>> welcome to the wednesday afternoon lecture. today we have dan littman who will talk about very intereting interaction between the microbiome and the immune system. he is an investigator at the howard hughes medical institute
and coordinator of the molecular pathogenesis program and kimmel professor of molecular immunology at the skirball institute at nyu. dan got his undergraduate at princeton and got m.d. ph.d. in molecular biology at wasu. following that did a post
doctoral time with richard axel at columbia, then set up his own inventory at ucsf while where for the next ten years he was assistant professor associate professor and full professor. but since 1995, he's been at the nyu school of medicine where he is both investigator of hhmi but
also kimmel professor. he's been recognized by a number of honors, notably election to the national academy of sciences in 2003 and the ross prize, and the coly prize. his investigations he's going to tell you about understand ways in which t lymphocytes acquire
functional properties during development in the thymus and how this is influenced busyingnals initiated by distinct commensal microbes in the intestin. so here we are at a time when one would be tempted to say and would say correctly, that there
is a revolution going on in our understanding of the microbiome and ability to assess what's happening particularly in the gi track but there also seems to be period of rapid advances in immunology an putting to two together is an example how sometimes the most exciting
science happens at those interfaces between disciplines and there could be nobody better to represent that this afternoon in this space than dan littman. please join me in welcoming dr. dan littman. [applause] >> thank you so much, francis.
it's really a great honor to be invited here. thank you so much for having me. i have a lot of very warm feelings for the nih for many reasons. i was here in 1976 just as i was starting my ph.d. work as part of my m.d. ph.d. because
advisers were finishing their post-doctoral fellowships here and they were in transition to st. louis and i helped them with the move after spending the bicentennial summer in washington which was a very special time. so as francis mentioned, i will
tell you about our work on trying to deconvolute how microbiota interacts with the immune system. and in particular with t lymphocytes. the title that i have here, i'm not exactly certainly we think the microbiota act as in
instructing t lymphocytes as to their fate. but also in deciding whether when they're on their way to one fate or another which branch they're going to talk. there's a lot of plasticity in the immune system and i will tell you a little bit about the
various options. we have heard of course a great deal about the microbiota during the past five to ten years. we now know the commensals within us contribute to insulin resistance or sensitivity, they have a very important role in obesity.
there's beautiful work in jeff gordon's group, a pioneer in this area showing that their microbiota that in twins who are discordant in terms of weight pre-dispose mice to the same phenotype as the human host. there are also examples of the microbiota influencing the
central nervous system and behavior. i will tell you a little more about this towards the end of my talk and the microbiota have been implicated in the from here, from the carrie laboratory has shown commensal microbes influence responses to cancer
therapy both chemotherapy and immunotherapy. we have a poor understanding what's going on there but we think some of the work we and others are doing to try to understand the responses of the immune system may contribute to eventually getting an
understanding there of how these microbes might be manipulated in treatment of cancer. our entry into this field molecule called r 1 gamma t, a transcription factor that is a nuclear receptor similar to glucocorticoid or estrogen receptors.
it has ligands of the -- in the cholesterol family, in the cholesterol biosynthetic pathway though the precise ligand is not identified. we think there are multiple ligands. this is a molecule we discovered years ago as being involved in
the thymus for survival of immature thymocites that give rise to cd4 and cd8 cells in the periphery. at the time we started working on this, we also found that it has a role in lymphoid organ development in development of patches and lymph nodes, but the
real excitement in this field came subsequently when it was found that our r gamma t has a role in certain classes of t-cells and in a lymphoid cells type 3 lymphoid cells. i will tell you a bit about each of these different cells, not so much about gamma delta t-cells.
what i will focus on is th 17 cells because it was really discovery of cells by several groups most prominently dan kua that set us on the way as i will tell you about briefly here. so when we first started working on r 1 gamma t and realized role in inductions of lymphoid organs
we found that it's expressed in a cell that was first described by (indiscernible) that eventually was named lymphoid tissue and user cell. this is found in the fetus for development of lymph nodes and secondary lymphoid organs but also found postnatally in
tertiary lymphoid organs such as a structure that occurs in intestine under the epithelial layer, this is called a crypto patch. what this consists of are these r gamma t positive innate lymphoid cells and it's surrounded by dendritic cells
and it's senses the state of microbiota across the single cell epithelium and upon signaling recruits b lymphocytes to become an isolated lymphoid follicle. and gerard, a post-doc in the lab made a knock in of gfp in the r 1 locust in homozygous
knock in mice which gene product is null, there are no structures frm bud in heterozygous mouse we can follow expression of gfp in cells of normal copy. (indiscernible) a new post-doc in the lab we missed earlier, that is that they're not just these bright gfp positive cells
that are innate lymphoid cells that don't have any other kinds of hematopoi ting markers for lineages but there is intermediate gfp populations here that are t-cells that express the alpha beta t-cell receptor, around the same time i received a call from dan who
told me they were look at these mysterious cells that were responsive to interleukin 23. in collaboration with (indiscernible) in dan cord's lab we began look at the cells and quibblingly were able to show -- quickly were able to show in this gfp intermediate
cells there were expression of intracellular interleukin 17 in the gfp r gamma t negative t-cells. there was no negative expression of il 17. this was the way we got into this field these were most abundant in intestine.
so it's been known since the mid 1980s about th 1 and 2 cells, i don't need to tell this adience about this. the late bill paul was a pioneer in this area, one of the heroes in the area of t-cell biology and bill's laboratory did a great deal on characterizing the
molecular features of these cells. the th 1 cells are dependent on the transcription factor d bit and they make interferon gamma, generally required to kill intracellular pathogens but also implicated in autoimmune disease and the th 2 cells are dependent
on transcription gamma factor 3, il 4, 5 and 13, very important for controlling infection with paracytic worms but they also contribute to allergy and asthma through production of cytokines. these were the cells though to be major types of helper cells until the mid 2000s, then with
the advent of discovery of cells that make il 17 they are dependent on gamma t it's clear these cells make other cytokines including into interleukin the 2. these are cells required to control extra cellular microbes i wrote pathogens, but not only
pathogens but also commensal microbes including both bacteria and fungi. the cells are very important repair of tissues at barrier services. primarily epithelial services and they are also implicated in autoimmune inflammatory disease,
there's enormous attention paid to these cells. and i will tell you a bit later about the various cytokines that are required for the generation of these cells. a fourth type of cd4 positive t-cell is regulatory t-cell. t reg cells differentiate,
develop in the thymus, thymic natural t regulars but also peripheral induced tregs that express fox p 3, dependent on transcription and all these cells are really essential for restraining these other types of t-cells and preventing autoimmune disease.
and as i will show you, the induced treg in intestin also express r 1 gamma t. the th 17 cells are ones that have really been at the center of studies of auto-immunity during the last half dozen years or so. and i told you about their
importance barrier surfaces and controlling infection with these pathogens here. but these cells have been implicated in these different kinds of inflammatory diseases. i should say psoriasis is the best validated because antibodies against interleukin
17 and interleukin 23 are tremendously effective in therapy for psoriasis. however that targets il 17 may not only by th 17 but also very likely by gamma delta t-cells. some of these other diseases also appear to be th 17 related. a lot of oar thread dees
arthropathies are dependent on th 17 >> rheumatoid arthritis has a th 17 component, inflammatory bowel disease search and seizure dependent on th 17 cells within some of the other cells that make interleukin 17. probably steroid resistant
asthma as well as ms have th 17 components. i put here as my question mark autism spectrum disorder, i will tell you about a model in mice. in which certainly in that model th 17 cells appear to have a role. so this is just an example of
what we are able to show as soon as we understood th 17 cells were most abundant in the intestin in the lamina propria. they're positive in the -- in terms of microbiota present there but when animals were treated with antibiotic or derived germ free there were
essentially no il-17 producing cells in the intestin of these mice. so this is work again of yvonne and he made observation that took us to study it is microbiota. he found the same strain of mice obtained from different sources
differed in the proportion of th 17 cells among cd4 cells in the intestine. so mice from jack's laboratory had few cells, 1 or 2% cd4 cells whereas mice from -- typically had 15% cells making interleukin 17a. what he was able to show was
help with collaborator kenya handa in japan, this was simply due to a difference in one single bacteria species and that is due to the presence of segmented fill mentos bacteria in mice from this source. this is a gram positive spore forming an aerobic bacteria,
different from taxons of bacteria described but probably it's related to the clostridia. these bacteria you can see in these films that cover the villi in the ilium. and they're completely absent in mows mice obtained from jackson laboratory.
you can see that in a transmission em here, these form segments and they embed themselves in the epithelium of the i willial in the ilium and cause a active polymerization action, we understand nothing about molecular mechanism involved here.
we're able to obtain fecal material from animals that were monoassociated with sfb and gavage germ free mice here they have no production of il 22 or il 17 in their cd4 cells and within a week of gavaging you see massive inductions of these cytokines this helps protect
mucosal barrier from pathogenic bacteria but at the same time make animals susceptible to developing arthritis in a mouse model we study in combination with diane and kristen, this is an example of the protection from ad hearing effacing bacterium there's considerably
less adhesion and there's also less weight loss in these animals colonized with citrobacter. on the other hand, the flip side is that in the kb by n mouse model of arthritis which diane and chris to have had shown no disease when animals were
derived germ free, colonization with sfb alone within only a few days now leads to thickening of the jointses and to inflammation and arthritis. and this is work of joyce wu in collaboration with evo yvonne in our lab. and what joyce was able to show
is that though this disease is dependent on antibody response and and body against a self-protein, it requires th 17 cell polarization. this is what sfb is achieving here. so this led us to consider that there maybe something similar in
humans with rheumatoid arthritis z to what have been observed in this animal model. jose sheer at the time was a rheumatology fellow at nyu, now on the faculty. and jose decided to look at introduced to the mouse gi track in the colon led to inductions
of regulatory t-cells and as i will show you many of these regulatory t-cells are specific for the microbial antigen elicited. these. so just to show you that work, actually this is work by mo shu who is a post-doc in our lack
and mo compared inductions of t-cells by segmented fill mentos bacterium and helicobacter. i told you about induces th 17 response under what we consider homeostatic conditions. there's no sign of inflammation in these animal, th 17 cells are abundant and do not promote
colitis. these are found typically in the ilium in the terminal ilium. sfb control nonnization. helicobacter hepatocuss is in the colon and it can induce a pathogenic type of t-cells response in models of colitis. so fiona has pioneered this
field and shown introduction of naive t-cells into animals that lack any t or b cells in rag knockout mice threads to colitis that is dependent on mixed th 17 th 1 response that cells making il 17 and interferon gamma. we want to see what is the t-cell response under
in her case this kind of colitis require there be colonization with bacteria, most often helicobacter hepatocuss, if one blocks il 10 production in helicobacter there's a similar inflammatory colitis. so what mo decided to do was to isolate t-cells from the
intestin of mice, colonized with helicobacter and identified the t criminal receptors of -- on these cells where elicited by helicobacter and make t-cell receptor transgenic mice as well as generated mhc tetramers to be able to track these particular which turned out to be specific
for heel doe backtor -- helicobacter antigen. he used a combination of t-cells specific for sfb and t-cell specific for helicobacter in animals that were colonized with both of these bacteria. so using eye toe type to distinguish between t-cells
specific for either bacterial antigens, i show you comparison between these here, because now when he recovers t-cells from the lamina propria in the mice you see the cells specific for sfb become r 1 gamma t positive cells, test are th 17 cells non-pathogenic.
helicobacter specific t-cells by and large become fox p 3 cells, shown here, which also express ro1 gamma t. this is the hallmark of these types of t-cells that differentiate in the intestin to become regulatory t-cells shown to have unique properties
different from other t-cells, the t-cells that do not express fox p 3 express chemokine receptors an transcription factor bcl of and these are follicular helper cells involved in the production of antibodies. so what happens when these are introduced, these cells are
introduced into il-10 knockout mice? i told you that in il 10 knockout mice, combination leads to colitis. what we see is these cells now differentiate into ro1 gamma t positive cells for the most part.
unlike the cell specific for sfb these cells are pathogenic. i will explain to you what that means. in a moment. now repopulated with these two types of t-cells, you see helicobacter specifc t-cells express not only il 17 on the x axis but also interferon gamma
on the y axis so this is one of the hallmarks of the pathogenic t-cells on the other hand the sfb specific t-cells in the il 10 knock out only express interleukin 17a and no interferon gamma. and when we have done rna sequencing to compare these
populations, also single cell rna seq we see there's hardsly any difference in the presence or absence of il-10 of these whereas helicobacter specific cells changed dramatically in properties after exposure to the il 10 deficient background. so this takes me to the topic of
pathogenic th 17 cells in these mouse models and whether there maybe such cells in human. there may be such cells in human and there have been called th 1 star cells by frederick salusto who first described them. these cells in the mouse are dependent on the interleukin 23
receptor. i told you that early on dan kua showed il 23 signaling leads to up regulation of interleukin 17 in vitro and it's required for pathogenesis in various mouse models of auto-immunity. so in particular there was an experiment done by a group in
london showing that using fake mapping that many of the cells that initially had expressed interleukin 17a went later to express interferon gamma and that process was dependent on interleukin 23 receptor. and only in the presence of interleukin 23 receptor could
these cells be pathogenic in a model in auto-immunity model of experimental autoimmune encephalomyelitis. we think something similar maybe occurring in human. because fredericka described cells that express chemokine receptors, ccr 6 which is a th
17 related receptor as well as cxcr 3, a th 1 chemokine receptor downstream of transcription factor tbet. these double expressing cells also produce interferon gamma, these are cells she called th 1 star. and joe casino via at
rockefeller found patient families that had immunodeficiencies with deficiency in ro1 gamma t they're known for ro1 gamma t, these cells were lacking and the individuals still had classical th 1 cells making interferon gamma in response to viral
infections. and both fredericka and john also found that these are cells that are particularly responsive to microbacteria. if these individuals from these families lacking ro1 gamma p were inoculated with bcg, they developed disseminated infection
and couple did not survive because the star cells are required for controlling these -- this type of bacterium. the microbacteria. what is special then about these inflammatory pathogenic th 17 cells? lynn w uk has taken a different
approach to look at these, that is to use a rag knockout mice which he reconstitutes bone marrow from il 23 receptor sufficient and receptor deficient mice and then subjects these animals to a couple of auto-immunity models as well as colonization with sfb.
the auto-immunity models are to look at central nervous system infiltrates, i told you about and sfb colonization i told you about. he did single cell rna sequencing in collaboration with -- who is at nyu and genome center.
looking initially at cd4 positive t-cells from the central nervous system in mice in which there was inductions he was able to cluster cells in to different categories shown here. cells had fox p 3 and interleukin 10, il 17 a, interferon gamma and gmcsf,
these are properties what have been defined described as when he looked at distribution of receptor sufficient deficient cells were under-represented in this pathogenic population here. but il 23 independent cells enriched here a regulatory type population we don't think
contributes to the disease. what happens then upon colonization with sfb? if we now look in the intestin at cd4 cells what we see is that there are regulatory t-cells that express fox p 3, cells here, which fall into the category of those that make ro1
gamma t and thymic derived and there what might be consider to be more classical conventional th 1 type cells that make interferon gamma over but there are no cells that make both interferon gamma and il you have instead these what we consider homeostatic th 17 cells
down in this category over here. so we're obviously very interested in trying to understand what regulates the cells going one direction or another. vijay has done beautiful work in this area in which he has identified several candidates
that maybe involved in this process. we're still trying to understand what the relationship is between induced regulatory t-cells and th 17 cells. and we would like to know, we know il 23 is required for these we think they differiate h from
what might be a homeostatic th 17 cells, we know that il 10 restrains this differentiation, and we don't know exactly where these treg cells come from though we think they likely polarize from naive cd4 positive we also know that these cells are absolutely required to
restrain this process of expansion of pathogenic th 17 we have data which if we completely eliminate these transcription factor in these cells, we're able to see the cells and see spontaneous so coming back then to what might be occurring upon
colonization, with sfb there's inductions we see inductions occurring first in the mesoteric lymph node within three days and a few pho days later these are found in the lamb ma propria where they are making interleukin a and f and il 22. we see cells distributing
broadly throughout the other parts of the intestine to distal lymph nodes and also to spleen. that may explain t-cells elicited here in the gut mucosa by microbiota may contribute in some ways to auto-immunity. that's been the goal to understand what it is that
trigger auto-immunity that colonized with certain bacteria. to do this we started taking a closer look at this inductions of ro1 gamma t and il 17a. during time colonization with sfb. i told you already that sfb is mostly in the terminal ilium.
we don't see any at all. we see a change in gene expression quicker -- and the genes highly epiregulated are the ma'am amyloid proteins 1 and 2, i will tell you about that in the next part of this talk duodenum and colon shown here he saw similar proportions of ro1
gamma t positive cd4 t-cells we don't have to stimulate the cells e vivo, we can take them directly out of the the lamina propria and look at gfp expression. we wanted to know what is it about the ilium that's special that allows inductions of
effector functions here. and he performed a series of genetic experiments and also antibody blocking experimentment one shown here we focused on the amyloid a proteins. we're able to use the t-cell receptor transgenic t-cells that are specific for sfb and
introduce them into mice that either were not colonized with sfb or were colonized with sfb. the cells were labeled within the intracellular cfc for proliferation. if we look in the mess tearic lymph node there's no inductions of cell proliferation in the
absence of sfb but with sfb there's good proliferation shown in animals recipients that lack the amyloid a protein genes there's still proliferation. maybe a little retarded but quite good proliferation. there's inductions of ro1 gamma t in the meds tearic lymph node
if we look in the illinois yum lamb ma propria, we see something different. look at il 17 a inductions you can see that in these t-cell receptor transgenic cells many cells make il 17a but few comparatively few of the cells make il 17a in the absence of
sero amyloid a. -- serum amyloid arc. this is due to a role of the microbiota, in this case particularly sfb activating the type 3 innate lymphoid cells to make interleukin 22. earlier randy longman in our laboratory found that
mononuclear figcytes or monocyte derived cells that make interleukin 1, 23 and tnf family member tnf sf 15 or tl 1a through production of these cytokine, able investigate the lymphoid cells to produce il 22 and il 22 acts to protect barrier.
miriam's lab at the same time found csf 2 or gmcsf also made by ilc 3 in response to microbiota contributes to homeostay at this and ex-- homeostasis and expansion of regulatory t-cells in the lamina appropriate preia. what they found is there was
inductions of il 22, upon if they lack interleukin 23 receptor, in this case the receptor primarily on innate lymphoid cells you can see upon colonization with sfb, there's less inductions of saa 1 and 2 in the epithelium of these mice. and this is looking in a log
scale at the rna level. this and others led us to the model that i show here. sfb colonization there's a local short circuit in which the monocyte derived cells making interleukin 23 act on receptor innate typhoid cells that acts on receptor in the intestine on
the -- in the epithelium, activating phospho3 and that leads the expression of saa 1 and 2 within the first few days of colonization. at the same time monocyte derived dendritic cells traffic to the draining lymph node, polarize t-cells through
mechanisms we don't understand yet, to become ro1 gamma t positive. these are distributed throughout and outside the intestin but here locally where we have serum amyloid a protein there's enduction of the il 17 of the effector cytokine.
let's propose a step 2 model in the differentiation and activation of th 17 cells versus specification in a draining lymph node and can be monitored by ro is gamma t expression. and inductions of effector function that in this case on serum amyloid a proteins to a
large extent. p p p p p i wouldn't say completely dependent, we say an example of endogenous adjuvant that might have a unique or special role on th 17 cells, this led us to endogenous adjuvants may change threshold for activation of
self-reactive t-cells that might be induced by sfb. that may lead to pre-disposed animals to autoimmune disease. that is a hypothesis we're following. so we focused because of these results on serum amyloid a known for many years in the clinic
because they are acute phase reactant, much like c reactive protein. these are lipoproteins associated with high density protein, and retina in the palacena. they're produced in liver for acute responses to infection.
enormous increases in circulating levels. they are also elevated in tumors in synovium of rheumatoid arthritis patient and in the serum in autoimmune disease and it's been reported in multiple sclerosis patients for example. it's a highly conserved family
of genes and there have been multiple reported receptors although it's unclear which might be physiologically important, toll like receptor 2, the formal peptide receptor 2 and cd 36 have been reported as receptor, so far we haven't found a function for any of
these in what i'm going to show you now. so jim young lee started looking at the role of saa which we found works directly on t with recombinant saa 1 and 3 he is able to see a really substantial increase in the proportion of cells that express
interleukin 17a even just in the presence of interleukin 6. interleukin 6 and tgf beta are cytokines that induce th 17 cells from naive cd4 positive t-cells and john o'shey's group here has looked at other conditions for looking at tgf beta independent differentiation
of cells with il 6 il 1 beta and il 23. you can see in these differentiation conditions, adding in recombinant saa 1 leads to major increase in the proportion of il 17 producing and what's really interesting is that if we now look at ro1 gamma
t expression here we see that even without any difference in ro1 gamma t expression in these cases, we see a big increase in il 17a, consistent with the idea of this is a second hit second signal that activates the effector function of the cell. so june decided to look in
models of auto-immunity to ask whether saa might be important and you look at saa knockout mice gotten from fred debeer in kentucky but introduced by crisper a mutation in the third saa gene saa 3 expressed by macrophages in the lamb ma propria upon colonization with
so he compared in the immunization model the wild type saa triple income out mice and what e h sees is -- triple knockout mice and he sees attenuation of disease in acute phase shown here but also chronic phase of disease here. i won't show you the data but we
know that this attenuation seem to be mostly a function of the absence of saa 1 and 2. whereas this more rapid recovery here, seems to be mostly function of saa 3. we're beginning to try to understand what might be going on there.
so you can see that after immunization with mog and adjuvant there's rapid increase in circulating saa 1 and 2, this is in the knock outs here for sa 1 and 2. and this is due to expression from the liver because you can see both saa 1 and saa 2 are
highly up regulated upon immunization primarily in the liver but there's nothing no activation of saa 3 in that organ. on the other hand looking at the central nervous system you see no effect of saa 1 and 2 but there is inductions of saa 3,
this turns out to be made by microglia and by monocyte derived cells that invade the central nervous system after immunization. so what june also noted was when he looked many the chronic phase of disease he saw animals had various levels of disease scores
but also varied in their circulating saa 1 and 2. he was able then to compare the level of saa 1 and 2 in the circulation to the disease score and he saw a remarkably strong correlation here the higher the circulating level of saas the greater the disease.
consistent with the saas having a role in the disease then he looked at another model of eae that was developed in vijay's lab using t-cell receptor transgenic mice which t-cell receptor is specific for a peptide of the myelin protein of mog.
so these 2d 2 cells are differentiated in vitro, polarized to become inflammatory th 17 cells with this combination of il 6 and tgf beta and subsequently interleukin 23. when introduced into wild type mice, that leads now to paralytic disease, to the eae.
he wanted to compare now to saa 1, 2, 3, knockout mice, i only show results with saa 3, there's a mild defect upon introduction into saa 1, 2, double knock out but you can see introduction into saa 3 knock out almost completely protects animals from the disease as shown by percent
of animals with disease and the average disease score. what's also interesting is gating on raw gamma t positive cells in animals sufficient or deficient for saa 3, you can see that even though there are gamma t positive cells in the knockout mice that go to cns many are 17a
consistent with the idea there might be a second signal coming from saa to facilitate production of effector cytokine in this microenvironment. so to put this two step model in context of pathogenesis, what we think then is that saa 1 and 2 in sfb specific t-cells
contributes to the cells locally becoming positive for interlocking 17a and f. on the other hand in the pathogenesis model of eae in the cns we see a role for saa 3 in the expression of the pathogenic cytokines and the ability of these cells to induce disease.
we think that i showed you example with sfb of these cells going to the lamb ma propria where they can be subject to this induction by saa, we think that similarly these cells make their way to the central nervous system but maybe by circuitous route.
we don't know whether they are derived from these cells first induced in the intestine, in this case potentially in the small intestine but i will show you in the eae model we think it's they are induced by saa produced by the liver. another important question here
is do the th 17 cells that go to central nervous system need to be cross reactive with th 17 inducing bacterial antigens. is there some kind of molecular mimicry involved in the autoimmune disease, i will be happy to discuss this after my lecture but our inclination is
to think that's not the case. we do think the microbiota are critical in this particular model of auto-immunity in the mog induced inducedeae as shownhere. we treat with ampicillin or advantage co-mycin. we protect them from the paralytic disease.
you can see the neomycin do not have this kind of an effect. so it's giving us the opportunity now to begin to look to see which bacterium are involved here, this is in mice that lack any sfb. this is not an sfb dependent effect.
to put into context we think the micriobiota do in fact program the cells to become ro1 gamma t positive in this polarization. thereafter immunization with mog in presence of adjuvant there's production of systemic sae 1 and 2, these proteins act on primed th 17 cells, activating them,
maturing them. they express now chemokine receptor crr 6 that allows to traffic to the central nervous once they get there, we think that they can cause local inflammation and positive feedback process activate production of saa 3 by the
myeloid cells there and now the saa 3 acts to generate the -- to induce the pathogenic cytokines and that leads them to chronic inflammatory disease. so these are all testable hypotheses now to try to understand how these cells make their way from being just prime
cells to being actually p pathogenic t-cells. so what i was telling you we think there's a two step model of the prime cells and effecter cells that can be generated by a variety of different means. and these could be homeostatic or pathogenic.
and contribute to autoimmune disease, e we think potentially harnessed for pro or antitumor affects. i will show you another example this two step model is important. that is a model for autism called the maternal immune
activation model. so this particular model was developed affidavit there were reports based on population studies that women who had had infections late in first trimester of pregnancy were more likely to have children with autism spectrum disorder.
the model developed was to either introduce a virus or later just polyic which activates toll like receptor 3 into pregnant mothers day 12 and a half of gestation. that leads to behavioral abnormalities in offspring persisting into adult life and
as well as to morphological abnormalities developing in the central nervous system. this is a model that was championed by the late paul patterson, caltech. and about a decade ago paul published a paper showing that this mia phenotype in the
offspring requires mother be specific for interleukin 6. this paper in journal of neuroscience. at that time a fabulous post-doc in my lab, june hu, joined the lab, he was a graduate student at caltech who knew paul patterson's work.
we were beginning to work on th 17 cells, this was now more than ten years ago. and we knew il 6 is involved in inductions of ro1 gamma t and th 17 cells so we propose this would be involved in this animal model, through inductions of th 17 cells.
so we got a pilot grant from simon foundation for autism research and began to look at this. there's three cardinal symptoms of autism, social interaction defect, communication defects and stereotypic or repetitive behavior, all can be modeled in
animal models in mice debatable whether they are related to what's going on in human. it is what we have available to work with. what june initially did was to reproduce paul patterson's result showing inductions of il 6 shortly after polyic injection
but he also showed inductions of interleukin 17a and in fact the interleukin 17a in circulation stays all elevated for longer than il 6 does. we can block with antibody against il 17. and importantly we have made a conditional knock out mouse for
ro1 gamma t which we could knock it out specifically in t-cells using the cd4 cre to cross to and you can see that after polyic in these mice there's no elevation of interleukin 17a in the mother. he was then able to use this as a tool as well as antibody
against il 17 to see what might occur. one of the first things he showed by doing in situ hybridization was that there is expression of interleukin 17 receptor a in the outer layers of the cortex in the fetus embryonic day 14 and a half, you
can see it here. you can see that after polyic there's actually an increase in experience here. it is known that il 17 receptor is regulated in the positive feedback matter by its own signal, il 17 binding. then we're able the start doing
behavioral assays in wild type and mutant mice in the offspring of wild type and mutant mice. this is the marble bearing assay, a repetitive behavior assay in which the mouse is asked to bury marbles in its bedding and there's a tremendous increase in this kind of
repetitive behavior after polyic treatment of the mother. we eliminate this in the model is lacking ro1 gamma t in her specifically t-cells. similarly one can look at social interaction assay, interacting with another animal here under this cup here, or inanimate
object here, a normal mouse spends twice as much with another mouse and with the inanimate object. and what we see here is after polyic, this is basically a 50/50 ratio, we basically restore preference for the social interacting mouse in the
-- if the mother is lacking ro gamma t in her t-cells. and there's also a nor foe logical defect that we observe that's quite severe that you see here in the cortex, sustaining with various markers for cells in developing cortex in the fetus day 18 1/2 six days after
injection of polyic, blocking il 17 in the mother basically abrogates this abnormality. finally, june with his wife gloria chow a neuroscientist together did injections across the uterus into ventricles of the fetuses and injecting il 17a or pbs or interleukin 6.
only after il 17a injection behavioral defects as well as morphological defects. and in this case we looked at behavioral defects of communication defect though we also saw other behavioral defects i told you about. you can see the morphological
defect as well here. and the last thing i wanted to tell you about is that this is requires microbiota specific microbiota in the mother. that's why i want to link the parts of my talk here. when june was recently able to do it, june has his own
laboratory now at u mass and wooser, he's looking at mice with or without sfb or mice colonized with a mix of bacteria from kenya honda, human bacteria from human that induced th 17 cells in mice. and was able to colonize a mothers and look at il 17
inductions and behavioral defects. you can see here after introduction of the human bacterium in this case and polyic there's a tremendous increase in il 17 production without increase in the introduction of human bacteria,
just black 6 flora from jackson laboratory, there's little il 17a induced by polyic. this corresponds to the behavioral phenotypes i told you about, the localization index, marble bearing, anxiety assay and interaction assay, social interaction assay.
what we can conclude is that in this model the microbiota are absolutely critical in the mother for the priming. and following that priming, of ro gamma t positive cells a second signal inducedded by polyic allows production of il 17 which crosses placental
barrier through mechanisms we don't understand and acts only ic receptor in the central nervous system, the fetus must have intact il 17 receptor in order for the phenotype to be manifested. and that leads both to the developmental phenotype, the
cortico phenotype as well as behavioral abnormalities. so this has interesting implications for humans. we don't really know whether there's anything like this in human, there's no question autism is caused in most of the cases by genetic influence but
there is a known environmental influence. some of that environmental influence could be due to the question here in maternal infection, activation in a subset of women, those women with a particular kind of microbiome that renders their th
17 cells more active larger numbers of ro gamma positive t-cells this is not something we can answer easily but i hope that this study is going to set the stage for doing more investigation in this area. i will stop there. this is june hu, the latest work
was done by san du kim, post-doc in his lab. all the work on the neurophysiology as well as histology in the cns was done by gloria chow at mit. we did much of our early behavioral work at nyu with chuck hull fer now at the
university of colorado collaborated with ken honda and cohhi in terms of microbiota and san woning kim made the conditional knockout mice for these are some of the folks that i mentioned as i went along. betty yang back in china, he was a south carolina for a while.
and he identified the specificity of the sfb induced wendy woning ucsd did much of the early work on saa inductions and its role on t lymphocytes, we collaborated with raul on single cell rna seq, fred debeer on 1 and 2 mice and kim koje on the microbiota studies.
and i will leave you -- stop there and be glad to take questions. thank you very much. >> the hour is a bit late but we should take questions starting >> great talk. do you think it has direct efects on regulates fox p 3
independent of affects of ro gamma? second part il 10 independent of ability to modulate ro gamma level? >> saa -- we haven't seen effect of saa on upregulation of stacks t 2. >> i was imagining
down-regulation. you're saying you get pathogenic th 17 cells and >> we really don't know if there's a difference how saa influences the homeostatic versus pathogenic. we don't know whether it's a difference between saa 1 and 2
and sae 3 in terms of signaling in the target cells. you could imagine there might be a difference there. we're looking at that now. all those experiments i showed you what we know basically. we can see these affects in vitro that are really dramatic
affects, we haven't looked to see whether saa inhibits regulatory t-cell inductions. we should probably look at that. >> that was fantastic. i wanted to touch on the question of antigenic mimicry, as you know we have some data that points to the fact that
this maybe important. you mention you think that if i understood correctly, saa might be changing the threshold of t-cell activation, through the t-cell receptor perhaps. but then later on in your talk you sort of discounted the possibility of antigenic
mimicry. expand on that. how does that go together? >> we have preliminary data that mog-specific t-cells that i showed you in that model in the knock out model, are still dependent on microbiota. even though they are specific
just for that peptide of mog. so it's very difficult to imagine there's fortuitous cross reactivity of a th 17 cell inducing bacterial antigen with mog. >> from who said the antigen affect built into the microbiota?
and angiogenic effect if such is present have to be on the same book? >> i don't think they need to be on the same. i think they can be separate. yes. and the adjuvant effect can be totally irrelevant to that
model, in the draining lymph node joint lymph node or cns draining lymph node, what have you. but it's still, it's still a hypothesis whether there's a difference in threshold. >> bringing it down to critical realm, with a well known
syndrome toxoplasmosis cmv, herpes, rubella and now zika entering the realm of cerebral abnormalities in terms of hypothesis, in terms of the th 17 in pregnant females in terms of these syndromes have there been any specific studies done? these would be perfect targets.
>> i'm not aware of that. after we got this result i called up tony fauci a year ago when zika was coming to the fore, i thought my god this could be it but obviously that's not the case. zika unfortunately cannot be -- if this was the way zika worked
it would be much more manageable clinically. it would be nice. but that's not how it works. but i'm not aware of anybody having done those kinds of studies with the other infections either. >> thank you.
>> very quickly. >> so considering the large pool and the variations of this regulatory t-cells in the brain versus the intestine, what is it so acute in the autosome? that has an effect other susceptible tissues. you mentioned maternal
local interleukin 17. is it good enough to make such a signature impact on inactivation of the cerebral -- >> you're asking can there be enough il-17 in the s laying in >> no, is there something else, what do you see in autism -- the spectrum, wanting the other
issues with the larger brain volume,is there anything else besides the causing inflammation and causing other affect? >> this is a very limited mouse model. we cannot extend this to human. there is no question il 17 can be sufficient here.
whether in vivo we don't know because when we bombard ventricle with il 17 it may not be actually what happening when fetus is in utero normally getting transplacental il 17. so there could be factors, i agree with that, we don't know that.
we know it's required in the it maybe sufficient. and it is a restrictive model even if it is like some human it will develop a small segment of the spectrum. outstanding work. >> if you would like to continue the conversation and also enjoy
coffee and refreshments please let's adjourn to to medical library but let's thanks professor littman again.