KS Community Lecture: Genome Editing Using CRISPR-Cas Systems

thank you all for coming to see you Boulder and for coming to the department and molecular cellular and developmental biology or we say mcdb is short for that and my name's Lee Niswander I'm the chair of the department of mcdb and i'm really happy to welcome you to what we really hope was going to be the first in a series of community lectures here in Boulder along with the Keystone symposia to really highlight important and timely discoveries in the life sciences and so before I tell you a little bit about our department I wanted to highlight this partnership that has been established with the Keystone symposia for molecular molecular and cellular biology and we were really excited when the Keystone folks came and asked whether we wanted to do this public lecture together and then we were definitely blown away when we heard who our first speaker was going to be you're going to hear from one of the the top and most renowned scientist in the world right now and you're gonna hear about discoveries that are really the kind of technological advances that are really going to set us on a whole new trajectory and really is so impactful probably the most impactful of of this century so let me tell you a little bit before we get into that about emcee to be and who we are where we came from and a little preview of where we're going so okay so the building that we're in now Gould Biosciences wasn't actually named after one of our Cu colors it's actually named after professor Larry Gould I'm so Prairie Gould is he's he's still part of mcdb he was here from almost the beginning and he's still an emeritus professor and professor Gould is a fellow of the National Academy of innovators and so during his time here at mcdb he came up with all sorts of new inventions he started a number of companies one of the things that he did was to essentially make a evolution in a test tube in a method that we call cell X and then that's been very important for the pharmaceutical industry and he also developed a number of biotech companies including a sinner gin neck surgeon oxygen and Soma logic which is just down the street so the other main campus that we have I'm sorry the other main building that we have on the campus is the Porter Biosciences building and this is named after dr. Keith Porter who is the founding member of MCD be starting back in 1968 and so that was exactly 50 years ago so we're turning 50 this year we're really excited I hope you guys will all stay tuned we're gonna have a number of celebrations this fall to celebrate our golden anniversary so Keith Porter just to tell you a little bit more about him he pioneered the the science of biological electron microscopy which allowed us to peer deep into cells at a really high magnification unprecedented levels and he was the first to take a photo of a cell at this at this level of resolution on an electron microscope and largely assured in the modern field of cellular biology he designed his own Micra tome in a way this hand cranked machine which could make very very thin tissue sections 80 nanometers and then could put that under this high voltage electron microscope to get images this microscope which was in mt DD mcdb was two stories high and went up only just a few in the in the world so if you remember back to your high school days you might remember hearing about the cellular structure that it's called endoplasmic reticulum or er and so what we see here is mitochondria and another mitochondria here but what you see is all these membrane structures that are studded with these little black dots and those little black dots our ribosomes there are protein machinery and and this is an image from dr. Porter that first about allowed him to see and to name the ER so over the past 50 years MC DB has made many many discoveries and cell and molecular biology life sciences but I wanted to just tell you a little bit of where we are now in MC DB and I'm gonna first come back to this er the sender plasmic reticulum and so in this movie here the ER is highlighted with a green stain and giya Voltas lab has discovered that the ER is essentially like the major trafficking center in the cell and so these little red dots are cargo cellular cargo that's being trafficked along the ER to get to where it needs to go in the cell and the ER as a votes lab also discovered as important in also shaping organelles within the cell and so in this case this red long thing here is actually mitochondria again that powerhouse of the cell and it's being unwrapped by the ER and then actually split in two to make two mitochondria so ray II studies stem cells and how they control growth and development of important tissues such as the skin and the Yi lab is discovering genes that when they go awry can give rise to disastrous consequences such as skin cancer Joel crawl is one of our newer scientists in mcdb and he's been interested in discovered that bacteria which are shown here use electrical activity to communicate with one another a lot like our neurons in our brains and so in this picture this is a group of bacteria that are sending out little electrical charges to one another they're actually being squished by jello and so this is kind of their sense of touch that they are seeing here and giving this voltage and then Zoe Donaldson is our newest faculty member and so she's interested in understanding the neuro bye ecology of complex social behavior and particularly that age-old question of let's love got to do got to do with it and so these little cute guys are prairie voles so I'm like most mammals prairie voles actually mate for life and so she's interested in using the advanced technology to understand the neural circuits and the molecules that are involved in making these guys faithful over over their lifespan and so I hope you'll keep in touch with us and and follow what's happening mcdb we are very excited and now I'd like to turn this over to the Keystone symposia to tell you about the enormous impact that they've had on life science research since they started in the 1970s and I'd like to introduce Dale Jarvis who's going to who's the chief scientific officer at the Keystone symposia to tell you about the mission and the vision of the Keystone symposia Thank You Lee and welcome everyone so Keystone symposia in case you're not aware is a nonprofit organization that's headquartered up in silver Thor in Colorado and like MC mcdb we've been around for a while and we haven't quite hit fifty years but we will in about three years so we were founded in 1972 and our mission is really to catalyze the advancement of biomedical and life sciences by bringing together that key thought leaders in in the life science fields at in-depth scientific conferences and many of our conferences are held in Colorado some of them at Keystone Resort although I will point out that we aren't not in a in a business affiliate of Keystone Resort we are a separate organization a non-profit and the community lecture series is is part of our effort to bring some of these thought leaders into the community and shared some of the really exciting breakthrough discoveries that are happening in the life sciences right now and and give the community a taste of the type of science that that is being discussed at our conferences and we were very very excited to bring dr. Fong Zhang to Boulder today to share some of his work I actually have a long history of working in in Boulder I actually was in this department mcdb I think I remember mcdb celebrating its 25th anniversary I came here as a postdoc in Carla Kierkegaard's lab this particular room I think didn't exist when I was a postdoc but the department has expanded and it's really actually fascinating to see what some of the new faculty are doing here but anyway happy to be back in in Boulder I've been the chief scientific officer stone for about a year and and and where as Lee mentioned we're really hoping that we can make and this community lecture series a regular event bringing it to Boulder so I'd like to introduce Fang Zhang he is a really eminent scientist he's a molecular biologist that's focused on better understanding brain function and trying to seek cures for neurological diseases and psychiatric diseases he has been a real pioneer in the area of genome engineering and actually so he started out I'm getting his undergraduate degree at Harvard University where he got a bachelor's in chemistry and physics in 2004 and then he went on to Stanford University where he worked with Karl Deisseroth group and he actually made some very key contributions to an emerging area called opto genetics and I don't know if we're gonna hear anything about opto genetics today but that's one of his other pioneering areas and that his lab is still working on he got his PhD in 2009 from Stanford and that was in chemistry and biophysics and then the Boston area was quick to snatch him up again he did a short fellowship at Harvard and then went on to the the Broad Institute of Harvard and MIT where he currently as a faculty member so he published a paper in early 2013 in science which is considered one of the one of the really key papers in precision genome engineering it was the first successful work in in genome editing and mammalian cells and that it is one of the most cited papers in genome editing his lab has has since gone on to further refine that methodology and that the 2013 paper was on CRISPR cast cast nine technology he's since gone to develop a technology that they're calling crisper CF CPF 1 which is a more refined methodology for editing DNA and and even more recently his lab has worked on a system called c2 c2 which is for editing RNA so it's it's a great pleasure to welcome him for this lecture as I mentioned he's a core member of the Broad Institute of MIT and Harvard he's also an investigator in the McGovern Institute of brain research at MIT he's the James and Patricia portress professor of Neuroscience at MIT he's an associate professor in the department's of brain and cognitive sciences and biological engineering at MIT and he's also a Robertson investigator at the New York stem-cell foundation so it is my great pleasure to welcome function first of all thank you so for the kind introduction but but even more so thank you for coming out here on Sunday afternoon you must have better things to do than that sitting here and listening to my talk it's really a great pleasure to be here and I wanted to share with you some of the work that we have been working on in developing genome editing technologies to be able to make very defined changes in the DNA of ourselves there are a lot of different and potential applications that range from advancing research to developing therapeutics to engineering crops so that they are drought resistant or virus resistant so that we can improve agricultural yield and also I will tell you about some of the ongoing activities of developing new technologies that have applications for Diagnostics and also continued development of therapeutic strategies for treating diseases so we'll get started with this slide here so over on the right this is the plot from the National Institute of Health and what you see here on the y-axis are a number of papers that have been published year by year over the past decade each paper here and reports a series of discoveries that link genetic differences to specific diseases now this is really still at a very tip of iceberg because as we start a sequence more and more human genomes will be able to identify many more genetic differences that are linked to specific disease so some of these genetic differences play what we call a causal role in the disease so that means if someone has a mutation then they will have a specific form of illness and so of those causative genetic mutations there are about a thousand that have been identified to date and so if you think about that if we know what mutation caused disease then you might ask the question why don't we just go into ourselves and get rid of them if we know what is a deleterious mutation let's go into the cells change it so that we can reverse it back to the healthy DNA sequence it turns out that to do that is fairly challenging that's one of the challenges that both my lab and also the lives around the world I have been working on to try to advance so the simple idea of how to go about changing DNA is here in this picture so before I explain what to you what this picture means the human genome has three billion letters long and so and we have or have twice at that so we actually have six billion letters because there are two copies of each DNA in our cell and so a mutation in the DNA can happen anywhere in this 3 billion letters and so if you think of the genome as a document of 3 billion letters then we need a way to be able to search in this document find where that the mutation is that that typo in the document and then be able to reverse it so that we can correct it and and restored the the healthy function of the cell how do we do that how do I search for it turns out Nature has invented a number of different ways to be able to search along pieces of DNA and these systems are called DNA binding proteins they by the name you you already see that their function is to interact bind to a specific DNA sequence so the basic idea for doing genome editing is is this if we can have a DNA binding protein a DNA binding domain that we can reprogram customize it to recognize search for a specific DNA sequence and if we link it to a protein called a nucleus then we can achieve genome editing what these nucleases do is that they can introduce a double-stranded break a cut in the DNA and so more than 15 years ago gene Haber and also Maria Jason two eminent scientists in the field of DNA repair discovered that whenever there is a DNA break in the genome that break triggers DNA repair machinery and allows the cell to repair that that DNA break and with much higher levels efficiency and so a lot of the work have really focused on how do we reprogram this DNA binding domain with significant ease so that we can for any sequence in the genome quickly and efficiently develop a DNA binding protein to find that sequence in the DNA and to be able to manipulate it so some of the earlier technologies to be able to search in the genome include seeing finger proteins and also tail proteins these are protein based DNA binding domains and what that means is for a thing finger protein there are different modules each one of these coloured sort of squarish things is one module for seeing finger protein each one of these naturally recognizes 3 DNA bases and you can mutate these proteins to get them to recognize a different set of 3 DNA letters when you do that then you can link them together to form what we call a seeing finger array so it's simply a linked a tethered system with multiple fingers and if each one recognizes 3 bases then a zinc finger rake had recognized a multiple of 3 bases based on a number of modules you have in it so for a 3 modules in finger you can recognize 90 na bases and for a 6 Marjorie can recognize 18 at DNA bases but even though this paper looks remarkably simple on paper it turned out is actually not so easy to use and why is that and that's because if we want to change the sequence that this module recognizes we'll have to change the protein and the way that a protein falls into a specific shape to be able to recognize that DNA depends on the interaction of many of the residues so when you change one it affects many other ones that this this module makes contact with and so the problems are becomes exponentially harder because we need to tweak one it changes other things and so it's not very predictable and so as a result to build a new missing finger protein that works well it can take several weeks to months and even then you are not guaranteed to develop a new that zinc finger that works well so a system that's a little easier than seeing finger is this thao protein here is simpler because each module recognizes a single DNA base and so that means you only need four modules rather than having to have 64 modules so this system is easier but still is hard because the way that each each mod each module recognizes the base is affected by the neighboring modules and also it's just a cumbersome process to be the engineer one of these proteins so we were working on both of these systems but because they were challenging I thought can we come up with a simpler system and turned out that Nature has him in fact invented other ways to be able to target DNA with remarkable ease and that system is something that's called CRISPR CRISPR stands for a clustered regularly interspaced short palindromic repeat but I would just call it CRISPR otherwise the talk will take forever that's a finish so what CRISPR is is an immune system in bacterial cells just like we have our only immune system our immune system uses proteins the bodies to recognize invading viruses or bacteria or other things Christopher uses RNA to recognize invading viruses so if you think of this gray box as a bacterial cell when a virus infects the bacterial cell it injects his own genetic information usually in the form of DNA or RNA and into that back to yourself so once the virus has injected itself into the bacteria then this system called CRISPR which is encoded in the genome of that bacteria gets activated and there are genes that encode CRISPR associated proteins which are called cast genes that get turned on and they start to become at these CRISPR proteins and these proteins working together with RNA that are derived from this repeat element called CRISPR array acts as the surveillance complex for recognizing that invading virus this so here is one example off of one of these surveillances there's a protein called cast 9 and when cast 9 works with a short RNA and that we called a single guide RNA and and so this single guide RNA is an engineered form of these natural and crisp RNAs to be able to recognize at the invading a virus DNA and so wherever this RNAi matches with the DNA so when blue one the blue virus DNA matches with the red guide RNA then the casting enzyme will make a double-stranded cut in the DNA so that it can inactivate the virus DNA but this same mechanism of searching for DNA sequence and then making a cut is exactly what we could and we may be able to use for doing genome editing and so what we have been working on then is trying to see can we harness the system engineers so that we can use it to edit specific sequences in human DNA so that we can achieve genome editing so I got a I got started with this work when I first read Sylvania my newspaper back in 2010 where he reported that wherever you have an RNA sequence which is highlighted here in green parrying the target DNA sequence which is in black if there is perfect recognition indicated by these vertical black bars then the Cassadine enzyme will make a double-stranded cut and within the DNA so exactly right here and in this double-stranded cut is is the end of that reaction it this enzyme doesn't go and start to make more cuts in the virus in the virus or in the plasmid DNA and so this is a very nice feature because this shows that you can program Cassadine with this RNA sequence to make a precise double strand break and that's suggesting if we can harness the system and make it so that we can program this enzyme to recognize a specific human sequence then we can go and and cleave at a specific location in the genome and then stimulate DNA gene editing and so working with a number of graduate students including law and also en we successfully introduced the Cassadine enzyme as well as engineered and the guide RNA so that you can recognize the target DNA in the human genome and then introduce the double strand break and when that happens they initiates the repair processes so that we can go in and either delete sequences or incorporate a new sequences in the genome so subsequent to that we then also working with additional members of the lab starting to interrogate the specificity the systems how precisely can we use this enzyme to be able to edit a specific sequence in the genome and also at the same time developing computational tools that allowed researchers around the world to be able to easily design guide RNA sequences so that they can target at their respective genes of interest so so I've said a lot but a picture is or a video is worth a thousand words so here is a video that shows you how the CRISPR counseling system works so in blue here this is the katadyn protein and then in red is the RNA that we programmed to recognize the specific sequence in the human genome so by introducing this complex through the human nucleus membrane and then getting it to recognize specific sequence in the human genome when the RNA when the protein opens up the DNA which is in blue and then when the RNA base pairs with that DNA to achieve recognition then caste name makes a double-stranded break so that we can stimulate at the inner repair so there are two ways to repair the DNA one of the ways simply rejoins the two ends of DNA together but the process of achieving this introduces a small mutation and this is quite useful for in activating a gene of interest sometimes our genes that are deleterious and we can use that to inactivate it the second way to repair DNA is more precise so here you have the two broken DNA ends you can provide the cell with a new piece of DNA that is similar to the two broken ends but carrying a new sequence that you want to introduce into the DNA and this could be a sequence that you want to swap into the genome to replace at the mutation and and the repair machinery is we incorporate this directly into the genome and so so this video shows that there there are two ways that we can go and edit the genome the first way is to make a cut allow the cell's reglue the ends together to make a in activating mutation that gets rid gets rid of a deleterious gene and second you can give it a template DNA and that allows you to switch in a specific sequence of interest in the genome so that's how we go about doing genome editing so in addition to - to edit in the genome turns out that these systems also have additional applications and and and so the first application that the video describes is using it as as the nucleus to cleave DNA and then stimulate DNA repair but then the second class of application is using this cassadine protein as a way to bind to specific DNA sequences so that we can bring other functionalities to a specific location on the DNA so that we can start to manipulate it and this is done by getting rid of these two domains on the Castlight protein that naturally are responsible for cleaving DNA so if we get rid of them then Cassadine will simply bind to DNA and not cleave it and then we can use this binding protein to bring effectors like things that can turn on transcription which allow us to turn a gene on or or turn off transcription by closing the chromatin condensing the DNA or we can bring in a fluorescent protein like GFP so that we can visualize the configuration of a specific location in the genome so so that we can track its motions as the cell is undergoing some type of biological function and there are many other ways to use at this basic system and these are all ways that biologists now are using to advance our understanding of different biological processes so following those initial work to develop the CRISPR Cassadine system for gene editing and we thought to further enhance the system and so one of the really exciting potential application of chrisberg has 9 is to use it as a therapy if we can deliver that into cells in specific tissue correct genetic mutations and then achieve a therapeutic outcome so that we can treat disease now one of the hurdles for for enabling that that goal is how do we deliver the Cassadine protein into the right tissue into the right cell efficiently so one of the challenge that the faces delivery is that the the Cassadine that we initially developed is quite large it's hard to fit it into delivery systems that we can use for delivery in to say cells in in the mouse and so working with members of my lab we examine the diversity of casts 9 so it turns out that cache 9 is a bacterial protein and many bacterial cells carried their own version of cats 9 and they come in all shapes and sizes some are larger some are smaller and the first one that we developed was pretty large is from the streptococcus pyogenes bacteria and then we also developed a smaller one from streptococcus thermophilus but that enzyme is not as efficient as the s pyogenes cast night so I thought maybe let's examine at this group over here the smaller bundle of cast nine some of them are smaller and if we can find one that works well and then we can more efficiently achieve gene editing in in animal models and so we examine a number of these different small cast nines and we eventually found a specific one and that's the sa cast night so sa stands for Staphylococcus aureus which is another human microbiome bacteria and so by taking this small cast night we can introduce it into a viral vector viral vectors are what we call gene delivery systems that people use for gene therapy for transferring genetic information into specific human cells so we can package the genetic information with cast night into this virus along with the RNA guide and that we use to target a specific gene of interest and so this allows us to produce viruses deliver cast 9 into a mouse model and then see can we achieve efficient or sufficient level of gene editing to achieve a therapeutic outcome so to test it we design a guide to target the gene pcsk9 pcsk9 is a it's a prominent therapeutic target because there's a small population of individuals who naturally don't have pcsk9 and they live long and healthy lives and and on top of that they have very low levels of cholesterol so they don't have high levels of heart disease risk and so drug companies have been developing either small molecules so chemicals or proteins to be able to deplete pcsk9 from human patients who have high risk for heart disease by reducing their question level and so we thought if okay newscast 9 guided by a RNA that targets the pcsk9 gene to e activate pcsk9 permanently then we may be of to achieve a long-term treatment for cardiovascular disease by mimicking this national mutation that's found in some individuals that have low heart disease risk so here is what a data looks like so the left panel shows gene-editing as measured in the mouse that we we injected and so these are different days after we injected the virus into this mouse and you can see we achieve about 50% editing in the liver cells of these mice and with hardly liver because that's where this pcsk9 protein is produced and so we were primarily interested in editing and hepatocyte in in the human liver so once we have that and then we thought if we're getting 50% editing that's pretty good does this translate into some kind of decrease in the pcsk9 protein level so here this plot will measure the amount of pcsk9 in the serum of the of these mice and what we find is that corresponding to the timeline here seven days after the injection pcsk9 became undetectable in the in the blood of these mice and comparing this with the control which is here in the grey line you can see that there is a very drastic decrease in the amount of pcsk9 in these animals so if you reduce pcsk9 then we should also be reducing a cholesterol level in in the mouse and so that's what the right figure shows this is measuring the amount of rest 'real in the serum of these mice and now what we see is also after seven days all of the experimental mice that received the gene editing reagent reduce cholesterol by 50% compared to the control animals and so when you can reduce cholesterol by 50% that's a significant advantage in terms of susceptibility to to cardiovascular disease so this shows you know it's promising but there are still a number of things that we need to do to continue to develop this system so that we can have an efficient and also precise and also safe gene editing to be able to achieve therapeutic outcome in patients so a lot of work is going on now to one achieve more efficient delivery and to making sure that a system it's safe and efficient so that when we injected into a human patient and we can achieve the desired outcome so one of the things that we have realized as we were developing the Cassadine technology is that there likely will be many other technologies out there that Nature has invented and if we can go in harness the natural diversity that exists in the bacteria that live around us we may be able to find other interesting systems that are suitable for advancing human health and so one thing that we have been doing for the past few years so explore bacterial diversity all these all these these these different microbes that live around us in our body in our environment and also even in extreme environments and that we don't we don't Olivia and and so some of the work we've been working on in this front has been to see are there additional crisp crisper cast systems that we can harness and to be able to advance the gene editing a technology and so what we did is we first examined the way that CRISPR Cassadine system are are present in bacterial cells turns out that there are two major classes of CRISPR systems there are the class one systems that use many different protein modules that work together to form a big complicated machine to perform the recognition and destruction of invading viruses in bacterial cells and then there are the simpler systems which are the class to crisper systems these have a single protein which is a much simpler machine to be able to recognize the the invading virus and also destroy it at the time of our study there was there were many different class one systems many complicated systems that scientists have discovered but only one simple system and that's the caste lysosome and that had been reported so we thought if there are many class one systems who's to say that nature didn't also invent additional simple class 2 systems to be able to provide by thorough defense so working with our colleague Eugene Koonin at NIH and also members my lab and we set out to perform a computational investigation to see can we identify new novel simple CRISPR systems that were maybe the harness and for achieving gene editing and so by developing a computational pipeline that uses some of the most well conserved elements of a CRISPR system to be able to identify other CRISPR systems that people haven't yet identified within were able to derive a number of candidate genes that that we followed up and to develop biotech applications and here is a summary of some of these CRISPR systems that have been identified today chiefly there are three major groups of CRISPR or single component CRISPR systems there are the cast 9s which are the gene editing system that I told you about and then they are a cast 12 which are also DNA editing enzymes these are systems that include a CPF one and and most interest interestingly in addition to these two DNA targeting systems and we also found that there are RNA targeting systems that are part of this a caste 13 family of crisper systems so so without telling you more about the DNA systems our focus probably the rest of my talk on on this RNA system and that's called cast at 13 it was previously called c2 c2 but just to make sure that everything lives in a happy crisper family would renamed it Tuvok has 13 so to figure out what castor teen does we took the the castor teen system from a bacteria that's pretty challenging to study and transferred it into e.coli we like e coli because it's easy to manipulate in the laboratory and we'll have a lot of tools that'll to study it and so once we transfer this into e.coli with indeed RNA sequencing to make sure to see whether or not these crispr RNA guys are expressed and we found that indeed they are which is very good sign it suggests that this is the system even after transplanting into into e.coli is likely still functional so based on the hypothesis that this is the RNA targeting and protein and within set up a screen to see can we validate that this is the RNA targeting system and - can we learn what is the rule that the system uses to be able to search for RNA and also target it and and so to test this we chose a bacterial virus called ms - phage and this is the virus that only exists in RNA form and so what shows it specifically because if we chose a virus that also exists in DNA form then when when this cancer gene system shows the offense of that virus we don't know whether or not it defended the DNA version of that virus or or a defendant against the RNA version of that virus but if we choose a RNA only form then if there is any defense activity that we observe then we know is defending against RNA and that means cancer teen is targeting RNA and not DNA and so the way we do this is we designed RNA guys that tile the entire length of this RNA virus genome and that way we can cover the entire virus we didn't know what are the rules that govern which sequence to target which one not which one to be not horrible so so by dialing that we basically saturate the entire virus so that something should work if it does work and so we synthesized all these RNA guys using inkjet a little printing technology and then cloned it into a library of DNA so we can introduce that into e.coli and then generate a library of e.coli where each cell carries a different guy sequence targeting a different position on this virus so now by by exposing this group of bacterial cells to this virus if any of these RNA guys was able to defend the virus the bacteria from the virus then that cell was arrived and so by looking at the surviving cells and sequencing and the guide that is left over in those arriving cells and then we can start to infer the rules that govern targeting of different virus sequences so what we found is that indeed some cells were able to survive which shows that it is able to target RNA and to by it by looking at all of the the RNA guys that provided a defense function we found that there is a simple rule that governs targeting which is that right next to the site on the are and the virus that's targeted it just needs to be flanked by a single base and that's either a C or a or a U and that makes the system quite easy to design and to be the target in in in bacterial cells so that's targeting the targeted sequence on the virus and what we found is that in addition to targeting the targeted RNA there's also a very interesting property we call the collateral activity and what this activity is is that if in a reaction you have the RNA that you're trying to target but also the guide RNA which recognizes that targeted RNA if both of these pieces are present then the cancer gene enzyme will get turn on so that it it will go and cleave any other RNA that's within the same reaction and so even if you have another piece of RNA that is not at all recognizable by the guide RNA as long as this recognition happened then this RNA will also get destroyed so in in bacterial in the life of bacteria we think that this provides a very unique advantage for for defending against very strong a pressure is an infection but biochemically this suggests that cancer teen operates in a different way as the DNA targeting systems because in addition to cleaving the target it also remains active and cleaves other RNAs in in the in the same reaction and so from this gel at the first a lanes simply show that as long as the target the RNA is present then that the RNA that is not being targeted can also get cleaved which are these different fragments here but if you don't give the target RNA so only the collateral RNA and also the guide RNA then there's no cleavage and this shows that it's a recognition triggered RNAs activity that goes and chews everything else so you may ask how does this fit in the overall context of bacterial defense and and this summarizes our understanding today so with the the C 2 C 2 or cancer gene system what we think is that there are two modes of immunity that this bacterial CRISPR system is able to provide the first mode of immunity is the standard immunity which is when the virus infects the bacterial cell cancer gene gets activated and destroys that virus and then a bacterial cell can live on happily ever after but if the virus is somehow able to evade the the initial immunity and starts to replicate itself then we're in a much more dangerous zone because if that virus is able to replicate and and spread to the rest of the population then you end up with an exponentially amplifying infection activity and so what cancer team does is that if it recognizes that somehow Eddy evaded the immunity defense then it will trigger what we call program cell death and that's this RNAi triggered recognition triggered destruction function of all the other RNA so that it can destroy all of the virus RNA through program cell death before this virus gets released into the community so so that's single bacterial cell that gets infected destroys itself and protects the populations quite altruistic for for this system to behave so this is all very interesting microbiology and you may be sitting there and asking what does this have to do with with you know improving human health and how can we use it and to make our lives better turns out that there are there several different ways to apply the system and the first application is that we can use the the collateral activity of castor teen to be able to develop a highly sensitive and also highly easy-to-use diagnostics a platform for either pathogenic viruses and bacteria or for detecting cancer at DNA so the way this works is is a system that we call Sherlock that basically takes as input biological samples usually in the form of urine saliva or or blood and we can extract DNA or RNA from this biological sample by performing p7 transcription we can convert either the DNA input or to RNA input into a big batch of RNA and then we can program cast thirteen to recognize the virus or the bacteria we're trying to detect so say we're trying to develop a Zika virus or influenza virus detection assay was simply program the cancer t enzyme with the RNA guide that recognizes that the influenza or the Zika virus sequence and by incorporating this cancer T enzyme and the program guide along with a reporter and that's just a short RNA with a Quinn sure and a fluorophore attached together so when they are attached when this RNA is in fact the quencher quenches the fluorescence so you don't see a color from this reaction but if the Cassidy molecule along with the guide that programs it would recognize the influenza virus or or the Zika virus when that recognizes the virus sequence then it will trigger the cancer T enzyme to become activated so that it can go and cleave these reporter RNA as well and so that's when you get the quencher released from the fluorophore and you get a fluorescent a signal from this reaction so using this you can get several orders of magnitude amplification you get amplification here and also amplification there so that we can't detect molecules with with sensitivity as low as one molecule per microliter which is the level of sensitivity you need to visit we've had early stages of virus infection so so this is how how the system works turns out that the cancer gene enzyme is very robust you can freeze dry this onto a piece of filter paper and then and then you can apply the biological sample onto it and then read out the fluorescence and so using this freeze-dried format and we show that you can detect as low as about 20 animals and so that's about 10 molecules per microliter of the of the material they're trying to detect and and so so this is already a much more sensitive and then say the over-the-counter HIV detection kits that you can buy from the store so that was for detecting viruses you can also detect a DNA and so over here with these two different plots what they show is that you can easily program this system with a guide that recognizes ecoli sorry by giving an eco-lodge genomic DNA or give me a Pseudomonas genomic DNA and you can program very specifically a guide that recognizes it with a ecoli sequence but if you give it guys I recognize other bacterial cells you you will not get a signal and this just shows that you can very accurately and also easily program the system to recognize different infectious material so that was all done using fluorescence based readout but it would be really nice if we can also have a colorimetric readout so that we can use this in the field so that we don't have a fancy for ammeter and too we have to read whether or not the virus is present and so along that direction we've been developing a lateral flow based method and where you can apply the sample to a paper strip very much like a pregnancy test strip and you can apply the simple let it flow through the strip and if we're trying to find his present you will get a new band a here in this paper strip and so so this for this latter float has we developed using the cancer T in a system and we can get a pretty good detection of Zika virus RNA and so these are different strips where we apply it and seek a virus sample and you can see we can get a pretty sensitive detection using this paper based at so so so using this and further developing the sensitivity of this lateral flow system and then it's possible to have over the counter or even at home Diagnostics applications for variety of different diseases this would be especially useful because once a person suspects that / that he or she has contracted an infectious disease the last thing you want that person to do is to go out into the subway travel to the hospital spread a virus every word that person travels is much better if the person can just tested at home and if that is the case call 9-1-1 and then apply quarantine procedures so that it so that we don't have a public health hazard in this case so that's something that we're working on to be able to to make it easier to be able to detect diseases so the second application of using this case routine enzyme is to use it to edit RNA as earlier I told you about editing DNA but turns out that there are a lot of benefits to be able to edit RNA and so one of the benefit is that the the DNA nucleus base genome editing is is not very efficient in cells that don't replicate in fact many of the cells in our body no longer replicate if they were replicating we will be growing taller intolerant bigger and bigger and so to have effective therapeutic efficiency we need to overcome this challenge and this is one thing that anything RNAi allows us to overcome the second thing is that RNA editing provides reversible modification of DNA sorry a reversible modification of the cellular property one of the one of the one of the concerns about DNA editing is that if it does go and make target edit so something that we don't want to edit those DNA edits are permanent but when we edit RNA we can overcome that particular consideration so RNA editing naturally occurs in in cells but what doesn't happen naturally is that it's hard to reprogram it to edit other RNA sequences and so by developing cancer 13 we now have a easy homing device that we can link to RNA editing complex so that we can direct it to write to edit any RNA sequence we want and and so the way we do this is we begin with a cast 13 protein just like how researchers inactivated the cast 9 nucleases so that it only binds to DNA we in activate the rnase domain so that it will bind to RNA but not cleave RNA and and so once we have this we have a RNA binding or RNA homing device that we can use to bring RNA editing enzymes to a specific location and so this is what we did we have this inactive or dead cancer teen and we use it to bring the RNA editing enzyme Adar to a specific sequence and then we can program which RNA base to change by missed pairing that RNA base with a particular cytosine on the guide RNA and you can achieve very precise adenosine to inosine changes so in essenes are read out by ribosome as guanosine so in fact people are making a through g changes in the cell and so this is just a more technical diagram showing you that if you have an RNA on top that you're trying to change this particular Dana zine you can design an RNA guide where you miss paired adenosine with the cytosine and that specifies which one to change and so when you do that even though there are other Edina scenes that are also bound by casts routine because they're not indicated with the mismatch that those are not edited at the same time and so it has to the system there are two different Adah are proteins in the human proteome there's eight are 108 are two both of them can mediate editing eight are two is a bit more active in so we chose to focus on a tower 2 and and so one of the really important things to look at is how specific is this RNA editing system and and so this yellow dot is what we want to edit but unfortunately when we sequence the whole transcriptome so every single are all the RNA is in the cell will find that there are whole bunch of other sites that are also edited which is very disappointing but we thought to figure out what's actually happening and we did that by performing a control experiment where we have a guide that doesn't target what we're trying to target and when we did that we found that many of the top many if not most of the off-target overlapped with off target sites with the targeting guide and that suggests is probably not these off targets are probably not indicated by the RNA guide but rather probably by the ADA enzyme or the Cassidy enzyme itself and so we thought based on this understanding we can maybe engineer this system and to make him more precise so this is repair v1 and you see there are about about 2,000 off target sites and so by looking at the crystal structure which was solved by Peter Beals group at UC Davis we were able to identify different residues that contact the RNA and then by mutating these different residues we thought maybe we can increase the specificity of this RNA editing system and we tested a large number of different mutations some of them like the t3 75g seem to dramatically increase the specificity that this is measured using a specificity so as a physics core and so I thought this is interesting maybe we can test this and to see does it make the RNA editing system a lot more specific so to test this we we did this experiment we're using the repair v1 you not only see editing of the targeted base but you also see editing in other location within the target site but when we are engineered with the t-3 75g mutation to make the repair v2 system we are now able to significantly clean up the editing event so you only see that targeted base edited but but not the other non targeted basis and then we are did more experiment so now we sequence the whole transcriptome at 125 times of coverage this is much higher than what people do typically for RNA sequencing and so now because we sequence are higher and we now see about $20,000 get sites with a repair view one system and then just for comparison this is the repair v2 system and you can see now we have 125 X reduction in the number of our target sites and so working on this working in this direction and we can make the system a lot more precise and one thing we're working on now is trying to bring this on target activity back up so that it has the same level activity as we probably want so decreasing our target increasing specificity increasing activity so so we're continuing to develop these gene editing systems there are a lot of things to do how do we increase specificity and efficiency how do we deliver this effectively to different organs so that we can get closer to developing them as therapeutics and then can we make other forms of RNA based changes not just a - I but I - G - G - a or C - you you to see all sorts of other changes and then and then evaluate this how does it work how well does it work in animal models so so that's what I have prepared for for the data today but but one thing that were continuing to look at is to continue in mind the bacterial and natural diversity that's around us this is a plot that shows the number of genomes that have been completed by 300 genomes have been completely sequenced and so what this really highlight is that what we looking at now is likely only the tip of iceberg there probably a lot of other interesting systems that remain this to be the waiters that there are waiting for us to discover and and as we study then we'll probably be able to develop more powerful by bio biological tools so our continuing to develop these technologies by the same time as the technologies become more mature we're now starting to apply them to both study and also develop ways that we may be able to treat neurological diseases including Alzheimer's and also a lot of the psychiatric diseases so that's all sort of this is really the most important slide on thank my group and mit umbrella who worked with me and also our collaborators and also the funding agencies that support our work over the past seven eight years thank you so much for for your attention [Applause] I'm happy to answer any questions you may have so create your 13 system is do you know what's actually structurally happening to the enzyme to make it a nonspecific RNA to greater and then is there on the flip side an off switch that you'll see to return to homeostasis very good question so so the question really is you know what is happening why is caster teen able to stay on and cleave other things while maybe cast an eye and see pf1 don't so I think part of the answer has to do with the way it Cleaves the target when caster teen Cleaves the target when cancer teen recognizes a piece of RNA and Cleaves it it doesn't cleave in the recognition site so it Cleaves outside the recognition site and so what that means is the guide RNA and the target RNA binding that duplex remains intact and that's what cancer he needs to be able to stay in this active conformation to be able to act his RNAs in contrast cast 9 or cpf 1 they cleave in the recognition site and so that means upon cleavage then the target is is destroyed and and so because it's destroyed this duplex that activates the protein into a DNA s is no longer there so you get silenced molecularly we don't have that how make structure for cancer teen yet so we don't really know you know how are the domains moving together to form this RNAs but we're working on that and hopefully we'll we'll have the detailed atomic structure soon you had a question so just last week I was reading that the some of the Americans who want to do clinic clinical trials are just beginning to get on down the pathway of getting approval but China has been at this for about 15 since 2015 do we know anything about any of the outcomes of those clinical trials that have been going on for a few years yeah that's a good question I have the same question actually so so so so there are a number of clinical trials going on around the world there's one happening in China there's a lot of work happening in the u.s. at UPenn in various companies focused on developing CRISPR base therapeutics and and so forth unfortunately we don't have very much information about what has been done and and what the readouts are from the trials happening in China what I know based on reading is that they are developing some type of cancer therapeutic using CRISPR to engineer immune cells so that you can put those engineering new cells back into cancer patients to be able to target lung cancer that's one of the trials that I that I read about in in the publish press beyond that I don't think they have released any data in the u.s. there are a number of groups working on developing CRISPR based therapies also for treating cancer probably in probably mostly in lymphoma for cancer but also for sickle cell disease for blindness and also for some metabolic diseases that having the liver I think there's a lot of progress being made and and different groups are filing for clinical trial applications in Europe or in the US so hopefully in the next year or so and we'll start to get some information about how well these technologies are working from from a clinical perspective yeah yeah so one of the things that we have been working on is to use castor teen for liquid biopsy so liquid biopsy is where you simply take blood from a patient and and by reading or detecting the cell-free DNA or RNA that's circulating in the blood you can make a determination of the disease State so we've been so a lot of groups are working out yeah these cell free DNA for for cancer to monitor and also track the progression of tumor development in the body we have been working on trying to do about castor teen as a Diagnostics method to read to read out are there to tumorigenic DNA that's you know in in the cell-free DNA fraction and and can we achieve quantitative readout so that we can not only detect it but also measure it and track the the growth of the tumor that's one of the applications that you might be able use cancer Tina for I have a couple questions over the first one is looking through the microbial diversity what else do you think you might find so for instance do you think you might find protein editing systems because they're proteinaceous infectious particles like prions right yeah so so that's one question and then the kind of related when you did the bioinformatics analysis what features about the other bacteria suggested that it was gonna edit RNA right really great questions so the first question is if you look into this microbial well what else can you find I think you can't potentially find anything you know nature is pretty amazing and Nature has had a long time to innovate and evolve and and refine a function of all sorts of proteins take them apart put them together miss mix and match domains from different proteins to form new proteins of new function I think it's very likely that we'll find you know other proteins that may have interesting biotechnology application beyond just DNA or RNA cutting enzymes there may be protein binding enzymes there may be protein cutting enzymes there may be sugar binding enzymes there may be sugar cutting enzymes things that can synthesize complicated chains of sugar molecules for cell surface recognition or communication things that neighbourhood lipid I think I think it's really a rich well that that we need to continue to explore the second question and and just and and one more point about the first question is that and what is really exciting is that based on these hypotheses of what might be a protein that can perform a desired function you can then start to mind the sequence databases look at all of the bacteria back through species at me a sequence look at their protein sequences and then see does any of them conform to the hypothesis and if it does then then your you can take it to the next step and start a study it so the second question is one more searching with CRISPR assistance what are the features that told us that it may be a RNAi targeting enzyme so so that's when we found these proteins actually this whole family of proteins that are fairly well conserved we then examine these proteins members of the same family to see what are the different domains that are most conserved if there are if within a whole family protein there is a domain that's highly conserved then that domain suggests a common function that's you know preserves throughout the whole family the whole family and so two of the domains that are highly conserved are the happen domains it stands for a higher neutrality eukaryotic nucleus domain and they are normally found in rnases or bacterial toxins and that's what toda is that this might be RNA talking the enzyme so then based on that hypothesis we can design a screen using the ms2 RNA virus and see that's it as a target are turning yeah you sort of touched on this in the in the beginnings I was just so for like diseases that kind of have that are undetectable early in life how long does it take for those effects to be reversed can you give me example of what these easy might be thinking about be like a later stage progeria that doesn't really affect people the early on in life but maybe later down the line yeah so this so I know about progeria but I don't know that much about I don't know a lot about late stage for Korea is that caused by the lemon okay so a lot of that I think the best way to do it is probably to t-type I do you by sequencing the genome or running a genetic panel to see that's a person carry that mutation that's probably the the best way to go in terms of how quickly say you edited the mutated gene the patient how long that will reverse the phenotype for progeria I don't know I think because it's a so well I guess we can pass the animal models to see is it a reversible phenotype some some genetic diseases may be reversible others may not be as easily reversible so I think you'll probably different case-by-case and some diseases can probably reverse much more quickly for a blindness disorder where there's a mutation that affects photo reception you might be able to break inside after you know enough protein has been produced enough for the edited you know protein from the edited DNA has been produced so you defend you know differ from diseases disease but some can be very quick the the pcsk9 example you saw the cholesterol level dropped seven days after the injection and and it just stayed there in the mouse yeah sorry I think that's a long-winded answer for yet for your collateral damage with the cast 13 our structured RNA is less susceptible to the cleavage so like a tRNA or though or are they similar that's a really good question we don't we don't have a very solid answer to that yet secondary structure does seem to to some extent affect cancer teen activity so if it really has a very strongly folded hairpin or something it may be less accessible but we also found is that excess susceptibility to to secondary structure varies from different orthologs if for different orthologues with cancer teen some cancer teens some members of cancer teen family from some bacteria are able to target a lot more broadly while others are much more sensitive so I think there will be a arranged and and that's interesting yeah so you had a question in there that's a good question so so there are so gene editing technology has been moving very very rapidly so there are two challenges that face mitochondrial gene editing I think probably actually probably the largest challenge is how do you get a tree and any machinery into a mitochondrial into mitochondria of course there are a lot of genes are encoding and nucleus and those are amenable to gene editing but if if the mutation is in a gene that's in the mitochondrial genome then that's harder and and for proteins you can traffic that into the mitochondria using a localization tag but for the RNA guide RNA we don't really know how to put that in yet the second challenge is that there are many mitochondria in the cell and so efficiency becomes a big issue so how do you how do you edit all enough mitochondria to make a difference so that so those are two challenges that's why it's difficult yeah but a nuclear encoded mitochondrial genes we can certainly edit yeah when the waning the black beanie you could use CRISPR to target various activities to specific sites in DNA have you or anyone you know working on some way of modifying histones using this technology yeah a lot of groups have done that so for example Charlie Gore's box group at Duke University has used cast night to bring H stack complexes to specific sites and that will modify the chromatin keith's John's group said Mass General Hospital has taken the Ted protein which which alters cytosine methylation on DNA and that will also change chromatin density or folding so so there are a lot of different ways and in people I supporting that yeah I'm just sort of amazed that this whole mechanism works your cert you said you're searching six billion letters yeah and my question is yeah how does it do it and how long does it take so gene editing can happen in a Cell you can probably detect gene editing you know six hours eight hours after introduce the enzymes into the sauce how old how does it do it exactly I don't think I have a very clear idea but it probably is a combination of having many molecules in a Cell so that it's not just one molecule searching through the six billion letters is a parallel search and many things at once and then at the same time there's probably some guidance by the enzyme because the way cast nine works is that it first finds a small motif which is gg once he finds the motif initiates unwinding of the DNA and then compares the RNA with the DNA if the RNA and DNA are drastically different of how they fall off and and the bounce of somewhere else that's kind of the general picture but exactly how you can probably model that what haven't done it but you may be able to you know do a back and little calculation to to see if we're thinking about it right great question I have kind of a dual multiplexing question so I'm wondering kind of at the let's say RNA where you went to our RNA and a disease process and we know so often that you know you'll have a whole cascade of genes that will get turned on so do you think that there's going to be ways to kind of multiplex so you can hit multiple RNAs and then that kind of on a technological side on your lateral flow assay when you're looking for Zika you know will you be able to kind of multiplex that as well and look for multiple viruses let's say yeah yeah great question so for early editing certainly our hope that you'll be able to enter in many different insights even edit many many different RNA even editing multiple positions on mRNA and that can be done by providing a cell with multiple guide RNAs each one directing has routine Adar to a specific sequence you want to edit so I think that's definitely doable and in fact the way ketene works is that naturally CRISPR guys are are present in the form of a repeated ray so it's many different guides concatenated together as a single long poly RNA cancer teen can go and cleave it chop it into individual guide RNAs so for editing you can introduce into the cell in long Concannon RNA and cancer teen can chop it up and then and then get directly by each one to edit a specific sequence under a second question of multiplexing the diagnostic readout I think that's certainly something that that would be feasible there are different ways of multiplex on ladder flow you can imagine having several different bands each one corresponding to a different thing you're trying to detect alternatively it may also be able to develop a a a dot array and read out where you have a matrix where each spot represents a different thing you're trying to detect and these are all different ways that that you might be able to to to use the system so much knows the graphs of the amount of papers that have been published recently in these new technologies interesting and I'm wondering what your thoughts are about this increase in this this rapid rate this exponential rate of papers being published in new technologies and the fairly stagnant level rate in which the general public is literate about these type of technologies in science in general and that growing gap I wonder if you have felt that pressure recently and how you as a scientist feel like you need to adjust in order to communicate these new technologies efficiently and effectively to the general public great thank you this is a really a fantastic question I think it's great that there are a lot of papers being published I think well that well that what I represent is that one our ability to conduct studies and in our tools are getting better so that scientists can make discoveries faster can proceed through experiments or multiple experiments using the same amount of time that it takes before and so our aggregate knowledge that's accumulating is growing exponentially which is fantastic it's great for the world great for developing medicine at the same time I absolutely agree with you that we need more education in into biology into bring the the biological literacy level up and that's something that I think all of us have a responsibility to do through our conversations with people around us through our formal and informal obligations to teach and to communicate science I think those are some of the things that that each of us can do our part to make the public understanding biology better by giving a community lecture like this and hopefully it will convey some new information about biology to the general community to the general public so that our understanding and our our appreciation of science can can increase and then people can continue to support the development of science and research I think those are some of the things that that we can all do but we could all do more - at the same time on that note I would actually love to have a show of hands of who here are non scientists thanks all for coming and then who are would consider yourself a scientist and what category have I missed sure yeah the other half of the strand besides the guide strand appears to have one or two short hairpin turns and maybe it looks like a nap temir is that the bind specifically to the cast enzyme and would variation of that many Optimas change the way that the enzyme functions and is that being looked at other than the Strand part which is selective or are things that can be done with yeah good question so so the guide RNA has two components there's the guide which is what they use that they're right to specify what DNA to recognize and then there's the the complicated structure that comes with a structure is it's just like a hammer it's what the protein recognizes to be able to form this RNA protein complex now there are many versions of Katz 9 1 from each type of by 2 yeah each one of them have a different guide RNA and it recognizes a different structure so so so in a way that's also nice because it allows us to multiplex the system you can have different caste nines within the same reaction and they won't recognize each other's guide RNA so that you can edit different things knowing that there's not going to be a lot of crosstalk between these different cation proteins there are groups that are working on studying what what what these recognition properties are are how they are mediated so structural biologists have resolved the crystal structure of different caste lines bound to different guide RNAs that give as I understand any of how the recognition happens biochemists have been mutating these RNAs to see how does that how there's a single base change in the secondary structure or how does the change of the entire loop affect the binding affinity of the protein for this RNA and and those are all very useful information for continued engineering of the system so then just a follow-up question if you are looking at therapeutic opportunities you'd want to be able to sequence and that might be I don't know an RNA could be 80 90 100 pairs individual letters yes how critical is it as you're making these synthetically or is it even being made synthetically that the whole sequence is is accurately made very good question so studies have shown that changes of a specific one or two bases in many positions will probably not affect the binding to that aren't it maybe we'll change the affinity by you know 5% and 10% but not not substantively enough to be able to make a difference on preventing it from editing the DNA and sometimes deleting a few bases will still be okay depending on where you are making the deletion or inserting sequences will will be okay too so it depends on where mutation is but but certainly there is tolerance for that and and from a perspective that tolerance makes sense to because if a single mutation is they will completely take out the CRISPR system then there is not very robust mutations happen as cells replicate and so so so there is some resilience that's built into into the system we'll take one more question so computers were a lot less accessible to most people back in the day and then became very accessible to everybody and now anyone can kind of really just hack and make just about anything out there do you kind of and I'm kind of slowly seeing that more and I'm molecular biology game cast iron CRISPR systems you know on do yourself kits and stuff like that online just wanted to see what your thoughts are on that and you know just kind of that trend very good question I was finding you when the question like this may come up it's a very good question in fact I was just on a panel a couple of days ago talking about drug technologies and and we're looking at artificial intelligence okay Joe body is looking at gene editing I think these technology is very powerful and and as they become more easy to use as they become more even more powerful than how we are going to be using it I think requires a lot of thought it requires a lot of sort of thoughtful ways of guiding the use of the technology it's not as simple and it's not a question that can be answered simply because there are a lot of complicated issues associate with it and so so the thing that I think we we need to do as we move forward is that we need to you know be careful about what we are doing with the technology but also be very thoughtful in terms of communicating to to people around as how the thing how it's being used and also sort of debate and talk about how we should use the technology I think that the most challenging aspect may come from unintended consequences of a proper desire to use the technology so for example you know you're using the system gene editing systems engineer a particular species with a gene drive with the hope of eradicating a disease is a very good goal and and if it works it will have a profound impact improvement to our health probably held around the world but there may be I'm saying there may be because we don't know there could be or there could not be there may be unintended consequences too because as you engineer a larger system around us how that system interacts with other complex variables around us to to give off other impacts on the environment and those are things that are much harder to predict so so I think that's where we need to be very thoughtful about how we are deploying the technology it's not the desired use but but really the unintended outcomes that come from a well-thought-out well-intentioned use of the technology thank you everyone and thank you so much [Applause] Phung will be heading up to Keystone he's one of the lead organizers for our precision genome editing conference which will happen Monday through Thursday this week you

Channel: KeystoneSymposia

Views: 38,229

Rating: 4.8299999 out of 5

Keywords: Keystone Symposia, scientific research, life sciences, Keystone, CRISPR, Feng Zhang, Community Lecture

Id: cpHB3LEWNOY

Length: 89min 53sec (5393 seconds)

Published: Wed Feb 14 2018