Scientist Stories: Feng Shao, Pyroptosis in antibacterial & antitumor immunity

9:00 AM in my, let's see, in my watch, 9:00 AM PST. And I hope that you are all healthy and well. And thank you for connecting to another Global ImmunoTalk. My name is Elina Zúñiga from the University of California, San Diego. And before I introduce our fabulous speaker today, I would like to remind you that next week our speaker is Dr. Eike Latz from Germany, so I hope that you can all join us next week as well. And so let me— I see that I have an echo, so let me Check this before I continue. Okay, I think that we should make it. Okay, and so now is time and my pleasure to introduce to you our fabulous speaker today, Dr. Feng Zhao. So Feng is native from China and is based in Beijing. Uh, he is an investigator and deputy director at the National Institute of Biological Sciences in Beijing. He was a chemistry undergraduate of Peking University and obtained his PhD from the University of Michigan in 2003. And before joining the National Institute of Biological Sciences in Beijing, In 2005, he was also a Damon Runyon Postdoctoral Fellow at Harvard Medical School. Dr. Xiao's research lies at the interface between bacterial pathogen and host inflammation, a very important topic for sure. And also very importantly, he has made really outstanding advances in this field.. For example, Feng identified many of the known cytosolic receptors for bacterial molecules, including caspase 11, 4, 5 for sensing LPS, and alpha kinase 1 for sensing ADP heptose, which is a precursor of LPS biosynthesis. These studies represented very important breakthroughs as they broke the dogma that LPS is only recognized by the membrane toll-like receptor 4 and explain why blocking just toll-like receptor 4 failed to treat sepsis and therefore reopening the door for new antisepsis and new anticholitis drug development. Another impressive example of Feng's research has been the discovery of the gasdermin family of pore-forming proteins and the redefinition of pyroptosis as gasdermin-mediated programmed necrosis. He identified gasdermin D, whose cleavage by caspase 1, 4, 5, and 11 determines pyroptosis, which is critical for septic shock and other inflammatory diseases. He also showed that gasdermin E is activated by caspase 3, which occurs mostly in non-cancer cells and contributes to cytotoxicity of chemotherapy drugs. And more recently, he demonstrated that pyroptosis is critical for phagocytes cytotoxicity of the target cells and that gasdermin activation can stimulate potent antitumor immunity. So, these were amazing breakthroughs, as I mentioned before, that have been published in top-tier journals and have been recognized by numerous awards that Feng obtained throughout the years, and as an example, we can mention the Future Science Prize, the Howard Hughes Medical Investigator International Early Career Award, and the Protein Protein Society Irvine Siegel Young Investigator Award, again, among many others. Feng is also a member of the Chinese Academy of Sciences and the Sherman National Academy of Sciences Leopoldina, as well as an associate member of, member of EMBO and a fellow of the American Academy of Microbiology. So I am sure we will learn much more about gasdermin and pyroptosis, and we will be delighted by Feng's talk today, which is entitled Pyroptosis in Antibacterial and Antitumor Immunity. So Feng, thank you so, so much for accepting the invitation to be a Global Immuno speaker. I know it's very late, in Beijing, so particularly thankful that you made the effort to connect live with us today. Thank you. So, as I mentioned to you, we always like to know our speakers better, and you had an amazing, tremendously successful scientific career, and I'm sure it was very fulfilling with all the impactful discoveries you have made. And so with I would like to ask you a question that may help young and maybe not so young scientists in the audience for their own career. And so the question for you today, Feng, is what would you advise to your 25-year-old self if you were to start your scientific career again?

Thanks, Elina. Thanks for bringing me to the Global ImmunoTalk. So at 25 years old, and I just started my PhD study at University of Michigan, and in retrospect, I think I do have that advice. I actually— so I had a— in the end, I had a pretty successful PhD career, published a Cell paper, Science paper, in less than 4 years, but But yet I think it was in my third year I almost quit, almost quit my PhD study because I was pretty successful during my undergraduate and master's study in China. When I went to Michigan, I did my PhD with Jack Dixon. I was a biochemist, so I was not so patient at that time. After 1.5 years, less than 2 years struggling, my project did not move forward at all. I was so worried that I may not be able to defend my thesis and get my PhD. So I started to lose confidence in myself and I almost started to switch my career to biostatistics. And actually, I did pick some classes course to take. And, but I stick for 2 or 3 months and then my project start to work. So I think if I did quit at that time, almost, I almost quit actually. If I did quit, I would not have a scientific career like what I have now. So, and the reason for that, I was not so patient. I think so. For young students, I think the most important thing during your PhD study is be confident in yourself, in your scientific pursuit, and try to learn as much as possible. And indeed, the publication at that stage is not that important. The most important thing is to learn how to do science, how to think about questions, and how to solve them. Problems and overcome the challenges you might encounter during your PhD study. So stick on it and be patient and be confident in yourself, and that's my advice.

That's terrific advice, Feng. I'm sure super helpful for students in the audience and even postdocs and faculty. At some point, you struggle during your scientific career. It's important to be patient and continue And yeah, thank you so much. I'm sure this is super helpful. So, Feng, we cannot wait to hear your talk and learning from you. So, if you wish to share your screen.

Okay, okay. I'll start to share my screen. Oh, can you see my screen now?

Perfect, perfect.

Thank you so much. Yeah, so welcome everyone to today's Global ImmunoTalk. And so it's pretty late in Beijing now, but I try my best to do a good job. Today I would like to share our work on pyroptotic cell death with you. And pyroptosis is is a quite hot topic at this moment. And I'll take you through some of the historical background and give you an update on the recent development in the field as well in my own lab. So I have to admit that I'm not trained as an immunologist. I was trained as a hardcore biochemist. And you probably will see during the rest of my my talk. So my lab started in 2005, and we originally studied bacterial infection, bacterial pathogenesis mechanism. And a few years later, I started to expand my research program to innate immunity. So we want to ask the question, how does the innate immune system deal with bacterial infection? So in the immunity, we know this toll-like receptor pathway, which is a paradigm in the immune sensing of microbial infections. So we realized at that time that the toll receptors, which are all on the membranes, and they can recognize extracellular bacterial infection. However, there is another group of bacterial pathogens, like Salmonella typhimurium, Shigella flexneri, even tuberculosis. And those bacteria, they invade into the cytosol, as this right picture shows you. So when those bacteria invade into the cytosol space of the host cells, you could imagine that the toll receptor sitting on the plasma membrane will not be able to recognize intracellular bacterial infection. So we then ask the question for the lab that, How does the immune system deal with bacterial infection in the cytosol space? Can we discover some novel receptors? Can we identify some new players in this intracellular immune sensing or defense pathway for intracellular bacterial pathogens? So here I'm going to show you a very simple experiment using this movie made by at Mayo in North Carolina. And in this experiment, we basically simply infect macrophages with intracellular bacteria like Shigella or Salmonella, as I mentioned. And at the same time, we add some propidium iodide into the culture medium. As you know, normally healthy cells will not uptake the DNA dye, the nucleic acid dye, the PI. However, after the cells have been infected by intracellular bacteria, you can see that the cells turn into red. The dye, the DNA dye, went into the nucleus and stained the DNA, and you can see that the nucleus is already overstained. And this simple experiment suggests that the infected cells undergo explosive lytic death, which we now call pyroptosis. So this not only occurred with those Shigella or Salmonella infections, but also many other intracellular or even extracellular bacterial infections. In addition, in immunology, we know there is another group of molecules called DAMPs, endogenous danger molecules, such as ATP. And they can also, like ATP, it can also stimulate macrophages to undergo pyroptosis. So, and this cell death, there has been two misunderstandings in the history for this kind of cell death. So the cell death was originally observed in 1986 and later on in 1992. But for a long time, it was disregarded as a bacterial virulence mechanism. So very intuitively that people think that bacteria want to kill us, right? But indeed, this was not true. This is not a bacterial virulence mechanism. The cell death is indeed a host immune response. The cells sense bacterial infection and then commit suicide to mount inflammatory response to defend against the infection. The second misunderstanding is that because in 1996, Atul Sekhriinsky's lab showed that this cell death requires caspase, specifically caspase-1. So for some people may know that in the '90s, caspase family is tightly associated with apoptotic cell death. So the caspase family has been deemed as a hallmark for apoptosis. So for a long time, this cell death was also disregarded as a special form of apoptosis. So indeed, caspase-1, which is key for pyroptotic cell death historically. So caspase-1 indeed was the first member of the caspase family discovered. And it was originally purified as the enzyme to process interleukin-1 beta in these two papers published in 1992 as a very classical biochemical fractionation study are very beautiful. So those investigators, they want to understand how the pro-IL-1, interleukin-1, is converted into mature IL-1 by a proteolytic cleavage. And they identified the proteolytic enzyme. The original name was interleukin-converting enzyme, which we now know as caspase-1. So by origin, caspase-1 was discovered in the context of inflammation. Cytokine biology. So it also processes into leukin-18, later on other people showed. So basically caspase-1 can induce pyroptosis and also process into leukins. So if we think this cell death is a host immune defense response, so we needed some oxygen sensors to activate caspase-1 in response to sensing of bacterial infection. And this goes to this so-called canonical inflammasome pathway. So the inflammasome is a cytoplasmic multi-subunit complex. And among the complex, we have receptors such as NRP3, NOL-like receptors. And they will recognize the infection signal or danger signals in the cytosol and then recruit adapter protein called ASK and then further recruit caspase-1 to form the inflammasome complex. So we identified the NAPs, the NOL-like receptors, as receptors for flagellin and the type 3 secretion apparatus in gram-negative bacteria. And we also identified and characterized the pyrin chromosome for sensing the bacterial toxin activity. Now, that was published a long time ago. And quite a few years ago, Joshua Elkington and other labs, and our lab, showed that caspase-11, which is very similar to caspase-1, is the most closest homologue or parallel of caspase-1. Caspase-11 in humans is 4N5, and those caspases, they can directly recognize lipopolysaccharide in the cytosol and then also stimulate pyroptosis but without the processing of interleukins. So this is so-called a non-canonical inflammatory pathway mediated by LPS activation of caspase-11 is only for pyroptotic cell death but not for cytokine maturation. As I mentioned, that this cell death was long time was regarded as mixed with apoptosis. So in this movie, I will just illustrate to you what is really the difference between apoptosis and pyroptosis. So on the left side, you will see the cells undergo typical apoptotic cell death. You see that the apoptotic cells, that they shrink, the cell body fragmented into small vesicles called apoptotic body. And the key, the critical information here is that for apoptotic cells, the membrane integrity is maintained. So without disruption of the plasma membrane. So, and at the same time, you have phosphatidylserine flipping from the inner leaflet of plasma membrane to the outer leaflet. So that serves as an eat-me signal for macrophages to clear the apoptotic cells, the debris of apoptotic cells. So basically, apoptotic cells will not trigger inflammation. It's a clean desk. On the right side, you see the cells undergo pyroptosis, and the cells do not shrink, but instead they swell. You see bubbles blowing out from the plasma membrane, and eventually the cells undergo burst with membrane lysis and then release cytosolic contents to trigger strong inflammation. Right? So, and also you release IL-1 beta and IL-18. So we now think is that apoptosis is more a clean desk for organism development, is more physiological response, while pyroptosis is more like a pathological response. And it is when the organism have a stress and then it may undergo pyroptosis to send out the signals to the immune system. To induce inflammation. So what is the execution mechanism for pyroptosis? And it requires caspase-1 or caspase-11, right? So 5 years ago, more than 5 years ago, our two graduate students, Tianyuan Shen and Yue Zhao, in the lab, they took advantage of the newly coming, at that time, the CRISPR-Cas9 technology and performed a genome-wide screen to look for genes that are required for caspase 1 and 11 to induce pyroptosis. And both screens hit the same gene called gastrin D. So basically gastrin D is absolutely required for both caspase 1 and caspase 11 to induce pyroptosis. So gastrin D encodes a protein about close to 500 amino acids. And three-dimensionally, you can see that it has two domains, N-terminal domain colored in blue, C-terminal domain colored in green. And these two domains actually bind to each other and they keep the protein staying in an auto-inhibited state. And between the two domains, there's a long linker region and there's a caspase cleavage site after this FLT-D sequence. So you can see, you can imagine that two caspases will cleave the protein in the middle. And then this N-terminal domain, when encounters the membrane, will undergo a dramatic conformational change to form this long, 4-stranded beta sheet. At the same time, this N-terminal domain will oligomerize and form pores, form a giant pore, and get inserted into the membrane. As you can see in this negative stain EM picture, there are multiple pores formed. By the N-terminal domain of gastronomic D. And this is the cryo-EM structure determined by Hao Wu's lab at Harvard. You can see there's a very giant pore. The inner diameter is about 18 nanometers. For each pore, you have 24, 27, or 28 protomers, this N-terminal domain. So, and the pore is quite large, as I mentioned, and we now know that this pore serves as a conduit for IL-1 beta or IL-18 to be released from the inflammatory cells. So, when a plasma membrane has multiple Gastrin-D pores, and it will disrupt the osmotic potential, and the water will come in, and the cells will swell, and eventually undergo burst. So, identification of Gastrin-D and the elucidation of the biochemical mechanism, basically, it gave a completed picture and what is really the pyroptosis. So pyroptosis is executed by a pore-forming protein, or gastrin-D, that is activated by cleavage by caspase-1 or -11. So how do caspase-1, -11 become activated to recognize gastrin-D? And I'll just very quickly summarize this paper to you, published last year. It was mainly done by a graduate student, Kun Wang, in the lab. So we determined the crystal structure of caspase-11 in complex with gastrin-D. So here is the C-terminal domain of gastrin-D, the inhibitory domain, right? And the caspase have this little— the dimer— caspase is a dimer— form this little beta sheet. To recognize the C-terminal domain of gastrin B. There is a really strong hydrophobic interaction. And because of this strong hydrophobic interaction, basically the caspase can recruit, recognize gastrin B specifically and easily. And this will allow and push the cleavage site, this FLT3 sequence, getting inserted into the catalytic pocket. Of the caspase. So, and this renders the tetrapeptide independent of cleavage. So as you can see in this data, if we mutated this FLT-D sequence into triple alanine followed by the aspartic acid, because all the caspases cleave after aspartic acid, and the dogma says that the, the, the, uh, three amino acids preceding aspartic acid is critical for for the substrate to be recognized by the caspase. However, when we mutate the FLT3 into triple alanine and do the in vitro cleavage by caspase 11 or caspase 4, you see that wild-type gastrin-D gets cleaved into two fragments, and the triple alanine mutation is comparably cleaved. There's little difference between wild-type and the triple alanine mutant of gastrin-D. And in the cells, you can see that the triple alanine mutation has no effect. The cells are still undergoing cell death with cleavage of gastrin-D. So structurally, basically, this exocyte-mediated enzyme-substrate interaction renders tetrapeptide-independent cleavage of gastrin-D. So the same mechanism also applies to caspase-4 as well as caspase-1, as shown by this co-crystal structure. Of caspase-4 and caspase-1 in complex of gastrin-MD is very much almost identical enzyme-substrate recognition mechanism. So as I mentioned, that pyroptosis was originally discovered in bacteria-infected cells, right? Cells undergo pyroptosis in response to bacterial infection. And this is not a bacterial virulence mechanism, but instead is a host immune defense response. So we made a knockout, gastrin D knockout mice, and infected mice with some gram-negative bacteria like Chromobacter violaceum, which is a gram-negative pathogen. And this bacteria is recognized by the NAP1LC4 caspase-1 gastrin D inflammasome pathway to induce pyroptosis. You can see that wild-type mice can can deal with, can resist, clean the infection. But if we knock out any gene in this canonical inflammasome pathway, the caspase-1 canonical inflammasome pathway, we knock out NRC4, if we knock out gastrin D, the mice become susceptible to chromobacter infection. They died within 1 week. And this is because you have overgrowth of the bacteria. In the liver and the spleen of the infected mice. So basically, if you knock out this pathway, Gastrin D or Caspase 1, the mice cannot clear the infection, and the bacteria, they just replicated heavily in the mice. And this is also true for the non-canonical inflammasome pathway. Using Burkholderia tereodensis as a model, you can see that Buccodaria, the wild-type mice can clear Buccodaria infection by the caspase-11, not compromise and succumb to the infection. And Buccodaria is an intracellular bacteria, is gram-negative, will leak out LPS to be recognized by caspase-11 and gastrin D. And there is a similar bacteria called Shigella, which is we have a long-time interest into this organism, Shigella flexneri, is also intracellular free-living bacteria, which inevitably exposes LPS to the cytosol, to LPS, to caspase 11. However, the wild-type caspase 11 mice showed no difference in response to Shigella infection. Both of them succumbed to Shigella infection.. And we were quite intrigued by this. And to cut the long story short, we found that Shigella has a virulence factor, a type 3 effector called OspC3. And this effector can, using a sugar donor, a sugar ligand, to post-translationally modify caspase 11 and caspase 4 and inactivate this caspase to block gastrin D activation. So if we delete OspC3 from C. galla, in this green line, you can see that when we infect the mice with the mutant bacteria, now the mice can survive very well and can clean the infection. And the bacterial load, you know, the bacterial load of the mutant OspC3 mutant bacteria is very low. Compared with that in the caspase-11 knockout mice. So, and this study highlights the importance of the LPS-caspase-11 pathway for defending against intracellular gram-negative bacterial infection, because that's why C. elegans has evolved a specific viral factor to shut down the caspase-11 pathway. And now using this OspC3 mutation, the mutant bacteria, we found that if we— we can see that the mutation can cause increase the antibody production, anti-Shigella antibody production. Because if you delete OspC3 from Shigella, now the host can activate the caspase-11, induce pyroptosis, in the mice. And after activation of this LPS-Caspase-11, caspase D, pyroptosis pathway, the mice can have a higher level of anti-circulatory antibody. And if you do the same experiments in Caspase-11 or Caspase D knockout mice, you don't see the increase in the antibody production in response to the Aspergillus C3 deletion mutant infection. And if we re-challenged the mice after the OspC3 infection, and the mice are better protected from re-challenged by wild-type bacteria. So, and this also depends on Gastric Phase 11 or Gastrin D. So basically, this study suggests that if we immunize mice with the attenuated OspC3 deletion signal mutant, and this will give us a vaccine-like protection. In a caspase-11, caspase-2-dependent manner. So this study shows that this non-canonical LPS caspase-11 pyroptosis pathway can promote anti-Shigella antibody production, which can protect mice from Shigella reinfection. So basically, these two infection studies highlights the importance of pyroptosis, the protective function, in the immune protective function of pyroptosis, right? However, we know that the immune response is always a double-edged sword, right? If the pyroptotic program goes excessively or is out of control, they will bring deleterious effect. Now, for a typical example is that if we inject mice with very high dose of lipopolysaccharide, could be as high as 50 milligrams per kilogram. And this will induce endotoxic shock, which mimics septic shock in humans. And you can see that wild-type mice die within 1 day in response to this high-dose endotoxin injection. But the Caspase-11 and Caspase-11d knockout mice, they can stay alive very well. So this data suggests that systemic pyroptosis is likely the key for septic shock.. And now we can— we think that if you block this pathway, either inhibiting caspase-11 or inhibiting gastrin-D, you may have a chance to develop antiseptic, a new antiseptic drug. And this has now been pursued in the biotech industry. What's more interesting is that we not only have gastrin-D. Gastrin-D indeed belongs to a larger gastrin family.. In humans, there are 5 members: gastrin A, B, C, D, E. And gastrin A has 3 genes in mice, A1 to A3. Gastrin C has 4 genes expanded into 4 genes in mice, for C1 to C4. And this family is highly conserved in lower organisms, and even coral and hydra. And now there are also papers showing that this family, this gastrin-like protein can also be identified even in bacteria and fungi. So all the family members, they share this two-domain architecture, N-terminal domain that can make pores that is normally inhibited by the C-terminal domain. But the difference is that other gastrin family members, they cannot be cleaved by caspase-1, caspase-11, but they can be activated by any other protease cleavage. So as long as you can cleave any place within this linker region, or you have another way to break up the two domains, to break up the autoinhibition, you can activate the pore-forming domain. So we think that for pyroptosis, the key, the hallmark is not caspase. It is gastrulina family of pore-forming protein. That is the hallmark of pyroptotic cell death. So now I'm going to tell you, summarize the story on gastrin E, another gastrin family member. So now, getting back to the apoptosis studies, we know that caspase-3 can activate apoptosis in response to DNA damage or in response to caspase-8 or TMF receptor activation. So, and this is true that if we damage the DNA, by treating cells with chemotherapy drugs such as cisplatin or doxorubicin, and the cells will undergo apoptosis. And you can see this very beautiful apoptotic morphology in HeLa cells. And this is due to activation of caspase-3. However, when graduate student Yipeng Wang expressed gastramine E, gastramine E in HeLa cells, generate stable cell line. We do the same kind of stimulation, and we did not see apoptosis, but instead the cells develop those characteristic bubbles and undergo burst, eventually die out of pyroptosis. And this experiment suggests that an ectopic expression of gastruline E in HeLa cells will switch, switch caspase-3-mediated apoptosis to Gastromycin E mediated pyroptosis. And why is that? And this is because Gastromycin E has the caspase-3 cleavage site in the interdomain linker region. And indeed, in vitro, we can show that caspase-3 can cleave, but not other caspases. Caspase-3 can efficiently cleave Gastromycin E in vitro into two domains, N-terminal pore-forming domain, C-terminal inhibitory domain. If we mutate the aspartic acid residue into alanine, we can block the cleavage, right, as expected. So now we can put the mutation into the HeLa cells to block a caspase-3 cleavage. So you can compare that this is the previous experiments that gastrin E expression switched into pyroptosis, but if we make a mutation the cells go back to apoptosis. So this data suggests that it is caspase-3 cleavage of gastrin-D that drives the cells into pyroptosis rather than apoptosis in the native HeLa cells. So, and this is because pyroptosis has much faster kinetics than apoptosis. So if we have caspase-3 activation, you go into apoptosis in the absence of gastrulina E. However, in the presence of gastrulina E, it will be quickly cleaved by caspase-3, and then it will switch the cell death to apoptosis because of the faster kinetics. So, and in fact, gastrulina E is not expressed in many cancer cell lines. In this NCI-60, we profiled, we have profiled probably 100 cancer cell lines. Only about less than 10% of cancer cell lines have gastrin E expression. You can see most, most of them do not express gastrin E. And this is because epigenetic silencing. So during cancer progression, tumorigenesis, the promoter of gastrin E, promoter of GASTR is methylated, is epigenetically modified, and then the gene gets silenced. But in some normal cells, in some normal cells and non-cancer cells, and this shows 5 primary human cells, you can see that 3 of them highlighted in red, such as epidermal keratinocytes, placental epithelial cells, they express endogenously GASTR. But these are umbilical vein endothelial cells. Endothelial cells, they do not express gastruline. So for the negative, gastruline-negative cells, if you treat it with DNA-damaging drugs such as doxorubicin, they undergo apoptosis, just as the dogma or textbook tells you. However, in these gastruline-positive cells, you treat it with doxorubicin or other chemotherapy drugs, you see the big bubbles coming out. They show very characteristic pyroptotic morphology. So this data suggests that in our body, we have normal non-cancer cells, some of them express endogenous gastrin E, and those cells will develop pyroptosis in response to DNA damage, in response to caspase 3 activation. This data also suggests that those DNA damage chemotherapy drugs, or even maybe radiation therapy used in cancer therapy clinics. And those agents' treatment, they induce cell death in vivo, not only involves apoptosis, which is a traditional view, but may also induce pyroptosis in vivo in cancer patients. And those cells, they are different in vivo from when we grow them in the culture system. So this slide, diagram, just summarizes to you that in cancer cells, they have epigenetic silencing of GASTR-1E expression, and they undergo apoptosis in response to DNA damage by chemotherapy drugs. However, we have normal cells or some outlier cancer cells, they have GASTR-1E expression, and in response to DNA damage by those chemo drugs, they develop pyroptosis, which is highly pro-inflammatory and causes tissue damage, inflammatory damage. So this pathway have— I think we have two functions, and it may— it probably contributes to the toxicity of those chemotherapy drugs. And it also may stimulate immune inflammation, immune response, and stimulate anti-tumor immunity, which I will show you later on. So we have two gastramins, gastramin D from the inflammasome pathway, which is activated by caspase 1 and 11, cleavage in the middle, and then release the pore-forming domain to induce pyroptosis. Gastramin E is activated by caspase 3, which probably functions in many biological, pathological contexts. And the consequence is the same, it will release the pore-forming domain and induce mediated pyroptosis. So the concept I want to share with you is that we usually tackle caspase, we usually tag caspase with some kind of biological function, which is not true. Caspase is simply a biochemical activity, just like a kinase. And think about it a little bit. So it's just a simple biochemical activity. It does not equal apoptosis.. And it's not the caspase, but it's the substrate that eventually determines the nature of cell death. Right? So if the caspase cleaves a pore-forming protein, such as gastrin D or gastrin E, then the cells will undergo pyroptosis. Right? So that's, I think, is a very important concept. We also found that many gastrin, particularly gastrin E and gastrin A and C, it's very hard to find their expression in cancer cells. They seem to be silenced, tend to be silenced in cancer cells. But why is that? So in this collaboration work with Dr. Zhibo Liu at Peking University, and he's a chemical biologist, and he developed a very sophisticated chemical linker And that can be, the linker can be cleaved by a small molecule, a phenylalanine mimetic called a tri-fluoride phenylalanine. So you can just imagine that this chemical linker is just like a rope. We can use this rope to lock gastrin, to link gastrin to nanoparticle. Okay, so nanoparticle will be engulfed or endocytosed by mammalian cells and particular cancer cells, right? And this gastramine is conjugated to nanoparticle and goes inside cytosol through the endocytosis pathway. And then we can have this small molecule which just like a phenylalanine can be imported into cytosol through amino acid transporter. So the lysine and this small molecule will cleave the linker, just unlock it, right? And then gastrin can be released and to make pores in the plasma membrane and induce pyroptosis. The nice thing here is that this nanoparticle conjugate and the small molecule, when you inject them into mice, they converge onto the tumor site. So that's why they call— we call it as a biosogonal reaction for tumor-selective protein activation, right? So if you inject the nanoparticle conjugate gastramine into mice, and then you inject the mice with a small molecule, and the reaction, the cleavage reaction will only occur on the tumor side. So we did this syngenic tumor model, this 4T1, a triple-negative breast cancer tumor model in the mice. And we, and then at day 6, we inject the mice with nanoparticle and then followed by the small molecule to release the caspase domain to induce pyroptosis. Then we did 3 rounds of induction of pyroptosis. And after 3 rounds of induction of pyroptosis, you can see that— we see a dramatic, very dramatic tumor shrinkage, tumor regression phenotype, and statistics show it's very striking. And we have all the other controls, we can even make a point mutation to knock out the pore-forming activity, and we don't see the tumor shrinking phenotype. And then either in the intratumor injection or intravenous injection, have the same effect. We also see the same effect in other tumor models, such as the EMT-6, another breast cancer tumor model. So you can, you can still imagine that, you probably think that the tumor shrinkage is caused by the tumor cell pyroptosis, per se, right? However, this is only partially true, or is not true. So this method, this chemical biology method that allows us to activate it, release gastrin specifically in the tumors in the mice. This method works highly efficiently in the test tube, this chemical biology reaction. But in the cell culture, we can only induce around 30% or 20% of cell death. So you can imagine that the cells has to endocytose the nanoparticle., and the same cells has to uptake the small molecule, right? And then the cleavage reaction can have to occur in the cytosol to release sufficient amount of gastrin to kill the cells. So you can see that, you can imagine that the efficiency is not, will be very high. So in the in vivo, we can only kill about 10% of tumor cells in the mice. But if we only kill 10% of tumor cells, where is the rest of 90% of tumor cells? Why they did not grow into tumors, right? So then we then did the experiments in the asymmetrical T-cell deficient mice. So in the wild-type Barmy C mice, the previous experiments, you see, when we induce pyroptosis, the tumor did not grow. We see a dramatic tumor shrinkage phenotype. The same experiments when we did in the ACEI-mega T cell deficient mice, there was absolutely no tumor shrinkage phenotype. So this data suggests that the tumor shrinkage caused by pyroptosis, a few cell pyroptosis, cancer cell pyroptosis, requires the T cells. So basically, we think that a few tumor cells undergo pyroptosis, And pyroptosis triggers inflammation, is part of the immune defense response. It would trigger inflammation and remodulate the tumor microenvironment, and then that will activate the T cell media and the tumor immunity. So even when we reduce the 3 rounds of pyroptosis induction in this 4T1 tumor model into 1 round of pyroptosis, it is not efficient. The inflammation is not efficient to induce sufficient antitumor immunity. So we know that this 4T1 tumor model is a cold tumor. It cannot be— is not sensitive to the checkpoint blockade of the NAPD1 treatment. So you see, we treat mice with NAPD1 antibody, there's no tumor shrinkage phenotype. However, after one round of induction of pyroptosis, we then are followed by, anti-PD-1 treatment, we now see a dramatic tumor shrinkage phenotype. So this data suggests that a little bit of tumor cell pyroptosis, although by itself is not sufficient to trigger a strong and robust anti-tumor immune response, but it can convert this so-called cold tumor into a hot tumor so that the tumor can now be sensitive to the checkpoint blockade. Therapy. So this data suggests that the gastrin activation-induced pyroptosis can trigger a very potent antitumor immunity. This explains that why the tumor cells tend to silence gastrin expression, because probably during tumorigenesis, there's a reprogramming, epigenetic reprogramming, And then they want to— the tumors want to silence those pore-forming proteins, gangstermins. And this probably explains very well that in the clinics that— and now there are a lot of combination trials using checkpoint blockade with chemotherapy or even radiation therapy that seems to have very good clinical efficacy. Because we show that the DNA damage induced by chemotherapy and/or radiation therapy can induce gastrin E activation. And gastrin E activation basically will give the same kind of pyroptotic response. And lastly, I want to share another piece of work that in this work, we want to understand that does gastrin— we asked the question, does gastrin-mediated pyroptosis normally play a role in anti-tumor immunity. So why we want to ask this? So we know that in anti-tumor immunity, downstream of TCR recognition of the MHC-presented tumor antigen, when you have PD-1 to release the checkpoint, right, and then that will activate the cytotoxic T cells. The cytotoxic T cells will then uses granzymes to kill the target tumor cells. So granzymes are sort of the executioner protein secreted by the CTLs and delivered into— through perforin-mediated delivery into the target cells, into tumor cells. And then granzymes are serine proteases. There are 5 granzymes in humans, 11 in mice. And historically, granzyme-mediated killing was thought to be apoptosis. Because in the '90s, the immunologist was highly influenced by the cell death people, that the cell death people think that granzymes are pretty much alike or similar to caspase. And the caspase induces apoptosis, and the granzymes probably induce, also induce a similar cell death like apoptosis. Indeed, granzyme B, the best study, the most extensive study of granzymes can activate, cleave caspase-3. So, and the textbook, immunology textbooks, tells you that granzymes will induce apoptosis and clear target cells, including tumor cells, through apoptosis. Is this really true? So in this experiment, I'll just show you that we use NK cells as a model. We can also use the CTRs. NK cells does not require TCR. Recognition, so it's relatively simple. So in 293T cells, we use NK cells to kill 293T cells in an in vitro co-culture system. So we loaded a fluorescence dye into 293T cells, and the dye gets in, then cannot get out. It's a green dye, unless the cells undergo lysis. So in the left side, you see the small cells are NK cells, they're trying to attack the 293T cells, but we did not see the green fluorescence leakage from 293T cells. So there's no lytic death, right? However, when the postdoc, Filo Tso, in the lab here expressed the gastrin B, the third gastrin I talked about today, into 293T cells, we did the same kind of killing, co-culture killing assay.. We also add PI, propylene iodide, into a culture medium. Now you can see that when NK cells attack this gastrin B expressing 293T cells, the green fluorescence disappears and the PI gets into the cells. So this suggests that gastrin B expression, but not any other gastrin expression in 293T cells, can also confer pyrotoxic killing of 293T cells by NK cells, as well as CTLs, which I did not show you. So why is that? So we have 5 granzymes, and to cut a long story short, we found that specifically granzyme A, granzyme A can efficiently cleave gastrin B into 2 fragments, the N-terminal domain and the C-terminal domain. So the cleavage also occurs at the interdomain linker region to release the N-terminal pore-forming domain. So very much like a caspase cleavage of gastrin D or gastrin E. So in the granzymes, specifically granzyme A made by CTLs or NK cells, and it can efficiently cleave and activate gastrin B, and this will release the pore-forming domain and make pores, and the target cells will undergo pyroptosis.. And we found that endogenously, there are 3— here I'm showing you 3 cell lines, and they are digestive tract cancer cell lines. And they have high level of endogenous gastrin B expression. If you give those cells granzyme A, all the cells will undergo dramatic pyrotoxic cell death. And if you make a mutation in this cleavage site, the lysine residues, you can block the pyroptosis, which I don't have time to show you. And so we also tested with cytotoxic T cells using the CAR-T cells, right? We know CAR-T cells has been approved for cancer immunotherapy, particularly the CD19 CAR-T cells. So we use the CD19 CAR-T cells to kill 293T cells that have gastrin B expression. You see that the cells develop those big bubbles, right? You see the leakage of green fluorescence from the 293T cells. And if you knock down, as long as you knock down granzyme A, you can block the— diminish the pyroptosis killing. And if you mutate the cleavage site, and there is no cleavage of caspase B, then you don't see the pyroptosis.. And we show the same kind of pyroptotic killing by TCRT cells. You see that in this melanoma cells, A375 melanoma cancer cells, and in response to TCRT cells killing, gastrin B gets cleaved and the cells undergo pyroptosis, lytic death. If you mutate the lysine residue, you don't get cleavage, you block the TCRT killing, pyrotoxic killing of melanoma cells. So gastrin B, where is it expressed? And we found that it's highly expressed in the entire digestive tract, starting from the lung to even to the appendix. It's also expressed in the lung epithelial cells as well as bladder epithelial cells.. And it's very interesting, we also found that gastrin B is transcriptionally induced by interferon gamma, and to some extent, TNF-alpha. For example, in ST29 cells, you don't see gastrin B expression, but if you treat it with interferon gamma or TNF-alpha, you see a dramatic transcriptional induction. And the same applies to many cancer cell lines. So gastrin B actually is a transcription target of interferon gamma. So we know that the cancer and the tumor immunity is driven by interferon gamma. So T cells secrete interferon gamma, and interferon gamma acts on tumor cells. It can induce PD-L1 expression. It can also induce gastrin B expression. So when gastrin B is expressed and the granzyme A coming from cytotoxic T cells was specifically cleaved the linker region in gastrin B and release the gastrin B and make pores, and the targeted tumor cells will undergo pyroptosis. So the most important thing is not only the death of tumor cells, and this death is different from apoptosis. It will trigger inflammation, it will remodulate the tumor immune microenvironment, and to further amplify the immune response to further stimulate antitumor immunity. So in this CT26 colon cancer tumor syngenic model, we also see gastrin B-mediated pyroptosis has a synergistic effect with NPD-1 checkpoint blockade treatment. So this data suggests that gastrin B, as well as gastrin E, and the work from Judy Lieberman's paper show a very similar story that Gastramine E in response to granzyme B activation also plays a role in antitumor immunity. So we think those gastramines may serve as biomarkers for cancer immunotherapy. If cancer cells express those pyrotoxic factors, they may have a better chance to respond to immunotherapy. And if we can develop small molecules to activate gastramine, It may have— it could be a next-generation cancer immunotherapy reagent. So I want to acknowledge the students and postdoc fellows in the lab who did all the work. Tianyu Zhen, Weizhou, they originally identified Gastrin D, and Yipeng Wang, Wenqing, and Shuyuan, they did work on Gastrin E. Jingjing Ding is a postdoc fellow, showed the pore-forming activity, and together with Hong Wang and postdoc fellow Qi Sun did the structural work. And the last piece of work on gastrin B was done by postdoc fellow Zhu Weizhou and graduate student Hua Bing. And Qian Wang from Dr. Tsubori Oh's lab did collaboration work on the chemical biology delivery of activation of gastrin.

And Fengchao Wang made knockout mice for us. And I'll stop here. Thanks for listening. Thank you so much, Feng. This is— I feel humbled and inspired by the tremendous work you have done in this field. Really, each of the discoveries you described today are so impressive on their own, so I'm sure the rest of the attendees will feel the same way. And now I would like to guide them to take the opportunity to interact with you in Twitter and ask questions. That can be done in an asynchronous manner. And so if you are watching this talk on YouTube, you can ask questions in Twitter after you watch the talk later.

And it's late in Beijing, so probably Feng will be answering questions tomorrow or—

Tomorrow, yeah, I'll do that tomorrow. Sounds great. Thank you. And so to ask questions, search for the account Global Immuno Talks, find the tweet that says, ask questions for Dr. Feng Zhao here, and reply to that tweet with your question and mentioning the hashtag Global Immuno. And then Feng will use the account Immuno Speaker to answer to the questions. And with that, we will close the session for everyone, and I hope that you can join us again next week.

So, Feng, again, thank you so, so much for presenting and being a Global Immuno speaker today.

Thank you so much.

Thank you. Thank you.

Yeah.

Bye. Bye, everyone. Thank you. Yeah. Have a good day.