What is CRSPR? With Sam Sternberg
No fancy sci-fi introduction is needed this time, folks, because today we have world-renown CRISPR expert Sam Sternberg. Sam worked with biochemist, author, and science celebrity Jennifer Doudna to refine CRISPR gene editing technology at UC Berkeley. The two wrote best-selling book A Crack in Creation about the discovery and what it could mean for humanity. Today Sam is working on CRISPR at Columbia University in New York. Sam was kind enough to take an hour or so out of his busy schedule to chat about CRISPR and how it’s being used to revolutionize science and medicine.
Dustin Driver: Hello, and welcome to Let’s Get Mental. I’m your host Dustin Driver. No fancy sci-fi introduction is needed this time, folks, because today we have world-renowned CRISPR expert Sam Sternberg. Sam worked with biochemist, author, and science celebrity Jennifer Doudna to refine CRISPR gene editing technology at UC Berkeley. The two wrote the best-selling book A Crack in Creation about the discovery and what it could mean for humanity. Today Sam is working on CRISPR at Columbia University in New York. Sam was kind enough to take an hour or so out of his busy schedule to chat about CRISPR and how it’s being used to revolutionize science and medicine. Without further ado, here’s my chat with Sam Sternberg.
Dustin Driver: What I’d like to start out with is a very basic description of CRISPR and how it works.
Sam Sternberg: Sure. Well, do you want to know about CRISPR in bacteria or CRISPR as a technology?
Dustin Driver: Let’s start with CRISPR in bacteria so that we can get an idea of how it works in nature before we talk about how it works in biochemistry or in a lab.
Sam Sternberg: Got it. The acronym itself, as you I’m sure know, stands for clustered regularly interspaced short palindromic repeats. That’s basically the term that was given to these specialized regions of the genome of DNA from bacteria that was first discovered in 1987, before the term CRISPR existed. By the early 2000s, researchers were seeing these same bizarre, repeating sequences in about half of all bacterial species whose DNA was being sequenced. So, the core feature is it’s these series of repeats where the same sequence is present many, many times, separated by the same exact length of intervening sequence. This was a pattern that hadn’t really ever been seen before, and so by the mid-2000s, it was recognized that this is pervasive. It exists all across the bacterial and archaeal kingdoms, but the function was a big mystery for 20 years. It was in 2005 when the first clues came out what it might be doing, and that was a result of the observation that if you focus not on the repeating sequences but the little snippets of DNA tucked in between the repeats, these were often a perfect match to viral DNA. So that was the first clue that CRISPRs had something to do with viruses and might be functioning as a viral immune system.
Sam Sternberg: These are bacterial viruses, so just like humans and other higher organisms, bacteria also have viruses that have the threat of infecting them, and these bacteria-specific viruses are actually one of the most prevalent forms of life on our planet. So you can go into pretty much any environment where you find bacteria, and you’ll find viruses there too that infect those bacteria, and they usually outnumber those bacteria about 10 to 1. So, it wasn’t a big surprise actually that bacteria might have a new way of defending against viruses, and just in the last couple of years, researchers are continuing to discover new immune systems that bacteria have evolved to fend off this threat. It’s kind of a cool area of research, because this is one of the oldest evolutionary struggles between different types of life. You have bacteria and the viruses that infect them, and over billions of years of evolution, both sides have evolved different mechanisms for attack and defense.
Dustin Driver: So this is pre what we think of as an immune system. This is so basic it’s on a genetic level, and these little bits of viral DNA between the CRISPR are used to identify attacking viruses as they enter the bacterial cell.
Sam Sternberg: Exactly. And it was appreciated before CRISPRs that bacteria have immune systems, but all of the ones before CRISPR fall into the bucket of being innate, meaning bacteria have these kind of defenses, but they can’t learn. They can’t adapt over time. Humans of course have adapted immune systems, so we can develop antibodies against new pathogens, and that’s one of the reasons why something like vaccination works, because you can actually vaccinate a baby with a small amount of attenuated or defective virus, and now that will lead to a permanent reservoir of immune cells that can detect and destroy that pathogen if it ever infects the human again later in life.
Sam Sternberg: Up until 2005, 2007, it wasn’t thought that bacteria had evolved this complex of an immune system to actually adapt over time. And that’s what’s actually special about CRISPR: it can learn. So by splicing sequences of DNA from a virus into its own genetic material, CRISPRs actually give the bacterium a permanent reservoir of information that it can recall any time in the future. Kind of like how you might keep biometric identifications like fingerprints or retinal scans stored so you can recognize certain individuals time and time again by those identifiers, bacteria are actually doing the exact same thing with DNA. They store little bits of DNA from the virus in their own genome so that they can recall that information to recognize those same viruses during a future infection.
Sam Sternberg: I think we got right back into the weeds, which I think you wanted me to stay out of, but the point being CRISPRs were discovered in bacteria, and it was discovered, after many years of research, that they enable bacteria to remember viral infections and use that information to destroy viruses during future infections. It was the full investigation of how that works that really led to this technology that we know of today as CRISPR-based gene editing.
Dustin Driver: Right. And this technology is basically taking that natural process, and instead of inserting the viral DNA that was used for identification, researchers can now insert pretty much any snippet of DNA they like.
Sam Sternberg: Exactly. In the landmark paper from Jennifer Doudna and Emmanuelle Charpentier, they made this switch where they swapped out the natural viral sequence for a user-defined sequence. In that first paper, they actually wanted to show that they could target any gene they wanted. They were targeting a jellyfish gene that codes for a fluorescent protein. That was just a convenient bit of DNA that was lying around in the lab, but they could take this gene from a completely different source and select arbitrarily different sequences of the gene to target with CRISPR, convert that information into a snippet of RNA -- the kind of functioning molecule that CRISPR systems use for that identification -- and then show that, indeed, these core components could now use this user-defined sequence to slice apart the jellyfish gene anywhere they wanted to.
Sam Sternberg: So that kind of was the clinching piece of evidence that this was in fact programmable, that whereas bacteria are using CRISPRs to learn from past infections to destroy viruses, that researchers could now program the same system to recognize and cut, and eventually edit, any sequence that they might be interested in. And of course now, in the era of genomics where we have full genome sequences for humans, for mice, for non-human primates, for a growing list of organisms, if you have the ability to program this machine to target and edit any genomic sequence of interest, it really opens up the possibilities to almost anything.
Dustin Driver: You were saying before that they used this jellyfish gene that codes for the glowing element of the jellyfish. That’s used in the lab quite a bit, isn’t it, to help actually see things under a microscope, or identify certain areas of cells?
Sam Sternberg: Yeah, GFP got the Nobel Prize, how many years ago now? 4 or 5 years ago. Actually, one of the Nobel Prize laureates for that is here at Columbia, where I just moved to for my own lab. A lot of these other technologies have now dovetailed with CRISPR. For example, CRISPR is used for gene editing, of course, but the possible applications have grown well beyond just making permanent changes to DNA sequences. Now researchers have combined the CRISPR components with this glowing protein to enable a new way of actually tracking specific chromosomes, specific genes, in living cells, where you might want to understand how different regions of the nucleus -- where the genomic information is stored -- actually changes its three-dimensional organization during development, during some process where the cell might be reacting to information. So, you can actually program CRISPR to attach itself to a very specific part of the genome, and as long as you fused CRISPR to this GFP protein, you now actually can see exactly the region where it’s attached, by looking at glowing bits of light under the microscope.
Dustin Driver: That’s genius.
Sam Sternberg: Yeah. So I think for me, as a researcher, what’s been really exciting is to see not only the gene editing technology develop, but all these other decorations of CRISPR, all these other ways that the core, fundamental components, which, in the end, just yield a way of targeting specific regions of the genome, all the different ways that that core functionality can be harnessed for all different kinds of applications.
Dustin Driver: So it’s not just for gene editing, but it’s valuable in all areas of biological research.
Sam Sternberg: That’s right. And you know, I just finished writing a review article on CRISPR/Cas systems and the way they’re being developed for technologies, and it used to be that we would use the term ‘genome engineering’ as a larger umbrella term which would encompass not just gene editing but also things like I just talked about, where you might image different regions of the genome, so just engineering to refer globally to all these different ways of manipulating genetic information. But in fact now, there have been even new developments harnessing different parts of CRISPR systems for things that don’t even fall under the bucket of genome engineering.
Sam Sternberg: For example, in the last couple of years, there’s been more research into the different flavors of CRISPR systems that are found in bacteria, not just those that utilize this Cas9 protein, which is the key protein that is the source of almost all gene editing experiments to date, but other enzymes, other proteins, that other types of bacteria use together with CRISPR, and it turns out that one of these newly discovered enzymes, called Cas13 -- so, the Cas part is the same, the number’s different -- this enzyme turns out to be amazingly useful for what’s called nucleic acid detection, or ways of detecting the presence of specific DNA or RNA sequences. And that’s actually now been developed by Feng Zhang and other researchers at the Broad Institute into a new tool for potentially detecting pathogens in human blood, or detecting particular mutations associated with cancer in some kind of diagnostic approach. And that’s not even genome engineering, that’s now kind of a new point-of-care tool where you could actually use it in the field for epidemiological studies on the spread of viruses or the spread of bacterial pathogens.
Sam Sternberg: I think there’s really no limit, at least in sight, as to the ways that these different molecular components in CRISPR systems can be harnessed in different ways.
Dustin Driver: It’s very exciting. It’s almost as if we’ve been given a bucket of Legos, so to speak. Biological Legos.
Sam Sternberg: Yeah, I love that metaphor. Actually, I brought in a classic line from Max Delbrück. He was a physicist in the early 20th Century who actually moved into biology and made some seminal discoveries in the field of bacterial genetics. He gave a lecture in 1946 where he described studies of bacteria and bacterial viruses as a playground where physicists were starting to join in and play around with these different things, to learn about the source of genetics and molecular genetics. I like that analogy, because I think, as a researcher entering the CRISPR field in the last 10 years, you can just see this amazing diversity in the different enzymes that are involved, the different kinds of RNA molecules, the differences between CRISPR systems and E. coli or Pseudomonas or Streptococcus, and it’s absolutely a playing ground, because you have all these different components that have different kinds of properties, and it’s just a matter of, first, understanding how they work, and then being creative in how you might be able to use them in different ways.
Dustin Driver: I also think it’s a little bit poetic that there was sort of a mechanical view of editing DNA in the early days, or maybe using a virus itself to edit DNA, and lo and behold, nature has a perfect way of doing that, that turns out to be a lot easier than anything that researchers had thought of previously.
Sam Sternberg: One of the lessons I think, absolutely, to take away from the study of CRISPRs, is often following your curiosity, and how nature has solved certain problems, is a much better strategy to uncovering new technologies than to think you might be able to invent it or build it yourself. And that’s been true time and time again in the area of biotechnology. We talked about immune systems and bacteria. It was actually the study of innate immune systems, three or four decades ago, that led to the discovery of a different class of enzymes called restriction enzymes, and those are completely different kinds of enzymes bacteria use to recognize and degrade viral DNA, and those actually paved the way for the entire revolution in recombinant DNA technology, different ways of splicing together and assembling artificial chromosomes that ended up spawning the development of the entire biotech industry, back in the 1970s. So I think, again, that was an example where the study of something fundamental in nature opened up this fertile playground for the development of a completely new technology, not by inventing enzymes from scratch, but by pulling from what nature had already solved itself over the course of evolution.
Dustin Driver: I guess there’s no substitute for millions and millions of years of evolution.
Sam Sternberg: That’s definitely true. You can’t compete with that, really.
Dustin Driver: You can’t compete with it. So now that we have CRISPR, what can be done with it, realistically? I know when I talked to Amy at OHSU, they were able to use CRISPR to, for lack of a better word, repair a single-gene disease that was in an embryo, but what about CRISPR treatments for people who are already sick? I know there seems to be a little bit of confusion about what CRISPR could be used for practically in medicine. Can you talk a little bit about that?
Sam Sternberg: Sure. I think it’s important to distinguish between how CRISPR has been already shown to be successful in the laboratory, versus what I think is still, in reality, an open question, which is how effective it’ll be in clinical trials and in actual patients. I start with a slide often when I give talks. You can pull from the hundreds of papers that have been published, just a small fraction, to appreciate how easy it’s become to repair disease-associated mutations in the laboratory. I would say even Amy’s work at OHSU and showing that in human embryos -- they showed that you can do it in human embryos, but technically speaking, it wasn’t yet done to the point where we actually see a reversal of any symptoms in a patient, and the same thing goes for a lot of the work that’s been done at the proof-of-concept level in cultured human cells, where you can take cells from a patient, you can take cells from a mouse model, you can take engineered cells that have a disease-associated mutation. It’s become fairly trivial to repair those mutations when you’re just growing the cells in culture, in the laboratory. The real challenge in translating that success into the clinic is going to be, for one, tackling this issue of delivery, of how do you now deliver CRISPR not into cells that you’re growing in a petri dish where they’re quite easy to manipulate, but actually access the fraction of cells that you might be seeking to edit in a living patient, where you have the immune system to combat against, where you have other natural barriers for getting genetic material or protein and RNA into cells, and then, frankly, you have the challenge of figuring out how many cells does one need to edit to have a reversal of the symptoms that might be causing a disease. We’re made up of some 30 or 40 trillion cells, there’s no way in hell that you’re going to be able to get CRISPR into even a tiny fraction of that, because we’re just too large. We’re made up of too many cells, and you can edit very many of them. So it’s a matter of going after diseases where you can edit a small number of cells, perhaps in a particular organ or a particular tissue type, and show that that has some consequence on the symptoms that that disease causes.
Sam Sternberg: I think one of the examples where this might be more accessible would be diseases of the blood. For example, a disease like sickle cell or beta thalassemia, where the primary effect is on the red blood cells that ferry oxygen throughout the body -- of course every cell in a patient’s body that suffers from one of those diseases will have the same mutations, but they only really wreak havoc in red blood cells, where the hemoglobin protein -- where that mutation exists, actually carries out its primary function. So even though those patients have that mutated copy of the gene all over their body, if you can edit it in the stem cells or their precursor cells that turn into red blood cells, you can have a massive improvement in their symptoms even though you’ve actually only edited a small proportion of their total cells. So for diseases like that, what researchers are pursuing are actually removing those stem cells, that turn into red blood cells, from the patient, editing them at the hemoglobin gene in vitro, so outside of the body, in the laboratory, and then returning those repaired stem cells back into the patient, where as long as they can engraft in the bone marrow and now serve as a future reservoir of repaired cells, that might be sufficient to alleviate many of the symptoms of the disease.
Dustin Driver: Got it. So an analogy would be, say, the stem cells within the bone marrow are basically red blood cell factories, and they’re broken in patients that have two copies of the sickle cell gene. So they’re able to remove that stem cell, repair it, repair the factory, and put it back into bone marrow, at which point it can start creating regular red blood cells that don’t have any of the same symptoms as sickle cells do.
Sam Sternberg: Yeah, that was much better stated than I did.
Dustin Driver: That’s really interesting, and it gets down to the point I think Amy was making, that there are a lot of these single-gene-mutation diseases that they are studying at OHSU, and there’s just one gene involved, and a lot of genetic diseases are like this, but for many other things there are multiple genes. I think there’s a little confusion about the amount of complexity involved once you have more than one gene involved in the system. Can you talk about what changes when, say, you have a disease where there’s one gene that causes the disease, and there’s another disease that maybe there are two or three or four. The level of complexity when you get into those other diseases makes it almost impossible, right, to even start treating them? Is that fair to say?
Sam Sternberg: It certainly increases the complexity substantially, and then there are other diseases that don’t even have particular mutations that are known -- there are diseases that result from other things beyond just point mutations in individual genes. They might be epigenetic in nature, or they might be defects in certain tissues that are a consequence of genetic factors but no amount of gene editing can address those, because it may happen much earlier in development.
Sam Sternberg: The number of diseases that will be targeted in these early years of gene-editing-based therapies are going to be confined to -- and I put “simplest” in quotes here, even though you won’t be able to see me putting it in quotes -- the “simplest” of diseases that really have a better monogenic, meaning a single mutated gene is at their source, and that it’s known that that mutation and that single gene is causative for the disease, because those are instances where if you know that you have a way to repair that mutation, at least the cells that you’ve edited will now be free of the source of that disease. That’s not to say that making multiplexed changes won’t be possible, and in fact, one of the areas of active investigation at the research level is how we can make CRISPR more effective at carrying out multiple changes simultaneously in a cell. And I’d say that’s where CRISPR has major advantages compared to previous technologies, which were much more difficult to multiplex because of the way that these editing platforms were constructed. I don’t need to go into the details, but the point is, because of the way that CRISPR uses a protein together with a guide RNA, it turns out to be much easier to program CRISPR to target multiple different genes, even within a single cell. Now you deal with some simple math where if you only have a certain efficiency editing one site, let’s say even 90%, which is pretty darn good, if you need to edit both chromosomes for that site, now you have to multiply .9 times .9 to get the percentage of cells that have both changes. And now if you want to talk about adding in a second gene, times two chromosomes, or a third one, the number of cells at a given efficiency of editing per site that have changes at every single chromosome you intended goes down pretty quickly. That’s where I think there are going to be challenges with getting this kind of multiplexed editing at a high enough efficiency where this turns out to really be actionable in a clinical setting.
Dustin Driver: So it becomes much harder to change more than one gene in more than one cell.
Sam Sternberg: Yeah, but of course there are caveats there. If you talk about doing this in stem cells, which can be grown indefinitely outside the laboratory, they’re immortal, now in theory one could make one set of changes, do sequencing to make sure you’ve made the change you want, derive clones from those stem cells where you know you have the cell that has exactly the changes you want, and in theory, you can now sequentially make changes to that same population and clonally derive new cell lines with changes made additively over time. That’s of course very far away from pursuing this in a clinical setting in patients, but if you’re talking about going back to those blood stem cells, you might be able to engineer changes sequentially in certain settings, maybe more in a research setting, but also potentially in a way that might be useful clinically down the road. If you can go to cells that you can culture outside of the patient’s body.
Dustin Driver: It’s inevitable we got into the weeds again, but let’s try to take a step back. I wonder if you can indulge me in a little bit more fun line of thought. I know in science fiction, they really love the idea of being able to use gene editing technology like CRISPR to create these fantastical, lab-grown creatures, and almost view DNA as like we were saying before, the Lego block concept. Is that in any way close to possible?
Sam Sternberg: Well, I have an example which has already been achieved, and it actually was achieved without CRISPR but with the kind of predecessor technology called TALENs. They edited cattle to no longer grow horns. They started with a cattle breed that grows horns naturally, they used TALEN technology -- I don’t know if they used it in fertilized eggs, or if they put it in stem cells and then did cloning, but the point is they used TALENs to edit a single region of the genome, and that converted the cattle into one that no longer grows horns. Now imagine that, that you can use gene editing technology and literally take away an animal’s capability to grow horns. That’s one of the largest macroscopic, fantastical changes you could imagine making in an animal. Could you also do the same thing to have a different animal grow horns, that wouldn’t normally? I think probably that’d be a little more complicated, but the point is, you can make a single change in the genome and the animal will or won’t grow horns as a result of that. If you take that as one of the examples, I think the more that we learn about the genome and how particular regions of the genome affect the resulting traits of an animal, we now have a technology that we can recreate those differences anywhere we’d like, pretty much. Whether that’s growing horns, growing different kinds of hair, the extent of muscle content -- that’s something that’s also been done time and time again in the research setting, editing a gene that’s involved in muscle formation and showing that you can have animals that result from this that have about one and a half times the normal muscle content because of that single edit. I think there are certain traits that we will be able to engineer using a tool like CRISPR, because we know exactly what the genetic basis is for these very macroscopically evident traits.
Dustin Driver: That’s fascinating. So, it actually is not too far from reality to think of an actual genetic library being able to assemble different traits into a single organism.
Sam Sternberg: I think it comes back to that playground analogy. I think the more we have this list of mutations and their effect on animals and plants, now we can go in with CRISPR and start, in theory, messing around with that catalog of mutations, and try to affect a new kind of animal or an animal with new types of traits that they might not have naturally out in the wild. It reminds me of a short story I read a couple years ago while I was researching the book, and I’m going to blank on the author’s name now, but the premise of the short story was some future time where you can go to a breeding specialist and request any custom animal. In this case the couple wanted a pegasus, a horse with wings, and yeah, that was easy for them. Of course we’re not there yet, but if we can start using CRISPR to change whether or not an animal grows horns, what might we be able to do in 50 years or 100 years or 1000 years? I don’t know. It’s tough to know where that kind of powerful technology hits a ceiling.
Dustin Driver: What are the biggest challenges right now? What would you like to see happen, say, in the next 5 years?
Sam Sternberg: I think from a core technology perspective, there’s still a lot to do in terms of perfecting the way that cells repair the cuts that CRISPR makes in the genome. This could get us back into the weeds, but I’ll try to simplify it by just saying, CRISPR cuts DNA, and the cell does the editing, but we don’t have perfect control over how that editing works. So, certain kinds of edits are easy, there are other kinds of edits that are much more challenging and tend to be much less efficient, and I think for a lot of the therapies that people imagine with a tool like CRISPR, the efficiencies of this more difficult type of edit will need to get much higher before we can talk about it being clinically feasible. There’s a lot of work that’s ongoing by some of my colleagues at Columbia and many, many researchers elsewhere to try to better understand the different factors that impact this precise type of repair, and use that information to make it much more efficient.
Sam Sternberg: That’s kind of a challenge with just the core technology. Then I think there’s going to be challenges in the delivery problem. Again, coming back to how do we go from editing cells grown in a petri dish to editing cells in a living patient, that’s going to require advances in how this is delivered into the body, what type of vectors is it delivered in, how do we make sure that’s safe and that it’s not going to provoke some kind of adverse reaction in patients -- and on that topic, there’s been some very recent work that’s not even published in a scientific journal yet, but it’s been released in the form of preprints online, that actually show that the human body in many cases has natural antibodies that are provoked by the very components from CRISPR that are being used in some of these therapeutic developments. The risk here is that the body might actually have natural defences against the CRISPR components that we might be using for medicine, and that’s not a good situation to be in. So, there’s right now ongoing work into really understanding this better, and I think that might be one bottleneck that we didn’t think about a year ago that might create an extra hurdle to getting CRISPR-based therapeutics into the clinic. I guess from your interview with Amy, there’s going to be this ongoing controversy over what are the applications where it’s not just about the technical feasibility but also some of the ethical controversy over what should and shouldn’t researchers and physicians be pursuing in terms of, let’s say, editing human embryos, or editing different kinds of germ cells that would make lasting changes in the human genome. I think that’s going to be a really interesting area to follow over the coming years, how not just scientists but many other stakeholders weigh in on this other controversial area of gene editing technology, and how even the government and regulators will react to some of these new areas that might have been fanciful topics of imagination 10 or 20 years ago but now are very real issues with the new power of CRISPR technology.
Dustin Driver: I can imagine it would be very difficult for two parents if they know there’s technology that exists that could make their children healthier, stronger, faster, smarter, and it becomes almost a slippery slope. Where do you stop in giving your children the best possible future that you can? Ethically, it’s very questionable.
Sam Sternberg: I talk to a fair number of people about this, and you find people on completely opposite ends of the spectrum when it comes to whether or not we should go in this direction. My research couldn’t be further from anything with human embryos, but one of the things I give Jennifer a lot of credit for is that, even being one of the early developers of the technology, she recognized that these are issues that are now provoked by the very tools that we’re studying and developing, and you can’t really walk away from that issue. It requires all of us, whether or not we’re scientists or members of the public, to really come to the table together and be a part of that conversation.
Dustin Driver: So, that being said, what are your current thoughts about editing the human genome to, say, beyond repair, but into ethically ambiguous territory of “improving” the human genome?
Sam Sternberg: I think the technology is not at a point where that would be defensible on safety grounds, because when you talk about enhancements, I think the risk-benefit assessment is very different, because any risk is -- the status quo you’re comparing to is a state of health or a state of being, let’s say, without disease, and so any downside to an intervention at that point would be very different if your point of comparison is a state of disease, where there is a major cost to not pursuing a therapeutic because that’s a state upon which you need to improve for wellbeing. I don’t know if I put that well, but my point being, I think the risk-benefit assessment changes a lot if you’re talking about enhancement, and I think we don’t know nearly enough about the technology to say that it’s safe enough to begin thinking about those kinds of applications.
Sam Sternberg: One of the issues I have a lot of difficulty with, thinking about new therapies coming out, is the fact that even with commercially available therapeutics for genetic diseases that have been recently approved by the FDA, these are very expensive. They cost up to a million dollars. I think when it comes to thinking about where we should be investing our resources, I’d rather see us tackle existing areas of health where we can do a lot more with a lot less than putting our resources into enhancements that would ultimately only be available to the wealthy at this point.
Sam Sternberg: That’s where I would say we are right now. I think it’s really hard to say where society will be in 100 years, and I think when you talk about enhancements and germline editing, I often think, I don’t know what we’re going to see in our lifetime, but it’s a completely different equation when you talk about 100 years or 1000 years in the future, where the technology will be far more advanced, and it’s going to be really hard to predict what’s going to be done then.
Dustin Driver: And society will be much different as well. Society continually evolves. I think it’s also interesting to think about whether or not it would be possible to even stop that sort of thing from taking place. One of my big fears is that, you know, we live in a world where money can pretty much buy anything, and if there’s enough money then maybe these sorts of experiments will happen.
Sam Sternberg: I occasionally get invited to speak at meetings with in vitro fertilization OB-GYNs and people that are working at clinics offering IVF and preimplantation genetic diagnosis and other assisted reproductive technologies, and one thing I’ve taken away from those meetings is there is a major demand in that space in terms of what people are willing to pay. I think that’s the side of this issue that makes me uncomfortable, is thinking about how consumer demand will impact the development of some of these technologies. I don’t know how I feel about it, because there are plenty of things we all benefit from that some people can afford and others can’t. We don’t live in a completely equal society, but somehow there’s something different when we’re talking about making decisions about the genetic composition of a future individual, and how that sets up an entirely new kind of inequality than the kinds of inequalities that we already have today, in the sense that they will have a genome that was to some extent chosen for them or improved upon by their parents, that will last them through their lifetime and the lifetime of any future children that they have. You can imagine this kind of propagating over time if we have now a system where some people can use technologies to change the genetics of their offspring, whereas others can’t. That definitely should give one pause when thinking about that outcome.
Dustin Driver: Definitely, it’s sobering to think it’s amplifying the existing inequalities, like you said. It’s a runaway train that we don’t even want to get started, I think. On the brighter side, though, there’s a lot to this technology in just making sick people healthy, or preventing diseases, just so people can live healthy, fulfilled lives, and normal lives, without being sick. I think that is where all the research is headed, from what I understand now.
Sam Sternberg: Absolutely. And I would add to that, also understanding disease. I think, as big as CRISPR may be for medicine, it’s being used a heck of a lot more in the basic research setting to understand disease by allowing researchers to mimic cancer-causing mutations or disease-associated mutations in cell lines, in animal models, in other model organisms, because it gives us a better insight into the cause and effect relationship between DNA mutations and the resulting properties of a cell. I think even without the use of CRISPR as a therapeutic modality itself, it’s completely changing the way we think about understanding disease, understanding basic biology, and I think the implications from that are going to be potentially as great as the use of CRISPR in medicine itself.
Dustin Driver: Yeah. And I think what we started with, the bioluminescent gene from a jellyfish being used to actually let researchers see how DNA replication is taking place, or see how cells are responding, I think that’s a perfect example of that.
Sam Sternberg: Absolutely.
Dustin Driver: That’s really great. I know you have a busy day ahead of you, and we’re already over time. I just want to thank you so very much for taking the time to speak with me today. I truly appreciate it. I could talk about this stuff all afternoon, and you’re going to go on to think about it all afternoon.
Sam Sternberg: That is true. Who’s next on your interview schedule? What comes after embryo editing, CRISPR tools, what’s next?
Dustin Driver: Artificial intelligence, of course. It’s what’s hot in science and technology now, right?
Sam Sternberg: CRISPR and AI, sure.
Dustin Driver: It’s the most fascinating. I think there might be some overlap there as well, from what I can tell. There’s some interesting theories into actually using AI in order to determine how gene sequences function or how they code for certain proteins. I think there’s a role for AI to play there.
Sam Sternberg: Have there been a couple papers recently using machine learning approaches to try to better model CRISPR guide RNAs? I think Microsoft, some mainly computer tech company, recently published a paper using machine learning in this way. Yeah, I think in biology there’s increasingly more and more instances where advanced algorithms and machine learning are going to really change the way we do science. Cool!
Dustin Driver: I think machine learning is changing the way we’re going to do everything.
Sam Sternberg: That’s probably true.
Dustin Driver: It’s fascinating. It’s exciting. Exciting times.
Sam Sternberg: Well, I look forward to hearing about that once that comes out.
Dustin Driver: Great. Thank you very much, I appreciate it. Have a great afternoon.
Sam Sternberg: Alright Dustin, good talking to you.