, they used a series of experiments to pick apart the crispr cas9 system. Once we discover that – and how – bacteria can cut DNA apart, we started morphing the system for synthetic biology applications. Can we use it to delete a gene? Insert new ones? Modify the DNA, or label it? Absolutely.
Why Do Bacteria Chop DNA?
Before diving into their experiments, I must ask – why do bacteria like S. pyogenes have this ability in the first place? What survival advantage is there to this ability to chop DNA?
Bacteria may be a small, simple form of life, but they cope with sub-life phenomena, just like we do. Viruses plague S. pyogenes, and attempt to take over the cell’s DNA replication system so that it can propagate its own destructive genetic material. But S. pyogenes has a trick up its sleeve. It has a piece of RNA constantly scanning DNA within the cell. As it scans, it’s looking for a short DNA sequence called a PAM. These nucleotides – about 2-6 base pairs long – are cut sites in the genome adjacent to a longer cut site that is more recognizable. That longer sequence is called a protospacer, and if it matches the copy stored in the RNA, the crispr protein comes along to chop it.
In other words, S. pyogenes stores a short sequence of viral DNA in its genome, so that an RNA fragment can be made to identify it. As humans, we get vaccines to protect us against viruses. Once our immune cells learn what a pathogen is (from an inactivated version in the vaccine), our immune systems can fight off infection. But bacteria don’t have the luxury of B cells and T cells, or an entire immune system response. They’re just one little prokaryotic cell, after all. So they fight it on the genetic level. And when they encounter a new virus, they copy a protospacer from its DNA and insert it into their own genomes for this RNA sensor to work.
Why Do We Care?
So there’s this bacterial system that chops DNA. Why do we care? Among other things, Doudna and Charpentier demonstrated that this system is programmable. In their 2012 paper, they made sequences to guide the crispr protein to a gene for encoding green fluorescent protein (GFP, or, most labs’ favorite cell label). If that’s possible – if you can just put any sequence in as a protospacer, provided that there’s that choppable PAM next to it – you can use crispr to easily cut apart virtually endless genomes, including viral, bacterial, and eukaryotic.
So why is this revolutionary?
Once the DNA gets chopped apart, it tries to repair itself. S. pyogenes relies on endless cutting until a mutation hides it from the guide RNA – and renders the virus harmless to the cell. But we can modify this system ever so slightly by introducing fragments of DNA that we want the cell to use to repair the breaks.
Add the crispr protein, the customized guide RNA, and genetic material to coax the cell towards some repaired sequence, and you can now delete, insert, label, replace, or otherwise modify the genome.
How Was This System Discovered?
Let’s look at a few of the experiments from the 2012 paper, A Programmable Dual-RNA–Guided DNA Endonuclease in Adaptive Bacterial Immunity. First, crispr cas9 is a big protein. They wanted to know which of the two different domains they saw actually did the cutting. How does a scientist go about deciding which does the action?
Well, you can mutate the gene for crispr – and they did just that. The team set up experiments testing if the crispr cas9 system cut DNA efficiently, reading the results on a gel. They used wildtype (WT) crispr, crispr mutated on the first protein domain, and crispr mutated on the second protein domain. A quick check of the DNA run on this gel (Fig. 2A) shows that if either domain is altered, crispr loses its cutting function. Further similar experiments showed that one domain was cutting the leading strang, and the other the complementary strand. Both are needed to fully chop the double helix.
So crispr itself chops the DNA. What does the RNA itself do? We know it identifies the sequence to be chopped. They didn’t know that yet, obviously. How do you figure out the function of some part of a system, without any clue as to why it is there?
The magic of controls, of course.
To figure out what’s necessary, they ran their template DNA with various components of the crispr system. If you through DNA into a test tube – and nothing else (except, say, maybe the magnesium necessary for the molecules to interact properly) – you are able to make sure nothing is happening outside of the variables you are studying. This is the negative control. If the DNA here is cut, we have missed some part of the crispr system.
What about a positive control? Well, we know that if you mix DNA up with the crispr protein and the RNA, it will cut. Sure enough, they observed cut DNA with thiese conditions (Fig. 1A).
What about the RNA? Turns out, the chopping action of crispr is not enough to cut the DNA. If you omit RNA, we don’t get cut up DNA. So even though crispr is there to cut, it seems that RNA is necessary for cutting to occur. Not only that, but further experiments showed that crispr wasn’t even binding to the DNA if RNA was absent. This told them that RNA’s job is to recognize the sequence.
If crispr cas9 works as a search-and-destroy team, guide RNA is the “search” and the crispr protein, the “destroy.”
Crispr cas9 was discovered in bacteria, in S. pyogenes. It has since been expanded to other species, and even different domains. In A Cas9-based toolkit to program gene expression in Saccharomyces cerevisiae, researchers put together a crispr system to work in yeast.
S. cerevisiae – Baker’s yeast – are also popular in the laboratory. Expanding their abilities into an easy crispr system would be very helpful for bioengineers everywhere! The typical synthetic biology toolkit – plasmids replicated or integrated in a cell – has its shortcomings, including instability, recombination, and low expression of the new DNA. Would adding a system like crispr make it not just easier to engineer yeast’s genome, but more efficient?
While the S. pyogenes enzyme crispr cas9 works well in yeast, the guide RNA (gRNA) needs to change. Now, gRNA has to be designed that will guide the cas9 to a place in the genome amenable to being cut. In this study, they found 23 places in yeast DNA where you can cut and insert new genes without hurting the cell. They then inserted fluorescent tags in each and every one of them – these random locations could change how brightly the tag fluoresced and even how easy or difficult it is to insert a new tag there. This kind of data is gold to bioengineers looking to put genes for some new process into the genome.
Promoters are another way to control gene expression rates, and they provide 37 of them. These promoters are places for transcription enzymes to bind to the DNA, so they control how heavily (and sometimes when) a gene is expressed.
Of course, this gene expression changes if you swap the media (cell food for the yeast). Is the cell in exponential phase (growing like crazy) or stationary (the population is at steady-state)? Gene expression changes for growth phase, too (see Fig. 3 and Table 1). All this would influence which promoter and cut site you prefer.
Genes, tags, promoters, cut sites – that’s a lot of genetic material to keep straight. And you have to use it to design multiple parts in your toolkit. The guide RNA, for guiding to the chosen cut site? That’s going on the plasmid containing crispr cas9 – the yeast will read that plasmid and make the cas9 enzyme, and the gRNA sequence. So you need primers for constructing that cas9/gRNA plasmid. Then you have to make the “donor fragments” – the DNA you want to insert. It will have genes or tags, the promoter, a terminator, and a “20-[nucleotide] guide sequence” on either end to match it to its cut site (see Fig. 1). Designing primers for all that, so I can build them all in my lab? That’s a lot. So they also wrote up a software for DNA design. Choose you genetic parts, it tells you what primers to order for each step.
These are the sorts of things to be developed when crispr is adapted into a new organism. Of course, now, scientists have gone beyond bacteria and yeast, and into plants and animals. It’s a rapidly growing field with promising abilities. Should it be used in humans? Not at all. But it can be used to expand yeast into making more for us than bread and beer.
Apel, A. R., D’Espaux, L., Wehrs, M., Sachs, D., Li, R. A., Tong, G. J., Garber, M., Nnadi, O., Zhuang, W., Hillson, N. J., Keasling, J. D., & Mukhopadhyay, A. (2017). A Cas9-based toolkit to program gene expression in Saccharomyces cerevisiae. Nucleic Acids Research, 45(1), 496–508. https://doi.org/10.1093/nar/gkw1023
Doudna, J. A., & Charpentier, E. (2014). The new frontier of genome engineering with CRISPR-Cas9. Science, 346(6213), 9–10. https://doi.org/10.1126/science.1258096
Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A., & Charpentier, E. (2012). A Programmable Dual-RNA – Guided. Science, 337(August), 816–822.