Skip to main content
Loading Explore guides…
Building Self-Signaling Networks · Acedra
Home
Explore
Biology
Building Self-Signaling Networks
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
63 segments
Copy
Hide video
0:07
Hi. My name is Wendell Lim. I'm a professor at the University of California San Francisco and an investigator at the Howard Hughes Medical Institute. Today I'd like to tell you about using synthetic biology to build self-signaling networks. One of the most amazing things about living systems is that they are able to monitor their environment and make complex decisions. The example we're showing here, this is a classic film of the human neutrophil, a white blood cell, and you can see
0:35
it's able to detect and chase that small little black dot which is a bacteria. So somehow it's detecting signals given off by that bacteria and it's using it to coordinate its entire cytoskeleton to give you this coordinated movement and eventually the phagocytosis of that bacterium. And so what we're interested in is trying to understand how is it that cells like this are able to read in many environmental inputs and then make these complex decisions about what they're going to do.
1:03
What's particularly interesting for us is the idea that these decisions that cells make are coordinated by networks made of molecules. And so we're interested in trying to understand how is that a system of molecules can actually work together to make these sorts of complex decisions. In addition, cells show many, many different types of responses to their environment. And so how is it that so many different regulatory behaviors and signaling behaviors have been able to
1:30
evolve. How it has been able to use this to get the kinds of diverse cellular behaviors that we see. Traditionally the way that the field of biology has attacked these sorts of problems is to take an interesting system like a cell that does something interesting and then try to dissect it. That is using things like a microscope or other tools to observe, describe and classify what's going on. This could be using a microscope to look at the structures that you see in the cell,
1:59
using genetics or genomics to look at the DNA sequences that are and genes that are involved in the process or structural methods to look at the structures of these molecules. And network methods to look at how they're arranged in networks as shown down here. But one of the problems that we have in today's post genomic era of biology is the fact that these approaches really help us to take a complex system and take it apart. So that's essentially like taking something like a radio
2:28
and then disassembling it. And what results is that we now find that we often have a really detailed and complete molecular parts list. But what we lack is we still have a relatively poor understanding of how one can assemble these sorts of parts into functional systems. And so one of the things that we're interested in is trying to complement traditional approaches of taking complex systems and breaking them down into their molecular parts. The traditional deconstruction
2:57
suppression biology with what we consider a synthetic biology approach. That is the idea of using, trying to understand how simpler molecular parts can be assembled and synthesized to create more complex systems. This is a lot like how an engineer for example, thinks about how to use simple electronic parts to put them together to assemble devices and systems like a computer or, or they can put them together in different ways to make other kinds of devices with different behaviors.
3:28
So really this is an approach that really is inspired by engineering but really is what we think goes on during the course of evolution. Where pre-existing molecular parts are put together in different ways that might generate new behaviors. And the goal here is that perhaps trying to understand this complimentary synthetic approach of how biological systems are put together could help us not just understand the details of any one system but really more about the logic and rules about how
3:58
in general cellular systems are constructed. Really this is about asking the inverse question rather than trying to take apart a cell. We're really trying to ask can we view cells for example as complex devices or robots that execute new decision making behaviors and can we learn how to program them much like this robot here which was programmed to actually look for and search for red balls In a field of other blue balls. Ok so what we're interested in is trying to build
4:30
and rewire new cellular responses. Why do we want to do this. First of all, we really feel that we can learn by building, that by tinkering with these systems and putting them together in different ways and in systematic ways, we can reach a deeper understanding of the molecular logic of how cell signaling systems are put together. The second thing is that if we really understand these rules and these principles of how evolution has used molecular parts to build new kinds of
4:57
cellular responses and behaviors, then maybe we can actually apply this to build designer cells that have new customized behaviors. This includes the idea that maybe we could build the cells that carry out very sophisticated therapeutic actions, like being able to seek out and kill cancer cells. More broadly this really represents a kind of a new way of trying to think about biology mostly and historically biology has focused on trying to study the biological systems that
5:27
exist and in a way synthetic biology has given us the capability of not just studying what exists but making different variants of these and asking what can exist or what cannot exist and trying to understand the boundaries between these. Even though this area of sort of synthetic systems, biological like systems may seem like they're not truly biological, what I would argue is that this systems still have to obey the same molecular and physical laws as evolved systems. And so
6:00
really trying to understand the entirety of these really can perhaps help us to understand those principles better and in addition it might very well be that within the sphere of what can exist there may be cells that show very useful behaviors that we can harness. So how do we go about trying to rewire cellular signaling systems. One of the key principles that I want to raise here is the principle of modularity. So back in the 80s and 90s when we were really starting to sequence
6:32
all of these genomes, one of the really striking findings was that many signaling proteins were built from simpler parts. That is if you look at a lot of the molecules involved in signaling we found that they had a lot of modular domains that were easily identifiable by sequence. And moreover that many of these domains were found repeated in different proteins and in different combinations. These domains really come in two flavors which I'll talk about in a second in more detail.
6:59
But one is catalytic domains things like kianases that carry out catalytic information and post translational modifications like phosphorylation. as well as interaction domains that are involved in binding and recognition. But the basic hypothesis that emerged at this time was that perhaps nature has evolved complex regulatory circuits by using these modules as kind of a tool kit to put together different types of devices. So let me tell you about the different types of domains that we see.
7:29
As I said one class are catalytic modules that really are involved in directly transmitting information. So a good example of this are the enzymes that are involved in phosphorylation. Kianases and phosphatases for example. So kianases will put on a phosphate onto a target protein that shown in green here and that phosphate might change the conformation or the activity of that protein leading to some sort of downstream effect. Now that phosphate can be taken off by of course
8:00
corresponding complementary eraser catalytic function, in this case a phosphatase. In addition to this kind of writer and eraser catalytic activity oftentimes there will also be a reader domains that can recognize and bind to say the phosphorylated moiety. ok so you have this ensemble of readers, writers and erasers. Another example of that that's exactly parallel are the enzymes that regulate GTPases. So GTPases can exist in two forms the GDP bound state and then the GTP bound state which is
8:33
really the active one. The transitions between these states are mediated by another set of writer and eraser catalytic activities. Gaps are a quantonucleotided change factors lead to putting on and activating the GTPase protein whereas gap proteins are GTPase activated protein lead to the hydrolysis of GTP and turn it off the system. And once again there are also often times reader domains that can recognize the specific activated form of the GTP. So these sorts of writer and eraser enzymes are
9:10
often put together. You can get cascades or other network linkages where for example one phospholation event will activate another enzyme that will in turn activate another enzyme leading to the sorts of networks that and circuits that we see in cells. Now the second important class of modules that we see in signaling domains in signaling protein are interaction modules that really play a more crucial role in directing and controlling the flow of information. so these catalytic
9:39
domains that I talked about are often combined with recognition domains, many of them are protein interaction domains. Like this SH 3 domain that recognizes a cognate linear peptide motif, in this case a proline enrichment type. there are many other sorts of domains that recognize different distinct types of linear motifs. In addition there are also recognition or interaction domains that bind to specific lipid species at the plasma membrane. So these are combined with the catalytic
10:09
domain control where these catalytic domains go and who they interact with. So how is it that catalytic and interaction domains can be put together to yield the diverse circuitry that we see. This slide shows two very simple but very prevalent mechanisms by which these domains can be put together to give you complex circuits. One is really based on the simple principle of recruitment assembly, that the cell is full of these different types of catalytic modules. and really a key question
10:40
is how do they efficiently interact with the right ones. And so what you'll find is often let's say protein X is a kianase and Y is its substrate. Then that, that they might have a modular interaction domain and recognition ligand fused to these partners which allows protein X to specifically interact with and modify protein Y. The importance of these sorts of recruitment and assembly interactions is really exemplified at the extreme by a class of proteins known as
11:12
scaffold proteins. These are proteins that have multiple interaction domains and they often combine to multiple signaling proteins like these shown here, X, Y and Z. They really help to coordinate a set of catalytic components to be at the right place in the right time, allowing them to interact in a very specific cascade. A second way that these modules can be used to lead to interesting behavior, is the construction of allosteric signaling switches using modular auto
11:40
inhibition but often times you'll find a catalytic domain say like a kianase that has other interaction domains within it. But often times in the basal state these interaction events will actually auto inhibit or interact with the catalytic domain itself or with other parts of the protein. and that can lead to inhibition either through directly sterically blocking the active site of the catalytic function or by conformationally contorting it so that it's inactive.
12:09
and if you get this sort of situation then oftentimes there are cases where a competing ligand that this disrupts this auto inhibitory interaction can then result in the allosteric activation of that catalytic function. in this case for example activation of a kianase. So we see these sorts of mechanisms over and over again in many different signaling projects with different types of catalytic interaction domains. but it really lends itself to the possibility that these modules and motifs
12:40
really might form a toolkit or a set of building blocks or a cellular language that we can use to create new functions and that evolution seems to have used this. we see this of the different combinations of domains that we see in evolved proteins. But also another piece of evidence comes from disease. so in diseases like cancer or in viral or other pathogenic diseases there's a lot of evidence that these sorts of modules have been hijacked and used to redirect catalytic functions to new
13:09
targets and other things like that. And it's in other words really generating signaling behaviors that actually are associated with a pathology or a disease. So a corollary of this is that if this is really true, then it argues that we should be able to use these modular words to say new things, that perhaps we can actually use them to engineer ourselves in a somewhat systematic and logical way. So if we can do this, what is it that we want to say and how do we say it.
13:39
So about ten years ago my group started to try to explore this question. and ask really could we start building new types of signaling proteins as well as signaling networks and cellular behaviors. so one of the first things that we did was actually ask could we build synthetic signaling switch proteins and we took a group of different catalytic functions and then a group of different interaction domains and then made a random combinatory libraries in which we
14:11
combined interaction domains with catalytic domains as well as the inter molecular ligands of these domains with the idea that we wanted to find out. Would any of these show inhibition and in a way that would allow them to now if we added competing ligands become activated and would they show interesting behaviors, regulatory behaviors like natural signaling switches. And in fact what we found is that when we did this a fairly large proportion probably about 30% of them actually showed
14:40
some form of auto inhibition and regulation and moreover when we looked at these in detail and looked at how the types of behaviors of integration between multiple inputs we saw a lot of different types of bulian behaviors, really we saw the diversity that mimicked the types of regulation that we usually see in natural signaling proteins. Arguing we think that this is, this type of domain shuffling really forms a way to get an evolutionary starting point for building new types of signaling
15:07
switches. Another example of looking at the modularity of signaling proteins comes from looking at signaling behaviors in the model cell system of yeast. So yeast is a single celled organism that has to make many decisions based on what is in its environment. So one of the things that yeasts do is they actually have two cell types and they can mate with each other but to induce that mating response, they have to detect a pheromone that comes from their partner and so there's a pathway
15:37
called the mating response that actually detects a mating pheromone. In this pathway shown in red, and then that leads to the activation of a set of mating response genes. Now a completely different pathway that yeasts have is a pathway that deals with hyperosmotic stress. So if you put yeasts into high salt environment it detects that change and then expresses a bunch of genes that help protect it against that osmolarity. So although these pathways are completely different, what
16:07
does link them is that they are both mediated by a set of kianases known as map kianases. The problem is that the cell is filled with many different types of kianases and quite a few of these mapped kianases that are very closely related. And so the question is if these two pathways which are physiologically distinct use shared or similar components, how is that they actually pass the information, transmit it in the right way and you don't get kind of cross talk so you get the wrong response.
16:38
And so one of the things that was discovered around this time was that there existed these scaffold proteins. For example, there's a scaffold protein that has multiple interactions remains that binds to the mating receptor as well as the set of kianases that are involved in the mating response. They essentially act almost as a bread board to wire together the circuits so that you get specific transmission of information from the mating receptor to the mating genes.
17:05
Correspondingly, there is also a scaffold protein associated with the osmotic response pathway which again binds to that receptor and then also assembles a distinct set of kianases that are involved in transmitting the information to the osmotic response genes. So in other words both of these pathways can coexist and be specific because there is a scaffold protein that essentially wires them together in a specific way. And so one of the hypotheses that came from this was that perhaps then
17:37
evolution can create new pathways and by wiring kianses and other catalytic functions together in different ways by generating new chimeric scaffold proteins. So Sanhing Park a post doc in the lab at the time decided to try to test this by asking could we make a chimeric scaffold protein that would assemble different kianses in these two distinct pathways. And what he found is that when he did this and made the right sort of combination of pathways of components, he was able actually to
18:06
make a scaffold that would when put into a living cell would now cause the cell to induce the osmolarity response when the cell was stimulated by the mating pheromone. So we could actually create a new input output pathway really mimicking evolution. In this case we had a pathway where the cells would only survive at high osmotic stress when we stimulated them with mating pheromones. So really supporting the idea that without changing actually the catalytic component
18:34
and really only changing the factors that were assembling them, we could create new types of input output wiring and behavior in the cell. Here's another example of rewiring that we have been able to do. So Ensom Levskaya recognized that in plants there are these really interesting protein interaction domains that where the interaction is regulated by light. So if you shine a certain wavelength of light 750 nanometers on this protein it will change conformations
19:06
and now tightly bind to its partner. And so what we realize is that again many signaling behaviors in mammalian cells are controlled by induced recruitment. So we asked could we actually replace many of those interactions. oftentimes a catalytic domain will be recruited to an activated receptor but could we replace that receptor now with this light responsive protein from plants and recruit a catalytic domain. For example, this GET to the plasma membrane
19:35
where it could act on its target GTPase and signaling output. could we actually control the cell now with light. And so this is an example of linking this light activated protein to a Rac GET so that's going to activate the protein Rac GTPase which itself activates actin polymerization. As you see here, what happens is that when we shine the light now on which is the red circle, on the cell, you see recruitment of that Rac GET and actually protrusion of the cell at that
20:09
point so we can actually make the cell follow the light by putting this novel chimeric protein that combines light controlled protein interaction with this catalytic activity. So what these experiments hopefully illustrate is that we really can use these sorts of signal modules to flexibly rewire and reprogram so they're signaling proteins as well as the networks and to generate novel cellular behavior. So one of the things that we become more interested in is asking the second question
20:45
about applications. Can we actually start harnessing this capability to rewire cells to, and put it to use. And one of the most exciting things for us is the idea of using this in therapeutics, in medicine. that mostly today when we think about medicine and therapeutics, we think about small molecules or biologics, which are very powerful and wonderful agents. But one of the things that limits them is they largely if you put them into the body they act in a systemic way. And then also they
21:17
largely can just either block or activate something and they can't actually make very complicated decisions. And that contrast with for example the cell, so a cell of the immune system for example which also essentially is a therapeutic. it fights disease, is able to move through the body and find where there is a problem. so it can be spatially targeted. It can also integrate sources of information from that environment and make complex decisions. But it can also execute complex actions
21:44
like this kind of migration or phagracytosis that we see with the neutrophil. And so one of the visions that we have is if we really understood these principles of how we can rewire self-signaling behaviors then perhaps we could actually start to customize cells so that they can actually carry out functions that are therapeutic. This would entail for example trying to change what sensors the cell has so that it can detect either disease or user provided inputs that we want
22:15
to also be able to change the self-signaling networks to control what's, how this information is processed. And how decisions are made. And then link these to the many interesting and powerful payloads or actions that cells are capable of to really yield now what we consider a much smarter therapeutic response. So one of the great places test beds to really explore this today is in the area of adoptive immunotherapy so immune cells, particularly a type of immune cell known as
22:50
key lymphosites or T cells are an ideal test bed for this kind of therapeutic engineering. These are cells that are able to for example recognize and kill a cell that's infected by a virus. so they're a very powerful activities. More importantly we know that we can actually remove T cells from a patient. We can genetically modify them ex vivo. We can expand them ex vivo. And then we can transfer them back into the patient. So this is pretty well established.
23:17
And so we can do a lot of different things that don't have to do with trying to modify things In situ. In addition, T cells can be used for many potential targets. They could be used to recognize cancer. They could be used to recognize chronic infection. And also they could be used to combat auto immunity. So how could we for example redirect T cells to recognize cancer. So if you look at a native T cell , the key receptor, a molecule that's used to mediate its action
23:52
is the T cell receptor. So if you look closely at that, that has a extra cellular domain and then key intra cellular signaling motifs. And so for example when the cognate peptide antigen is bound by that extra cellular domain it will lead to phosphoralation of these intra cellular motifs which leads to the recruitment of various downstream signaling proteins that then in turn lead to the activation of that T cell. And for example killing of the target cell. What was shown in the 80s
24:22
by a number of people including Zelig Esshar was the idea that you could actually make synthetic or chimeric T cell receptors so the idea here was could you take an antibody that recognized a antigen that was only expressed on the surface of a tumor cell. And then remarkably they found that if they took this and then added the intra cellular portion of the T cell receptor that this would actually get activated when it recognized the cell with this antigen and lead to T cell activation.
24:52
So you could actually make these kinds of synthetic receptors in this case known as chimeric antigen receptors or CARs that are built of different extra cellular recognition domains things that recognize tumor antigens But then link together different sorts of signaling inter cellular signaling device that control the activation of the T cell. And now 20 years later this process is refined somewhat and they're really quite powerful versions of the chimeric antigen receptors
25:23
that can recognize tumor antigens and stimulate T cell activation so we have these modular synthetic sensors that can recognize disease antigens and redirect T cell activity towards these cells. So the idea is that we can take these CAR modified T cells put them back into the patient and they should be able to move around the body and seek out those cells that have the cognate antigen and then kill them. So although again this seems like science fiction what we now know in
25:54
today's world, over the last several years is that we actually can use this to attack certain B cell cancers. this has been remarkably successful in a number of studies shown here. And that there have been quite a few patients that have had a form of B cell cancer that is non responsive to chemotherapy, that have been cured by using this sort of approach with modified engineered T cells. So this is tremendously exciting but there are still problems with this particularly
26:26
what we have shown is that we can take T cells, these powerful cells and redirect them. But often times there are many strong side effects because these cells are very strong in terms of their immuno response. They secrete a lot of molecules called cytokines and really leads to a big systemic immune response. So although we can redirect them, right now oftentimes they are too strong or poorly controlled. So we really need to learn not only how to redirect
26:53
this beast, but also how to tame the beast. And so this is where we have been trying to use synthetic biology to generate engineered T cells that show much more controlled behavior. And so one example is the idea that could we actually make T cells that are remote controlled. T cells that only switch on when for example a physician provides a drug. We'd still want these cells to recognize the tumor cells but we want the secondary switch to be layered over this.
27:22
That allows the user or the physician to control when and to what extent they are activated. And so Chia Wu, a post doc in the lab, has been doing studies at trying to build switchable chimeric antigen receptors. So this is the conventional chimeric antigen receptor that of course has an extra cellular antibody that recognizes the tumor antigen and also has these intracellular signaling motifs. What we decided to do was to ask could we actually build an additional switch
27:51
into this receptor by creating a split version of it. so could we take the recognition component but then take off some of the key signaling modules and put those in a separate molecule but then bring them together using modules that only heterodimerize when we add a drug. Ok so the idea here is that the T cell would only get activated when it had this combination of both recognizing the cognitive antigen as well as having the drug provided by the user.
28:23
So fortunately we have been able to construct a number of different versions of this that work and one of them is shown here. And so these are some actual assays based on these cells so what's plotted here is T cell activation monitored by release of a cytokine called IL 2. And so what you see is when we look at the conventional CAR cells, CAR T cells that if we stimulate with the cognate antigen which is shown in green that the cells get activated.
28:53
And they actually don't care whether or not the drug is present or not, which is indicated by the orange square. In contrast when we move to this on switch CAR T cells bearing this on switch CAR we see that they only get activated when there's a combination of both the cognate antigen in green as well as the drug in orange. So here is a case where the physician could actually control when the cell is activated as well as we found the timing and the dosage or potency of that response.
29:21
What's shown here now is time lapse microscopy of these switchable CAR T cells attacking the tumor cells so what you see here in the first movie is the cells, the T cells are shown in color and this is one recognizing a cognate, a tumor cell with the correct antigen. But this is in the case without drug and so you see that they interact but over the course of this 30 minute movie they don't actually kill that target cell. In contrast we now take the same cells and mix them with cancer
29:55
cells and then add the drug what we see is this. And we see the cells now engaged with the target cells. But then immediately you see this cell bleding and then this blue dye is indicative Of a cell undergoing afactosis. So that cell immediately starts dying and once that cell is on its way to death this T cell reorients contacts the second cell, reorients its receptors to that. forming a different synapse and then actually starts killing that. you see the blething as well as the uptake of the
30:23
blue dye. So within this 30 minute period where the in the absence of drug the cell doesn't do anything, doesn't kill anything we see this, in the same time when the drug is added they are able to potently and serially kill multiple target cancer cells. So we think this is tremendously exciting. We think that in today's world, now that we have the capabilities of engineering and modifying cells and we understand much more about the principles of how Signaling networks are put together
30:57
that we're on the verge of really exciting applications where we can use this toolkit of modules to engineer different types of cells that can sense disease signals and environmental or niche signals in the body as well as user signals and integrate them to give you customized therapeutic action. And so we think this is a really exciting period for applying synthetic biology to these sorts of mammalian cells, to immune cells to create exciting and hopefully very useful therapeutic functions.
31:29
And so that's the end of this lecture. I'd like to thank all of the people from my lab who did all of this work. a very creative and inspiring group of people as well as all of our collaborators and colleagues. But I’d like to thank you for watching this video. end
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat