A Researcher's Experience with CRISPR

test tube drop.jpg

Dr Karin Purshouse shares her excellent insight into using CRISPR in the lab, having worked with it herself as a researcher in the USA. Karin is currently working as a doctor in the UK, whilst attempting to juggle a side-line in research. Read her blog here.


When I tell friends that I did research on something called CRISPR, the first response is often a raised eyebrow and a belief that I’m making some terrible potato-related joke.  Since the publication of the Human Genome Project over a decade ago, scientists have been trying to utilise our knowledge of the entire human DNA sequence to understand and tackle human disease.  Progress has been slower than many anticipated – will CRISPR/Cas9 help to take off the brakes?

Why is CRISPR/Cas9 interesting?  If we can induce double stranded DNA breaks in very specific places, we can cut out chunks of DNA, or even manipulate DNA repair mechanisms to insert small base changes.  That means we can develop isogenic cell lines – cells which only differ from each other by virtue of the specific gene of interest.  The off target effects of siRNA and viral gene manipulation make this an extremely attractive option for scientists conducting basic research looking to understand the impact of genetic changes on disease initiation and progression.  Ultimately CRISPR can be used to screen for novel drug targets as well as screen the drugs themselves, such as the programme currently underway at Astrazeneca.  Another exciting potential area of use would be to eliminate genetic diseases at the source – what if you could cut out known gene defects at the embryonic stage of fetal development, or implant corrected differentiated cells back into a patient? 

But, as with most things that sound too good to be true, the same could be said of CRISPR.  For one thing, it is sold as being far easier, cheaper and quicker than its predecessors, (transcription activator–like effector nucleases) TALENs and Zinc Finger Nucleases (ZFNs) – in reality, achieving successful on-target cleavage can be extremely challenging, and the optimisation stage can take many months.  Selection and identification of successfully mutated cells can be even more challenging than the initial cleavage process.  That’s not to speak of the high number of on- and off-target effects that seem to vary with each guide RNA and target DNA, and what’s worse is that we have no effective method of screening for particularly the off target effects.  The high number of off-target effects identified in the first embryo-based trial of CRISPR/Cas9 should be a major red flag in considering this technology in humans. 

Telling other researchers that you’ve done work on CRISPR is usually met with a heightened level of enthusiasm first, and bafflement that I ‘got it to work’ second.  As a clinician I came to glioma research because frankly it’s a really horrible cancer. Our treatment options are archaic and the outcomes are terrible compared with many other cancers.  That’s why I chose to do a glioma based research project as ‘gap year’ from medicine – to, as my American supervisor put it, ‘actually try and beat this disease rather than just improve ‘progression free survival’’.  I hopped across the pond and spent a year in a translational laboratory at Yale University.  I thought I’d learn about a novel, cutting edge area of science and perhaps make something useful for the oncology world.  In reality I nearly lost my mind trying to induce a single base mutation into brain cells to characterise a common yet poorly understood mutation. The biggest challenge was that not only was I trying to cleave the DNA in my cells in a very specific place, but I was also trying to persuade my cells to repair the break via homologous recombination with a surrogate strand of DNA.  Designing a guide RNA and Cas9 and getting DNA cleavage proved relatively straightforward.  The real challenge was inducing a mutation, and then finding the small number of cells that actually contained the mutation.  In doing so, it also became very apparent that CRISPR/Cas9-induced cleavage results overwhelmingly in error-prone DNA repair (non-homologous end joining), which can mess up even successfully mutated cells.  I threw numbers at the problem, devising a high volume, high throughput screening method to test nearly 3000 single cell clones.  I got one, but only one, single cell clone that had worked.  That cell line will hopefully reveal the secrets of this particular mutation by serving as an isogenic cell line which can be directly compared with wild-type cells, and provide a basis for high throughput drug screening.  As a clinician, it’s pretty exciting to think that those cells could result in a drug that I might use to treat gliomas as an Oncologist.   So yes, I nearly lost my mind screening so many cells, but that’s a pretty strong motivator!

CRISPR/Cas9 remains a buzz topic in genomic research.  The question is whether it will allow the promise of the human genome project to be realised, or whether it is simply too non-specific to be safe?