Nice little write-up. If I may offer a couple of comments. You're point about off-target effects is a good one, but I feel like we've already largely solved that problem through Cas9-nickase. Or at least drastically reduced it as a problem. Also, the development of base-editing and point-editing from CRISPR promises even greater control over the type of mutations we introduce. Not being limited to simply knocking out the gene, but fine-tuning the sequence.
That being said, I feel like you overstate the potential of CRISPR therapies in the short term. I think you yourself recognize the limitations of CRISPR for therapeutic purposes give our current level of knowledge of molecular regulatory processes, given that you spent a good chunk of the end of the article discussing a cancer therapy that doesn't involve the use of CRISPR at all. But the idea of the p53 cancer therapy highlights the major roadblock for for CRISPR therapies: i.e. we can be confident in our hypothesis that additional p53 genes may significantly reduce the risk of cancer because we already know what role p53 plays in the cell.
The reason the Human Genome Project didn't produce the medical benefits that many people thought it would is because mapping the genome didn't tell us what any of those genes did, nor how they interacted together to produce the cellular and organismal phenotype. CRISPR runs into the same problem. If we don't already know what a gene does then knocking it out runs the risk of producing any number of unintentional effects. Not an ideal scenario for a medicine.
Of course, CRISPR is also uniquely positioned to help us solve that problem. More almost a hundred years now our preferred method of figuring out the function of a gene has been to break it and see what happens and CRISPR has made that process so, so much easier. So in some ways, CRISPR will help bring about its own therapeutic relevance by speeding up the process by which we determine gene function.
Thank you for your comments! I largely agree with what you have to say. Regarding base and prime editors, I would like to see more clinical data before concluding the off-target problem is solved, more or less. But it certainly seems true that these methods and tools are more likely to be utilized for wide spread therapeutics as CRISPR becomes more medically mainstream.
On the topic of p53, it does not necessarily require a CRISPR system (e.g., could use viral vectors or transposon systems) like you said. In my mind though, the likely scenario in the real world uses CRISPR with either an HDR template or an integrase-like system for its more pinpoint modification of the genome than say lentiviral or AAV based approaches. Particularly since in this sci-fi-like scenario we would be inserting this gene into healthy people’s genomes. This could have certainly been made more clear in the write up. I do think that CRISPR therapies are overhyped in the short term (5-10 years) and under hyped in the longer term (30-50 years) as the technology matures. I tried to convey that via the “in our lifetime” timeframe but I understand how that could have been made more clear.
I whole-heartedly agree with the last point you make about our limited understanding of the human genome as a living system of information and CRISPR’s potential in unlocking more intel. I felt that was slightly outside of the scope of this post, maybe a part 2 could cover this and other missed points?
Thank you again for your feedback, I appreciate you taking the time to read and respond!
Nice little write-up. If I may offer a couple of comments. You're point about off-target effects is a good one, but I feel like we've already largely solved that problem through Cas9-nickase. Or at least drastically reduced it as a problem. Also, the development of base-editing and point-editing from CRISPR promises even greater control over the type of mutations we introduce. Not being limited to simply knocking out the gene, but fine-tuning the sequence.
That being said, I feel like you overstate the potential of CRISPR therapies in the short term. I think you yourself recognize the limitations of CRISPR for therapeutic purposes give our current level of knowledge of molecular regulatory processes, given that you spent a good chunk of the end of the article discussing a cancer therapy that doesn't involve the use of CRISPR at all. But the idea of the p53 cancer therapy highlights the major roadblock for for CRISPR therapies: i.e. we can be confident in our hypothesis that additional p53 genes may significantly reduce the risk of cancer because we already know what role p53 plays in the cell.
The reason the Human Genome Project didn't produce the medical benefits that many people thought it would is because mapping the genome didn't tell us what any of those genes did, nor how they interacted together to produce the cellular and organismal phenotype. CRISPR runs into the same problem. If we don't already know what a gene does then knocking it out runs the risk of producing any number of unintentional effects. Not an ideal scenario for a medicine.
Of course, CRISPR is also uniquely positioned to help us solve that problem. More almost a hundred years now our preferred method of figuring out the function of a gene has been to break it and see what happens and CRISPR has made that process so, so much easier. So in some ways, CRISPR will help bring about its own therapeutic relevance by speeding up the process by which we determine gene function.
Thank you for your comments! I largely agree with what you have to say. Regarding base and prime editors, I would like to see more clinical data before concluding the off-target problem is solved, more or less. But it certainly seems true that these methods and tools are more likely to be utilized for wide spread therapeutics as CRISPR becomes more medically mainstream.
On the topic of p53, it does not necessarily require a CRISPR system (e.g., could use viral vectors or transposon systems) like you said. In my mind though, the likely scenario in the real world uses CRISPR with either an HDR template or an integrase-like system for its more pinpoint modification of the genome than say lentiviral or AAV based approaches. Particularly since in this sci-fi-like scenario we would be inserting this gene into healthy people’s genomes. This could have certainly been made more clear in the write up. I do think that CRISPR therapies are overhyped in the short term (5-10 years) and under hyped in the longer term (30-50 years) as the technology matures. I tried to convey that via the “in our lifetime” timeframe but I understand how that could have been made more clear.
I whole-heartedly agree with the last point you make about our limited understanding of the human genome as a living system of information and CRISPR’s potential in unlocking more intel. I felt that was slightly outside of the scope of this post, maybe a part 2 could cover this and other missed points?
Thank you again for your feedback, I appreciate you taking the time to read and respond!
Your blog is on absolute fire.
Thank you so much Niko! Your blogs and work at Asimov Press has been a huge inspiration