Hello everyone. I’m Jianjin Shi, a PhD student from Dr. Feng Shao’s lab at National Institute of Biology Sciences. What we study is a deadly disease called sepsis. Sepsis can kill many people around the world, but currently we do not have any drugs to treat this disease. And what we study, and what we’ve discovered, might provide new treatment for this deadly disease. To begin with, I would like to introduce… what is sepsis? Sepsis, by definition, is systemic inflammation caused by infection. As shown here, this old man had a tooth infection, but didn’t get proper treatment with antibiotics, and later this infection spread into the bloodstream and his body mounted a very strong immune response, just like unleashing a very strong army. Sometimes, this army can cause collateral damage to his own body and lead to sepsis. So, what do you feel after you have sepsis? You may feel fever, chills, you can breathe rapidly, and your heart beats rapidly, and you can may experience confusion, disorientation, as well as nausea and vomiting. So, why is susceptible to sepsis? In general, people with weakened immune sytems are more susceptible to sepsis, including the elderly, pregnant women, children and infants, and people with chronic illness, including AIDS, diabetes, and cancer. This man lost his spleen after an accident and therefore lost the majority of his immune system, and several years later he had sepsis. And this little girl also had sepsis because her immune system is not strong enough yet. Luckily, these two people survived sepsis. But sepsis doesn’t only kill susceptible people, it can kill anyone, including you and me. This young lady, she was a model from Brazil, and she was healthy. In the beginning, she had a urinary tract infection, which is a quite common infection, and later the infection developed into sepsis, and within only a few days she died in the hospital at the age of 20. And one scary thing for sepsis is that sepsis can kill people, and kill people rapidly. So, why should we care about sepsis? In the United States alone, over 750,000 people develop sepsis. Among them, over 200,000 people die. That is more than the population of Salt Lake City. And sepsis kills more people than breast cancer, colon cancer, and AIDS combined. And sepsis also accounts for at least one third of all hospital deaths. Because sepsis generally need special care in the ICU, it is a very expensive disease, and in the United States alone in 2011, sepsis cost more than 20 billion US dollars. But most people don’t even know about this disease. And sepsis has three main stages: in the beginning, it’s called sepsis. You have a local infection that’s in the lung or other places, and this infection breaks your immune defense and gets through to the bloodstream, and your body mounts a systemic inflammatory response. So, this is stage one. And in stage 2, you will experience several organ dysfunctions, and this is called severe sepsis. And in the last stage, people will have multiple organ failure, as well as a sudden drop of blood pressure, and this is called septic shock. And for septic shock, there will be a 50% mortality rate. One thing I want to mention is that sepsis is caused by our own reaction to the infection, but not due to the pathogen. Unlike other deadly diseases, sepsis has no drugs. Even after decades of clinical drugs, none of them succeeded. So, I think that most important thing is that we don’t know enough about sepsis. So, what causes sepsis? This is still an open question to the scientific community, and I think maybe… I think scientists are making progress now towards understanding what causes sepsis. It begins with the late 19th century. Then, Richard Pfeiffer, a German military doctor who worked with Robert Koch, a very famous bacteriologist, at that time what they found is that injection of heat killed bacteria, which caused cholera… and injection of dead bacteria can cause sepsis in guinea pigs. He then hypothesized that there are some toxic substances inside the dead bacteria. It took about fifty years to find this toxic substance, the lipopolysaccharide, or LPS for short. LPS is the major component of the cell walls of nearly all Gram-negative bacteria. As shown here… this is a scanning electron microscopy picture of E. coli, a Gram-negative bacteria, and this is what the cell wall looks like. The most abundant molecule on the cell outer membrane is LPS. If you zoom in, this is what the LPS molecule looks like. It contains two parts: one is the sugar part, the other is lipid A. The lipid A, as shown in yellow, is the active part of this LPS molecule. If you inject LPS or lipid A into mice, those mice will develop sepsis. So, the question is, how can LPS cause sepsis? If you remember what I told you before, sepsis is caused by our own reaction to infection, not by the pathogens. So the similar question is, how can we respond to LPS? This is the pathway that people think plays an important role in LPS sensing. The membrane-bound receptor, called TLR4/MD2, can directly interact with LPS molecules, just like the eyes of a cell. When the eye sees an LPS molecule, it can trigger the expression of a series of proinflammatory genes, including cytokines. And people use to, for a long time, thought about that… that these cytokine armies are actually the cause of human sepsis. And more than ten clinical trials have been performed with sepsis patients with targeting these cytokine army molecules, but none of them have succeeded, and later on people thought about… what about inhibiting this eye? If you blind the eye of the cell to LPS, then you will block all the cytokines from being produced. So, this is a molecule people developed to inhibit TLR4, the eye of our cell, and indeed this molecule can inhibit the production of these cytokine armies, but after years of trying and after a billion dollars spent on this project, and after recruiting of thousands of people on these clinical trials, this is what they got. As shown here, by looking at the survival rate of sepsis patients, the Eritoran, the drug that can blind the eye to LPS, saves no more people that than placebo. So, it failed. But why? Did we miss something? This is indeed the case, as shown here. Recently, people have identified another pathway that can recognize LPS, inside the cell, in mouse macrophages. In this pathway, there is an unknown LPS sensor, which can recognize LPS and lead to a signal to a gene called capase-11 in mouse macrophages, and caspase-11 is a proinflammatory caspase. It has two domains. One is a CARD domain at the N-terminus, and the other is the protease domain at the C-terminus, and the protease looks like molecular scissors that can cut through other protein substrates. And activation of caspase-11 can lead to proinflammatory cell death, and this cell death may eventually lead to sepsis, because this cell death has a much stronger effect than the production of the cytokine armies. So, I will call this cell death “unleashing the special forces”. This is what this cell death looks like. As you can see here, in the beginning cells look fine. Then all of a sudden, the cell just blows up and it releases all of the cellular contents, which is very proinflammatory. You can see nearly all of these cells died after triggering this proinflammatory cell death. The importance of this pathway was further highlighted by the fact that caspase-11 knockout mice are resistant to sepsis. As shown here, wildtype mice… nearly all wildtype mice died within a day after you induce sepsis with LPS, and caspase-11 knockouts are still very resistant to this disease. And thinking about that… caspase-11 has a functional TLR4, which is the eye that can recognize LPS outside the cell. This suggests that this pathway that senses LPS inside the cell plays a more important role in sepsis, at least in mice. So, let’s summarize what we have known before we get into this field. There is a pathway that can recognize LPS when LPS gets into the cell, and this sensor is not known. And this sensor can activate caspase-11 in mouse macrophages and lead to proinflammatory cell death, and therefore may trigger sepsis. But we care more about humans, and because humans do not have caspase-11 genes, so the first question we want to answer is, does this pathway, which senses LPS once it gets into the cell, exist in humans? To begin to study this question, we need an efficient method to deliver LPS to the inside of the cell. This is what we use, called electroporation. If you put a cell into an electric field and then give an electric shock, you can punch holes in this cell membrane, and then these molecules that are outside the cell can get into the cell through these holes. And this is what will happen to the cell membrane. In the beginning, the cell membrane is okay, and after electroporation you can really see these holes on this cell membrane. And after electroporation, and more importantly, the cell can recover from this electroporation. So, this is the method that was used to deliver LPS, and if you deliver LPS to a human immune cell called U937, which is a human monocyte, you can see here the cell just blows up. It pretty much looks like the mouse macrophages. So, this blow up will release all the cellular contents, unleashing the special forces, which can trigger sepsis in humans. And if you perform electroporation with control ligand, then the cells are just fine. We can also measure the cell death. You can see here, if you deliver LPS into the cell, it can cause about 80-100% of cell death, but control ligands do not cause any cell death. So, in human, we do not have capase-11, but we have two other closely related genes, called caspase-4 and caspase-5. We first detected the expression level of these two genes in human monocyte cell lines. As shown here, we can easily detect the protein, as well as the mRNA, of caspase-4, but we cannot detect any caspase-5 expression, even using more sensitive methods that detect mRNA. So, the next question we want to answer is, does this pathway depend on caspase-4? We then used a method that is a small molecule that can inhibit, transiently inhibit, caspase-4 expression in human cells. By using this small molecule, we can see here, by transient knockdown of caspase-4, this LPS-induced proinflammatory cell death is totally blocked. So this data suggests that LPS can activate caspase-4 in human cells. And quite unlike the mouse studies… in mouse, caspase-11 is only expressed in macrophages, that is, an immune cell. And for human cells, we have found that several other non-immune cells also have caspase-4 expression, as well as they can also respond to LPS inside the cell. So, let’s summarize the previous two slides. We have found that, in human, we also have this pathway that can sense LPS that gets into the cell, and that this pathway can activate caspase-4, rather than caspase-11 in mouse macrophages. And this pathway also can induce the proinflammatory cell death, unleashing the special forces, and may cause sepsis in humans. This is a possible reason for why the TLR4 blockers failed in clinical trials, because [those drugs] only blocked the pathway that can sense LPS outside the cell, and data from mouse suggests that the caspase-11 pathway, which recognizes LPS inside the cell, plays a major role in sepsis. And it is possible that targeting caspase-4 in humans might be the right target. The next question, and most important question for this pathway, is, what is the LPS sensor? Because if you know what the sensor is, you can design small molecules to inhibit this sensor, just like they did on TLR4. So, after trying and trying… we almost tried everything we could, but we cannot find this direct receptor. Then, one day, after characterizing the biochemical function of caspase-4 and caspase-11, we got some hints suggesting that caspase-4 and caspase-11 might directly bind to the LPS molecule. This is the data and this is an assay called a pulldown. So, there is a molecule A and a molecule B. If you pulldown molecule A and you get both, it suggests these two molecules have an interaction. If you pulldown molecule A and only get molecule A, then it is possible that the two molecules do not bind each other. By using this method, we can see that both lipid A, the active part of LPS, or LPS, can bind to caspase-4 in human and caspase-11 in mouse. And control ligand, which is lipopeptide, or MPD, the muramyl dipeptide, cannot bind to these two caspases. By using similar assays, we found that the N-terminal CARD domain, which represents about 90 amino acids, is the LPS binding domain. As shown here, the full length protein can bind to LPS and the CARD domain can also bind to LPS, but if you delete the CARD domain the C-terminal protease domain can no longer bind to LPS. So, there are no hypothetical LPS sensors other than caspase-4 and caspase-11. The direct sensors are caspase-4 and caspase-11. This is quite surprising and this is the big deal. So, for caspase proteins, no one has shown that this protein can be a direct sensor for a molecule. This is the first case. And the next question we want to answer is, what happens after caspase-4 and caspase-11 recognize LPS? So, this is the data we have. In the normal condition, without any treatment, the caspase-4 protein migrates as a monomer in our polyacrylamide native gels, and if you incubate LPS or lipid A with caspase-4, this protein sticks together to form large oligomers, as shown here. Control ligand does not have this activity. We also have similar results with mouse caspase-11 proteins. So, this is what we see: after binding with LPS, the caspase-4 in human and caspase-11 in mouse become oligomers. If you’re familiar with other caspases, for example, caspase-1, caspase-8, or caspase-9, these proteases are all activated through protein complexes, but not self-assembled protein complexes. So, the next question we want to answer is, does caspase-4 in human and caspase-11 get activated by this self-assembled protein oligomers? This is the assay we used. We first incubate LPS with this caspase-4 or caspase-11, and then monitor the protease activity. Remember, the C-termini of these two proteins are proteases. So, as shown here, incubation of LPS leads to the oligomerization, and by monitoring the protease activity, you can see that, for caspase-11, we have about 20-fold protease activity increase, and for caspase-4 we have about 60-fold protease activity increase. So, let’s summarize what we have found. There is a pathway… when LPS gets into the cell, it can be recognized directly by caspase-4 in human and caspase-11 in mouse, with the N-terminal CARD domain. This activation may well lead to the oligomerization of caspase-4 and caspase-11, and caspase-4 and caspase-11 actually get activated through this oligomerization. Because these caspases are proteases, when they activate they can cleave other substrates, and maybe cleaved substrates may cause something to lead to this proinflammatory cell death. And remember I said earlier that proinflammatory cell death releases the special forces, which may trigger sepsis. Since this pathway is very important in a mouse sepsis model, we have hypothesized that this pathway might also play an important role in human sepsis. So what I’m working on is performing a large-scale small molecule screen which contains 300,000 small molecules. We are aiming to find some molecules that can inhibit this proinflammatory cell death, as well as the whole pathway. We hope, in the near future, our small molecule compound might provide new treatment for this deadly disease. Okay, here’s my acknowledgements. I’d first like to thank by supervisor, Dr. Feng Shao, for his guidance and support during these years. And I would also like to thank my major collaborator, Dr. Yue Zhao, and all other members in our labs. And this is Dr. Feng Shao, and this is Dr. Yue Zhao. And thank you for your attention!