hi my name is beth stevens i'm an hhmi investigator at boston children's hospital in the broad institute and our research is focused on understanding how the immune system helps sculpt developing neural circuits and in particular we've been focusing on this resident immune cell called microglia that i'm going to tell you about today so i think it's become increasingly clear that a large number of molecules traditionally associated with the adaptive in the innate immune system are actually expressed normally in the healthy brain especially during development when the brain is wiring up so neurons and glial cells make a lot of these molecules and more and more that we look the more we realize that we've learned a lot from how these molecules function by going back and learning what they do in the immune system i'm going to give you a couple examples of that today so fundamentally one of the questions that my lab has been focused in on is how it is that specific synopsis the connections between neurons are eliminated both in health and disease now during development synapse elimination or synaptic pruning is a normal developmental process in which excess synapses or excess connections are permanently pruned away while others get strengthened and maintained now this is illustrated here in a series of of images golgi stains from the human cortex that just shows that there's a process in which synapses are it is an exuberant synaptic connectivity in this case it's six years old in the cortex but by 14 years old in a cortex you can see sort of a decrease in the connectivity this this is what we would call pruning now this is not an all or nothing process this happens in different critical periods at different times um areas like the prefrontal cortex that are involved in executive function and working memory they actually continue to prune pretty late into late adolescence early adulthood but areas like the visual system and sensory systems tend to prune or remodel relatively early so this is a key process that's necessary for precise brain wiring and defects in this process of pruning are thought to underlie neurodevelopmental and neuropsychiatric disorders including schizophrenia and autism but it's been pretty hard to test in people in part because we don't have a way or a biomarker to be able to go back and study pruning in people over time which is why fortunately we have animal models that have been able to provide us with some of this insight in addition to the neurodevelopmental neuropsychiatric disorders it's becoming increasingly clear that synaptic loss and dysfunction is a hallmark of unfortunately normal aging but also of many cns neurodegenerative diseases including alzheimer's disease so while each of these diseases listed here are quite different with respect to the age of onset and the clinical symptoms um even the circuits they affect could it be that there could be some common mechanisms or common pathways that converge to enable to drive the synaptic vulnerability and this is the question that we've been really interested in the lab which is why we think or hypothesize that understanding how synapses are normally pruned during development could provide new insight now as i mentioned synaptic elimination is a normal developmental process it's been studied by many investigators over the years in many different circuits this is just an oversimplification of what i mean when i say pruning this is a typical postsynaptic neuron initially innervated by lots of inputs or axons that are relatively weak starting out and then through this process called pruning you can see some of these inputs the blue ones get pruned away while the red ones get strengthened to maintain this idea of use it or lose it uh this is coin a coin by collis schatz uh and we know that this process is regulated by neuronal activity meaning the axons or the inputs that are able to drive the postsynaptic cell more effectively tend to win the competition at the expense of its less active nearby neighbor in this case the blue one so we know it's activity dependent much of the work in the field has largely focused until recently on neuronal signals that regulate this process and there's a large number of those mechanisms that i'm not going to talk about today but more and more evidence over the last decade or so are starting to implicate glia the non-neuronal cells of the brain including microglial cells and and also astrocytes which i'm not going to focus on today these are cells that are also critically important in synapse development and function but today i'm going to focus on microglia and what you can appreciate from this cartoon is how dynamic these cells are in the brain so microglia make up about seven to ten percent of the cells in our brain and unlike other glia and other brain cells they're the only cell not born in our brains and so fate mapping studies by miriam rod's lab a number of years ago showed that microglia actually derived from the neural uh from yolk sac progenitors myeloid cells that enter the brain quite early in development and that these cells actually become what we now know as microglia when they're exposed to brain signals and when you can look into the brain and image them which is what we've been doing and the field has been doing you can start to really appreciate how dynamic these cells are and how much they're really interacting with synapses so thanks to some tools that were developed by stefan young's lab initially that can use genetic engineering to label or to label microglia with a gfp this is a fractal kind receptor gfp mouse crossed to a a td tomato mouse that's enabled you to label the neurons in red this is a two-photon imaging experiment by janelle wallace in our lab when janelle's actually as you can see here these microglia processes are dynamically associating with these synapses and the other thing to mention is that microglia are particularly phagocytic during development and it looked to our eye that microglia might be engulfing some of these inputs and others have made similar observations over the years but we needed a way to be able to study this and test this more directly and to address this question we decided to turn to a circuit in which pruning has really been well studied on the developing visual system of the mouse i mean in particular we're studying the the circuit between the retina and the visual thalamus the retina geniculate circuit now this is a circuit that's really been well studied thanks to pioneering work by carla schatz and many others in the field chenfei chen where this circuit can be both labeled and studied anatomically using tracers so for example we can put polar toxin or anterograde tracers into the eyes of a mouse that enables you to track the projections into the lgn here and what you can see is that initially early in development a postsynaptic neuron in the lateral geniculate nucleus the relay station here in the thalamus is initially innervated by the axons from both left and right eye retinal ganglion cells you can see here so they've got they're binocularly integrated but over these next two weeks you can see the segregation or this refinement such that in the mature lgn there's no overlap and there's a shift from binocular to monocular innervation now again this has been well studied and we know that this process is regulated by neuronal activity so the question then is are microglial cells playing a role in this refinement process and if so what are the mechanisms that are regulating it so if you look in the lgn in and mouse in which the inputs have been labeled with cholera toxin what you can see is a microglia shown here zoomed in interacting very intimately with the axons from both the left and the right eye so the question was are they actually phagocytosing or engulfing these inputs so to address this question dory schaefer in my lab actually did an experiment where she measured the engulfment did some 3d imaging and could then remove the inputs that weren't within the microglia and through 3d rendering surface rendering what we observed is that almost every microglia throughout the lgn during this process of eye segregation and pruning we're engulfing both left and right eye inputs and you can even have we showed evidence of that actually engulfment into their lysosomes as well and more um interestingly even by em we can show that they seem to be sort of nibbling off or phagocytosing the presynaptic terminals of these retinal ganglion cell axons and it's happening in a developmentally regulated way meaning between postnatal day five and about post anal day 10 is where we see a lot of this engulfment but after about p10 this really diminishes significantly now the big question now since i told you already that that this process of pruning especially in the circuit is activity dependent we wondered could microglia be responding to activity dependent cues could they be engulfing less active inputs preferentially for example so to address this question again dory labeled the inputs with cholera toxins shown here in left in red and blue and then she used tototoxin which blocked sodium dependent action potentials injected that into one eye to drive this activity dependent competition and then asked the question do microglia preferentially engulf the the weaker or the less active inputs and she did the experiments a couple of different ways but in both cases we saw evidence of an activity dependent or preferential engulfment of the less active eye inputs now this was an important experiment because it shows or it suggests that microglia aren't just acting as the cleanup crew here but that they were actually responding to or are being instructed by some sort of activity dependent signals and then the question then was what are those signals we think there are probably many signals but i'm going to tell you about uh one of them at least initially so thinking about how this might work in the immune system raised the question of could there be actually cues that were that were instructing microglia to engulf the less active inputs and what might those cues be so when i was a postdoctoral fellow and and with ben barris at stanford we unexpectedly discovered a role for the classical complement cascade these are a group of innate immune molecules traditionally associated with clearance of pathogens or debris in the periphery we show that some of these molecules including c1q and downstream c3 this is a part of the classical complement cascade that these molecules both at the protein and the rna level were expressed in the healthy developing brain and quite interestingly subsets of synapses in the lgn and the visual system but also throughout the brain were co-localizing with these complement molecules and what we went on to show and this is just shown here in this high resolution imaging you can see subsets of synapses labeled here in blue and red these are pre and postsynaptic markers you can see subsets of these synapses were also co-localizing with c1q in green and we went on to show that if you blocked or genetically blocked using knockout mice either the complement molecule c1q or c3 or the receptor on microglia and it turns out that the only resident cell in the brain to express complement receptors in the healthy brain at least is microglia when we blocked either the receptor or the molecules these mice had defects in i specific segregation or synaptic refinement and microglia were only about half as good at eating these synapses which identified complement as one of the molecules or the instructive signals that was regulating this process of pruning so this raised a number of questions and also propose a model that we're still working on but i wanted to take you through kind of what we're thinking about at this stage you can imagine a system where during development some of these inputs in this case the blue ones are targeted or tagged by complement molecules and that that's an instructive signal that's telling microglia engulf those inputs but it also raises the question of why microglia aren't engulfing the red inputs or for that matter the whole axon why might there be specificity and what molecules could be regulating this process could there for example be molecules that are also saying to the microglia don't eat me or essentially protective molecules that prevent microglia from engulfing so again we went back and we learned from the immunologists and we thought well what molecules like this actually prevent phagocytosis in the immune system and it turns out that there's a large number of these molecules in the immune system that instruct macrophages not to engulf these are both a complement inhibitors there's a whole group of these molecules that regulate the classical complement cascade at different levels but there's also a group of molecules not so creatively called donate me signals in the immune system and so emily lehrman a graduate student in my lab decided to ask well are some of these molecules especially the donate me signals actually expressed in the developing healthy brain and could they actually be some of the signals that are working in concert with complement to regulate pruning so in the immune system as i mentioned a healthy cell or a cell cell has a large number of these donate me signals that label their surfaces and essentially they act as stop signals for circulating macrophages which express receptors that recognize these don't eat me signals essentially they're stop signals or or sort of protective molecules that even if complement was all over the cell it would prevent phagocytosis in the context of disease or apoptosis it turns out a lot of these donate me signals become down regulated and then molecules like complement then can enable the engulfment of those of those either apoptotic or non-cell cells now could similar molecules actually be labeling synapses and so emily in particular focused on a molecule called cd47 which is a well studied donate me signal in the immune system and one of the reasons why she did that is because it turns out that cd47 was highly expressed or is highly expressed in retinal ganglion cells but in many neurons throughout normal healthy development and into adulthood and microglia express a receptor called serp alpha at high levels especially again during these developmental pruning periods so just as an example i just wanted to sort of give you a sense of some of this this work on still ongoing when we looked at mice that lacked either cd47 or the receptor serp alpha these actually these microglia were over pruning in other words they were engulfing too many synapses or or the wrong number of synapses and so loss of function of either the molecule cd47 or the receptor and microglia led to increased engulfment of microglia in the in the same system i just told you about the lgn and these mice also had enhanced pruning excess fewer numbers of synapses and we actually think they also have some some behavioral defects as well so we have much to do to try to understand mechanism but this is just another example of molecule in this don't eat me signal family that we think might be acting in concert with compliment to give specificity to the circuit so if you look at this the story put together you can imagine during this activity dependent remodeling process the working model is that microglia have a lot of these receptors and because they're dynamically surveying the environment they can essentially read out the combination of cues on those inputs or synapses so synapses that have complement but lack the don't eat me signals might be preferentially eliminated and work now in the lab is designing tools and and new model systems to be able to interrogate this and even be able to watch these molecules interacting in real time during this process of activity dependent pruning this also begs the question what is complement binding to at synapses these are secreted molecules but yet we're seeing some sort of specificity in terms of of some synapses that are labeled with complement and others that are not what are the receptors or binding proteins that might recruit complement to those synapses but not others and again how does activity regulate this process of complement binding or complement expression in the brain what are some of the upstream signals and neurons that's leading to this process and could the same mechanisms that i just told you about today could they become dysregulated in development or in the adult brain and could this lead to pathological synapse loss which again is a hallmark of a lot of these disorders that i that i mentioned earlier and so this sort of suggests a model that maybe some of the same molecules and mechanisms especially relating to the immune system that are normally helping to prune synapses during normal development could become dysregulated either through genetic or environmental perturbations that could then lead to aberrant or pathological synapse loss now this is a hypothesis but it worked over the last many years by my lab and now many other labs have provided some evidence to support this at least in animal models in alzheimer's disease glaucoma frontal temporal dementia models and others now i'm going to tell you now in the second part of my talk about some work to support this hypothesis in alzheimer's disease and there's a number of reasons why we decided to focus on alzheimer's disease first one is because the fact that there's already evidence from humans and in animal models that synapse loss does happen and it's a relatively early event in fact it's predicted that synaptic dysfunction and synapse loss is happening in the pre-clinical phase of alzheimer's disease and there's evidence to suggest this is also happening in various models of alzheimer's disease as well so synapse loss is also the strongest correlate of cognitive decline in alzheimer's disease yet we still don't have a lot of work in the field or evidence that really points to specific mechanisms that drive the synaptic vulnerability so understanding that mechanism i think is going to be critically important and we wondered could some of the same pathways we've been studying during development be playing a role in this early synapse loss the other thing to support a potential role for complement if you look in the literature it's long known that complement is upregulated in alzheimer's in humans and in animal models in fact you can see from from these studies complement not only becomes upregulated but it also co-localizes with amyloid plaques in the in these brains so that's been well known but it's always been thought to be sort of late stages and maybe secondary to neuroinflammation and disease and damage and we also know of course that microglia play critical roles in alzheimer's disease they are not only important in neuroinflammatory events but they're also important in phagocytosing or clearing those amyloid plaques so the question is you know putting all this together could microglia be actually mediating some of the synapse loss through some of these mechanisms including the kalman cascade now it's well known that microglia dramatically alter both their morphology and function in the context of diseases like a.d they go from this really beautifully ramified microglia that i told you about in the first part of my talk and they become quite different and quote-unquote activated um they increase a lot of phagocytic receptors they become more lysosomal there's more lysosomal activity but the problem with this sort of shift in morphology is it really doesn't tell us when these cells change or by at least by morphology and these handful of markers whether they're actually contributing to the detrimental or specific functions in alzheimer's disease now it's been a thought for the most part until recently that microglia dysfunction was sort of a secondary event but merging genetic studies especially in late onset alzheimer's disease more and more are implicating microglia and myeloid cells as being key drivers in alzheimer's disease at least late onset alzheimer's disease in fact almost half of the late onset ad risk genes which are over here a lot of these are common variants actually expressed in or enriched in myeloid cells and even apoe which we know is a critical risk factor in alzheimer's disease is indeed upregulated and expressed in microglia both in ad human ad brains as well as mouse models so putting all this together this has sort of changed changed the way we've been thinking about microglia in the context of alzheimer's disease and so the big question that we've also been focused on is do microglia contribute to synaptic and cognitive impairment in alzheimer's and if so how what are the mechanisms so going back to what i told you about in the first part we specifically asked could microglia and the complement cascade be contributing to this early synapse loss and this is actually important because at least in animal models unlike humans you can kind of rewind the clock to the beginning before this pathology and you can ask when do these things start to change and there are certain parts of the brain including the hippocampus that are that are vulnerable and they're vulnerable to early synapse loss so putting this together my postdoctoral fellow at the time now a a pi in her own lab at ucl soyun hong in collaboration with dennis selco and cynthia lemire we set out to ask the following question is early synapse loss mediated by microglia aberrant reactivation of this microglia pruning and so using established genetic alzheimer's mouse models we we set out to test this question and the first thing that soyun did is she looked in the brains of these alzheimer's mice asked you know how soon does synapse loss happen establish that window and she showed using both anatomical and immunohistochemical approaches and high resolution imaging by abstaining these mouse brains that are alzheimer's versus control with antibodies for pre and postsynaptic markers in red and green you can you can you can actually quantify the number of synapses this way and doing so she found that there's quite an early loss of synapses that happened prior to amyloid deposition in these amyloid or alzheimer's mouse models and moreover when she looked at the hippocampus in the frontal cortex of these mice there was a significant increase in the expression and localization of complement proteins including c1q which you can hopefully see here that if you zoom in on the hippocampus there's an increased expression and localization of c1q shown here in green and interestingly and this was happening at least early stages in vulnerable brain regions like the hippocampus where we didn't really see this in the cerebellum or the striatum for example and this is just quantified here so this is interesting because it's showing an early increase in complement expression and localization to vulnerable brain regions in two different amyloid mouse models moreover if we zoom in and do high resolution imaging much the way we do in development so ion observed that a large number of these synapses in the hippocampus were co-localizing with these complement proteins c1 q and c3 and shown here so what are the consequences is having complement binding to synapses is this is detrimental or beneficial we hypothesize that blocking complement or removing c1 q or c3 in these animal models could be protective of synapse loss and cognitive decline so to address this question we crossed amyloid or ad mouse models to c3 knockout mice this is mice that lack that downstream protein c3 that tags the synapses and then we compared synapse loss and cognitive function in these different genotypes those app models that have c3 and those that lack c3 and what we found is more synapses in the mice that lack c3 this was actually in a three to four month old animals shown here but quite interestingly when we then looked many months later so this is just showing significantly more synapses quantified the way i told you before by looking at co-localized psd 95 and synapsin we could see significantly more synapses in the c3 deficient app p prisonel and mice and even um in the aged ad mice after a year we could still see significant more significantly more synopsis which is important because it's saying that this is happening early and it's it's also being observed at late stages of the disease so what about functional uh and behavioral consequences so in collaboration with cynthia lemire's lab she carried out some behavioral and cognitive testing in these mouse models and showed and interestingly and importantly that c3 deficiency from these amyloid models was protective against cognitive defects despite the fact that these animals had more amyloid increased amyloid plaques and i think this is a really important point because it's seeing that and consistent with the fact that complements involved in plaque clearance that's been shown by others in the field this is saying that even though these mice have more plaques they're protective they're protected from some of their their cognitive impairment and that's also relates to their protection with synapses so this is sort of suggesting two different processes here that may be independent it's also raising the question that this if you're starting to think about when you would want to target some of these pathways that context matters so early on in the process blocking complements especially before you start to get flax seems to be beneficial but then we also know that later stages in the disease when there's a lot of amyloid plaques you know you want to actually have that complement of microglia clearing the plaques and this is really um now starting to want us to dive deeper into understanding what the mechanisms are that regulate this process at different stages and in particular we want to understand even more so uh than ever before what complement is binding to with these synapses and we want to know how complement is interacting with molecules like amyloid and activity to regulate this process now the question is is microglia part of the story is microglia mediating the synapse loss and much like we showed in development what soyun showed is if we blocked the receptor on microglia this complement receptor we also protected the synapse loss and also showed some benefits in this model as well so this really together suggests that the same pathway that's normally regulating synaptic pruning during development can become aberrantly activated to contribute to aberrant or pathological synapse loss at least in animal models and we're now looking in human alzheimer's brain especially at the earlier stages to see if we can see evidence of changes in some of these pathways both in the brains and in the cerebral spinal fluid or csf of alzheimer's patients to try to see if we can validate this mechanism in humans but clearly microglia have complex and diverse roles i told you only about one of them today which is synaptic pruning and i largely focused on one pathway complement but we weren't know from the work from many in the field that many different mechanisms and signals are likely mediating different aspects of this process and of course while that's quite interesting biologically it also is a challenge for starting to think about how we might target both microglia and the complement system and some of these molecules in different stages of disease progression so this really goes back to the question i i led with earlier which is how and when do microglia contribute to alzheimer's disease especially now that we know from emerging genetics that microglia really are not just responding to neuronal dysfunction but they may also be drivers in the disease and even though much of the fields focused on the plaques and amyloid which is critically important the idea that these genes that have been identified by genetics many of which we don't even know their normal biology their normal function we need to better understand how these cells and these pathways are normally functioning in microglia before we can really understand how it goes orion disease and again this is really brings back the fact that it's not as simple as it seems because um microglia actually have many and diverse roles both in normal health normal development and in disease and more and more evidence is suggesting that they exist in different states throughout the lifespan of a mouse i just told you about one function pruning but microglia also play roles in synaptic plasticity they're critically important in synapse function but also other functions in the brain and we also know that when there's a challenge whether that's a local challenge or a larger more large-scale genetic challenge microglia can change state they can go from this homeostatic surveying microglia that i just told you about in the first part of my talk and they can shift into this pathological state for they can also move into an inflammatory state but we appreciate that inflammation can be both detrimental or beneficial depending on context they can shift into a more phagocytic state but that again can be both beneficial in the context of clearing a plaque or detrimental in the context of clearing a synapse aberrantly so how do we get at this question and a challenge has been that the field still lacks tools that enable us to both track these state changes especially across disease or development and manipulate and target specific populations or states in microglia which is ultimately what we want to be able to do um so how do we go about this and i think this is where emerging technology including single cell sequencing is really opened up the field especially when you combine that with genetics which is now kind of laying the groundwork for some of the pathways and genes that we can start to focus in on and pioneering work that began with ito admits first paper that applied single cell sequencing to alzheimer's mouse models and human brain um showed that indeed when you sequence and break the cells into individual cells and read out rna using single cell sequencing there is indeed specific disease associated populations or states of microglia that can be observed in the brains of alzheimer's patients and in alzheimer's mouse models but even beyond alzheimer's there's evidence that this is happening across multiple diseases and so this is indicating and identifying markers or signatures that denote disease-associated states in microglia so the question now is how do we understand what these states are doing one of these disease-associated states based on the markers that came out of the single cell could be used to spatially map these populations identify where in the brain they exist many of them are actually localizing to the microglia that surround the plaques which have been observed anatomically for years but now we actually have markers that can actually identify those cells and of course this also opens up a lot of critical questions in the field which is how dynamic are these states are they reversible for example once a microglia shifts into this disease-associated state do they have the ability to shift back how plastic are they what functions are each of these different populations how do we link changes in transcriptional state to specific functions of microglia can we use the information gleaned from single cell sequencing to develop new tools and probes to be able to track and target specific populations especially non-invasively where we can imagine trying to track this in over time across disease progression what's the impact of environmental challenge or aging in the context of microglia state changes and of course what is the impact of specific genetic variants or mutations that are being identified and although i focus today on alzheimer's disease more and more work is implicating microglia in other neurodegenerative diseases as well including parkinson's disease als and others so really we're in a position and it's an exciting time in the field where we there probably are now many potential paths to therapeutics that we didn't have before the old approach might be oh let's just block inflammation or let's block all microglia clearly that's not going to work because there's lots of diversity in states so we really want to start to think about how to target the detrimental but not the beneficial states in microglia and the idea that you can use some of the single cell sequencing that are identifying these markers we could use to then target and then manipulate just the populations and then hopefully be able to shift them back into a more beneficial state so there's much to do in the field and i i wanted to just end with the fact that we also have really lacked biomarkers from microglia and in particular i think it would be a real game changer if we had ways to be able to non-invasively track microglia changes in the brain across disease progression there are some pet ligands or imaging biomarkers that are suggestive of microglia they seem to be more a marker of inflammation of some sort this tspo pet ligand for example but it's not specific enough for microglia so perhaps some of the molecules and markers that we're identifying in the single cell data might identify new markers that we could use to develop new pet ligands and we specifically want to start to develop biomarkers both imaging and also cerebral spinal fluid or csf biomarkers that can enable us to be able to track microglia and myeloid cell state changes across disease progression especially early when synapses and cognitive impairment is beginning so that's an example of how understanding really basic functions of microglia that for us began in development is providing new insight into synaptic vulnerability in the context of disease and how then using unbiased and other new technologies and tools could then enable us to try to better understand and validate these mechanisms across multiple species including human so i'd like to end by thanking all of the amazing collaborators and lab members both past and present some key collaborators generous funding and i want to end with my human microglia in my lab that have really made a lot of the work i told you about possible so thanks very much you



Please enter your comment!
Please enter your name here