Chapter 6. Better, CRISPR Homemade Genomes

Derek Jacoby

The point of synthetic biology is editing genomes—adding and removing functionality until the organism you are creating behaves exactly as you want it to. Since this behavior is determined by DNA, these additions and deletions must take place in the genome.

When I was first getting into synthetic biology in 2008, some friends and I formed an iGEM team at the University of Victoria. Everything we worked on was based around BioBricks: biological units of defined functionality. BioBricks contain specific DNA sequences at each end of the DNA so that they can be precisely snipped out with restriction enzymes and then put back together in the order you want them. I remember being amazed at the specificity of restriction enzymes—to be able to cut a piece of DNA at a very precise point confers an amazing ability to move segments of DNA around as functional units and create genetic circuits. However, all the circuits we created had to be on small, circular segments of DNA called plasmids. This introduced some significant limitations: we could only create new circuits, existing functionality in the chromosomal DNA was inaccessible to us, the plasmids had to contain antibiotic resistances, and if we ever placed the new constructs in an environment without that antibiotic selection, our hard-won changes would be gone in just a few generations. So everything we created felt temporary. And it all came about because our palette of restriction enzymes was fairly limited. This is rapidly in the midst of changing.

Before talking about how that change is occurring, let’s delve into restriction enzymes.[5] Evolved as a bacterial defense mechanism, a particular bacteria generates an enzyme that cuts a specific segment of DNA. Most frequently, it cuts DNA possessed by an invading bacteriophage—a virus that infects a bacteria—so that the phage DNA is disrupted without harming the bacteria itself. Our menu of restriction enzymes is derived from these evolutionary battles between bacteria and phage and are named after the bacteria they are isolated from. For instance, EcoRI (pronounced “Eco-R-One”) is the first restriction enzyme isolated from E. coli and cuts precisely at the sequence GAATTC. About 600 of these enzymes are available commercially, which provides a significant set of options for how one can design and manipulate synthetic DNA. But it does no good when trying to cut a segment of DNA that does not match up to one of the available enzymes, so it is of very limited utility in editing natural chromosomal DNA.

Restriction enzymes are DNA-binding proteins, and the activity of a protein depends largely on its shape. Although protein structural prediction is improving, it is still a very difficult problem, so creation of artificial restriction enzymes to cut at a particular point is nontrivial. Some approaches to creating sequence-specific DNA binding proteins do exist, though. The first to arrive on the scene were ZFNs (zinc finger nucleases). More recently, TALENs (transcription activator–like effector nucleases) provide another alternative. In both of these cases, though, the resultant DNA binding protein is rather cumbersome to create and suffers from problems such as off-target binding. In practical terms, this means that it is difficult and expensive to create a new zinc finger or TALEN to target a specific DNA location for in situ genome editing.

In late 2012, however, the first paper was published on adapting for gene editing a defense mechanism that exists in Streptococcus pyogenes called CRISPR (Clustered Repetitively Interspaced Short Palindromic Repeats).[6] In combination with a protein called cas9, CRISPR binds and cuts DNA at a specific location. Unlike restriction enzymes discussed thus far, however, the exact sequence at which DNA is bound and cut is determined by a short sequence of RNA. This means that no complex protein engineering is required to target a new specific section of DNA, just a short section of RNA. This RNA is most commonly provided as DNA in a plasmid, allowing the cell machinery itself to create the guide RNA to target the CRISPR-cas9 complex appropriately. It is functionally a programmable restriction enzyme! In the year since the first publication of this approach, the CRISPR-cas9 system has been used to cheaply and effectively edit DNA in humans,[7] mice,[8] zebrafish,[9] yeast,[10] bacteria,[11] and vascular plants.[12] There is reason to believe that it is a fully generalizable system for in situ genome editing. It has even been used already to make germline edits in mice, which were successfully passed down to subsequent generations.[13]

Although a detailed discussion of the mechanism of action of CRISPR-cas9 systems is beyond this quick introduction,[14] it is worth briefly noting the highlights of the system to better understand its strengths and limitations. First, there is a particular DNA sequence needed, known as the PAM sequence (proto-spacer associated motif), which corresponds to the particular nucleotide sequence NGG. As an indication of the frequency of this motif, there are approximately 642,000 sites on the yeast genome where the CRISPR complex can bind with maximal efficiency. Some elements of the system can be manipulated to bind with reduced efficiency if an exact PAM sequence is not located near your target of interest, so most genes can be targeted with some degree of efficiency.[15] Next, the guide and spacer RNAs have some length requirements to ensure optimal specificity. But beyond these bioinformatic tasks, the production of a highly specific CRISPR-cas9 complex consists merely of the synthesis of a few dozen bases of oligos and inserting the CRISPR and cas9 plasmids into your target of interest, a far cry from the complexity of previous approaches.

In all the cases discussed here (zinc finger nucleases, TALENs, and CRISPR-cas9), the end result of this specific targeting is a break in the DNA at that point. Either in only a single strand[16] or in both strands. The goal of this single- or double-stranded break is to enable new genetic material to be inserted, or to disable an existing gene. The insertion of a new genetic sequence depends on a DNA repair process known as homologous recombination—essentially built-in DNA repair machinery to knit the DNA back together. If you provide compatible DNA donor sequences, a certain portion of the time they will get stitched into that specific point of breakage in the chromosomal DNA. Or, if you cut at two adjacent points in the genome and do not provide a compatible donor DNA sequence, the target will get stitched back together without the segment you want deleted.

We’re entering an exciting era in genetic engineering where we have tools accessible to not only create new genetic circuits on plasmids, but also to edit the existing chromosomal DNA of a wide variety of organisms, and to do it using tools that are low cost and accessible. At BioCurious, we’ve begun (but not yet seen results of) CRISPR-cas9 editing of the E. coli genome. You’ll hear more about those efforts in future reports. But there’s a concern: the tools to work with CRISPR-cas9, and many other related technologies, are the purview of academic researchers, and their commercial use requires individually negotiated licenses. As a nonprofit institution, BioCurious is able to order materials from Addgene, a repository of research plasmids, for research and educational use of those materials at BioCurious. As an educationally focused nonprofit, BioCurious is essentially acting in the role of a research university in supervising the material transfer agreements (MTAs) that protect the intellectual property rights of the original depositors of biological material. It’s a model that seems to be working well, but blurs the line between DIY biology and the support and legal structures that underpin institutional biology. Along with safety and regulatory concerns, the legal status of the essential tools and organisms that form the core of modern synthetic biology are geared toward institutions rather than individuals. It’s a core challenge of DIY biology to figure out how to work with those structures in such a way that individual freedoms to explore and learn are balanced with the needs that led to the development of the institutional control structures in the first place.

[5] Pingoud, A. and A. Jeltsch. "Structure and function of type II restriction endonucleases." Nucleic Acids Research 29 (2001): 3705–3727.

[6] Jinek, M. et al. "A Programmable Dual-RNA–Guided DNA Endonuclease in Adaptive Bacterial Immunity." Science 337 (2012): 816–821.

[7] Mali, P. et al. "RNA-Guided Human Genome Engineering via Cas9." Science 339 (2013): 823–826.

[8] Wang, H. et al. "One-Step Generation of Mice Carrying Mutations in Multiple Genes by CRISPR/Cas-Mediated Genome Engineering." Cell 153 (2013): 910–918.

[9] Hwang, W. Y. et al. "Efficient genome editing in zebrafish using a CRISPR-Cas system." Nature Biotechnology 31 (2013): 227–229.

[10] DiCarlo, J. E. et al. "Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems." Nucleic Acids Research 41 (2013): 4336–4343.

[11] Jiang, W. et al. "RNA-guided editing of bacterial genomes using CRISPR-Cas systems." Nature Biotechnology 31 (2013): 233–239.

[12] Feng, Z. et al. "Efficient genome editing in plants using a CRISPR/Cas system." Cell Research (2013). doi:10.1038/cr.2013.114

[13] Wang, H. et al. "One-Step Generation of Mice Carrying Mutations in Multiple Genes by CRISPR/Cas-Mediated Genome Engineering." Cell 153 (2013): 910–918.

[14] Barrangou, R. "RNA-mediated programmable DNA cleavage." Nature Biotech 30 (2012: 836–838.

[15] DiCarlo, J. E. et al. "Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems." Nucleic Acids Research 41 (2013): 4336–4343.

[16] Mali, P. et al. RNA-Guided Human Genome Engineering via Cas9. Science 339, 823–826 (2013).