Editing the Genome Quickly and Efficiently: Round 1

So much of biology depends on manipulating the genome. Since the human genome has been sequenced, the clear next step is to identify the function of each gene encoding portion of the genome. Gene mutations give us a direct idea of the role that gene plays within the body.

Many different fields within biology have been abuzz recently because of the enormous steps in genome editing which allows for easier manipulations of the genome. The ability to mutate individual genes or introduce exogenous DNA, additional DNA not from the organism you are putting it into, in the hopes of altering the genome has been studied for nearly 40 years. The field has been building onto itself, with each step improving the last, and recent advancements allow for efficient alterations unlike we have seen before.

This will be a two part blog post. I would like to start by doing an incredibly brief catch-up of the last 40 or so years of genome editing, although it won't give the field or any of the individuals involved nearly enough credit. This will show where we started and the huge advancements of the last decade while the next blog post will go into greater detail about the recent explosion of the Crispr-Cas9 system.

Altering the genome started humbly. Initially, cell lines were simply exposed to DNA constructs containing a gene of interest they wished to insert. This was put directly into the media bathing the cells. This led to an infection rate of about 1 in 1 million cells. Soon thereafter, the DNA constructs were injected directly into cells using incredibly thin needles. By using this method, the cell would incorporate the DNA construct, and even pass the message onto daughter cells. This would work in about 1 out of 3 cells that were injected. This was a huge breakthrough, although injecting into single cells is painstaking and is more difficult to turn into a large scale experiment.

These methods, however, only allowed for random integration of a gene of interest. The next step would be to see whether they could target specific loci in the genome for integration. By randomly inserting DNA constructs like we saw previously, scientists realized they could target those inserted DNA constructs and reverse the phenotypes the random inserted gene created. This meant they could target specific sequences in the genome, albeit inefficiently (roughly 1 in 1000 cells would be rescued). These sequences would loop into the genome using homologous recombination. This could even be done in embryonic stem cells, then put into a blastocyst and transplanted into a pregnant mouse. This would hopefully create an offspring with the mutation of interest.

More recently, there has been targeting, not against the gene itself, but the product of the gene. Using RNAi (similar methods include shRNA, and siRNA), they can target the mRNA created by the gene of interest and degrade it. This is a very powerful technique. It is easier transmitted into multiple cells, and mRNA is easier to target than the genes themselves. However, it comes with several caveats: there are off target effects, the level of knockdown is variable from experiment to experiment, and the knockdown is transient. So, although it is relatively simple to use, it is not ideal for large scale experiments.

Within the last 10 or so years, there have been enormous progress in the development of nucleases, enzymes able to cut the phosphodiester bond that makes up the backbone of DNA. These nucleases are attached to signals that are attracted to specific sequences of DNA so they can be targeted to precise locations in the genome.

Zinc Finger nucleases (ZFNs) use a DNA binding motif called a zinc finger (because of its shape, and because it contains, you guessed it, Zinc). These zinc fingers have been engineered to bind the 64 unique three-amino-acid sequences that can be found in DNA. So one would simple have to identify the sequence in the DNA where they want to cut, and create the appropriate sequence of zinc finger domains to match the sequence (see image).

Another, similar, targeted system for DNA cutting are Transcription activator-like effector nucleases (TALENs). They work in a very similar way to the ZFNs. TALE proteins occur naturally in Xanthomonas bacteria strain. The protein contains a pattern of 5 different proteins that recognize one of the 4 bases of DNA. When the sequence of interest is determined, one can design the TALENs to recognize that sequence by altering the order of the 5 different proteins.

The next step to both ZFNs and TALENs is adding a nuclease to the end of the DNA binding motif so the DNA can be cut where the protein has been bound to. This forces the cell to repair the DNA which will often lead to mistakes in the sequence, and thus, mutations in the gene of interest. In the end, these techniques allow for a directed and specific cut of the DNA in live cells that often leads to mutation.

This concludes Round 1 of what will be a two part post on genome editing. Next week we will look at some of the caveats with ZFNs and TALENs, and why Crispr/Cas9, the next iteration in directed mutagenesis, has so many people talking.

References:

Capecchi, M.R. Gene Targeting in mice: functional analysis of the mammalian genome for the twenty-first century. Nature Reviews Genetics, 6, 507-512 (2005).

Gaj, T., Gersbach, C.A., Barbas III, C.F. ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering. Trends in Biotechnology, 31, 397-405 (2013)

Image credits:

Both imagines of Zinc Fingers and Talens come from the Gaj et al. paper in Trends in Biotechnology.