Scientist Stories: Akiko Iwasaki, The Innate Immune System and Defense from Viruses

Colleagues and friends, I'm Michael Kashkarian, the outgoing chair of the trustees of the Cushing Whitney Medical Library Associates. The Library Associates and I are very glad to have you spend your afternoon with us. A few small details. The lecture is being livestreamed and recorded, so please do as I just did, silence your cell phones and/or any other noise-making devices, including your voices, I assume. Please be sure to join us afterwards in the Beaumont Room for a reception. I'm pleased to present our 71st Annual Associates Day lecturer, Akiko Iwasaki. Dr. Iwasaki received her PhD from the University of Toronto and her postgraduate training from the National Institutes of Health. While still a graduate student at the University of Toronto, she was among the first to demonstrate the role of circulating antigen-presenting cells in eliciting the immune response to vaccines, which at that time would always injected into muscle. And people thought that the muscle had Langerhans cells or some other antigen-presenting cell which then would initiate the immune reaction. Dr. Iwasaki joined Yale University as a faculty member in 2000 and is currently the Waldemar von Zetwitz Professor of Immunobiology and Molecular and Cellular Developmental Biology. She's a principal investigator at the Howard Hughes Medical Institute and she was he was elected to the National Academy of Sciences just last year. Akiko Arisaki's research focuses on the mechanisms of immune defense against viruses at mucosal surfaces and the interplay of innate and adaptive immunity. Her laboratory is interested in how our innate immune recognition of viral infections leads to the generation of adaptive immunity and how adaptive immunity mediates our protection against the subsequent viral Please welcome Dr. Awasaki.

Thank you so much. Thank you very much. Can you hear me?

No?

Okay, maybe I'll stand here. Okay, can you hear me now? I'm very honored to give the 71st annual lecture of the Christian Whitney Library Association. Today what I'd like to do is to give you my— take you through my scientific journey by discussing some of the highlights, some of the findings that are made by the members of my laboratory. So today I'd like to give stories of 3 separate topics. The first study has to do with the underlying cause of infectious disease. How do infectious viruses cause cause disease in humans. Second story has to do with the warfare between retroviruses and us. Vertebrates have been colonized by retroviruses ever since the beginning of evolution. And so about 500 million years of battle has been fought between the host vertebrate and the retroviruses. So I'd like to speak on the endogenous retroviruses and how that interacts with the immune system. And finally, I'd like to touch on this very important topic of the is the lack of vaccines against major pathogens. Viral diseases like HIV, herpes simplex virus has no effective vaccines right now. And so that's the title of my talk today. Can we learn from studying viral infections to make better vaccines against these kinds of agents? So how do we develop a new vaccine that actually works? So with respect to how viruses cause diseases, many viruses, can infect the host. And some viruses are completely asymptomatic while others cause disease. It's thought that viruses cause disease by causing tissue damage. So the target cells that are infected by the virus is killed by the infection and therefore causing damage to that particular tissue. However, it is equally possible that it's the immune response to the virus that causes the damage. And understanding which of these scenarios are true for a particular viral diseases is important in trying to deal with such infectious diseases. So the first story has to do with genital herpes infection. So it's been known for many years that herpes simplex virus causes genital herpes by infecting the genital mucosa. And this virus replicates in the epithelial layer of the vaginal tissue. And it infects the innervating neurons, the sensory neurons that innervate this tissue. When it becomes dormant, the virus can establish latency within the dorsal root ganglia, and it can stay there for the rest of your life. It does stay there for the rest of your life. And the activation of this latent pool of virus occasionally back to the genital mucosa is what causes the lesion and disease. And it was thought that this infection is confined to these kinds of tissues, the vaginal mucosa and the dorsal root ganglia. However, work from our laboratory has shown that it's not the case. That in fact the virus can travel from the dorsal root ganglia to the enteric neurons and there infecting the neurons that control the peristalsis of the colon. And that results in the block of peristalsis and therefore mice actually die of toxic megacolon and not viral encephalitis or some other diseases that are caused by the infection of the central nervous system. So this led to this understanding that in fact viral diseases can be extended to unexpected consequences. In this case, genital herpes results in chronic constipation. Now this, when we published this, this was thought of as a sort of a mouse finding that in humans this doesn't happen. But it turns out to be that's not the case. I'm getting emails emails and phone calls from people all the time, almost daily, telling me that they have developed chronic constipation just at the time of their acquisition of genital herpes. And in 3 cases, people have contacted me and they didn't know they had herpes. Sometimes this herpes infection is completely asymptomatic. And these women took acyclovir and cured themselves of chronic constipation. So it turns out that this type of disease may be actually occurring in human hosts. And finding this for an unexpected outcome provides us with a medical tool to deal with these types of infections. Another example that I like to give is how Zika virus causes infection. As you know, Zika virus transmits by the bite of mosquitoes. However, it's also known that it's sexually transmitted to an infected person to an unaffected partner. In most cases, these types of diseases are transient and acute and are cleared in healthy adults. However, in rare cases, people develop Guillain-Barré-like syndrome, which is an autoimmune disease that attacks the peripheral neurons, thereby debilitating the person's ability to walk and control their motor functions. In another case, we know very well from the epidemics in Brazil in 2015 that In pregnant mothers that are infected with Zika virus, microcephaly and other neurologic disease can result from these kinds of infections, as well as cases of spontaneous abortion as a result of Zika infection. So we want to understand how these diseases actually manifest. Particularly, we need to understand how sexual transmission of Zika virus leads to diseases and how the host deals with these kinds of infections. So we asked, how does Zika virus cause fetal demise? And this was a question that was tackled by a graduate student, Morayuki, who developed the first model of vaginal transmission of Zika virus infection whereby she demonstrated that the virus can enter the vaginal mucosa and replicate very well within this tissue. So unlike if you were to inject the same amount of virus into the skin, in which case the virus is cleared, the vaginal mucosa can host these viruses to high titers of replication even in wild-type animals. And she showed that in pregnant mice, if you were to infect those mice with Zika virus vaginally, this infection then results in infection of the fetal brain. And in a follow-up study, she demonstrated that the fetal demise actually occurs as a result of maternal immune response. To Zika virus. So when a mother is infected with the Zika virus, she will produce these type 1 interferons, which are antiviral factors that usually protect the host. Unfortunately, in the case of pregnancy, however, this interferon acts on the fetal tissue, especially the placenta, to arrest its development and thereby resulting in fetal demise. So this work laid a sort of overarching understanding of the mechanism of congenital disease. That can occur as a result of viral infection during pregnancy. The next question is how does Zika virus cause the Guillain-Barré-like syndrome? So our work demonstrated that the central nervous system, including the brain and spinal cord, is in fact infected by the Zika virus. But it's not the virus infection that makes the Guillain-Barré-like syndrome. In fact, it's the cytotoxic T-cell response, which is the immune response against the Zika virus, that results in the killing of the infected neurons, resulting in the paralysis of these animals. So this provided the first insights into how Guillain-Barré syndrome might result from a Zika virus infection by demonstrating it's the cytotoxic T-cell, which is the host immune response, that ultimately results in killing of the neurons that paralyzes these animals. Animals because until then it was thought that it was the antibody cross-reactivity between the Zika virus and the neuron antigen that resulted in this disease.

So I hope I was able to demonstrate that by studying these infectious diseases, the mechanism underlying such diseases, in some cases it's the virus infection that kills the target target tissue such as the enteric neuron in the case of herpes simplex virus that results in the disease manifestation such as chronic constipation. And this is a very unexpected sort of cause of disease for chronic constipation which currently there's very little is understood about how this occurs and millions, actually billions of people suffer from chronic constipation. Whereas in the case of Zika virus, We demonstrated that it's the immune response to the virus and not the virus infection of the target tissue per se that causes disease, both in the case of pregnancy, in eliminating or arresting the fetal development by type 1 interferon of the mother, and in the case of the Guillain-Barré-like syndrome in which the CD8 T cells, cytotoxic T cells, attack the infected neurons. So why is this important to study, disease manifestation? Well, first, in order to treat infectious disease, we need to understand what aspect of infectious disease we need to target. Because the target strategy will greatly depend on whether it's the pathogen, the virus itself that's causing the disease, or whether it's the immune response to the virus. Is causing the disease. And by studying this fundamental infectious disease biology in animal model will give us insights into diseases that were previously completely unexpected such as the constipation resulting from herpes simplex virus. So now I'd like to move to the second topic which has to do with endogenous retrovirus. And the immune system. So over the last 100 years or so, the immunology community has focused on bacteria as pathogens. How do these pathogenic bacteria make us sick, such as Salmonella or E. coli outbreak? However, in over the last 15 years or so, the community has shifted its focus on the microbiome, which is the collection of bacteria that inhabits us, which outnumber us by about tenfold. And so this has brought us enormous insights in how microbiome controls health and disease in variety of settings. However, how the virome controls the host and vice versa is much less understood. Part of the reason is that the microbiome is easy to identify because we can use a specific tag, which is people use 16S RNA, ribosomal RNA sequencing, to identify the nature of the bug that is inhabiting us. In the case of virome, it's very difficult because there are no conserved sequences that we can target to identify the virome and to study them, let alone to eliminate one virome versus another. So this is a question that we became interested a few years ago. And we focus on the most fundamental of all viruses, which is the endogenous retroviruses, because it has integrated into our genome over millennia. And so this is just demonstrating how vast the real estate that is occupied by the LTR retrotransposons, which are remnants of exogenous retroviruses that have integrated into our genome, ancestral genome, compared to the 2% of our genome occupied by protein coding sequences. So the vast majority of the community focuses on this 2% of the protein coding gene, whereas we wanted to study what this 8% of our genome is doing, which is occupied by these retroviruses that have integrated into our genome. And just to point out that not only LTR retrotransposons transposons, but there are vast majority of this genomic real estate, which is about 40%, occupied by these transposable elements, which are jumping genes that can go from one locus to another. And again, much less is studied about what these things are doing in our body. So we focused first on this autonomous LTR retrotransposons, which These look just like the genome of a retrovirus. Essentially, these are the retroviruses that have integrated into our germline ancestors over millions of years, and they basically contain the same kinds of genes that an HIV virus might contain, gag, pol, and envelope, and they are able to express themselves and sometimes recombine with each other. In mice, they can make infectious retroviruses by recombining from one locus to another. And so what we wanted to dig in is the immunology at the genomic level. How does the immune system control these kinds of elements within the— from within the genome? And there are at least two types of ways we deal with this. One is to silence the expression of these retro elements by binding to specific locus, conserved sequences that are found within these retroviruses. And the other is to chop up the nucleic acid intermediate, replication intermediate that are generated from these elements. In both ways, we deal with suppressing the reemergence of these kinds of viruses. So, this question about how does the immune system control endogenous retroviruses at the genomic level was tackled by a graduate student, Rebecca Treger, who solved a 50-year-old puzzle by cloning genes that are responsible for controlling these elements in the mouse genome. And these turned out to be two related CROD-containing zinc finger proteins that can recognize and suppress the expression of these genes. And they turned out to be the locus that controls lupus susceptibility in the animal as well. So there's a link between endogenous retrovirus expression and lupus disease. So the way— to make this long story short, essentially what Rebecca found was that, as I mentioned, she found two different crab-containing zinc finger proteins that have these fingers, literally fingers that project and bind specific DNA sequences. And she found a conserved region near the LTR of vast majority of mouse endogenous retroviruses that can bind and control the expression of these genes. And we named them SNURVs, SNURV1 and 2. So in wild-type animals, animal, this KRAB zinc finger protein can recruit CAP1, which can silence the expression of these endogenous retroviruses by imposing epigenetic marks. And so the basal level of these endogenous retroviral gene expression is quite low. Whereas in the case of lupus-prone animal, these KRAB zinc finger proteins are missing and therefore where the expression of these ERVs are elevated without any control. And this resulted not only the expression of these viral genes on the surface of the cell, but also resulting in the secretion of these viral genes that then end up becoming a target of antibody that recognizes these kinds of envelope genes. And that's what's causing the lupus disease in the animal. Animal. So these nerves that are usually bound to the— our genomic loci at the LTR region of the endogenous retroviruses are missing in the lupus-prone animals, which results in the expression— overexpression of these viral genes, including the envelope gene, which then becomes the target of autoantibody, and it's the antibody antigen complex that deposits into the kidney that causes lupus nephritis in the mice. So this story demonstrated that the epigenetic regulation of ERVs is a critical step in preventing diseases such as lupus. Now, how do we study ERVs in, in the human population? The major challenge The biggest challenge in studying these ERVs is that there is no way to know the identity of many of these ERVs because they're so highly repetitive and similar to each other. There's no way of saying this ERV came from this gene locus versus another. So Maria Tokuyama in the lab solved this problem by making a program called ERVmap whereby you can take your RNA sequencing data, and this is publicly available and free to anyone who wants to use this, and plug it into this website and out comes the sort of identity of all the ERVs that are expressed in a particular sample of interest. And so she made this first tool in order to characterize ERV expression in the genome-wide level, and now we can do some amazing things with the CERV-MAP, which a couple of examples I'll show you. So every cell type turns out to have different sets of ERVs that are expressed to different extent. So you can take a fibroblast or epithelial cells or macrophages, T cells, whatever cells of your choice, and ask what are the ERVs that are expressed. It turns out to be distinct between all of these cell types. In fact, you can use the ERV information alone to project the different subsets of cells that are present in our body. And so Maria was able to show that just by looking at the ERV expression pattern alone, she was able to cluster different cell lines in different locations within these kinds of analysis. And moreover, when she looked at the blood cells collected from lupus patients, she found these ERVs to be highly expressed. So there are about 124 different ERVs that are upregulated in the lupus patient peripheral blood cells. And this is just demonstration of the different kinds of ERVs that are expressed in the lupus patients that are highly upregulated in red compared to the healthy control, which have these dark blue lines. Indicating that they have low expression. And not only that, this kind of links what we found in the mouse model where we demonstrated that in the mouse lupus, it's the lack of these nerve genes that elevates the envelope expression that becomes the target of autoantibody. In the case of lupus, Maria demonstrated similar overexpression of endogenous retroviruses. And something something akin to SNURF is regulating the expression of these genes. We're actively looking for what these genes are. But in any case, these ERVs are elevated, but do they become target of autoantibody? So in order to address this issue, Maria with Armin, who is an undergraduate student in our laboratory, made these proteins that are expressed by the highest expressed ERVs in the lupus patients. And she was able to show that these kinds of antigens that are basically envelope protein from the endogenous retroviruses become target of antibody that can stimulate leukocytes within our circulating blood cells very well, stimulate them to excrete these toxic molecules that would be detrimental to to the host, perhaps driving these lupus-like diseases. So, in this program, what we demonstrated is that we now have a tool to study these endogenous retroviruses, which is the most fundamental member of the virome that we carry. Second, we discover the two genes that control ORF expression in the mouse. Without which we develop lupus-like disease. And finally, we saw similar elevation of ERV expression in human lupus patients, which also become the target of autoantibody reaction. So these collectively demonstrated a link between, again, completely unexpected link between endogenous retrovirus expression to lupus disease in two different types of models. So whether ERV envelope is driving the lupus disease or not, this is a sort of target of our research in the future. And what is controlling ERVs in the human is also unknown at this point. So now I come to the third topic, which has to do with lack of vaccines against very major infectious agents, viruses that transmit and cause diseases. So major challenges in current vaccine strategy for which we don't have successful vaccines such as HIV and HSV is that current vaccines rely on antibodies. And many viruses are known to evade antibody responses. So herpes simplex virus expresses different glycoproteins on the surface that completely block the activity of antibody. HIV can mutate its surface antigen very quickly to evade the recognition by antibody. And influenza virus does this very well. Every year it changes the coat protein slightly so that the previous year's antibody can't detect the coats very well. So all of these viruses have figured out the immune system them way before we even knew antibodies. So that's a problem for developing vaccines against these kinds of agents. And the other thing to remember is that most infections start from a local tissue, likely the mucosal surface of the nose, the genital tract, the gut. These are the mucosal surfaces that are favored by these viruses, and these kinds of current vaccines are not designed to generate antibody responses at these sites. So current vaccines really fail to establish not only antibody but T-cell responses in these sites of infections. However, we don't know whether circulating memory T-cells may be sufficient to control these kinds of infections. If the virus evades the antibody response, immune response? Can we rely on the circulating T cells to protect us from further viral challenge? So we addressed this issue. So in picture form, what we're studying is upon immunization, which is given to the— intramuscularly, we develop T cell responses. T cells circulate throughout the body and seed many organs. But they fail to seed certain organs such as the brain, the skin, the airway epithelia, and female reproductive tract are devoid of memory cells, T cells, that circulate and see these tissues. And this is a major problem because sexually transmitted pathogens enter through the genital mucosa and yet we are not establishing T cell responses at the site of viral entry. Entry for many of these pathogens such as HIV, human papillomavirus, and herpes simplex viruses. They all enter through the general mucosa and yet none of the conventional vaccines are generating T-cell response at the site of entry of the virus. And these are staggering number of people who are currently infected by these pathogens in the hundreds of millions. And so how do we deal with this problem? Well, first we have to ask the question, if we were to just vaccinate people with a conventional vaccine and we generate these circulating memory T cells, are they going to be sufficient to protect the host? So in order to address this, we've developed a mouse model of genital herpes to study this in which we give animals an attenuated herpes virus intravaginally and then we know that that after weeks there are memory T cells that are resident within the vaginal tissue. So what we can do now is to take an animal that's been immunized this way with lots of memory cells in the genital mucosa and fuse it to an animal that has never seen the vaccine before. And this is a sort of an old technology known as parabiosis in which we we suture the skin between the two animals and these two animals share the circulation, the blood circulation. So whatever that's in the blood of this one mouse will transfer to the other. So in this setting, what we have is this mouse which only has circulating memory T cells, whereas this mouse has circulating memory T cells as well as the resident T cells in the genital tissue. So we can use these animals to ask ask whether circulating memory cells are sufficient to confer protection against herpes. So what we can do is, well, first of all, this is a baseline control where if we parabiose naive animals together and then infect one of them with herpes, all of these animals succumb to infection within about a week. However, if we were to immunize the animal in which only circulating memory cells are memory T cells are present, then many of them succumb to infection. Whereas if we were to infect with lethal herpes into the animal that has this tissue-resident memory pool, 100% of these animals are protected. So these experiments demonstrated that circulating memory T cells are only partially protective against genital herpes disease. Whereas the tissue-resident memory cells confer 100% protection against challenge by lethal herpes. And we showed that the reason for the advantage of the tissue-resident memory T-cell is because they can secrete this antiviral protein known as interferon gamma rapidly upon infection. And this timing as well as the robustness of of the interferon gamma response conferred by the tissue resident memory cell is what controls the viral infection very well. Whereas the circulating memory cells, it takes them a while to get to the tissue, find the antigen, start secreting interferon gamma, and it's too late by that time because the virus has entered the neuron already and there's no turning back from there. In addition, we discovered these new lymphoid clusters known as memory lymphocyte clusters that are key in maintaining this kind of tissue-resident memory cells within the vaginal mucosa. And what we find is this local cluster of T cells that are kept there by macrophages that are stimulating these T cells by presenting their viral antigen and having a feed-forward loop to maintain these types of T-cell in the tissue for a very long time. And these are the MLCs, these structures are what is protective against the challenge by genital herpes infection. So Norihumi Iijima who did this work discovered this new structure that maintains these tissue-resident memory cells in the genital tissue and that really confers 100% protection against herpes disease. So, however, even though it's a very, very effective mode of immunizing a host, which is to use this attenuated herpes virus locally, this is not going to fly in the face of FDA as a childhood vaccine. So we needed to come up with a safer way of eliciting similar response. We knew from the work of Yusuke Naganishi that that you can actually call in killer T cells by a two-step process in which this is during primary infection with herpes simplex virus. CD4 T cells enter the tissue first. They see the antigen. They start secreting interferon gamma and this induces chemokines which are these attractants that would recruit the killer T cells into the tissue. And so we thought maybe we can override the requirement for infectious vaccine by making a similar setting by mimicking this kind of activity by just giving the mice a vaccine which is a conventional vaccine that's injected intramuscularly to generate T cells but then we can pull in these T cells using the same chemokines that nature uses to recruit these kinds of killer cells. And this led to the development of a new vaccine strategy called prime and pull in which we prime the animal with a vaccine and then we generate these effector killer T cells which have these chemokine receptors that can be used to lure them into the tissue of interest, in this case the vaginal mucosa, by providing recombinant and chemokines that are specific for these receptors that are expressed by the killer T cells. So as a proof of principle, Hannah Shin in my lab tested whether this kind of prime and pull strategy can recruit the killer T cells into the tissue. To do this, she immunized animals subcutaneously, just sort of what we do normally in the case of vaccines. Pull these T cells into the vaginal tissue using recombinant chemokines. So because these vaccines are given in a distal site, not the vaginal tissue, they do not elicit local tissue-resident memory cells by themselves. However, we were hoping to recruit the T cells with recombinant chemokines and then we analyzed the distribution of the T cells in the host animal. Animal. And what we saw was actually a very remarkable increase in the number of T cells that are specific to herpes virus that's elicited by the vaccine that can be recruited by this pull strategy into the animal. And prime alone was completely unable to do this, which we knew from our previous studies. And of course, if you infect the animal intravaginally, traditionally with a recombinant virus, of course, you can elicit a very good tissue resident memory cells. But we wanted to do this in a more safer manner. And in fact, the prime and pull strategy only affected the population of CD8 T cells within the vaginal tissue and not anywhere else. So this is a very fine local control of recruiting T cells and maintaining them over weeks weeks. But the important thing is we need to be able to protect the host animal against the disease. So Hannah took these prime and pull immunized animals and then challenged them with wild-type herpes into the vaginal tissue. And here what I'm showing is the disease score or the survival curve of animals that have received prime and pull and virtually it's completely diminishes any disease that's resulting from the challenge. And 100% of the animals survive this kind of challenge if they are immunized by prime and pull compared to the prime alone. So this is sort of the conventional vaccine that is given. You can see that the disease is elevated here and many of these animals succumb to infection. So Hana's work led to a development of a new vaccine vaccine, a whole new vaccine that overcomes the failed vaccine strategies that have been tried in thousands of people. So herpes vaccines have already been tried in phase 3 clinical trials by a variety of different companies, all of which unfortunately have failed. And we think we can rescue these vaccines using the prime and pull strategy. So we're taking this to the to the next level to see whether we can use the prime and pull vaccine to treat an existing disease. So what I showed you just now in the mouse model is a preventative vaccine. But can we do this in already infected people? And I showed you there are hundreds of millions of people already infected with herpes around the world. So what we can do is to use a guinea pig model of genital herpes herpes. And the advantage of using guinea pig model is that they have a naturally recurrent genital herpes disease, unlike the mice which do not. And so we can mimic better what happens during infection in human using guinea pigs. And so we infect a whole bunch of guinea pigs through the vaginal tract using herpes, wild-type herpes simplex virus, and then we let them become fully infected. So these animals have full lesion reactivation, the whole gamut of diseases that you might expect from a viral infection. And then we divide these animals into 4 different arms. Some of them we give no vaccines or no pull. And another group we give the prime, which is the HSV-2 glycoproteins that have been tried and failed in the clinical settings. And then we also give with pool, which is a chemical compound called Imiquimod, which stimulates the innate immune system to induce the chemokines of interest, in our case, the CXCL9 and CXCL10. And so we compared the ability of these kinds of vaccines to protect the host. And what we see is that the prime and pool group pretty much completely shut down any recurrent diseases in the animal that received this vaccine 2 weeks after the infection is already ongoing. Whereas the prime alone, which is the green bar here, green dots here, confers very little protection over just PBS control. So this is the conventional vaccine that have tried and failed, and this is our vaccine. So you can imagine that this vaccine approach will have a very promising kind of treatment option for women who are already infected with genital herpes. So we're now taking this to the next, next level, which is to ask whether we can use the prime and pull vaccine to prevent cancer formation in women who are already infected with HPV, which is the human papillomavirus that causes cervical cancer. So we are collaborating with with two fantastic clinicians, Alessandro Santin and Sangini Sheth, who recruit patients who have these cervical intraepithelial neoplasia, which is a precancerous lesion in the cervix that could develop into a cervical cancer. We're recruiting these women to give the prime and pull vaccine. In this case, we give prime, which is the Gardasil 9 vaccine that's FDA-approved and are giving to children. Children already. But the Gardasil 9 vaccine alone is already known not to work in the setting of women who already have HPV. But then we're adding this pool step, which is the Imiquimod applied to the cervix, which induces the chemokines of our interest. And what the interim analysis is showing is that compared to the observation arm in which about a third of the women women go on to progress in disease. The pool-only group has some effect, but the Primal Pool group of women, none of them have developed progressive disease and they have all reduced the level of the lesions during this clinical trial, indicating that this may be a promising new way of dealing with HPV infection. Infection. And so these people, along with Jean, who is a postdoc in my laboratory, is really trying to understand how we can use this prime and pull strategy to prevent cancer from forming in the cervix of women who already have HPV infection. And they're really revolutionizing the way we treat precancerous lesions in the cervix. And this often detected in women of childbearing age. And this kind of preventative strategy can really preserve the integrity of the cervix, which enable women to go on and have successful pregnancy later on. So why this is important, we think that the development of the new vaccine strategy that have— that can rescue a failed vaccine strategy is a very promising way of approaching this problem. We don't have to invent new vaccines for every disease. We just have to basically combine the prime to a pool to a given tissue of interest. And many existing failed vaccines can be rescued, we believe, using the prime and pool vaccine. And the prime and pool can be applied to vaccines against cancer. We've already shown that for HPV, but even for cancers that are not virally driven. As long as we understand which antigens are expressed in those cancers, we can theoretically do prime and pull against a variety of different cancers. So just to finish up, what is the outlook for the future? We think that the studying viral infection and disease mechanism It's really a great way to learn about the immune system and host physiology in general. And I hope I was able to demonstrate that using these basic research tools such as the mouse model, the guinea pig model, and this in vitro system allows us to gain insights that can ultimately be applied to improving human health. And we need to leverage this natural immune immune response against viruses for vaccine development. So Prime and Pull is one of those examples, but I believe that we can gain insights into how the immune system naturally protects against viral diseases and then apply that to different kinds of vaccine strategies. And similar thought process can be applied to understand and treat cancer. So all the examples that I gave I gave you today is about viruses, but there are many similarities between cancer and viruses. And we might be able to learn from viral diseases and immunity against viruses and apply that insight into the treatment of cancers as well. So in final reflection, I'd like to take just 2 minutes to kind of discuss my other passion, which is to increase diversity in the science and technology and engineering and math. And this slide just demonstrates that about 50% of the graduates from PhD programs in medical schools are about women and men, about 50/50. And yet if you look at the deans and department chairs and full-time professors, the percentage drops to 6%. 16%, 15%, and 21%. And this is true all across this country. It's not just a problem of certain schools. And I think that what I've learned over the years is that diversity is what drives excellence. And we must provide— promote women and minorities to excel in science and encourage them to succeed because we are really kind of an untapped resource of talents that are wasted out there without this kind of support. And it's not just words. I think we need to invest money in infrastructure that supports women and particularly mothers to succeed in science, because not having childcare and child support, especially during early pregnancy and early after after having a child is where the pipeline kind of drops, right? So the women who are in the system drop out mainly after having a child. And this is a major problem that we can solve. We know how to do this. I think we need to invest into these kinds of programs to enable the rest of the 50% of the population whose talent is being untapped. And I also talk about many of my thoughts in my Twitter, so if you want to follow me, you can, you're welcome to do so. And this is the most important slide, which is all the lab members that have kind of made my scientific life such an enriching environment, and past and current members of the laboratory that kind of makes me happy every time time I wake up in the morning and want to come to work. And scientific discovery is really quite addictive. I wouldn't want to do any other job. This is a wonderful environment to do science. And I want to thank all of you who have contributed to the joy in my life as a scientist. So thank you very much for your attention.