This blog article builds upon ideas from: Science 29 February 2008: Vol. 319 no. 5867 pp. 1215-1220.
Gene and genome editing has been hot in molecular biology since the combination of DNA restriction enzymes and DNA ligases in the late 1970s ushered in the recombinant DNA revolution. Restriction enzymes reliably cut DNA in very specific places allowing for DNA ligase enzymes to paste a known sequence into another place. This restriction enzyme/ligase technology remains dominant in molecular biology today but is being supplanted by a method commonly known as Gibson assembly. In Gibson assembly, synthetic DNA gene constructs are created with common overlapping ends which are stitched together in a single reaction. In 2008, Daniel Gibson used this reaction to compile thousands of synthetic DNA constructs to create the first synthetic genome organism based on the bacteria, Mycoplasma genitalium (Science 29 February 2008: Vol. 319 no. 5867 pp. 1215-1220). M. genitalium has a genome of only 580,000 DNA base pairs whereas the smallest human chromosome (#21) has 48 million base pairs. Gibson assembly was already pushed to the extreme limit in making the M. genitalium genome and it is unclear how long synthetic genomes can be built with current methodology. We will need new ideas for making longer structures relevant for replacing human chromosomes. Still, this was a big step up from restriction/ligase reactions since synthetic DNA constructs could be created by automated machines instead of being cut from naturally occurring sequences.
Over the years the synthesis of longer and longer DNA chains has progressed. For example, a company called Cambrian Genomics was developing a DNA laser printing method, which will print out sequences up to 30,000 base pairs at cost far lower than than current technologies on the market. If companies like Cambrian Genomics and Gen9 can bring their technology to the wider life science market, we will be able to print out re-engineered mammalian chromosomes at greatly reduced rates.
<While writing this blog posting, I did some follow up research on Cambrian Genomics’ progress toward DNA laser printing. Apparently, the founder and main brains behind the company, Austen Heinz, raised $10M USD for further development in the of Fall of 2015. As an advertising ploy, Austen often made inflammatory statements regarding the future uses of DNA printing including living creature creators, designer babies and synthetic probiotics. His company helped fund a startup called Sweet Peach Probiotics creating commensal bacteria that he incorrectly implied would make vaginal probiotics that smell like peaches. His comments were considered juvenile and sexist by the media which apparently made some of his investors pull their money. With a history of depression and a downturn in his business fortunes, Austen committed suicide at the end of May, 2015.>
It’s not as simple as printing out a sequence of DNA and putting it into a cell for gene editing as some may imply. Longer DNA chains such as chromosomes are bulky beasts. For example, an unraveled human chromosome 1 would stretch out to be about 8.5 centimeters long (at 0.34 nanometers per base pair). By comparison, a mammalian cell is about 0.000005 to 0.000025 cm in diameter. Cells overcome these problems by condensing chromosomes into densely coiled structures called chromatin. Since DNA is negatively charged, it requires positively charged companions to neutralize the overall superstructure to achieve condensation. These positive charge companions are called histone proteins. Chromosomes wind and writhe around histone proteins in grape like structures called nucleosomes which compact tightly to condense chromatin from the centimeter range down to the sub-micrometer range. Chromosome DNA does not exist on its own; it is always found tightly twisted in complex with histones. Therefore, if one were to synthesize a large chromosome it must first be packaged similar to natural chromatin before it can be transported. This technology does not currently exist outside a living cell. This limits the current deliverable DNA size to only tens of thousands of base pairs. Furthermore, I hypothesize that a chromosome would be unstable for its intended use if it is not packaged with histones in a chromatin structure *as* it is being synthesized. This is because different kinds of histone proteins have different functions when bound to different regions of a chromosome. Some histones organize packaging into highly condensed structures and others form very loose structures to facilitate gene activity. For example, histones bound to DNA during the synthesis phase of mitosis must be loosely bound to allow the DNA polymerase to create a copy. In another example, one of the two copies of the X-chromosome in normal human females cells becomes highly condensed and inactivated in a complex called a Barr body. Point being, chromatin is dynamic with different flavours and topologies created by changing the landscape of DNA and histone protein variants. These need to be programmed in during assembly before an artificial chromosome will have any real function in a human cell. This probably means that the synthesis of a chromosome, DNA sequence verification, packaging into chromatin and further packaging in a vehicle must occur at the same time.
So the first example of gene editing was the use of restriction enzymes in recombinant DNA technology. The second was the use of Gibson assembly to create the first synthetic genome. Concurrent to Gibson assembly were a few other systems being tailored to create specific breaks or insertions in a genome. These were the zinc finger nucleases, transcription factor activator like effector nucleases (TALEN), type 2 introns and the CRISPR/Cas9 complex. These systems can be tailored to bind to a specific DNA sequence and then produce a DNA cut site. We therefore have the cut event of cut and paste genome editing. For many reasons that I won’t go into, the Cas9 system has become the dominant system for producing this cut event.
The Cas9 system, with its tailored RNA sequence called the guide RNA, binds to a specific site in a genome and creates a double stranded DNA cut. When a cut event occurs it creates two large sections of chromatin which immediately begin migrating away from each other by diffusion. Chromatin size and bulk slows this down considerably, but the occurrence of the two ends finding each other again is problematic. DNA to be inserted in the cut site would require that the right DNA ends find themselves in the nucleus not once but twice. This problem is not currently being addressed in the field. To begin with, it is exceedingly difficult to bring new DNA segments into the nucleus at the same time as the Cas9 system. So today, the most a lab can hope for with gene editing is to introduce one or two cut sites to remove a gene segment. This Cas9 system came from bacteria and will need further engineering to function well in a mammalian nucleus. Cas9 evolved to cut invading bacteriophage virus DNA which have structures completely unlike mammalian genome structures. In conclusion, what we have today with the “genome editing boom” is a cut button without a paste button: Mostly useless.
In order to make useful genome editing a reality we need to be able to hold two chromatin cut ends close together. Next, we need to be able to hold a new DNA segment in the right place for it to be correctly pasted into that cut site. To complicate the matter, the nucleus of a normal cell is very dense. We will need to think about a system which will actively pull the new DNA segments into place prior to assembly as opposed to relying on diffusion to get the right strands together. Of course, the new DNA needs to be transported into the nucleus in the first place which, depending on the size, is not feasible with current technology. This is a seriously complicated task which requires efficient nuclear targeted synthetic virus delivery technology.