| 1 | The Royal Institution | The cell programming revolution – with Mark Kotter | 88590 | 2545 | 169 | 72.0 | positive | 29:22 | [Music] [Applause] thank you thomas um for this kind invitation and a big thanks to the royal institution for having me here to give this talk these are unusually challenging times and the question is whether it's appropriate to talk about science in the context of such geopolitical events this but these events have extraordinary impact now on people and the long-term and positive changes on our societies are often driven by innovation so my background i'm a neurosurgeon i look after spinal cord injury patients partially because i'm curious but also because the tools that i have to treat my patients are very limited i went into research at the moment i have screws and rods that i can put into their spines but there's nothing i can do to give them back the use of their hands or walking so we need to innovate and the thing i'm really very excited about are cell therapies that's the concept of using a cell as a therapy and these cell therapies can replace lost cells but they can also interact with their environment so you can call them intelligent drugs the last wave of innovation however happened in physics at a very particular moment in time when physics transitioned from a descriptive science so scientists were observing and recording to a predictive science a special moment was the introduction of calculus to describe the laws of motion maths gives you a lens to look into the future we can now calculate how long it takes an apple to fall from the tree until he hits the ground and this together with the creation of more and more sophisticated tools sparked waves of innovation from telescopes to steam engines spacecraft and computers and computing again introduced a phase shift since ada lovelace wrote the first computer script we've built computing power that allows software to transform the world but here's the limitation of current computing the hardware determines the software not vice versa but what if the software could determine the hardware what if you could use computer code to control the physical the physical nature of objects it's already happening right in front of us it's called biology so we need to move away from the traditional questions of biology how does it work how is it built and how did it get there to a new paradigm in biology in which a code script is the kernel of biological theory in this paradigm the nucleus is the hard drive it contains the dna and these genes interact with each other to create programs so-called gene regulatory networks the entire space of programs you could call the operating system of a cell and at the core of this is the genetic code this is the way that information translates into the physical reality into protein and here you see a wonderful depiction of the inner makings of a cell and all these different proteins that form molecular machines at the top you can see the nuclear membrane with nuclear complexes they're very long structures they're proteins that provide structure to the cells and then they aggregate to form organelles on the right you can see an energy producing organelle called mitochondria but do we have the tools to program biology well we started doing so about 10 000 years ago when we started to select plants and animals with traits that are desirable and then in the 1700s we learned how to hybrid hybridized plants and today nearly all of the vegetables and grain that we consume have been somehow influenced by us in the 1800s we learned about the concepts of evolution and the laws of inheritance in the 1950s we discovered the structure of the dna and in the 1970s we discovered the first restriction enzyme that's a protein that can cut dna in very particular sites and that allows you to engineer biology so we've used it to generate bacteria that create insulin and insulin is a very important drug now and that treats patients with diabetes in the 1980s the tools became more refined we started to use them to to create plants modify plants in the 1990s we learned how to clone mammals um some of us will remember dolly the sheep in the 2000s we sequenced the human genome 2010 we built very refined tools for engineering dna crispr the molecular scissors that allows you to cut the dna exactly where you want it to cut and we've also learned how to activate programs in cells and i'll come to that so the tools are very sophisticated but what about the maths again sequencing and other omics technologies in the recent years have helped to translate biology into numbers and attending today's conference here at the conference about programming cell identities i learned that maths is really progressing as well i think we're not far off predictive models in maths so for 50 years we've been engineering biology and at the center of this was greg winter sir greg winter engineer developed a technology that allows you to engineer antibody drugs um in in bacteria and this allowed us to create a second generation of drugs so the first generation of drugs were small molecules and like aspirin this second generation is a lot more sophisticated and more recently we also started to program other microbes to produce materials biodegradable materials such as spider silk and have been used in this very high spec running shoes so we've been programming lower species but is it possible to also program higher species like mammals here we need to go back to the 1980s the heydays of molecular biology and you see harold weintraub here who discovered a gene a transcription factor that's a gene that regulates other genes which he called myod and you observe that if you activate that gene cells switch fate they go from a skin cell into a muscle cell that's completely contrary to the dogma of developmental biology so developmental biology is still the predominant dogma in science it's based on the understanding that a stem cell forms the origin of all cells in the human body and in order to produce a neuron or a skin cell it has to differentiate into more and more specialized cell types waddington proposed a model by which a marble you can look at the picture marble running down a valley and it hits certain branch points where he has to choose these are called self-made choices in order to ensure that not one single cell type is missing during development these choices are based on probabilities stem cell biologists are studying the chemical cues that enable us to bias these self-made choices in order to nudge cells down particular trajectories and to create particular cell types there are two problems with this approach number one development takes time it takes about nine months until a baby is born and second these self fade choices there are a lot more of them on the way to become specialized cell types so you're accumulating a lot of probability and chance events and that causes inconsistencies today long protocols with lots of inconsistencies are the major but bottleneck for manufacturing cells and it's the reason why we don't have cells a stem cell based therapies in our clinic yet but harold weintraub proposed a completely different paradigm if a single transcription factor can switch cell fate we have to relearn biology unfortunately harold died 49 years old and his workforce forgotten for nearly 30 years until shinya yamanaka surprised the world with a new discovery which earned him his nobel prize so yamanaka discovered that a combination of four transcription factors is able to program a skin cell back into a stem cell and this stem cell behaves just like an embryonic stem cell so this allows us to create stem cells from every individual each of us has their own repair kit and the second thing it did was take away all the ethical constraints in the use of embryonic stem cells there is no need to chop up an embryo finally this is an early stage he maybe even opened the door to cell rejuvenation now marius wernick was in the audience was inspired by this work and asked the question whether this paradigm can be generalized so he showed that you can take a skin cell and program it into a brain cell and then he took a stem cell and programmed him to into a brain cell and then i have no idea why you did this marius you took a liver cell and turned it into a brain cell in any ways i think this showed that this a perhaps a generalizable principle but there's still a problem with cell reprogramming it's often very inefficient this is also what we found when i started my lab at the university of cambridge and the objective that i had was to program human oligodendrocyte precursors that's a specialized cell in the brain that's really important in repair processes of the spinal cord and the brain so i was joined by matthias pavlovsky a very talented phd student who now has his own lab in germany and within a fairly short period of time he figured out a transcription factor combination that allows him to produce these cells and you can see a beautiful depiction up here the problem was that this was a very rare event about one in a million cells turned into oligodendrocytes so the dogma was you need to refine the transcription factor combination the program isn't right but his data shows something entirely different it suggested that there's a new problem that we were facing called gene silencing so the stem cell recognized that someone is wanting to activate a new program that is inconsistent with their current state and they shut it down so we decided to tackle this problem and for a few years we only made incremental you know change improvements in fact it's quite hard actually um i used nearly all resources that we had we put the credibility of the lab the survival of the lab at stake and we were quite desperate until one moment um matthias called me and showed me the following video so here you see human stem cells converting into brain cells within days so 10 times faster as previously and the incredible thing here was that every single cell in this culture turned into a neuron and these neurons are the brain cells that connect to each other you can see fine processes and then they start signaling electrical impulses and and reach maturities that you haven't seen before so my first reaction is this this is not possible he probably had photoshop this video these students in cambridge are pretty smart so he took me down to the lab we looked at the cultures and it was real so i thought probably the stem cell that we were using was broken um so let's take another stem cell um takes six months he came back it worked again and then thought okay this is a maris vernick protocol i mean this is probably very unique and therefore um this is still a glitch but over the years our lab company and collaborators have been able to show that it works again and again and again for various different cell types now let's look at that if you think about disease every disease manifests at a cellular level so every cell is potentially a treatment or a model to study this disease so the impact of this is quite incredible the way we achieved this was by inserting the genetic program of the cell into specialized locations in the dna so-called genetic safe harbor sites and they protect the cell but they also protect the program from silencing this creates a control system that allows us to switch on genetic programs at will so the cells can't escape and this allows you to create all sorts of different cell types like the neurons that we've just had a look at so the next challenge here is how do we identify the cell type programs that encode each of the cell types every human cell has twenty thousand genes and at any particular moment in time approximately ten thousands are switched on and these determine the function and the identity of the cell of the twenty thousand genes two thousand genes are transcription factors these are the genes that regulate the activity of the other genes and themselves and what we've seen from studies emerging over the past decade that about one to six of these transcription factors encodes a cell identity so every cell has their own unique combination so the question is how do we read this out if you look at the number of experiments that are required um this is pretty large eight point eight times ten to the sixteenths um i've calculated this approximately thousand scientists working for three thousand years that's going to be difficult so we had to come up um with with something else and to do so to think bit deeper about this problem we entered the collaboration with the london institute of mathematical sciences so they were able to show that the way that dna rna and protein interacts in a cell it creates a set of constraints that limits the possibilities and already reduces this by orders of magnitude and we've also been able to refine the experimental approach by breaking down these pools of transcription factors into smaller pools and ways of screening them we're now at a point where we actually can read out all cell type programs in one big effort it's not going to be cheap but it's doable but imagine the impact that this could have so this is a darpa style project and i'm really keen to get this off the ground the next question is how is the cell type actually determined and defined in the genome so there's what intense landscape again and there are really two possibilities on the left you could imagine that a cell is entirely predefined in the genome so it sits in its own little valley and on the right is the opposite hypothesis a cell type is a label that we attach to a combination of sub programs so if the second we're right we should be able to combine programs from different cell types and we again have to go back to the 1980s to harold weintraub's work because he asked that question and he showed that we can combine a cell program of a muscle cell and a pigment cell a melanocyte and he also was able to create some sort of hybrid of a brain cell in a muscle cell and we've done similar things using transcription factors combining a brain cell with an immune cell this confirms that the operating system of cells is created by sub-modules and it opens the opportunity to create cell types that actually don't exist potentially have therapeutic use so then the next question that we always get asked is how close are these reprogrammed cells to the real cells that we have in our body traditional stem cell biology has struggled to produce cells that resemble the cells in our body because often they are stuck in an embryonic or early immature phenotype but if you want to do a therapy or if you want to do research often you'd like to have the mature the adult cell type so this picture up here is a landscape that we've created with data from the human cell atlas so the human cell atlas is a international consortium that is trying to chart all cells in the human body and they're using a technology called single cell sequencing so they can read out thousands of genes in single cells and so if you look at the top these are this is a picture this is a landscape of uh liver cells the cells that you can find in the liver at the top you can see adult hepatocytes liver cells and at the bottom they're the fetal the embryonic early stage counterparts and in the middle there are all sorts of other cell types that you can find in the liver from immune cells to fibroblasts etc etc we can now create liver cells with transcription factors then we can project them onto this landscape and what you can see here for the first time we were able to produce fully mature hepatocytes in the top their fetal counterpart and some of the intermediate stages as well so now we've got about a technology that allows us to program biology what are we going to do with it i'm a medic so my my primary aim is to use this technology for medical applications and the technology that we've created in the lab has been spun out into two companies one of them is bitbio is the leading cell reprogramming company right now we've been able to to raise more than 200 million dollars so far and the objective is to create cells that we can use to democratize science give these to scientists but in particular to also create a new set of affordable cell therapies we are incredibly privileged to work with some of the leading innovators and this includes the chair of our board hermann hauser he's been at the center of two new industries he's one of the co-founders of arm so chip industry but it also sparked the sequencing industry by backing the technology that now most people are using and i've already told you about sir greg winter and maris vernick i'm also extremely privileged to work with roger peterson who is one of the forefathers and founders of human stem cell research and the cambridge stem cell institute so with a consistent and scalable source of human cells we can change the way that the drugs are developed let's think about alzheimer's we've not been very good at creating alzheimer drugs and one of the main reasons is because there's a difference between the animals that are used for research and the human beings that suffer from alzheimer's in fact there is not a single mouse on this planet that suffers from alzheimer's this is a entirely human condition so in order to create alzheimer's drugs scientists engineer the mice so that they get something that looks like alzheimer's and then companies run a very efficient drug discovery process and at the end of that when the drug enters in the clinic we learn that it doesn't work in fact we learn that probably what's been created in the mouse isn't quite what we need to treat so the way to fix this is to use human cells the cells that are actually affected by the condition but what i'm most excited about are cell therapies this is the first child that has been treated by a cell therapy this therapy uses blood cells immune cells from the patient and engineers them in order for it to recognize cancer so she's been cured from a blood cancer this is a very complicated process at the moment it's extremely expensive one treatment costs hundreds thousands of pounds imagine we can use this software approach to create cells at scale and make cell therapies affordable this therapy was based on t cells they're not very good at going to solid tumors they're rejected there's another immune cell type that is much better suited to that called macrophages so here you can see macrophages that have been programmed with our technology and they are fighting bacteria as they eat them they turn red this is just the beginning think about combining cells in order to print organs and organites or using nerve cells and integrate them into bionic devices to control artificial limbs moving beyond medicine the second company uses the technology to create cells for the manufacture of meat this will not only allow us to fine-tune our stakes but it also will tackle some of the challenges associated with farming like the use of antibiotics causing bacteria to become resistant or the greenhouse gases that are associated with farming and the risk that another virus transfers from animals to human beings causing another pandemic but this approach might take us beyond our planet because i don't think it's feasible to take a cow up to the station that we're going to build on mars finally we may be able to use this technology to create new kinds of computers here you see a living computer chip made of brain cells that are learning to play pong so when software determines the hardware anything seems possible thank you you | ↗ |